chive.org/details/journalofanatomy52anatuoft JOURNAL OF ANATOMY ¢ si ahs WE ee eagl JOURNAL OF ANATOMY (ORIGINALLY THE JOURNAL OF ANATOMY AND PHYSIOLOGY) CONDUCTED, ON BEHALF OF THE ANATOMICAL SOCIETY OF GREAT BRITAIN AND IRELAND, BY PROFESSOR THOMAS H. BRYCE, UNIVERSITY OF GLASGOW PROFESSOR EDWARD FAWCETT, UNIVERSITY OF BRISTOL PROFESSOR J. P. HILL, UNIVERSITY OF LONDON PROFESSOR G. ELLIOT SMITH, UNIVERSITY OF MANCHESTER PROFESSOR ARTHUR KEITH, ROYAL COLLEGE OF SURGEONS, 7 LINCOLN’ S-INN-FIELDS, LONDON, W.C. VOL. LII. : THIRD SERIES.—VOLUME XIII. pas, -ailel WITH PLATES AND NUMEROUS ILLUSTRATIONS IN TEXT: LONDON: CHARLES GRIFFIN AND COMPANY, LTD., EXETER STREET, STRAND. 1918. aon wie : a% * : 7 « SP apy ce gam ee i Uk, le ee * 7 FIRST PART—OCTOBER 10917. ‘ PAGE Tur EvoLurion OF THE TETRAPOD SHOULDER GIRDLE AND Fore-Lims. By Lieut. D. M. S. Watson, M.Se., R.N.V.R. 1 A Case or Accessory LUNGS ASSOCIATED WITH HERNIA THROUGH A CONGENITAL Derect OF THE DiApHRAGM. By E. A, Cockayneg, D.M., F.R.C.P., and R. J. GLApstong, M.D,, F.R.C.S., F.R.S.Ed. . : : 64 Form and Function or TrerH: A TueEory or ‘‘Maxtmum Swear.” By D. Muse SOAW ) 42.3... ee 97 Tue EARLIEsT STAGES OF DEVELOPMENT OF THE BLOOD-VESSELS AND OF THE HEART tn Ferrer Empryos. By Cuune-Cuinc Wane, M.D., Ch.B, Edin. . =. . ~~ 107 SECOND PART—JANUARY Io18. - Tue EARLIEST STAGES OF DEVELOPMENT OF THE BLOOD-VESSELS AND OF THE HEART IN Ferret Embryos. (Conclusion.) By Couna-Cu1ne Wane, M.D., Ch.B. Edin. 137 Tur REPRODUCTIVE ORGANS OF CETACEA, By Professor ALEXANDER MEEK : ; 186 THe PrimorDIAL CRANIUM OF ERINACEUS EUROPZUS. By Professor EDWARD FAWCETT . 211 CoNGENITAL STENOSIS OF THE PULMONARY ARTERIAL VALVE, WITH PATENT FoRAMEN OVALE, IN AN OLD MAN: WITH REMARKS ON THE VALVULAR Factor IN CARDIAC * Action, By ALExaNDER Morison, M.D., F.R.C.P. . «ew eC vi Contents THIRD PART—APRIL 1018. PAGE Tue INTERNAL MAMMARY LYMPHATIC GLANDS. By E, Puinip STIBBE : ‘ : 257 On ABNORMAL SEXUAL CHARACTERS IN Twin Goats. By Miss EstHer RIcKARDS and Professor F. Woop JoNEs. F - F Risch 3 : : : ‘ 265 Stupy oF AN Opopymous Kitren. By J. A. Prres pg Lima, M.D. . : : ; 276 THE EXAMINATION OF A SKELETON OF KNOWN AGE, RACE, AND SrEx. By Miss KATHLEEN F, LANDER, B.Sc. . : ; : 3 é ‘ ; : ate 282 THE PEcroRALIS Minor: A MorpHoLocicat Stupy. By Miss KaTHiEEn F, LANDER, B.Sc. - ; : é , : : 4 ; ; : f ; : 292 OBSERVATIONS ON THE OMOHYOID MusciE.’ By THomAs WALMSLEY, M.D. : 4 319 THe RepucrION OF THE MAMMALIAN FisuLta. By THomas WALMsLEY, M.D. . , 326 THe RELATIONS OF THE GLOSSOPHARYNGEAL NERVE AT ITs EXIT FROM THE CRANIAL Cavity. By E. Joyce Partriper, M.R.C.S., L.R.C,P. % ; : ‘ A 832 MULTIPLE ANOMALIES IN A HuMAN HEART. By HERNANI Bastos MONTEIRO . : 335 Muscutar Action. By W. Corin Mackenzin, M.D., F.R.C.S. Ed. . : ; ‘ 343 FOURTH PART—JULY 1018. "Tux SuBLINGUA AND THE PLIcA Fimpriata, By Professor Frepenic Woop Jonks , 345 A Hirnerro Unpescrrvep MALFORMATION oF THE Heart, By H. Buaxeway, M.S., F.R.C.8. . ‘ ; ; ; 4 : 5 . ‘ eres ‘ ‘ ; 354 Some Inpices AND MEASUREMENTS OF THE MopERN Femur. By J. R. D, Howrey, M.D., 8c.B. . . 2 ; : ; F ° ; : 5 ‘ ; 3 363 Tux Muscies nevarep To THe BrancntAL Arongs IN RAIA cLAvATA, By Epwarp Pures Avis, Junr. PMP aces ee ee eee Tux PaimonpiaALn Cranium or PactLopnoca WeppeLii (WeppeLi’s SEAL), AT THE 27 um. O.R, Lexorn. By Professor Epwarp Faworrt, M.D. Edin, , : » 412 : "+ ae in ae a i) 7 Contents ON THE STRUCTURE OF THE HumAN Soteus Muscie. By Doveras G. Rein, M.B., Ch.B, Edin., M.A. Trin. Coll. Camb. Tue HomoLocy oF THE MAMMALIAN ALISPHENOID AND OF THE ECHIDNA-PrERYGOID. By H, Leicuron Kesteven, D.Se., M.B., Ch.M. . THE HEART oF THE LEATHERY TURTLE, DERMOCHELYS (SPHARGIS) CORIACEA, WITH By Cuas. H. A Nore oN THE SEPTUM VENTRICULORUM IN THE REPTILIA, O’DonocuuE, D.Sc., F.Z.S. InpEx To Vou. LII. vii PAGE 442 449 467 481 JOURNAL OF ANATOMY Ss THE EVOLUTION OF THE TETRAPOD SHOULDER GIRDLE AND ; 3 FORE-LIMB. By Lieut. D. M.S. Watson, M.Sc., R.N.V.R., Lectwrer : in Vertebrate Paleontology in University College, London. From the beginning of the study of osteology the Tetrapod shoulder girdle has been discussed by innumerable anatomists. Left by Cuvier and Owen in a state-of perfect simplicity so far as concerned its morphology and nomenclature, it has gradually become more and more confused through the investigation of its development and the discovery of new fossil material. This confusion is partly genuine, depending on actual and real differ- ences of ‘opinion about the homologies of its elements, and partly depends on misuse of names. I have had evidence. that a thorough study of the literature of \this region by one who is not personally acquainted with actual material leads\inevitably to the acceptance of views which cannot be made to agree with any-reasonable and consistent scheme of its evolution. In my opinion morphological questions must be approached from below— _ that is, from the standpoint of the truly primitive members of the phylum— and the changes in relation of the individual regions of the organs con- sidered traced upwards from this beginning. The series so constructed must then be checked by a study of the development in such forms as can be obtained, and finally, if it be possible, a functional explanation must be found for the observed veries-of changes: ~~ Tt is my purpose in this paper to study the mode of origin of the mam- _ malian shoulder girdle in this way, because material for this study is more abundant and more readily treated than that for other groups, and because, as the nomenclature of the mammalian shoulder girdle is the standard to which others must be reduced, such a study is an essential preliminary to their investigation. I have in a previous paper endeavoured to show that the embolomerous } amphibia are the most primitive as they are the oldest known Tetrapods, _\ and that from them the: Cotylosauria and the rachitomous amphibia are alike derived. The well-known Embolomeri are primitively aquatic animals, whilst \ VOL. LII. (THIRD SER. VOL. XIII.)—OCT. 1917. a \ \ ~t \ 2 Lieut. D. M. S. Watson the best-known Rachitomi (Eryops, e.g.) and all the early Cotylosaurs are essentially land animals, the advances in their structure being almost all such as fit them for a thoroughly terrestrial life. As I shall treat of the shoulder girdle of Embolomeri in another con- nection, this paper will begin with the rachitomous amphibia and pass on through the Cotylosauria and various types of Anomodonts to the mam- malia, along a series which is now recognised by the majority of paleonto- logists as a true although approximate phylogenetic series. For convenience the paper is divided into two parts, the first dealing with facts and interpretations only so far as they concern individual forms, the restoration of the musculature and the possible motions being the more important items under the latter head. The second part deals with changes and their functional meanings. RacuHiIToMous AMPHIBIA. The shoulder girdle is now known in many forms of land-living rachitomous amphibia. Eryops, Cacops, Trematops are well-known Artinskian (Lower Permian) forms agreeing in main features. Eryops as the largest type may be briefly described here; the whole structure has been figured by Cope and Case (Carnegie Inst. of Washingtun Pub. 146, 1911, pp. 99-100, pls.) and Moodie. The “primary” shoulder girdle consists of a single large element on each side. This bone has a large scapular — blade fying over the ribs of the anterior part of the round body of the | animal. From the outer surface of the bone rises a powerful process whose posterior surface forms the anterior end of the glenoid cavity. The articular surface of this cavity forms a strip, shallow dorso-ventrally, but of consider- able length; its anterior end faces backwards. In the middle of its length it lies at its most dorsal position and faces directly outwards. Finally, posteriorly it faces outwards and forwards and somewhat dorsally. The — coracoidal end of the bone lies ventral to the glenoid cavity, and forms a small thin plate which is somewhat turned in towards the mid-ventral line. The scapula-coracoid is pierced by three foramina, tirst accurately de- | scribed by Williston. These are the swpraglenoid, passing through from > the back of the scapula above the glenoid cavity to the inner surface, where it opens into a pit; the glenoid, opening from the outer surface just below the anterior part of the glenoid cavity and passing through the bone to open into the same pit as the supraglenoid; and the precoracoid foramen, opening through the precoracoidal region of the bone, The pre- coracoid foramen transmitted the supracoracoid nerve ; what passed through the other openings is quite uncertain, but they perhaps received extensions | of the synovial cavity surrounding the head of the humerus. The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 8 The interclavicle is a small oval plate of bone lying in the mid-ventral line in the region of the anterior margin of the coracoidal end of the “ scapulo-coracoid.” The central region of the ventral surface of the interclavicle bears a somewhat faint ornament of pits, similar to that of the dorsal surface of the head, and showing beyond all question that the interclavicle is a dermal bone lying in the skin and having no muscle insertions on its ventral surface. The clavicle is a slender bone, bent so as to form about a quadrant of a circle, with an unornamented upper end of nearly circular section, but bearing a groove on its posterior surface which receives the lower end of the cleithrum. The lower end of the clavicle is flattened and expanded antero-posteriorly so that it forms a relatively broad plate of bone which overlies the recessed antero-lateral part of the ventral surface of the interclavicle, and like that bone has its outer face covered with a pitted ornament. The cleithrum is a very large bode forming a cap over the dorsal end of the scapula, and provided with a long slender process applied to the front edge of that bone and itself covered below by the upper end of the clavicle. The humerus is a short, very massive ‘idue with an exceedingly short shaft and much- -expanded ends, whose broad surfaces make a considerable angle (about 90°) with each other. The head of the humerus in old, well-ossified individuals shows a sharply marked articular area in the form of a long strip winding round the cylindrical proximal end like the thread of a screw. The only other features of the humerus to which it seems necessary to call attention here are the large, well-developed entepicondyle, the presence of a special very strong process arising from the shaft above and behind the ectepicondyle, and the general excellence of the ossification in old individuals. The peculiar shape of the glenoid cavity and of the articular strip wound round the head of the humerus shows that the humerus has its motions very strictly limited; it is incapable of any rotation on its long axis, and that axis must be so placed that it lies nearly parallel to the ground, but even at its position of extreme depression the distal end is higher than the glenoid part. Finally, its only possible motion is such that a point on the distal extremity when viewed directly laterally moves along a small segment of a large circle placed parallel to the animal’s sagittal plane, with the chord of the segment placed nearly horizontally ; in other words, as the humerus moves forwards its outer end is first slightly depressed and then raised again. This type of motion, in which the humerus is restricted to one definite 4 Lieut. D. M. S. Watson track both in its forward and backward motion, is common to very many early Tetrapods, and its significance and the musculature which goes with it will be considered in this paper under the head of Pelycosauria. The enormous size of the upper part of the cleithrum in the majority of Rachitomi shows that that region was of functional importance, and it seems probable that it gave origin to the scapular part of the deltoid, which muscle may in these animals have formed a continuous sheet having its origin from the outer surface of the cleithrum and the posterior and lateral margins of the clavicle and interclavicle. The rachitomous amphibian Archegosaurus is known by a very large number of specimens, which allow us to trace the origin and progress of ossification in the shoulder girdle. In fully-grown old individuals, which are not at all common, the scapulo-coracoid greatly resembles that of Eryops, having a large © scapular blade and a smaller coracoidal part lying below the glenoid cavity. The articular surface is a screw-shaped strap like that of Eryops, except that it is abruptly truncated posteriorly, in the specimens I have seen, obviously owing to the persistence of a small cartilaginous region. The cleithrum is quite similar to that of Eryops, and the clavicle and interclavicle only differ in the widening of their ventral surface, which is apparently an aquatic adaptation. The evidence of young specimens shows conclusively that the scapulo- coracoid is really a single bone, ossification beginning in the scapula blade just above the glenoid cavity at the hinder margin, spreading gradually through the scapular part, and only invading the coracoidal region in — individuals which are about three-quarters of the largest size. The evidence given by the less abundant material of Actinodon, which has recently been described by Thévenin (Ann. de Palwontologie, 1910), supports the conclusion that the scapulo-coracoid of certain rachitomous amphibia may be only a single bone, the ossification of which begins in the scapula. In this connection it may be recalled that no one has demonstrated the presence of sutures in the scapulo-coracoid of any Lower Permian amphibian. COTYLOSAURIA. So far as its skull is concerned, the Lower Permian Cotylosaur Seymouria is the most primitive of known reptiles. The limbs of this animal, both fore and hind, are also so similar to those of Eryops and its allies that when their bones were first discovered isolated they were deseribed as an amphibian, Desmospondylus, and their The -Evolution of the Tetrapod Shoulder Girdle and Fore-limb = 5 true position was only realised by Professor Williston when he obtained a complete skeleton. Considering this extraordinary primitiveness in nearly all regions of the skeleton, it is remarkable that the shoulder girdle is in some ways specialised. The scapulo-coracoid resembles that of Eryops in its general characters, and has the same type of screw-shaped glenoid cavity. The extreme posterior end of this cavity was, however, unossified, being cut by a cartilagé face in the only described specimen. There is a definite suture between the scapular and coracoid regions, running so as to divide the anterior region of the articular surface of the glenoid cavity horizontally. There is no cleithrum. The clavicle and interclavicle on the whole resemble those of Eryops, but the latter bone has a larger area, with a short posterior stem and produced lateral angles. The clavicles extend up to the top of the scapula, no doubt secondarily following the loss of cleithra. The Diadectids are a group of Cotylosaurs which, though more ad- vanced than Seymouria, are still in many ways very primitive reptiles. The scapulo-coracoid is a bone extremely similar to that of Eryops, agreeing with it in all important features, including the form and position of the glenoid cavity. No specimen has been described as showing any sutures separating the scapular from the coracoidal ends. The interclavicle is a long straight flattened bone whose anterior end is surrounded by the inner ends of the clavicles, which have firm sutures with it. The clavicle is a massive bone bent round to fit the body, and slightly expanded ventrally where it articulates with the interclavicle and its fellow. Neither clavicle nor interclavicle is in any way ornamented, and - all their surfaces may have served as muscle attachments. The cleithrum is a slender bone partially capping the scapula. .The Diadectid humerus is very short and massive, differing from that of Eryops only in the presence of a foramen piercing the very large entepicondyle, and in the slightly more proximal position of the tuber for the deltoid insertion. The powerful process in the ectepicondylar region which probably gave origin to the supinator longus or brachialis anticus is identical in the two forms. The Captorhinide are an important group of Cotylosaurs closely related to the ancestry of the Anomodontia, and hence of the mammals. No well-preserved shoulder girdles have been described, and none is avail- able to me. In the large form Labidosaurus, there are three elements in 6 Lieut. D. M. S. Watson each side of the cartilaginoid shoulder girdle, the glenoid cavity being screw- shaped and much resembling that of the Pelycosaurs. The humerus has expanded ends and is twisted, but differs from those of Eryops, Seymouria, and Diadectes in being mueh more lightly con- structed and lacking the powerful process for the brachialis anticus. Judging from the well-ossified head of the humerus of Captorhinus, the glenoid cavity in that form is not typically screw-shaped, but in the absence of well-preserved shoulder girdles it is unnecessary to discuss its possible motions. The Pariasauride, a family of Cotylosaurians, includes a series of large animals of very heavy build whose generic separations are still largely undefined. In the shoulder girdle and humerus they present such marked differences that no general statements are possible, and it is necessary to describe individual types in some detail :— | Embrithosawrus sp.?, the individual of which the skull was described by Watson, Proc. Zool. Soc., 1914, pp. 155-180, figs. 1, 2, 4,5, from the Tapinocephalus zone, Middle Permian, of Hottentot’s River, Dist. Beaufort West, Cape Province (fig. 1). The scapulo-coracoid of this form has a short narrow scapula blade with whose anterior margin a small cleithrum is fused. At the level of the lower end of this bone the front margin of the scapula is thickened and everted, forming a distinct acromium, which terminates below in a margin which passes inwards to the front face of the scapular blade so as to leave a distinct rounded notch, the first representative of the mam- malian supraspinus fossa. Posteriorly the scapula is thickened, and with the posterior upper corner of the precoracoid, which is not visibly de- limited by suture, forms a powerful outstanding process which carries the anterior end of the glenoid cavity. The glenoid cavity is screw-shaped, but differs from that of Diadectes, ete., in having a much deeper articular surface in proportion to its length ; its anterior part faces downwards, backwards, and slightly outwards; the posterior part faces directly outwards. The lower margin of the glenoid cavity is notched, the notch forming the upper border of a small pocket from the bottom of which the precoracoid foramen starts and passes through the bone. Below the humeral articulation the precoracoid and coracoid form a large flat area, The clavicle is a massive bone articulating with the front surface of the acromium and forming a quadrant of a circle, so that its lower end is applied to the front face of the interclavicle and just touches its fellow of the opposite side. The interclavicle is T-shaped, its flat front surface giving attachment The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 7 to the clavicle. The stem is long, narrow, but slightly swollen posteriorly, - and of considerable thickness. The humerus of this individual was not preserved. The type specimen of Bradysawrus bani (Seeley), from the Tapino- cephalus zone, Middle Permian, has been well described by Seeley, Phil. Trans., B, vol. 183, pp. 311-370, pls. 17-23. Fig. 1.—Right scapulo-coracoid and cleithrum of Hmbrithosawrus sp., 7 external surface. x 75. The clavicle and interclavicle are indistinguishable from those of the specimen described above, but the scapulo-coracoid is very different. The scapula blade is much narrower, the acromium less €verted and more markedly distinct from the rest of the bone. The glenoid cavity _is still serew-shaped, but its articular area is much deeper in proportion to its length, and its anterior end looks much more directly downwards— in fact, it scarcely faces at all backwards. Finally, the whole lower end of the bone is much shorter from front to back, the precoracoidal region in particular being much reduced. The humerus of a species of this genus is very well preserved in 8 Lieut. D. M. 8. Watson a skeleton I collected on Hottentot’s River, which shows that, as Professor Seeley suspected and Dr Broom has already shown, the two ends, instead of being nearly parallel as in the British Museum specimen, really make a considerable angle (about 50°) with one another (fig. 2). Fie, 2,—Left humerus of Bradysaurus sp. Dorsal, ventral, and po views, to show especially the screw-shaped surface for articulation with the glenoid cavity, x }. When correctly articulated, the humerus stands out at right angles to the animal’s body, with the broad plane of the upper end inclined down- ., wards in front and the broad plane of the distal end nearly parallel to the ground, as it is placed on Professor Seeley’s figure. ; ; The lower end of this, as of all other Pariasaurian humeri, is remark- able for the very small relative size of the entepicondyle when the bone is compared with that of any other Cotylosaur. The radius and ulna, The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 9 which are well preserved in both the skeletons available, fit their articula- tions on the humerus with great accuracy, and completely justify the position given them by Seeley. In the normal position they stand at right angles to the expanded lower end of the humerus, and are placed - nearly perpendicular to the ground, separated from their fellows of the other side by the length of the two humeri plus the distance between the two glenoid cavities. The vertical position of the ates arm, by enabling the main stress in its components to be a compression on the radius and the tension in the flexor muscles of its palmar surface to be very small, is the explanation of the very small entepicondyle of the humerus of the animal. The deep sigmoid notch and pronounced olecranon strictly limit the possibility of motion of the fore-arm to a small flexion and extension in the plane containing these bones and the main axis of the humerus. Embrithosaurus and Bradysaurus are of Tapinocephalus zone age—that _ is, Middle Permian. In the Upper Permian Cistecephalus zone of South Africa and of the Dwina region of Russia we meet other Pariasaurs. Of _ these, the commonest South African type is Propappus, whose shoulder girdle as described by Broom resembles that of Bradysaurus in all ways, including the shape of the glenoid cavity. The scapulo-coracoid of one of Professor Amalitzki’s Russian Pariasaurs is represented in the British Museum by a cast from which fig. 3 is drawn. It closely resembles that of Bradysaurus in many ways, but differs in the higher position of the acromium, and especially in the complete loss of the screw-shaped glenoid cavity, the articular surface being now represented by a shallow pit whose posterior margin lies at the extreme posterior edge of the bone. The shoulder girdle of Pariasuchus, from the Cistecephalus zone of the Niewveld, as it has been described by Broom and Haughton (Ann. So. Afr. Mus., vol. xii. p. 22, pl. iii.) is of very great interest. The scapula blade resembles that of any other Pariasaurian in having a pronounced outstanding acromium, and the general shape of the pre- coracoid and coracoid region is also Pariasaurian, but the glenoid cavity is quite unlike that of any other member of the group. Instead of facing outwards and being formed by a depression of the outer surface of the shoulder girdle, it faces backwards and notches the posterior surface so that its position is shown from the visceral surface. This new position and shape of the glenoid cavity means that the mode of moving the fore- limb differed very considerably in Pariasuchus from that in all other known members of the group, and explains the remarkably individual form of the humerus, which has the antero-external corner of the head to which 10 Lieut. D. M. S. Watson the m. supracoracoideus and m. scapulo-humeralis anterior were attached much produced. The position of the mounted skeleton photographed by Broom and Haughton (op. cit., pl. v.) is, I believe, nearly correct, and is one into which it is quite impossible to place Bradysaurus or Embrithosaurus without, as in the Cape Town specimen of the latter genus, dislocating all its joints. 4 \) \ wy \y a) Ihe Hy y hi fy ad, Fie, 3.—Left scapulo-coracoid of Pariasawrus Karpinskyi, Amalitz., external surface. x. The Pariasaur bones at my disposal do not show the muscle inser- tions sufficiently plainly to allow me to endeavour to reconstruct the musculature. PELYCOSAURIA. The group Pelycosauria includes a large number of considerably diverse forms, some of considerable size, others comparatively small. The shoulder girdle and fore-limb consequently present considerable variations through- out the group, but the fundamental plan is identical in all. The Evolution of the Tetrapod Shoulder Girdle and Fore-linb — 11 Dimetrodon, a large advanced form, is on the whole the best known, and is described here as a standard. The scapulo-coracoid is a large bone with a large flat scapula blade lying parallel to the ribs; its anterior edge is thin, whilst the hinder part is much thickened and rounded. The lower margin of the scapula is divided into two parts, the anterior having a long and close suture with the precoracoid, the posterior a more open suture with the coracoid. At the posterior end of the suture between the scapula and precoracoid the two bones are much thickened, forming a stout process bounded below by a deep fossa from Fie, 4.—Right scapula, precoracoid, and coracoid of Dimetrodon sp. External (A), internal (B), and posterior (C) views. x4. which the precoracoid foramen starts. The posterior face of this process forms the anterior part of the glenoid cavity, facing backwards, slightly downwards, and scarcely at al] outwards. The glenoid cavity has a narrow screw-shaped articulating face resembling that of Eryops, the posterior half of which lies entirely on the coracoid; this region faces upwards and outwards. This glenoid cavity differs from that of Eryops in that the articular > surface is wider in proportion to its length and that the anterior part lies higher than the posterior. Below the level of the glenoid cavity the coracoid and precoracoid are continued down towards the middle line, forming a large flat area. On the posterior edge of the coracoid is a powerful inwardly and backwardly directed process, presumably for the insertion of the coraco-brachialis longus. 12 Lieut. D. M. S. Watson On the visceral surface the scapulo-coracoid shows a large generally smoothly-lying surface bounded behind by the thickening of the posterior margin of the scapula and coracoid. From this a subsidiary ridge runs forwards on the precoracoid so as to form a deep fossa, lying over the suture between the precoracoid and scapula, into which the foramina supra- genoideus and precoracoideus open. On the lower side of the ridge on the precoracoid is a very pronounced depression partly on the precoracoid and partly on the coracoid, which appears to have housed a large subcoracoid muscle. Thei inner surface of the posterior upper part of the scapular blade is’ depressed and separated by a low ridge from the rest of the surface; it passes by a distinct wide depression to the posterior margin of the scapula, and probably denotes the position of the belly of a large sub- scapularis muscle. Membrane Bones. The interclavicle is a long straight flattened bone lying on the ventral surface of the animal and extending far behind the coracoids. At its anterior end the interclavicle is widened into a lozenge-shaped head, whose lower face is largely covered by the clavicle. The clavicle is a bone which articulates with the front edge of the | scapula as a narrow bar which widens as it is followed downwards until it expands into a broad flat plate underlying the interclavicle. A cleithrum _ does not occur in Dimetrodon, but a small splint representing this bone occurs in the nearly allied Clepsydrops natalis and in Edaphosaurus. Humerus. The humerus of Dimetrodon is a stout bone with much-expanded upper and lower ends placed nearly at right angles. The proximal end of the bone is crossed obliquely by a narrow articular strip which begins on the dorsal surface above the deltoid crest and finishes ventrally at the ' posterior margin of the expanded head. From the anterior edge of the lower face of the expanded upper end a powerful plate-like delto-pectoral crest projects downwards. From this crest a low ridge crosses the shaft and becomes continuous with the posterior margin of the enormous entepicondyle. This region is pierced by a large foramen lying high up towards the shaft. Placed right at the distal end of the bone in the direct line of the shaft, and lying almost entirely on the anterior and lower surface of the expanded end, is a very distinct hemispherical articular surface for the head of the radius. If the radial condyle be placed facing upwards, the equally distinct face for the ulna lies in the main internal but partly below it; it forms a some- iii yk eT tis | - following manner :— The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 13 what spheroidal face mainly on the end, but also to a slight.extent on both broad surfaces, of the humerus. The ectepicondyle is massive and directed somewhat backwards; proximal to it and lying in front so that the two enclose a distinct channel is a small but powerful process for the brachialis anticus. Fore-arm. The radius is a straight rod with a hemispherical cup for the condyle on the humerus. The ulna is chiefly remarkable for the possession of a ‘very large olecranon and of a deep sigmoid notch. Freedom of Motion of the Fore-limb. The narrow screw-shaped articular surfaces of the glenoid cavity and the head of the humerus fit extremely accurately in old individuals, and fix the position and possible motions of the humerus with great exactness. When the humerus projects at right angles to the body, its principal axis in Dimetrodon lies nearly parallel to the ground. The bone is incapable of any rotation about its axis whilst it retains a fixed position, and can only move so that a point on the distal end describes a segment of a circle whose chord is higher anteriorly than posteriorly. When the humerus is carried backwards the anterior margin is depressed relatively to the posterior. The humerus can scarcely be placed much in advance of a right angle to the animal’s principal plane, and its range in motion is extremely limited ; it can perhaps turn through about 50°. The fore-arm has considerable freedom of motion on the humerus. At its normal position it stands at right angles to the expanded lower end of the humerus, and makes an angle of about 30° with the ground. It can be moved in and out through a somewhat small angle restricted by the deep sigmoid notch of the ulna, whilst the spheroidal shape of the ulnar articulation on the humerus allows of some movement in a plane at right angles to the long axis of the humerus. The walk of Dimetrodon must have taken place somewhat in the When the animal is standing with the left fore-leg in the normal position, the right humerus is advanced, raising its outer end and its anterior edge; the fore-arm, lying nearly parallel to the ground, is extended as much as possible and the palm placed on the ground. The humerus is then retired and the elbow flexed. ‘The combined motions raise the glenoid cavity, owing partly to the depression of the outer end of the humerus and partly to the depression of its anterior edge; they also swing the body 14 Lieut. D. M. S. Watson across so that the hand gets nearer the middle line. Finally, the fore-arm is pulled backwards on the humerus, still further raising and driving forward the animal. At or before this stage the left arm is advanced and repeats the same process. The whole anterior part of the body is thus swung from side to side at each stride. The width of the track is very considerable, being equal to the width between the distal ends of the two humeri, and the stride is very small. Measurements suggest that a rather small Dimetrodon might make a track 16 inches wide, with a stride of perhaps 6 inches to 8 inches. | The bones of Dimetrodon at my disposal show certain muscle insertions. clearly, but none shows all of them. I collected from the Craddock bone bed, however, the upper end and part of the distal half of “a small primitive Pelycosaur humerus (? Varanosaurus) which shows all the muscle insertions diagrammatically. These agree closely with such as are known in Dimetrodon, and there can be no doubt that the musculature and mechanics of the two animals are quite similar. The positions of the insertions of the muscles on this humerus will be best understood from the figures; their determination rests on a comparison with Sphenodon, whose musculature must be very similar to that of the Pelycosaurs (fig. 5). The important features are the large size of the insertions of the deltoid, pectoral, and supracoracoid muscles; the somewhat distal position of the powerful insertion which seems to*be for the latissimus dorsi and scapulo- humeralis posterior; and the well-marked confluent insertion of the sub- scapularis and subcoracoideus. The coraco-brachialis brevis is short but wide, the coraco-brachialis longus much more considerable. The muscle insertions on the lower end of the humerus cannot be safely distinguished. There are powerful flexor insertions on the huge entepi- condyle, and the dorsal surface of the ectepicondyle has a very powerful insertion for an anconeus quartus. Like other ‘letrapods, Dimetrodon is supported by its limbs, which are compound levers held in place by muscles which act as tensile members. The stress developed in the muscles of the fore-limb of Dimetrodon may be grouped as follows :— 1. The palm resting on the ground, contraction of the flexor muscles attached to the entepicondyle and to the carpal region raises the humerus if that bone does not revolve on its axis. The anconeus quartus, passing from the ectepicondyle to the upper end of the olecranon process of the ulna, assists this motion. 2. The elbow is flexed by the biceps and brachialis internus. 3, The elbow is extended by the triceps. Th ee Nes The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 17 lies somewhat more ventral to the posterior part than in the more advanced genus. The undetermined Ophiacodont scapulo-coracoid from the Craddock bone bed figured .by Williston (p. 50, fig. 268) generally resembles that of Ophiacodon, but is of importance because it shows that in these animals the precoracoid plays only a small part in the glenoid cavity. DEINOCEPHALIA. Apart from those of the Pelycosaurs, the oldest and-most primitive shoulder girdle of a mammal-like reptile which is known is that which, found in the copper-bearing sandstones of Middle Permian age in the department of Orenburg and the Urals, was referred by Seeley, with great probability of correctness, to the Deinocephalian Rhophalodon. (Only two types of scapulo-coracoids have been described from the Orenburg beds, one being certainly Labyrinthodont. Only Labyrinthodonts, Deinocephalia, and a single very small Cotylosaur are known from other bones.) The shoulder girdle of “ Rhophalodon ” is represented in the British Museum by casts of the glenoid region of the right side and the scapula broken off through the top of the glenoid cavity of the left side. These fragments are of identical size, and appear to have belonged to the same individual; they were figured by H. v. Meyer (Palwontographica, Bd. xv. pl. xvii. figs. 1-2, pl. xviii. figs. 1-2) as Eurosaurus, a Labyrinthodont. From these casts my figures have been drawn, the outline depending on a badly preserved specimen figured by Eichwald and Seeley. The scapula has a large considerably curved blade with a thin anterior border and a massive thickening along its posterior edge. The ventral end of the scapula has a long straight suture with the precoracoid, and a much shorter connection with the coracoid. The glenoid cavity is definitely screw-shaped, but its articular surface is relatively wide, and its anterior end, which is carried almost entirely by the scapula and only to a very small extent by the precoracoid, faces downwards, outwards, and scarcely at all backwards. The posterior part of the glenoid cavity faces upwards, and is entirely supported by the cora- coid. The outer surface of the precoracoid is raised into a very definite ridge running forward from the upper anterior corner of the glenoid cavity. Ventral to the ridge the bone is depressed, forming an open pocket from which the precoracoid foramen starts. On the visceral surface the thickened posterior margin of the scapula forms a distinct ridge, which is sharply marked off below by the deep pocket into which the precoracoid foramen opens. The precoracoid is crossed by VOL. Lil. (THIRD SER. VOL. XIII.)—OCT. 1917. 2 18 Lieut. D. M. S. Watson a low ridge, the surface of the lower part of the bone and of the coracoid being depressed to form a shallow hollow. No complete well-preserved humeri are known, but fragments differ chiefly in their small size and greater slenderness from that of Titanosuchus, a South African Deinocephalian. They indicate an elongated humerus with wide proximal and distal ends set at a considerable angle to one another. The humerus of Brithopus, which may probably be Rhophalodon, has a comparatively small entepicondyle and an ectepicondylar foramen. Fie. 6 —‘‘Rhophalodon,” Deinocephalian from the Middle Permian sandstones of the Dist. of Kargalinsk in the Urals, Outer (A), inner (B), and posterior (C) views of the right scapula, coracoid, and precoracoid, Shaded portions from two casts of right and left bone in the British Museum ; outline from Professor Seeley. x 4. The material at our disposal does not allow of a definite statement of the freedom of motion of the fore-arm in Rhophalodon, but the form of the screw-shaped glenoid cavity shows that the humerus had only a limited power of conical motion and was in the main restricted to motion in a plane lying nearly at right angles to the animal’s axis and passing down at an angle of about 45° with the ground from the front to the back. The shoulder girdle of the large South African Deinocephalia, which are of somewhat more advanced structure and probably later in date than Rhophalodon, has been described by Dr Broom (Phil. Trans., Ser. B, vol. 206, pl. i. figs. 1-5) and by the present author (Proc. Zool. Soc., 1914, ST Ie a Ue a a ee Se a ee ec ee eet eee "7 r a we y The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 19 p-. 764, fig. 11). The scapula of the huge Tapinocephaloid Phocosaurus is a flat bone with slightly thickened posterior edge; it has a-long straight suture with the precoracoid, and posteriorly bears the upper part of the glenoid cavity. The precoracoid has long sutures with the scapula and coracoid, and takes no part in the formation of the glenoid cavity; its outer surface is flat, there being no ridge descending from the upper edge of the glenoid cavity; the precoracoid foramen is small. The coracoid is a relatively large bone carrying the posterior and lower part of the glenoid cavity, which faces upwards and outwards. The inner surface of the shoulder girdle is a smooth curve with a small sharply depressed pocket into which the precoracoid foramen opens. The humerus is massive, with widely expanded upper and lower ends set at a considerable angle with one another. The head is straight when viewed from its broad plane, the articular region being cylindrical, ex- tending round for some distance on to the dorsal surface of the bone. The ulna has a short olecranon, but the sigmoid notch is shallow. If, as is justified by all analogy, we place the ventral edge of the pre- coracoid and coracoid parallel to the middle line of the animal, then the scapulee slope forwards so that, as in the living Monotremes, the dorsal end of the bone lies considerably in advance of its ventral portion. The cylindrical head of the humerus and corresponding shape of the glenoid cavity suggest that there was considerable freedom of motion about an axis lying nearly horizontally but inclined down at about 30° in front— that is, that the humerus could be raised and depressed in a plane at about 60° with the ground. In a direction at right angles to this the possible motion must have been very small. Finally, there can have been little or no rotation of the humerus about its own axis. In the normal position the humerus stood nearly parallel to the ground, and with its axis directed outwards and backwards at about 60° to the animal's length. THERIODONTIA, Sub-order GORGONOPSIA. Only one Gorgonopsian shoulder girdle has been described, and that shortly and incompletely by Haughton and Broom, Ann. So. Afr. Mus., vol. xii. pp. 80-32, pl. vi. In this form, Seymnognathus tigriceps from the Upper Barmian Ciste- cephalus zone, the scapula is very short and broad, thickening behind and below to support the upper part of the glenoid cavity, which faces outwards, downwards, and backwards. The precoracoid is a rather small squarish bone having long sutures with the scapula and coracoid, and perforated by a small foramen near its upper posterior corner. The bone seems to be otherwise featureless. The coracoid is large, longer than the precoracoid, 20 Lieut. D. M. S. Watson and carrying the lower half of the glenoid cavity, which faces outwards and upwards. Behind the glenoid region the coracoid is produced back- wards in a long, powerful process. This scapulo-coracoid generally re- sembles that of a Tapinocephaloid, but differs in the more anterior position of the lower part of the glenoid cavity, which also is much more nearly vertical in position, facing more outwards and less upwards than in the more primitive group. — j \\ Nf Fic. 7.—Right scapula, with fragments of the precoracoid and coracoid attached, of Inostranzevia, from a cast R. 4035g in the B.M.N.H, There is a loosely articulated cleithrum in the Gorgonopsid Scylacops according to Broom. The clavicle is a large curved bone articulating with the front edge of the scapula and passing below the interclavicle ventrally. The interelavicle is a wide flat sheet of thin bone with a pronounced ventral ridge rising to a special elevation. The only other Gorgonopsid available to me is a cast of the huge Russian Upper Permian Inostranzevia, which differs exceedingly from Seymnognathus. It has a large thin scapula blade, much curved to surround the animal's thorax, and very broad dorsally. This blade narrows and thickens as it is The Evolution of the Tetrapod Shoulder Girdle and Fore-limb = 21 traced downwards until it is terminated by the upper part of the glenoid cavity, which appears to be somewhat mutilated. In front of the thick posterior part of the scapula, just above the glenoid region, is a very deep fossa lying on a special thin part of the scapula which projects forwards and whose lower end articulates with the precoracoid. - The coracoid is destroyed, and I am uncertain how much of the edge of the precoracoid is original margin. Gorgonopsid humeri have been well described by Owen, Cat. Foss. Rep. S. Africa, B.M., 1879, pl. xix., Cynodrakon major; and by Broom and Haughton in Seymnognathus (op. cit.). The Gérgonopsid humerus is comparatively slender when compared with those described in the preceding part of this paper; it has expanded distal and proximal ends which make an angle of about 30° with one another, so that when the head is placed horizontally the anterior face of the lower end looks downwards and slightly forwards. The head of the humerus when viewed from the dorsal aspect is straight, the articular surface covering the medial part of its proximal surface commencing abruptly below and extending round so that a little of it is visible on the dorsal aspect of the bone, although from the ventral surface it is quite hidden. From the anterior corner of the head the very powerful delto-pectoral crest runs down the shaft of the bone so that its anterior margin is at right angles to the head. This crest is deep, and stands at a considerable angle to the dorsal surface of the proximal _part of the bone, directed downwards. It terminates distally in a smoothly rounded margin continued as a ridge over the entepicondylar foramen to the entepicondylar edge of the lower end of the humerus. The distal end of the humerts bears on its lower surface towards the ectepicondylar side an almost hemispherical articular condyle for the end of the radius. External to this is the articular face for the ulna, which lies almost entirely on the distal end of the bone. Above the radial condyle a thin extension of the bone forms a very marked ‘ supinator crest. This subsides into the dorsal surface of the shaft of the ‘bone, and at its extreme proximal end is perforated by a small ect- epicondylar foramen. Experiment with the bones of Scymnognathus in Cape Town shows that when articulated the humerus stands out from the body. so as to lie with its distal end inclining backwards at about an angle of 60° with the animal’s principal plane, and with the axis of the bone nearly parallel to the ground. The possible motion lies in a vertical plane. When articulated in this way the radius and ulna lie side by side and pass down to the ground at an angle which may be about 45°. 29 Lieut. D. Memeaualeen THEROCEPHALIA. The only shoulder girdle material yet described is that of Ictidosuchus, which Broom described. None is available to me; there are in the British Museum, however, a good many scraps of humeri in bad association to Therocephalian skulls, to which they undoubtedly belong. These all resemble the humerus of the fore-limb described by Seeley as SN ae which certainly belongs to some member of the sub-order. These humeri seem to be structurally similar to the Gorgonopsid humeri described above, but differ from them in their great slenderness ; a difference which extends to all the other limb bones, and shows that as a group the Therocephalia are remarkable for their light build and probably great agility. CYNODONTIA. © The Cynognathid shoulder girdle is represented in the British Museum by the magnificent scapulo-coracoid of the type specimen of Cynognathus crateronotus (described by Seeley, Phil. Trans., B, vol. 186, pp. 59-148, 1895), which lacks only the lower parts of the coracoid and precoracoid. In addition, the young skeleton of a Cynognathid, described by Seeley as Microgomphodon in error, shows both scapule and the anterior margin of a precoracoid identical with those of Cynognathus, and imperfect clavicles and interclavicle. In South Africa, in the Cynognathus beds of Winnaarsbaaken, Dist. Albert, Cape Province, I collected fragments of a Cynognathid skeleton, including well-preserved but imperfect clavicles, interclavicle, precoracoid and coracoid, and the lower end of the humerus. There are in the British Museum three perfect and several fragment- ary Cynognathid humeri, many exquisitely preserved. As these humeri, of which the largest is twice as long as the smallest, agree exactly in shape, and as there is direct evidence that the proportions of all Cynognathids are similar, I have enlarged the perfect precoracoid of my own specimen to an extent determined from a eomparison of the lower end of a humerus belonging to it with the upper end of the humerus of the type specimen of Cynognathus crateronotus, thus obtaining a — complete scapulo- coracoid., The scapula is long, broadest at the dorsal end, and during life inclined forward. The main part of the outer surface forms a deep channel bounded by the everted anterior and posterior margins, which form strong out- ae - The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 23 standing ridges. Near its lower end the anterior border is produced into a low and not much thickened acromium. This process does not stand out freely from the bone as in Dicynodon, mammals, and even Embritho- saurus, but is only marked off by the depression of the outer surface of the scapula below it to form a deep fossa, the whole of whose surface Fic. 8.—Cynognathus crateronotus, Seeley, type specimen. Right scapula, with apaceus pre- coracoid and coracoid, inner surface, x 4, is visible in a direct lateral view, and which in no way undercuts the acromium; the anterior margin of this fossa is continuous above with _ the acromium, and below with that of the precoracoid. Although slightly damaged in the type specimen, it is obvious that it was thin and sharp as it is in the little skeleton described by Seeley as “ Microgomphodon.” The external surface of the scapula is completed by two small ears arising from the visceral surface of the everted anterior and posterior borders and extending outwards so as to form small anterior and posterior 24 Lieut. D. M. S. Watson fossze at the upper end of the bone. The inner surface of the scapula is saddle-shaped, flat or nearly so dorsally, and very strongly convex in an antero-posterior direction in the middle of its length. The face with which the precoracoid is articulated is short, straight, and fairly thick. The precoracoid of the ‘Winnsadieases specimen, which is very well preserved, has separated from the scapula at its suture. It is a small bone, perforated at about the middle of its length, but near to its suture with the — Fie. 9.—Cyn seni ?gen. Right ‘Sapna and fragment of coracoid, lower end of right clavicle, and interclavicle. From the ‘Cynognathus zone of Winnaarsbaaken, Dist. Albert, Cupe Province, S. Africa, x 3. ‘coracoid, by a rather large foramen. It extends forwards some way in front of the suture with the scapula; this margin is somewhat damaged proxi- mally, but shows by its rapid thinning that not more than two millimetres can be missing. The perfect distal part supports this conclusion, The ventral margin of the bone forms nearly a quadrant of a circle, and is raised into two distinct thickenings, from each of which a ridge runs on to the outer surface of the bone; these thickenings are sharp-edged, and agree exactly in character with those on the dorsal edge of the scapula of an old adult Echidna, On the inner surface the foramen through the precoracoid ends in the The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 25 lower extremity of a small deep fossa lying on both precoracoid and scapula. The coracoid is nearly completely preserved in the type specimen, its anterior part also in the Winnaarsbaaken specimen. It is much larger than the precoracoid, and has a long, bent suture with the scapula. The outer surface of the coracoid is much more extensive than the inner, and the suture with the precoracoid is not at right angles to the surface. The glenoid cavity is carried entirely by the scapula and coracoid ; the upper part on the former bone faces outwards and downwards at an angle of about 45°, the lower half of the cavity facing nearly directly outwards. _ Both are nearly flat, and have no curvature in an antero-posterior direction. The lower end of the right and what is probably the upper end of the left clavicle are well preserved in the new specimen. The upper end is bent to form a quadrant of a circle; its lower end is circular in section, whilst the upper end is flattened in the plane of the curvature of the bone and ridged for attachment to the acromium. The lower end is nearly straight, ‘and gradually widens from its cylindrical shaft. This part of the bone rests on the lower surface of the anterior part of the interclavicle. The interclavicle is represented by its perfect anterior end, connected by matrix with a part of the right lateral border so as to give the shape of © most of the posterior part. The bone forms a thin slightly concavo-convex plate with a pronounced ridge down the middle line of its ventral surface. Anteriorly the bone ends in a narrow massive rod of triangular section, whose lateral margins suddenly widen to form short blunt lateral processes, to which the clavicles are attached. Between these processes the median ridge is lowered so as to form a transverse groove, behind which it rises to a summit from which it steadily descends posteriorly. The posterior part of the bone is very wide, nearly as much across as the two clavicular arms. The humerus generally resembles that of the Gorgonopsia, but presents some differences of importance. The head is expanded, the articular surface occupying the middle of it; anteriorly the deep delto-pectoral crest rises from its lower surface; caudally it ends in a thickening covered with muscle insertions. The cylindrical articular surface of the head does not extend on to the ventral surface, but is largely exposed in dorsal view. The proximal part of the shaft of the bone is flattened, with a concave ventral surface bounded on the outer side by the powerful delto-pectoral erest so as to form a very broad bicipital groove. The dorsal surface is rounded and is crossed obliquely by two grooves which mark the insertion of the lateial humeral head of the triceps and the brachialis internus. 26 Lieut. D. M. S. Watson The delto-pectoral crest is long, and its distal end subsides abruptly into the shaft, being continued by a ridge which forms the bar across the entepicondylar foramen and runs on to the inner margin of the expanded lower end of the humerus. The lower end of the humerus is expanded, its broad plane lying nearly parallel to that of the head. The radial condyle is confluent with B A TETANY mn! ay 4 AN i} at f \\, AA h, | AN i \"'| \" \ \ | i Fie, 10.--Left humerus of a Cynognathid. Dorsal (A) and ventral (B). x 4. that for the ulna, and forms a cylindroidal surface lying almost entirely on the lower face of the bone. The ulna articulation is extensive; on the front face it is confluent with that for the radius, but does not extend so far proximally. It passes over the distal end of the humerus on to a rounded boss on the dorsal surface, The ectepicondylar region is thick and massive, and gives origin only to a small supinator crest. There is a small ectepicondylar foramen autos ing the dorsal surface of the shaft at about the level of the lower end of the deltoid crest. The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 27 Normal Position. _ Actual experiment with the perfectly preserved scapulo-coracoid of Cynognathus and consideration of the shape of the lower. border of the precoracoid and coracoid have convinced me that the Cynognathid shoulder girdle leant forward at the top just as does that of Echidna. When placed in what is obviously its natural position, the glenoid cavity looks outwards and backwards. The shape of the head of the humerus, and especially the great production of its articular surface on to the dorsal surface, show that normally it was carried parallel to the ground. The obliquity of the head on the shaft makes the plane containing the humerus in its normal position lie at about 45° to the animal’s principal plane— that is, its elbows are only slightly turned out. The cylindrical shape of the humeral articulation shows that the bone can have had freedom of motion only in the vertical plane. When the humerus is placed in its normal position, the expanded lower end lies very nearly parallel to the ground. Unfortunately, I have no bones of the fore-limb. The articulations for them on the humerus suggest that these bones had considerable range of motion in the plane including the long axis of the humerus and standing at right angles to its expanded lower end. The series of Cynognathid humeri at my disposal show nearly all the muscle insertions clearly in one or other specimen; all that are clearly determinable are represented in fig. 11. The determination of these muscles depends on a comparison with the Pelycosaur humerus described in an earlier part of this paper, which itself agrees with Sphenodon and with Echidna. It seems unnecessary to describe these insertions in words. Knowing the insertions of the shoulder muscles on the humerus, we are in a position to investigate the musculature of the scapulo-coracoid, in which the insertions are not very clearly shown. It is at once obvious that the great concave external face of the scapula gives origin to the scapulo-humeralis anterior, whose humeral insertion lies immediately below the lower end of the groove. There is a very marked insertion for the scapula head of the triceps in the form of a narrow crest on the _ posterior face of the scapula just above the level of the top of the glenoid cavity. The scapulo-humeralis posterior lies caudal to the scapula head of the triceps, and arises from the posterior and medial surface of the great posterior crest forming the posterior side of the deep groove on the outer face of the scapula; it no doubt extends up to the top of the bone in front 28 Lieut. D. M. S. Watson of the little posterior wing which extends the area for the insertion of the serrati. é Professor Seeley and all subsequent authors have held that the anterior wing of the scapula forms a mammalian prespinous fossa occupied by a supraspinatus muscle running downwards under the clavicle below the acromium to be inserted on the head of the humerus just in advance of the articular condyle. She. _CorBra.Br LU, in| NY We Wal x 1 I WNYI N \\ S SN { Fic, 11.—Right humerus of a Cynognathid, ? Diademodon. x. On these drawings I have indicated by thick dotted lines the muscular insertions visible on all the specimens of this bone available, (A) dorsal, (B) ventral aspects. Reference letters as before, with :— Anc. Quar., anconeus quartus origin; Ex. Car, Rad, L., origin of extensor carpi radialis longus ; Fl. Car, Rad., flexor carpi radialis origin; Fl. Car, U1., origin of flexor carpi ulnaris ; For, Ket. Cond., ectepicondylar foramen ; Sup. Long., origin of supinator longus. ©, section of the head along the axis, to show the articular surface, included between the two dotted lines, This view seems to me to be mechanically impossible, for it would involve the muscle or its ligamentous continuation gliding over the slightly concave margin of the scapula below the acromium and very markedly changing its direction at this point; as this margin is-exceedingly thin and sharp-edged, it seems incredible that it ever did act as a giding surface in this way. In any case, the fact that, although smaller, the caudal wing of the scapula is exactly similar to the anterior wing, suggests that this latter expansion has arisen to provide aétachment for a large omotrachelian or (and) anterior part of the serratys series, j/ The. Evolution of the Tetrapod Shoulder Girdle and Fore-limb 29 The deep fossa below the acromium on the outer surface of the scapula I hold to have been occupied by a separated part of the scapulo- humeralis anterior, just as the very similar fossa in Inostranzevia un- doubtedly was. An obscure ridge on the visceral surface of the scapula, running from above downwards and backwards at about the middle of the bone, seems to mark the insertion of the subscapularis, whose principal insertion may have been on the anterior and inner surface of the outwardly _ reflected anterior margin of the scapula, the muscle winding round the inner surface of the bone and passing directly downwards to its insertion on the dorsal surface of the lesser tuberosity of the humerus. The supracoracoideus was obviously inserted in the whole free border of the external face of the precoracoid, and ran directly to its insertion into the proximal part of the ventral surface of the delto-pectoral crest. The coraco-brachialis had an insertion just as in Echidna or Sphenodon. Mode of Walking. It is most convenient to discuss the action of the shoulder musculature in two parts, first keeping the fore-arm fixed nearly at right, angles to the humerus. With the bones in this position and the animal standing on its two hind-legs and the left fore-leg at its extreme backward position, the right humerus is depressed by the action of the coraco-brachial and pectoral muscles. Owing to the shape and position of the glenoid cavity, this drives its lower end forwards and inwards so that the hand approaches the animal’s middle line well in advance of the glenoid cavity. If the hand be now placed on the ground, the action of the scapulo-humeralis anterior, scapulo- - humeralis posterior, latissimus dorsi, and to a less extent the deltoid and subscapular muscles, raising the humerus, will not only raise the animal but drive it forwards and to the left side. At a certain point of this motion, when the hand is vertically under the glenoid cavity, the animal will be balanced on this limb; if the motion continues, the pectoral and coraco- brachialis come into strain as the animal, as it were, falls forward. If now, by the action of the triceps, the fore-arm is extended on the humerus, the animal is still urged in the same direction, and further depression of the humerus will bring the arm to its extreme backward position. At this stage the left arm is advanced and repeats the motions above described, while the right hand is picked up by the action of the extensors, brachialis © -anticus, biceps, etc., the fore-arm flexed, and the humerus driven forwards by the coraco-brachialis, pectoral, deltoid, and supracoracoideum. Thus in this type of walk all the muscles proceeding from the animal’s body to the humerus are actually used at one stage or another to drive the 30 Lieut. D. M. S. Watson animal forwards, none serving solely for support or to oppose rotational couples produced in the humerus by the action of the fore-limb musculature. DICYNODONTIA. The Dicynodonts are a side branch of the mammal-like reptiles, very specialised in their skull structure, and leading to no other group. Their Fic, 12.—Right scapula, coracoid, and precoracoid of Dieynodon halli, Watson, from the outer (A), inner (B), and posterior (C) aspects. From the type specimen, x 3, shoulder girdle and fore-limb are, however, of great importance, because in them for the only time in known reptiles we get a typical supraspinous muscle of Monotreme type. The group has a long history, and slight changes of proportion and of the position of the humerus take place within it. Dieynodon is best represented by the type skeleton of D. halli, Watson, found by the writer in the lower part of the Cistecephalus zone at Kults Poort, Dist. Beaufort West. The bones of this individual are magnifi- cently preserved and undistorted, with the exception of one clavicle and of a notch in the lower border of one coracoid formed by the pressure of a humerus which lay on it. All the bones of the fore-limb were heaped The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 31 together, the right radius and ulna being naturally articulated at right angles to the broad surface of the lower end of the humerus, and standing up perpendicular to the bedding of the rocks. _ ‘The scapula has a slender blade, very much curved in a transverse plane so as to fit over the rounded body of the animal. The blade is throughout thick, but becomes more massive at the lower end, where it furnished the upper half of the glenoid cavity. The anterior margin of the scapula is produced directly forwards into a slender acromium, which is terminated below by a margin continued by a definite ridge on to the outer face of the bone. Below the acromium the scapula extends forwards so as to form the articulation for the precoracoid. The conditions are such that the plane of the surface of this part of the scapula if continued upwards passes below the acromium so as to leave a smoothly rounded groove leading downwards from the inner face of the scapula to the outer surface. The anterior margin of the scapula above the acromium forms a flat roughened face, directed forwards, with which the cleithrum articulated. The anterior part of the lower face of the scapula presents a flat surface with which the precoracoid articulates; this margin on the inner side is crossed by a deep small depression which continues the fossa on the visceral*surface of the precoracoid into which the foramen piercing that bone opens. . The posterior part of the lower end of the scapula is divided into two parts—a small inner face with which the coracoid articulates, and a much larger area facing downwards, backwards, and outwards, which is the upper part of the glenoid cavity. The precoracoid is a small square nearly flat bone which articulates by straight sutures with the scapula and coracoid. Near the junction of these sutures the bone is perforated by the precoracoid foramen, which opens into a small deep fossa on the inner surface. The outer surface of the bone is shallowly concave; the inner surface is similarly concave but is crossed obliquely by a low broad ridge. The coracoid articulates by a long close suture with the precoracoid and by a short loose articulation with the scapula; its outer surface bears the lower glenoid surface, looking upwards and outwards and carried backwards by a special outstanding process. The cleithrum is not preserved, but was undoubtedly a slender slip of bone, articulated with the whole anterior border of the scapula above the acromium. The perfectly preserved right clavicle is a slender bone with a round shaft bearing a sharp ridge along its caudal surface. The lower end of the 32 ; Lieut. D. M. S. Watson bone is slightly expanded, its inner surface being recessed to receive the ridge on the interclavicle with which it articulates. The upper end of the clavicle is expanded in a plane at right angles to the lower so as to rest on the outer surface of the acromium, and undoubtedly also to a large extent on the lower end of the cleithrum. Its inner surface is excavated so as to clasp the bones with which it articulates. The anterior part of the interclavicle is preserved: it forms a wide flat plate with a ridge along the middle line of its ventral surface and two low ridges starting from this and passing out at right angles. These give attachment to the clavicles, which nearly meet in the middle line. So far Fie. 13. _Right ¢ clavicle of Dicynodun haili. A, from behind, lower surface upwards, scapula end to the left. B, ventral surface, anterior border upwards. x . as preserved, the bone is extremely similar to that of a Cynognathid illustrated in fig. 9. " The sternum is not preserved, but in the very closely allied D. microtrema, Seeley, is a hexagonal plate quite similar to that ascribed by Owen (Q.J.G.S., vol. xxxvi. pp. 414-424, pls. xvi. and xvii.) to Platypodo- saurus. : Material of D. microtrema shows exactly the mode of articulation of the clavicle with the scapula and the relation of the coracoid to the sternum. | If the scapulo-coracoid of D. halli is placed with the ventral edge of the precoracoid and coracoid parallel to the middle plane of the animal, the upper end of the scapula is thrown forward exactly as in Echidna and Ornithorhynchus; that this position is the normal one is shown at once by articulating the clavicle, whose connections at both ends were The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 33 certain. It is quite impossible to place these bones in any other position -so as to give a possible form to the animal’s neck and thorax. That this position is the normal one in Dicynodonts is shown by numerous skeletons still connected by matrix, in which the upper end of the scapula always lies far in advance of the lower. The humeri of D. halli are perfectly preserved. They have expanded upper and lower ends at an angle of about 60° with each other. Fie, 14.—Left humerus of Dicynodon halli, x. Viewed ventrally the head is straight, only its middle portion being occupied by the articular surface, which is scarcely visible in this view, but which extends well over into the dorsal surface. From the anterior end of the head the enormous delto-pectoral crest is developed as a flat plate standing nearly vertically when the skeleton is naturally articulated. The posterior end of the head is thickened to form a powerful lesser tuberosity from which a strong muscular ridge runs down the inner side of the bone to terminate in a large well-roughened process for the insertion of the latissimus dorsi and scapulo-humeralis VOL. Lil. (THIRD SER. VOL. XIII.)—OcT. 1917, 3 34 Lieut. D. M. S. Watson posterior. The shaft is short and thick, crossed on its ventral surface by a ridge which continues the lower extremity of the pectoral crest and extends on to form the bridge over the entepicondylar foramen. The expanded lower end of the bone bears on its front (ventral) surface a small round but flattened condyle for the head of the radius. The ulna condyle lies very largely on the end of the bone, but extends on to both {/ Fie. 15.—Right radius and ulna of Dicynodon halli. A, humeral ends, radius on the right. B, posterior surface. x 4. broad surfaces. The ectepicondyle is massive, and bears on the sharp outer edge a very distinct process for the origin of the supinator longus. The entepicondyle is of considerable size, but presents no interpretable traces | of muscle insertion. The radius is a straight slender bone with a slight concavity on its proximal end which fits the flattened condyle on the humerus. The ulna is a still more slender bone, whose expanded upper end lies partly behind the radius and is divided into two parts—the outer formerly continued by cartilage to form an olecranon, the inner articulating. with the condyle on the humerus. The bones of D. halli and closely allied species at my disposal do not The Evolution of the Tetrapod Shoulder Girdle and Fore-limb — 35 show muscular insertions sufficiently clearly to demand a detailed discus- sion, but give evidence that the whole arrangement agreed in its general structure with Cynognathus One feature, however, of great importance is that there must have been a distinct though small supraspinatus inserted on the inner surface of the scapula and passing downwards over the Fig. 16.-—Right side of the shoulder girdle and fore-arm of Dieynodon halli, Watson, in natural position, from the type specimen. The specimen stands on a rectangular surface, the hinder edge of which is under the animal’s middle line. x #. - smooth notch below the acromium to be inserted on the greater tuberosity, which lies just below. The bones of D. halli are so well preserved that there is no difficulty in _ articulating them and studying their movements. When the shoulder girdle is in its true position, the humerus in its extreme anterior position (fig. 16) stands out nearly at right angles to the body and lies with the outer end of its axis pointing upwards. 36 Lieut. D. M. S. Watson The fore-arm stands at right angles to the broad plane of the lower end of the humerus, coming down to the carpus at an angle of about 45° with the ground. The flat radial condyle suggests that the possible amount of flexion and extension of the fore-arm was small, and I think that there is little or no motion possible in a plane at right angles to this. The humerus can be moved only in one plane, or at any rate its possible motion out of this plane can only be very small, so that it can be depressed and moved backwards, through a considerable arc, at about 45° to the ground. Such motion raises and drives forward the animal, and a flexion of the fore-arm at the same time throws the animal towards the side considered, sufficiently to get the centre of gravity within the triangle formed by this fore-foot and the two hind-feet. The very short fore-arm and its limited motion on the humerus suggest that the fore-foot never . actually lay below the animal’s body, and that as a measure for increasing the stability the hind-foot of the opposite side was in its most advanced position when the fore-foot was retired. The order of putting the feet down would be: left fore, left hind, right fore, right hind. Consideration of these possibilities suggests that the track was wide and the stride fairly long. Fortunately, we have actual tracks which were in all probability made by Dicynodonts. In the Cutties Hillock beds (Upper Permian) of Elgin, Scotland, which have yielded a large number of individuals of the Dicynodont Gordonia and Geikiea, whose skulls and fragmentary skeletons agree with the South African Dicynodons such as D. halli in all important features, I found a fore and hind footprint of a form which is represented by numerous well- preserved tracks in the Cummingstone beds of the neighbourhood and in numerous other localities in Britain. For a description and figures of these tracks see H. G. A. Hickling, Manchester Memoirs, vol. liii, No. 22, pls. i.—iv. The footprints of these tracks are short, five-fingered and clawed, quite similar to those which should be made by Dicynodon feet (e/. Owen, Q.J.GS., vol. xxvi. pl. xvii. fig. 5). The tracks are wide, the prints of the feet of opposite sides being distant from each other, and the stride as long as the width of the. track; the smaller hind footprints often follow closely on the fore foot- prints of the same side, and the whole agrees very closely with the type of walk inferred from the characters of its bones to have been used by Dicynodon, In fig. 17 I have illustrated the seapulo-coracoid, interclavicle, and sternum of a small Dieynodon from the Endothiodon beds of Beaufort The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 37 West. In this individual the scapula is even more markedly inclined forward than in D. halli, and the precoracoid is considerably bigger, both relatively to the scapula and to the coracoid. The complete interclavicle is unusual in its considerable contraction behind the arms which give attachment to the clavicle, agreeing in this feature with other interclavicles of small Dicynodons from the same beds, and differing from the larger later forms. Fic. 17.—Left scapulo-coracoid, interclavicle, and sternum of a small Dicynodon from the Endothiodon zone of Beaufort West. ~ x 1. The sternum is a single pentagonal element giving evidence of the attachment of two sternal ribs, but otherwise a mere flat plate. Several. specimens of large and small forms give definite evidence that it was not followed by any other ossified sternal element. The very large Dicynodont Kannemeyeria, from the Cynognathus beds (Lower to Middle Trias) of South Africa, is represented by splendid material in the British Museum, from which I have drawn a few bones in figs. 18 and 19. Its shoulder girdle on the whole resembles that of D. halli, but. the ‘notth below the acromium for the supraspinatus is better developed, and Fic. 18.—Kannemeyeria sp. A, right scapula, precoracoid, and coracoid. B, right scapula. ©, right cleithrum. x #4. 7 \ BY)! j : (\\ | \\\ ‘\ } \ | : 1 A Fic, 19,—Kannemeyeria simocephalus (Weithofer). Left humerus. x 4. The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 39 the precoracoid and coracoid are smaller in proportion to the scapula and the coracoid larger when compared to the precoracoid. The glenoid cavity faces very much more directly backwards than in D. halli, the articular face on the coracoid lying almost entirely caudal to the scapula. The humerus generally resembles that of D. halli, but differs in the narrower lower end. and in the much more extended and better rounded condyle for the radius, and in the fact that the expanded ends lie much more nearly in one plane, being placed at an angle of about 30°. The bones of the fore-limb are similar to those of D. halli. There is no doubt that in Kannemeyeria the humeri did not project so strongly from the animal’s body as in the earlier form, and that in consequence the walk more nearly resembled that of Cynognathus. LYSTROSAURUS. _ The aquatic Dicynodont Lystrosaurus much resembles Dicynodon itself in all the bones of the shoulder girdle and fore-limb, but is remarkable, at any rate in some species, for the very large size of the supraspinatus notch below the acromium. MONOTREMES. The shoulder musculature of the Monotremes has been repeatedly described, best by W. J. S. M‘Kay (Proc. Linn. Soe. N.S.W., vol. ix., 1895, pp. 263-358, pls. xx.-xxiii.), who reviews all previous work. The morphology of the shoulder girdle was discussed by J. T. Wilson and M‘Kay (Proc. Linn. Soc. N.S.W., vol. viii, 1894, pp. 377-387, pl. xxi.), who cleared up the whole problem of the homologies of the borders and surfaces of the scapula with those of an ordinary mammal in a clear and convincing manner. I wish especially to express my indebtedness to this paper, which first led me to the views of muscle homologies which are. adopted in this paper. The most significant features are :— That there are epicoraco-brachial and epicoraco-humeral muscles attached to the anterior coracoidal element which are unrepresented in other mammals; that the long head of the biceps is attached to the same bone; that the spine and acromium form the front edge of the scapula, the insertion of the supraspinatus being entirely on the inner surface; and - that the subscapularis arises largely from the outer surface of the scapula. C. Westling (Bihang till K. Afd. Svenska Vet.-akad. Handlinger, Bd. xv. Afd. iv. No. 3, pp. 3-71, pls. i-vi.) gives good figures of the humerus of Echidna with the muscle insertions marked. 40 Lieut. D. M. S. Watson TRICHOSURUS. R. Broom in 1899 made the very interesting discovery that in small pouch young of Diprotodont marsupials the coracoid extended inwards as a large cartilage and finally became continuous with the sternum. Subsequently he published an illustrated account of the development of the shoulder girdle in these animals, with many graphical reconstructions from serial sections. Later he discussed the Polyprotodonts. Although examination of series of embryos completely confirmed Broom’s results, I thought it would be interesting to see the structures concerned in the solid and to get further Acr. Cr. Omo.St GWT MYL i a we fi Oy) Uf } He mua ee Oe, MLZ Za jue Ie | 5 am Sc é RUM. 3c. OT: Fic, 20.—Wax-plate model of the shoulder girdle of an 11°5-mm., Trichosurus embryo. Right side of the shoulder girdle from without. x 33. Aer., acromium; Cl, clavicle; Hum., head of humerus; Omo. St., omosternum; Sc., scapula ; St., sternum, light on the musculature. I therefore made a wax-plate model of an 11°5-mm. embryo of T’richoswrus vulpecula from one of Professor J. P. Hills’ series. In this model, which is illustrated in figs. 20 to 24, the skeletal structures are shown on the right side and the shoulder mus- culature (except the rhomboid) on the left. The upper end of the scapula lies opposite the last three cervical and first dorsal vertebrae; it inclines very slightly backwards as in the adult. The blade of the scapula is a thick sheet of cartilage of considerable breadth, with a concave inner surface and an outer surface bevelled off to the anterior margin, and with a slight concavity on the middle of its face. Ventrally the scapula narrows, and from its anterior margin above the glenoid cavity the very large hook-shaped acromium arises by a narrow neck. The acromium is at first directed outwards, then turns downwards, The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 41 and finally points inwards; it is throughout a thick rod, and lies in a _ transverse plane of the animal. The lower end of the scapula turns towards the middle line, having a massive tuberosity overhanging the glenoid cavity, into which the ligamentous long head of the biceps is inserted. The coracoid is a quite massive cartilage, with a low prominence Fie, 21.—Trichosurus as in fig. 20, Right side of the shoulder girdle from behind. Reference letters as before, with :— Cor., coracoid. on its caudal surface just medial to the inner lip of the glenoid cavity. The inner end of the coracoid is in wide cartilaginous connection with the sternum and first dorsal rib. The glenoid cavity is concealed by the head of the humerus in the model and is still not completely formed, the split separating it from that element being not yet carried through ; it is a large shallow concave area facing downwards, backwards, and outwards. The sternum is a massive cartilage whose lateral borders fuse with 42 Lieut. D. M. S. Watson the coracoid and first rib, behind which the width is very much reduced, so that the presternum forms a pronounced knob. : The clavicle is already considerably ossified and shows no trace of cartilage, consisting only of dense connective tissue. It takes the form of a nearly straight rather massive rod, the tissue of which it is composed \ Gag Hum. LCCor. CL. Omadt Vic, 22.—Trichosurus as in fig. 20. Right side of the shoulder girdle from in front. Reference letters as before, with:— ~ L, Cl. Cor., coraco-clavicular ligament. . being continuous with the perichondrium of the acromium at one end and that of the sternum at the other. The omosternum at this stage and in the embryo from which the model was made consists entirely of dense connective tissue exactly similar to that of the clavicle. It forms a small triangular patch with straight anterior surface lying ever the inner ends of the clavicles and sternum, Anteriorly it is very markedly separated from these elements by deep slits passing inwards from its lateral sides (cf. fig. 25), but i The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 43 posteriorly it becomes continuous with the thick perichondrium of the ventral surface of the sternum. In another embryo of the same size (11°5 mm.) it still contains no trace of cartilage, but does show the very beginnings of ossification in - two points of its antero-lateral regions. The humeri are still entirely cartilaginous, being of course much shorter and thicker than in the adult. Théy stand directed backwards and ‘outwards, and ‘with their distal ends more dorsally placed than the proximal. The head is large, rounded, and faces very largely dorsally, TRAP. Fick. IRac. Suede INE OP Devtr. Pecr, Bra. Ant EHTAL ScTr. DevSe. TerMa. Fig. 23.—Trichosurns as in fig. 20. Left shoulder musculature from outside. Refereice letters as before, with :— Acr. Trac,, acromio-trachealis ; Del. Cl., clavicular deltoid; Del. Sc., scapular deltoid ; FE, H. Tri., external humeral head of triceps; Inf. Sp., infraspinatus; Pect., - pectoralis; Sc. Tri., scapular head of triceps; Sup. Sp., supraspinatus; Ter. Ma., teres major ; Trap., trapezius. the greater and lesser tuberosities being well marked and much separated so as to leave a wide shallow bicipital groove. Musculature. So far as they are represented in the model, the serrati arise from the tips of the parapophyses of the 4th, 5th, 6th, 7th cervical vertebrae just. outside the vertebraterial canal and from the outer margin of the proximal part of the first dorsal rib; how far backwards they extend I have not determined. They rapidly approach one another, forming a single sheet of muscle which is inserted into the dorsal part of the peri- - chondrium of the front edge of the scapula, into the imner surface of 44. Lieut. D. M. S. Watson that cartilage just below its dorsal border, and into the connective tissue covering its caudal edge. The omohyoid muscle is a thin strip passing under the serrati and inserted into the visceral surface of the scapula ventral to the insertion of these muscles. The acromio-trachelian is a powerful muscle inserted into the anterior surface of the outer part of the acromium and widening rapidly as it passes forward over the supraspinatus. The trapezius is a thin sheet inserted into the ventral surface of the clavicular end of the acromium, some of its outer fibres extending on to become continuous with the clavicular deltoid. The cleido-mastoid and cleido-occipital are small muscles ‘of circular section inserted into the front surface of the middle of the clavicle. The sterno-masteid is a large muscle arising entirely from the lateral part of the front face of the omosternum. The subclavius is a single flattened muscle arising from the first dorsal rib just dorsal to its fusion with the coracoid, and passing directly forward over the coracoid and scapula to be inserted into the extreme outer end of the clavicle and into the fascia surrounding the supraspinatus as it plunges under that bone. Two or three parts of the pectoralis major are shown in the model. The first arises from the median line of the ventral surface of the sternum and omosternum, and passes outwards to join the lateral clavicular part. The second part is really only composed of the partially separated deep layers of the first. The third arises from the outer end of the clavicle, and faces backwards and outwards, becoming continuous with the first until finally they are all inserted into the pectoral crest of the humerus. The pectoralis minor is a small muscle arising from the ventral surface of the coracoid and passing almost directly backwards to its insertion into the posterior part of the sternum. Throughout it lies deep to the pectoralis major. The clavicular deltoid arises from the lower part of the acromium, and is continuous with some of the outer fibres of the trapezius; it passes from here directly backwards to the humerus. The scapular deltoid arises from the thick sheet of connective tissue which joins the acromium and separates the supra- and infraspinati. It lies over the latter muscle, and runs downwards and backwards to its — humeral insertion. The supraspinatus arises from: a very small area of the outer surface, from the anterior margin and largely from the inner surface of the blade of the scapula. It passes downwards between the acromium and the ee + ™ The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 45 glenoid part of the scapula to be inserted on the outer tuberosity of the humerus. The infraspinatus has its origin from the outer face of the blade of the scapula, and passes downwards caudal to the acromium to its insertion near but dorsal to the supraspinatus on the great tuberosity. The teres major is a thick muscle having its origin from the extreme posterior strip of the outer face of the scapula and the posterior edge of that cartilage. It passes downwards caudal to the scapular head of the triceps, and with its inner side in contact with the subscapularis and serratus. Acer. Ct. Omoor. Oty Our. OP. Acr. Trac. Hum. Bull Cor. Co. Pecr. Trae = Dew.C. EH Tre Fic, 24,—Model of Trichosurus as in fig. 20. Shoulder girdle and left shoulder musculature from below. Reference letters as before, with :— Bi. L. H., long head of biceps tendon ; Cl. M., cleido-mastoid ; Cl. Oc., cleido-occipital ; Co., 1st dorsal rib ; 0. ‘Hy., omohyoideus. The sub-scapularis arises from the inner surface of the scapula below the omohyoideus and serrati and behind the supraspinatus. It lies dorsal to the subclavius, and finally winds round over the glenoid region of the scapula to its insertion into the lesser tuberosity. The biceps arises by a long ligamentous head from the tuber of the scapula in front of the glenoid cavity and from the process in the caudal surface of the coracoid. The two parts soon meet and have their usual course. The coraco-brachiales, brevis and longus, arise near the coracoid head of _ the biceps from the posterior and dorsal surface of the coracoid, and cross that bone in their course to be inserted into the lower surface of the humerus. | The triceps and brachialis anticus present no feature of special interest 46 Lieut. D. M. S.- Watson except that their insertions extend well up to the head of the humerus. There is a well-marked coraco-clavicular ligament connecting the anterior face of the scapula, where it passes into the coracoid, with the outer end of the clavicle. There are no traces of subcoracoid or supracoracoid muscles. I have confirmed Broom’s account of the breaking down of the inner end of the coracoid into a ligament in older individuals of Trichosurus. ti Cornu. Sus.Cu. Sup. Sp. Cu. Omo.St. | Huw. Fic. 25.—Photograph of a transverse section of an 11°5-mm. Trichosurus, to show the distinctness of the omosternum. Reference letters as in preceding figures. In a preparation of a much larger embryo (18 em.), which Mr Tan was good enough to make for me, the ligament passes from the end of the coracoid to the outer edge of the presternum just in front of the insertion of the first rib. The cartilaginous coracoid is still quite large, forming half the glenoid cavity, and contains a single bone. Discussion OF THE EVIDENCE PRESENTED IN THE FOREGOING PAGEs. The chief outstanding problems of the shoulder girdle of the animals discussed in this paper are :— 1. Are the clavicles of a mammal strictly homologous with those of Labyrinthodontia 7 The Evolution of the Tetrapod Shoulder Girdle and Fore-limb — 47 2. What is the fate of the interclavicle in Theria ? 3. Which of the two coracoidal elements of Cynognathus is the mammalian coracoid ? 4. Is the “epicoracoid” of the Monotreme asad: girdle homologous with the anterior coracoidal element in early reptiles, or is it a neomorph corresponding to the “epicoracoid ” of a lizard ? 5. With what muscles of a reptile do the supra- and infraspinati and teres major of a mammal correspond ? 1. The doubt whether the mammalian clavicle is homologous with that of reptiles was first introduced by Gegenbaur, who was a profound believer in the hypothesis that if two bones developed, one in cartilage and the other in membrane, they could not be homologous. Gegenbaur found that in man the clavicle develops in cartilage, and hence argued that it could not correspond solely with a membrane bone of dermal origin such as the clavicle of the lower Tetrapods was known to be, but must also include a vestige of the old Urodele precoracoid. Broom (1899, Trans. Roy. Soc. Edin., vol. xxxix. pl. iii. No. 29) showed that in man the ossification begins before there is a real cartilage in the precartilaginous clavicular basis, and hence with perfect justice pointed out that it was really a membrane bone. Faweett (Jowrn. Anat. and Phys., vol. xlvii. pp. 225-234) has given an excellent account of the whole process of ossification, extending previous descriptions. He shows that ossification does take place in the “curious tissue, generally alluded to as a peculiar form of precartilage,” before the appearance of any genuine cartilage cells, but starts from two centres, perhaps correlated with the muscle insertion, and that subsequently _ cartilage appears in which ossification then takes place. These facts seem to establish the view that the cartilage in the human clavicle is really a neomorph, analogous to that which develops in the mam- malian lower jaw (the reptilian dentary, a pure membrane bone) and in the mammalian pterygoid and other facial bones. The clavicle of Dicynodon differs in no respects except shape, and its inner connection with the inter- clavicle, from the Therian clavicle, and, like that bone, undoubtedly gave origin to tne Jeltoid and insertion to the trapezius and cleido-mastoid muscles. There is no doubt that the clavicle of Dicynodon is strictly homologous with that of Cynognathus, which in all its relations, and even in shape, agrees with that of the Pelycoseurs, where in some forms (Ophiacodon) its lower end is ornamented on the eter surface, and undoubtedly lay in the skin just as does the similar bone in Seymouria and the large amphibia. 48 Lieut. D. M. S. Watson We may therefore regard it as certain that the Therian clayicle is homologous with that of Stegocephalia, and contains no other elements whatsoever. Se tt 2. Interclavicle. The interclavicle in the rachitomous amphibia and in Seymouria is a flat plate shown by its ornamented ventral surface to lie in the skin. From this bone, obviously of purely dermal origin, there is a complete series, passing through the Pelycosaurs, Deinocephalia, and Gorgonopsia to Cynognathus, where it is a broad flat plate with a median ridge which ~ no doubt separated the insertions of the pectoral muscles. That this element is homologous with the interclavicle of Echidna has never been doubted ; their relations to the clavicles and sternum carry conviction. In lizards (Fiirbringer, Jena. Zeit., vol. xxxiv. pl. xv.) the cleido-mastoid and trapezius have an insertion on the interclavicle, caudal of the clavicle, and the muscle is continued by the interclavicular-sternal ligament to the sternum. In Sphenodon there appears to be no insertion of this muscle into the interclavicle, although the ligament is still present. In Echidna, the superficial portion of the m. episterno-cleido-mastoideus is inserted into the anterior face of the interclavicle, and the sterno-mastoid passes over this bone to an insertion on to the ventral surface of the sternum. Both Monotremes have the power of depressing the head to an enormous extent, until the lower surface of the mandible rests on the ventral, surface of the chest in Echidna. It seems very probable that these two muscles, which have an origin from contiguous parts of the skull, and run down the neck together, are really separated parts of one muscle, originally inserted on to the interclavicle, and that the more superficial part has shifted its insertion backwards as an adaptive modification to allow of this great depression of the head. If this is so, both must be derived from the inter- clavicular part of the cleido-mastoid of lizards, in which indeed the inser- tion is functionally carried back by ligaments to the sternum. The Theria in their adult state present no definite traces of an inter- clavicle, but the conditions shown in my model of the 11:5-mm. Trichosurus shoulder girdle suggest that this element is really represented by tue osternurm (a view already held by Gegenbaur so far as concerns insecti- vores). The mass of dense connective tissue representing this element is clearly marked off trom all others in front, and has a definite sharply defined shape. Its relations to the inner ends of the clavicles, coracoids, and sternum are altogether those of an iuterclavicle, and it gives insertion to the sterno-mastoid, which, as I have tried to show above, the interclavicle originally did. Finally, the fact that ossification commences in it before The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 49 ‘the appearance of any cartilage cells shows that i is the repre- _Sentative of some membrane bone, and none except the interclayicle ever appears in this region of the Tetrapod shoulder girdle. 3. The Homology of the Mammalian Coracoid. It is now generally believed on abundant evidence that the Tetrapods: arose from the Rhipidistia—that is, the Osteolepid fish. In these animals an ossification in the cartilaginous shoulder girdle usually occurs, which in good material of Rhizodopsis and Megalichthys which I have examined consists of a single small bone on each side, bearing, in the latter genus at any rate, a deep hemispherical glenoid cavity, below which it is extended towards the middle line. In the embolomerous amphibian Pteroplax the scapulo-coracoid is _ a single plate-like bone of somewhat irregular shape, with a distinct scapular blade. : In the rachitomous amphibians Archegosaurus and Actinodon, there is conclusive evidence that the scapulo-coracoid is a single ossification, and no evidence has been brought forward to show that any Lower Permian rachitomous amphibian had more than one bone in each side of its carti- laginous shoulder girdle. There are thus considerable grounds for believing that the reptiles are descended from ancestors which had only a single scapulo-coracoid element surrounding the glenoid cavity and extending into both scapular and coracoidal regions. It is possible that this condition is preserved in the Diadectids, where no specimens have shown any sutures in this region. Seymouria is in nearly all ways the most primitive of known reptilia; the only feature which is unquestionably advanced is the loss of the cleithrum. It is remarkable that in this animal there are only two elements in the scapulo-coracoid; as there can be little doubt that some reptilian ancestor had only one, it is not improbable that when (for some quite obscure reason) the single bone was broken up in reptiles it was at first ossified from two centres, a third being subsequently added. Thus in this character Seymouria may actually be primitive. Other groups of Cotylosaurs (Captorhinids, etc.) have clearly two coracoidal elements. Case (1910, Carnegie Inst. of Washington Publ. 55, p. 157) endeavoured to show that the Poliosauride are the most primitive group of Pely- cosaurs. and may be regarded as morphological ancestors of the larger forms allied to Dimetrodon, “ Theropleura” being an intermediate form. I have recently supported this view by quite new evidence drawn from the brain-case and stapedial and temporal regions which was not available VOL. Lil. (THIRD SER. VOL. XIII.)—OCT. 1917. 4 Pd 50 Lieut. D. M. S. Watson at the time of Case’s discussion (Bull. Amer. Mus. Nat. Hist., vol. xxxv. arts. xxxi. and xxxil. pp. 611-648, 1916). I have shown (Proc. Zool. Soc., 1914, pp. 749-786) that the Deinocephalia are the most primitive as they are the oldest known group of South Fic, 26.—Series of Anomodont scapulo-coracoids reduced to the same size. A, Varanoops, after Williston ; B, Ophiacodont, gen. nov,, after Williston ; C, Dimetrodon ; D, *‘ Rhophalodon” ; E, Phocosaurus; F, Seymnognathus tigriceps, from the figure of Broom and Haughton ; G, Cynognathus, from the type specimen and the original of fig, 8. African mammal-like reptiles. Comparison of faunas suggests that the Deinocephalia from the copper-bearing sandstones of the Urals are older and more primitive than the well-known South African representatives of the order. Finally (ef. Proc. Zool. Soc., 1914, pp. 1021-1038, and Ann. and Mag. Nat. Hist., ser. 8, vol. ii. pp. 65-79, 1913), there can be no reasonable The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 51 doubt that the Gorgonopside are ancestral to the Cynognathide, being in all ways more advanced than the Deinocephalia. If we place in correct time order, and in the sequence suggested by the known advances in structure of other parts of the skeleton, a series of drawings of the scapulo-coracoids of animals belonging to the groups referred to above, reduced to the same absolute size, we see in an extremely clear way that throughout the histor of the group the anterior coracoidal t element tends to b Varanoops the posterior coracoid is unossified and the anterior one enormous; in Cynognathus the anterior is very much smaller—in fact, scarcely more than half the size of the posterior. General considerations derived from a study of many similar series, which show a steady change of character in a definite direction, make it extremely probable that in still later members of the stock the anterior coracoidal element will disappear entirely. This series of figures (which are accurate drawings made with great care) shows also that, whilst in the early forms, up to and including the Deinocephalian “Rhophalodon,” the anterior coracoidal element supports part of the glenoid cavity, in all the later members the articular surface lies entirely on the scapula and posterior coracoidal element. The evidence that the mammals arose from this group of reptiles is now so extensive and so varied in character that this hypothesis is as * well supported as any phylogenetic speculation. The resemblance of the ’ cartilaginous coracoid of the 11°5-mm. Trichosurus to the bony coracoid of Cynognathus is so close in all ways as to establish their homology ; and as the one bone developed in the coracoid of the marsupial takes part in the formation of the glenoid cavity, there can be no doubt that this element, the true_ mammalian _coracoid, is homologous with the ior element in the Anomodont shoulder girdle; for it is inconceivable that , the anterior element, which is undergoing steady reduction in this group and is finally excluded from the glenoid cavity, should suddenly reverse its evolutionary trend, functionally replacing a bone which is steadily growing larger and playing an increasing part in the glenoid cavity. That this view is correct is indicated also by the fact that the marsupial coracoid gives origin to the coraco-brachialis and short head of the biceps, which were presumably attached to the posterior element in Anomodonts, whilst the long head of the biceps, originally attached to the anterior element, has migrated into the scapula, and the subcoracoid and supra- coracoid muscles attached to that element have totally vanished. We are therefore justified in calling the posterior element coracoid, and i reecorscaid, ae 52 Lieut. D. M. S. Watson 4. The Homology of the “ Epicoracoid ” of Monotremes. The “epicoracoid” of Monotremes was regarded by Owen and other early workers as the homologue of the precoracoid of the early reptiles. Later authors—Parker, and at the present time Gregory (Ann. New York Acad. Sci., vol. xxvi. pp. 317-383, pl. iv.)—believe it to be a neomorph homo- logous with the imperfect ossification or calcification which is so largely developed in the large anterior cartilaginous continuation of the coracoid of lizards. This view appears to be founded mainly on the fact that it does not articulate with the scapula as the precoracoid does in Anomodontia. In Cynognathus (figs. 8 and 9) the small precoracoid projects very considerably in front of the scapula, and only meets that bone in a relatively short suture. The anterior hook formed by this bone in the Cynodonts is on the whole strikingly like that formed by the “epi- coracoid” of Monotremes. That the Monotremes have never lost their precoracoid, as the other mammals have done, is, I think, demonstrated by the occurrence in them of epicoraco-brachial and epicoraco-humeral muscles exactly corresponding in their attachments and relations, and, if we may use Sphenodon as a term - of comparison, in their innervation to the m. subcoracoideus and m. supra- coracoideus of the Anomodonts. If, then, the preservation of these muscles . shows that the Monotremes never completely lost the precoracoid, it seems an unnecessary complication to refuse to see in their epicoracoid (a probably secondary enlargement of) the precoracoid. The fact that the long head of - the biceps arises from the “ epicoracoid” in Monotremes, instead of from the scapula, seems to afford an additional reason for regarding that bone as the precoraceid. This secondary enlargement of the precoracoid in Monotremes is due to the fact that in them the rounded. head of the humerus allows of a considerable antero-posterior motion of that bone in a horizontal plane, which does not occur in Cynognathids but is a special adaptation connected with the mode of burrowing found in both Monotremes, which demands such motion, to throw the earth excavated backwards and outwards, and to enable the animals to get their hands as far or nearly as far forwards as the tip of the snout. This motion is performed largely by the epicoraco- brachial and epicoraco-humeral muscles attached to the precoracoid, 5. Determination of the Homologous Mesactes im Reptiles and Mammals. In the descriptive part of this paper I have examined in considerable detail the structure, possible motions, and musculature of the shoulder in mahy forms. The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 53 In the series of figures in fig. 26 I have indicated by a thick dotted line the direction of motion of a point connected with the distal end of the humerus. This series shows how in a perfectly regular way the direction of motion of the humerus changes from a plane parallel to the ground in Varanoops to one at right angles to it in Cynognathus. This change of direction is dependent on the fact that the anterior part of the glenoid cavity, which at first lies very low down—in fact, nearer to the ground than the caudal end—gradually rises when traced through the three families of Pelycosaurs, the Deinocephalia, ete. The process is as if the anterior and posterior parts of the glenoid cavity were twisted round on an axis passing through the middle of its length, the cranial end being raised and thrust backwards, the caudal end depressed and moved towards the head. The exact resemblance between the insertions of the scapulo-humeralis anterior on the humerus of a Pelycosaur and of Sphenodon raises a strong presumption that in the Permian reptile, as in the living one, this muscle had its origin from the outer face of the scapulo-coracoid over the scapulo- precoracoid suture. In Cynognathus (p. 27) there is conclusive evidence that this muscle occupied the whole exterior face of the scapula between the two strong flanges of that bone. This being the case, it follows that this muscle has been twisted round in the same way as the glenoid cavity, so that it no longer runs antero-posteriorly but dorso-ventrally. This alteration of the origin of the scapulo-humeralis anterior necessarily involves a similar change of the scapulo-humeralis posterior, which, from an origin on the outer surface of the blade of the scapula in Pelycosaurs, comes in Cynognathus to have an origin on the posterior edge of the scapula. Its insertion in both forms is on the dorsal surface of the posterior edge of the humerus below the head and in close connection with the latissimus dorsi. In Cynognathus its origin and insertion are identical with those of the teres major in both Monotremes and the 11°5-mm. Trichosurus, and I think the homology of the muscles is clear. In Pelycosaurs (p. 16) the muscles fall into two groups: one solely concerned with the support of the animal, and with resisting rotation of the humerus; the other group, the supracoracoid and scapulo-humeralis anterior, clavicular deltoid, and the subcoracoideus, coraco-brachialis longus, and posterior part of the pectoralis, are concerned with the advance and retirement of the humerus, and the animal is driven forwards solely by them.. _ The subcoracoideus and supracoracoideus can have no function except 54 Lieut. D. M. S. Watson the antero- posterior motion of the humerus on a plane parallel to the ground ; and when in Cynognathus this motion, if not totally impossible, at any rate becomes of very slight importance to the animal, the muscles are reduced, the precoracoid, whose main function is to give origin to them, being of correspondingly small size. Comparison of the discussion of the mode of walking in Cynognathus (p. 29) with that of a Pelycosaur (p. 13) shows that the former is superior in all ways :-— 1. The track is narrower, and hence the animal’s weight is not thrown so violently from side to side. 2. The stride is longer. . 3. The motions at the elbow are seecacien to a single plane, and the more vertical position of the fore-arm allows of a great reduction of the flexor muscles of the fore-arm. — 4. All the muscles passing from the animal’s body to the arm are used at one phase or another to drive the animal forwards; in particular, the large scapulo-humeralis anterior plays a very important part in pulling the humerus back at the beginning of the stroke. Thus by these modi- fications Cynognathus gains a far more effective progression, and does it with a much smaller mass of muscles whose centre of gravity tends to be nearer to the body. The Theria present a still further advance on Cynognathus. In them the humerus swings round until it lies in a plane nearly parallel to the sagittal plane of the animal, and the relatively long legs allow of the feet being placed under the centre line of the body and retained there through- out their motion, so that the body scarcely sways at all from side to side The absence of any great stress resulting from the animal’s weight tend- ing to move the fore-arm At right angles to the plane, including the axis of the humerus, and normal to the broad surface of the lower end of the humerus allows of a very great reduction of the flexor and extensor muscles, the animal’s weight being largely taken by the triceps. The existing musculature of Therians is probably not directly explicable on any mechanical theory postulating ordinary walking as the end to be obtained, because there are strong reasons for believing that all Theria, both marsupials and eutherians, have passed through an arboreal stage at an early period of their history. To the tree-living adaptations then acquired are perhaps to be attributed, amongst many other things, the round head of the humerus, the comparative slenderness of that bone, the great reduction of the coracoid, and the division of many originally single muscles into distinct parts, which occur in mammals. . The Monotremes are unique amongst living Tetrapods in that the The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 55 acromium and spine form the real anterior edge of the scapula, there being a well-defined m. supraspinatus arising from the inner surface of the bone and running down through the fossa formed by the scapula, clavicle, and precoracoid to be inserted on the greater tuberosity of the humerus in close relation to the infraspinatus. Study of the development in man (W. H. Lewis, Amer. Jour. of Anat., vol. i., 1901, pp. 145-182, pls. i—xi.) and Trichosurus (p. 40) shows that in these types in early stages the supraspinatus arises from the inner surface of the scapula and from its anterior edge, a fact which establishes a strong presumption that the Monotreme condition is really primitive. That this view is correct is rendered almost certain by the fact that the conditions in Dicynodonts, the only reptiles with a typical supraspinatus, are identical with those in Monotremes so far as concerns these statements. The Dicynodonts are, however, different, and more primitive than the Monotremes in retaining an anterior part of the scapula, which articulates with the precoracoid and lies on the inner side of the lower part of the supraspinatus. In Cynognathus, where I have shown (p. 28) there is no supra- spinatus, the acromium lies in the same position as in Dicynodon, and that part of the scapula which articulates with the precoracoid is obviously homologous in the two genera. Now, in Cynognathus that part of the outer surface of the scapula which lies below the acromium is sharply depressed, and affords origin to a special separated part ofthe m. scapulo- humeralis anterior inserted on the dorsal surface of the humerus close up to the head and extending forwards to the region of the greater tuberosity. Expansion of the muscle, either in length to give increased range in action, or in thickness to increase its power, can only take place either by extending the scapula forwards, which makes its pull more horizontal, or by winding the muscle over the edge and giving it an insertion on the inner surface of the scapula, a course which gives it a more vertical action. As the mode of walking in Cynognathus calls for no increased motion of the humerus in a horizontal plane, it is evident that the latter is the modification which is most likely to ensue. Once the detached lower portion of the scapulo-humeralis anterior has attained an origin from the inner surface of the scapula, its pull, transmitted over the edge of the scapula between the acromium and the precoracoid suture, will tend to emarginate this border, producing the notch which we actually find in Dicynodon. Further reduction of the precoracoid leads directly to the condition in Monotremes. This account of the origin of the supraspinatus from a detached lower anterior portion of the scapulo-humeralis anterior, and of the infraspinatus 56 Lieut. D. M. S. Watson from the rest of that muscle, is consistent with their innervation in mam-. mals and Sphenodon and their neighbouring insertion on the humerus. In Theria the supra- and infraspinati to a large extent change their function. In Cynognathus, where their insertion is on to the dorsal surface of the humerus distal to the head, they raise that bone—a function also efficiently performed by the teres major, latissimus dorsi, and scapula — deltoid. In Theria, by the migration of their insertions on to the lateral surface of the greater tuberosity, and the throwing forwards of that region in front of the head of the bone, they become depressors of the humerus, taking over in a more simple way the function of the coraco-brachial muscles, which are hence reduced in size and somewhat altered in function, the coracoid shortening rapidly in sympathy when it no longer needs to extend far down and inwards to give origin to a muscle depressing the humerus. The only outstanding mammalian shoulder musele whose origin is unknown is the teres minor. CONSIDERATION OF EVIDENCE COMMONLY REGARDED AS OPPOSED TO SOME OF THE CONCLUSIONS REACHED ABOVE. In the preceding discussion I have carefully refrained from discussing the views of previous authors, because it is impossible to make a long argument intelligible if it is constantly interrupted by references to other views often themselves involving knowledge of a long series of argumen- tative statements, Discussion of the shoulder girdle has long ranged round the homo- logies of “the coracoidal elements, and has in my opinion been led into most unprofitable paths by extended comparison with the Anuran shoulder girdle. The Anura are a group of extraordinarily specialised animals of totally unknown ancestry, and why their shoulder girdle, quite unlike that of all other Tetrapods, should commonly have been taken as a primary term of comparison, as if its morphology were well understood, is obscure except for the idea that it could be made to square with Gegenbaur’s view that the mammalian clavicle had a cartilaginous base and was a representa- tive of the preci region ‘of the cartilaginous shoulder girdle of Urodeles. As we now know (vide p. 47, antea) that the mammalian clavicle is considerably ossified before any cartilage cells appear in its matrix, and as it can be traced by a close series of forms down to the dermal clavicles of “aa WS Ta TT ee ae ae ele Cee e ‘. , ™ ' bs, The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 57 the Labyrinthodonts and fish, this view of Gegenbaur’s and the huge _ literature which rests on it have become only of historical interest. Professor G. B. Howse (Jowrn. Anat. and Phys., vol. xxi., 1887, pp. 190-198, pl. viii.) in an important paper adopted Gegenbaur’s view that the mammalian clavicle included a representative of the precoracoid, and was hence driven to the necessity of finding a new homology for the epicoracoid of Monotremes, which he showed to ossify in a suturally separated sheet of cartilage late in life. In this paper Howse showed that in the rabbit the coracoid process contains two centres of ossification—a posterior which forms part of the glenoid cavity, and an anterior which does not enter into the articular surface but gives origin to the coraco-brachialis and biceps short head. Subsequently (in Proc. Zool. Soc., 1893, pp. 585-592) he showed that these two bones occur in many orders of mammals, including man, where the glenoid epiphysis is the posterior element. Professor Howse argued that the comparative constancy of this element shows that it is really of morphological significance, and that it corre- sponds to the posterior element in Monotremes, the universally occurring anterior element being the epicoracoid. It is impossible to bring con- clusive evidence either for or against this view, but the evidence of muscle attachments considered on p. 52, and the total loss in Theria of those muscles whose origin is from the epicoracoid of Monotremes, is so strong as to leave no real doubt that Howse’s view is incorrect, the posterior bone being a neomorph analogous to the cotyloid bone of the mammalian acetabulum. The view adopted by Howse, that the “epicoracoid ” of Monotremes is not homologous with the precoracoid of Urodeles, really rests on Gegenbaur’s incorrect belief that the mammalian clavicle includes a precoracoid representative. Strictly speaking, there is no precoracoid in Urodeles, but only a cartilaginous process to which the term is by courtesy extended. No one who compares the cartilaginous shoulder girdle of such a ' Urodele as Cryptobranchus with that of Deimetrodon is likely to doubt that the precoracoid of the reptile is ossified in a cartilage strictly homo- _ logous with the precoracoid process of the amphibian. The series of forms discussed in this paper demonstrates the identity of the precoracoid of a Pelycosaur and the Monotreme epicoracoid. The name coracoid is therefore strictly applicable to the posterior element of the early reptile shoulder girdle. The Cuvierian name epicoracoid applied to the anterior element in the Monotremes is the technically correct name for the anterior element of the old reptilian shoulder girdle; but owing 58 Lieut, Do Meee leon to the fact that this term has been constantly applied by many anatomists to the unseparated cartilaginous extensions of the coracoids of lizards, and of the coracoidal ends of the shoulder girdles of frogs, in which true bones are never found, it seems advisable to discontinue the use of the word in its original sense to avoid grotesque confusions. The term precoracoid is unfortunately subject to very similar draw- backs, having—in addition to its legitimate use for the anterior part of the coracoidal cartilage in Urodeles and bones which develop in a homologous region in early amphibia and reptiles—been applied by Parker, and especially Fiirbringer, to a mere process of the single coracoidal element in lizards and birds—a totally incorrect and quite unnecessary use. In this paper I have shown that many early reptiles and amphibia are characterised by a remarkable screw-shaped glenoid cavity, which restricts the humerus to a single definite motion, and through a long series of forms have traced the changes, all in the direction of greater mechanical efficiency, leading from this type to the mammalian condition. I regard it as certain that this type of glenoid cavity is the primitive form, or rather that it represents a condition through which all groups of amphibia- and reptilia have passed. In fig. 27 I have drawn of the same size two shoulder girdles of Pariasaurs from the Tapinocephalus zone and two from the later Ciste- cephalus zone. ‘This series shows a rotation of the glenoid cavity, and probably, from the high position of the acromium in the later forms, of the m. scapula-humeralis anterior, exactly similar to that occurring in the mammal-like reptiles. With this change is also shown a gradual thrust- ing of the elbow towards the body, which in the last form, Pariasuchus, has gone to a remarkable extreme, the glenoid cavity facing very nearly backwards. Many other groups of reptiles and amphibia show slight traces of deriva- tion from this characteristic early form; including, for example, the living Cryptobranchus, where the strange form of the articular face of the glenoid cavity, discovered by Anthony (Jnt. Cong. Med., 1913), receives a ready explanation by such descent, and amongst reptiles the little lizard-like Broomia with only a single coracoidal element. In this paper I have not so far considered any one of the living reptiles or birds. These animals are remarkable in that they never have more than one coracoid bone—-a fact that has been needlessly obscured (especially by Fiirbringer) by the description of a mere process of this bone as pre- eoracoid, and sometimes by the penn acy AC erat of the superficially calcified epicoracoid. There is no doubt that there really is only a single coracoid in lizards, Te aT, CS ae Oe en ee eee ee oe en ee eT Ee a ee Ne ee Core ee ay a TT = ee The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 59 Sphenodon, Chelonia, and crocodiles; no author, so far as I know, having ever claimed to see more than a single ossific centre. The problem now arises, to which of the two coracoidal elements of the early reptile shoulder girdle is this bone homologous. Until a few years ago, all authors believed that the lizard coracoid was the posterior element, and was thus entitled to its name. In 1911, however, Professor S. W. Williston (American Permian Vertebrates) described the shoulder girdle of Seymouria and Varanosaurus, in which the posterior element, the true coracoid, is unossified, and showed that the precoracoid in these forms agrees exactly with the single coracoid of a lizard in shape, relations, and its perforation by a supracoracoid foramen. This view has been accepted by ~ Fic, 27.—Series of Pariasaur scapulo-coracoids, A, Embrithosaurus; B, Bradysaurus baini, after Seeley ; C, Pariaswurus Karpinskyi; D, Pariasuchus, after Broom. Broom (Anat. Anz., vol. xli. pp. 625-631), and seems to me to be certainly correct. Thus the single bone in lizards, etc., is the precoracoid, Williston and Broom both assume that the primitive condition is to have two coracoidal elements, and that Seymouria and Varanoops have lost the posterior. In view of the facts (1) that Seymouria is much the most primitive known reptile, (2) that Varanoops is one of the most primitive of all Pelycosaurs, (3) that the precoracoid steadily decreases in size in mammal-like reptiles whilst the coracoid grows, and (4) that the embolomerous and Lower Permian rachitomous Labyrinthodonts have only a single ossification in each side of the cartilaginous shoulder girdle, as do the Osteolepid fish,—it seems to me more probable that the true coracoid has not yet been acquired in Seymouria, Varanoops, and lizards, than that it has been lost. The old view, that the lizard coracoid is the homologue of the posterior 60 Lieut. D. M. S. Watson element, will be supported, if anyone should think such work worth while, by arguments drawn from the Chelonia, Plesiosaurs, and birds. In opposition to the old view of Owen, that the strong process which runs inwards from the scapula towards the middle line in tortoises is an exaggerated acromium, many authors, of whom Fiirbringer is the most important, have held that this process is the precoracoid. It is extremely probable that but for comparisons with the Anura this view would never have arisen. The scapula including the process is always single, and according to W. K. Parker (“A Monograph on the Structure and De- velopment of the Shoulder Girdle and Sternum,” Ray Soc., 1868, p. 141) ossifies from a single centre. The recent description by Jaekel (Palaeon- tologisches Zeitschrift, vol. i., 1914) of a shoulder girdle of @ Triassic Chelonian in which the“acromium is small and there is no trace of a precoracoid of any kind, removes any doubt that Owen’s view was per- fectly correct. In the case of Plesiosaurs, the huge acromium which in n both phylogeny and ontogeny grows forwards from the anterior edge of the scapula until it meets its fellow of the opposite side and extends back to the coracoid, has been interpreted by Fiirbringer, Hulke, and Lyddeker as precoracoid, because of its resemblance to that of Chelonia. The series of growth stages of the shoulder girdle of Cryptocleidus figured by Andrews (Ann. and Mag. Nat. Hist., ser. 6, vol. xv. pp. 338-346), when considered in connection with the Nothosaurian shoulder girdle, show beyond all ques- tion that the entire bone is scapula. (In a later paper I shall give a mechanical explanation of the form of an Elasmosaurid shoulder girdle, derived from an analysis of the stresses to which it is subjected.) The only bird which really enters into consideration is Struthio, in which Parker and Fiirbringer recognise, in the bar of bone which extends down to the sternum in front of and separated by a vacuity from the coracoid, a precoracoid. No evidence has been brought forward to show that this bar ossifies separately, and its interpretation as precoracoid again rests on prior conceptions. In any case, a hypothesis of skeletal homologies which requires to be bolstered up with evidence from the frogs, turtles, Plesiosaurs, and birds, and these groups ‘alone, has little to recommend it, as the first group is obviously degenerate, the second extremely modified by the unique shell, and the last two extraordinarily specialised for very unusual habitats. . \ oe SEE ee Le Pay ae PLOT oe a eee Le eee ae eae ee ee ey rere Kieren a ; ie The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 61 SomME CONSIDERATIONS REGARDING THE ORIGIN OF THE TETRAPOD FORE-LIMB FROM THAT OF FISHES. The remarkable shoulder girdle and fore-limb which is shown in this paper to provide the starting-point from which all those of mammals, reptiles, and amphibia are derived is at first sight very dissimilar to that which we should expect to be directly derived from a fish fin. In it the restriction of the humerus to a definite track and the great freedom of motion at the elbow are marked differences from the freedom at the joint between the shoulder girdle and proximal radial in fish and the slight freedom between this radial and the rest of the fin. This type of limb, however, leads to a grouping of muscles and to the restriction of the function of each group to one definite purpose, either support or progres- sion, which probably involves a much less elaborate correlative apparatus in the nervous system than the mechanically superior condition in Cyno- gnathus, where every muscle is used for both support and progression. This more satisfactory arrangement is made possible in Cynognathids by the highly developed cerebellum which occurs in that group (Ann. and Mag. Nat. Hist., ser. 8, vol. xii. pp. 217-228). Any fish climbing out of water and progressing on the land depends on its fins, first for progression, which is easily obtained by moving ‘the fin horizontally when its distal end rests on the ground, and second, in a more advanced stage, for raising its body from the ground; this last - function will clearly follow the actual shuffling along. It is evident that the original motion of punting the animal along will be carried out at the original fish fin joint between the scapula and the proximal radial; and as (a priori) the body of the fish is not raised, this motion will be strictly in a plane parallel to the ground. It seems prob- able that to this stage belongs the initiation, of course in a simple form, of the elongated screw-shaped glenoid cavity. When it becomes necessary to raise the body, the problem of correlation of muscle movements is much simplified if motion producing such effect does not take place at the same joint as that driving the animal along. The obvious way of doing it is to depress the front edge of the fin, which will be done by muscles running from the axis of the Euthenopteron-like (short archipterygoid) fins along the pinnately arranged preaxial branches. I have endeavoured to show (Anat. Anz., vol. xliv. pp. 24-27, 1913) that the primitive axis of a Tetra- pod limb runs through the ulna, ulnare, 4th distal carpal, and 4th digit, the radius being the proximal preaxial branch. Depression of the anterior border of the fin of the type postulated above will develop muscles running from the first axial element along 62 Lieut. D. M. S. Watson’ the first branch, from the second segment along the second branch, and so on. I see in the muscles passing from the entepicondyle and ulna to the carpal region in Tetrapods the successors of these muscles, and a remarkable confirmation of this view appears in the fact that in Urodeles no muscles pass from the radius to the palmar surface of the distal part of the limb. I therefore hold that the fore-limb of Eryops is very much what we might expect to find in an animal not long removed from the fish ancestor of Tetrapods. = SUMMARY. In this paper typical shoulder girdles and fore-limbs of most of the families of Cotylosaurs and Anomodonts are described, the muscle insertions being indicated and the first attempt made to determine the shoulder musculature, possible movements, and methods of walking. Examination of a series of forms of the Anomodonts arranged in correct time and morphological order shows a gradual steady change from the conditions found in Pelycosaurs, in common with most other Lower Permian Reptilia and Stegocephalia, where the articular surface of the glenoid cavity _is a screw-shaped strap corresponding to a similar surface of the humerus and restricting the motion-of that bone-to a simple backwards and forwards motion in a plane parallel to the ground, with the middle position when the bone is at right angles to the animal’s length, to those of Cynognathus, . where the humerus moves in a vertical plane at an angle of about 45° with the animal’s length. It is shown that this change is associated with corre- - sponding modifications changing the action of all the muscles, and resulting in a mechanically much superior animal. From a consideration of the same series and of the Osteolepid fish, Laby- rinthodonts, and Cotylosaurs it is suggested that originally the Tetrapods had only a single ossification, the scapula, in each side of the cartilaginous shoulder girdle; that a second, the precoracoid, is then added, and finally a third, the coracoid, which steadily increases in size in comparison with the precoracoid in the mammal-like reptiles, until in Therian mammals it alone remains, the precoracoid and the muscles connected with it disappearing after becoming unnecessary. Evidence is led to show that the “epicoracoid” of Monotremes is the Anomodont and Urodele precoracoid, and that the omosternum of Theria is the old reptilian interclavicle. A theory, derived from the conditions in the Anomodonts, is produced to explain the origin of the mammalian supra- and infraspinatus muscles from the reptilian scapulo-humeralis anterior, and of the teres major from the scapulo-humeralis posterior, The Evolution of the Tetrapod Shoulder Girdle and Fore-limb 63 It is suggested that the lizards split off from the other reptile stocks before they acquired the true coracoid, their coracoidal element and that of Sphenodon, Chelonia, and birds being the precoracoid. Old views of the homologies of the elements of the shoulder girdle are then critically discussed. Finally, it is shown that shoulder girdle and fore-limb of Eryops may have arisen directly from that of fish, under the influence of intelligible forces. In conclusion, I have to express my indebtedness to Professor A. Dendy, ‘through whom I had the opportunity of dissecting Sphenodon, and Pro- fessor J. P. Hill, to whom I owe a specimen of Echidna. The whole work really rests on material which I collected in Texas and South Africa, with the assistance of grants from the Perey Sladen fund. A CASE OF ACCESSORY LUNGS ASSOCIATED WITH HERNIA THROUGH A CONGENITAL DEFECT OF THE DIAPHRAGM. By E. A.. Cockayne, D.M., F.R.C.P., and R, J. Guapstong, M.D.,: F.R.CS., F.R.S.Ed., University of London, King’s College. THE combination of accessory lungs with congenital diaphragmatic hernia is a very rare abnormality. The association of these two congenital - abnormalities, however, although rare, is not necessarily uncorrelated or fortuitous, and it appears to throw considerable light on the manner in which accessory lungs are produced. , The specimen which forms the subject of this paper was obtained from a female infant born prematurely at the eighth month. The child at the time of birth was alive and moved, but did not breathe. It weighed 5 lbs. 7 oz., and measured 18 inches in length. The “ post-mortem” examination revealed a normally developed right lung, with its three lobes of average size and shape. -The left lung was much smaller than normal. It consisted of an upper and a lower lobe, each of which was much but proportionally reduced in size, though otherwise they appeared normal in shape. Below the left lung (fig. 1) were two accessory lungs, which lay one above the other between the lung and the stomach. Both were completely covered by pleura, and were attached by pedicles, below the root of the left lung and between the cesophagus and the thoracic aorta. The cesophagus was - considerably anterior to the attachment of the pedicles. Both accessory lungs were of a pale pink colour, and contrasted strongly with the deep red of the right and left lungs. The upper accessory lung was roughly pear-shaped, its longest diameter being 3-4 cm. The slightly flattened pedicle was about 1°75 cm. in length and 1 em. in width. Its point of attachment just in front of the aorta was more than 2 cm. below the lower edge of the left pulmonary ligament. In the pedicle ran two arteries, arising independently from the thoracie aorta, one arising from the right side and one from its left. Two accompanying veins opened into the vena azygos major. The nerves were apparently derived from the great splanchnic. The lower accessory lung was irregularly lobed and fan-shaped, but when in sitw it was folded upon itself. It measured 4 em. in length, its vertical diameter when opened out was 4°5 em., and in thickness it varied from 0°75 to 12cm. The pedicle was 1°75 em. long and 1 cm. wide. Its A Case of Accessory Lungs 65 attachment to the aorta was some distance below that of the other pedicle and lay only a short distance above the level of the lower end of the cesophagus, though as in the case of the other pedicle at a considerable distance from the cesophagus and in close relation to the anterior aspect of the aorta. The pedicle contained an artery which sprang from the left side of the thoracic aorta, and a vein which entered the hemiazygos vein. A nerve entered it from the great splanchnic. The total length of the right lung was 5:6 em., and of the left 3:1 cm. Left lung. —~ — — ——— Satie — i — — Base of right Upper accessory — lung. lung. ete egy Lower accessory £23 RN Siaskinem lung. > Fie. 1.—Accessory lungs of human fcetus viewed from behind. Thus each of the accessory lungs was but slightly smaller than the right lung, and actually larger than the left lung. Microscopically the upper accessory lung showed very ntmerous _ bronchioles lined by columnar ciliated epithelium; alveoli, lined by rather low cubical epithelium, and a thick interalveolar stroma of connective tissue containing elastic fibres. The lower accessory lung showed a division into lobules séparated by connective tissue, which contained arteries surrounded by a sheath containing incompletely developed striated muscle, and lymphoid tissue which was also incompletely developed. A bronchus was present, in the wall of which lay two plates of hyaline cartilage. In the central part of each lobe were numerous bronchioles, lined by ciliated VOL. LII. (THIRD SER. VOL. XIII.)--OCT. 1917. 5 66 Dr E. A. Cockayne and Dr R. J. Gladstone epithelium, and in one region partly by ciliated columnar and partly by simple cubical epithelium. A section of the left lung showed a much looser alveolar structure, with much less interalveolar tissue; and the number of bronchioles was very much smaller in comparison with the number of alveoli. On making a comparison of the structure of the accessory lungs with that of embryonic or foetal lung tissue, we found that their histological characters did not resemble any stage in the development of the normal lung tissue. At the stage of development when the bronchial system is as extensive as in the accessory lungs the alveoli are not yet differentiated. At a later stage when the alveoli are differentiated, the bronchiolar system is proportionately much less extensive, and the interalveolar stroma much thinner. THE DIAPHRAGMATIC HERNIA. An opening, 2.5 em. in diameter, was present in the left side of the diaphragm. Through this opening there protruded upwards into the left pleural cavity two separate groups of organs (figs. 1 and 2):— (a) Stomach. First part of duodenum, Part of the gastro-hepatic omentum. A tongue-shaped lobule of the liver continuous with a peritoneal fold attached to the under surface of its left lobe, and the left triangular (or lateral) ligament. The spleen, and an accessory spleen. Part of the body and tail of the pancreas. (b) A loop of intestine consisting of :— Ileum. Ceecum and appendix. Ascending and transverse colon. Splenic flexure and commencement of descending colon. Mesentery and mesocolon. The right and left kidneys with the suprarenal glands lay below the diaphragm. ‘The opening in the diaphragm was situated in front of the left suprarenal gland. It lay behind and below the left lobe of the liver, and to the left of the cesophageal opening. It was bounded medially and behind, by the left crus of the diaphragm; while in front and to its left side it was limited by a flat muscular band (fig. 2), which was continuous medially with the left crus below the apex of the heart, while laterally it was continued backward in the left hypochondrium to the posterior A Case of Accessory Lungs 67 abdominal wall, where it was attached behind the left suprarenal body. The serous membrane covering the left crus of the diaphragm, when traced upwards, was continuous with the mediastinal pleura; when followed downwards, it was seen to pass on to the stomach, from the anterior surface of which it could be followed to the right, through the hernial orifice, as the anterior layer of the gastro-hepatic omentum, and to the left and downwards as the anterior layers of the gastro-lienal and gastro-colic ligaments. The membrane lining the under surface of the muscular band forming the anterior boundary of the hernial orifice was continuous medially with the left triangular ligament of the liver and the peritoneum Fic. 2.—Diagram representing the position of the herniated structures, and the opening in the diaphragm. covering the anterior wall of the abdomen. Laterally the costal pleura was continuous with the peritoneum of the left paracolic fossa. A crescentic fold, the phrenico-colic ligament, containing an accessory spleen, marked the boundary between the pleural and peritoneal cavities. LITERATURE. Although a considerable number of cases of accessory lungs have been reported, their occurrence must be considered as a rare event. We have been able to find in the literature of the subject 29 cases, including the one we have described, as occurring in man, and 3 in calves. Of the human, 23 are thoracic and 6 abdominal, or subdiaphragmatic. The 3 found in calves were all abdominal in position. This classification does not include tracheal and cesophageal cysts, or lung tissue contained in teratomata. 68 Dr E. A. Cockayne and Dr R. J. Gladstone Seven of these cases, including our own, were associated with a con- genital defect of the diaphragm on the left side and displacement of abdominal contents into the thoracic cavity. Of the remaining cases in which no defect of the diaphragm was present, 19 were upon the left side and only 3 upon the right. Cases of Accessory Lung occurring on the Right Side. One of these was described by Diirck (7). The accessory lung was situated in the lower part of the right pleural cavity, in which it lay free, with the exception of a short round pedicle which anchored it down to the angle between the posterior chest wall and the diaphragm. The right lung appeared normal. The second, described by Simpson (28), was a case in which the accessory lung, measuring 2 x 1} in. in diameter, was attached by a long pedicle to the mediastinal pleura; it lay in contact with the cesophagus, and at the level of the tenth thoracic vertebra. The right lung was small, but otherwise appeared normal. The accessory lung was situated on the diaphragm, between it and the base of the lung. The pedicle contained a large artery springing from the aorta, a vein which entered the vena azygos major, and nerves derived from both vagi. The histological structure was that of lung tissue showing cyaue degenera- tion. The left lung was normal in all respects. Finally, Herxheimer (11) described a right-sided specimen in which the accessory lung was situated in a totally different position. It lay above the right lung and was connected by a bronchus originating from the trachea. It was found in a male infant, who died aged three weeks. ‘ases of Accessory Lung occurring on the Left Side. The remarkable preponderance of cases of accessory lung occurring — on the left side corresponds with a similar preponderance of left-sided congenital diaphragmatic hernia. Moreover, the fact that among 29 examples of accessory lungs these two conditions are associated in 7, indicates that there is a causal relationship between the two. This relation we shall discuss later, when dealing with the etiology of acces- sory lungs. Of the left-sided cases four specimens possessed pedicles which connected the accessory lung to the left lung, but in none of these was there a connection of the bronchial system of the true lung, with the bronchial tubes of the accessory lung; and the accessory organs were all described as being atelectatic. In one case, however, described by Joest (14), which A Case of Accessory Lungs 69 was found in a calf, the mouth of an open bronchial tube was visible on the cut surface of the pedicle of the accessory lung, and according to the report of the individual who removed the specimen at the post-mortem examination, it was originally connected with the tracheal system. The air cells in the greater part of the accessory lung did not contain air; the part, however, in immediate relation with the severed bronchus con- tained distended air cells. Although we have been unable to find a record of a case in the human subject in which the bronchial tubes of the accessory organ were directly continuous with the bronchial system of the normal lung, cases have been published which indicate a primary connection with the normal lung, viz.: A specimen described by Lewisohn (18), of an acces- sory lung in a human subject which was connected by a pedicle as thick as a lead pencil with the under surface of the lower lobe of the _ left lung. This pedicle contained a vessel opening into the pulmonary vein. There was, however, no connection with the bronchial system, and a second pedicle connected the accessory lung to the diaphragm, and contained an artery springing from the aorta. There was also a vein emptying into the vena azygos. This double connection of an accessory lung with the pulmonary and systemic circulations raises the question as to which of these is the primary and which is the second- ary connection. A case of Végel’s (30) and another of Dubler’s (6) indicate that the primary connection is the pulmonary. Both these cases were abdominal. In Dubler’s case a diverticulum containing a bronchus passed upwards through the aortic opening into the left pleural cavity, where it ended blindly. In this and also in Végel’s specimen there was a deficiency in the bronchial tree of the left lung, one of its main branches being absent. These relations were regarded by Végel as establishing the pulmonary origin of the accessory lungs in these two cases, and also of the accessory lungs having been cut off from the parent stem. His observations appear to be confirmed by a case which was reported by Hammar (10), who in reconstructing the lungs of an 11‘7-mm. human embryo found that a small accessory lung, and an epithelial cyst, were separated from the upper lobe of the left lung. Corresponding with these sequestrated parts, he found a deficiency in the bronchial system of the upper lobe of the left lung. An adhesion between the separated part and the dome of the left pleura was also present behind the apex of the lung. This was probably of secondary origin, and if growth had continued, vessels would have developed within it for the nutrition of the separated part. 70 Dr E. A. Cockayne and Dr R. J. Gladstone ETIOLOGY. Before discussing 1 in detail the various factors which are, or appear to be, concerned in the production of accessory lungs, it will be necessary, in the first place, to give a summary of the various theories which have been put forward by previous writers to account for their origin, and to include in this summary certain additional propositions which have occurred to us, and which we consider it necessary to discuss in conjunction with the former. In the second place, we have considered it essential, in order to give a clear conception of the merits of these different theories, to give a brief réswmé of certain main features in the development of the lungs, and of the later stages in the development of the diaphragm. In this account we have recorded certain observations made by one of us on the relative positions of the lungs, diaphragm, and adjoining viscera in different phases of their development; also the relative positions of the pleural recess for © the reception of the infracardiac lobe of the right lung, and the right pneumato-enteric recess, in rodents and carnivores; and finally, the mode of closure of the pleuro-peritoneal openings in the diaphragm. These observations are based partly upon the study of serial sections of human and other embryos, and partly on a model reconstructed by one of us (R. J. Gladstone)! of the lungs, diaphragm, and adjacent organs of a 16-mm. human embryo, the serial sections of which were kindly lent for the purpose by Professor J. Ernest Frazer. This work has been undertaken, not only with the object of attempting to solve the problem of the origin of accessory lungs, but also in order to explain, if possible, the statistical correlation which exists between the formation of accessory lungs and congenital diaphragmatic hernia; and further, the greater frequency of both these ‘conditions on the left as compared with the right side. Also because it is obvious, even from a casual inspection of cases of congenital diaphragmatic hernia, that the position of the parts forming the hernia, relative to the structures bounding the cavity into which it has protruded, cannot be explained on the assumption that a simple arrest of development has taken place at any particular phase in the development of the embryo. A further object which we have had in view has been to study the mode of closure of the pleuro-peritoneal opening of the diaphragm. ' Details of the latter not directly concerned in this paper will be described in a later communication. A Case of Accessory Lungs 71 Theories on the Origin of Accessory Lungs. 1. Accessory lungs are derived from the blastema of the pulmonary groove, a portion of which at a very early stage of development is separated from the rudiment of the lung and grows independently. 2. They originate as a separate outgrowth from the cesophageal portion of the foregut, and are not derived directly from the pul- monary bud. 3. They originate at a somewhat later stage of development as the result of adhesions taking place between the growing lung bud and _ the ccelomie mesothelium. That this usually but not always takes place on the left side, in the region of the pleuro-peritoneal passage, and at the time when the lung bud is in relation with the stomach, Wolffian ridge, supra- renal body, and that part of the septum transversum which covers the left lobe of the liver. 4, A portion of the growing lung bud becomes caught in an opening or recess, such as that for the “infracardiac,” or “azygos lobe” of the right lung. It becomes secondarily adherent to the parietal pleura, and after- wards the adherent portion becomes separated and forms an accessory lung. 5. They arise as atavistic structures which are homologous with the pulmonary air-sacs of birds. 6. They originate as teratomatous inclusions. The material belonging to-one of us, which we have made use of for this study, consists of serial sections of the following mammalian and avian embryos and foetuses :— Human. Length inmm. Estimated age. Length in mm. Estimated age. ITS. 9 4 weeks | TS. 25 74 weeks T.S. 10 4, ,, T.S. 32 9 im tog 11 5 a T.S. 35 94, 2T:S. 16 6 5 TS. 41 10 5 TS. 22 7 a T.S. 45 103 __s=z, Carnivores. T.S. Cat, 14 mm. | T.S. Dog, 45 mm. 1 T.S., transverse section ; C.S., coronal section ; L.S., longitudinal section. 2 Specimen kindly lent by Professor J. E. Frazer, from which we have reconstructed - the model represented in figs. 11 and 14, showing the diaphragm and the relations of the pleuro-peritoneal openings at this stage. 72 Dr E. A. Cockayne and Dr R. J. Gladstone Rodents. C.S. Mouse, 4°5 mm. (in situ). | CS. Guinea-pig, 6 mm. (im situ). oh: eee Dries By Mra OFS af 15: te 3 a ee Gai Bone FF ie 1b, LS. Rabbit, 9, TS. eS: tee LS... 2k Ae oe | : Chick. T.S. 1st day (2 series). : T.S. 3 days. L.S. 1st day. T.S. 4th day. | L.S. 2nd day. | T.S. 4 days. . T.S. 38 hrs. . T.S. 44 days. T.S. 48 hrs. T.S. 54 days. T.S. 3rd day. Development of the Lungs. The first indication of the development of the respiratory system appears in the human embryo when it has attained a length of 2°5 mm, (Embr., Rob. Meyer, Normentafel No. 7). A longitudinal groove, the laryngo-tracheal sulcus, is formed in the epithelium lining the ventral wall of the pharynx; this terminates caudally in an unpaired swelling, the rudiment of the lungs. The suleus extends from the last branchial pouches to the region of the hepatic diverticulum. The unpaired rudiment of the lungs soon becomes bilobed, and the pulmonary and tracheal parts then become separated from the gastric and cesophageal portions of the foregut by the development of two grooves, which are situated one on each side of the foregut. These project inwards and meet across the lumen of the tube, subdividing it into two parts, an anterior or tracheal and a posterior or cesophageal. Two epithelial tubes are thus formed which subsequently become separated by the interposition of surrounding mesoderm. ‘The mesoderm which immediately surrounds the epithelial tubes and takes part in the formation of the walls of the trachea and cesophagus, afterwards becomes further differentiated by the appearance of muscular tissue in both tubes and of hyaline cartilage in the larynx, trachea, and bronchi. Leaving out of consideration the further develop- ment of the larynx, it is found that the two lobes of the bifurcated caudal extremity of the lung rudiment grow in a dorsal direction on each side of the lower end of the cesophagus into the isthmus of the ccelom, and come into relation with the septum transversum, In this situation the lung A Case of Accessory Lungs 73 buds pass on each side beneath a verious arch, which is formed by the union of the posterior cardinal vein with the duct of Cuvier. The concavity of this arch is occupied by a crescentic fold of the ccelomic wall. The dorsal horn of the crescent joins the posterior wall of the isthmus of the ccelom (dorsal parietal recess of His) medial to the upper end of the Wolffian ridge, and lies in front of the posterior cardinal vein. The ventral horn is connected with the pericardium above, but below the level of the heart passes downwards on to that part of the septum transversum which covers the dorsal aspect of the liver. On the right side this bal - NS I ry Ge a cs Ws Liver Septum transversum. - --} ~+~~ WXZA--AN 3 ea 1 ules fig NN ae ~ Pericardial cavity. Peritoneal cavity. -~~~-\- = Cctee if ->+-4+7-- Inf. vena cava. Pleural cavity. . ._—-— $s EY ay Hieurn- partentdial Dorsal and ventral pillars of a the pleuro-peritoneal arch. - Se at rea Fig. 3.—Transverse section of human embryo (9-mm.) at level of fifth cervical segment, showing the relation of the lung buds to the pleuro-peritoneal arch. attachment is postero-lateral to the inferior vena cava (fig. 3). This crescentic fold, which in the earliest stage of development is termed by Mall (66) the “pulmonary ridge,” becomes differentiated later into an upper and ventral part, the “pleuro-pericardial membrane,” and a lower and dorsal part, the “pleuro-peritoneal membrane” (Uskow (80), Brachet (53) ). That portion of the duct of Cuvier which becomes enclosed within the pericardium differs somewhat in its relations on the two sides, but in both it comes to lie ventral to the root of the lung, and it maintains this position in the adult. Thus we find the superior vena cava on the right, and the “ligamentum cave sinistre,” with the vestigial fold of ! On account of this relation to the Wolffian ridge, this structure has been named by Keith (65) the Wolffian fold. 74 Dr E. A. Cockayne and Dr R. J. Gladstone Marshall, on the left side, in front of the root of the lung. Further, the terminations of the posterior cardinal veins, represented in the adult by the vena azygos and intercostal part of the superior intercostal vein, lie behind the root of the lung. The development of the pulmonary arteries and veins is well advanced in the 10-mm. stage (His), and their relations to the main bronchi definitely fixed. The artery of the left side passes down behind the left bronchus, and on the right side in front of the eparterial bronchus but behind the hyparterial bronchus. The single pulmonary vein lies below the bifurcation of the trachea and in front of the cesophagus. It receives two tributaries on each side: the upper pair pass in ‘front of the main or hyparterial bronchus, which thus intervenes between the pulmonary vein in front and the artery behind. The root of the lung is thus at this early stage fixed in its position, relative to the heart and main blood-vessels. The position, however, of the roots of the lungs, and of the heart, septum transversum, and liver, relative to the vertebral column, is by no means fixed. The change in the relative position of these parts has been admirably shown by Uskow (80) to be due to the forward growth of the vertebral column rather than a descent of the heart and diaphragm from the neck into the thorax. This view of the change which takes place in the relative position of the parts has also been put forward by Professor F. Wood Jones (63) in an article entitled “The Functional History of the Ccelom and the Diaphragm.” The forward growth of the vertebral column is obviously associated with the early and rapid growth of the brain and spinal cord. The latter elongates both in the cephalic and caudal directions from a central region, which we may regard as a virtually fixed point, situated somewhere in the region of the umbilicus. The elongation first takes place in the are of a circle, so that the longitudinal axis of the embryo, which in the 2°5-mm. stage is almost straight, becomes curved to such an extent when the embryo has attained a length of 7 mm., that the tip of the tail is almost in contact with the fore-brain. In an embryo of 2°71 mm. described by Mall (66), the pericardial ccelom lies anterior to the otic vesicle, and the eighth cervical myotome is behind the posterior boundary of the stalk of the umbilical vesicle. As development pro- ceeds, the brain and spinal cord, with the nerves issuing from. them, grow forward round the dorsal aspect of the heart and pericardium, carrying with them the vertebral column. Thus the bifurcation of the trachea is found opposite the fourth spinal nerve in an embryo of 6°38 mm., and to have the position shown in the following table relative to the bodies of the vertebra, at the stages indicated under the heading, “Length of Embryo.” a A Case of Accessory Lungs | 75 Length of Embyro. Centrum of Vertebra, 9 mm. 6th cervical. 10 =; 7th ;} : 4 A 1st thoracic : os Ree 2nd _sS, I 181; 3rd ‘ RE 4th To avoid confusion, however, we shall notwithstanding this divergence in our conception of the direction in which the growth-changes take place, a Ventral mesentery. Liver. — LV.C. Superior recess of lesser omental sac. Lateral mesentery. Adhesion of __ left lung. (Esophagus. Fic. 4.—Transverse section of a 9-mm. human embryo showing adhesion of the left lung to the diaphragm. adopt the usual description of a descent of the lungs and diaphragm relative to the vertebral column. In embryos from about 6-10 mm. length the lungs project below the pleuro-peritoneal membrane into the peritoneal cavity, where they come into relation with abdominal organs such as the liver, stomach, and the upper end of the Wolffian ridge. At this stage the lungs may become adherent to these organs, as is shown in fig. 4, of a 9-mm. human embryo in which the left lung is adherent to the septum transversum covering the liver. The area of contact of the lungs with the liver in the 7-mm. and 1l-mm. stages is well shown in two models reconstructed by Professor Peter Thompson (79). ‘The models also show the relation of the lung buds 76 Dr E. A. Cockayne and Dr R. J. Gladstone to the omentum minas (ventral mesentery) and the plica vena cava inferior. After this period the lungs come to lie entirely above the plane of the pieuro-peritoneal membrane and of the pleuro-peritoneal orifice. This appears to be due to the more rapid descent of the dorsal attachment of the diaphragm, relative to the vertebral column, as compared with the rate at which the lungs grow downward. After this period, therefore, should an accessory lung be formed by adhesion, and subsequent separation of the adherent portion, it would be intrathoracic rather than abdominal or subdiaphragmatic in position. In the 10-mm. stage the main subdivisions of the bronchial system are well established, and also the subdivision of the lungs into their constituent lobes. The right lung, moreover, shows a distinct subdivision represent- ing the “azygos” or “infracardiac” lobe of rodents and carnivores. This causes an elevation which, as has been shown by Blisnianskaja, is distinctly visible on the surface in a human embryo of 138 mm. A model of the lungs reconstructed by one of us from a embryo 16 mm. long, however, shows only a slight rounded elevation in the position of the azygos lobe; and in another model made by Blisnianskaja from a 17-5-mm. embryo, the surface of the lung has become quite smooth, and only shows the main fissures. Normally, therefore, there is at no stage of development an isolated lobule of the right lung projecting as a tongue-shaped process into a pleural recess (subpericardial sinus) behind the inferior vena cava, such as is present in the mouse or guinea-pig (figs. 5, 6, 7, 8). Moreover, when an infracardiac lobe (fig. 9) is found as an abnormality in the adult human subject, the lobe, although it may be distinctly marked off by deep fissures, does not project beyond the mediastinal surface of the lung. In the 15-mm. stage the lungs have descended for a considerable distance. The upper extremity of the pleural cavity now corresponds in level to the neck of the first rib, and its lower to the neck of the tenth or eleventh rib. The lower pointed end of the lung reaches the ninth rib or interspace, The lungs at this stage are small compared with the size of the heart and liver, and are confined to the posterior part of the thoracic cavity. They present a posterior convex surface, and a slightly concave anterior surface; the latter is subdivided into two parts, viz.: an upper, the future mediastinal surface, which is in relation with the pericardium and phrenic nerve; and a lower, which is in contact with the septum transversum, covering the dorsal surface of the liver, and with the pleuro- peritoneal membrane. ‘This part will become the base of the adult lung. The apices of the lungs have now become distinct from the roots, which in earlier stages formed the highest points of the lungs. The apices at this period are on a level with the second ribs and the bifur- A Case of Accessory Lungs 77 (Esophagus, f_. Ventral mesentery. Inferior vena : ; ; Besa ; cava, " 7 7 pin ‘ Infracardiac lobe. Fic. 5.—Transverse section of a mouse embryo (13 days), showing the relations of the infracardiac lobe of the right lung to the ventral mesentery, cesophagus, and the inferior vena cava. Mesenchyme. — (Esophagus. Pleuro-peritoneal -~ Infracardiac lobe. membrane. ~ Inf. vena cava. Fic. 6.—Transverse section through a cat embryo, showing the large amount of loose mesenchyme which lies between the parietal pleura and the thoracic wall. Infracardiac __ _ lobe. Communication between right = pleural cavity and the recess for the infracardiac lobe. — Inferior vena cava. (sophagus. — Fic, 7,—Coronal section of guinea-pig embryo, showing the relations of the infracardiac lobe of the right lung to the diaphragm, pericardium, and inferior vena cava. Pericardium, ————-~ contre PAS « im : - Ea ar eapeces ee Dorsal part of diaphragm. Fic, 8.—Median sagittal section of rabbit embryo (15 days), showing the relation of the infracardiac lobe to the pericardium and diaphragm. A Case of Accessory Lungs 79 cation of the trachea. Whether in the living embryo they extend to the full height of the cervical dome of the pleura is somewhat doubtful. The summit of the pleura corresponds to the neck of the first rib, and the difference in level of the apex of the lung and the highest point of the pleural dome, as seen in serial sections, is probably largely due to contraction. If the apices of the lungs do completely occupy the pleural dome in the living embryo, they will have already reached their per- manent position relative to the thoracic wall, and it might be thought that any further growth of the lung above the level of its root would be in a downward direction from the apex. It must be remembered, however, that there is a growth in the cranial direction of the spinal cord and vertebral column relative to the root of the lung. This is Fic. 9.—Human lung, showing ‘‘ azygos” lobe. accompanied by a shifting of the vertebral ends of the ribs and of the intercostal vessels and nerves in the same direction. There is, further, a marked elongation of the cesophagus, of the trachea, and of the pneumogastrie and phrenic nerves. The heart and pericardium, with the main blood-vessels and roots of the lungs, remain relatively fixed in position, and thus appear to sink down from the neck into the thoracic cavity. There is, however, a true growth of the lung in a cranial direc- tion above the level of its root; if this does not occur, we must assume that there is a “migration” downward of the hilum, on the mediastinal surface of the lung. Moreover, as evidence in favour of the upward growth, we may mention the occasional separation of an accessory lobe of Wrisberg, by the vena azygos major, which vessel lies at the bottom of the fissure dividing the accessory lobe from the main upper lobe of the lung. The vena azygos has obviously retained its original position relative to the root of the lung, whereas a portion of the lung has grown upward to its medial side. 80 Dr E. A. Cockayne and Dr R. J. Gladstone In ‘the later stages in the development of the lungs, their anterior margins grow forward in the interval between the pericardium and the thoracic wall, and appear to separate these from one another; finally, the anterior borders of the lungs meet in the middle line behind the sternum. The upper pericardial part of the anterior surface of the embryonic lung thus becomes directed medially. The lower border of the lung grows downward, this growth being accompanied by a separation of the peri- pheral part of the diaphragm from the deeper layer of the thoracic wall. Left pleural cavity reduced in size owing to the movement of the cesophagus to the left. Pleuro-peritoneal membrane. Right pneumato- Recess for infra- i enteric recess. cardiac lobe of right lung. Inferior vena cava. = Fic. 10.—Transverse section of a mouse embryo (18 days), showing the relation of the right pneumato-enteric recess (infracardiac bursa) to the recess for the infracardiac lobe of the right lung. The lower diaphragmatic part of the original anterior surface of the lung now becomes the base. The extension downward of the phrenico- costal sinus of the pleura which accompanies the downward growth of the lower border of the lung corresponds in the 16-mm. stage to a plane which lies a short distance below the level of the pleuro-peritoneal opening (fig. 11). After the opening has become closed the sinus extends further down- ward, and ultimately reaches a plane below the level of the neck of the last rib, which is considerably below the site formerly occupied by the foramen. The original posterior surface of the embryonic lung, which A Case of Accessory Lungs = Se was situated in a plane anterior to the bodies of the vertebra, grows backwards on each side of the vertebral column. This change in posi- tion, which follows a similar change in the position of the ribs, gives rise to the rounded posterior border of the definitive lung. It is pre- ceded by a remarkable change in the mesodermal tissue which lies between the parietal pleura and the ribs (figs. 6 and 15). This becomes extremely loose in texture, so that it resembles the mucoid tissue of the umbilical cord. At a later stage this tissue disappears and the parietal pleura then comes into contact with the ribs. With regard to the histological features which are characteristic of Dorsal mesocardium. ~- Pericardial cavity. }_ _ Pleuro-peritoneal membrane. cit -- Dorsal mesentery. a R. pleuro-perito- es neal opening. Fic. 11.—Sketch of unfinished model of the diaphragm of a 16-mm. human embryo seen from behind, Note the position of the pleuro-peritoneal openings and that the left pleural cavity is narrower and more irregular in shape than the right. developing lungs, the most important from the standpoint of the formation of accessory lungs are, firstly, the large proportion which the connective tissue bears to the subdivisions of the bronchi in the early stages of development as compared with the condition at birth and in the adult, and secondly, the periods at which specialised histological elements such as muscle or glands make their appearance in normal development. According to Kolliker and Merkel, lobulation is present in a foetus of 20 weeks. In the fourth month cartilaginous plates and rudimentary glands are developed in the primary bronchi, and muscular tissue is also distinguishable in the larger bronchi, which are lined by ciliated epithelium. At the sixth month the terminal bronchi are expanded into alveoli, lined- by a low, flattened epithelium, and the interbronchial and interalveolar connective VOL. LIl. (THIRD SER. VOL. XIII.)—OCT. 1917. 6 82 Dr E. A. Cockayne and Dr R. J. Gladstone tissue is much reduced. Elastic fibres, according to Linser, appear in the vessels in the third month, in the fourth month in the larger bronchi, and in the seventh month in the connective tissue—stroma. As already stated, the structure of the accessory lungs in the case we have described does not correspond to any particular phase in the normal development of the lung, and it appears that a small portion of pulmonic tissue which has been separated from the main part, and has acquired new connections, has an extraordinary potentiality of growth. It can not only increase in size until it considerably exceeds the size of the lung from which it has originated, but it can also produce formed elements such as muscle, cartilage, and elastic tissue, which were not present at the time when it was separated from the parent tissue. The striated muscular tissue which was present in the lower af the two accessory lungs, in the specimen we have described, may have been derived from that part of the foregut from which the upper part of the cesophagus is developed, or from the pre-muscle tissue of the septum transversum. As, however, there was no direct connection with the cesophagus, and no other histological character which in any way resembled the microscopical structure of the cesophagus, we are inclined to think that the striped muscular tissue of the accessory lung was derived from the pre-muscle tissue of the septum transversum, to which the growing lung bud had become adherent. The Later Stages in the Development of the Diaphragm. The next changes in development which we shall consider concern the diaphragm and the production of the hernia. In a 16-mm. embryo kindly lent by Professor J. E. Frazer the pleuro-peritoneal openings are still un- closed (figs. 12 and 13). They are, however, reduced to small, slit-like aper- tures with rounded margins. The opening on each side is situated lateral to the upper end of the suprarenal body, and above and medial to the superior extremity of the Wolffian ridge (fig. 14). On the right side it is in relation below with the liver. On the left it lies above an interval between the suprarenal body and the left lobe of the liver. This interval is occupied by the stomach (fig. 12). The left pleuro-peritoneal opening is therefore unprotected on its under aspect by any solid organ, whereas the right is in relation with the liver. At this stage the intestinal loop still occupies the umbilical ccelom, and it is not normally retracted until a considerably later period, when the foetus has attained a length of about 40 mm., whereas the pleuro-peritoneal openings close in embryos 18-19 mm. long. The final closure of the opening appears to be due to proliferation of R. pleural cavity. . Suprarenal glands. Genital and Wolffian ridges. R. pleuro-peritoneal _ opening. Small omentum and Caudate lobe of liver. Fic. 12.—Transverse section through 16-mm. human embryo, showing the position of the R. pleuro-peritoneal opening. Dorsal L. pleuro-perito- : mesentery. neal opening. Plenro-perito- neal membrane. Cardiac orifice Inf. v. cava, of stomach. Fig. 18.—Transverse section of 16-mm, human embryo, showing the relation of the L. pleuro-peritoneal opening. 84 Dr E. A. Cockayne and Dr R. J. Gladstone the epithelium covering its margins. The epithelium here is thickened, | and stains more deeply than elsewhere. Further, the cell elements are more rounded and project irregularly from the surface of the membrane. The exact relations of the muscle fibres of the diaphragm in the early stages of their development are difficult to determine, but it appears that the pre-muscle mass in which the phrenic nerve ends, in a 9-mm. human embryo, and which is situated just below the corresponding infra-hyoid. mass, in the neck (Lewis), later becomes separated from the infra-hyoid rudiment. The latter remains in the cervical region, where it may be seen L. umbilical R. umbilical : vein. vein. Area of attach- ment of liver. Opening for inf. we ger Pleuro-perito- - neal membrane. (@sophagus. r i) L. pleuro-peri- SR areL ip -— toneal opening. R. pleuro-peri- SRASSS GZ Se SY SS ii / Genital ridge. toneal opéning. SN Pl, é / ic, Ae Se Is Wolffian ridge. R. suprarenal By | : Rg gland. ‘ Cares 0 i Si. ag tah Dorsal mesen- ~~ tery. Facne Penn) Fig. 14.—Sketch of unfinished model of diaphragm reconstructed from a 16-mm. human embryo seen from below. passing forward in front of the upper part of the heart and pericardium (fig. 15), whereas the phrenic nerve courses downward in the body wall lateral to the anterior and common cardinal veins, and behind the sub- clavian vein into the pleuro-pericardial membrane. In embryos of-about 15-mm. length it ends in a thickened mass of tissue (fig. 15) at the junction of the pleuro-pericardial membrane with the septum transversum, and here it divides into two principal branches, an anterior and posterior, which apparently correspond to the definitive branches in the adult, and to the costal and vertebral segments of the muscle. In embryos of from 20 mm,, onward, when muscular tissue is distinctly recognisable, the muscle fibres of the costal segment appear to be continuous laterally with the m. transversus abdominis, whereas the fibres of the vertebral segment A Case of Accessory Lungs 85 * course downward in the dorsal mesogastrium to form the crura of the diaphragm. The triangular gap between the costal and vertebral attachments of the muscle is completely closed by the pleuro-peritoneal membrane. The .pleuro-peritoneal opening, as previously stated, appears to be closed by PMOMIA WEG: Whe as Sit ot ie Se Je muscle mass. 5 DD Ce ———— Nasal cavity. Phrenic nerve. ____ Extrapleural «mesenchyme. A OD nN a5) , a Fic. 15.—Sagittal section of rabbit embryo (16 days) passing through the pleuro-peritoneal Negi and showing the termination of the phrenic nerve in the pre-muscle tissue of the diaphragm. proliferation of the coelomic epithelium bounding the aperture. It seems probable, therefore, that the muscle fibres grow downward in the pleuro- peritoneal membrane separating pleura from peritoneum, and also into the dorsal mesentery; and further, that the closure of the pleuro-peritoneal opening is independent of the triangular gap, the hiatus diaphragmaticus, which is left between the adjacent edges of the muscular sheets. Having considered the more important developmental conditions which 86 Dr E. A. Cockayne:and Dr R. J. Gladstone must influence the production of accessory lungs, and of congenital dia- phragmatic hernia, we shall now briefly discuss the different theories, already mentioned, which have been advanced to account for their occurrence. 1. One of the most attractive of these hypotheses is, that accessory lungs are derived from the pulmonary blastema of the original laryngo- tracheal groove, a portion of which at a very early stage of development has become separated from the parent tissue and has taken on an in- dependent growth. We think it quite possible that this mode of origin may account for the formation of accessory lungs in certain atypical situations, but it is difficult on this hypothesis to offer any explanation for the preponderance of cases of accessory lung on the left side as compared with the right, or to account for the high proportional frequency with which accessory lungs are associated with a defective development of the diaphragm on the left side and congenital diaphragmatic hernia. 2. The same objection applies to the hypothesis of Eppinger and Schauenstein, who regard them as independent pulmonary organs, arising from an additional outgrowth from the foregut, in the region of the “dorsal parietal recess” and “septum transversum.”’ Bert and Fischer (8) support these authors by pointing out that cysts occur in the wall of the cesophagus, some lined by both ciliated and squamous epithelium, some by ciliated epithelium alone, and others lined by ciliated epithelium which have plates of cartilage in their walls. They regard these as transitional to true accessory lungs. It is, however, more probable that such cysts arise from the pulmonary groove, from which they become separated owing to the rapid elongation of the foregut which takes place in the formation of the cesophagus. Their position in connection with the lower end of the gullet and in the vicinity of the cardiac orifice of the stomach is thus readily explained. The occurrence of cysts lined by ciliated epithelium in relation with the trachea, and sometimes in communication with its lumen, is an additional point in favour of this view of their origin. (Esophageal cysts also are quite as often situated to the right or in the middle line as to the left of the cesophagus. 3. The theory that accessory lungs are formed by an adhesion taking place between the growing lung bud and the eccelomic epithelium, and that a portion of the adherent lung tissue subsequently becomes separated -off from the main part and takes on an independent growth has much in its favour. It appears to afford an explanation of the greater frequency of accessory lungs on the left side as compared with the right, and also of the association of these with a defect in the development of the left side A Case of Accessory Lungs 87 of the diaphragm, and congenital hernia of abdominal organs through this defect. As already mentioned, we have ourselves observed such an adhesion present in two human embryos, one 9 mm. (fig. 4) and the other 10 mm. longest diameter (C.R.M.). The adhesion was in the same situation in both, namely on the left side, between the lung bud and the septum transversum, covering the dorsal aspect of the liver. Both specimens were macerated and probably pathological. Adhesions of the lung to the parietal pleura may, however, be observed in perfectly healthy embryos, eg. we have noticed an extensive adhesion of the lower and medial portion of the left lung to the parietal pleura which lines the niche between the cesophagus and its dorsal mesentery in a 16-day rabbit embryo. The tissues in this specimen appear to be quite healthy and are well preserved. Whether such adherence of the lung to the adjoining tissues in this situation is primary and due to delay in the separation of the lung bud from the mesodermal tissue of the mesentery, or is a secondary condition, does not affect the argument that accessory lungs may arise from such an adherent portion of the lung becoming separated off from the main organ so as to form an independent structure. Secondary adhesions, moreover, normally take place in the same situation as in the human embryos mentioned above, but on both sides in the chick embryo, and are permanent. Their. occasional and probably temporary occurrence in the human embryo may therefore possibly be atavistic rather than pathological. In the chick embryo these adhesions are associated with the development of the pneumato- enteric recesses. According to Lillie (93), “the accessory mesentery grows ” forward on each side in the interval between the “lateral mesocardium ” and the ventral mesentery and adheres to the septum transversum. Later the entodermal lung sacs grow into the accessory mesenteries, and thus lie lateral to the pneumato-enteric recesses. On the left side the accessory mesentery ceases opposite the lower extremity of the lung, but on the right side it is continued back by the mesentery of the inferior vena cava as far as the middle of the stomach, where it is attached to the septum transversum at the superior lateral angle of the liver. The very much greater frequency of accessory lungs on the left, as compared with the right side, points to there being some mechanical condition which is present on the left side but not upon the right, and we would suggest that there is a tendency for the lower end of the developing lung bud to adhere to the structures in relation with the pleuro-peritoneal passage at the time when the septum transversum and liver are descending from the upper to the lower part of the thoracic portion of the trunk. At this stage a rapid change in the relative position of the parts takes place on the left side. The lower end of the left lung, which at first is in 88 Dr E. A. Cockayne and Dr R. J. Gladstone close relation with that part of the foregut which will become the fundus of the stomach, and with the left lobe of the liver, becomes separated from these by the pleuro-peritoneal membrane, which is developed in continuity with the lower part of the pulmonary ridge. The lung thus comes to lie above and dorsal to the membrane and the plane of the pleuro-peritoneal orifice. We believe that it is at this critical phase of development, involv- ing the above-mentioned change in the relative positions of the adjacent organs, that the lung is apt to become adherent to one or other of the neighbouring structures. A factor which we would suggest may be responsible for the adhesion of the lung bud to the ccelomic epithelium taking place more frequently on the left side than on the right is that the normal movement of the stomach to the left, which commences in embryos about 4°5 cm. in length, carries the lower end of the cesophagus, with its dorsal and ventral mesenteries, and the lung buds also to the left (see figs. 4, 6, 10), thus reducing the size of the left pleural passage and widening that of the right. The left pleural space is thus distinctly narrower than the right, and in serial sections of embryos, in which as a result of “fixation” the lung buds are seen to have retracted away from the thoracic wall, the lung appears to more completely fill the pleural cavity on the left side than on the right. At a later stage in development the width of the left pleural cavity is still further reduced by the projection of the heart and pericardium more to the left than the right side. Another factor which would favour adhesion on the left side is that the mesodermal lobes formed round the second dorsal bronchus and the terminal part of the stem bronchus of the left lung, lie in especially close relation to the lower end of the cesophagus, and occupy a niche formed between its dorsal mesentery and the thoracic wall. This is the position in which accessory lungs are commonly found, as for instance the case which we have described. This relation of the above-mentioned lobes to the ceso- phagus is well shown in a reconstruction by Hammar (90) of an 11'8- “mm. human embryo. If such an adhesion should fail to break down, a small portion of the lung tissue might become separated and remain in relation with the liver, stomach, diaphragin, or dorsal mesogastrium. The separated part, cut off from the bronchial system and pulmonary circulation, would draw its blood- supply from the vessels of the part to which it has become adherent; thus we find accessory lungs with arteries derived from the left phrenic artery, coeliac axis, or thoracic aorta, and the veins passing into the portal system, hemiazygos,and azygos veins, ‘The position of the pedicles, and the vessels contained therein, varies according to the site of the original attachment, A Case of Accessory Lungs . 89 and may be subdiaphragmatic (intra-abdominal) or intrathoracic.. The specimen described by Lewisohn (18), and previously mentioned, in which an accessory lung was connected with the left lung by one pedicle, and with the thoracic wall by a second pedicle, appears to us to afford very strong evidence in favour of the origin of accessory lungs by adhesion to the ccelomic epithelium lining the pleural passage or covering an adjacent organ such as the liver or stomach; and the theory is further supported by the cases described by Dubler (6) and Vogel (30), and that of Hammar, in which an accessory lung adherent to the cervical pleura was associated with a defect, corresponding in position to it,in the bronchial system of the adjacent portion of the lung. 4. A modification of the last-mentioned theory is that which attributes the origin of accessory lungs to a portion of the lung becoming caught in an opening or recess such as that for the “infracardiac” or “azygos” lobe of the right lung. The part thus caught, being subsequently separated from the main portion, contracts adhesions to the wall of the sac. The position of the pedicle, however, in the cases which have been recorded is usually posterior to the cesophagus, whereas the azygos lobe passes to the left, behind the inferior vena cava but anterior to the cesophagus (figs. 5 and 6). Further, although in the human subject an “azygos lobe” is represented in the bronchial system of the embryo, and may occur as an abnormality in the form of a separate lobe in the adult, there appears to be no separate recess in the pleural cavity for its reception. The upper end of the right pneumato-enteric recess, which is described by Broman in human and other mammalian embryos as being cut off to form the infra- cardiac bursa, is an independent structure quite distinct from the recess (sinus subpericardiacus) which contains the infracardiac lobe of the right lung (fig. 10). We consider, therefore, that though the azygos lobe may be cut off in the way suggested in those animals in which it normally exists, this view of the origin of accessory lungs cannot be entertained as far as the human subject is concerned. We think it quite possible, however, that an accessory lobe such as Wrisberg’s could be completely separated off from the parent stem and thus form an accessory lung. 5. The theory that accessory lungs are homologous with the lower pulmonary air-sacs of birds is one which deserves careful consideration, for it is into the “accessory mesenteries” previously mentioned that the entodermal air-sacs of the avian embryo extend, and if the adhesions of the lungs described above as occurring in the human embryo are regarded as atavistic in nature, the accessory lungs arising as a result of these 90 Dr E. A. Cockayne and Dr R. J. Gladstone adhesions might be regarded as homologous with the air-sacs. We are inclined, however, to regard the condition in the human subject as patho- logical rather than atavistic, more especially on account of the frequent association of accessory lungs with other defects, which may be attributed to faulty nutrition of the embryo. 6. The last hypothesis which we shall consider is, that accessory lungs originate as teratomatous inclusions. Although lung tissue has been found along with other tissues or organs in the teratomata, or mixed tumours, these nearly always occur in the form of a definite tumour, surrounded by a cutaneous or fibrous capsule, and though they have been ' found in the mediastinum of the thorax (inclusio mediastinalis), or within the abdomen in the region of the umbilicus (inclusio abdominalis), they are not found free in the pleural or peritoneal cavities. Thus, whether they are regarded as parasitic in nature, or as aberrations in the develop- ment of a single foetus, they differ essentially in character and situation from the typical accessory lungs, and we consider that they belong to a totally different category. CONGENITAL DIAPHRAGMATIC HERNIA. Passing now to the anatomy and mode of development of the congenital diaphragmatic hernia, so much has already been written on this subject that it will only be necessary for us to draw attention to certain points which so far as we know have not been previously considered. It is commonly stated that a congenital diaphragmatic hernia is not a true hernia, but rather the persistence of an early developmental phase, in which the pleural and peritoneal cavities are continuous with one another, and the stomach is intrathoracic in position. Now, in those cases in which the stomach and colon are both present in the hernia, although it is not usually precisely stated, there must be two loops forming the hernia: (1) consisting of the stomach and sometimes the first part of the duodenum, with the body of the pancreas and spleen, and (2) the terminal part of the small intestine, caecum, appendix, and the colon as far as the splenic flexure. The descending and terminal parts of the duodenum and the head of the pancreas are intra-abdominal and approximately normal in position. Further, the cesophageal opening of ‘the diaphragm, which is situated in that part of the diaphragm which is developed from the mesentery of the foregut, is complete and separate from the hernial aperture, It is, moreover, at its normal level. The cesophagus, therefore, does not run a direct course to the highest part of the stomach, as it would do if the condition was a mére persistence of the early embryonic A Case of Accessory Lungs 91 condition. The cesophagus passes down usually to the level of the ninth thoracic vertebra, it then, after having pierced the diaphragm, becomes continuous with the proximal limb of the “stomach loop,’ running a recurrent course upwards from the cardiac orifice to the fundus. It is “ Peritoneal Peritoneal itoneal __ _ cavity. Liver. Liver. Caudate process. Sup. spend of omental bursa. - Portal vein. ae meee Inf. vena cava. Liver, —|- —- senerepettie. {ya neal membrane. cavity. Pleuro-perito- Pleural cavity. neal membrane, ~ Pleural cavity. Fic. 16.—Transverse section of a human embryo (22 mm.), showing the relation of the pleuro-peritoneal membranes to the suprarenal bodies and liver. The outlines of the suprarenal glands are indicated by interrupted lines, obvious, therefore, that the “stomach loop,” with the body and tail of the pancreas and spleen, must have been displaced upwards from the abdominal cavity into the thoracic. Secondly, the “intestino-colie loop” is developed from a part of the primitive canal, which in the early developmental stages lies outside the body of the embryo altogether in the “umbilical ccelom,” and retains this position until long after the time when the pleuro-peritoneal Openings in the diaphragm are normally closed. They therefore must 92 Dr E. A. Cockayne and Dr R. J. Gladstone - also be displaced upwards through the hernial aperture, and form a true hernia into the thoracic cavity. : With regard to the nature of the hernial aperture and the period at which the hernia takes place, allusion has already been made to the unprotected area in the region of the left pleuro-peritoneal opening between the left suprarenal body and the left lobe of the liver. This area, after the opening has become closed (18—20-mm. stage) (fig. 16), is covered merely. by the pleuro-peritoneal membrane. Later, muscular fibres are developed in the membrane which separate pleura from peritoneum. If the normal closure of the opening should be delayed, the muscle fibres if traced downwards will be found to diverge on each side of the aperture, so that the costal fibres will bound it laterally, the vertebral or crural fibres medially. In most cases of congenital diaphragmatic hernia it is probable, however, that there is not only a delay in the closure of the opening, but a defect in the development of the diaphragm. In extreme cases this appears to be primary, but in others it might be explained by the pressure of the herniated viscera. It is possible that the cause of the delay in the closure of the pleuro-peritoneal opening in cases in which accessory lungs are associated with a hernia may be the adhesion of the lung to the liver, stomach, or ecelomie wall. It is, however, obvious that in the vast majority of cases of congenital diaphragmatic hernia the cause of the delay in closure of the opening, or imperfect development of the diaphragm, must be attributed to a defective nutrition, probably due to disease of the chorion or placenta, since the condition is frequently associated with other defects in development. _ Those cases in which the herniated organs are covered by a membrane are probably to be explained by a failure in the development of the muscle fibres; the thin pleuro-peritoneal membrane is thus unable to resist the pressure of the viscera and becomes stretched out over them. It is quite possible, however, that the membrane instead of yielding would be ruptured, in which case a secondary opening would be formed in the diaphragm, at the margins of which the pleura and peritoneum would become continuous, thus resembing a persistent pleuro-peritoneal passage or isthmus of the ecelom. SUMMARY. The principal points which we have discussed in the preceding pages and which we wish to emphasise are the following :— 1. Accessory lungs are derived from the embryonic tissue of the “pulmonary groove,” or of the “ lung buds.” . 2. A portion of the embryonic pulmonary tissue, after having become _ A Case of Accessory Lungs 93 adherent to a neighbouring part or organ, from which it draws its blood- supply secondarily, may be separated from the parent tissue and grow independently. 3. We believe that the small size of the left pleural cavity, and the close relation of the mesodermal lobes of the left lung to the lower end of the cesophagus (see p. 88), would favour adhesion of the lung taking place on this side, and this may account for the greater frequency of accessory lungs on the left as compared with the right side. 4. It is probable that accessory lungs do not ordinarily originate as a secondary and independent outgrowth from the cesophageal portion of the foregut. 5. The adhesion of the lung bud to the septum transversum covering the liver, or to the wall of the pleuro-peritoneal passage in typical left-sided cases of accessory lung, may interfere with the normal retraction of the lung bud from the abdomen into the thorax, and cause a persistence of the pleuro-peritoneal opening. We are thus able to explain the occasional intra-abdominal position of accessory lungs, and the association of acces- sory lungs with left-sided congenital diaphragmatic hernia. 6. The closure of the pleuro-peritoneal openings is normally due to proliferation of the mesothelium covering the edges of the openings. 7. This closure of the pleuro-peritoneal openings is independent of the development of the muscular fibres of the diaphragm. 8. At the time when the pleuro-peritoneal openings normally become closed, the intestinal loop is situated within the umbilical ccelom, and it appears probable that congenital diaphragmatic hernia is due to the persistence of the opening, and takes place at the time, or after the return of the intestines into the abdomen. 9. The hernia usually consists of two loops: (1) The stomach, with the spleen and part of the pancreas. (2) The terminal part of the small intestine, and the colon as far as the splenic flexure. The duodenum and head of the pancreas are nearly always intra- abdominal. 10. A congenital diaphragmatic hernia is a true hernia, both in those cases in which the contents of the hernia are covered with a serous membrane and those in which they lie free in the pleural cavity. 11. A congenital diaphragmatic hernia cannot be regarded as a simple arrest in development, or persistence of any particular stage in the for- mation of the embryo. In conclusion we wish to express our gratitude to Professor J. E. 94 Dr E. A. Cockayne and Dr R. J. Gladstone Frazer for the loan of his serial sections of a 16-mm. human embryo from which the model of the diaphragm figs. 11 and 14 has been re- constructed, and to Professor Barclay Smith for kindly revising the manu- script, and the kind interest which he has taken in the work. BIBLIOGRAPHY. Accessory Lunes. (Those with diaphragmatic hernia marked *.) (1) Bau, Arbecten des zweiten allrussischen Kongresses der Veterindr-Aerzte in Moskau, 1910, ii. p. 548. (2) Beneke, Verhandl, deutsch. Path. Gesell., Jena, 1905, ix. p. 202. (3) Berr and Fiscugr, Yrankfurter Zeitschr. f. Path., 1910, vi. p. 27. (4) Campaccl, quoted by Griser. (5) Darter, J., Bull. Soe. Anat. de Paris, 1888, lxiii. pp. 892-897. (6) Dusier, Medizinische Gesell. in Basel, March 1888. (7) Dirox, Miind. med. Wochenschr., 1895, xlii. p. 456. (8) * von Goéssnitz, Jenaische Zeitschr. f. Naturwissenschaft, xxxviii. p. 619. (9) *Grtser, Grore, Beitr. zur path. Anat. u. z. Allgemeinen Path., Jena, 1914-15, pp. 491-500. (10) Hammar, Bertr. path. Anat., Jena, 1904, xxxvi. p. 518. (11) Herxuetver, Centralblatt Path., Jena, 1901, xii. p. 529. (12) Hueenin and Soret, Bull. Soc. Anat. de Paris, 1888, lxiii. p. 862. (13) Humpurey, Jowrn. of Anat. and Physiol., London, 1885, xix. p. 345. fia} von Joxst, Bericht iiber das Veterindrwesen im Konigreich Sachsen, 1905, pp. 294-295. (15) Kapuay, S., Jnaug. Dissert., Kinigsberg, 1913. (16) * Kaup, quoted by Grizer (two cases). (17) * Koun, /naug. Dissert., Konigsberg, 1913. (18) Lewisoun, Centralblatt Path., Jena, 1903, xiv. p. 869. (19) Luparscn and Osrertac, Ergebnisse der allgem. Pathologie, 1902, viii. p. 267. (20) Pauxut, E., Arch. f. Wissensch. u. prakt. Thierheilkunde, Berlin, 1912-13, xxxix. pp. 352-374. (21) QuenseL, Nord, Med. Arkiv, Stockholm, 1900, xi. pp. 1-8. (22) Rexroréik, Zeitschr. du k. k. Gesell. dw Aerzte in Wren, 1861, xii. p. 14. (23) * Rossmann, Jnaug. Dissert., Kénigsberg, 1904. (24) Roxrransky, Lehrb. der Path. Anat., Wien, 1861, iii. p. 44. (25) Ruan, Gesell. f. Geburtshiilfe wu. Gynaek. in Berlin, 1878, xii. 3. Berlin klin. Wochenschr., 1878, xxvii. p. 401. (26) Sacus, E., Gynaek. Rundschau, Berlin u. Wien, 1912, vi. pp. 853-860. (27) Seursam, Arch, path. Anat., Berlin, 1905, clxxx. p. 549. (28) Simpson, G, C. E., Journ. of Anat. and Physiol., 1907-8, xlii. p, 221, (29) Sprinoer, Prager med. Wochenschr., 1898, xxxi. S. 393, (30) Vionn, Virehow’s Archiv, Berlin, 1899, pp. 155 and 235. A Case of Accessory Lungs 95 (31) Vorsin, Arch. de méd. expérimental et de Vanat. path., Paris, 1903, xv. _ —p. 228. iv. (32) Wxcuszere, Centralbl. f. allg. Path. u. path. Anat., Jena, 1900, xi. p. 595. MEDIASTINAL AND BronowiaL Cysts with CrtiATED EPirHELIUM. (33) Hermann, Prager medizin. Wochenschr., 1890, xv. S. 146. (33a) Muyer, Virch. Archiv, 1859, xvi. p. 78. (34) Srituine, Virch. Archiv, 1888, cxiv. p. 557, (35) SrérK, Wiener klin. Wochenschr., 1897, x. p. 25. (36) Vircnow, Virch. Archiv, 1871, liii. p. 444. (37) Zaun, Virch. Archiv, 1896, exliii. p. 173; tbed., p. 416. Liver Cysts with CriiaTeD EPIrHELIUM. (38) Epertu, Virch. Archiv, 1866, xxxv. p. 478. (39) Frrepreicn, Virch. Archiv, 1857, xi. p. 466. (40) Menken, Berichte iiber Arbecten aus dem pathol. Institut der Univ. Wiirzburg, p. 72. (41) Von ReoxiincHausen, Virch. Archiv, 1881, Ixxxiv. p. 425. (42) Zaun, Virch. Archiv, 1896, cxliii. p. 175. (ESOPHAGEAL Cysts. (43) Kurn, Virch. Archiv, 1910, cci. p. 135. (44) Kraus, Vothnagel’s Handbuch d. Spez. Path. (Erkrangungen der*Mundhohle u. der Speiserdhre). (45) Mour, Ziegler’s Beitrdge, 1909, xlv. p. 333. (46) Rav, “ Kasuistische Mitteilungen von der Prosektur des Katherinenhospitals in Stuttgart,” Virch. Archiv, 1898, clili. p. 26. (47) Trespx, Arbeiten aus dem path. Instit. Posen, Wiesbaden, 1901. (48) von Wyss, Virch, Archiv, 1870, li. p. 143. (49) Zaun, Vireh. Archiv, 1896, exliii. p. 171. DEVELOPMENT OF D1APHRAGM AND CoNGENITAL DiaAPpHRAGMATIC HERNIA (see also Accessory Lune6s *). (50) Bary and Miusr, J'ewtbook of Human Embryology, 1909, p. 377. (51) Banuantyng, Manual of Antenatal Pathology, ii. p. 477. (52) Biscuorr, Archiv f. Gynaek., Berlin, 1885, Ixx. p. 437. (53) Bracuxr, Bull. de l’ Acad. Roy. de Méd., xlix., Brux., 1906. Ergebnisse der Anat. wu. Entw., 1897, iii. pp. 886-936. (54) Carrutuers, Lancet, 1879, ii. p. 503. (55) Crayron-GrEENE, see Harris, W. * oer Constantin, DantEL, Bull, Soc. Anat. de Paris, 1901, p. 422. 57) Durante, see Porak. (58) Frazer, J. E., and Rossins, Journ. of Anat. and Physiol., 1. p. 75. ° (59) Fucus, Sitz. Ber. die deutsch. Naturwiss.-med., 1898. (60) Grapstonr, Arch. Middlesex Hosp., 1908, xii. p. 38. (61) Godssnivz, Diss. Med., Jena; u. Jena Zeitschrift f. Naturwiss., xxxviii., N.F. 31, pp. 619-672. 96 A Case of Accessory Lungs (62) Harris and Crayton-GREENE, Brit. Med. Journ., 1912, i. Pp. 367. (63) Jones, Journ of Anat. and Physiol., 1913, xlvii. p. 282, (64) KerpeL, Vormentafeln zur Entwicklungsgeschichte der Wirbelthiere, 1908, 8 H., S. 67. (65) Keiru, Journ. of Anat. and Physiol., 1905, p- 243. (66) Maut, Johns Hopkins Hosp. Bull., 1901, xii. p. 158 ; and Keibel and Mee Human Embryology, vol. i. pp. 523-547. (67) Meyer, Journ. of Anat. and Physiol., 1914, xlviii. p. 153. (68) Monks, Brit. Med. Jowrn., 1914, i. p. 708. (69) Murray, Lancet, 1905, ii. p. 1472. — (70) Nau, Bull. Soe. ‘Anat. de Paris, 1903, p. 504. (71) PATERSON, Brit. Med. Jowrn., 1888, ii. p. 1207. (72) Piper, Anat. Anz., 1902, Ixi. ’D. 531. (73) Porak et DURANTE, Bull. Soc, Anat. de Paris, 1901, p. 354. (74) Prentiss, Textbook of Human Embryology, 1915, p. 188. (75) Ravn, Arch. f. Anat. u. Physiol. : Anat. Abt., 1889, 8. 123. Arch. Anat. wu. Entw., 1896. (76) Riscupiern, Australasian Med. Gaz., 1913, xxxiii. pp. 359-387. (77) Rosinson, Proc. Anat. Soc., 1900, xxix. » Journ. of Anat.and Physiol., xxxiv. (78) SwaEn, Journ. d’ Anat. et Physiol, 1897, xxxili.; Bibliograph. Anat., 1899, pp. 32, 222, 525. (79) THompson, Journ. of Anat. and Physvol., xlii. p . 170, and xlviii. p. 222. (80) Usxkow, Arch. J. mkr, Anat., 1883, xxii. 8. 143. (81) Vécrt, Amer. Journ. of Med. Sci., 1913, exlv. (82) WATERSTON, Journ. of Anat. and Physiol., 1915, 1. p. 24. DEVELOPMENT OF THE LUNGS. (83) Assy, C., Der Bronchialbaum der Sdugethiere und des Menschen, Leipzig, 1880. (84) Battey and Mirimr, 7'extbook of Embryology, 1909, p. 366. ’ (85) Brisnranskasa, Diss., Zurich, 1904. (86) Broman, Die Hntwickelungsgeschichte der Bursa Omentalis, Wiesbaden, 1904, p. 403 et seq. (87) Fut, Amer. Journ, of Anat., 1906, vi., i. (88) Griiber, Beitrdge path. Anat., Jena, 1914, lix. SS. 491-518. (89) Grosser, Orro, in Keibel and Mall’s Manual of Human Embryology, pp. 446-493. (90) Hammar, Beitrdge 2. path. Sens: u. allgem. Path., 1904, xxxvi. (91) His, Arch, f. Anat. u. Physiol.: Anat, Abt., 1887, S. 89. (92) Keiru, Journ. of Anat. and Physiol., 1905, xxix. p. 243. (93) Litt, The Development of the Chick, New York, 1908, p. 325. (94) Linsgr, Anat. Hefte, 1900, xiii. 8. 307. (95) Merk, in Bardeleben’s Handbuch der Anat, des Menschen, 1902, vi. S. 98. (96) Meyer, Anat. Anz., 1910, xxxvii. S. 449, (97) Naratu, Bibliotheca Medica, Abt. A., Anat., iii, Stuttgart, 1901, s, 354. (98y Warterston, Journ. of Anat. and Physiol., l. p. 24. F FORM AND FUNCTION OF TEETH: A THEORY OF “MAXIMUM SHEAR.” By D. Mackintosh Suaw, Royal Dental Hospital, London. THIs concept, “maximum shear,’ has been applied to explain in a.more precise manner than has hitherto been done the functional meaning of many specific features in the shapes of teeth. It offers also, in the case of man’s dentitions, an explanation of the chewing mechanism that. fits all the observed dynamical conditions and actual results. If it be granted that teeth were evolved mainly for their utility in manipulating food, then a close and persistent study of dental mechanism from the “machine and tool-action” point of view is a very rational and promising method of interpreting the physiological meaning and value of any morphological - feature or detail in teeth. It would seem, a priori, that the definiteness and constancy that characterise each specific feature in the form of a tooth might well be directly related to a like definiteness and constancy in function or functions. If a morphological feature remains, in a clearly recognised sense, fixed and distinctive throughout that particular species, the functional requirement that evoked and preserved it is admitted to be, in several obvious instances, similarly fixed and distinct. But the same relationship most probably exists in the less obvious instances, and possibly in all, or nearly all instances, if only we could discover and verify the relationship. What here follows is an attempt to explain some of the results of my endeavours in this direction during the past twelve years. Attention for the present is confined to man’s teeth, where the opportunity for inquiry and investigation is much more favourable than we would hope for in other animal dentitions. In the dentition, of perhaps every species of asad the form and mechanism of the teeth are influenced—one might almost say determined —by two conditions that are fundamental: (1) the nature of the food; (2) the degree of its reduction. In regard to the first, we know something of the nature of man’s food, although unfortunately very little is known with certainty as to the nature of his “staple” foodstuffs during the period in which the main features of the normal present-day dentition were evolved and established. Regarding the second, the degree of reduction of food, VOL. LIl. (THIRD SER. VOL. XIII.)—OCT. 1917. 7 98 Mr D. Mackintosh Shaw it is well ascertained that man’s teeth are adapted to (that is to say, they can and do) comminute hard and tough food to a fine degree. Such distinctive characters in the teeth as hardness and sharpness must correspond to qualities of hardness, toughness and the like, in the food. It appears, then, that in investigating man’s dental mechanism the problem is mainly concerned with the behaviour of hard and tough material when subjected to certain stresses. There is perhaps no subject . to which a greater amount of close scientific study and experiment has been devoted than the behaviour of hard and tough material under various stresses. An immense body of knowledge has been built up by engineers and mathematicians and applied to useful purposes, but, so far as I am © aware, neither the knowledge gained nor the methods of investigation have been at all seriously applied to the study of dental mechanisms. True it is that the engineer, the physicist, and the mathematician worked mainly at problems relating to such hard materials as metal, wood, ete.; but they were and are prepared to use the same methods as far as possible in dealing with any material or structure that might come before them for inquiry, and no odontologist will suggest that a new kind of physics or mechanics is called for in the study of dental mechanism. The same principles and methods will apply; although a distinct and somewhat peculiar difficulty arises from the fact that the engineers’ inquiry into strains and stresses is mainly directed towards preventing the break- down of material or structures, while our particular object is to find out how Nature went about to facilitate the breakdown of material with economy of effort. Material is strained and may rupture or break under the action of one or more of three principal stresses: (1) compression; (2) tension; (3) shear. In a given case rupture may be produced as a response to one only of these three principal stresses, or two or the three kinds of stress may be so combined and intermingled that it is difficult to say whether a particular one predominates. More often failure takes place as the result of combined stresses, one of them playing a dominant part in the rupture. A point on which it should be worth while to fix attention is this :— hard and tough material is more easily (requiring less effort) broken up by a shearing stress than by either a compressive or tensile stress. This is true for many metals and alloys, and for many other materials is so well recognised that the estimated “safe limit of stress” is given as consider- ably lower for shear than for tension or compression. The tests upon which these results were based were of course only applied to materials employed for constructional purposes. No such tests have been applied Form and Function of Teeth: A Theory of “Maximum Shear” 99 to foodstuffs. But, relatively to metals, among the constructional materials wood may be regarded as approaching in some degree more towards the range of the moderate hardness and toughness found in foodstuffs. For two woods, oak and pine, the following values have been given for the breaking stress in lbs. per square inch :— Tensile. | Compressive. | Shearing. Wood,oak. . ./| 15,000 10,000 2300 » pine e ; 12,000 6,000 650 It does not seem needful at this stage to present further evidence that, as we pass under review the action of stresses on materials possessing those qualities of hardness and toughness that lie between the extremes of metal on the one hand and fibrous foodstuffs on the other, there is found a marked and increasing effectiveness in the power of shearing as compared with tensile or compressive stresses. Further, in any device for the break-up or reduction of material by shearing stresses, by a very simple adjustment a time factor can be intro- duced by which the resistance at each instant is lessened, with a consequent reduction in the magnitude or intensity of the force required. A familiar example is a pair of scissors, where the edges of the blades are so inclined to one another that the material is attacked and severed gradually, or bit by bit, and not instantaneously at every place, as would occur if the edges were parallel and coincided. It is not usually practicable to secure this particular advantage (in economy of effort) when tensile or compressive stresses are used. Returning now to the natural chewing mechanism, it is an obvious and familiar fact that man’s incisors are adapted for shearing. But in written references to the functions of the cheek teeth—whether made by odon- tologists, zoologists, or evolutionists—there seems to be a fixed belief, either implied or stated positively, that the cheek teeth act mainly by “crushing” or “grinding.” There is a certain handiness (as well as vagueness) about these terms, “crushing” and “grinding,” that makes them quite allowable in popular descriptions of teeth: unfortunately the same attributes have helped to perpetuate their use in scientific writings. In drawing attention to the fact that man’s cheek teeth normally act mainly by shearing and not by crushing stresses, I would urge that the distinction is far from being a merely verbal or academic one. My hypothesis of “maximum shear” may now be stated as follows :-— The teeth of man, alike in regard to many of their specific morphological 100 Mr D. Mackintosh Shaw features, the manner of interaction between the opposing rows, and the precise character of the jaw movements that are habitual and effective, » are all normally adapted and used to secwre in effect the dominant — condition that the shearing stresses—not the compressive stresses—are at a maximum. The conditions that are here more or less indispensable for effective shear are: (1) a certain amount of “overlap” arranged for between the opposed shearing edges ; (2) a certain degree of sharpness in the shearing edges. Now both of these prime conditions, although usually acespted as specially and distinctively characterising the anterior teeth only, are found on examination to hold also—and in at least equally effective degree—for the cheek teeth.. The amount of overlap (distance between the shearing edges when the teeth are in occlusion) may be somewhat greater in the anterior than in the posterior teeth; but that makes no difference in a close shear, the effective action of which ceases the instant that an opposed pair of shearing edges have just passed one another. Then as to the sharpness, I find that the sharpest edges are not to be found on the incisors, but on the “cusped” or posterior teeth. It should be borne in mind that it is normal wnworn teeth that are at. present under consideration. Examination and experimental tests dismiss any possible doubt that the statical conditions for efficient shearing—the form and relationship of tool” blades, edges, and surfaces—are present and no less marked in the posterior or cusped teeth than in the incisors. And turning to the dynamical and kinematical conditions, we find that they are at least quite as favourable for efficient shearing in the cheek teeth as in the incisors. In strict fact these conditions are by one remarkable adaptation made more favourable for the cheek teeth, inasmuch as in the general mechanism of the “machine ” there is a sliding and guiding contact between the canines that ensures for all the cheek teeth the precise alignment of the shearing edges that is required for efficient shear: the edges are thus also prevented from “clashing,” or being needlessly deteriorated. I have observed and verified this “fine adjustment” in many sound normal dentitions, and, what is indeed remarkable, have found it also working almost to per- fection in some hundreds of dentitions that were far from normal in arrangement, During mastication, the outstanding functional effect of direct protru- . sion Of the mandible is to bring the lower incisors into shearing relation- ship with the upper incisors; so likewise the outstanding functional effect of lateral deflection of the mandible is to bring the shearing edges of the Form and Function of Teeth: A Theory of “Maximum Shear” 101 lower cheek teeth into precise alignment with the edges of the upper ones. (Diagram fig. 1, a, b, and’c.) THE DIRECTION of the shearing thrust between any opposed pair of teeth will be in and along the general plane in which the overlapping and coincident surfaces of the shearing blades lie. The inclination of those surfaces (that come into sliding contact-and determine the direction of the thrust) varies with the individual or group type in the series, being somewhat vertical on the incisors and becoming progressively more horizontally inclined towards the end-member of the molar group. With advancing age and wear those surfaces or planes become more flattened or horizontal, and the shearing edges are gradually shifted outward. And it is a well-observed fact that as the cusps become worn and flattened, the character and direction of the jaw movements become altered; the lateral Fie. 1.—-Diagram showing direction of effective normal thrust in (a) incisors, (b) 1st premolars, and (c) molars. deflection of the mandible becomes more marked, and the direction of the thrust more horizontal (fig. 3). The “maximum shear” hypothesis affords a clear explanation of the alteration in the character of the jaw movements in advancing age, wherein the worn and deteriorated machine endeavours, by a more and more laterally-directed thrust, to close up the failing shear and maintain the optimum shear available. The shearing edges thus far spoken of are all on the external blades of the tooth row, where they lie in a smooth and practically continuous curve from one 3rd molar to the other. What about the internal row of cusps? it may be asked. It can be demonstrated that the presence of the internal row of cusps does not interfere with or prevent the shearing action of the external cusps. What then are the functions of the internal or lingual row of cusps ? To answer that question with any degree of preciseness it will be convenient to consider first that pair of opposed lingual cusps that come next to the anterior single-bladed teeth—the lingual cusps of the Ist premolars. Let us imagine that a piece of food (of a kind that requires 102 Mr D. Mackintosh Shaw hard and sharp teeth to operate on effectively) has been bitten or sheared off by the incisors or anteriors. We may permit ourselves the assumption that the morsel is passed on to the premolars for further reduction. Let the diagram. (fig. 2) represent a pair of opposed Ist premolars in an observed functional relationship, with a portion of the food morsel ‘in the position indicated. The lingual cusp of the 1st lower premolar is so short Fic, 2.—Showing function of lingual cusp during chewing stroke. and small that—without “ prejudice” to any theory of its function—it may for the moment be disregarded and attention concentrated on the bold lingual cusp of the upper premolar. What is the upper lingual cusp actually doing? The accepted notion that lingual cusps were evolved to act as “crushers” is not so firmly established by proofs—or indeed so supported by any published evidence— Fie. 8. —a, unworn ; 6, worn ; c, more worn. that we must refrain from making a fresh examination and inquiry. The upper lingual cusp is doing very little in the way of crushing; in point of fact it is doing even less crushing than might seem to be represented in the diagram, inasmuch as the two teeth do not, of course, come into occlusion with their morsal. surfaces directly opposite, but with one rather in front of the other. A good deal of observation and experiment has satisfied me that this prominent cusp does something more important and necessary than a little doubtful or adventitious crushing; it holds (and helps tlie Form and Function of Teeth: A Theory of “Maximum Shear” 103 tongue to hold) the food fragment in proper position for effective shearing by the precisely-aligned external blades. In any shearing mechanism there must be some provision, either made or existing, for holding in proper position the material to be operated on. The smaller the fragment (up to a certain limit) the more elaborated or refined must be the devices required for holding it in position for further reduction—as in the molar teeth. There is firm ground for believing that a close study of this paramount requirement for food-holding devices or adaptations will prove of real value in the more rational interpretation of many of the varied cusp forms and cusp arrangements in different kinds of animal dentitions. Returning to the masticating mechanism of man’s premolars, my pro- position that the external cusps are adapted mainly for shearing and the internal cusps mainly for food-holding will, it can be shown, bear well a more searching examination. The differences in form that distinguish the di Wy, % eh jst. Fie, 4, outer from the inner cusps require examination in detail. Fig. 4 shows the morsal or occlusal aspect of the left upper premolars. It will be observed that the cutting edges of each buccal cusp lie almost entirely in a plane, the continuous line of the cutting edge being bent in that plane only at the cusp point. On the other hand, the cutting edges of each lingual cusp do not lie in a plane, but are crescentic and curved outwards from the cusp point. Now in both cases those are exactly the special modifications that would be of prime utility in the functions I have referred them to—the “straight,” sharp buccal blade for shearing, and the curved (pointed, but more rounded) lingual blade for food-holding; more strictly—in the latter case—to help the tongue in food-holding, the lingual cusp being smaller and very smoothly rounded off away from the tongue, and the concavity of the erescentic cutting-edge being turned outward to hold the food against the inwardly-directed shearing thrust. That the tongue by itself is a very inefficient instrument for holding small food fragments in position can be attested by anyone who has had the misfortune to lose all his cheek teeth. Then the work done—by the anterior teeth only—is tedious, and performed in a conscious and niggling manner ; the free, rhythmical activity of normal function is conspicuously 104 Mr D. Mackintosh Shaw absent. The tongue, helped by the lips, etc., can hold the larger fragments in position without risk of wounding itself; for finer reduction the assist- ance of an inner row of special “holders” is required. On the molar cusps trwe crushing takes place also, but the effective so- called “ crushing” and “ grinding” which actually disintegrates the food is in reality—and even with much-worn teeth—mainly a shearing action. To the food-holding interpretation of man’s lingual cusps it may be objected that in many animal dentitions (selenodont, etc.) the buccal cusps are crescentic, and also have the concavity of the erescent turned outward. That objection might be valid and might even throw light on the functional meaning of man’s cusp-forms if only something definite were known of the functional meaning of the (apparently) like features in those animal dentitions. But no definite interpretation of those specific features has yet been put forward, and the vague and general notion of a “crushing” function has so far proved to be almost completely unproductive in the field of comparative odontology. In this line of inquiry comparative work will continue to be somewhat barren in results until we search out in each type or species the two prime factors already mentioned. (1) The physical nature of the food: size, hardness and toughness, surface texture (rough- ness or smoothness, dry or slippery, etc.); distinct and more-or-less fixed differences in configuration, such as grasses, that are large in one dimension only—length; leaves that may be large in two dimensions—length and breadth ; and fruits, seeds, roots, animal tissues, etc., that may be relatively very large or very small in all three dimensions. (2) The degree of reduc- tion or comminution that has been effected when the food leaves the mouth, Assuming that, in regard to these primary factors, adequate or the best procurable information had been gained, one would then be in a fair position to examine the general structure and mechanism of the dental apparatus, and ascertain the particular functions of its various parts. In even the simplest kind of tool there is invariably more than one (often three or more) requirement of construction and material found to be essential to its utility; e.g. a lancet must have hardness, sharpness of edge, and thinness of blade. And in a machine or contrivance of moderate com- plexity we can usually, after examining it at work, pick out and correctly assign to separate parts of it distinct and separate functions. By way of summarising in some kind of related order some results of investigation on the mechanism of mastication in man the following table is presented :— Form and Function of Teeth: A Theory of “Maximum Shear” 105 MANn’s DENTITION AND MAstTIcaTING MECHANISM. Relation between Nature of Work done and the Form and Function of Interacting Tissues. 1. Physical nature of food :—Hard, with some toughness ; average or general size — moderate in all threedimensions. 2. Degree of its reduc- tion :—Finely com- ( minuted, (a) Blade edges of optimum sharpness, so opposed and moved as to break up food chiefly by shearing stresses. (b) Characteristics of surface shape adapted to protect gum tissue from injury by hard fragments during the shearing thrust. (c) Tongue-shielding adaptations, the “ umbicle.” Lingual cusp-edges retreating from lingual side. (d) Conditions or devices that are effective in conserving the optimum sharpness of cusp- edges :—the guiding function of opposed canines and anteriors; deep grooves formed by the convergence of convex walls. F (a) Freely movable soft tissues or organs that re- place and hold the food in position between the tooth rows: tongue, cheeks, lips. - (b) Lingual cusps that hold food in right posi- tion during shearing stroke. (c) Order and arrangement of blades and cusps adapted for normal reduction of food to be progressively increased oie incisors to molars. (ad) Morphological features ‘canon and special) so interacting as to allow of “chip. clear- ance,” and prevent the impacting and inter- dental wedging of food particles or morsels. Most of the function interpretations summarised in the above table have been noted and described in some detail in previous papers’(Dental Record, Aug. 1909, Sept. 1912, and Aug. 1914; Natwre, July 23, 1914; and a paper presented at the International Dental Congress, 1914, Section of Anatomy and Physiology). It is not implied that some or any of these interpreta- tions apply exclusively or uniquely to man; no more is here predicated than that certain distinc tive features, relationships, movements, etc., were observed to be very effectively associated with certain useful results, and 106 Form and Function of Teeth: A Theory of “Maximum Shear” that after prolonged search no other or different kind of associated utility could be observed or—in some cases—even conjectured. This is at least as high a degree of probability as exists in the numberless cases where adaptation to a particular function has been posited of an organ or part and accepted as a settled fact. Statements made upon the functional meaning of the tooth-forms of even familiar animals are often based upon guesses and conjectures which could only be tested and verified, if at all, by prolonged and tedious research, so far unattempted. And in the controversies that turn upon the evolution of mammalian teeth, eminent zoologists, when discussing the origin and gradual progress in the phylogeny of a new or changing morpho- logical feature in particular teeth. do not hesitate to support the argument by saying that the modification arose and persisted because of its utility in this or that more or less vaguely specified way. In the case of man’s teeth, the new results summarised in the above table can be definitely verified or rejected by an investigation in which there should be no insur- mountable difficulties. As the physical factors are almost entirely mechani- cal and relate to dynamics rather than to statics, a fair knowledge of and interest in the precise working action of tools and simple machines would be specially useful, if one may say so. The findings tabulated under 1 (@), (c), and 2 (d) are interesting from the anatomical point of view; 1 (6), 2 (b), and (c) should be of some value to the morphologist; and the last three, considered together, offer a solution to the baffling problem of how to trace, in the lower steps or minute beginnings, any definite utilities by which an evolving cuspule might be supposed to climb into prominence and an obvious crushing or other function. In particular, the way in which a scarcely discernible but quite effective gum- shielding feature may become modified and enlarged, so as to take part in food-holding and other utilities, can be clearly followed out in the existing variations among normal human teeth. But adequate description of these, and a clear presentation of the related facts observed, would require another paper. THE EARLIEST STAGES OF DEVELOPMENT OF THE BLOOD- VESSELS AND OF THE HEART IN FERRET EMBRYOS. By Cuune-Cuine Wana, M.D., Ch.B. (Edin.), Acting Lecturer and Demonstrator in Anatomy, University College, London; late Carnegie Research Fellow in Embryology, and Assistant in the Department of Anatomy, Edinburgh University. INTRODUCTION. BEFORE the details of the reconstructions which were made are given and the conclusions to which they tend are discussed, it seems advisable to state briefly the main results of the observations which have previously been made by other investigators regarding the formation of the intra- embryonic vessels and blood-cells, and the earliest stages of development of the heart in mammals generally. It is to be noted that the phenomena observed are, of necessity, intimately associated with the early stages of development of the pleuro-pericardial cavity, and are closely connected with the mode of formation of the pre-umbilical portion of the body of the embryo. To facilitate references to be made hereafter in this communication, the ferret embryos are classified into stages. Thus in Stage I. the deserip- tion deals principally with the blood-cells and vascular endothelium; in Stage II. the account relates chiefly to the first appearance of the heart rudiment as a single transverse vascular channel situated caudo-ventral to the pleuro-pericardial cavity; in Stage III. it is shown how the single heart tube is converted into two lateral heart tubes; in Stage IV. the _ statement refers mainly to the conditions of the two endothelial tubes and their relationships to the muscular wall of the heart; and in Stage V. the site of fusion of the two heart rudiments to form an unpaired heart tube is indicated. As far as the technique and the histological conditions of the ferret embryos, to be immediately described, are concerned, the description given 1 An abstract from a thesis for the degree of M.D., for which a gold medal was awarded by the University of Edinburgh. 108 Dr C. C. Wang in Stage IV. may be applied equally well to all the other specimens selected for the purpose of this investigation. The only differences to be found in ~ this respect are chiefly in the matter of staining and in a few other minor points, such as, whether the embryo was detached or not from the uterus before being sectioned. BLoop-VESSELS AND BLOOD-CELLS. It may be mentioned at once that-the question of the origin of the vascular - rudiment is one of the most obscure in the realms of comparative embryology. Even the lowest vertebrates, which present the greatest simplicity in their structures and whose development is most easily understood, have failed to throw satisfactory light on this question. ; There is a great diversity of opinion regarding the germ layer from which the vascular endothelium and blood-cells arise. In the literature it may be found that certain competent investigators have in each verte- brate class claimed the vascular endothelium and blood-cells to be derived from entoderm, while other workers of equal authority have found the vessels and blood corpuscles to arise from the mesoderm. It is to be noted, however, that in no case has an author stated that the blood-cells and vascular endothelium are derived from different germ layers. . Ziegler (87) expresses the view that “the system of blood-vessels and that of the lymphatic vessels are produced in their first fundaments from remnants of the primary body cavity (blastoccel), which at the general distribution of the formative tissue (mesoderm) remain behind as vessels, lacune, or interstices, and are closed by that tissue and incorporated in it.” This author therefore agrees with Biitschli (82) that in all metazoa the blood vascular system has its origin from the blastoccel. On the other hand, Felix (97) inclines to the belief that the circulatory system is, from a developmental point of view, closely related with the ccelom. In reptiles, Strahl (’83), in birds, K6lliker (84), in Selachians, Ziegler ('92), and in mammals, Kélliker (’84), all claim that the first vessel rudiments are found in the mesoderm and not between the mesoderm and entoderm. The current view which is held by a great number of investigators is that in embryos of the higher vertebrates, the first vascular rudiments which can be identified as the forerunners of the blood-vessels and blood-cells, appear, at first, in the form of localised cell cords lying upon the yolk-sac between the mesodermic tissue and the entoderm, His (00) gives the name of “angioblast” to these cell cords. Upon the question as to whether the angioblast is to be looked upon as a derivative of the mesoderm, or as an offshoot from the entoderm, opinions again diverge. In support of the mesodermic origin, the names Development of Blood-vessels and Heart in Ferret Embryos 109 of Maximow (09), Weidenreich (10), Evans (12), Riickert and Mollier (’06) are of prominence, while the advocates of the entodermic theory are to be found in the persons of Kélliker (82), Robinson (92), Keibel (88), and Van der Stricht (99). Miss Parker (15), in her recent investigation into the development of the heart in Marsupials, admits the possibility of the entodermal origin of the endothelium. In Fundulus, Stockard (’15) finds that vascular endothelium does arise in sitw in many parts of the embryonic body in which blood-cell rudiments are not present, and that independent blood islands are found on the yolk-sac. The persistent claims that vascular endothelium has the power to change into various types of blood corpuscles have been disproved by the recent experimental work of Stockard (°15). After considerable study and eareful observations, Stockard observed nothing that would indicate that the vascular endothelial cells possess the power to change into the blood- cell type, nor could he find any evidence to indicate that cells having once assumed even the earliest blood-cell type are capable of metamorphosis to form endothelial cells. He firmly believes that, in Fundulus at least, endothelium is incapable of giving rise to any type of blood-cell. Whatever its origin, the angioblast, according to current views, is, in the majority of amniota, chick for example, found lying between the mesoderm and entoderm in the form of cell cords in the area vasculosa immediately surrounding the embryonic shield. It is believed that the peripheral part of each angioblastic strand soon resolves itself into an uninterrupted network of endothelium, and the central part into clusters of blood-cells. The endothelial cells which enclose the blood-cells continue to divide and produce vascular sprouts which appear, at first, as solid cords but later become hollow (Hertwig (92), Minot (’12)). Around the periphery of the area vasculosa, in the majority of animals, the vitelline plexus resolves itself into a broad circular vessel—the sinus terminalis, which is continuous round the margin of the area except at the cranial end where it terminates on each side in a vessel which enters the embryo. The vascularisation of the splanchnic layer of the mesoderm gradually extends through the extra-embryonic region of the zygote until it covers the whole extra-embryonic region, where it forms an intermediate layer between the entoderm and mesoderm. Some of the larger channels of the area vasculosa, even in the early stage of its formation, converge to form a single vessel on each side, which enters the embryonic body through the splanchnopleure and ultimately joins the venous end of the heart rudiments when they are developed. These are the two vessels previously mentioned as connected with the cranial end of the sinus terminalis, and are known as the omphalo-mesenteric (vitelline) 110 Dr C. C. Wang veins. It is believed that other channels of the area vasculosa on each side grow towards the median plane of the embryo, and as they approach the region of the notochord their extremities fuse to form a longitudinal vessel which is the dorsal aorta of that side. The dorsal aorta ultimately unites with the arterial end of the heart. In this way the vitelline circulation of the embryo is completed and plays an important role in all vertebrates in supplying the growing embryo with nutritive materials from the yolk-saec, which is comparatively large. In reptiles and birds, a second circulation, as it were, develops in con- nection with the allantois, and persists during incubation; in mammals the allantois is incorporated with the placenta which establishes the com- munication between the embryo and the mother, and the vessels which correspond to the allantoic vessels in reptiles and birds become associated with the placental circulation (vide infra). Consideration has been confined so far tiainhy to the part played by vessels which in origin are extra-embryonic. The next question to be considered is whether the early blood-vessels in the body of the embryo itself are formed by an ingrowth of the vitelline plexus which is, by origin, composed of extra-embryonic blood-vessels, or whether, on the other hand, the intra-embryonic vascular system, or at least a part of it, arises im situ from the germ layers of the body. The problem is difficult, and it is not surprising that various conflicting views have been formulated regarding the precise mode of origin of the intra-embryonic blood-vessels. Thus, the names of Hertwig (92) and His (00) have been identified with the idea that the early blood-vessels in the body of the embryo itself are formed by a budding or ingrowth of the endothelial wall. of the vessels from the extra-embryonic vascular area, and the name of Sobotta ('02) is associated with the belief that there is an outgrowth of the vessels from the body of the embryo to the wall of the yolk-sac. Riickert and Mollier (06), on the other hand, maintain that the embryonic vascular stems, or at least a part of them, arise in sitw from the mesoderm of the embryo. Other investigators, basing their opinion on the results of a series of experiments on growing chick embryos, adhere to the conviction that, even after the destruction of the yolk-sac vessels of one side, the heart, the aorta, and the other vessels are found to develop on both sides in the embryo. On the other hand, Vialleton (92), His (00), and Evans (’09), who have investigated the intra-embryonic blood-vessels in birds, and to whom we are indebted for a comprehensive knowledge of the formation of the caudal portion of the dorsal aorta, come to the conclusion that the greater part of the dorsal aorta in the bird is formed from the medial margin of the Development of Blood-vessels and Heart in Ferret Embryos 111 vitelline plexus which has grown into the embryo in the manner already referred to. Tiirstig (’84) also has noticed the frequent early connection of the primitive dorsal aorta with the vitelline plexus in mammals. With regard to the development of the cranial portion of the aorta, on the other hand, various opposing views are held; thus, His (00) attributes it to the result of a further growth of the same extra-embryonic vitelline piexus which forms the caudal part of the aorta, but which is reduced to a capillary chain growing headwards, eventually turning ventrally over the blind end of the fore-gut and fusing with the cranial portion-of the heart tube. Lewis (04) and Bremer (12) both arrive at a somewhat similar conclusion, namely, that in the rabbit embryo, the dorsal aorta, the aortic arch, the conus arteriosus, and the lateral heart are all parts of an original network of angioblastic cords derived from the extra-embryonic plexus of blood-vessels. On the other hand, Mollier (’06) believes that the notion of His (75) and Vialleton (92) is not nearly so probable as that the individual intra-embryonic vessel cells arise im loco and thus form the vascular nets. ‘Recently the local origin of vascular endothelium received additional support from the experimental results recorded by Miller and M‘Whorter (14) on the origin of blood-vessels in the chick embryo. Such a view is further strengthened by the still more recent experimental evidence pre- sented by Reagan (’15), which shows the origin in loco of vessels in isolated parts of chick embryos, and by Stockard (15), which claims that in Teleost embryos there can be no doubt that the heart endothelium and aortz arise im situ within the embryo. It is to be noted that all these experiments just quoted confirm the earlier results of Hahn (09) on the origin of vessels in the chick, and that the formation of intra-embryonic blood-vessels is nuch more extensive and important than has formerly been supposed. Though the development of blood-cells lies without the scope of the present communication, a few remarks may be made regarding the close relationship of these cells to the vascular endothelium. Nearly all in- vestigators on this subject assert that blood-cells and vascular endothelium arise from either the mesoderm or entoderm. In reviewing the literature on this point no definite statement has been found which might suggest _ that blood-cells and endothelium develop from different germ layers. The possibility of these structures having an independent and separate origin cannot, however, be overlooked, as will be seen later. Maximow (09) believes that the endothelial cells and blood-cells are closely related, and arise from a common stem cell in the blood island, and may continue to arise from such a cell during later development. Stockard 112 Dr C. C. Wang (15), on the other hand, states “that vascular endothelium forms in per- fectly normal fashion within the heart and head regions of embryos without circulating blood, but in no case in early or late stages was the endothelial lining of the aorta or other vessels capable of giving rise to any type of corpuscles: Yet the power to form blood corpuscles was abundantly present in the same embryos as shown by the huge numbers: of blood-cells within the blood- forming regen ae intermediate cell mass and yolk islands.” Maximow (’09) states farther: that the intra-vascular primitive blood- cells are not only increased by mitosis but are added to also by the pro- liferation of the same kind of cells from the fixed endothelial cell of the primitive vessels. The assumption is based on the fact that clusters of blood-cells are often seen adherent to the endothelial wall of the blood- vessels. Maximow thinks that these clusters of cells may arise by the proliferation of the endothelium. Minot (’12), however, disagrees with Maximow, because he finds that there is no continuity of the protoplasm of the cells either in the rabbit or in man; also, because mitosis of the endothelium -in the neighbourhood of the clusters is almost invariably wanting; and, finally, because the endothelial nuclei are differentiated, while the nuclei of the cells of the clusters are not differentiated. Minot regards the cells composing the clusters as solely primary wandering cells, Stockard ('15) concludes that endothelial lining is utterly incapable of giving rise to any form of blood-cell. As in mammals, before referred to, so in man, in connection- with the vascularisation of the yolk-sac, or, according to Fetzer (10), even at a period before any vascular rudiment on the yolk-sae proper can be distin- guished, there develop, in the belly-stalk and chorion of the embryo, highly characteristic strands of spindle cells, which repeatedly exhibit the nature of having the appearance of a double row of nuclei and of possessing a distinct lumen. This has been observed by Graf Spee (96) in the embryo von Herff of 37 mm. The strands of spindle cells have been claimed to form endothelial cells eventually. In young human embryos, without any vessels or blood islands on the yolk-sac, Jung (07) and Herzog (09) have called attention to the aggrega- tion of cells, sometimes arranged round a lumen, situated at the periphery of the mesoderm of the yolk-sac and belly-stalk in the neighbourhood of the extra-embryonic area, In slightly older specimens with recognisable yolk-sac vessels, irregular spaces in the mesoderm, some lined with en- dothelium, some without any definite lining, have been observed by many authors, and recently Grosser ('13) and Debeyre ('12) have independently described, beside the irregular spaces, true blood islands in the belly-stalk Development of Blood-vessels and Heart in Ferret Embryos 113 near the allantois. In human embryo 1:17 mm. described by Frassi (’08) there is an abundance of well-formed vascular rudiments on the ventral surface of the yolk-sac, and Frassi states that with little or no difficulty vessels can be detected also in the belly-stalk and chorion. The next phase of development of. the human vascular system is illus- trated by the well-known embryo Glaevecke 1:54 mm. of Graf Spee (’89, '96). Here again, as in the preceding stage, vascular rudiments are seen on the yolk-sac and in the chorion, but, in addition to these, it is possible to note the first intra-embryonic vascular rudiments. In embryos of 5 somites and upwards it is observed that the © vitelline plexus has established its communications cranially with the heart by means of two channels, the vitelline veins, reaching as far as the first intersegmental cleft, as Dandy (10) first showed, while caudally, in the region of the unsegmental mesoderm, the branches of the vitelline arteries form a plexus of capillary-like vessels from which, as shown by Felix (10) and confirmed by Evans (12), the umbilical artery takes its origin. THE HEART AND PERICARDIUM. In mammals much has been done to throw light upon the develop- ment of the heart, notably by His (’81, ’85, ’86) and Bischoff (’42, 52), in rabbits, Born (’89), in dogs, Bonnet (’91, ’01, 07), in pigs, Keibel (’88), and in ferrets, Yeates (11), but many gaps must be filled up before it is possible to obtain a clear conception of the details of its formation. Tandler (12) says “the earliest developmental processes of’ the heart, especially in so far as they concern the formation of the endothelium of the heart and vessels, are unknown in the human embryos, but probably one will not be far astray in assuming that the earliest rudiment of the human heart is essentially similar to that of the mammalia.” The earliest stages of development of the mammalian heart are undoubtedly intimately associated with the development of the blood-vessels, but concerning the latter various opposing views have been formulated and already been dwelt upon in the beginning of this communication. It is clear that the precise mode of development of the blood-vessels is not yet definitely established, and a short survey of the early stages of the development of the heart will show that our knowledge of that subject also is deficient. In mammals, according to Mollier (’06), the first rudiment of the heart is the appearance of a number of cells, which are discernible in embryos of 2-3 primitive somites. The vascular cells appear between the entoderm and mesoderm on both sides not far from the median plane of the embryo, at first in the distal portion of the VOL. LIL. (THIRD ‘SER. VOL. XII1.)—OCT. 1917. 8 114 Dr C. C. Wang head. They are responsible for the formation of the endothelium of the heart tubes only, the remaining constituents of the wall of the heart— that is, the myocardium and the epicardium—being derived from that part of the visceral coelomic wall which has been designated by Mollier (06) the heart-plate or cardiogenic plate. It is generally held that the first aggregation of the vascular cells of | the mammalian heart is paired and is situated ventral to the ccelomic cavity. By a process not yet satisfactorily explained, spaces soon make their appearance in the vascular cell mass, and when these spaces coalesce, two endothelial tubes are thus formed, one on either side of the median plane of the embryo. A fusion of the two endothelial tubes next takes place, and the unpaired heart tube is formed from the paired heart rudiments, but exactly how this fusion is brought about, opinions differ. It was for long believed (Balfour (81), Hensen (76), Hertwig (’92), Kolliker (61), and it is still .held by some (Tandler (12), Bryce (08), Bailey (12), Wilson (14), H. von W. Schulte (715), and others) that in mammals, as in birds, the two endothelial tubes, out of which the heart is formed, appear at a time when the lateral folds, which are said to form the ventral wall of the throat, are only just visible, that, as the lateral folds of the splanchnic walls increase, the two halves of the heart, enclosed within the hitherto symmetrical and laterally placed pleuro- pericardial cavities, become carried medially and ventrally until they fuse on the ventral side of the fore-gut, and that the heart is therefore provided, at least for a time, with a ventral and a dorsal mesocardium. . In the chick, a ventral mesocardium is recognisable, but this is due, as Robinson (02) points out, to the relatively late penetration of the mesoderm in the cranial region. In amphibians lateral folds have been described, but it is erroneous to presume that such folds, which, by virtue of their fusion ventrally, form the ventral wall of the fore-gut, really occur in mammals. In the latter, the pericardial mesoderm appears in the pericardial portion of the embryonic area, and it is there completely differentiated into somatic and splanchnic layers before the head bend is developed; there is therefore a single pericardial cavity to begin with, which extends from side to side along the cranial boundary of the embryonic area. As the head bend develops, the single pericardial cavity is reversed, and it is carried into the ventral wall of the fore-gut, where it forms a U-shaped tube which communicates at each end with the general eccelom, The heart rudiments are formed in the splanchnic layer of the pericardial mesoderm; therefore, after the reversal of the area they lie in the dorsal wall of the pericardial cavity attached only Development of Blood-vessels and Heart in Ferret Embryos 115 by a dorsal mesocardium to the ventral wall of the fore-gut, but they are never, at any time, connected with the ventral wall of the pericardium by a ventral mesocardium. Rouviére (’04), on the other hand, while he agrees with Robinson as to the absence of the ventral mesocardium in mammals, gives a different account of the process which leads to the closure of the fore-gut. He describes the formation of the lateral pleuro-pericardial canals, which grow cranially round the cranial end of the. brain-plate and fuse to form a continuous channel. ‘The splanchnopleure forming the caudal wall of the pleuro-pericardial cavity now forms a continuous fold, which Rouviére calls the cardiac fold and which he describes as growing actively backwards as a whole. Griiper (12), in a description of the _ growing processes in the developing chick, asserts that there is consider- able evidence in support of the view that the margin of the fore-gut (umbilical orifice) moves caudally concurrently with the growth of the head fold cranially. Miss Parker (’15), in her studies of the early stages,in the development of Marsupials, summarises her statement by saying “that while the initiation of head-fold formation is in all probability due to the forward growth of the brain-plate, there occurs also an active backward growth of the anterior intestinal portal (umbilical orifice).” Shore (’89) describes that the head fold of the chick embryo results from a growth of the head forwards over the diblastic part of the blasto- derm, and that a “folding off” does not occur, at any rate at first. Recently Watt ('15), in his investigation into two young twin human embryos with 17-19 paired somites, states: “From the atrial canal the ventricle continues on at the left and runs far forward in the pericardial cavity, when it is strongly flexed ventrally and turns caudad as it reaches the median line. It is here attached to the pericardial wall by a short stretch of ventral mesocardium, the only portion of this structure which is still present.” Though a small piece of dorsal mesocardium is depicted in his paper (plate 3, fig. 4, and plate 4, fig. 2), there is nothing to indicate the existence of the ventral mesocardium which Watt describes. STaGcE I. The material for this stage consists of three embryos. Selected sections of each of these specimens have been photographed to show the conditions and relationships of the blood-cells and vascular endothelium. 116 Dr C. C. Wang (a) Ferret Embryo, 1:15 mm. (F.C.Q.2.(Z) )- General Description. The germinal area of this embryo measures 14 mm. No mesodermic somites can be detected, and there is no indication of a heart rudiment. The head fold has not yet begun to develop, and no intra-embryonic blood-vessels can be found in the specimen. There is, however, a shallow neural groove which terminates at the primitive caudal and subsequently dorsal end of the bucco-pharyngeal membrane. The groove broadens out caudally. Beneath the caudal end of the neural plate it becomes the chordal canal. The chordal canal — Mesoderm. ese s) ped 4% Fic. 1,—Showing blood-cells, 500, terminates caudally in a mass of cells which fuse with the ectoderm at the cranial end of the primitive streak. Blood-Cells. Though no intra-embryonic blood-vessels can yet be found in this specimen, extra-embryonically there are solid clusters of blood-cells which are not surrounded by any endothelium (fig. 1). In none of the clusters of blood-cells is it possible to detect that the peripheral layer of the cells resolves into endothelium. The cells forming the clusters are spheroidal in shape, lying between the mesoderm and the entoderm. They are pro- vided with large well-stained nuclei, and are very often adherent to the entodermal cells, which exhibit characters similar to those of the. blood- cells. No lumen can be found in any of the blood clusters. The mesodermal cells spread out in a thin layer to cover the adjacent Development of Blood-vessels and Heart in Ferret Embryos 117 yolk-sac. ‘They are spindle-shaped, and are attached to each other by long protoplasmic processes. Fig. 1 shows a portion of the extra-embryonic area in which a cluster of blood-cells is seen lying free between the mesoderm and the entoderm. The cells forming the mesoderm are, as noted, spindle-shaped, whilst the cells constituting the entoderm are spheroidal and exhibit other characters which are similar to those of the blood-cells. Mitotic division occurring in one of the entodermal cells can be detected in the neighbourhood of the blood clusters (fig. 1). It is proved that this is not a singular occurrence by the fact that a similar phenomenon is again seen in the entoderm of another embryo (Stage IT. (0), figs. 15b, 17a and b). (b) Ferret Embryo, 1°6 mm. (F. 1904, QZ., U2). General Description. This embryo measures 1'6 mm. after it has been cut. There is, as yet, no mesodermic somites. The heart rudiment is absent, and no intra- embryonic blood-vessels can be detected. There is, of course, no head fold. A primitive streak is, however, present, and there is a primitive groove. The notochord is tubular at its caudal end. In parts its ventral wall opens into the yolk-sac. Its caudal extremity is fused with the ectoderm at the cranial end of the primitive streak, as in the previous specimen. Cranial to the primitive streak a faint neural groove is present on the surface of the embryo. It terminates, as in the previous case, at the bucco- pharyngeal membrane cranially. Blood-Cells. In this specimen blood-cells are found abundantly on the wall of the yolk-sac between the mesoderm and the entoderm. The blood-cells are spheroidal in shape, provided with large well-stained nuclei, as in the preceding specimen. They are arranged in solid clusters, without any lumina in them, and are devoid of any endothelial coverings (figs. 2a, 2b, and 3). The majority of the cell clusters are found to be adherent to the entodermal cells, which exhibit characters similar to those of the blood- cells (fig. 2b). The mesodermal cells, covering the adjacent yolk-sac and in the neigh- bourhood of the blood-cells, are arranged in a thin layer which is not in contact with the entoderm. They are spindle-shaped, and are connected with one another by long protoplasmic processes, as previously noted (fig. 2a). Mesoderm. Blood-cells. 1 Entoderm. Fig. 2a.—Showing blood-cells. x 500. Mesoderm, Blood-cells, Entoderm. Fig. 2b.—Showing blood-cells. x 500. Blood-cells, — Entoderni. Fie. 8.—Showing blood-cells, x 500, Development of Blood-vessels and Heart in Ferret Embryos 119 (c) Ferret Embryo, 1:74 mm. (F. 1904, Q.A.A. U1) with 3 Somites. General Description. The primitive streak is well marked, and is notched at the caudal part of its extent by the primitive groove. The mesoderm, covering the caudal part of the embryonic area, is thickened, and indicates the position of the allantoic mesoderm. Cranially the mesoderm of the primitive streak fuses with the caudal end of the chorda. Mesoderm, -cells. Ento- Endo- derm. thelium. Fie. 4.—Showing blood-cells. x 500. There is a broad, shallow neural groove which narrows cranially, and its walls become much thickened in the position which is occupied by the trigeminal ganglion. The groove gradually disappears, and becomes con- _ tinuous cranially with the bucco-pharyngeal membrane. There is no heart rudiment. The two pleuro-pericardial canals are present one on each side of the embryo, but these have not grown across the median plane cranially ; consequently there is no pleuro-pericardial cavity. No intra-embryonic blood-vessels can be detected. Blood-Cells and Endothelium. Extra-embryonically clusters of blood-cells are found scattered over the greater part of the yolk-sac, to which they are often adherent (fig. 4). The ‘ Mesoderm. Blood-cell. ' M Fie. 5.—Showing blood-cells. x 500. Y 120 Dr C. C. Wang characters of the blood-cells, the entoderm and the mesoderm are similar to those already seen in Stage I., (a) and (6) specimens. The mesodermal cells (fig. 5) are, however, more flattened, and are connected with each other with longer protoplasmic processes than those observed in the previous specimens. In addition to the blood-cells, endothelial cells can be detected here and there in the extra-embryonic region lying between the mesoderm and entoderm. These endothelial cells, unlike the blood-cells, are spindle- shaped (fig. 4), and can be traced in some cases to their mesodermal origin. Stace IT. The material for this stage consists of two embryos, one measuring 1:97 mm. in length with 5 somites, and the other 2°3 mm. with 6 somites. (a) Description of the Graphic Reconstruction of the Heart and Cranial Portion of a Ferret Embryo 1:97 mm. in Length with 5 Somites. (F.B.A.A., G.A.) This is the youngest specimen of the series of ferret embryos selected for the purpose of reconstruction in this investigation. Its total length measures 1°97 mm. after preparation and embedding. It may be mentioned that the sections, each of which is 10, in thickness, are perfect, and that the histological condition is excellent. General Description. No plastic reconstruction of the embryo was made, for it appeared that a graphic reconstruction of the heart would be sufficient for the purpose in hand. The neural groove and the primitive streak are both present. The neural tube is deepest at the brain region, where it shows thickenings which correspond to the positions of the trigeminal and otic ganglia. The cranial end of the neural groove gradually shallows until it disappears at the primitive caudal end of the bucco-pharyngeal membrane. Vascular System. In this specimen the heart is represented merely by a transverse blood channel which lies across the median plane and unites the cranial ends of the two vitelline veins (figs. 6a and b). It is bounded caudally by the bucco-pharyngeal membrane, and cranially by the pleuro-pericardial cavity (fig. 6a). The cranio-caudal diameter of the heart rudiment is 20y, its breadth, 1204. The rudiment of the pleuro-pericardial cavity is present. It is that portion of the ecelomie space which crosses the median plane of — Development of Blood-vessels and Heart in Ferret Embryos ~ ER ga eRAOe ISAM TNS PA ae Fic. 6a.—Graphic reconstruction. x 200. P., pericardium ; H., heart rudiment (primary union); P.p.c., pleuro-pericardial canal ; V.v., vitelline vein; D.Ao., dorsal aorte; Pl., plexus; Buc., bucco-pharyngeal membrane. Mesoderm. Ectoderm. Pleuro-peri- cardial canal. ™ Entoderm. x 500. Heart rudiment. Fic. 6b.—Transverse section through primary heart tube. 121 122 Dr C. C. Wang the embryonic area cranial to the rudiment. of the heart and is closed eranially and caudally, and on each side it is connected with the pleuro- pericardial canal (figs. 6a and 7). Together with the pleuro-pericardial Ectoderm, euro-pericardial cavity. Mesoderm, Pleuro-pericardial cavity. Entoderm. Fic. 7.—Transverse section through pericardium. x 500, canals it forms an inverted U-shaped canal (fig. 6a), which lies dorso-cranial to the vitelline veins and the heart rudiment (fig. 8). Each vitelline vein is placed ventral to the corresponding pleuro- Mesoderm. Ectoderm., Pleuro-pericardial Entoderm, canal, Heart rudiment, Fic, 8.—Transverse section through pericardium and heart. x 500, pericardial canal and lateral to the dorsal aorta of the same side (figs. 6a and 9). The two veins converge cranially, and each terminates in the corresponding end of the heart rudiment cranial to the bucco-pharyngeal membrane (fig. 6a). There are two rudiments of the dorsal aorte (fig. 6a). They run caudo- Development of Blood-vessels and Heart in Ferret Embryos 123 cranially one on each side of the medullary groove. They are still more or less plexiform in character, and they terminate blindly at their cranial extremities some distance caudal to the heart rudiment. Communications between the dorsal aorte and the corresponding vitelline veins are described by Bremer (’12) in a 3:4 mm. rabbit embryo. Such communications (fig. 6a) can be traced in the caudal portion of the ferret embryo at this stage, which, so far as its general development is concerned, is considerably younger than Bremer’s embryo. Caudally the _ dorsal aorta breaks up into a plexus spreading over the wall of the yolk-sac. There is, of course, no ventral mesocardium, and the head fold and fore- gut are not yet developed. As far as the pericardial cavity and the pleuro- ’ pericardial canals are concerned, this specimen does not differ, to any great Pleuro-pericardial canal, Ectoderm. Mesoderm. y =e ¥ x a > are Rd ee raga > ° wig See, 7 eS Entoderm. Dorsal aorta. ine vein. : Fig. 9.—Transverse section through embryo, _ x 200, extent from what has been described in Dasywrus viverrinus (7°5 mm. vesicle) by Miss Parker ('15). The heart of the ferret embryo is, however, in a more advanced stage of development than the 7°5 mm. Dasyurus, in which the heart rudiment is represented merely by some scattered angioblast cells and strands of cells, and in which the vitelline veins terminate eranially at the level of the caudal limit of the trigeminal rudiment. In the ferret specimen under consideration the walls of the vessels of the vitelline plexus are in direct continuity with the adjacent splanchnic mesoderm (figs. 10a, b, and c), a fact which is of interest in association with the mesoderm origin of the walls of the vessels. (6) Description of a Ferret Embryo 2°3 mm. with 6 Pairs of Somites. (F. 1904, B.G.a.) General Deseruption. The primitive streak is present. Its cranial extremity is continuous with the notochord, which shows indications of the notochordal canal, and the notochord has begun to dovetail with the entoderm. More cranially 124 Dr C. C. Wang Vitelline plexus, & Eetoderm, Ceelom. > Mesoderm. Vitelline plexus, wl Entoderm. toe all Fic, 10b.—Transverse section through embryo, showing endothelium. x 375, Vitelline plexus, Fic, 10¢,—Transverse section through embryo, showing endothelium, 375. Development of Blood-vessels and Heart in Ferret Embryos 125 the notochord is entirely fused with the entoderm, and for some ten sections cranial to the primitive streak it can scarcely be distinguished except by the height of the cells from the entodermal cells. The neural groove extends caudally to the primitive streak. It is wide and shallow at the caudal end, but deepens and narrows as it passes cranialward between the laterally placed mesodermal somites, and cranial to the somatic region the walls of the neural groove show thickenings which correspond in positions to the auditory and trigeminal areas. The neural groove then gradually shallows until it disappears at the caudal end of the bucco-pharyngeal membrane. Vascular System. This embryo exhibits several features which were not present in the preceding specimens. No reconstruction was made at the time, but a Ectoderm, Mesoderm. " Pleuro-pericardial . ; ae Entoderm. canal, Heart rudiment. Fic. 1la.—Transverse section through pericardium and heart. x 200. Ectoderm. Mesoderm. x Pleuro-peri- cardial canal. Sop a “ ines a ap). pikodaem: ies sallanak Fic. 11b.—Transverse section through pericardium and heart. x 500. study of the sections shows that the embryo as regards general development is slightly in advance of the previous specimen. The heart and pericardium - 126 Dr C. C. Wang (fig. lla), have practically not changed, the former lying ventral to the latter. Figs. lla and b illustrate the union of the two heart tubes Ectoderm. Mesoderm. Entoderm. Pericardium. Fic. 12a.—Transverse section through pericardium. x 200. - Ectoderm. Mesoderm. Fic. 126.—Transverse section through pericardium. x 500. Ectoderm. Mesoderm. Pleuro-pericardial Entoderm, Pleuro-pericardial canal, canal, Fic, 12c.—'lransverse section through pleuro-pericardial cavity. 500. across the median plane, and immediately cranial to this the pericardium is seen stretching transversely to communicate on either side with the Development of Blood-vessels and Heart in Ferret Embryos Ectoderm. ' Mesoderm. Entoderm. i Dorsal aorta. Vitelline vein. Plexus. Fic. 13.—Transverse section through embryo. x 200. Mesoderm. Pe ah SE tl Intra-embryonic vascular endothelium. Fic, 14a.—Transverse section showing endothelium, x 500. 3 weet Pabae oleae oie ei een Entoderm. Mesoderm. ~ Entoderm. Intra-embryonic vascular endothelium. & Fie. 14b,—Transverse section showing endothelium. x 500. 127 128 Dr C. C. Wang pleuro-pericardial canal (figs. 12a, b, and c). In the 1:97 mm. ferret embryo (Stage II. (a)), it has been noted that there is a communication between the Mesoderm. oe .. eS A 75 tas os pees eS Entoderm. Blood-cells, Fic. 15a.—Showing blood-cells. x 500. pel” ee tion br Sore ae } Entoderm, Blood-cells. Fic. 15b,—Showing mitosis in entodermal cells to form blood-cells. x 500. vitelline vein and the dorsal aorta. Such a communication is seen also in this specimen (fig. 13). Some of the intra-embryonic vascular endothelium can be traced to its origin from the mesoderm (figs. 14a and b), Blood-cells. e - Entoderm. == Endothelium. Fic. 16.—Showing endothelium and blood-cells. x 500. Mesoderm. Blood-cell. . on. . Badkoderan: Mitosis, Fie. 17a.—Showing mitosis in entoderm to form blood-cells, x 500. Blood.-cells. : ~ Mesoderm. Blood-cell. Mitosis.» » Entoderm. N . Fig. 17).—Showing mitosis in entoderm to form blood-cells. x 500, * VOL. LU. (THIRD SER. VOL. XIII.)—OCT. 1917. ¥ 130 Dr C. C. Wang In this specimen it is possible to detect that some of the intra-- embryonic blood-cells are not surrounded by endothelium (figs. 15a and 5), whilst others are partially engulfed by flattened vascular endothelium (fig. 16). Figs. 15b, 17a and 5b, taken from sections of the caudal end of this specimen, show distinctly mitotic divisions in the entodermal cells in the neighbourhood of some blood-cells. . There is no ventral mesocardium, and the fore-gut has not yet begun to develop. Stace III. The material for this stage consists of one embryo which is 2°3 mm. in length and with 9 somites. Description of the Graphic Reconstruction of the Heart and the Cranial Portion of a Ferret Embryo 2°3 mm. in Length with 9 Somites. (F. Ap. 16/28/08.) General Description. The amnion is closed caudally. The allantoic diverticulum and the allantoic mesoderm are both present. The neural groove extends to the caudal end of the embryo, but terminates, however, at some distance cranial to the tailamnion-fold. It narrows and deepens as it passes cranially. The rudiment of the otie ganglion is distinct, and the rudiment of the trigeminal ganglion is likewise recognisable. The head fold has begun to form, and a portion of the fore-gut is also defined. There is no indication of the primary optic vesicle. Vascular System. The heart of this embryo has not yet. been reconstructed in wax, but the graphic reconstruction (fig. 18) which has been made shows that, in length, this specimen is identical with the Stage II. (b) embryo, but in its general development it is decidedly in a more advanced stage than the preceding one, as the central part of the transverse rudiment of the heart seen in Stage II. (fig. 6a) has, in this specimen, begun to break up, for two non-vascular loculi have divided it incompletely into a cranial and a caudal portion (figs. 18, 19a-e). Apparently the division of the central part of the rudimentary transverse heart proceeds still further as develop- ment goes on, until it is completely separated into right and left halves, for in Stage IV. (fig. 27) the heart rudiment is represented by two separate longitudinal endothelial tubes which lie side by side and are in contact in their middle third. Cranially two vessels, one on each side of the median plane, run cranial- 131 Development of Blood-vessels and Heart in Ferret Embryos (‘olpeULayos-tutag) “QOL x "SSRLU-[[90 9JBIpOUtI9zUT ee [etprvoltsad-orne[g *Teueo [VIpreorsod -oina[d jo a3pa yng t *WnIpIwoLieg — “ulnIpieotied YSNoAY. Waas ULOA DUTTIOIIA ~ *;BuRd [eIpavolled-oine[g -—— --—--+—------=—-=- ~ *[euRd [VIpceoLtad-o.ne|d jo dadel [eqotred Jo aBpa yng f ‘dn Suryeoiq quawrpns 4180 ‘aqyuOg a eee ‘Soqn} Jlvoy [eLoyR, OMY (Loy 07 dn Faryeerq aqnyz yrvoy Arvurud Futmoyg—'gl ‘vig ‘SSBUI-[[99 OBI pouLIE4,UT *(eueo jetprvolied-oimeg “HOA OUTLTOFLA | “[eueo [eIpsvolied -ornetd jo a8pe yng "e108 fesioq = “agWIOg eee oo “unIpaBoleg “uunIpreottied ([3NOA1YA UdOS UDA 9UTT[9ITA { _ SRS Toes tree ere —"yeuro [erpavotied-ouns[g *peueo [erpavolsed-oimotd Jo Jake] peqotied jo odpea yng *soyoie O1910V 132 Dr C. C. Wang Medulla. Dorsal Ectoderm Fore-gut. aorta. H Entoderm. Parteaidieaa: Heart rudiment Ectoderm. | united across, | Vitelline Pleuro- vein. pericardial canal. Fic. 19a.— x 100. Vitelline Heart rudiment vein, | broken up. Pleuro-pericardial canal. Fie. 19b,— x 100. VWitellite- seas 27 Heart rudiment eokkdn up. Pleuro-pericardial canal, Fic. 19c,— x 100. od Sis Vitelline vein. Heart rudiment broken up. Pleuro-pericardial canal, Fira, 19d,— x 100, Development of Blood-vessels and Heart in Ferret Embryos 133 ward from the heart rudiment. They arch round the cranial end of the fore-gut and form the first cephalo-aortic arches, which terminate dorsally in the corresponding dorsal aorte (figs. 18 and 20). Caudally the heart rudiment receives the two vitelline veins. It is obvious that the fore-gut has begun to develop pari passw with the head fold. The pericardial cavity is much wider and longer than it is in the pre- ceding specimen, measuring 870. in the transverse diameter and 140, in the antero-posterior direction. The two dorsal aorte are well developed and run parallel to each other, one on each side of the median plane of the embryo, not far from the medullary groove (figs. 18, 19a-e). Here again there is nothing to indicate the presence of a ventral mesocardium. This stage of development, as far as the heart is concerned, appears to fall in between the Stage II. Dasywrus viverrinus (8°5 mm.) and the Heart rudiment — Vitelline vein. united across. Pleuro-pericardial canal. Fie. 19¢,— x 100. Stage III. Perameles nasuta (75 mm., 11 somites) of Miss Parker (’15). In the Dasyurus the endothelial tubes have come actually in contact with each other at their extreme cranial ends, and presumably have united across the median plane of the embryo. It should be noted that the significance of this connection between the two vitelline veins across the median plane was not dwelt upon, and was considered only as being remarkable by Miss Parker, who states also “that lateral and caudal to the median union, each endothelial tube gives rise to the first aortic arch, which follows the antero-lateral margin of the gut almost to the median plane, and there becomes continuous with the corresponding dorsal aorta, the two aorte being well developed at this stage.” It is quite possible that what has been taken for the first aortic arch in Miss Parker’s Dasyurus specimen, as in the case of Eternod’s human embryo (’95, 99), may, after all, prove to be the plexus which lies between the dorsal aorta and the vitelline vein, and which has been described by Bremer (’12) to be present in the rabbit embryo of 5 somites. If this is 134 Dr C. C. Wang the case, the Stage I]. Dasyurus may be regarded as being similar with the Stage II. (a) ferret embryo, in so far as the development of the heart, the vitelline veins, and the dorsal aorte is concerned. Dorsal aorta. Fore-gut. Entoderm. Medulla. mee Pleuro-pericardial Heart rudiment ‘ united across. Fie. 19a’.—Showing primary heart tube. x500. (Enlarged from fig. 19a.) Fore-gut. Medulla. % : Ectoderm. ee’ - ‘ Heart rudiment. Ejeary pees Heart rudiment. canal, Fic, 196’,—Showing primary heart tube breaking up. 500, (Enlarged from fig. 19d.) In the Perameles (7°5 mm.), the two heart tubes lie separate from each other ventral to the closed fore-gut. At the level of the umbilical orifice they diverge and lie on each side of the open gut in the dorso-medial wall Development of Blood-vessels and Heart in Ferret Embryos 135 of the pericardium. From the cranial extremity of each endothelial tube there arise two vessels, one of which runs cranially and laterally towards Fore-gut. Entoderm, Medulla. Heart rudiment. Heart Pleuro-peri- Heart Ectoderm. rudiment. cardial canal. rudiment. Fic. 19¢’.—Showing primary heart tube breaking up. 500. (Enlarged from fig. 19c.) Fore-gut. Entoderm, ~~ Heart rudiment. rudiment. Pleuro-peri- Ectoderm. cardial canal. Heart rudiment. Fie, 19d’.—Showing primary heart tube breaking up. 500. (Enlarged from fig. 19d.) the lateral margin of the gut and then parallel with this margin. It loops round the cranial end of the gut, joins the dorsal aorta, and thus consti- 136 Development of Blood-vessels and Heart in Ferret Embryos tutes the first aortic arch. The other vessel is small, and runs caudally and laterally, lateral to and almost parallel with the heart tube. This, Fore-gut. Entoderm. 4 aN ae ——— —— — — --- Umbilical orifice. Arrow in pleuro-pericardial canal. Mesodermic tissue. Left sinus venosus. — ~— - Cut surface of myocardium. Right endothelial tube partially exposed. Left ventricle. - Fie, 30.—Cuudal view of plastic reconstruction of ferret embryo 2°5 mm, x 100. it pursues a course caudally along the dorsal wall of the fore-gut as the dorsal aorta (figs. 21 and 25). It is a relatively wide vessel, its calibre throughout being distinctly greater than that of the bulbus cordis. Immediately dorsal to the dorsal aorta there is situated on each side of the embryo a series of apparently isolated sections of a minute blood- vessel (figs. 28a-d). These capillaries lie close against the medullary tube. This vessel appears to be the vena capitis medialis of Grosser ('95), which Miss Parker ('15) also has found to be present in her Stage III. Perameles nasuta 75 mm. Intersegmental offshoots from the dorsal aorte: in the region of the caudal somites are described by Miss Parker in Perameles nasuta, but no such offshoots from the dorsal aortee were present in the ferret specimen under consideration, In the caudal region of the embryo, the Development of Blood-vessels and Heart in Ferret Embryos. 149 two dorsal aortz become continuous with the vitelline arteries, which spread themselves out in a plexiform manner on the wall of the yolk- sac (fig. 24). This stage of development of the dorsal aorta in the ferret agrees, in some ways, with the 13-mm. human embryo described by Eternod (95, ’99). The vitelline vein (fig. 27), which opens into the caudal end of the sinus venosus, so far as it lies on the embryonic region, runs at first transversely towards the median plane in the substance of the septum transversum immediately cranial to the umbilical orifice. As it approaches the median plane it changes its course suddenly, making a sharp bend upon itself cranially, and terminates, as has already been indicated, in the dorsal extremity of the caudal part of the corresponding sinus venosus. It is applied so closely to the margin of the umbilical orifice that its caudal border causes a distinct bulging of the boundary of the orifice (fig. 27). For the sake of comparison, and for the purpose of bringing out the chief differences, as far as the development of the heart is concerned, between this specimen and that of Yeates (15), the description of the endothelial tubes of the latter given by Yeates might be quoted: “As the endothelial tubes course through the primitive cavity of the heart (the primitive cavity of the heart means the muscular tube—the myoepi- cardium) they are separated by a variable but distinct interval from the myoepicardium The myoepicardial tube has previously been stated to _ present constrictions at each extremity, at the sinuatrial junction and at the atrio-ventricular canal. The endothelial tubes present corresponding constrictions. In the region of these constrictions the endothelial tubes are relatively close to the primitive myoepicardium, whilst they gradually recede from the heart wall as the middle of each of the three primitive cavities of the muscular heart is approached: In other words, the muscular cavity is more expanded, and the endothelial is more tubular between the constrictions. The endothelial tubes are in contact medianly in the cranial two-thirds of the atrium, in the ventricle and in the region of the bulbus. In the atrio-ventricular canal they are not only in contact but have fused and are partially absorbed, so that their cavities communicate across the median plane. On the other hand, in the region of the sinus venosus and of the truncus, the tubes are free and separate from each other. The portions. of the tubes which are in contact within the atrium are con- nected by delicate endothelial strands with the inner aspect of the ventral wall of the atrium along the crest of the irregular ridge, which has been spoken of as the remains of the primitive cardiac septum. The ventri- cular limbs of the endothelial tubes are subdivided into ventricle, bulbus 150 Dr C. C. Wang cordis, and truncus arteriosus by two faintly marked constrictions. The endothelial heart, therefore, consists of not only sinus venosus, atrium, and ventricle, as in the muscular heart, -but also of bulbus cordis and truncus arteriosus.” . It will be noticed, then, that Yeates’ specimen is decidedly in a more advanced stage of development, since the two endothelial tubes have partially fused in the region of the atrio-ventricular canal, much in the same way as the Stage V. specimen, to be immediately described. Subdivisions of both of the endothelial tubes into sinus venosus, atrium, ventricle, bulbus cordis, and truncus arteriosus have been observed also by Yeates in his specimen; but, unlike those of the Stage IV. 25-mm. ferret embryo, these subdivisions correspond in the main in positions to those exhibited in the muscular (myoepicardial) wall of the heart tube. One point which is clear is that the endothelial tubes differentiate into their various subdivisions more completely than the muscular tube. It should be pointed out that the second aortic arch in Yeates’ specimen has already made its appearance. It is to be noted that, as far as the external appearance of the heart is concerned, the ferret embryo at this stage of development shows certain prominent features which are in many respects identical with those of the dog of a similar stage of development, and which have been investigated by Bonnet (01). In his paper it has not been possible to find a com- prehensive -description of the development of the heart of the dog. In fig. vi. (Anat. Hefte, 1901, Bd. xvi.) Bonnet depicted a dog embryo 57 mm. long with 10 somites. There the heart is represented by two endothelial tubes which are distinctly separated from one another and are both bent towards the right side in the ventricular region; each endothelial tube has, for its caudal continuity, the corresponding vitelline vein, which runs latero-medially and at the same time cranially towards the sinus venosus There seems to be no distinctive demarcation between the sinus venosus _ and the atrium, but between the atrium and the ventricle there is a constriction both on the lateral side and the medial side. The ventricle looks very much dilated in the figure, so much so that, when viewed from its ventral aspect, the whole heart appears to be formed by the distended ventricle, with the bulbus cordis and atrium attached to it cranially and caudally respectively as mere appendages, The bulbus cordis is clearly constricted off from the ventricle, and its calibre is barely one-third of that of the ventricle. Development of Blood-vessels and Heart in Ferret Embryos 151 STaGE V. The material for this stage consists of one embryo which measures 3°14 mm. in length. Description of the Graphic Reconstruction of the Heart of a Ferret Embryo 3°14 mm. in Length with 13-14 Somites. (F. 15d. (e).) The general development of this embryo is so similar to that of the Stage IV. specimen as to merit no separate description. The specimen is slightly older than the Stage IV. ferret embryo. , As in the case of the other embryos described, the state of preservation of this embryo is perfect. In length it undoubtedly exceeds the Stage IV. specimen, but when other measurements are taken, the fact is revealed that the embryo in question is relatively a small one. The pericardial cavity measures 4004 in its cranio-caudal diameter and 840 from side to side. No plastic reconstruction of this embryo has been made. The graphic reconstruction of the heart, however, shows that the two endothelial tubes have united in part of their extent (figs. 31 and 32). The fused portion, extending through some sixteen sections of 10u each, appears to be the ventricular part and lies more to the right side (figs. 31 and 32). Cranially the fused ventricle divides into two vessels, each of which represents the bulbus cordis (figs. 31 and 32) and becomes continuous with its correspond- ing dorsal aorta by looping round the cranial end of the pharynx, thus constituting the first cephalo-aortic arch (fig. 31). The paired atria run into the fused ventricle cranially, and each receives its sinus venosus caudally (fig. 34). The vitelline veins are very much in the same stage of development as those observed in Stage IV. (compare figs. 27 and 31). The right vein pursues a more transverse course latero-medially, and terminates at its corresponding sinus venosus. In this, as in the preceding specimens, there is no trace of a ventral mesocardium. The dorsal mesocardium is, however, present in this specimen (figs. 32, 33, and 34). It may be observed that at this stage of development the ferret heart, though resembling very closely the heart of the Stage V. Perameles obesula (19, viii., 08) described by Miss Parker, yet differs from it in many respects. In both cases the two endothelial tubes have partly fused. In the ferret the fusion occurs in the ventricular region (fig. 31). In the Perameles obesula, according to Miss Parker, the bulbus is the only portion of the heart in which the endothelial tubes have actually fused at this stage. 152 Dr C. C. Wang AoA. eS ca RE Fie. 31.—Graphic reconstruction of the heart of ferret embryo 3°14 mm., ventral view. x 100 Ao.A., aortic arch; D.Ao., dorsal aorta; B., bulbus cordis ; A., atrium; S.v., sinus venosus ; 5 Pe pericardial reflection ; Viv. , vitelline vein ; Vv. , ventricles (fused). Medullary tube. Dorsal aorta. — —— Dorsal mesocardium, Fused ventricular portion of the two endothelial tubes. Pericardial cavity. Fig, 32,——Transverse section phen fused ventricular region of 3°14-mm. ferret embryo. x 100. Development of Blood-vessels and Heart in Ferret Embryos 153 Asymmetry of the two endothelial tubes has-also been. noted in Parameles embryos. Miss Parker observes that whilst the left heart tube Medullary tube. t Dorsal aorta, Fore-gut. Dorsal mesocardium. Left bulbus cordis. Right bulbus cordis, Pericardial cavity. Fic. 33.—Transverse section through bulbus cordis. x 100. Dorsal mesocardium. Dorsal aorta, Fore-gut. \ Medullary tube. / > Right vitelline vein, Caudal reflection of pericardium Left sinus at the level of septum transversum. venosus. Fie. 34.—Transverse section through sinus venosus. x 100. is practically straight, the right tube shows well-marked curvatures. In the ferret it has been found that the two tubes, prior to fusion, appear to have been shifted as a whole towards the right side (fig. 27), and that they remain in this position even after partial fusion has taken place (fig. 31). VOL. Lil. (THIRD SER. VOL. XIII.)—JAN. 1918. > 154 Dr C. C. Wang Two pairs of aortic arches are found arising from the fused bulbus in the Perameles embryo, but only one pair of vessels can be recognised in the ferret embryo at this stage, and these come from the yet unfused bulbus (fig. 31). In both instances the ventricular portion is the most dependent part of the endothelial tubes. Stage VI. Macropus ruficollis 5: 2 mm. of Miss Parker is distinctly older and is ina more advanced stage of development than the Stage V. ferret -embryo, but it is interesting to note that in Macropus the right and left heart tubes are fused except in the region of the simus venosus, where they remain separate. Three pairs of aortic arches are described. The fused heart has already begun to acquire the S-shaped curvature, so that the bulbus arteriosus lies dorsal to the cephalic extremity of the ventricle. ‘The bulbus arteriosus is continued into a short median ventral aorta which bifurcates to form the first pair of aortic arches. The second and third pairs of aortic arches arise from the median ventral aorta immediately caudal to its bifurcation. The atrial limb of the S is carried into position dorsal to the ventricle. On the other hand, in the ferret of 3:14 mm. the two heart tubes are only fused in the middle parts of their extents. | DISCUSSION. Origin of Blood Cells and Vessels. The transformation of the blood-cells into red and white corpuscles lies outside the scope of the present communication, in which the develop- mental relationships which the blood-cells have in common with the vascular endothelium will alone be considered. In the paragraph dealing with the development of, the extra-embryonic vascular rudiments in mammals, it has been pointed out that in embryos of the higher vertebrates the earliest vascular rudiments have invariably been described by most authors as appearing, at first, in the form of localised cell cords (“angioblast” of His). Thus far all observations on this point incline to support the work of His (’00). Further, in the literature of the past twenty-five years there are numerous descriptions and illustrations of the origin of blood-cells from the vascular linings. In 1892 Schmidt described the transformation of individual endothelial cells into white and red blood corpuscles. In support of Schmidt’s view, Maximow (’09) states that the endothelial cells and blood-cells are closely related and arise from a common stem-cell in the blood islands, and may continue to do so from sueh a cell during later development. The most damaging evidence against Maximow’s view is to be found Development of Blood-vessels and Heart in Ferret Embryos 155 in the recent work of Stockard (’15), who, after having conducted a series of experiments on Fundulus, comes to the conclusion that endothelial lining of vessels is utterly incapable of giving rise to any form of blood- cells, and that vascular endothelium arises in loco in many parts of the embryonic body in which blood-cell rudiments are not present. If Stockard’s view is correct, it necessarily follows that, even if the two groups of cells have a common origin, they are not interchangeable nor can one replace the other: nevertheless it is the common opinion, based upon the results of many investigators, that the angioblast cells produce both blood-cells and endothelium, but there is no agreement as to whether the angioblast cells are derived from mesoderm or from entoderm. Nearly all investigators in this field of work have assumed that blood cells and vessels have a common origin which some attribute to the mesoderm and others to the entoderm. The facts now to be recorded do not support the view of a common angioblastic origin of both endothelium and _ blood-cells, and in this communication the term angioblast will be restricted to. the progenitors of the blood-cells. | It has been pointed out that in the ferret embryos, Stake I. (a) and Stage II. (6) (figs. 1, 15b, 17a and 5) it is possible to identify mitotic division in the entodermal cells in the neighbourhood of angioblastic clusters, and there is no evidence to show that, in the ferret, endothelial cells are capable of giving rise to blood-cells. In figs. 2a, b, 3, 4, 5,sand 15a it is to be observed that angioblast cells are in abundance on the yolk-sac. These are frequently adherent to the entodermal cells, which, if not in direct protoplasmic continuity with the blood-cells, are, in many cases, in close contact with them. It is to be further noted that where the apposition of these cells is intimate, it is impossible to distinguish the angioblast cells from many of the entodermal cells, the nuclei, size, and shape of the two kinds of cells having a close resemblance. On the other. hand, a great dissimilarity exists between the blood-cells and _ the neighbouring mesodermal cells, for in the former the cells are, without exception, spheroidal in shape, their nuclei are large, staining more deeply, and the protoplasm is comparatively small in amount; whilst in the latter, the cells are usually spindle-shaped, their nuclei have mostly differentiated and taken on a lighter stain (figs. 3, 4,5, and 15a). Further- more, in the ferret embryos angioblast cells have been demonstrated in _ regions of the yolk-sac where invasion of the mesoderm has not yet taken place. The above facts appear to indicate that, in the ferret at least, if _ not in all the other mammals, the origin of angioblast cells from the entoderm is highly probable. 156 Dr C. C. Wang With regard to the yenetic origin of the vascular endothelium, numerous investigators have recorded wandering mesenchymal cells upon the yolk- sac. Stockard (15) claims that in Fundulus these wandering mesenchymal cells ultimately give rise to four different kinds of cells—the endothelial cells, the black chromatophores, the brown chromatophores, and the blood- cells. Another current view is that after the so-called “angioblast” has made its appearance, the vascular endothelium arises from the cells of the blood islands by a rearrangement of the peripheral cells of the blood islands to form lining endothelium and the central ones to remain as blood-cells, and that further extension of the endothelium is brought about by buddings of the endothelium which appear, at first, as solid cords but later become hollow. There is so far no evidence to show that the peripheral cells of the angioblast group in the ferret are capable of being transformed into endothelial cells. It is to be noted that in the ferret, endothelial cells, whether in the form of solid cords or grouped together with a lumen, are invariably spindle-shaped from the very beginning. Ziegler (’87), as will be remembered, maintains that the system of blood-vessels and that of the lymphatic vessels are produced from the remnants of the blastoccel which remain behind as vessels, lacunz, or interstices. Felix (’97), however, inclines to the belief that the circulatory system is, from a developmental point of view, closely related with the ccelom. In connection with this question Stockard (15) states: “The vessels arising from independent mesenchymal cells in the space of the blastoccel in the teleost yolk-sac entirely overthrow any notion that vessels arise ontogenetically as portions of the ccelomic epithelium. The vascular lumen is originally continuous with the primary body cavity, the segmenta- tion cavity, and never with the secondary body cavity or ccelomic cavity.” It is clear that these authors agree, at least, that the origin of the endothelium is from the mesoderm. In the ferret it is possible to demon- strate that endothelial cells take their origin from the splanchnic layer of the mesoderm. In some of the sections of the ferret embryo of Stage II. (a) (figs. 10a-d) there are indications to support the view of Felix (97) that portions of the ecelomie space surrounded by mesoderm may be cut off to form vascular endothelium and to lie between the mesoderm and entoderm. The opinion that endothelium develops independently of blood-cells is furthered strengthened by the observations of Stockard (15), who finds that vascular endothelium arises im loco in many parts of the embryonic body of the Fundulus, in which blood-cell rudiments are not present, and that independent blood islands, having no connection with the intermediate Development of Blood-vessels and Heart in Ferret Embryos 157 _cell-mass, are found on the yolk-sac, and even in extremely young embryos blood islands may appear on the ventral yolk surface at a great distance away from the intermediate cell-mass. He maintains, also, that early blood islands are invariably destitute of endothelial walls. Though favour- ing the view that vascular endothelium and blood-cells are independent of each other in their mode of development, Stockard firmly bélieves that both types of cells are derived from wandering mesenchymal cells. This author summarises his statement by saying: “The differences among the four types (endothelial cells, black chromatophores, brown chromatophores, and blood-cells) produced are from the standpoint of our present knowledge in all probability due to the potential differences among the apparently similar mesenchymal cells from which they arose. The four types includ- ing endothelial cells and erythrocytes we must consider from an embryo- logical standpoint as arising from different mesenchymal anlagen.” It will be observed then that Stockard, whilst admitting, on the one hand, (a) that endothelial cells are quite different from blood-cells in shape, in position, and in the period of migration; (6) that the former develop independently of the latter; and (c) that blood-cells when first formed are devoid of endothelial surroundings, claims, on the other hand, that both types of cells have a common parent trunk—the wandering mesenchymal cells. In the Fundulus this is perhaps true, but the evidence produced by the ferret shows the conditions are not the same. In the ferret there are indications to show that vascular endothelium is mesodermic in origin and that blood-cells are, at all events in the first instance, derived solely, by proliferation, from entodermal cells. Such being the case, it is obvious that the sources of origin of the blood-cells and vascular endothelium are distinct, and that these two different vascular rudiments cannot be considered to have a common origin. If the biphyietic origin of blood-cells and vascular endothelium is to be accepted, two more points still remain to be solved, namely, how, when, and where the first blood-cells enter the circulation. This has been variously described not only in embryos of different species but probably even among embryos of the same species. Ziegler thinks, however, that just beyond the lateral plates in the plasma-filled spaces of the yolk-sae which lie between the periblast and ectoderm, the first blood-cells project into the circulation. Stockard describes that in Fundulus embryos the earliest blood-cell formation occurs in the yolk-sac blood islands. The cells in these islands continue to divide until they become surrounded by endo- thelium. As to how these blood-cells are’ provided with endothelial covering, Stockard makes the following statement: “A growing vascular tip may be observed at certain stages to come in contact with a group of 158 Dr C. C. Wang erythroblasts, or actually a blood island unsurrounded by vascular endo- thelium. The tip of the vessel seems to disorganise to some extent, and its cellular elements slowly surround the group of corpuscles which are later taken into the cir Sulanes as the current becomes established in the includ- ing vessel.” Unfortunately the ferret embryos, at present worked? upon, provide no definite evidence on this point, but it is quite clear that angioblast cells are formed outside the embryonic area, and that blood-vessels are formed inside the embryonic area, and are at first. devoid of blood corpuscles. Moreover, it has been pointed out that there is no evidence in the very | young embryos dealt with that any blood corpuscles are formed by division of or budding from the endothelial walls of the independently formed blood - vessels. It would appear therefore that the earliest -blood - cells probably enter the embryo from the pepiery. Intra-Embryonic Blood- Vessels. In the review of the literature on the subject of the origin of the intra- embryonic blood-vessels, it has already been indicated that the problem has proved to be one of the most difficult in the development of the vertebrate animals. Numerous’ conflicting views have been advanced regarding the precise mode of the origin of the intra-embryonic blood- vessels. Thus His (00) and Hertwig (92) have associated themselves with the theory that the early blood-vessels in the body of the embryo are formed by a budding or ingrowth of the endothelial lining of the vessels from the extra-embryonic vascular area, and Sobotta (02) supports the belief that vessels in the embryo develop in situ, and those on the wall of the yolk-sac are secondary as a result of an outgrowth from the intra- embryonic blood-vessels. Rabl (’86), on the other hand, pointed out the possibility of the vessels of, at least, the cranial region, if not the whole vascular system of the embryo, having been formed by the extension of the paired heart rudiments when these are developed. Recently Riickert and Mollier (06) maintain that the embryonic vascular system, or at least a part of it, arises in situ from the mesoderm of the embryo. Felix (97) states that in birds, the aorta and certain veniplexuses all arise i Loco. The question is therefore still open, and each view invites further criticisms or support. In birds the caudal portion of the dorsal aorta is, according to Vialleton (92), His (00), and Evans (’09), formed from the medial margin of the vitelline plexus which has grown into the embryo in the manner already indicated in the beginning of this communication, In the ferret, Stage II. (a), 1:97 mm., it has been found that the caudal portion of the dorsal aorta has established its communication with the Development of Blood-vessels and Heart in Ferret Embryos 159 vitelline plexus. Precisely how the caudal end of the dorsal aorta in the ferret is developed, no definite statement can be made, but .as far as evidence goes, it is probable that this part of the dorsal aorta arises much in the same way as described by Vialleton (92) and His (00). For the development of the cranial portion of the dorsal aorta, on the other hand, various opposite views are held. His (’00) attributes it to the result of a further growth of the same extra-embryonic vitelline plexus which forms the caudal part of the aorta, but which is reduced to a capil- lary chain growing cranially, eventually turning ventrally over the blind end of the fore-gut and fusing with the cranial portion of the heart tubes. In support of this theory Lewis (’04) affirms that all intra-embryonic blood- vessels of rabbits are apparently derived as offshoots from the extra- embryonic network of vessels in the splanchnopleure of the yolk-sac, the vitelline plexus ending medially in the embryo in the form of two vessels —the dorsal aortz. Quite recently Bremer ('12) states that in the rabbit embryo of 5 somites, the dorsal aorta, the first. aortic arch, the conus arteriosus, and the lateral heart are all parts of an original network of angio- blastic cords derived from the extra-embryonic plexus of blood-vessels. Riickert and Mollier (06) maintain that the cranial portion of the aorta is developed in sitw from the mesodermic cells of the lateral plate of the mesoderm of the cranial region of the embryo. In support of the autoch- thyonic origin of the cranial portion of the dorsal aorta, the work of Huntington (10, 14) and M‘Clure (’10, 12) may be cited. Recently this view is further strengthened by the results of the experiments of Miller and M‘Whorter (14) on the origin of blood-vessels in the chick embryo. Further support is to be found in the more recent experimental evidence presented by Reagen (15), which shows the origin im loco of vessels in isolated parts of chick embryos, and by Stockard (15), which claims beyond doubt that in Fundulus embryos the heart endothelium and aorta arise im loco within the embryo, and here there are no vessels, nor even meso- derm, present on the yolk-sac in the cranial portion. Fig. 6a represents the graphic reconstruction of the vascular system of the cranial portion of the Stage II. (a) ferret embryo. In this specimen ‘the heart rudiment is represented merely by a transverse blood channel which lies across the median plane and unites the cranial ends of the two vitelline veins. The pleuro-pericardial cavity, together with the pleuro-pericardial canals, has already been described as having the shape of an inverted U-shaped canal which lies dorsal to the vitelline vein and the heart rudiment. Two rudimentary dorsal aortz can be made out in this specimen. They run caudo-cranially one on each side of the medullary groove. They are still more or less plexiform in character, and they terminate blindly 160 Dr C. C. Wang at their cranial extremities. The absence of the first aortic arch, which is so conspicuously seen in the next stage, deserves particular notice. It is clear that at this stage of development in the ferret, the heart rudiment and the caudal part of the dorsal aorta are present, but the connection, 4.e. the cranial dorsal aorta, the first aortic arch, and the conus arteriosus, between the heart and the caudal dorsal aorta, is still wanting (fig. 6a). A stage further in the development of the cranial portion of the dorsal aorta is illustrated by the Stage III. embryo (fig. 18). Here the dorsal aorta is seen to have established its connection with the heart rudiment through the first aortic arch. But exactly how this connection takes place, there is no evidence from which to form any definite conclusion. It is as yet impossible to decide whether the first aortic arch and the conus arteriosus when developed, as seen in the specimen just referred to, should be attributed to the result of a cranialward growth from the dorsal aorta, or as the direct outcome of an extension of the heart rudiment growing round the cranial end of the fore-gut to join the dorsal aorta and to constitute the conus arteriosus and the first aortic arch. All that can be said is that probably coinciding with the formation of the head fold the two dorsal aortz are carried, pari passu, cranialward over the cranial end of the fore-gut, and possibly, as the result of a further growth from the blind ends of the aorte towards the heart rudiment, these structures establish their communications with the heart. If this contention re- presents precisely what really takes place in the ferret embryo, the conus arteriosus and the first aortic arch must be considered as being the result of a further growth from the dorsal aorta. But if the other theory is to be accepted, that is, that the development of the conus arteriosus and the first arch is due to an extension of the heart growing round the fore-gut, then the development of these parts of the vascular system does not conform with the statement made by Bremer (12) relating to the early development of the blood-vessels in the rabbit embryo of 5 somites. ‘This investigator asserts that the dorsal aorta, the first aortic arch, the conus arteriosus, and the lateral heart are all parts of an original network of _ angioblast cords derived from the extra-embryonic plexus of blood-vessels. Another mode of origin ‘which cannot, however, be overlooked, is that; after the heart rudiment and the dorsal aorta have been laid down, the remaining parts of the main vascular system of the cranial region of the embryo may develop vm sitw from the mesoderm. The view that parts of the intra-embryonie vascular system arise in sitw cannot be ignored, for there is an overwhelming accumulation of conclusive evidence to indicate that the formation of the intra-embryonie blood-vessels is much more extensive and important than has hitherto been supposed. Development of Blood-vessels and Heart in Ferret Embryos 161 From whichever point of view the development of the dorsal aorta, the first arch, the conus arteriosus, and the heart is to be looked upon, the fact remains that these structures do not develop simultaneously. ‘This is clearly shown in fig. la, in which the heart rudiment and the dorsal aorta at this stage of development are all represented, yet there is nothing to indicate or to represent the future first aortic arch and the conus arteriosus. It is certain that, in the ferret, mesoderm is laid down in the whole of the embryonic area before any blood-vessels appear in the area. It is also certain that some of the splanchnic mesoderm cells arrange themselves into an anastomosing plexus of larger and smaller strands. One of the larger strands lies lateral to the notochord on each side and ultimately is trans- formed into the dorsal aorta, whilst previously a strand along the anterior border of the area is transformed into the channel of communication between the vitelline vein of opposite sides. The dorsal aorta and the transverse communication between the vitelline veins are at first quite separate from one another, except that they are both connected. with the surrounding splanchnic mesoderm. As the head fold develops the dorsal aorta and the transverse channel between the vitelline veins become united and the first aortic arches are formed. How the union occurs is not clearly shown by the specimens, for intermediate stages are wanting. It might be by growth caudalwards from the vitelline veins, or cranialwards from the aortz, but it is more probable that arches of communication are formed im sitw from the mesoderm, and that the whole process of blood-vessel formation in the ‘embryo is one of transformation of cords of splanchnic mesodermal cells into tubes. The transformation takes place first in the anterior and posterior ends of the embryonic area, but extends more rapidly forwards than backwards, hence the dorsal aorte grow from behind forwards and the last formed parts of the great vessels are the aortic arches. Development of the Human Vascular System. Some points in the early development of the human vascular system may now be discussed, although I have only had the opportunity of examining two very young specimens. According to Evans (12), it is certain that in man, long before any vascular rudiments are found in the body of the embryo, and at a time before any mesodermic somites are formed, typical vascular rudiments are detected irregularly scattered, at first, over the surface of the ventral pole of the yolk-sac only, but on account of its comparatively small size the Yaeeslariestion of the whole surface of the yolk-sac is soon completed. It is generally believed that, as in other vertebrates already studied, 162 : ; Dr C. C. Wang these vascular rudiments make their appearance as nodular swellings of that part of the wall of the yolk-sac known as the area vasculosa, and are cell clumps lying between the mesoderm and the entoderm. It is claimed also that very shortly after their appearance, the peripheral cells of these cell clumps arrange themselves to form endothelium while the central ones remain as blood-cells. : In young human embryos it has been possible to demonstrate that, at a period before any vascular rudiments on the yolk-sac proper can be distinguished, there develop in the belly-stalk and chorion of the embryo indisputable blood-vessels which appear, at first, as strands of spindle cells possessing a lumen. This has been described by Fetzer (10), and also | observed by Graf Spee (’96) in the embryo Von Herff of 37 mm. Others (Jung (07) and Herzog (’09)) have called’attention to the aggregations of | endothelial cells in the belly-stalk. True blood islands in the belly-stalk near the allantois have been described also by Grosser (113) and Debeyre (12). Frassi (08) also is in favour of the view that well-formed angio- blastic cords can be detected on the ventral surface of the yolk-sac and in the belly-stalk and chorion. I have examined very carefully the whole series of sections of “A very early human ovum embedded in the uterus” described by Johnstone (14), _ and have found that true vascular endothelium in the form of isolated cords is present in the ventral pole of both of the twin vesicles near their attachments to the blastocysts. Dr Johnstone, however, considers these endothelial cords as merely localised thickenings which he was unable to denote definitely as the precursors of vessels. Judged in the light of observations made by other = cainionie it is not unreasonable to look upon what have been considered as the “tere localised thickenings” by Dr Johnstone as in reality true vascular endothelial cords. The reason for this belief is the fact that, in position and in their general characters, the cell cords or “thickenings” in question bear a close resemblance to those which have been described in other early human embryos as endothelium by Fetzer, Graf Spee, Jung, Frassi, and others. Recently Bremer (14) has stated vs in human embryos, the earliest blood-vessels appear separately in the yolk-sac and in the belly-stalk in the form of multiple rudiments which are for the greater part funnel-shaped invaginations of the surface of the mesoderm. By a partial fusion of the walls of an ingrowth, a portion of the ecelom, bordered by mesoderm, may be cut off as a separate cavity, lying deep within the substance of the belly-stalk. This investigator, therefore, believes that the endothelium arises either by delamination from the walls of such a detached portion of Development of Blood-vessels and Heart in Ferret Embryos 163 the ccelom, or by direct extension, in the form of an angioblastic cord, from the mesothelial ingrowth. Most authors believe that the early development of the vascular rudi- ments in the belly-stalk and chorion in human embryos happens before the yolk-sac proper exhibits any vascular elements. That this should be the case is due to the fact that in human embryos the vitelline circulation is of secondary importance. The belly-stalk and chorion, on the other - hand, constitute the primary connection between’ the embryo and the placenta, and are therefore the first to be vascularised. This deviation from the ordinary type of development is but one of the remarkable series of variations with which man is distinguished from his fellow-creatures. All facts, therefore, tend to point that in man the first vascular endothelium is laid down in the belly-stalk and chorion. This furnishes an additional evidence in favour of the biphyletic origin of blood-cells and blood-vessels. The next question to be considered is whether the vascular- isation of the yolk-sac proper is to be regarded as the extension of a further growth from the vascular rudiments in the belly-stalk and chorion, or whether the process arises im situ by separate vascular rudiments. Bremer (14) thinks that the vascularisation of the yolk-sac proper is a separate manifestation, but it is quite possible that the first vascular endothelium of the yolk-sac is the result of an extension from the endothelium of the belly-stalk and chorion. Unfortunately, Johnstone’s specimen gives no definite indication either of the origin of the cell cords or of their extension. Development of the Heart and Pericardium. In the literature dealing with the development of the heart and _peri- cardium, much has been written regarding the developmental processes of these structures in mammals, but in man much is required yet before a comprehensive knowledge of the developmental phenomena can be obtained. It is still a speculation if the earliest rudiment of the human heart is essentially similar to that of the mammalia. Tandler (’12) inclines to the belief that they are similar. It isnot to be expected, at present, that the comparison of the development of the human heart with that of the mammalia will throw much light upon the subject until the various developmental processes of the different members of the mammalian embryos are first understood. In mammals, as has been stated previously, the first rudiment of the heart is the appearance of a number of cells—the angioblast of His, which are distinguishable in embryos of 2-3 primitive somites. These vascular cells appear between the entoderm and. mesoderm in the cranial region of 164 Dr C. C. Wang the embryo on each side not far from the median plane. They are responsible for the development of the endothelial heart tubes only, the remaining structures of the wall of the heart being derived from that part of the visceral ecelomie wall which has been designated by Mollier (06) the heart-plate. It is generally believed that the first appearance of the vascular cells of the mammalian heart is bilateral and is located on the ventral aspect of the pleuro-pericardial canals. Hensen (’76) is credited as being the first who observed this bilateral origin of the heart. A fusion of » the two endothelial tubes next takes place, and the paired heart rudiments are therefore transformed into an unpaired heart tube, but the precise mode of this transformation is, up to the present, swh Judice. As far as the bilateral origin of the heart rudiments is concerned there seems to be still a little doubt. In all vertebrates that have been investi- gated all writers incline to believe that the heart rudiments originate in two lateral parts, but precisely how these two parts are brought together to form an unpaired heart is much a disputed point. As already pointed out (vide swpra), some believe (a) that in mammals, as in birds, the two endothelial tubes, out of which the heart is formed, appear at a time when the lateral folds which are said to form the ventral wall of the throat are only just visible; (b) that, on the formation of the lateral folds of the splanchnic walls, the two halves of the heart, enclosed within the hitherto symmetrical and laterally placed pleuro-pericardial cavities, become carried medially and ventrally until they fuse on the ‘ ventral aspect of the fore-gut; and (c) that the heart is therefore provided, at least for a time, with a ventral and a dorsal mesocardium. Professor Wilson of Sydney (’14), writing in favour of the presence of a ventral mesocardium in human embryos, asserts that the human heart, like the amphibian, has, at a certain period of development, a ventral mesocardium. ‘This author bases his conclusion on the result of the ex- amination of a series of sections of a human embryo measuring 1°78 mm. catalogued H3 in his series, which was aborted and is admittedly in an indifferent if not bad state of preservation. In addition it is stated that the specimen was cut “in a plane intended to be ‘transverse to the long axis of the embryo, but which turned out to be distinctly oblique.” The obliquity appears to have taken place in two planes instead of one, namely, a ventro-dorsal and a lateral. No reconstruction of the heart of the embryo has been made to verify his statement, and the number of somites has been hesitatingly determined to be three pairs. In fig. vi. of Wilson’s paper an incomplete “septum” in the peri- cardial cavity has been named by him the “septum proprium interperi- cardiacum.” This he looks upon as evidence of the bilateral origin of Development of Blood-vessels and Heart in Ferret Embryos 165 the pericardial cavity, which, according to his observation, ends on each side “blindly without establishing any communication with other ccelomic cavity such as occurs later, ¢.g.in Mall’s embryo, No. 391.” In the same figure the heart appears to have an attachment to the ventral wall of the pericardium which presumably Wilson calls the ventral mesocardium. It is quite possible, as far as the figure shows, that the so-called septum proprium interpericardiacum is merely a fold of the pericardial wall which has never, at any time, formed a true septum. The statement made regarding the blind termination of the pericardial cavity on both sides of the embryo is contrary to all known facts established by other observers on this structure in mammals. The narrow communication between the pericardial cavity and the coelomic cavity has probably been overlooked by Wilson in his specimen on account of the poor histological details of the sections, or the pleuro-pericardial canal might have been obliterated in such a way as to render the communication unrecognisable. If the ob- servation of Wilson, is the true interpretation of the condition of the peri- cardium at this stage of development, then the balance of evidence in favour of the primary communication of the pericardial cavity with the general ccelomic cavity in mammals is totally upset. But the unsatisfactory condition of the sections, from a histological point of view, unfortunately affords ground to doubt the accuracy of the observation, otherwise this embryo opens a new field for further investigations into the true nature of the pericardial cavity at this particular stage of development in the human embryo. ‘The connection between the heart tube and the ventral wall of the pericardial cavity in Wilson’s specimen can be feasibly explained in the following way. If the obliquity of the plane of section is such that it passes through the septum transversum as well as the heart tube situated immediately cranial to it, it will be seen that what appears to be the ventral mesocardium is really a part of the septum transversum. More- over, it is quite possible that what Wilson considers to be the two endo- thelial tubes, may after all turn out to be the two vitelline veins passing through the septum transversum to reach the sinus venosus cranially, and that the specimen he has been dealing with might have been pathological. I was fortunate enough to have at my disposal also the second young human embryo which has been described by Dr Johnstone (’14) in his. “ Contribution to the Study of the Early Human Ovum.” When the sections came to my hand, it was found that the series was unfortunately not entirely complete; consequently a reconstruction of the heart of this in- teresting stage had to be abandoned. Nevertheless the series serves one particular purpose very well, namely, that it shows that this specimen has been cut in a plane not dissimilar to that of Wilson’s embryo, and that the 166 Dr C. C. Wang heart tube in Dr Johnstone’s specimen bears a close resemblance to that of Professor Wilson’s specimen, in so far as the anatomical relationships of the heart are concerned. Though no description of the heart of the embryo has been made by Dr Johnstone in his paper, a few very interesting points can be detected by an examination of the serial sections. At first glance the section of Johnstone’s specimen, from which fig. 35 is taken, appears to indicate that the heart has a ventral mesocardium connecting the ventral part of the heart tube with the ventral wall of the pericardium. But on closer examination, what seems to be ventral mesocardium is found to be really Fore-gut. Medulla. Dorsal aorta, Dorsal mesocardium. Pericardial cavity. — Endotnelial tube. Mesodermic tissue. * * 4 ‘ Ps Vitelline Septum vein. transversum. Fie. 35. —Human embryo. x 100. part of the septum transversum, for when this structure is traced forwards and backwards in the series, it is found that the vitelline veins traverse it to join the sinus venosus, Fig. 36 shows that the vitelline vein passes through the septum transversum to reach the heart. Cranially to this there is nothing to indicate a ventral mesocardium, the cavity of the pericardium extending without interruption from side to side (fig. 37) Professor Robinson has very kindly looked through the sections of Dr Johnstone's specimen with me and has confirmed the above statement. In the chick, on the other hand, a ventral mesocardium is recognisable, but this is due, as Robinson (’02) points out, to the relatively late penetra- tion of the pleuro-pericardial canals into the mesoderm in the cranial region. The pleuro-pericardial canals do not extend round ‘and unite in front of Development of Blood-vessels and Heart in Ferret Embryos 167 the medullary plate in early stages, but only at a later stage do they penetrate into the floor of the fore-gut after the mesoderm has been Mesodermic tissue. Muscular heart tube. Vitelline vein. : Septum transversum. Fic. 36.—Septum transversum in fig. 35 magnified. x 500, Medulla. Dorsal aorta. Fore-gut. Pericardial cavity. Dorsal mesocardium, Pericardial cavity. . Endothelial heart tube. Muscular heart tube. Mesodermic tissue. Fic. 37.—Human embryo. x 100. formed. The lateral cavities therefore do not at once become continuous, but remain separated from each other by a double layer of mesoderm which constitutes the ventral mesocardium. 168 Dr C. C. Wang According to Robinson (’02) the pericardial mesoderm appears in the pericardial portion of the embryonic area, and it is there completely differentiated into somatic and splanchnic layers before the head bend is developed; there is, therefore, a single pericardial cavity to begin with, which extends from side to side along the cranial boundary of the embryonic area. As the head bend develops, the single pericardial cavity _is reversed, and it is carried into the ventral wall of the fore-gut, where it forms a U-shaped tube which communicates at each end with the general ccelom. The heart rudiments are formed in the splanchnic layer of the peri- cardial mesoderm ; therefore, after the reversal of the area, they lie in the dorsal wall of the pericardial cavity attached only by a dorsal mesocardium to the ventral wall of the fore-gut, but they are never, at any time, con- nected with the ventral wall of the pericardium by a ventral mesocardium. Rouviere (04), whilst agreeing with Robinson as to the absence of the ventral mesocardium, describes the formation of the lateral pleuro-peri- cardial canals which grow cranially round the cranial end of the brain- plate and fuse to form a continuous channel. Miss Parker (15), in her investigation into the early stages in the development of Marsupials, modifies Rouviére’s opinion, saying “that while the initiation of head-fold formation is in all probability due to the forward growth of the brain-plate, there occurs also an active backward growth of the anterior intestinal portal (umbilical orifice). This process is associated with the rapid expansion of the pericardium which occurs at this period of development, and which brings about the backward and inward growth of the layer of splanchno- pleure limiting the pericardium.” Against the view that a backward growth occurs in the cranial margin of the umbilical orifice is the contention of Robinson (02) that “the orifice (of the umbilicus) is not reduced in size during the early stages of develop- ment by the convergence of its margins towards a central point. This being the case, no tucking off of the embryo from the surface of the ovum can occur; on the contrary, what does occur is almost the exact opposite of such a process, for the margin of the area remains as a relatively slow- growing region, whilst the embryonic and extra-embryonic portions of the wall of the ovum rapidly increase in extent. Under these circumstances, it follows that the margin of the embryonic area will soon appear as a ring between the upper or embryonic, and the lower or extra-embryonic parts of the ovum, both of which have expanded beyond it in all directions.” The evidence afforded by the ferret material I have described shows that there is no ventral mesocardium in the ferret, and all the indications it gives are strongly opposed to the idea that any part of the gut closure is affected by the fusion of the lateral folds. Development of Blood-vessels and Heart in Ferret Embryos 169 In the Stage II. (a) and (b), in which the head fold of the ferret embryo has not yet appeared (figs. 6a and b, 1la and 5), the pleuro-pericardial cavity is present in the median plane, and immediately caudal and ventral to this the two vitelline veins communicate with one another across the median plane. There is no evidence that they were even separate from one another, but the union may be called the primary union of the heart rudiments ; it should be specially noted that the term “ primary ” proposed signifies that the right and left parts separate from one another and are again fuséd by a “secondary” union. To this point reference will be made subsequently in connection with the discussion of the formation of the unpaired heart rudiment from the two separate endothelial tubes. At Stage II. of development the separation of the right and left parts of the rudimentary heart has commenced, and the pleuro-pericardial cavity and the heart rudiments are all represented before there is any indication of the formation of the fore-gut or the head fold (fig. 6a). It should also be noticed that when the heart rudiments and the pleuro-pericardial cavity first make their appearance, the former invariably lie ventral to the latter (figs. 6b and 8). In the subsequent stages of development (Stage III. and onwards) a change takes place in their position, with the result that the heart rudiments occupy a plane dorsal to the pleuro-pericardial cavity, and are connected with the ventral aspect of the fore-gut only by the reflection of the splanchnic wall of the pleuro-pericardial cavity. The reversion of positions of the heart and the pleuro-pericardial cavity can only be explained, as suggested by Professor Robinson (’02), by a forward growth of the head fold to form the fore-gut with the cranial margin of the umbilical orifice remaining stationary. With the development of the head fold the heart rudiments suffer a rotation round a transverse axis over the pleuro-pericardial cavity. Consequently when the two heart rudiments which became separated are again brought together to form one endo- thelial tube, it is connected with the ventral surface of the fore-gut only by a dorsal mesocardium, which is formed by the reflection of the splanchnic layer of the pleuro-pericardial cavity over the lateral and ventral surfaces of the two heart tubes; the portion of the splanchnic wall of the pleuro- pericardial cavity in which the endothelial tube lies is in the meantime pushed ventrally, and thus the dorsal mesocardium is elongated. There is no indication of a ventral mesocardium. : Since the time of Hensen (’76), the first appearance of the heart rudi- ments has been considered to be bilateral, that is to say, the heart rudiments develop as two separate endothelial tubes, one on each side of the embryo not far from the median plane and on the ventral aspect of the pleuro- pericardial canals. Recently the bilateral endothelial tubes have been VOL. LII. (THIRD SER. VOL. XIII.)—JAN. 1918. 12 170 Dr C. C. Wang traced to the stage of angioblastic cords, which have also been described as lying between the mesoderm and the entoderm. Spaces soon make their appearance in the vascular masses, and when these coalesce, two endothelial tubes are thus formed, one on each side of the embryo, ventral to the pleuro-pericardial channels. In the ferret the coalescence of the vascular endothelium is to produce one endothelial tube (the “primary” union of the heart rudiments) lying across the median plane caudo-ventral to the pleuro-pericardial cavity (Stage II. (a), figs. 6a and 6). This occurs, it should be remembered, at a time before the formation of the fore-gut has made its appearance. When this “primary” endothelial tube is traced laterally, it is found to communicate with the two vitelline veins (fig. 6a). A search through the literature on the early stages of development of the mammalian heart has failed to discover any account of this “primary” union of the heart rudiments. Miss Parker, in the course of her investiga- tion into the early stages in the development of Marsupials, notices the early or “primary” union of the heart rudiments across the median plane in Stage II. Dasywrus viverrinus (85 mm.); its significance, however, has not been explained by her. That the “primary” union of the heart rudiments to form a single endothelial tube is not a singular occurrence due to any abnormal or pathological conditions, and that this stage of development of the heart is an important one is proved by the fact that a similar phenomenon repeats itself in another ferret embryo (Stage II. (b), figs. lla, b, and 12a, 6, c), in which the heart and the pleuro-pericardial cavity exhibit features similar to those observed in the Stage II. (a) ferret embryo. It has to be emphasised once more that the “primary” union of the heart rudiments across the median plane as a transverse vascular channel lying caudo-ventral to the pleuro-pericardial cavity occurs at a period before any indication of the head fold or the formation of the fore-gut can be detected. The next stage of development of the heart in the ferret embryo is represented by the 2°3-mm. embryo (Stage III.). In this specimen, which is obviously slightly older than the Stage II. (a) and (6), the medial part of the “primary” endothelial tube shows indications of splitting into two lateral endothelial tubes, for two non-vascular loculi divide its central portion incompletely into a cranial and a caudal portion (fig. 18). Ap- parently the destruction of the central part of the rudimentary transverse heart (“primary” union) proceeds still further as development goes on, until it is completely separated into a right and a left half; for in the Stage IV. ferret embryo the heart rudiments are represented there by two separate longitudinal endothelial tubes lying side by side close together (fig. 27). It is again to be noted that the head fold and the fore-gut of Development of Blood-vessels and Heart in Ferret Embryos 171 this ferret embryo (Stage III.) have already begun to develop, since at the cranial end two vessels, one on each side of the median plane, run eranially from the heart rudiment. They arch round the cranial end of the fore-gut and form the first pair of aortic arches, which terminate dorsally in the corresponding dorsal aorta (fig. 18). In the comparison of this stage of development of the heart in the ferret with that of other mammals of a similar stage of development, the Stage III. Perameles nasuta (1 8) of Miss Parker may be cited; the total length of her specimen, after partial flattening under cover glass, from the anterior margin of the brain-plate to the hinder extremity of the primitive streak is, according to Miss Parker, 7°5 mm. In her specimen two endo- thelial tubes are depicted lying side by side in the pleuro-pericardial cavity. Cranially each endothelial tube becomes continuous with the ventral aorta, which runs underneath the ventral surface of the fore-gut. As the cranial extremity of the fore-gut is reached, the ventral aorta bends dorsally round the blind end of the fore-gut and communicates with the dorsal aorta, so forming the first aortic arch. Whilst still in the pleuro- — pericardial cavity the heart tube is described as giving off a lateral branch which is presumably the second aortic arch. What is most striking and interesting in Miss Parker’s specimen is, that “in the median space between the anterior ends of the endothelial heart tubes, are a number of scattered angioblast cells lying between the splanchnic mesoderm and the entoderm.” Miss Parker believes that these cells possibly represent the primordia of the capillaries and afford an instance of the origin of angioblast cells from the splanchnic mesoderm after the establishment of the definite endothelial heart tubes. Judging from what has been observed in the ferret embryo of Stage III. it is perhaps more correct to interpret these angioblast cells as being the remains of the once “primary” union of the heart rudiments. This “primary” union is well developed in the ferret embryo of Stage II. (a) and (b), and also in the Stage II. Dasywrus viverrinus (8°5 mm. A) of Miss Parker. The explanation for the breaking up of the “ primary ’ union of the heart rudiments to form two separate endothelial tubes is, however, not apparent in the light of our present knowledge of the development of the mammalian heart. It is conceivable that partly as a result of the development of the fore-gut cranialward, in the manner pointed out by Robinson ('02), and partly also due to the rotatory movement of the heart rudiments from a position ventral to the pleuro-pericardial cavity to a position dorsal to it round a transverse axis, the cross channel, that is, the “primary ” union of the heart rudiments is, at first, put on the stretch, and finally separated into two endothelial tubes lying side by side, each of which becomes continuous with its corresponding dorsal aorta through the 172 Dr C. C. Wang conus arteriosus and the first aortic arch. In the ferret embryo of Stage III. (fig. 18) the connection between the heart and the dorsal aorta has already established itself before the heart (after its “primary ” union) has again completely separated into two halves. In the case of the Stage II. Dasyurus of Miss Parker, the “ primary” union of the heart rudiments is described as giving rise to the first aortic arch, which follows the antero- lateral margin of the gut almost to the median plane and there becomes continuous with the corresponding dorsal aorta. But what has been taken for the first aortic arch by Miss Parker may, after all, prove to be the plexus between the dorsal aorta and the vitelline vein, which has been observed by Bremer (’12) in the 3-4-mm. rabbit embryo. It will be noticed, then, there is evidence to show that, at a period before the formation of the fore-gut, the first rudiment of the heart in the ferret is single and is situated in the median plane of the embryo, caudo-ventral to the pleuro-pericardial tube cavity, in the form of a transverse endo- thelial tube which is destitute of any blood-cells. Laterally it is in direct communication with the two vitelline veins, one on each side of the embryo. But exactly how this “ primary” transverse heart tube is formed it is impossible to make any definite statement. The contention of His (700) does not fully account for the origin of this “primary” union of the two vitelline veins, for, according to His (’00), two endothelial tubes are directly formed from the ingrowth of the two vitelline veins which, by a further growth, communicate with the two dorsal aortz. Possibly the “primary” union of the heart rudiments is the result of an ingrowth from the vitelline veins which, instead of linking up each with its corresponding dorsal aorta, as His would have imagined, have grown across the median plane. In this way the U-shaped vitelline system is, at least for a time, not connected with the longitudinal aortic system cranially. This is exactly what is seen in the ferret embryo of Stage II. (a) (fig. 6a). It is also possible, and perhaps more probable, that the “primary” union arises im situ and is afterwards joined to the two vitelline veins. From whatever point of view the development of the heart is looked upon, it is clear that in the ferret the first rudiment of the heart appears as a cross channel situated ventro-caudally to the pleuro-pericardial cavity, and is connected only with the venous system, which is composed of the two vitelline veins laterally. The arterial system, that is, the dorsal aorte, remains, at least for a time, distinct and unconnected with the heart rudiment. No explanation can at present be offered as to why the heart rudiment, when first represented, should only be united with the venous circulation. It is evident that further light in this field of investigation is required before a solution can be obtained, Development of Blood-vessels and Heart in Ferret Embryos 173 Quite recently Professors Robinson and Gibson (’16), in their description of a reconstruction model of a horse embryo twenty-one days old, mention that “the allantoic blood-vessels consist of a number of dilated capillaries which form a coarse network on each side. Each lateral network receives two branches from the caudal end of the dorsal aorta of the same side, and it terminates, at the caudal end of the allantoic mass, in a terminal trans- verse sinus from which the umbilical veins take their origin. But, in addition to the connection with both umbilical veins through the terminal sinus, each vascular network also communicates directly with the umbilical vein of the’same side.” The terminal transverse sinus at the caudal end of the allantoic mass of the horse and the transverse heart rudiment situated in the cranial end of the ferret embryo may together represent at one time a portion of an original complete ring of circulation. The next phase of development of the heart is represented by the separation of the “primary” heart rudiment into two distinct endothelial tubes lying closely together, one on each side of the embryo, not far from the median plane (Stage IV.). Coinciding with the development of the heart, the head fold appears, and as the formation of the fore-gut proceeds, the heart rudiment suffers a reversion with regard to its relation with the pericardium, for it is seen that when the fore-gut is formed the heart is found on the dorsal aspect of the pleuro-pericardial cavity and is attached to the ventral surface of the fore-gut. Each tube is covered by the splanchnic wall of the pleuro-pericardial cavity on its lateral, ventral, and medial surfaces. Dorsally they are connected with the ventral aspect of the fore-gut by the reflection of the wall of the pleuro-pericardial cavity. Ventrally the pleuro-pericardial cavity passes from side to side (fig. 27). It is to be noted that the two tubes lie far apart from each. other cranially and caudally where they emerge from the pleuro-pericardial cavity. The distance between them is greater caudally (fig. 26) than cranially (figs. 27 and 28d). It is obvious that at a stage further, when the two heart rudiments are brought together to form one endothelial tube, it is connected only with the ventral. surface of the fore-gut by a dorsal mesocardium and no ventral mesocardium can possibly be developed. Cranially each endothelial tube can be traced to its corresponding dorsal aorta through the first aortic arch, and caudally each tube com- municates with its corresponding vitelline vein in the region of the septum transversum (figs. 21, 25, and 27). It is generally held that the rate of growth of the two endothelial tubes exceeds that of the pleuro-pericardial cavity. Consequently, as the result of the different rate of growth, the two tubes, instead of having a 174 Dr C. C. Wang — straight appearance, are thrown into loops before the “secondary” union takes place. Miss Parker (15) thinks, however, that at this stage of development the pericardium grows rapidly in length and decreases in width so that the heart tubes are brought together by longitudinal stretching of that part of the pericardial wall which lies between them. In the ferret it is clear that this is not the case; the two endothelial tubes seem to grow more rapidly in length than the pericardium, with the result that loops are formed with constrictions here and there to mark out the different — subdivisions of the heart (atrium, ventricle, bulbus, ete.), before any fusion of the two tubes takes place (Stage IV. fig. 27), these subdivisions appear to remain even when the two endothelial tubes have partially united (Stage V. fig. 31). The appearance of these loops in the endothelial tubes speaks against the theory of stretching advanced by Miss Parker. For, if ‘it were true, the result of any stretching of that part of the pleuro-peri- cardial wall which lies between the medial borders of the two endothelial tubes would, in the first instance, be the undoing of the loops, or the pre- vention of their formation, before any approximation of these tubes could be effected. . In the ferret (Stage IV.) what has really happened is this, that as the result of the rapid growth, the two endothelial tubes are thrown into loops, and as the result of a further growth of the two tubes medially _towards each other, the part of the pleuro-pericardial wall which lies between them is pushed ventrally. When fusion of the two endothelial tubes takes place, the unpaired heart is, therefore, attached only to the ventral wall of the fore-gut by the dorsal mesocardium. Ventrally the splanchnic wall of the pleuro-pericardial cavity passes from side to side across the ventral aspect of the fused heart tube, there being no fusion of this part of the pleuro-pericardial wall, as supposed by Miss Parker. Asymmetry of the heart tubes, either before or after fusion takes place, has been noted in the ferret embryos (Stages IV. and V., figs. 27 and 31), as in the Stage IV. Perameles nasuta and also, in a measure, in the Stage V. Perameles obesula (10, viii.,’03) of Miss Parker and likewise in the 5‘7-mm. dog embryo of Bonnet (01). There seems to be a tendency for the two endothelial tubes, even before fusion, to curve and to be shifted distinctly as a whole to the right side with the concavity of the bend facing the left. In Stage V. (fig. 31) the fusion of the two endothelial tubes in the ferret takes place in the ventricular region, and the fused part lies clearly on the right side of the median plane. The fact of this oceur- _ rence contradicts, in a very convincing manner, the theory of stretching of the pleuro-pericardial wall, for the fused portion of the two endothelial Development of Blood-vessels and Heart in Ferret Embryos 175 tubes should occupy the median plane of the embryo if the two tubes were really brought together by the uniform stretching of that part of the splanchnic wall of the pleuro-pericardial cavity which lies between them. There is no evidence to prove that the fusion of the two tubes, occurring on the right side, may be due to an uneven strofehing of the dorsal wall of the pleuro-pericardial cavity. In the Stage IV. Perameles nasuta (2 P) of Miss Parker the descrip- tion of the heart in the text does not seem to agree with her illustration— Plate 1, fig. 5. In the text it is stated: “They (the two endothelial tubes) have fused at their cephalic extremity, the fused portion extending through some eighteen sections and representing the most closely approxi- mating portions of the endothelial tubes. . . . From it is derived the bulbus (conus) arteriosus... . . Posterior to this fused portion, the endothelial tubes lie close together but unfused for a considerable portion of their length and then diverge widely and pass into the vitelline veins.” In her illustration, on the other hand, it appears that the fused portion represents really more than the conus arteriosus. Possibly the ventricles also have fused across, because in the figure it shows that at least the cranial half, if not more, of the two endothelial tubes have fused. If this is the case, it is difficult to explain why in the next stage of development (Stage V. Perameles obesula (10, viii., 03) of Miss Parker) the bulbus is the only portion of the heart in which the endothelial tubes have actually fused, and the ventricles have become once more separated. In the text and the figures illustrating Stage IV. Perameles nasuta (2 P) and Stage V. Perameles obesula (10, viii., 03) of Miss Parker, there are no data to be found upon which the magnification and length of the embryos in question can be gauged. But as far as one can judge from the illustrations alone, the figures of Stage V. Perameles obesulu are undoubt- edly of a higher magnification although probably taken from an embryo of a greater length than-that of the Stage IV. Perameles nasuta. If it is so, an explanation is needed to account for the decrease in length of the ‘fused portion of the two endothelial tubes seen in Stage V. It is unfortu- nate that these important data should have been omitted by Miss Parker in her paper, for their absence diminishes the value of her communication when an attempt is made to use it for the purpose of solving the points under consideration. Hitherto the heart tube with its time-honoured S-shaped character has been the subject of much dissertation. The literature on the development of the S-shaped heart rudiment is so voluminous and confusing that only a brief summary on this subject can be given. All writers have laid great stress on the subsequent development of the fused endothelial tube which 176 Dr C. C. Wang is now covered with its myocardial coat. It is generally believed that the next phase of development of the fused heart tube, consisting as it does of an inner endothelial tube and an outer myocardial covering, is the acqui- sition of the well-known S-shaped form. The caudal portion of the single heart tube grows cranio-dorsally and to the left to form the atrial limb, and the cranial portion grows caudo-ventrally and to the right to form the ventricular limb. In the Stage IV. ferret embryo model (fig. 22), it has been shown that the external configuration of the muscular wall of the heart does not necessarily furnish a true index to the condition of the development of the enclosed endothelial tubes, for it has been remarked upon that although the muscular coat of the heart rudiment may have acquired the familiar S-shaped appearance, yet the two endothelial tubes have not begun to fuse. Moreover, the segments exhibited in the muscular tube do not correspond in the least with the dilatations and constrictions of the underlying endothelial tubes. Furthermore, in the dog, it has been pointed out by Bonnet, and in Marsupials by Miss Parker, that primary divisions of the endothelial tubes into sinus venosus, atrium, and ventricle occur before they have completely fused to form a single tube. Similar divisions of the heart tubes are present also in the Stage IV. ferret embryo (vide supra). These facts therefore show that the primary divisions of the heart rudiment take place in the endothelial tubes ata time when no fusion can be detected, and that the convolutions of the S-shaped muscular coat of the heart do not necessarily represent the different segments of the underlying endothelial tubes. In connection with the fusion of the two indothalial tubes there is one more point which requires attention. Many investigators have noted the asymmetry of the heart rudiments at this period of development. Devia- tion of the two heart tubes to the right, before and after fusion, has also been observed in the mammalian heart. In the ferret these features are present. The significance of the asymmetry and deviation of the two endothelial tubes have, so far, not been fully explained by those who have - made these observations. It has been noted that the ventricular portion of the heart is the first to fuse, and is situated on the right side of the embryo (Stage V. fig. 31). The fused portion in fact represents the junc- tion between the ventricular and the atrial limbs of the future S-shaped heart tube. As fusion proceeds cranially and caudally, more parts of the heart tubes are taken in to form the two limbs of the S-shaped heart rudiment until the two tubes are completely fused. The result of this fusion is to produce a ventricular limb which is situated ventrally and to the right side and an atrial limb which is directed dorsally and to the left. It should also be remembered that in the space betweerl the myo-. Development of Blood-vessels and Heart in Ferret Embryos 177 cardium and the endothelium of the heart tubes there is a more or less thick mass of loose mesodermic tissue which separates not only the former from the latter but also the two endothelial tubes from one another before their fusion (Stage IV. figs, 27 and 28a, b,c). This mass of loose tissue, though playing an important role in the establishment of the sino-ventri- cular bundle (Mall (’12)) in the subsequent development of the heart, is essentially a passive element at this stage of development. It permits, on the one hand, the growing endothelial tubes to assume their various bend- ings, constrictions, and dilatations independently, and adapts itself, on the other hand, to the characteristic curling of the muscular tube into an S- shaped appearance. This point is clearly shown in the Stage IV. ferret embryo in which the two endothelial tubes have, as already noted, dif- ferentiated into their various divisions in advance of their fusion (fig. 27), whilst the muscular wall of the heart has seemingly acquired the familiar S-shaped form (fig. 22). It is obvious, therefore, that owing to the interposition of this more or less thick mass of mesodermic tissue, the endothelial tubes do not follow, in a faithful manner, the various curvatures and bulges of the myocardial covering of the heart, and it would be a mistake to try to determine, at this stage of development, the true nature of the two endothelial tubes by the simple examination of the condition and shape of the muscular coat, without ascertaining, at the same time, the various features of the under- lying endothelial tubes and comparing these with those exhibited by the muscular covering. It may therefore be concluded that, at least, from the first appearance of the paired heart rudiments as two endothelial tubes to the time of their “secondary ” fusion, the myocardium has little or no common relationship with the underlying endothelium; that. the two structures are quite in- dependent of each other, as far as their individual growth is concerned ; that the various constrictions and dilatations of the two endothelial tubes have a definite significance; that these constrictions and dilatations fore- shadow the site and limits of the future sinus venosus, sino-atrial canal, atrium, atrio-ventricular canal, ventricle, and bulbus cordis; and that the muscular tube comes into conformity only at a later period. SUMMARY AND CONCLUSIONS. A. The Blood Cells and Vessels. Hitherto it has been the general belief that blood-cells and vascular endothelium in mammals are derived from a common origin which, accord- ing to some, is mesodermic, and others, entodermic, and that, whichever 178 Dr C. C. Wang may be the source of origin, the vascular rudiment appears, at first, as angioblastic cells lying between the mesoderm and entoderm. It has been claimed also that the peripheral part of the angioblast soon resolves itself into an uninterrupted network of endothelium, and the central part — into clusters of blood-cells. It has been stated further that the endothelium so formed is capable of giving rise to new blood-cells. The facts revealed by the study of the early stages in the development of the ferret point to the conclusion that, whilst blood-cells and vascular endothelium are closely related to each other and are found invariably between the mesoderm and entoderm, there is evidence to show that, in the ferret, the origins of these two vascular elements are separate and distinct—the blood-cells arising from the entoderm and the vascular endo- thelium from the mesoderm. Blood-cells develop first extra-embryonically in the area vasculosa in the form of clusters of spheroidal cells which are provided with large and round nuclei and with a comparatively small amount of protoplasm. These are for the most part found adherent to the entoderm in the neigh- bourhood of their origin before they are engulfed by the endothelium, and are, in most instances, identical in structure with the entodermal cells ‘where the contact is intimate. On the other hand, the cells which form the endothelial rudiment are mesodermic in origin and are, without exceptions, spindle or flattened in shape. They are generally connected with one another by long slender protoplasmic processes, the result of which is to form a network of endo- thelium. This network is capable of extension by buddings which grow either into the embryo or on to the yolk-sac. The blood-cells are next engulfed by the vascuiar endothelium which grow round them, and in this way they are taken into the circulation. B. The Intra-Embryonic. Blood- Vessels. The caudal portion of the dorsal aorta communicates, at a very early ‘stage of development, with the yolk-sac circulation through the vitelline arterial plexus. Cranially the dorsal aorta stops short and remains un- » connected with the heart rudiment as long as there is no head fold or fore- gut. This part of the dorsal aorta makes its appearance in association with the development of the head fold and fore-gut. The conus arteriosus and the first aortic arch develop after the “primary” heart rudiment, the two vitelline veins, and the dorsal aorte are all represented, and at the time when the formation of the head fold and fore-gut has begun, The vitelline veins are two in number, one on each side of the embryo. a Development of Blood-vessels and Heart in Ferret Embryos 179 They lie for the greater part of their extent ventro-medial to the pleuro- pericardial canals. They grow from the wall of the yolk-sac into the cranial extremity of the embryo. They are, at an early period, united across the median plane to form the “ primary” heart rudiment which lies between the pleuro-pericardial cavity dorso-cranially and the bucco- pharyngeal membrane caudally. A short distance caudal to the “ primary ” heart rudiment, each vitelline vein sends out offshoots medially to anastomose with the cranial end of its corresponding dorsal aorta. C. The Pericardium and the Heart. Before there is any indication of the head fold or the formation of the fore-gut, the pleuro-pericardial canals have grown across the median plane of the cranial end of the embryo to form the pleuro-pericardial cavity which, in relation with the “primary” heart rudiment, lies cranio-dorsal to it. As the head fold and fore-gut develop by growing cranially, a rotation of the pleuro-pericardial cavity and the heart rudiment round a transverse axis takes place, with the result that the former occupies a position ventral to the latter and the “primary” heart rudiment is now found ventral to the fore-gut. In the subsequent stages of development the “ primary ” heart rudiment divides into two endothelial tubes, each of which is covered laterally, ventrally, and medially by the splanchnic wall of the pleuro-pericardial cavity and is attached only dorsally with the ventral aspect of the fore-gut by the reflection of the pleuro-pericardial wall. As a result of the growth of the two endothelial tubes towards the median plane, the part of the splanchnic wall of the pleuro-pericardial wall which lies between them is thereby pushed ventrally. When fusion of the two endothelial tubes occurs, the “ secondary ” heart tube is consequently attached only dorsally to the ventral surface of the fore-gut by the dorsal mesocardium. Ventrally there is no fusion of the pleuro-pericardial wall, and therefore no ventral mesocardium can possibly be developed. At an early period when the pleuro-pericardial cavity has not yet reversed its position, and when there is no indication of the formation of the head fold or the fore-gut, the heart rudiment appears as a transverse blood channel—the “primary” heart rudiment—situated in the median plane ventro-caudal to the pleuro-pericardial cavity and cranial to the bucco-pharyngeal membrane. Laterally the “primary” single heart rudiment communicates with the two vitelline veins, one on each side of the embryo. At this stage of development the cranial extremities of the two dorsal aorte are found, for a considerable distance, caudal to the heart rudiment. 180 Dr C..C. Wang In the subsequent development of the embryo when the head fold and the fore-gut make their appearance by growing cranially, the pleuro- pericardial cavity and the “primary” heart rudiment undergo a rotation round a transverse axis, with the result that their positions are reversed, so that the heart rudiment now lies dorsal to the pleuro-pericardial cavity. The division of the “primary” single transverse heart rudiment. into two longitudinal endothelial tubes is due to the fact that, at this period, the fore-gut grows rapidly cranially in length and ventrally in width, together with the expansion of the pleuro-pericardial cavity in all directions, so that the transverse heart rudiment which lies between these structures is at first put on the stretch, and is subsequently divided across into two endothelial tubes, a right and a left. Concurrently with the development of the head fold and the fore-gut, the two dorsal aortz establish their communications with the “primary” heart rudiment even before the latter is completely separated into two endothelial tubes. The next phase of development of the heart is the appearance of the two separate endothelial tubes, one on each side of the median plane. Each tube is continuous cranially with the dorsal aorta and caudally with the vitelline vein. As development proceeds they tend to grow towards the median plane, and in this way the part of the splanchnic wall of the pleuro-pericardial cavity which lies between them is pushed ventrally. The two tubes remain in contact, but not fused, with each other for a time. Owing to the fact that the endothelial tubes grow more rapidly than the pleuro-pericardial cavity, the former are thrown into loops which are separated by constrictions. These dilatations and constrictions indicate in caudo-cranial succession the future sinus venosus, the sino-atrial canal, the atrium, the atrio-ventricular canal, the ventricle, and bulbus cordis. The most dependent and approximating parts of the endothelial tubes are the ventricular portions. They incline more to the right side of the embryo, and are the first to become fused and to form the “secondary ” single heart tube. , ' The myocardium assumes the familiar S-shaped appearance at a stage when the heart is still in the condition of two separate endothelial tubes. The segments and sulci, appearing on the surface of the myocardial tube, do not correspond, in any way, with the dilatations and constrictions of the two underlying endothelial tubes. ACKNOWLEDGMENTS. I take this opportunity to express my thanks to Professor Arthur Robinson for the use of the ferret embryos from his collection which he Development. of Blood-vessels and Heart in Ferret Embryos 181 placed at my disposal. To him I am also grateful for helpful advice frequently asked and readily given. For the two early human embryos I am grateful to Dr R. W. Johnstone of the Midwifery Department, Edinburgh University, who, as stated, has kindly permitted me to make use of the serial sections. To the Carnegie Trust my sincere thanks are due for pecuniary support which has greatly facilitated the completion of this part of my research, and a grant to defray the cost of illustrations. The earlier part of the expenses incurred whilst working at the reconstructions has been borne by a grant from the Earl of Moray’s Fund, for which I desire to put on record my gratitude. The illustrations taken from sections of embryos and from drawings have all been photographed by Mr W. Watson of the Royal College of Physicians Laboratory, Edinburgh, under my personal supervision. Last, but not the least, I am indebted to Mr J. T. Murray for the drawings of figs. 18, 21, 22, 25, 27, and 30, which represent the original reconstructions. BIBLIOGRAPHY. 1912. Barney, F. R., Zext-book of Hmbryology, 2nd ed., Bailey and Miller, pp. 222-224. 1881. Batrour, F. M., Comparative Embryology, vol. ii. p. 524. 1842. Biscuorr, Lntwicklungsgeschichte des Kaninchens, Braunschweig. 1852. Biscnorr, Hntwicklungsyeschichte des Meerschweinchens. 1891. Bonnet, R., Grundriss der Entwicklungsgeschichte, Berlin. 1901. Bonner, R., ‘“ Beitriige zur Embryologie des Hundes,” Erste Fortsetzung, Anat. Hefte, H. 51, Bd. xvi. ; 1907. Bonner, R., Lehrbuch der Entwicklungsgeschichte, Berlin, 1889. Born, G., ‘‘ Heart, Rabbit,” Archiv fiir mikroscopische Anatomie, Bd. xxxiii. pp. 284-378. 1905. Bremer, J. L., ‘‘ Description of a 4 mm. Human Embryo,” Amer, Jour. of Anat., vol. v., No. 4, pp. 459-480. 1912. Bremer, J. L., “The Development of the Aorta and Aortic Arches in Rabbits,” Jowr. of Anat., vol. xiii. pp. 111-128. 1914. Bremer, J. L., ‘‘The Earliest Blood-vessels in Man,” Amer. Jour of Anat., vol. xvi., No. 4, p. 464. 1908. Bryor, T. H., Quain’s Anat., vol. i., “‘ Embryology,” p. 62. 1882. Burson, O., “Ueber eine Hypothese bezuglich der phylogenetischen Herleitung des Blutgefassapparates eines Theil der Metazoon,” Morphol. Jahrbuch, Bd. viii. 1910. Danny, W. E., Amer. Jour. of Anat., vol. x. 1912. Depryre, A., Jowrnal de l’ Anat. et de la Physiol., vol. xlviii. p. 448. 182 Dr C. C. Wang 1895. Erernop, A. C. F., ‘Communication sur un ceuf humain avec embryon excessivement jeune,” Arch, [tal. de Biologie, xxii.: XIe Congrés International des Science Médicales, pp. xii-xiv. 1899. Erernop, A. C, F., “Il y a un canal notochordal dans l’embryon humain, Anatomische Anzeiger, xvi. pp. 131-143. 1909. Evans, H. M., “On the Development of the Aorta, Cardinal and Umbilical Veins, and the other Blood-vessels of Vertebrate Embryos from Capillaries,” Anat. Record, vol. iii. pp. 498-518. 1912. Evans, H. M., ‘ Description of the Vascular System present in Early Human ‘ Embryos,” Keibel and Mall, Manual of Human Embryology, vol. ii. pp. 571, 588, 595. 1897. Feurx, W., “ Beitraige zur Entwicklungsgeschichte der Salmoniden, Anat. Hefte, Bd, viii. 1910. FELIX, W., “ Zur Entwicklungsgeschichte der Rumpfarterien des mensch- lichen Embryo,” Morph. Jahrb., Bd. xli., Heft 4, 8S. 577. 1910. Ferzer, “ Ueber ein ‘durch Operation gewonnenes menschliches Ei, das in seiner Entwicklung etwa dem Peterschen Ei entspricht,” Ver. Anat. Ges., Erg. Heft des Anat. Anz., Bad. xxxvii. S. 116-126. 1907. Frassi, L., ‘Ueber ein junges menschliches Ki in situ,” Arch. fiir mikr. Anat., Bd. Ixx. S. 482. 1908. Frassi, L., ‘‘ Weitere Ergebnisse des Studiums eines jungen menschlichen Eies in situ,” Arch. fiir mikr. Anat., Bd. lxxi. S. 667-694. — 1912. GrAper, L., “ Beobachtung von Wachstumsvorgaingen an Reihenauf- nahmen lebender Hiihnerembryonen nebst Bemerkungen iiber vitale Farbung,” Archiv fiir Entwicklungsmechantk der Organismen, Bd. xxxiii. 1895. Grosser, O., and Brezina, E., ‘‘ Ueber die Entwicklung. der Venen des Kopfes und des Halses bei Reptilien,” Morph. Jahrb., Bd. xxxiii. 1913. Grosser, O., and Brezina, E., “ Ein menschlichler Embryo mit Chordal Kanal,” Anat. Hefte, Bd. xlvii. S. 653-686 1909. Haun, H., “‘ Experimentelle Studien ueber die Entstehung des Blutes und der ersten Gefisse beim Hithnchen,” Arch. fiir Entwicklungsmechantk der Organismen, — Bd. xxvii. 1876. Hensen, V. von, “ Beobachtungen ueber die Befruchtung und Entwick- lung des Kaninchens und Meerschweinchens,” Zeitschr, fiir Anat. u. Entwicklungs- geschichte, Ba. i. 1892. Herrwia, O.,.“ Text-book of the Embryology of Man and Mammals, pp. 176-186, 542-548. 1909. Herzoc, M., Amer. Jour, Anat., vol. ix, p. 361. 1875. His, W., “Der Keimwall des Hiihnereies und die Entstehung der para- blastischen Zeller,” Zeitschr. f. Anat. u. Entwicklungsgeschichte. 1881. His, W., ‘Mitteilungen zur Embryologie der Saugetiere und des Menschen,” Archio J. Anat. u. Phys. 1885. His, W., Anatomie menschlichler Embyronen, LIL. i, Leipzig. 1886. His, W., Bettrdge zur Anatomie des menschlichen Herzens, Leipzig. 1900. His, W., “ Lecithoblast und Angioblast der Wirbeltiere,” Abh. d. math.-phy. Klasse d. Kyl, Stich, Ges, d, Wiss., Bd. xvi., Leipzig. 1910. Huntineron, G. S., “The Phylogenetic Relations of the Lymphatic and Blood Vascular System in Vertebrates, and the Genetic Principles of the Develop- ment of the Systemic Lymphatic Vessels in the Mammalian Embryo,” Anat, Rec., vol. iv. p. 339. »” »” Development of Blood-vessels and Heart in Ferret Embryos 183 1914. Huntineton, G. 8., “The Development of the Mammalian Jugular Lymph-sac, of the Tributary Primitive Ulnar Lymphatic, and of the Thoracic Ducts from the view-point of recent investigations of Vertebrate Lymphatic Ontogeny ; together with a consideration of the Genetic Relations of Lymphatic and Hemal Vascular Channels in the Embryos of Amniotes,” Amer. Jowr. Anat., vol. xvi. 1914. Jounstonz, R. W., “Contribution to the Study of the Early Human Ovum, based upon the investigation of (I.) a very early Ovum embedded in the Uterus,-and (II.) a very early Ovum embedded in the Infundibulum of the Tube,” Jour. of Obstetrics and Gynecology of the British Empire, May, pp. 231-276. 1907. June, Miinchen med, Wochenschr., Jahrg. liv. p. 1343. -1888. Kerpet, F., ‘ Die Entwicklungsvorginge am hinteren Ende des Meer- schweinchenembryos,” Arch. f. Anat. u. Phys. 1908. Keren u. Evze, Normentafel zur Entwicklungsyeschichte des Menschen, Jena. 1861. K6éuuiker, A., Yntwicklungsgeschichte des Menschen und der hoheren Thiere, Leipzig. 1882. K6uuiker, A., Die Entwicklung der Keimbldtter des Kaninchens: Fest- schrift, Leipzig. 1884. K6uuKer, A., “Die embryonalen Keimblitter und die Gewefe,” Zeitschr. Siir wiss, Zoologie, Bd. Xxxiv, 1904. Lewis, F. T., “The Intra-embryonic Blood-vessels of Rabbits from 83-13 days,” Amer. Jowr. of Anat., vol. iii., No. 1, pp. 12-13. 1908. Low, Auex., “ Description of a Human Embryo of 13-14 Mesodermic Somites,” Jour. of Anat. and Phys., vol. xlii. (ser. iii., vol. iii., part iii.), pp. 237-251. 1912. Matt, F. P, “On the Development of the Heart,” Amer. Jour. of Anat., vol, xiii. pp. 249-298. 1912. Maun, F. P., “A Human Embryo 26 days old,” Jour. of Morph., vol. v. pp. 459-480. 1909. Maximow, A., “ Die Frithesten Entwicklungsstadien der Blut- und Bindege- webzellen beim Siiugetierembryo, bis zum Anfang der Blutbildung in der Leber,” Archiv fiir mikr. Anat., Bd. 1xxiii. 1910. M‘Cuurs, C. F. W., ‘‘The Intra-intimal Theory and the Development of the Mesenteric Lymphatics in the Domestic Cat (Felis domestica),” Verhandl. der Anat, Gesellsch. ; 1912. M‘Crurg, C, F. W., “A few Remarks relative to Mr Kampmeier’s Paper on the Value of the Injection Method in the Study of Lymphatic Development,” Anat. Rec., vol. vi. 1914. MiniEr, A. M., and M‘Wuorrer, J. E., “ Experiments on the Develop- ‘ment of Blood-vessels in the Area Pellucida and "Embryonic Body of the Chick,” Anat. Ree., vol. viii. 1912. Minor, C. S., ‘The Origin of the Angioblast and the Development of the Blood,” Keibel and Mall, Manual of Human Embryology, vol. ii. p. 499. 1906. Mouuier, S., ‘‘ Die Entwicklung der Gefiisse im Embryo,” Handb. vergl. u. expt. Entwickl. d. Wirbeltiere, herausg. von O, Hertwig, Bd. i,, Kap. v., S. 1261. 1915. Parker, Karuarine, M., “The Early Development. of the Heart and Anterior Vessels in Marsupials, with special reference to Perameles,” Proceedings of the Zoological Society of London, pp. 459-499. 1886. Rast, C., ‘‘Ueber die Bildung des Herzens bei Amphibien,” Morph. Jahr,, Bd. xii. 1915. Reaaan, “ Vascularisation Phenomena in Fragments of Embryonic Bodies completely isolated from Yolk-sac Blastoderm,” Anat. Rec., vol. ix., No. 4. 184 Dr C. C. Wang 1892. Roptnson, A., ‘‘Observations upon the Development of the Segmenta- tion Cavity, etc.,” Quart. Jour. of Miscroes. Science, N.S., vol. xxxiii. 1902. Rosrnsoy, A., ‘‘ The Early Stages of the Development of the Pericardium,” Jour. of Anat. and Phys.. vol, xxxvii. p. 15. 1916. Ropryson, A., and Grson, A., ‘‘ Description of a Reconstruction Model of a Horse Embryo Twenty-one Days Old,” Transactions of the Royal Society of Edinburgh, vol. li., part ii. (No. 8), p. 340. 1904. Rouvikre, H., “ Etudes sur le développement du péricarde chez le lapin,” Jour. de V Anat. et de la Phys., xl. 1906. Rickert u. Mouser, “Die Entstehung der Gefasse und des Blutes bei Wirbeltieren,” Handb. vergl. u. expt. Entwickl. d. Wirbeltiere, herausg. von 0. Hertwig, Bd. i. S. 1019. 1892. Scumipt, M. B., “Ueber Blutzellenbildung in Leber und Mily, etc.,” Zeigler’s Beitrage, Bd. xi. 1915. Scuuurs, H. von W., “The Fusion of the Bilateral Anlagen of the Heart and the Formation of the Bulbo-ventricular Loop in Embryos of the Cat,” Proceed- ings of the Amer, Assn. of Anatomists, p. 72. 1889. SHore, T. W., “The Proamnion and Amnion in the Chick,” Jour. of Anat. and Phys., vol. xxiv. = 6. 1902. Soporra, J., ‘‘ Ueber die Entwicklung des Blutes, des Herzen und der grossen Gefiissstdmme der Salmoniden nebst Mitteilungen ueber die ase i der Herzform,” Anat. Hefte, 63 (Bd. xix., Heft 3). 1889. Sper, Grar, “ Beobachtungen an einer menschlichen Keimscheibe mit offener medullarrinne und canalis neurentericus,” Arch. f. Anat. wu. Phys., Anat. Abt., S. 159. 1896. Sprx, Grar, “ Neue Beobachtungen ueber sehr friihe Entwicklungsstufen des menschlichen Eies,” Arch. f. Anat. u. Phys., Anat. Abt. 1915. Srockarp, Cuarues R., “The Origin of Blood and Vascular Endothelium in Embryos without a Circulation of the Blood and in the Normal Embryo,” Amer. Jour. of Anat., vol. xviii., No. 2. 1915. Srockarp, Cuarues R., “A Study of Wandering Mesenchymal Cells on the Living Yolk-sac, and their Developmental Processes: Chromatophores, Vascular Endothelium, and Blood-cells,” Amer. Jowr. of Anat., vol. xviii., No. 3, pp. 525-594. 1883. Srrant, H., “ Ueber die Anlage des Gefiisssystems in der Keimscheibe von Lacerta agilus,” Maré. Sttzwnysber. 1899. Srricut, C. van per, “ L’origine des premiéres cellules sanguines et des premiers vaisseaux sanguins dans l’aire vasculaire de chauves-sauris,” Bull. de U Acad. Roy. de méd. de Belgique, t. xiii. p. 4. 1912. Tanpier, J., “The Development of the Heart,” Keibel and Mall, Manwal of Human Embryology vol. ii. p. 534. 1884. Torstic, J., “Untersuchungen ueber die Entwicklung der primitiven Aorten,” Schri/t herausgegeben v. d. Naturf.-Ges, bei d. Univ. Dorpat, Ba. i., Dorpat. 1892. Viauieron, L., “ Développement des aortes chez lembryon de poulet,” Jour. de VAnat. et de la Phys., t. xxviii, p. 1. 1913, Waniin, Ivan E., “A Human Embryo of 13 Somites,” Amer, Jowr of Anat., vol. xv., No. 3, pp. 319-332. 1915. Wart, J. Crawrorp, “ Description of Two Young Twin Human Embryos with 17-19 Paired Somites: Contribution to Embryology, No. 2,” Carnegie Institution of Washington Publication, No, 222, pp, 5-44. Development of Blood-vessels and Heart in Ferret Embryos 185 i EBLO, WEIDENREICH, F., ‘‘ Die Morphologie der Blutzellen und ihre Beziehungen zu Einander,” Anat. Rec., vol. iv. p. 317, 1914. Witson, J. T., ‘ Observations upon Young Human eateyes, ” Jour. of Anat. and Phys., vol. xl viii. (series iii., vol. ix., part iii.), pp. 315-351. ‘1911. Yeates, Tuos., “ Studies in the Embryology of the Ferret,” Jour. of — — Anat, and Phys., vol. xliv. (ser. iii., vol. vi.), pp. 319-335, 1915. Yeares, THos., “Studies in the Embryology of the Ferret,” Studies in ——— Anatomy, Birmingham University, pp. 71-107. 1887. Zizcuer, H. E., “ Die Entstehung des Blutes bei Knochenfischembryonen,” Arch. f. mikr. Anat., Bd. xxx. 1892. Zizauer, H. E., und Zircier, E., “ Beitriige zur Entwicklungsgeschichte von Torpedo,” Arch. f. mikr. Anat., Bd. xxxix. VOL. LI. (THIRD SER. VOL, XIII.)—JAN. 1918. 13 THE REPRODUCTIVE ORGANS OF CETACEA. By Professor ALEXANDER MEEK, Armstrong College, Newcastle-on-Tyne. THE Cetacea have become modified for aquatic life, and it has been an interesting task for the morphologist to trace the changes which have been brought about by a more and more complete adaptation to the habitat. The reproductive organs are peculiarly interesting in this respect, for all the events connected with reproduction take place in aquatic and pelagic conditions. The general disposition of the organs and their general morphology - are already known from the investigations made by Weber, Guerin, Rapp, Van Beneden, Hunter, Owen, Cleland, and recently by Hepburn; but the modifications they present in relation to function do not appear to have been clearly stated, and I find it necessary, therefore, to give a re- statement of the facts of structure with reference to some of the common Deiphinide. About 1889-1890 I had the opportunity of examining male foetuses of various toothed whales in the laboratory of Professor D’Arcy W. Thompson, Dundee, and it was the intention to publish the results in the series of researches which were at that period issued from his department. But financial difficulties intervened and the series came to an end. I have the permission of Professor Thompson to incorporate in this paper the figures which were made so long ago. This year, 1917, the fishermen at Cullercoats have captured many porpoises in the salmon drift-nets, and the investigation of some of these gave the opportunity of re-examining the structural modifications. | Although this species is the one which has been most frequently examined, it will be convenient to describe as fully as is necessary the organs in both sexes. The sexes are easily distinguished. In the female the anal and vulval openings are close together, and may indeed occupy the same recess. On each side of the vulva is the slit-like opening in which is lodged the teat of the mammary gland. In the male the similar opening of the preputial pouch is centrally situated, and the anal opening is far removed from it posteriorly. In front of the latter is a small aperture in which the reduced teats of the mammary gland are situated. The Reproductive Organs of Cetacea 187 Phocena communis. Female. (Figs. 1 and 2.) The females examined, four in number, measured 4 ft. 3 ins. to 4 ft. 5 ins. in length, and the anterior end of the vulval opening was situated about 2 ft. 5 ins. to 2 ft. 7 ins. from the anterior end. The vulva varied _ from 3 to 44 ins. in length. The anal opening is usually separated from the vulval opening by a ridge, but, as has already been stated above, it may actually be enclosed in the vulval aperture. In all respects, except at the uterine end of the vagina, the organs have a simple structure. The elongated, only slightly wrinkled ovaries are situated on the dorsal proximal aspect of the Fallopian tube. The Fallopian tube is much coiled, and ends in a uterus which, meeting its fellow, descends without fusion to open just above the os in a small part of the uterus common to both tubes. The cervix descends into the vagina as a short cylindrical structure. It is beset externally with numerous deep, longitudinal folds which are continued into the canal leading from its free flattened vaginal surface to the uterus. The vagina is much modified in the region next to the uterus; the ventral wall is continuous with, or occupies practically the same plane as, the uterus, but the dorsal wall is distended into a cavity which forms a prominent swelling. Associated with this expansion are certain folds of the vaginal walls, which clearly subserve an important function. The transverse folds of the mucous membrane of the upper part of the vagina are highly developed, and in addition two immense folds, almost encircling the vagina, convert the upper part of the tube into a recess which may be called the spermathecal recess (fig. 1). The upper fold springs from the ventral wall, thus serving to occlude the cervix and the os. The other is a similar transverse outgrowth of the dorsal wall, which carries the opening from the vagina into the spermathecal recess to the ventral side. The folds which define this chamber are clearly liable to be moved backwards and forwards, but it is important to note that in their normal disposition they form a passage from the vagina proper to the sper- mathecal recess of the vagina, which begins as a narrow opening on the ventral side and is continued between the two folds as a passage which opens into the recess dorsally. Immediately below the folds the vagina is a spacious cavity, wrinkled internally by transverse and longitudinal shallow grooves, and it has a smooth and velvety surface. The transverse grooves are more prominent in the upper region of the vagina, and the longitudinal in the lower, a few of these latter extending into the vulva. The aperture between the vagina and vulva is a very small one. The sphincter vagine muscle is well 188 Professor Alexander Meel developed, encircling the tube in this region, and under usual conditions practically closes the passage. The external opening extends from the anus, or immediately in front of the. anus, forwards as a long slit leading into a flattened, roomy cavity, rapidly narrowing to the opening into the vagina. The lateral walls are more or less grooved, and the anterior angle is occupied by the clitoris. The latter forms a prominent ridge, and ends near the vaginal opening in a distinct glans-like projection. Anteriorly it divides into two arms or crura, which, however, have nothing to do with the corpora cavernosa. Each corpus arises from the pelvic bone in the same region as that of the male, and, descending on the internal side of the erector muscle, the two fuse as urethra. blood vessels » 4 a) —_ Fic, =f, Longitudinal sectional view of reproductive organs of feniale porpoise (half natural size) ; 2, sectional drawing of clitoris to show disposition of corpus cavernosum (natural size). the muscles do above the clitoris. The single corpus cavernosum, resulting from the fusion, enters the clitoris near the middle of the body of that organ and is bent backwards towards the glans. It presents another distinct curve in front of the glans, which it enters, after narrowing con- siderably, and occupies its ventral free region. The corpus in the clitoris is surrounded by a thin sheath and by erectile and fibrous tissues. The disposition of these structures will be found to be important when the structure of the penis is considered. The glans, it will be noted, is not a true glans, as it is traversed by the corpus cavernosum ; but its position and appearance serve to indicate that it is the glans invaded and reinforced by the corpus cavernosum. The clitoris is surrounded by a preputial fold which, beginning above the glans in the urethral prominence, extends on each side to the anterior end of the clitoris, where the two folds approxi- The Reproductive Organs of Cetacea 189 mate but do not fuse. The folds are especially prominent anteriorly. The disposition of the preputial fold is interesting, since it homologises the external part of the clitoris with the free part of the penis which occupies the preputial pouch of the male. Immediately beyond the clitoris is the opening of the urethra, defined as a medium, short, longitudinal opening between two transverse grooves. It occupies the anterior part of a median prominence just internal to the glans clitoris, and which marks also the beginning of the preputial fold. As has already been remarked, on each side of the external opening there is a slit which lodges the cylindrical teat of the mammary glands, on the free end of which there is one opening. erector clitoridis sphincter vaginae. levator ani reputial clitori , ore pouch : urethra anus Fig. 2.—Pelvic and associated muscles of female porpoise. (Natural size. ) The muscles of this region are interesting (fig. 2). The pelvic bone is smaller in the female than in the male, but associated with it are the usual muscles and others connected with the-external opening. The erector clitoridis springs practically from the whole outward border of the bone. The two muscles practically surround the corpora cavernosa at their origin, the: mesial side next the vagina only being uncovered, and the muscles fuse as the corpora do, just above the clitoris, both muscles thus being concentrated in the sheath of the corpus and the neighbouring region of the clitoris. The levator ani arises from the internal border of the pelvic bone and is inserted into the lateral walls of the lower part of the rectum. The sphincter vaginz is well developed. It encircles, and is con- nected to, the lower part of the vagina, and it also embraces the urethra. The lower half only is seen in fig. 2, the upper half being covered by the 190 Professor Alexander Meek erector muscle. Below it comes into close association with the clitoris. The retractor clitoridis arises from the rectum, and is associated at its origin likewise with the sphincter; and passing forwards in a tunnel between the erector, the levator ani, and the sphincter, as a long tendon, it expands into a muscle above the clitoris and is inserted at the base of the preputial pouch. The female organs are thus, apart from the spermathecal folds of the upper part of the vagina, simple and typical. As will be seen, their condi- tion, and especially the presence of the spermathecal recess, helps us to understand the remarkable modifications in the male. The narrow entrance to the vagina may be said to be an adaptation to occluding the cavity of the vagina from the sea, and the spermathecal recess to preventing the semen leaving the vagina on the withdrawal of the penis. The spermathe-_ retractor ligament Fic. 3.—Diagram to show disposition and relative size of mature male organs of porpoise measuring 4 ft. 5 ins, cal recess is a remarkable modification, since it has taken place in spite of the equal periodical difficulties connected with the birth of the young. The deep folds of the upper part of the vagina permit of considerable dis- tension, and are evidently modifications connected with birth. Phocena communis. Male. (Figs. 3-7.) In the male the testes undergo an enormous development in the summer, as will be apparent from fig: 3, which indicates the relative disposition of the organs in a mature male measuring 4 ft. 5 ins. They also increase — greatly in weight; one testis in this male weighed 500 grams,in another 600 grams, and in still another 700 grams. On the other hand, in a ‘male measuring 4 ft. the- testis was only 2} ins. long and the weight 1-2 grams. Associated with the huge testis the epididymis is highly developed. The duct becomes much folded posteriorly to the testis, and the folds are still retained as the vas deferens passes to the space between the bladder and the ureter. The two vasa deferentia approach one another behind the bladder and pass downwards. side by side to open separately The Reproductive Organs of Cetacea 191 on the eminence of the floor of the urethra, called the verumontanum. Between the openings, and slightly posterior to them, the verumontanum presents another, or a pair of openings which lead into a blind tube lying between and behind the terrninal ends of the vasa deferentia. The opening is crescent-shaped, and may be single or paired, but in each case the cavity is a single one occupied by a slight ridge projecting from the floor. From its position this tube, with its single or double opening, answers to the uterus masculinus or sinus pocularis. In the porpoise the sinus has usually two apertures on the verumontanum. The vasa deferentia and the epidi- dymes are highly developed and are capable of containing a large quantity of sperms. It is to be noted that seminal vesicles are absent. In this region of the urethra the numerous ducts of the prostate glands open along parallel lines in the depression on each side of the veru- montanum, below the level of the openings of the vasa deferentia. They thus lie near the angle where the first sharp bend of the urethra takes place. The verumontanum begins at the base of the bladder as a narrow ridge expanding to receive the ducts above mentioned opposite the prostate glands, and it narrows again in the region occupied by the ducts of the prostate gland. Cowper’s glands are absent. The urogenital duct beyond the prostatic region is surrounded by penial structures, the disposition of which is highly interesting. The duct itself is enclosed by a corpus spongiosum, which, beginning in a bulb, extends to the tip of the penis. The corpora cavernosa arise from the pelvic bones, and in passing between those bones unite almost immediately to form a single corpus cavernosum, which extends likewise to nearly the end of the penis. The corpus cavernosum consists of a vascular part surrounded by a tough fibrous sheath, and the disposition of each of these in the various regions will be plain from an inspection of fig. 6. The sheath expands postero-medially near its origin into a process, rectangular in section. The process invades the accelerator muscle; the two processes approach one another below the prostate gland and almost meet behind it. In this region the processes are connected by a thin fascia. It will be seen from fig. 4 that at the anterior end the corpus cavernosum is suddenly reduced to a rounded shape together with its sheath, and is thrown into a distinct fold, being bent downwards and to the right side before running forwards to the end of the penis; the meaning of this will be discussed presently. The anterior end of the penis is enclosed in a preputial pouch, the skin of which, both that of the pouch and that covering the penis, is remark- ably thin. That of the pouch is thrown into longitudinal folds allowing of lateral distension. The opening is like that of the vulva, but it is situated practically in the middle of the body, and the pouch is also con- 192 Professor Alexander Meek stricted to a narrow opening at about the same depth as that of the vulva. The free part of the penis occupying the pouch is, because of the bend in the corpus cavernosum mentioned above, sharply resolved into two regions— a proximal, continuous with and having the same structure as the main body of the penis, and a distal, which is long, narrow, and almost round in section. The urethra opens by a longitudinal slit on the ventral aspect of this latter part of the penis near the tip (figs. 4 and 6). The whole structure, from the prostatic part of the urethra to the free part of the penis_in the resting condition, is thrown into (1) the fold mentioned above at the origin; (2) the double fold shown in fig. 3, involving adrenal kidney Gaseeren ae =e eS epididymis orpus cavernosum mm i) a a lly, a 7 < > 2 retractor ieaene Z preputial pouch Fm 12 3 4567 8 9 Fic. 4.—Longitudinal sectional view of reproductive organs of male porpoise. (Half natural size.) c¢ cav., on the right, points to the tendon-like process of the sheath of the corpus cavernosum which invades the accelerator muscle. nearly the whole of the proximal part of the penis behind the preputial pouch. The corpus cavernosum of the clitoris presents similar folds; compare figs. 1 and 4, The penis is highly developed, and the muscles associated with it are strong and massive. The erector penis muscle arises from the whole length of the large pelvic bone, and is inserted into the sheath of the corpus cavernosum as far forwards as near to the beginning of the second fold mentioned above. ‘The two muscles enclose the corpora cavernosa at their origin, and each corpus is also attached to the pelvic bones by a strong ligament extending to near the anterior extremity of the bone. The accelerator urinw muscle is especially strongly developed around the bulb and the prostate; the fibres are concentrated in the process of the sheath The Reproductive Organs of Cetacea 193 of the corpus cavernosum mentioned above. The processes and the fascia connecting them serve to separate the prostate gland above from the bulb ut. masc, ureter 4 Fie. 5.—General dissection of the reproductive organs of male porpoise. 1, The general relationship of the organs in the adult, but immature, state; 2, the preputial pouch and penis ; 3, the bladder and verumontanum region of the urethra ; 4, dissection of verumontanum. below. A strong, double, elastic retractor ligament extends from the rectum, and the transverse process of the sheath of the corpus cavernosum, forwards 194 Professor Alexander Meek below the accelerator muscle to be inserted into the base of the preputial pouch. This is the general structure with reference to the penis in the retracted 1. 0.6cm. 2. 0.8cm. 3. 2.5cm. Cie ies > ) priate g elena arene 10, 12.0cm. 11. 7.Ocm. Fic, 6.—Transverse sections of penis of porpoise. The numbers below fig. 4 indicate the levels of the successive sections. The sheath of the corpus cavernosum is unshaded. condition, when it is withdrawn within the preputial pouch. In a state of erection the penis may be protruded to a length which brings the tip as far forwards as opposite the pectoral fin, even the pectoral girdle. As will be seen from fig. 3, its direction during erection is forwards and only at a The Reproductive Organs of Cetacea 195 small angle outwards; and as will be seen further from fig. 7, the erection brings about a remarkable change in the disposition of the proximal and distal parts of the preputial portion of the penis. They become still more sharply defined, the proximal part being stiff and hard, like the main part of the penis posteriorly, while the distal, terminal part remains pliable, especially where the two parts are joined together. In attaining this ‘condition a corkscrew-like movement is given to this terminal = of the penis. It has already been noted that the vagina is relatively short. In the female figured it measured, from the opening of the vagina to the second fold, 7°5 cm., and in another female about 9 cm. The latter is just about the length of the distal end of the erected penis. It is evident, therefore, that this is the only part of the penis which enters the vagina. The pairing of these creatures must be a difficult proceeding, and the Fre. 7.—The preputial part of the penis when erected, above structural modifications are meant to make it effective. They apparently range themselves in ventral apposition in such a manner as to bring the respective openings together. The head of the female will therefore be in advance of that of the male. The latter may be able to assist matters with his pectoral fins, and the female by flexing the caudal fin so as to form a groove for the body of the male. When erection takes place, the first part of the penis to emerge is the narrow, free end, which will therefore enter the vulva, and the rotation which is being communi- cated to it will enable it to gain access to the vagina; the blunt end of the proximal part of the penis will abut against the front wall of the vulva, adapting itself to the clitoris. The direction of the free end of the penis is such that it will pass the first fold of the vagina and bend upwards the second. The outlet of the penis is on the ventral side near the tip, and consequently the discharge is projected into the spermathecal recess. The bulb is compressed as well as the prostate by the accelerator muscle, and it is evident, therefore, that the terminal part of the penis is subject to a slight expansion during ejaculation. It is evident, also, that as a result of the erection the female is carried forwards and some little distance away 196 Professor Alexander Meek from the male, and at this period the bodies of the two may be subject to some degree of rotation. This is met by the joint in the penis at the base of the copulatory part, allowing for a wide range of movement. From this it may be gathered that discharge follows immediately, for any movement will likely result in separation. When the discharge has taken place and erection ceased, the retractor ligament withdraws the penis once more into the preputial pouch. In all the males examined a discharge, rich in spermatozoa, could be obtained by simply pressing the penis, and sometimes even without. Living sperms were obtained in this way a day after death. The immature males similarly yielded a discharge, but the ejaculate did not contain sperms. In both cases the ejaculate contained other cells, as prismatic cells from the vasa deferentia and lymphocytes. The immature males are apparently capable of conjugation, for they have been captured in the same part of the net as the females. The fact that the urogenital duct is filled with the spermatic fluid in the mature male, and with a discharge in the immature male, points to a recent copulation. But this simple and obvious conclusion does not meet the conditions. A male was brought in injured but alive, and remained in the floor pool for a couple of days before it died. In this case also the urogenital duct was fully occupied by sperms. It cannot be said, then, that copulation had taken place sufficiently recently to account for this, for urination must have occurred in the meantime, and that this was so was indicated by the small quantity of urine in the bladder. We may take it that in the act of micturition the bladder is emptied, and it is the function of the accelerator urine muscle to clear the urine from the urogenital duct of the penis. The difference between this act and that of discharge during copulation is that the pulsations of the accelerator eject the contents of the urogenital duct, which is filled immediately by the simultaneous peristaltic contractions of the vasa deferentia. In the case of urination there is no supply to draw upon; in the other there is a supply. We can thus account for the penial part of the duct being filled with sperms after copulation, but it is evident that the duct will be cleared of sperms at the next urination. We can only account, then, for the presence of sperms in the urogenital duct in the above cases by a choice between two explanations. The first is that the pressure of sperms into the duct is steadily maintained by the contraction of the vasa deferentia apart from copulation; the second is that erection and discharge has taken place as a result of drowning. The first explanation is not satisfactory, for in the resting condition the penis is not at all likely to be thus so completely occupied by the spermatic fluid injected. It is not likely, either, that the duct will be in this The Reproductive Organs of Cetacea 197 way filled and emptied at every act of urination. It is also necessary to take note of the fact that in no case had the females associated with the males been in copulation, for the vagina and the spermathecal recess were free of sperms. Yet in the case of two of the females a discharge was freely issuing from the narrow opening of the vagina, bathing the clitoris. In all the females the spermathecal recess was occupied by a thick, viscid uterine discharge, although the latter may have had no connexion with the immediate act of copulation. The absence of sperms in these added evidence of significant value pointing to the second explanation, and that both sexes are affected. There cannot be any doubt, however, that pairing takes place in the summer, although, so far, we have not obtained a female which shows that it had actually taken place. This may be due to the female immediately after the event quitting the inshore waters. The penis, then, is greatly modified in the porpoise, the copulatory part being differentiated from the rest, from which it is separated by a joint allowing of a wide range of movement. In the state of erection it is still pliable, and it may therefore be said to be adapted to obtaining easy access to the vagina, to changes of position of the bodies of both the male and the female, and also to yielding to pressure if access to the vagina has not been obtained. It is probable, therefore, that the preputial poudl has a vaginal import- ance, and, in the case of successful copulation, the copulatory part of the penis is, because of the continuity to begin with between the preputial pouch and the vagina, prevented from coming into contact with the water, as the rest of the free part of the penis must do during erection. The prevalence of porpoises near the shore during the summer months when drift-net fishing is taking place, and their absence at other seasons, have led to the conclusion that their advent is due to their feeding on salmon. In the case of all those that were examined it is interesting to point out that food was only obtained in the stomach of two, a female and an immature male, and in each case it was gadoid, probably young whiting: much digested. There is no doubt that at this season, July and August, the porpoises are pairing, and their appearance near the shore at this period may be described, therefore, as a migration for that purpose. It is also fairly clear that in the region where the salmon-fishing takes place males are more commonly captured than females. It seems warrantable to conclude that the males frequent the inshore waters and the females enter the region when they come into season, departing after pairing has taken place. Mr W. J. M. Menzies of the Scottish Fishery Board has informed me that 198 Professor Alexander Meek four porpoises caught in the salmon nets on the shores of the Moray Firth and which he examined, were all males. Porpoises are common both in the Atlantic and Pacific, and no doubt in both. regions an inshore migration takes place for the birth of the young and for pairing. Lagenorhynchus albvrostris. An abnormal male. This white-beaked dolphin, which measured 5 ft. 2 ins., was caught in the salmon nets on the north side of the mouth of the Tyne, and was landed at Cullercoats on August 24, 1911. The urogenital organs were removed and preserved by Mr Storrow. They were labelled “male,” and they prove on examination to be essentially male, but they are so abnormal in position and.structure as to warrant a somewhat detailed description. In the first place, the preputial pouch, instead of being central in position, is far back, close to the anus, thus in the position of the vulva. The free part of the penis is very short, only 3} ins. long, but has the gradually tapering shape and the structure of the species. The pouch is proportion- ally short; its wall is furrowed, but is more like the vagina in structure. It is constricted externally into a narrow neck, and the anterior angle of the external opening is occupied by two ridges which almost suggest a clitoris in appearance. The position of the mammary glands could not be ascer- tained, as the posterior part of the pouch had been removed, but it is more than probable that they were paired and situated on either side of the preputial pouch. In association with the short pouch and the near proximity of the rectum, the retractor ligament only measured about 1? in. It was inserted at the base of the preputial pouch. The proximal part of the penis between the preputial pouch and the prostatic region is also strikingly short, although it presents the usual folds of the normal condition. It is flexed ventrally from the region of the prostatic ducts, and is then directed — dorsally to the left side and again to the right before it enters the pouch. This is the normal disposition, but the whole structure is dwarfed. The urogenital duct is that of the male, and ends in a horizontal slit terminally ; and this is also the normal condition. A small prostate is present, and the ducts open as usual below the openings of the verumontanum. The expanded portion of the verumontanum presents the paired openings of the vasa deferentia, and below them the A-shaped opening of the uterus masculinus. But what makes the specimen particularly interesting is the presence of small testes, each with an epididymis and a coiled Wolffian duct, together with a Miillerian duct. The Miillerian ducts meet and fuse, but they are altogether unconnected with the uterus masculinus, which latter has the normal disposition (fig. 9). It is difficult to imagine the circumstance which The Reproductive Organs of Cetacea testis vas def‘ > corp. cav.Q)\ retrac tor penis prepu ial pouch Fic. 8.—The general disposition of the organs of an abnormal male Lagenorhynchus albirostris, in ventral view. (Half natural size.) v., Verumontanum ; v.d., opening of vas deferens; u.m., uterus masculinus ; p.g., opening of prostate gland ; m.d., Miillerian duct. 199 200 Professor Alexander Meek has led to a disappearance of the oviduct in the interval separating the uterus masculinus from the Miillerian duct, and this feature raises, therefore, some degree of doubt as to the identity of the sinus pocularis with the Miillerian duct. The bladder and the kidneys and ureters are in striking contrast (as will be seen from fig. 8) with the rest of the structures as regards size. Sections were made of the testis at different levels. They were found ~ ,Millerian duct. ‘vas deferens uterus masculinus Fic. 9.—The relationship of the Miillerian and Wolffian ducts in the abnormal Lagenorhynchus. to present numerous tubules of the usual character, but no ova were to be seen. Lagenorhynchus albirostris. The normal male organs. Fortunately, I have drawings of the normal male organs of a foetus of Lagenorhynchus albirostris, which are produced here in fig. 10, and which will be described briefly with especial reference to the abnormal specimen an account of the organs of which has just been given. These were made long ago, as has been stated, in the Zoological Department of the University College, Dundee. The disposition of the organs in the body is already well known,! and it is almost exactly the same as in the porpoise. The penis, it will be observed, is situated forwards from the anus, and in front of the latter are the small paired openings of the mammary glands. Internally, the organs except in their general relationship, are very like those which have just been described in the abnormal specimen. The testes are elongated, the epididymes are large, each projecting beyond the testis in a caput epididy- mis (the latter is somewhat exaggerated in the abnormal specimen). ‘The ! Hunter, Structure and Qeonomy of Whales, Cleland, Jowrn. Anat. and Phys., vol, xviii. p. 334, The Reproductive Organs of Cetacea 201 vasa deferentia pass as usual from the epididymes to open separately in the verumontanum. The uterus masculinus has a A-shaped aperture immedi- ately below the openings of the vasa deferentia, and the shape of the adrenal = E> ureter\ preputial pouch j penisf» ; 2 / \- aves deferens i, testis/ J ‘a y, © 4 a Ve rumontanum mA-vas def. suterus masculinus wo/vas def, ut, mn. » vas def, + * 7 ee Fic. 10.—The normal] male organs of a feetal Lagenorhynchus albirostris. 1, The disposition of the penis, mammary glands, and anus ; 2, the penial muscles ; 3, the organs from the ventral aspect ; 4, dissec- tion of verumontanum. ee NS anus mam, gland opening is also that of the lumen of the structure, for a crest projects into the cavity from the floor throughout its length. The crest becomes gradu- ally reduced towards the blind end of the single sinus. Cleland (loc. cit.) stated that in the white-beaked dolphin “there are four openings into the floor of the first part of the urethra. The two VOL. LIIl. (THIRD SER. VOL. XIII.)—JAN,. 1918. 14 ® 202 Professor Alexander Meek upper are the openings of the vasa deferentia; and immediately beyond these are two larger openings distinct one from the other, and both leading up into a single sinus pocularis.” Such a variation, it has been noted above, is also found in the porpoise, the cavity in both cases being a single one, but the opening single or double. The ridge, it will be noted, on the floor is more prominent in Lagenorhynchus than in Phocena. The verumontanum, in both the normal and the abnormal specimens, is a longitudinal ridge on the floor of the urethra. It begins near the base of the bladder as a double ridge on which the ureters open, fuses to a single ridge, and expands opposite the prostate gland to receive the openings just mentioned, and narrows again to gradually disappear in the urogenital duct. On either side of this latter narrow part the numerous openings of _ the prostate glands occur in rows. The general structure and musculature of the penis are the same as those of the porpoise, and require no further special description. The urethra opens, as has been stated, in both the normal and the abnormal specimens, terminally by a horizontal slit. It will be seen from the drawings that the free part of the penis rapidly narrows, but it is occupied throughout its length by a single corpus cavernosum, which ends some little distance from its extremity. With regard to this, Cleland (loc. cit.) made the significant observation that “in the white-beaked dolphin the combined corpora caver- nosa, about 2 ins. from the free extremity, are suddenly compressed as if by removal of more than half their substance on the ventral side, and then expand again to taper to the point.” This is so like the formation of the penis of the porpoise as to give the impression that the function is ‘ essentially the same. Cleland added, “It is such an arrangement as will allow the slender end to be folded down on meeting with resistance in the erect condition”; but, as has been pointed out, the jointed end of the penis of the porpoise is the only part which gains access to the vagina, and it may be that this is the case also with reference to Lagenorhynchus. Notes on the Male Organs of other Species of Delphinida. A few notes may be added here with reference to one or two other species which were dissected in Dundee. In a foetal beluga, Delphinapterus (Beluga) leuwcas, about 12 ins. long, the penis was Pound to be rounded and tapering as in Lagenorhynchus. The verumontanum presented three openings, the vasa deferentia and the uterus magsculinus. ‘he latter is interesting, as distally it is stretched out into a flattened cavity (fig. 11), A fcetal narwhal (Monodon monoceros) was dissected, and the general disposition of the organs is shown in fig. 12, and from these it will be noted that the penis has a similar shape to those already described. The Reproductive Organs of Cetacea 203 From the drawings in my possession the conditions in the dolphin (Delphinus delphis) appear to be fundamentally different. The penis is Fic. 11.—The vasa deferentia and uterus masculinus of a foetal beluga (Delphinapterus leucas), evidently short and only slightly tapering (fig. 13). There is some evidence in the drawing of a change in the corpus cavernosum near the extremity similar to that described in the other species, but it would be necessary to kidney’ rectip ire Fic. 12.—1, A general dissection of a foetal narwhal (Monodon mouoceros) to show disposi- tion of male organs ; 2, the reproductive organs in ventral view. It will be noted that the bladder has been reflected backwards, examine the species afresh to establish the point. The pouch is evidently proportionally short. There appear to be only two openings on the veru- montanum, those of the two vasa deferentia The two openings are ap- proximated into a crescent-shaped aperture leading into the separate vasa 204 Professor Alexander Meek deferentia, a narrow septum separating the terminal portions of the ducts, in which no trace of a uterus masculinus was found (fig. 14). | The penis of Kogia breviceps was described by Benham,’ and it is evident that in it the corpus cavernosum suffers a reduction in size anteriorly, Vas, def, retractor| | Fic. 13.—Penis and preputial pouch of Fic. 14.—The verumontanum of the dolphin (Delphinus delphis). dolphin. becoming circular in section. This modification of the preputial part. of the penis appears therefore to be general in whales. GENERAL CONSIDERATION. A. Morphological. The reproductive organs of the Cetacea, so far as may be gathered from a consideration of the foregoing species, are of a generalised mammalian type, with special modifications connected with the pelagic life of the group. It is in connexion with the muscles relating to these organs that the pelvic bone has been preserved, and it is worth noting that it is smaller in the female than in the male. It would be interesting to know to what extent the spermathecal recess is developed in the females of other species. The separation of the penis of the porpoise into a special copulatory jointed to a non-eopulatory part is a remarkable modification, and has been led up to, doubtless, by the penis becoming narrow anteriorly, as is the case ' 1901, Proc, Zool, Soc. (ii.), p. 128. The Reproductive Organs of Cetacea 205 in so many toothed whales. There is no true glans, but it appears to be the fact, from a consideration of the structure of the clitoris, that the glans has been secondarily invaded by the corpus cavernosum. In other respects the organs do not arouse a great deal of interest. The absence of Cowper’s glands is certainly worthy of remark, for they are so constantly presentin mammalia. They are absent also in Arctoid Carnivora. The absence of seminal vesicles is not of such moment, for it is a feature which is not uncommon in primitive types, as monotremes, marsupials, the sloth, and in seals and other Carnivora. It is not necessary, in the case of a . group so modified as the Cetacea, to make a point as to the abdominal position of the testes. The uterus masculinus is generally present as a single tube lying between the terminal ends of the vasa deferentia. The opening of this single sinus pocularis is usually one in the feetal stage, but may be, and usually is, double in the adult. The condition of the sinus in association with independent Miillerian ducts in an abnormal male Lagenorhynchus is noteworthy. A sinus pocularis does not appear to be represented in primitive mammals, and its presence here in a reduced condition is thus of interest. But the question of its homology will have. to be reconsidered. The preputial pouch must be considered also as a primitive feature, and in relation to the condition of life of the Cetacea its presence is plain. From the fact that the penis is everted in some of the (and especially in the younger) foetuses, it may be held to be secondary in origin. Associated with the pouch is a retractor ligament which in some cases appears to be more or less muscular in composition. In the porpoise it is a strong double elastic ligament extending from the rectum to the preputial pouch. Both the pouch and the retractor muscle are represented in the female. The muscle is especially interesting, for it originates in a tendon, expands above the clitoris into a muscle, and narrows again to be inserted into the prepuce. The presence of a retractor may also be said to be a primitive feature. The erector penis and the accelerator muscles are strongly developed in the male. The erector clitoridis, though small in comparison with the corresponding muscle of the male, is well developed, and the sphincter vagine (accelerator) is also a prominent muscle encircling the opening of the vagina. It comes into association with the erector and also with the levator ani. Sebaceous glands occur in the preputial pouch in both sexes, and their secretion in the case of the male is always evident. 206 Professor Alexander Meek B. Physiological. It is more than probable that the act of micturition necessitates the protrusion of the anterior end of the penis from the preputial pouch. If this be the case, it means that the penis is erected to some degree. The bladder is strong and muscular, and no doubt can be completely emptied by the contraction of its muscular wall, and it is evident that this is followed by the long urethral duct being emptied by the action of the accelerator muscle. In the case of the female there appears to be little doubt that urination . is accompanied by an erection of the clitoris. As will be seen from figs. 1 and 2, the urethral duct, in the resting condition, points towards the posterior wall of the entrance into the vagina. The corpus cavernosum of the clitoris, as will be seen from fig. 1, is so constructed and disposed as to make it apparent that during erection the main part of it will be straightened, and this will have the effect of bending forwards and down- wards the glans, thus bringing the opening of the urethra to a ventral position and directed outwards. At the same time the distension of the whole structure will result in the clitoris and the urethral opening being carried downwards. In this case it is not necessary to consider that any muscular effort is required except that of the bladder, but the terminal end of the urethra is enclosed in the sphincter vagine. It has been suggested that the males frequent the inshore waters during the pairing season, and that the females enter the region as they come into season. Whether this be the case or not, it is evident that the migration is due to a periodical development of the gonads. During the summer porpoises are very common inshore, and they may frequently be seen swimming in pairs. Mennell and Perkins! refer to their “grotesque gambols on the surface of the water, especially during the herring season.” As has been stated, we have not yet come across a female which has been in copulation. But the attempt has been made to indicate how the act is accomplished in pelagic conditions, and what is said about the por- poise will apply in the main to the other Cetacea It is not necessary to repeat that description here, but a few words may be added as to the physiological changes which are more or less specialised in the porpoise and other Cetacea. The vulva is prepared for the reception of the penis by the descent of the erected clitoris, the glans being rotated forwards. This puts the ! Trans. Tyneside Field Olub, vol. vi. 1863-64, * A summary of the scanty knowledge we possess is given on p, 219 o0f Millar's Mammalia of Great Britain and Ireland, vol, iii. The Reproductive Organs of Cetacea 207 retractor clitoridis on the stretch, and this muscle may be said to join synchronously in the pulsations of the erector muscle. The erection of the penis similarly brings about an elongation of the retractor ligament. It is apparently a gradual process, for complete erection appears to be postponed until copulation. When the penis is protruded the free end of the copulatory part enters the vulva and is directed into the vaginal opening distending the sphincter muscle. The whole of the copulatory part of the penis gains access to the vagina; the main body of the penis, culminating as it does in a prominent knob, only reaches the vulva where it abuts against the clitoris. It has been noted above that in the case of two of the females a dis- vharge was seen to be issuing from the narrow opening of the vagina. The exact nature of the glands which give rise to this has not yet been determined. It is probable that in association therewith the sphincter muscle of the vagina is thrown into a pulsation. If this takes place, as is more than likely, on the withdrawal of the penis, water may thus be intro- duced into the vagina, but the folds which guard the spermathecal recess will obviate any danger of the ejaculate being lost. The levator ani in both cases may be said to help by keeping the parts taut. The presumption that erection and discharge may take place as a result of drowning raises interesting physiological questions as to the factors. It cannot be said to be the result of direct stimulation of either of the centres. It seems to be due to want of oxygen or perhaps more prob- ably to the effects of the general fatigue produced by struggling. An examination of other cases of the kind, in relation to this, may point to the cause. Notes ON LITERATURE. The peculiar disposition of the corpus cavernosum in the porpoise does not appear to have been described by any of the many authors who have examined the male organs of this species. The fact that there is no true glans has generally been noted. Beauregard and Boulart,! and also Daudt ? have found, as I have found, that the preputial pouch is a secondary formation ; but it is not necessary because of that to call the pouch pseudo- preputial as Daudt proposes. The musculature in the case of the Greenland whale has been described by Struthers,* and in the case of the Grampus by Turner.‘ 1 Journ. d. Anatomie et de la Physiologie, 18 ann., 1882. 2 Jenarsche Zeitsch., Bd. xxxii., 1898. 8 Journ. Anat. and Phys., vol. xv., 1881. * Ibid., vol. xxvi., 1892. 208 Professor Alexander Meek Daudt failed to see the preputial pouch in the female porpoise, but indicated its presence in other toothed whales, Beluga and Hyperoodon, and he observed it, as others have done, in Balenoptera. He appears, however, in some cases, to have confused the body of the clitoris with the preputium. Daudt has described the two folds which form the spermathecal recess in the porpoise, but without indicating their significance, and his figure does not clearly indicate their disposition. Watson and Young! found eight such folds in Beluga, and Daudt said there were ten, the cervix being the first of the series. With reference to this genus I can now say that the penis is pointed, the free portion being spindle-shaped. Sections of the penis of a foetal Beluga measuring 17°5 em. show that, as in Lagenorynchus, the corpus cavernosum suddenly becomes reduced in size and rounded in section. This reduced part of the corpus cavernosum springs from the dorsal region of the body, and extends near to the extremity. The penis, as in the porpoise and as in Lagenorynchus, is therefore resolved into a proximal fixed part and a distal flexible part. This is associated in the case of the Beluga female with a succession of structures which resemble the cervix in structure, and probably in function. With their development the cervix has disappeared as such, and is probably represented by the small fold at the uterine end of the vagina which marks the first of the series. Daudt found that in Hyperoodon there were four to five such folds, and in Balenoptera there were about ten. There seems to be some degree of variation in Baleenoptera with regard to the number and the disposition of the folds, for in the species described by Beauregard and Boulart the number is less than that described by Daudt, and the lower folds appear to undergo some degree of anastomosis. THE FEMALE ORGANS OF THE GREENLAND WHALE (BALANA MYSTICETUS). Since the foregoing was written I have examined the female organs of a foetus, measuring 41 cm., of the Greenland whale. This specimen was pre- sented to the Natural History Society of Newcastle in 1835 by T. R. Batsom, and is preserved in the Hancock Museum. So far as I know, the organs have not been described with reference to this species, and they present some features of interest. The vulva is quite small, measuring only 5 mm. in diameter, and is rounded in shape. The anal opening is situated about 5 mm. posteriorly, and the minute openings of the mammary glands are placed on either side but somewhat ' Trans. Roy. Soc, Edinburgh, vol, xxix., 1879. The Reproductive Organs of Cetacea 209 anterior to the vulva. The vulva was found to be occupied by a clitoris, which is elongated and coiled. It is thus so far like the penis in shape as to indicate that the penis is long and tapering, as it is in so many of the Cetacea.! The clitoris springs from the base of a shallow preputial pouch and it is evident, therefore, that it represents, as in the porpoise, the free preputial part of the penis. The preputial pouch completely surrounds the clitoris, and is continued by folds with the eminence on which the urethra opens. Beyond the vulva theré is a long, roomy vagina. Near the vulva the wall of the vagina is thrown into longitudinal folds, and these can be traced, especially on the dorsal side, along rather more than clitoris Fic, 15, —The urino-genital organs of a female foetus of the Greenland whale, measuring 41cm. (x 2.) half its length. Otherwise the lower part of the vagina has a smooth wall in which pit-like depressions occur at intervals. The upper part of the vagina is occupied by a series of transverse folds similar to those which have been described in the species mentioned above. There are seven complete rings, a trace of an eighth at the upper end of the series, and at the lower end two incomplete folds on the right and left sides. They are traversed by longitudinal folds, which are especially well marked at the upper end and are almost absent from the incomplete folds at the lower end. . As in the other cases in which such folds have been found, the condition is accompanied by the absence of the cervix. If it be the fact, as the structure certainly indicates, that the cervix has disappeared, we have in these cases the development of a series of folds which have replaced the 1 Struthers, Journ. Anat. and Phys., vol. xv., 1881. 210 The Reproductive Organs of Cetacea cervix in function. The vagina has thus been peculiarly modified in a large number of the Cetacea, in a manner which appears to be without parallel in other Mammalia, and is even strikingly pererene from the conditions in the porpoise. In the light of the foregoing description it may be suggested that they have a double function. The penis will have to pass through a number of the folds at least, and it is probable, therefore, that they have been developed for the purpose of promoting the discharge. It will be noted that the folds are directed downwards and they are highly muscular. It is possible, therefore, that the folds are expanded and contracted, and depressed and raised, and such movements*would have the effect of sending the spermatic fluid upwards, to and towards the uterus. Even in the case of the porpoise it is quite possible that the two folds are thrown into similar movements. The uterus a short distance above the folds divides into the two horns, which are bent ventrally and backwards. The Fallopian tube presents a few folds, and, like the large funnel into which it opens, is directed forwards. The ovary is large, and it is beset with numerous furrows as it is in Balenoptera. The broad ligament is occupied on each side by around | fat body, such as was found by Beauregard and Boulart in Balenoptera. THE PRIMORDIAL CRANIUM OF ERINACEUS EUROPAUS. By Epwarb Fawcett, M.D., Professor of Anatomy wn the University of Bristol | THE specimen from which the model illustrating this communication was made I owe to the kindness of Prof. J. P. Hill, of University College, London. In total length it was 25 mm., and was cut into sections of 25 microns thick, stained before cutting with alum-cochineal, and then on the slide with Mallory’s triple connective-tissue stain. Being an Insectivore, it was thought of interest to examine it in detail and compare it with the mole. The cranium may be divided into an axial and an appendicular, or a _ neural and a visceral part, neither division being quite satisfactory. The former is that more particularly associated with the encephalon, the latter the skeletal basis of the visceral arches. We may therefore consider seriatim the neural chondrocranium, the visceral cartilaginous skeleton, and finally the bones where they are found. ; THE PRIMORDIAL NEURAL CRANIUM.2 This is perhaps most conveniently considered as consisting of the following parts, viz. :— 1. A central stem, to which are appended, at more or less constant positions, 2. Appendages to the central stem ; 3. Lateral structures ; 4. Commissures binding these lateral structures together ; 5. Dorsal structures, i.e. structures which lie dorsal to the encephalon and form a cartilaginous roof to the cavum cranii. * 1. THE CENTRAL STEM. (Pls. 1, 2.) The central stem stretches at this stage uninterruptedly from the anterior margin of the foramen magnum to the tip of the nose. As far * The expenses of this research were largely defrayed by a grant from the University of Bristol Colston Society’s research fund. : re ace * * In order to justify this mode of description, several plates are appended. They ree the condition as seen in the 11-mm. mole, 12- and 17-mm. Tatusia, and ?19-mm us taurus. 212 Professor Edward Fawcett forwards as the nasal capsule it is much wider than deep; but once it has become included within the nasal capsule, its depth suddenly increases; its width at the same time diminishes, so that from now onwards as nasal septum the depth of the central stem is greater than its width (PI. 5). Constituent Parts of the Central Stem.—Morphologically the central stem appears to be made up of the following parts from behind forwards, viz.: (i.) a part in some relation to the notochord—the pars chordalis (Pls. 1, 2); (ii.) a part related of the hypophysis and its stomodeal duct—the pars trabecularis (Pls. 1, 2); and (iii.) a part which, lying between the optic foramina, then as traced forwards between the orbital cavities, and finally within the nasal capsule as the nasal septum, may be alluded to as a whole as the interorbito-nasal septwm or pars interorbito-nasalis (Pls. 1, 2). In Talpa (Pl. 22), and in Tatusia (Pls. 16, 17, 18) and Cavia, these parts chondrify independently of one another. Whether they do so or not in Erinaceus has not yet been determined, for want of material; but it is most probable that they do. In the calf (Pl. 13) and man (Pls. 19, 20) the pars chordalis chondrifies independently of the pars trabecularis, but I do not know if the latter is independent of the interorbito-nasal septum. At the stage now being described, as all these parts are fused together and no histological difference can be made out anywhere in the central stem, their sites of union can only be surmised. Detailed Description of the Central Stem.—The pars chordalis (Pls. 1, 2) may be looked upon as a quadrilateral plate, wide and concave behind as it forms the anterior boundary of the foramen magnum. The postero- lateral angles are fused with the exoccipitals, only the hypoglossal canal indicating any demarcation between the two, and there is no developmental distinction since the exoccipital is a direct outgrowth from the chordal plate; but, for later convenience, ossification taking place as it does inde- pendently in the exoccipital and extending medially towards the basi- occipital centre, one may speak of this cartilage as the exoccipital. The lateral margins of the chordal plate are partly free where they look towards the foramen jugulare and partly directly fused with the cochlear capsule ; anteriorly the chordal plate is fused with the pars _ trabecularis, and an imaginary line drawn transversely through the middle of the cochlear capsule may be taken as indicating the site of fusion between the pars chordalis and the pars trabecularis. The Pars trabecularis (Pls. 1, 2)—This, assuming its commencement to be as stated above, is of large size, and very slightly hollowed to receive from above the hypophysis. It may be regarded as a quadrilateral plate with an imaginary posterior border at the aforesaid junction with the pars The Primordial Cranium of Hrinaceus ewropwus 213 chordalis. Its anterior boundary is also an imaginary one, formed by drawing a line transversely through the central stem immediately behind the hinder root of the ala orbitalis (Pl. 1). From its postero-lateral angles on each side are given off the following parts: from behind forwards, first the posterior trabeculo-cochlear commissure, fused with the chordo-cochlear commissure (Pl. 1), next the anterior trabeculo-cochlear commissure (P1. 1) which runs back to the lateral side of the apex of the cochlear capsule (Pl. 1). Between the anterior and posterior trabeculo-cochlear commissures the carotid foramen (Pls. 1, 2) is found. Then follows the ala temporalis (Pls. 1, 2), which at this stage is fused at its root with the anterior trabeculo-cochlear commissure. The lateral border of the pars trabecularis is now free and forms the medial margin of the anterior segment of the spheno-parietal foramen or fontanelle (Pls. 1,2). In the centre of the pars trabecularis a large circular foramen is met with; this transmits the stomodzal duct and is the cranio-pharyngeal canal (Pls. 1, 2). It is to be noted that the*main mass of the hypophysis always lies behind this foramen. Two cartilages lie below the pars trabecularis, and are developed independently of the neural chondrocranium. These are the cartilaginous masses developed in connection with the pterygoid bone (Pl. 2). Each is a somewhat ovoidal mass whose long axis is directed sagittally. Each is placed infero-medial to the processus pterygoideus of the ala temporalis and separated from it by the vidian (parabasal) nerve. No pterygoid bone is developed at this stage in connection with it, but on its supero- lateral aspect a dense crescentic mass of connective tissue, which occupies the position one usually associates with the pterygoid bone, is found. This mass, too, lies immediately behind the os palatinwm. The mass of cartilage just described has on its outer side the tendon of the tensor palati muscle, and is evidently the cartilage of the hamulus. The Pars wnterorbito-nasalis (Pls. 1, 2,5, 6).—This in Erinaceus is of quite unusual width, and receives at its sides the two roots of the ala _ orbitalis, between which the optic foramen lies on each side. Perhaps the hinder root or limb of the ala orbitalis is really connected with the pars interorbito-nasalis through the ala hypochiasmata (P\. 2), but that could not be determined at this stage. Free of the two limbs of the ala orbitalis, the lateral free margins of the pars interorbito-nasalis now form the medial boundary on each side of the orbito-nasal fisswre (Pl. 1), then this interorbito- nasal septum becomes included within the nasal septum. The nasal septum will be described in detail with the nasal capsule. 214 : Professor Edward. Fawcett 2. STRUCTURES APPENDED TO EACH SIDE OF THE CENTRAL STEM. From behind forwards are :— (1) The exoccipital cartilage or pila occipitalis ; (2) The auditory capsule ; (3) The ala temporalis ; (4) The ala hypochiasmata and the ala orbitalis ; (5) The lateral nasal capsule. (1) The Exoccipital Cartilages (Pls. 1, 2). Each may be taken as extending from an imaginary line drawn through the medial margin of the hypoglossal canal in an antero-posterior direction. _Each passes backwards and outwards at first, having a free antero-lateral border which forms the hinder boundary of the foramen jugulare (Pl. 2), Can. semic. ant. Fossa subare, int. Ampnila can, semic, post. Vagus. N. hypogloss. Pars chordalis. Fic. 1.—Frinaceus ewropeus, All the sections are coronal ones. and a postero-medial border which forms the lateral boundary of the primary foramen magnum (Pls. 1, 2). Beyond the lateral limit of the foramen jugulare, the exoccipital cartilage is projected under the pars _ canalicularis of the auditory capsule in the form of lamina—the lamina alaris (P\. 2), which, for the most part separated by a narrow fissure—the The Primordial Cranium of Hrinaceus ewropwus 215 recessus swpra-alaris (Pl. 1 and fig. 3)—from the auditory capsule, is terminally fused with that capsule, so that the: fissure above mentioned can only be seen from within, and that only after removal of the lateral sinus. The recessus supra-alaris is itself occupied by dense cellular tissue. At its anterior extremity the lamina alaris is fused with the infero-lateral margin of the pars canalicularis of the auditory capsule, and it projects sufficiently below that region to be recognisable externally as a processus paracondlyoideus (Pl. 2), to whose under surface the musculus rectus capitis lateralis is attached (fig. 3). ~ Cart. supraocc. Sin. lat. Prom. semic. ant. Fossa subare. ext. Crus commune and more medially duct. endolymph. Can.semic. post. . Proc. paracond. Pars chordalis. * Atlas. Axis. me Se & Fic. 2.—Erinaceus ewropeus, In a coronal section of the auditory capsule (figs. 1, 2) above the point of fusion with the processus paracondyloideus one can see the posterior semicircular canal. Before the hindmost limit of this canal is reached, the exoccipital cartilage has once more freed itself from the auditory capsule and is thus separated from it by a narrow fissure, the fisswra occipito- capsularis inferior (Pls. 3, 4, 7, 8), and the paracondyloid process has come toanend. This fissure is of comparatively short extent, because the ex- occipital cartilage once more fuses with the pars canalicularis, but this time with the postero-medial aspect of the posterior pole of that capsule, so that a recess is produced between the two, which is visible from the exterior. The region of fusion between the posterior pole of the pars canalicularis of the auditory capsule and the exoccipital cartilage may be termed the occipito-capsular commissure (Pls. 7, 8). Here the exoccipital cartilage comes obliquely to an end (Pls. 4,8). Its medial margin forms a consider- able part of the lateral boundary of the primary foramen magnum. e 216 Professor Edward Fawcett (2) The Auditory Capsule (Pls. 1, 2, 7, 8). This structure taken as a whole may be described as a truncated three- sided pyramid, placed very nearly at right angles with the central stem, but with a slight inclination forwards. Together with the pars chordalis of the central stem, the auditory capsules form the floor of that part of the cavum cranii which lodges the parts derived from the hind brain—a part which is as large as the rest of the cavum cranii. The auditory capsule may be regarded as consisting of two main parts: Comm. orb.-par. Sin. lat. Can. semic. ant. Foss. subarc. ext. Crus commune, Duct. endoly. Can. semic, post. Recess. sup. alaris. Proc, paracond, Pars chordalis, Fic. 3.—Frinaceus ewropeus. one, more posterior, forming the basal part of the pyramid, and which, because it contains in its interior the semicircular canals, is termed the pars canalicularis (Pls. 1, 2,3, 4, 7,8); the other, which is directed forwards and medialwards, because it contains mainly the cochlea is the pars cochlearis (P\s. 1, 2,7). The small size of this latter part, and especially that which actually contains the membranous cochlea, is very striking. The auditory capsule is moored to other parts of the neural chondro- cranium at certain precise spots; thus, the pars cochlearis is fused along the whole of its medial border to the central stem (Pls. 1, 7), and no basi- cochlear fissure is present at this stage at all events, in which respect ee en ee oe oie i wees yee ee ~ The Primordial Cranium of Hrinaceus ewropeus 217 Erinaceus differs markedly from Talpa (Pl. 9). The posterior trabecluo- cochlear and the chordo-cochlear commissures have fused together to form a common commissure just lateral to the apex of the cochlear capsule. The anterior trabeculo-cochlear commissure connects the cochlea with the postero- -lateral angle of the pars trabecularis, and at the same time this commissure is fused with the proximal portion of the ala temporalis (Pls.°1, 2/7). The pars canalicularis inferiorly is fused with the par- eccipital process of the exoccipital (Pls. 7, 8), and again, by the inner aspect of its posterior pole, is fused with the anterior edge of the upper end of that cartilage (Pls. 7, 8). From the anterior half of the upper border of the pars canalicularis there projects in an upward direction a thin plate, the parietal plate (Pls. 3, 7,8). This, by its anterior margin, is fused with the orbito- parietal commissure (Pl. 3), which connects the parietal plate thus with the ala orbitalis. The posterior margin of the parietal plate is fused with the supra-occipital cartilage of the corresponding side (Pls. 3, 7, 8), but the under margin of the latter cartilage is separated from the hinder half of the upper margin of the pars canalicularis by a well-marked fissure—the fisswra oceipito-capsularis superior (Pls. 3, 7, 8), a fissure which at its anterior end transmits the lateral jugular vein and is in itself blocked up by the down-coursing lateral sinus. In the intervals between the various commissures mooring the auditory capsule to neighbouring parts of the chondrocranium, vacuities are met with; thus, antero-lateral to both pars canalicularis and pars cochlearis, the spheno-parietal fontanelle is seen (Pls. 1, 2, 3, 4, 7); anterior to the apex cochlee the carotid foramen (foramen caroticum) (Pls. 1, 3,7). Infero-medial to the capsule one sees the jugular foramen continued laterally as a recessus supra-alaris (Pls. 1, 2, 7,8). Finally, above the posterior pole of the pars canalicularis, between the hinder half of the upper margin of the pars canalicularis and the anterior process of supra-occipital cartilage, is the superior occipito-capsular fissure (Pls. 3, 4, 7, 8). If we now examine in detail the parts of the auditory capsule, com- mencing with the pars canalicularis, we find that— The Pars canalicularis represents the basal part of the capsule (Pls. 1, 2, 3, 4, 7, 8). It contains in its interior the utricle and the semi- circular canals; it owes much of its external form to its contents, as we shall see later. For descriptive purposes we may regard this region as having four surfaces, which from their direction may be termed lateral, anterior, medial, and inferior. The lateral surface (Pls. 3, 4, 8) is irregularly quadrilateral, having an anterior, a superior, a posterior, and an inferior border. The anterior border is somewhat irregular in outline. Commencing above, one notices VOL. LII. (THIRD SER. VOL. XIII.)—JAN. 1918. 15 218 ’ Professor Edward Fawcett a slight projection—the tegmen tympani (Pl. 3)—which, compared with that in a much younger stage in Talpa, is very small; below this there is a broad sulcus, and below this a notch which is the outer end of the fossa incundis (Pl. 3), in which, in the model, the processus posterior of the incus cartilage is observed lying. Below this a projection, the -crista parotica (Pl. 3), is met with, and from the lower part of this there projects in a downward and a forward direction the processus styloideus (Pl. 3). Behind and below the root of the processus styloideus there is a broad notch, through which the facial nerve appears. This notch is in the position of the primary stylo-mastoid foramen, and it may also be taken as at the anterior limit of the inferior border of the lateral surface of the pars canalicularis. The inferior border, commencing as said at the hinder part of the site of attachment of the root of the processus styloideus, is first notched by the facial nerve, behind which it is notched by the stapedius muscle to form the incisura stapedii; behind, this border is fused with the outer edge of the processus paracondyloideus of the exoccipital, a slight groove (PI. 8), seen from the lateral aspect, separating the two superficially. The posterior border (Pl. 8) stretches from the hindmost part of the site of fusion of the aforesaid parts and runs almost vertically upwards to reach the posterior pole of the auditory capsule or cwpula posterior. . In this part of its course it is separated from the exoccipital cartilage by a narrow fissure—the fisswra occipito-capsularis inferior (Pls. 3, 4, 7, 8). At the posterior pole the pars canalicularis is fused with the exoccipital cartilage (Pls. 7, 8). The swperior border (Pls. 3, 4, 7, 8) stretches from the posterior pole in an arched fashion as far forward as the root of the tegmen tympani. — In its hinder part it is free and forms the lower boundary of the superior occipito-capsular fissure, whilst in its anterior half the superior border gives origin to the parietal plate, which rises vertically here from it. Both this border and the posterior one are rounded, each being formed by the bulging of the contained semicircular canal, the former by the superior vy. anterior ; the latter by the posterior semicircular canal. The Surfaces of the Pars canalicularis.—The lateral surface (Pls. 3, 4, 8) . is quadrilateral; at its upper margin and for some distance below is a semi- cylindrical swelling reaching practically the whole length of the upper part of the surface, and caused by the superior v. anterior semicircular canal, therefore called the prominentia semicireularis, superior or anterior (Pls. 3, 7). Along the posterior border of the surface the prominentia senicircularis posterior (Pls, 3, 7) is met with. This reaches from the posterior pole to the neighbourhood of the paracondyloid process of the exoccipital, and from near the lower end of this prominence another of ers - i Ce alias a. ee eee ee ee ee a eS re The Primordial Cranium of Hrinacews ewropeus 219 considerable width runs upwards and forwards towards the upper end of the crista parotica and is the prominentia semicircularis lateralis (Pls. 3, 8), which is caused by the lateral semicircular canal. Between this pro- minence and the prominentia semicircularis superior v. anterior is a large hollow, which may be termed the fossa subarcuata superior externa (Pls. 3, 9). This fossa is occupied by connective tissue. No opercular process is developed in connection with the lateral surface, in which respect it is unlike that of Talpa (Pl. 10). The anterior surface (Pl. 7A) is bounded, in the main, laterally by the prominent crista parotica. It commences above, below the tegmen tympani, is below this grooved by the facial nerve—sulcus facialis,—and further down, in addition, by the stapedius muscle. As a matter of fact, a large part of the fossa stapedii, which ends below at the incisuwra musculi stapedii, lodges the stapedius muscle just below the point where the facial nerve turns forward and round the outer side of the root of the processus styloideus. The medial surface (Pls. 1, 7 ; figs. 1, 2, 3) is next in point of size to the lateral surface. It is inclined a little upwards and somewhat towards the middle line. It shows just above its middle a fairly well-marked depression, surmounted by the prominentia semicircularis anterior. This hollow is the fossa subarcuata anterior interna. It contains loose connective tissue, and at some distance from it is the flocctlar part of the cerebellum, which therefore at this stage does not occupy it. Below this fossa is a large somewhat oval vacuity which leads into the interior of the pars canalicularis and in the recent condition is occupied by the crus commune of the labyrinth and by the ductus endolymphaticus (fig. 3). This, when reduced in size, as it doubtless is at a later stage, becomes the foramen endolymphaticwm aqueductus vestibuli (Pls. 1, 7) and conducts the ductus endolymphaticus. Below this vacuity the inferior border of the medial surface is seen. It is rounded, and at its anterior part somewhat prominent, forming here the prominentia utriculo-ampullaris posterior. The upper border of the medial surface is rounded for the most part and formed by the prominentia semicircularis anterior (fig. 1). This ends below the anterior part of the spring of the parietal plate in a prominence, the prominentia utriculo-ampullaris anterior (Pls. 1, 7). The wmferior surface (Pl. 7; figs. 1, 2, 3) might also be termed the infero-medial surface, because it is to a large extent inclined towards the cavum cranii. It forms the roof of the recessus swpra-alaris and to a slight extent that of the foramen jugulare. It is hidden from view by the lateral sinus when that is in position. At its most inferior and lateral part it is fused with the upper aspect of the lateral part of the paracondyloid process. Immediately beneath this surface in the recent 220 Professor Edward Fawcett condition the vagus, glossopharyngeal, and accessory nerves are seen (fig. 1), and the auricular branch of the vagus is given off in this region. A slight fossa swbarcuata posterior (Pl. 7; figs. 1, 3) is present. The Pars cochlearis (Pls. 1, 2, 7).—This may be divided into a cochlear and a vestibular segment (Voit). These two segments contain respec- tively the cochlear duct and the saccule. The “vestibular” segment is characterised by the presence of a number of foramina which open into it from the exterior; thus, on the lateral aspect, the fenestra vestibuli (ovalis), closed by the foot of the stapes, is seen. On the inferior aspect a large vacuity, the common opening of the fenestra cochlew (rotunda) and foramen perilymphaticum (aqueductus cochleew) are met with. These a Prom, semic. ant. 3 on ~---—-— Fossa subarc. ext. £ —— -—- be, Utriculus. =. 1 me Pol Can. semic. lat. 4 i r Sacc. perilymph. : } Fic. 4.-—Hrinaceus ewropeeus, two components are beginning to be separated from one another by a spur of cartilage growing from the posterior wall of the common vacuity towards the anterior wall; but when complete separation takes place (if indeed it do) I know not. The part of the vacuity destined to become the fenestra cochlee is closed by a dense sheet of connective tissue, evidently the primordium of the secondary tympanic membrane, whilst that part which will become foramen perilymphaticum (aque- ductus cochlew) transmits veins from the cochlea and the ductus _peri- lymphaticus, which latter at this stage is quite solid (fig. 4). The median wall of the “vestibular” segment of the pars cochlearis shows the internal auditory meatus (Pls. 1, 7), which has already become divided up into an upper and a lower part by a crista falciformis (Pls. 1, 7; fig. 5). The upper part is divided into an anterior foramen for the ‘ { | ee The Primordial Cranium of Hrinaceus ewropwus 221 facial nerve— foramen faciale, and a posterior part—the foramen acusticum swperius, through which passes the upper branch of the vestibular division of the auditory nerve, whose branches run to the utricle, the ampulle of the superior (anterior) and lateral semicircular canals, as well as to the upper and hinder part of the saccule. The lower segment of the internal auditory meatus is not at this stage divided up into secondary foramina. The lower division of the vestibular part of the auditory nerve supplying the lower and fore part of the - Can. semic. ant. ; a Can. semic. lat. N, facialis. Stapes. Art. staped, Proc. styloid. Fic, 5.—Hrinaceus europeus, saccule, and ampulla of the posterior semicircular canal, runs with the cochlear division of the auditory nerve through this lower segment of the internal auditory meatus. The upper anterior wall of the meatus auditorius internus is formed by the commissura suprafacialis (Pls. 1, 3, 7)—a bridge of cartilage which stretches from the “cochlear” segment of the pars cochlearis backwards over the genu of the facial nerve to reach and fuse with the pars canalicularis. Before leaving the “vestibular” segment of the pars cochlearis, a word or two may be said regarding that surface of it directed towards the primary tympanic cavity. It is, as has already been mentioned, perforated by the fenestra vestibuli, which in the recent 222 Professor Edward Fawcett condition is occupied by the foot of the stapes (Pl. 7A; fig. 5). Below this fenestra a large prominence—the promontoriwn—is met with, which indicates the position of the first turn of the cochlea. This is grooved about its middle, and in a vertical direction, by the stapedial artery. From the upper margin of the promontorium a sulcus runs forwards and medially towards the medial side of the apex cochlex; this is the sulcus caroticus; it ends at the foramen caroticum (Pl. 7A). It corre- sponds for a considerable part of its extent with the sulcus septalis, a suleus which corresponds itself with the septwm spirale in the interior of the cochlear capsule. Art. staped. Gang. Gasser. Gang. genic. Incus. Musc. tens. tymp. Man. mallei. N. ch. tymp. Art. carotid. Caps. coch. Tubu. eustach, vig Parstrabec. 7 I Pituitary (pars nervosa). Fic. 6.—Zrinaceus europeus. The cochlear part of the pars cochlearis is very small (Pls. 1, 2, 7).! In this respect Erinaceus agrees much with Talpa (Pl. 9), and is in very striking contrast with the cat, in which the cochlear segment of the pars cochlearis is of enormous size; so large, in fact, are the cochlear capsules in the cat that they compress to a remarkable degree the chordal plate (Pl. 11), from which they are separated by a large basi-cochlear fissure, but they nearly meet in the middle line over that plate. The “cochlear” segment in Erinaceus has three surfaces, viz. a large imfero-lateral, which is in continuity directly in a forward direction with the corre- sponding surface of the “ vestibular” segment of the pars cochlearis, and ' I am informed by gamekeepers and by mole-catchers that the sense of hearing is poorly developed in the hedgehog. ' * 4 » P r See ee ee ee ee oe ee a a ae The Primordial Cranium of Hrinuceus ewropwus 223 is marked by the sulcus septalis and the sulcus caroticus, both of which tend to divide it into a lower medial part corresponding with the first coil of the cochlear duct, and an upper more lateral part which, projecting forward as the cupula or apex cochlez, contains the remainder of the cochlear duct. Opposite the middle of the suleus caroticus there is a deficiency (Pls. 2, 7A) in the cartilaginous wall of the cochlear segment evidently a “pressure” deficiency which is doubtless made good later. Over the outer side of the cupula near its basal part the great petrosal nerve descends, and after a very short course joins the carotid plexus of the sympathetic accompanying the carotid artery. The swperior surface Comm, orb.-par Gang. Gasser. N. mandib. 5 Tee Squamosum, ~~ Int, art. disk. Condyle. Cart. Meckel. ' t i i i Gang. otic. { : Ala temp. | Gang. sph. par. j ' i | Pe] Pars trabec. Can. hypophys.’ Fic. 7.— Hrinaceus europeus. of the “cochlear” segment is small and in part overlain by the Gasserian ganglion, more especially in its outer half and posteriorly, where the suprafacial commissure is being given off. The remainder of the superior surface is free and helps to join the floor of the cavum supra-cochleare of Voit. The medial surface is in relation with the hypophysis, a structure of enormous size in Erinaceus (fig. 6). This surface is fused in the whole of its length with the central stem, no basi-cochlear fissure being present (Pl. 7). From the region of the cupula the anterior trabeculo-cochlear commissure runs forwards and inwards on the latero- anterior aspect of the carotid foramen to join the stalk of the ala temporalis. 224 Professor Edward Fawcett a (3) The Ala temporalis (Pls. 1, 2, 3, 7). This comparatively small cartilage springs from the postero-lateral angle of the pars trabecularis by a narrow stalk common to it and the anterior trabeculo-cochlear commissure. At first it runs directly outwards; but once the commissure above mentioned has left it, it begins to expand in a forward direction and forms a, for the most part, horizontally directed plate which breaks up into three processes (Pls. 1, 2, 3, 7), which are from Ala orbit. N. optic. “ Central stem.” Art. int. max. Pteryg. ext. Palatinum. Proc. ptys. of ala temp. Palate alate. Cart. Meck. Pteryg. int. Fie, 8.—Hrinaceus ewropeus. behind forwards a postero-lateral, an antero-lateral, and an anterior. Between the postero-lateral and antero-lateral processes the mandibular division of the fifth nerve leaves the cranial cavity (fig. 7), the notch through which it goes being consequently the homologue of the foramen ovale, immediately below which is placed the otic ganglion. The antero- lateral process is especially the site of origin of the upper head of the pterygoideus externus muscle, which arises from its lower aspect and is coincident with it. The anterior process is the forward continuation of the medial thickened edge or processus pterygoideus of the ala temporalis ; it juts forwards between the lower head of the pterygoideus externus and on its medial aspect the greater part of the pterygoideus internus (fig. 8). The Primordial Cranium of Erimaceus ewropeus 225 (4) The Ala orbitalis (Pls. 1, 2, 3; fig. 8). The ala orbitalis is a large triangular plate of cartilage whose base lies laterally and is included in the lateral structures enumerated above as entering into the neural chondrocranium. It forms part of the floor of the cavum cranii as well. By its anterior basal angle it is fused with a backwardly prolonged process from the frontal prominence of the pars intermedia of the lateral wall of the nasal capsule, forming the spheno- ethmoidal commissure (Pls. 1, 8). Its posterior basal angle is prolonged backwards to fuse ultimately with the parietal plate and form the orbito- Ala orbitalis. Ala orbitalis. Muse. rect, lat. Palatinum,. Duct. nas. pharyng. Cart. Meckel. Ala hypochias. Cavum oris. Fie. 9,—Hrinaceus europeus, parietal conmissure (Pls. 1,3). The apex of the ala orbitalis is directed towards the central stem (hinder part of pars orbito-nasalis); before reaching the central stem the apical region bifurcates into an anterior and a posterior branch (Pl. 1). The former reaches the central stem in front of the optic nerve ; the latter reaches the central stem—? the ala hypo- chiasmata—behind the optic nerve, so that between the two limbs the optic foramen is formed. The optic foramen here is really a tunnel (fig. 9) whose floor is formed by the ala hypochiasmata, or at all events what corresponds with that structure in other mammals.. The upper wall of the tunnel seems, so far as my experience goes, intrinsic to Erinaceus. It seems to give origin especially to the obliquus superior of the eyeball. The lower wall of the tunnel, the ala hypochiasmata proper, gives origin to the rectus system of muscles, especially the inferior and lateral recti. The anterior 226 Professor Edward Fawcett border of the ala orbitalis is for the most part free, and forms the hinder boundary of the orbito-nasal fisswre (Pls. 1, 2, 3), through which the nasal nerve enters the cranial cavity from the orbit to reach ultimately the cribriform plate of the ethmoid. Just lateral to the optic foramen a deep notch is found in the anterior free border of the ala orbitalis (Pls. 1, 3), but it does not transmit anything. The posterior border of the ala orbitalis is free, and forms at once the common anterior boundary of the spheno-parietal fontanelle and that segment of it which, lying anterior to the ala temporalis, is the representative of the sphenoidal or superior orbital fissure and foramen rotundum combined, through which run the maxillary and ophthalmic divisions of the fifth nerve, the third nerve, the sixth, as well as the fourth. It may here be mentioned that the maxillary division of the fifth nerve does not perforate the ala temporalis, so that a foramen rotundum is absent; as is the case in most mammals. The outer surface of the ala . temporalis is marked by a horizontal suleus in which the supraorbital artery lies, down to the level of which reaches the frontal bone. The Ala hypochiasmata (Pls. 2, 3; fig. 9)—This has already been alluded to as probably the lower half of the optic tunnel, and is well marked in Erinaceus. It gives origin to the rectus system of muscles of the eyeball, and at this stage seems to be associated with the hinder branch of the apex of the ala orbitalis. How it develops in Erinaceus I know not, but in Tatusia novemcincta it is developed independently. Possibly that is its usual mode of development, and recently Kernan has shown this to be the case in man. (5) The Lateral Nasal Capsule. This, though truly a lateral appendage to the central stem, will be described with the nasal capsule. 3. LATERAL STRUCTURES. These are the ala orbitalis, the parietal plate, and the anterior process of the supraoccipital cartilage. The ala orbitalis has been fully described, and nothing more need be said concerning it. The Parietal Plate (Pls. 3, 7, 8), already mentioned in connection with the pars canalicularis of the auditory capsule, is a very simple structure, and may be considered as a quadrilateral plate fixed by its lower border to the anterior half of the upper margin of the pars canalicularis of the auditory capsule (fig. 5). Its anterior margin is free, and helps to form the posterior boundary of the spheno-parietal fontanelle, Its superior margin is likewise free. Its superior margin forms the anterior boundary .of the superior ee ee ee ee ae Se en Rt 4 a The Primordial Cranium of EHrinaceus ewropeeus 227 occipito-capsular fissure. Its antero-superior angle is fused with the orbito- parietal commissure, whilst its postero-superior angle blends with the anterior process of the supraoccipital cartilage. It is covered laterally at this stage by the parietal bone, but only in its upper half. The lower half of its outer surface is covered by a thick layer of perichondrium, from which arises part of the temporal muscle. The Swpraoceipital Cartilage (Pls. 1, 3,’4, 7, 8).—This cartilage is of great interest in Erinaceus at this stage, because it perhaps gives us the key to the morphological history of what will later be alluded to under the name tectum synoticum or tectwm posterius. It may be described as a triangular plate whose apex is directed down- wards towards the pole of the pars canalicularis of the auditory capsule (Pl. 8). Its anterior basal angle is fused with the postero-superior angle of the parietal plate, whilst the posterior basal angle stretches medially over the cavum cranii to meet its fellow in the middle line, fuse with it, and form the so-called tectum synoticum or tectum posterius, which in Erinaceus is very narrow. The under margin of the supraoccipital cartilage is free, and forms the upper boundary of the superior occipito-capsular fissure. The upper margin is free and somewhat irregular; whilst the lower margin, at first near the apex, connected by fibrous tissue with the oblique upper edge of the exoccipital cartilage and more medially for a short distance fused histologically with it, is then continued medially as the upper margin of the primary foramen magnum. In the calf at the 19-mm. stage (Pls. 12, 13) and at the 40-mm. stage (PI. 14) the various phases in the formation of the tectum synoticum from the enlarging supraoccipital cartilages of the two sides can be followed out in their entirety. Whether this plan is followed in other mammals remains to be shown. The term tectum synoti- cum is not altogether a happy one, and had better be abandoned for the term tectum cranii posterius (Pl. 8). Its connection with the auditory capsule is indirect, i.e. through the parietal plate, and it is also clearly a secondary one. In man there are two tecta (Bolk, Fawcett): one a very wide one, the more posterior, therefore called the tectum cranii postervus, from the middle of whose anterior border a processus ascendens arises (Fawcett ; in pig, Mead); the other, the tectwm cranw anterius, is very slender and quite isolated, not reaching the parietal plate on either side, nor being in any way connected with the tectum cranii posterius. Recently, I have observed two tecta in the cat: one certainly the ordinary tectum posterius, the other small, median, and anterior to this, which may be an isolated processus ascendens or may be looked upon as a tectum anterius. It seems to be quite clear that this region of the chondrocranium is in a progressive condition, as Gaupp has stated. In Dasyurus viverrinus (Pl. 15) 9°5-mm. 228 Professor Edward Fawcett stage two bars run across the middle line, but both appear to be parts of the tectum posterius, or has the anterior one reappeared in isolated form as in man and cat? This is scarcely probable. 4. THE LATERAL COMMISSURES. These are, from before backwards, first the spheno-ethmoidal (Pls. 1, 3) (of large size), which connects the anterior basal angle of the ala orbitalis with a backward projection from the frontal prominence of the pars inter- media of the lateral wall of the nasal capsule. Next follows the orbito- purietal commissure (Pls. 1, 3), connecting the posterior basal angle of the ala orbitalis with the antero-superior angle of the parietal plate. Finally we may describe the union of the anterior angle or process of the supraoccipital cartilage with the postero-superior angle of the parietal plate as the occvpito- parietal commissure (Pl. 3); or of course the tectum posterius, stretching as it does from parietal plate of one side to that of the other, may be looked upon as the commissura posterior. 5. THE NasaAL Capsule (Pls. 1, 2, 3, 4, 5, 6). This is an enormous structure. Its antero-posterior length amounted in the model to 200 mm., whereas the total length of the central stem behind was only 125 mm. It consists of a central part, the septum nasi, and two lateral appendages (as Fischer says, it may be compared with a double-barrelled gun). When the whole is viewed from above (Pl. 1) it has a very characteristic pear- shape, of which the larger end or bulb is placed backwards. The whole capsule is more complete than usual, recalling that of the mole (Pl. 9), but it is relatively not so long. The main openings into it are the large sub- cerebral vacuity on each side of the septum which, at this stage, has not been bridged over by a lamina cribrosa; the fenestra (incisura) narina (Pls. 2, 3,4) which is placed rather laterally than apically; and the long fenestra basalis (Pl. 2) which is seen below. A small fenestra dorsalis (Pl. 1) is present just behind the apex of the anterior part of the nasal capsule, recalling somewhat that seen in the mole (PI. 9), but it is more apically placed than in the latter. No foramen epiphaniale could be found after a most careful search, but a small eribro-ethmoidal canal is present. The nasal septwm (Pl. 5) is that segment of the pars interorbito-nasalis which is included in the nasal capsule. It is of great length; is partly subcerebral—say one-third; the remainder is precerebral. The subcerebral part is very narrow from above downwards; posteriorly it is also narrow a ee a The Primordial Cranium of EHrinaceus ewropeus 229 from side to side (fig. 10); thicker below, however, than above. When traced in the forward direction it rapidly increases in height, its upper border rising in height whilst the lower remains more or less constantly at the same level. The greatest height is reached at the junction of the subcerebral and precerebral parts of the upper border. At the hindmost part of the subcerebral segment of the upper border a cartilaginous bar is given off on each side to connect the septum with the lateral wall of the nasal capsule. This bar I think well to name the tectwm nasi posterius (Pls. 1,5). Fischer Comm, sph. eth. —- N. olfact. Sept. nasi. — Pars post. Eth. turb. I. —fees Lam. trans. post. - Fic. 10.—Hrinaceus ewropeus, has named the corresponding bar in Talpa the planwm antorbitale. That term I think better to use in another sense and for another region. At the junctional region of the upper border of the subcerebral part of the septum with the precerebral part two prominences arise on each side. Perhaps they represent a cartilaginous crista galli. The precerebral part of the upper border of the septum nasi is attached laterally in its whole length to the tectum anterius (Pl. 1; fig. 11), and as each half of this tectum in passing laterally from the septum arches first in an upward direction, it follows that the precerebral part of the upper border of the septum lies at the bottom of a longitudinal median sulcus, swleus dorsalis nasi (Pls. 1-5). From behind forwards this part of the upper border for 230 Professor Edward Fawcett two-thirds of its extent slopes downwards and forwards, forming roughly an angle open below of about 135° with the subcerebral part of the border (Pl. 5). But beyond this point the anterior third runs more horizontally forwards. This part at its anterior end turns suddenly down at right - angles to form the anterior border of the septum, and from its lateral margin the wing-like expansions pass in an arched manner forwards, then outwards, and finally backwards to form the ewpula anterior (Pls. 2, 3, 4, 5) or median alar cartilage on each side, from which there projects the free lateral edge and about its middle a peg-like process, the processus ewpularis or processus alaris medius. The inferior border (Pls. 2, 3, 4, 5) is of great length, and is thicker than either the superior or the anterior border, Recess, ant. Crista semic. Eth. turb, I. , a and lat. nas. Max. turb. Vomer. Fic. 11.—Zrinaceus europeus. especially in its hinder two-thirds. For about its hinder two-thirds it is comparatively straight and continues almost uniformly the plane of the under aspect of the central stem behind it, being bent at its commencement only slightly at an angle with the latter; but as one reaches the anterior extremity of its hinder two-thirds it rises upwards somewhat and becomes greatly thickened (PI. 5); then bending downwards at right angles to its original course it joins the anterior third of the lower border, which is much thinner and appears to bifurcate to form more posteriorly the laminz transversales anteriores, and more anteriorly two triangular plates which are almost a replica of those found in the mole. Perhaps they may be called processus laterales ventrales (P|. 2). The apex of each is separated by a narrow fissure, continuous with the lower segment of the fenestra narina, from the processus alaris swperior v. lateralis (Pl. 2). From what The Primordial Cranium of Erinaceus ewropeus 231 has been said, the whole septum viewed from the side has a somewhat triangular form, of which the base is formed by the lower border. _ The whole septum is comparatively thin save at its lower border, and in the neighbourhood of the junction of the posterior two-thirds with the anterior third it greatly thickens, and in coronal section has a somewhat pear-shaped form (fig. 12), the larger end of the pear being downwards. The septum at its lower border (Pls. 2, 5) has numerous important relations. Commencing behind, there are at each side of this border the following structures: first, the lamina transversalis posterior, separated by a narrow fissure from the septum (fig. 10); next, the corresponding lamella _ of the vomer, which stretches from the lamina transversalis posterior as Duct. gland, lat. nas. Atrio. turb. Duct. naso-lac, -- Proc. trans. — Cart. parasep. ant. - Incisivum. Fie, 12.— Hrinaceus ewropeus. far as the medial side of the hinder end of the organ of Jacobson; next, the free part of the organ of Jacobson ; next, the medial lamella of the anterior paraseptal cartilage (fig. 13); then a thick band of connective tissue (fig. 12), which stretches between the thick part of the lower border of the septum and the narrow cylindrical stalk of the medial lamella of the paraseptal cartilage and the median part of the hinder edge of the lamina transversalis anterior; next, the lamina transversalis anterior, which is directly continuous with the septum; finally, the base of the processus lateralis ventralis (fig. 14). The lateral part of the nasal capsule (Pls. 6, 3, 4) may be described as consisting of the following parts: a roof, composed of a small tectum posterius ; a large tectum anterius; a lateral wall (paries nasi); a floor; an anterior wall (cupula anterior); and a posterior wall (cupula posterior). The roof (Pls. 1, 5, 6) is partly subcerebral and partly (mainly) pre- 232 Professor Edward Faweett cerebral. The subcerebral part of the roof is complete posteriorly, forming the tectum nasi posterius,—a bar of cartilage, quite flat as seen from the — cavum cranii (Pl. 1), which stretches outwards from the upper border of the subcerebral part of the septum to the lateral wall. The remainder of the subcerebral part of the roof, represented in older stages by the lamina cribrosa, is wanting at this stage. The precerebral part of the tectum is complete save near the apex, where it is perforated by the small fenestra dorsalis (Pls. 1, 5), which in the recent condition is filled with fibrous tissue and a blood-vessel. In its whole length this part of the tectum — Naso-turbin. — Duct. gland. lat. nas, — Max. turbin. Organ Jacobs. -— Cart. parasep. ant. Fic. 138.—Hrinaceus europeus. (tectum anterius) is slightly convex from side to side, the two convexities producing the median dorsal sulcus, at the bottom of which the upper margin of the precerebral part of the septum lies. Almost insensibly the tectum passes into the lateral wall or paries nasi (figs. 11, 12, 18, 14). The lateral wall, as seen from the exterior (Pls. 3, 4), shows the three divisions already mentioned in connection with roof, viz. the pars posterior pars intermedia, and pars anterior. These parts are separated from one another by curved sulci, thus—the pars posterior is separated from the pars intermedia by the sulcus lateralis posterior. This is a curved sulcus, convex forwards, which stretches from the upper margin of the lateral wall of the nasal capsule behind the attachment of the spheno-ethmoidal com-— missure forwards, then downwards, and finally somewhat backwards to reach the lower margin of the lateral wall. It is not very well marked, Pe) es Pe ee eee | | | The Primordial Cranium of Hrinaceus ewropoeus 233 especially below, but it coincides in the interior with the main attachment of the first primary ethmo-turbinal. The sulcus lateralis anterior is like- wise not very well marked, but it may be traced in a downward and forward direction from the cribro-ethmoidal foramen to the lower edge of the lateral wall. Of the three divisions of the lateral nasal capsule, the pars anterior is in length almost the equal of the pars intermedia and pars posterior taken together. The pars posterior commences behind at the cupula posterior, and is triangular in shape, the apex being at the said cupula and the base at the sulcus lateralis posterior. Its upper margin is subcerebral, and here forms ----- Atrio. turb. —— Duct. nas. lac. f.. Lam. trans. ant. Fic. 13a.—EZrinaceus europeus. the outer wall of the olfactory vacuity. Its lower margin is sharply_ defined, and when traced inwards is seen to be continuous with the outer edge of the lamina transversalis posterior. This margin in the complete condition is in relation with the os palatinum. The lateral surface of the pars posterior is flattened or even slightly hollowed out, and, being in - relation to the orbit, is called planum antorbitale; but the planum ant- orbitale extends beyond this region on to the posterior aspect of the pars intermedia, and it obscures the postero-lateral sulcus somewhat by doing so. The pars intermedia is placed between the antero-lateral and postero- lateral sulci, and possesses the greatest height of the three parts into which the lateral wall is divided. It shows three prominences, viz. a prominentia superior or frontalis, a prominentia inferior or maxillaris, and between VOL. LIL. (THIRD SER. VOL. XIII.)—JAN. 1918. 16 234 ae Professor Edward Fawcett the two in front a prominentia anterior. This last is ill marked. Each of these prominences is caused by a corresponding hollow on the internal aspect. To the prominentia superior or frontalis the spheno-ethmoidal commissure is attached, while from the front of the prominentia inferior v. maxillaris the inferior oblique muscle of the eyeball arises. | The sulcus lateralis anterior, which delimits the pars intermedia anteriorly, corresponds in the interior with the greater part of that important landmark the crista semicireularis (Pl. 6). ~ The lower half of the anterior surface (Pl. 4) of the maxillary promi- nence is covered by the maxilla, and at this stage the hinder part of the frontal prominence is covered by the frontal bone ; and as these bones, so far - Tect. nas. ant. =--~— Proc. alar. sup. -- Proc. lat. vent. CEarren ee he : : «" Sos athe a iy Fic. 14,—ZHrinaceus ewropeus. _as I know, always have this relation to the prominences in question, I have given the alternative names to them. The inferior margin of the pars intermedia is slightly inrolled, and forms in its anterior part a portion of the maxillo-turbinal (Pl. 6; fig. 13), which, however, chiefly belongs to the pars anterior. The pars anterior (Pls. 3, 4) is of considerable length, as has already been stated; but it is of only moderate height, a height which is pretty uniform throughout. Superiorly, it runs into the tectum anterius at a rounded superior border; posteriorly, it is separated from the pars intermedia by the antero-lateral sulcus. The inferior border can be divided into three parts; the hinder part, equal in length to the other two combined, is inwardly projected as a shelf, and along its greatest convexity runs the naso-lacrimal duct. Here too the mucous membrane 2, 2 OG ‘Pa ee ee Pee eee ° ad Se Se ‘ Pee ee” oo © Leavy ee ye oe a ee eee ee ae The Primordial Cranium of Hrinaceus ewropweus 235 of the nasal sac projects sufficiently low down to be visible from the lateral aspect, but that is not shown in the model. There is no lamina infraconchalis such as occurs in the rabbit (Voit) or in the model of -Microtus, fromm which the mucous membrane has been removed (Fawcett). One may see at some depth the greater part of the anterior paraseptal cartilage as well as the organ of Jacobson, which rests in its concavity (Pl. 3). The anterior end of this part of the inferior border ends at the lamina transversalis anterior; the next stage of the inferior border is fused with the outer edge of that lamina; a deep gutter, in which the naso-lacrimal duct is lodged, marks the line of junction of lamina with the inferior border (fig. 12). In front of the lamina the final part of the inferior border becomes free, forming the lateral margin of a small notch, the incisura pretransversalis, through which the naso-lacrimal duct passes to fuse with the mucous membrane of the nasal sac. The anterior border of the pars anterior forms the hinder margin of the fenestra narina, which in Erinaceus is directly laterally ; it is incurved about its middle, forming a somewhat triangular plate which is the anterior end of the atrio-turbinal (PI. 2). The fenestra narina (Pls. 2, 3, 4) is a somewhat dumbbell-shaped vacuity directed laterally. Its anterior border is formed by the cupula anterior, from whose middle there projects laterally and in a somewhat downward direction a cupular spine. Its posterior border is formed by the anterior margin of the pars anterior, and, as the cupular spine and the anterior end of the atrio-turbinal project towards one another some- what opposite the middle of the fenestra, it is constricted into an upper and a lower large segment.. The fenestra is bounded below by the processus alaris superior, a triangular piece of cartilage which is somewhat bent, so as to be eonvex externally. Behind this plate of cartilage the naso-lacrimal duct reaches the nasal sac. : The floor (solum nasi) (Pl. 2) of the nasal capsule is incomplete, more so than it is in the rabbit or Dasyurus or Microtus. There is a large vacuity in it, the fenestra basalis; moreover, a narrow fissure, represent- ing the septo-paraseptal fisswre of other mammals, intervenes between the anterior paraseptal cartilage and the septum, and between the median edge of the lamina transversalis posterior and the septum. The fenestra basalis is even relatively larger in Erinaceus than it is in the above- mentioned forms, because the paraseptal cartilage is incomplete: all that narrow cylindrical part such as is seen in the animals before cited being absent. The structures entering into the solum nasi are the following: behind, the lamina transversalis posterior; in front, the lamina trans- versalis anterior, from whose inner end is projected backwards half way 236 Professor Edward Fawcett along the medial aspect of the fenestra basalis the bilaminar anterior paraseptal cartilage. These parts may be examined seriatim :— . The lamina transversalis posterior is a triangular plate whose apex is at the cupula posterior, whose base is free, forming the posterior boundary of the fenestra basalis, and whose median end lies along the infero-lateral margin of the septum nasi, separated from it wholly by a narrow fissure— the posterior moiety of the septo-paraseptal fissure (fig. 10). The median edge of the lamina near the cupula is covered by the perpendicular plate of the os palatinum, whilst anteriorly the same edge is covered by and is perhaps being commandeered by the ossifying corresponding lamella of the vomer. The outer side of the lamina is fused with the inferior margin of the pars posterior of the lateral wall, and when viewed from above it can be seen that the second primary ethmoid-turbinal springs from the site of fusion of the two. (I have used the term “fusion” here to express continuity of the two structures, not to infer that they were ever separate.) The region to which the lamina transversalis posterior belongs is one of very. great interest, and it has been discussed at con- siderable length by Eugen Fischer in his description of the skull of Talpa. It cannot be said that anything like settlement has been reached regarding its fate, but what is certain is that the cupula posterior extends backwards, being pushed backwards by the backward growth of the nasal mucous sac. It finally becomes jammed against the point of the apex of the ala orbitalis [(?) ala hypochiasmata], and takes a part in the formation of the cartilaginous body of the sphenoid; and so much is the median part of the body—z.e. the part derived from the hinder end of the pars inter- orbito-nasalis (presphenoid)—compressed between the cupule of the two sides that it is reduced to a very narrow septum when ossified, which projects in the middle line as the rostrum of the sphenoid and in front as the ethmoidal crest. By what mode exactly the cupula is replaced by bone is not as yet certain, but a glance at the fully-formed sphenoidal turbinal at a time before that has become fused to the rest of the body of the sphenoid gives a facsimile of the cupula as seen from below. Whether, then, the sphenoidal turbinal ossifies from the cartilaginous cupula or from its perichondrium or from the fibrous remnant of the former, and whether by one centre or by many, as Cleland has described, remains uncertain. The Lamina transversalis anterior (Pl. 2).—This is of large size and quadrilateral in form; moreover, it is convex downwards and about a horizontal plane. Medially, it is to a large extent in direct continuity with the lower margin of the septum nasi; but as this border is traced backwards it becomes free of the septum, and from its hinder edge a narrow rod The Primordial Cranium of Hrinaceus ewropeus 237 of cartilage runs backward to connect it-with the anterior paraseptal cartilage, to which we will return later. The lateral border is directly continuous with the under edge of the anterior part of the lower border of the pars anterior of the lateral wall of the nasal capsule, and at the line of junction a deep groove—the naso-lacrimal suleus—lodges the naso-lacrimal duct (fig. 12). The anterior border is V-shaped and free; the apex of the V is the incisura pretransversalis, through which the. naso-lacrimal duct reaches the nasal mucous sac. The posterior border of the lamina trans- versalis anterior is free, and forms the anterior boundary of the fenestra basalis. From its medial end, at its junction with the medial border of the lamina, the narrow cartilaginous stalk of the anterior paraseptal cartilage is given off, whereas from its lateral edge an extraordinary cartilage passes outwards almost horizontally, but with a slight curve (whose concavity is upwards) (fig. 12), as far as the outer contour line of the pars anterior of the lateral wall of the nasal capsule. This bar of cartilage I have not met with in any other animal. Between it and the junctional region of the outer edge of the lamina transversalis anterior and the lower edge of the lateral wall of the pars anterior of the nasal capsule, the naso-lacrimal duct enters the naso-lacrimal sulcus from behind, so that the root of this bar conceals the duct from view as seen from below. The bar itself is covered below by the body of the os incisivum. From its direction the bar may be called the processus transversalis (Pl. 2). _ The Anterior Paraseptal Cartilage (P1.2).—This cartilage commences in a narrow stalk which runs back from the angle of junction of the medial with the posterior border of the lamina transversalis anterior. This stalk is placed at some depth from the general inferior plane of the lamina and main mass of paraseptal cartilage, so that the appearance is that of a deep notch in the paraseptal cartilage. This notch is occupied by the connect- ing stalk of bone between the main mass of the os incisivum and its “paraseptal” process. The main part of the anterior paraseptal cartilage between this notch and stalk is bilaminar, consisting of a large medial lamella which is outcurved below to form a gutter.in which rests the organ of Jacobson (fig. 13). The medial lamelle of the opposite paraseptal cartilages are very closely approximated by their medial surfaces, only a certain amount of connective tissue and the “ paraseptal” processes of the ossa incisiva intervening in the anterior half of the inter-paraseptal space (Pl. 2; fig. 13). Each medial lamella is suspended from the side of the lower part of the nasal septum by dense cellular tissue. The anterior end of the bilaminar part projects downwards and forwards and lodges the duct of the organ of Jacobson. This duct, distinguished from the organ itself by the different character of the epithelium lining it, ends ultimately 238 Professor Edward Fawcett at the inner side of the middle of the naso-palatine duct. The body of the organ projects backwards beyond the hindmost limit of the paraseptal cartilage for a short distance (Pl. 5), and it is in the space between the two “organs” that the anterior ends of the vomer appear (PI. 2). The Medial Aspect of the Lateral Wall of the Nasal Capsule (Pl. 6).— This, like the lateral aspect, may be divided into three parts, and these parts are much more readily distinguished from one another from this aspect. The parts are as before: a pars anterior, a pars intermedia, and a pars posterior. The pars anterior is only imperfectly separated from the pars intermedia by a crest, the crista semicircularis, which stretches from above in the region of the cribro-ethmoidal foramen, which has perforated it, in a downward and forward direction for half the depth of the lateral wall of the capsule; the crest then becomes very low, running down- wards and backwards to end in the lateral wall of the maxillary recess. This lower part of the crista semicircularis may be looked upon as the representative of the ‘processus uncinatus, when that exists, of the later ethmoid. The pars intermedia is the deepest segment of the lateral wall. It is bounded behind by the first primary ethmo-turbinal, and reaches back to the cupula nasi posterior. Considering these parts now in detail, and commencing with the pars anterior, it may be said at once that it is by far the longest segment. It is roughly quadrilateral in outline, deepest behind, narrowing about its. middle to again deepen somewhat in front. Its upper border stretches from the upper end of the crista semicircularis as far as the upper edge of the fenestra narina, and is slightly concave on its upper aspect. It is thickest behind. Its lower border or boundary is formed of several parts: in front, by the processus alaris inferior; behind, that of the lamina trans- versalis anterior, which is here upeurved to form a large part of the atrio- twrbinal; behind the lamina transversalis anterior the lower border is formed by the maxillo-turbinal for a considerable distance—in fact, in almost the remainder of its extent; its final part posteriorly is formed: by the lower edge of the recessus glandularis, of which more will be said later. ‘The anterior boundary is short, and its middle is swung medially in the form of a triangular plate which forms the commencement of the atrio-turbinal. The posterior boundary or border is only well marked in its upper half, where it is formed by the crista terminalis. The general medial surface of the pars anterior is markedly concave from above downwards, but from end to end shows about its middle a slight convexity. This region shows us three turbinal cartilages, which are the atrio- The Primordial Cranium of Hrinaceus ewropeus 239 turbinal, the maxillo-turbinal, and the naso-turbinal. The atrio-turbinal (P1. 6) commences anteriorly as a triangular inrolling of the middle of the posterior wall of the fenestra narina. At the middle of its base a small vacuity is seen; the lower basal angle of the triangular plate is continuous with the ridge on the upper aspect of the lamina transversalis anterior ‘caused by the sulcus naso-lacrimalis above described, so this ridge is a part of the atrio-turbinal. Posteriorly the atrio-turbinal is directly continued into the maxillo-turbinal without any notch of separation such as exists in the rabbit (Voit) and the water-rat (Fawcett). The maxillo-turbinal (Pl. 6) is the medially-rolled lower edge of the pars anterior behind the plane of the lamina transversalis anterior. It may in this case be regarded as the direct backward continuation of the atrio-turbinal. It does not quite correspond with the whole of the lower free border of the pars anterior, which lies posterior to the lamina transversalis anterior. If that border be traced backwards with the microscope, section by section, it will be found that the mucous membrane which covers it, and which forms the actual conchal projection into the nasal cavity, gradually dies away and ceases to project at a spot perpendicularly below the lower end of the crista terminalis. The naso-turbinal (Pl. 6; fig. 18) is of small size, and may-be traced forwards from the front of the lower end of the crista semicircularis about midway between the upper and lower borders of the medial surface of the lateral wall of the nasal capsule. Unlike that of the rabbit and of the water-rat, it is attached by its whole length to the lateral wall, and at this stage reaches from the crista semicircularis only one-third of the way towards the anterior margin of the lateral wall (Pl. 6). It partially divides the cavity of the nose into an upper and a lower channel. The lower channel has been named in the rabbit the sulcus swpraconchalis (Voit). It lodges here a considerable part of the lateral nasal gland. The pars intermedia (PI. 6) is entered by a crescentic somewhat wide fissure. Anteriorly it is imperfectly cut off from the pars anterior by the crista semicircularis, but below that crest it is continuous with the pars anterior through the recessus glandularis. . Posteriorly it is limited by the first primary ethmo-turbinal, which overhangs its medial side to a very considerable extent and so reduces the entrance into the general nasal cavity to a crescentic fissure. Below, the pars intermedia is bounded by the junction of the first ethmo-turbinal with the lower border of the lateral nasal wall; whilst above, the pars intermedia opens into the cavum cranii through the large olfactory vacuity. : The general concavity of the pars intermedia is divisible into a series of recesses, of which the main ones are the recessus superior v. frontalis and the recessus inferior v. maxillaris. These recesses are imperfectly 240 Professor Edward Fawcett separated from one another by the anterior root of the first ethmo-turbinal. In front of this root they communicate with one another and form the point of communication; a small but somewhat deep recess, the recessus anterior, projects forwards under cover and therefore to the lateral aspect of the lower half of the crista semicircularis proper. A fourth recess, but very shallow, is found at the antero-inferior part of the recessus maxillaris ; this recess, the recessus glandularis (fig. 11), is continuous posteriorly with the recessus maxillaris and anteriorly with the sulcus supraconchalis of the pars anterior. These various recesses produce corresponding elevations on the exterior of the pars intermedia, but these are not very readily distinguished on that surface, either because the model is not so well made as it might be or because the corresponding prominences are not well marked. The recessus frontalis v. superior (Pl. 6) is of large size; freely com- municating above with the cavum cranii, overlapped medially by the upper half of the first primary ethmo-turbinal, and below partly shut off from the recessus maxillaris v. inferior by the anterior root of the turbinal just mentioned, it communicates anteriorly with the recessus anterior, and below, in front of the anterior root of the first primary ethmo-turbinal, with the recessus inferior. Its outer wall is marked by one low frontal turbinal which divides it imperfectly into an upper and a lower channel. A special bundle of olfactory nerves arises from the mucous membrane lining the frontal recess and enters the olfactory bulb. The recessus mawillaris is a large and-deep recess lying under cover of the lower part of the first ethmo- turbinal and below its anterior root; it is crossed in front of its middle by a low-curved ridge in continuity above with the crista semicircularis and on the outer wall of the recess. This ridge, which I have before mentioned as being probably homologous with the root part at all events of a processus uncinatus, cuts off from the recessus maxillaris a part in front and below in which is lodged the lateral nasal gland (fig. 11); and the actual fossa in which the gland lies is the fourth recess of the pars intermedia, viz. the recessus glandularis. This recess, as has been before mentioned, is directly continued into the suleus supraconchalis. The pars posterior v. ethmo-turbinalis is characterised by the presence in it of well-marked ethmo-turbinals. These in Erinaceus are two in number at the stage modelled. In general outline it is quadrangular in form, having therefore four borders, which are a superior, an inferior, an antero- superior, and an antero-inferior. ‘The superior and inferior borders meet at an acute angle posteriorly—at, in fact, the cupula posterior. The superior border is in great part free, and forms the posterior boundary of the orbito- nasal fissure. At its inner end it is, in its upper three-fourths, fused with The Primordial Cranium of Hrinaceus ewropeus 241 the nasal septum and only artificially displayed by section. The lower part is free and directed towards the septum nasi—separated from it, how- ever, by the posterior moiety of what has been previously called the septo- paraseptal fissure. The inferior border in its anterior half is formed by the lower edge of the lateral wall of the pars posterior, in its posterior half by the inner edge of the lamina transversalis posterior. The antero-superior border is formed by the upper half of the free edge of the first primary ethmo-turbinal, the antero-inferior border by the lower half of the free edge of the same turbinal. ‘Two turbinals are found in connection with this region which from before backwards are the first and second primary ethmo- turbinals (I use the term primary in the sense of priority of development). At this stage there are no secondary ethmo-turbinals. By means of the second primary ethmo-turbinal the general cavity of the pars posterior is divided into two meatiis, viz. a superior or posterior behind the second primary ethmo-turbinal and an inferior or anterior in front of it. The primary ethmo-turbinals (PI. 6), as has been said, are two in number, viz. first and second. ‘The first primary ethmo-turbinal is of large size, and extends between the upper and lower borders of the pars posterior. It is attached to the lateral wall by three roots, viz. an upper, lower, and an anterior, all of which, with the exception of the anterior, are continuously attached to the outer wall. The anterior root is bridge-like, being attached only at its anterior and posterior extremities. It runs forwards into the pars intermedia and incompletely separates it into the recessus frontalis and the recessus maxillaris. The upper and lower roots are much in the same continuous line, attached to the lateral wall, a line of attachment which is indicated on the exterior by the sulcus lateralis posterior. From these roots of equal length the main plate is formed which stretches from upper to lower borders of the pars posterior; the medial surface of the plate so arising forms a large part of the lateral wall of the pars posterior, and the free edge of the plate, which looks forwards as well as a little medialwards, forms the most anterior limit of the pars posterior. This free edge really consists of about three equal parts, joined at two angles, of which the lower is the more prominent. The upper part of the free edge is almost horizontal and directed inwards and forwards. The next part bends downwards and forwards to the angular projection above mentioned, whilst the lower part of the freé edge runs downwards and somewhat backwards to the lower edge of the lateral wall of the nasal capsule. It is the angle of junction of the middle and lower parts of the free border which constricts the opening into the pars intermedia. The medial surface of the first primary ethmo-turbinal is marked by a faint ridge which runs downwards and forwards and is a hint at a subsequent 242 Professor Edward Fawcett bilamellar condition of this turbinal, such as frequently exists in other animals. It would seem from the specimens at my disposal that that uni- lamellar condition is the primary one ontogenetically and phylogenetically. The second primary ethmo-turbinal is almost a replica in small form of the first, but it has no anterior root, and it shows no sign of being bilamellar. Its root stretches from the upper free border of the pars posterior to the junction of the lamina transversalis posterior with the lower edge of the lateral wall of that region (Pl. 6; fig. 10). Its anterior free edge repeats in smaller form that of the anterior edge of the first primary ethmo-turbinal. By it the cavity of the pars posterior is divided into two meatiis, and it — develops later than the first ethmo-turbinal. It is likely that it owes much of its development to the backward growth of the nasal sac. Whether more than two primary ethmo-turbinals are developed in Erinaceus, I know not. The position of the ethmo- bastioiale especially the first, and of the crista semicircularis strongly suggest that these structures play an important part in the modelling of the nasal capsule, and that they, by acting as strengthening rods, as it were, or internal buttresses to this wall, prevent its being outpouched at their sites of attachment; hence, as the nasal sac grows it tends to cause the lateral wall of the capsule to bulge outwards at three spots, viz. in front of the crista terminalis, between it and the first ethmo-turbinal, and behind the first ethmo-turbinal. In this way the various divisions of the exterior are brought about. Morphology of the Parts of the Solum Nasi.—Some remarks may be made regarding the morphological history of certain structures forming the solum nasi, more especially of the lamina transversalis anterior, the para- septal cartilages, and the lamina transversalis posterior. In the majority of mammals below man, perhaps in all, the lamina transversalis is present, and in the majority it connects the septum with the lateral wall, a true zona annularis being formed; but in Lepus this lamina is stated by Voit—and I confirm his statement—to be separated from the septum ay a narrow fissure. The posterior paraseptal cartilage seems to be present i in some forta: or other in all mammals hitherto examined. Its presence is usually denied in man, but in a model made by me‘of the nasal capsule of a 65-mm. embryo, it is certainly present, though small. It is formed by the hollowing out of the planum antorbitale due to the backward growth of the nasal sac, and the floor of the hollow is the lamina. Its form is usually triangular, with base forwards and straight; but if concave, its inner basal angle projects forwards as the posterior paraseptal cartilage. The anterior paraseptal cartilage in the lower mammals commences anteriorly at the inner part of oir The Primordial Cranium of Hrinaceus ewropeus 243 the hinder border of the lamina transversalis anterior by a transversely or vertically deep root, depending upon the animal; it grows backwards by the side of the jower part of the septum and ends posteriorly in a sharp point close to the inner end of the posterior wall of the nasal capsule. When that wall comes to be hollowed out and thrust back by the backward growth of the nasal sac, the lamina transversalis posterior is formed as the floor of the hollow and the posterior end of the anterior paraseptal cartilage fuses with its inner basal angle; by the still greater thrusting back of the posterior wall of the capsule, and perhaps too by intrinsic growth, this junctional region is drawn out into a slender cylindrical rod. Such is the mode of development in Dasyurus; it is similar in Lepus, and in all where there is a complete common paraseptal cartilage. Now this common para- septal cartilage has on its medial side the corresponding lamella of the. vomer, and the vomer tends to invade and surround the narrow cylindrical part and, as we rise in the scale of mammals, to replace it altogether. That condition is well seen in the cat, and in such a condition the vomer makes use of the perichondrial connecting link between the posterior and anterior paraseptal cartilages—in other words, ossifies at its expense. This is very well seen in the ferret and in man. Meanwhile certain changes are taking place in regard to the anterior paraseptal cartilage. As already said, that cartilage arises by an either transversely wide or vertically deep root from the anterior paraseptal cartilage. In Dasyurus it. is wide transversely, in the ruminant it is deep. But in Erinaceus and Talpa the commencement of this cartilage is quite narrow; it looks in fact, as seen from the side, as if there were a large notch in it below. This notch is occupied by the commencement of the paraseptal process of the pre- maxilla, which process is continued backwards along the median side of the anterior paraseptal cartilage, which behind the premaxillary notch has become bilaminar. When we examine the anterior paraseptal cartilage in the ferret and in the cat, we find that the narrow connecting piece above the paraseptal process of the premaxilla is absent, and the bilaminar part is left stranded by the side of the lower border of the septum, in direct continuity with neither lamina transversalis anterior nor with the posterior paraseptal cartilage (if that exists). Its isolation has in each case been brought about by the presence of bone. It still retains the bilaminar form, and the organ of Jacobson rests in the hollow between these lamin. When we reach man, not only has the anterior paraseptal cartilage become isolated as in the cat or ferret, but it has lost its lateral lamella, and it no longer retains any relation to the organ of Jacobson, for that lies stranded high up on the septum nasi. So complete is the separation between paraseptal cartilage and organ that Michalkovics supposed that the organ 244 Professor Edward Faweett is not really the organ of Jacobson, but the duct of a septal nasal gland. There can, however, be little doubt that this structure is the organ of Jacobson. In man a still greater change has taken place in the cartilage of the solum nasi, for the lamina transversalis anterior has disappeared, so that in him the fenestra narina and the fenestra basalis have run together to form one large vacuity, the rostro-ventral fisswre of Gaupp. THE VISCERAL SKELETON (PI. 3). This consists of, from above downwards, the mandibular arcade, next the hyoid arch, then the first, second, and third branchial arch cartilages. The mandibular arcade consists, from behind forwards, of the incus cartilage, the malleus cartilage, and Meckel’s cartilage, the two latter being in direct histological continuity with one another. The incus cartilage is of the usual form, with a short posterior process lying in a fossa incudis above and medial to the crista parotica, and a long processus which descends somewhat medially and articulates through the medium of a well-marked meniscus of connective tissue with the head of the stapes cartilage. The front of the body of the incus articulates in the usual way with the head of the malleus cartilage. The malleus cartilage is massive, and sends downwards and then suddenly medialwards a strong manubriwm; from the front of the malleus cartilage Meckel’s cartilage passes forwards with a slight bend downwards and finally upwards in a generally convergent direction towards its fellow of the opposite side. It preserves in an unusual degree a cylindrical form in the greater part of its forward course. It soon comes to lie against the medial side of the bony mandible—at, in fact, the root of the processus coronoideus,—and it continues its course onwards on the medial side of the body of the mandible, becoming somewhat flattened from side to side as it: nears the anterior extremity of that bone; then, about the spot where the bone comes to an end, the cartilage fuses with its fellow of the opposite side, and the two fused cartilages project forwards for some distance beyond the bone in a somewhat conical spinous process. Nowhere at this stage is the cartilage enclosed by bone, nor does it show any sign of ossification at this stage. The hyoid cartilage. This consists of two parts: a long hinder part fused directly to the lower edge of the crista parotica and passing forwards from thence with a somewhat sinuous course; having nearly reached the anterior extremity of the thyro-hyal, it suddenly ends, to be succeeded by an ovoidal cartilage which is clearly the cerato-hyal. The stapes, which et a ee ee ee ee ee ee ee ee ee ee ae ee The Primordial Cranium of Hrinaceus ewropeus 245 developmentally belongs to this arcade, is at this stage freed from any connection with it, and is only remarkable in being perforated by a very large stapedial artery. The thyroid arcade consists of a large thyro-hyal which posteriorly is directly continuous with the upper hinder angle of the thyroid cartilage ; anteriorly it. articulates with both the cerato-hyal and the body of. the hyoid (Parker’s basi-branchial). The thyroid cartilage, or cartilage of the second branchial arch, is of large size, and consists of two small ale fused near the lower part of their medial borders. The postero-superior angle of each ala is directly continuous with the hind end of the thyro-hyal, whilst the postero-inferior angle articulates by a well-marked inferior cornu with the side of the ericoid cartilage. Each ala is perforated by two large foramina, which lie in the same vertical line. The cartilage of the fifth arch, viz. the cricoid cartilage, is of the usual form of completed cricoid, and articulates in the usual way with the thyroid cartilage and the two arytenoids. The large size of the thyroid cartilage is in striking contrast to that in the mole, in which the ala of the thyroid is in the form of a cylindrical rod, from which, if Symington be correct, one may conclude that the vocal powers of Erinaceus are much more highly developed than those of Talpa. THE OssEOUS SKELETON (Pls. 1, 2, 3, 4). The following bones are now ossified: the parietals, frontals, nasals, premaxille, maxille, palatines, vomer, squamosals, and mandible. The os paretale (Pl. 3) is of comparatively large size, covers the upper half of the parietal plate of cartilage, the whole length of the orbito- parietal commissure and a part of the ala-orbitalis as far down as the sulcus for the onward continuation of the supraorbital artery. The os frontale (Pl. 3) is comparatively small, overlaps the ala orbitalis as far down as the same sulcus, and then passes forwards along the outer surface of the spheno-ethmoidal commissure as far as the frontal prominence of the pars intermedia of the lateral wall of the nasal capsule, which it covers only to a slight extent posteriorly. The os nasale (Pls. 1, 3) is very small at this stage, and lies upon the hinder part of the tectum of the pars anterior of the nasal capsule. The os ineisivum (premaxilla) (Pls. 1, 2, 3, 4) is well formed, and consists of a body which underlies the processus transversus of the lamina trans- versalis anterior. The body gives upwards and backwards from the hinder and outer part of its upper surface a short process, obviously the com- 246 : Professor Edward Fawcett mencement of the frontal process of the os incisivum, and a gap separates this from the corresponding process of the maxilla. From the medial end of the body of the premaxilla a long paraseptal process (Pl. 2) is given off; this passes at first medially into the notch between the medial lamella of the anterior paraseptal cartilage and the lamina transversalis anterior, no doubt causing that notch, as Fischer observed in Talpa. The paraseptal _ process now goes directly backwards on the medial aspect of the medial lamella of the anterior paraseptal cartilage, and is so continued for about — half the length of that lamella. It is not possible to say whether it ossified independently or not. The mawilla (Pls. 1, 2, 3, 4) is of large size, and consists of a body with frontal, palatine, zygomatic, and outer alveolar processes. Between the outer alveolar and the palatine process tooth-buds are present. The body is perforated by a very large infraorbital foramen (Pl. 4), through which runs the infraorbital nerve. Above the foramen the broad frontal process ascends over the lower part of the maxillary prominence of the pars intermedia of the nasal capsule, and at the same time over the naso- lacrimal duct. The palatine process is of very large size, passes medially and horizontally towards the middle line, forming a floor to the greater part of the fenestra basalis of the solum nasi; from the junction of the palatine process with the body of the maxilla the external alveolar process depends. Between the anterior edge of the palatine process of the maxilla and the posterior edge of the corresponding process or plate of the pre- maxilla is the primary choana, through whose inner end the naso-palatine duct and the terminal part of the duct of Jacobson’s organ project. The zygomatic’ process of the maxilla is a great length, and passes backwards half-way under the cavity of the orbit to end in a free pointed extremity. To its medial side tooth-buds are found, so that perhaps the term alveolo- zygomatic proccss would be a better one for it. The zygomaticum is only ossified to the very slightest extent, and no attempt has been made to model it. It, however, may be seen with the . microscope behind the pointed end of the alveolo-zygomatic process of the maxilla. There seems to be some difficulty about the zygomaticum of insectivora. In some cases it is said to be absent, eg. Centetes; in others very small; but Fischer, in his description of the model of Talpa, says “the jugal is found lateral to the maxilla, passing backwards from the latter as a cylindrical arch (fig. 2). The infraorbital nerve runs over the attach- ment of this arch. In the full-grown animal it is attached by two limbs to the maxilla; the upper limb here fails. It has not yet reached the squamosal behind, but small masses of bone in the tissues show its future course. * a ‘ r * a ae sealed i el he The Primordial Cranium of Hrinaceus ewropwus 247 In my own model of Talpa (19 mm.) (Pl. 21) there certainly is a cylindrical rod passing backwards from the body of the maxilla below the infraorbital nerve, and there is no upper limb going from it to the maxilla to complete the infraorbital canal. In my Erinaceus model (25 mm.) this cylindrical rod likewise goes back from the maxilla, and has an upper limb completing the infraorbital foramen. But this I have already described as the alveolo-zygomatic process of the maxilla, for I cannot see that it is separate from the maxilla; nor is it so in Talpa, not even at the 15-mm. stage. Then too its relation to the tooth-buds, I think, proclaims its - maxillary nature, since some of the alveoli are formed in part of it. The jugal or malar bone, then, is nearly absent at this stage; what there is of Squamosum. ZySomaticum Maxilla . Fic. 15.—Erinaceus ewropeus. Right lateral aspect of adult skull. it occupies a little, a very little, of the gap between the alveolo-malar zygomatic process of the maxilla and the anterior end of the squamosal. In the adult Erinaceus (fig. 15) a quite well-marked zygomaticum is present, but it occupies, perhaps, a somewhat peculiar position. What happens is that the alveolo-zygomatic process of the maxilla articulates with the squamosum, and the zygomaticum is placed lateral to that articula- tion, hiding it from view as looked at from theside. Both alveolo-zygomatic process of maxilla and zygomatic process of squamosum as they approach one another are bevelled at the outer surface, and in the resulting concavity the zygomaticum rests. The squamosal (Pl. 4) is a small bone which commences behind opposite the middle of the outer side of the short process of the incus cartilage ; it increases rapidly in height, and perhaps reaches its maximal height opposite the incus-malleus joint. Here its upper margin widens sufficiently to form a surface, over which lie the hinder fibres of the temporal muscle. The bone rapidly diminishes in size as it is traced forwards, and it ultimately 248 Professor Edward Fawcett terminates opposite the hinder edge of the processus coronoideus of the mandible. For a considerable part of its length it is separated from the condyle of the mandible by the still cellular discus articularis, above and below which no joint cavity is as yet demonstrable. . The palatinum (PI. 2: fig. 8) is a somewhat curious bone, consisting of a palatine process placed for the most part in front of its vertical plate and only connected with the latter by a narrow stalk, and that to the lower part of its interior edge. The palatine process is imperfectly ossified, and from a common stalk connected with the vertical plate it spreads forwards under the lamina transversalis posterior of the solum nasi towards the palatine ~ process of the maxilla, from which it is, however, separated by a considerable interval. The vertical plate really arches somewhat upwards, forwards, and inwards to reach the infero-lateral aspect of the “central stem,” and stretches forwards from the level of the cranio-pharyngeal canal as far as the under aspect of the hinder part of the cupula nasi posterior. It bounds the hinder part of the naso-pharyngeal duct laterally (fig. 8). On its supero-lateral aspect is placed the spheno-palatine ganglion. From its infero-lateral angle the pterygoideus internus arises. The periosteum, in which the upper edge of the vertical plate is embedded, seems to be practically directly continuous with that of the hind end of the vomer. The vomer (Pl. 2) consists here of two long narrow lamelle, each of which commences behind at the medial edge of the corresponding lamina transversalis posterior, to which it is closely applied (fig. 10). Where the lamina ceases, the vomer is continued forwards infero-lateral to the septum nasi to the interval] between the hinder end of the organ of Jacobson and the septum. Here the two separate plates of which the vomer is formed come to lie very close to one another, and it is here that they will first unite. From its position there can be no doubt that the vomer is primarily a covering bone to the medial edge of the solum nasi—in other words, to the common paraseptal cartilage when that exists and to the lamina transvers- alis posterior. Its relation to the septum as a covering bone—i.e. when the two lamelle have united below the septum—is clearly a secondary one. The hinder end of each lamellew reaches very close to the vertical plate of the palate bone behind; the anterior end of the vomer is separated by a considerable interval from the hind end of the paraseptal process of the premaxilla. The mandible (Pls. 3, 4) consists of two separate halves, each closely applied to the lateral aspect of the corresponding Meckel’s cartilage. The constituent parts of each half are a body splitting above into outer and inner alveolar walls, a coronoid process, an angle, and a condyle. The common mass from which angle coronoid process and condyle spring may ‘ op Le od ee ee Sa Ae The Primordial Cranium of Hrinaceus ewropeus 249 be termed the ramus. The body is pointed anteriorly, and does not reach so far forward as the end of Meckel’s cartilage ; about a third part of the way back from the anterior end a long oval mental foramen is met with, placed much nearer the upper than the lower border. No sign of any accessory cartilage is met with either at the anterior extremity or in the coronoid process, nor is any to be observed in the angle. The condyle is composed of dense cellular tissue which contains bone in its core, and over this cellular condyle a discus articularis of wedge-shaped form is recognis- able, but no joint cavities are present (fig. 7). No other bone such as one associates with the craninm is as yet ossified. COMPARISON BETWEEN THE CRANIA OF ERINACEUS AND TALPA. The differences between the cranial and the visceral skeleton of Erinaceus and Talpa are not many, but some are very striking. The nasal capsule, especially its pars anterior, is relatively shorter in Erinaceus than in Talpa, and there seems to be no foramen epiphaniale perforating the tectum anterius. That in Talpa is very sinuous and difficult to follow. There is no opercular process on the pars canalicularis of the auditory capsule such as exists in Talpa (Pl. 15). The Basi-cochlear fissure is absent in Erinaceus. The cranio-pharyngeal canal remains as a large circular aperture in Erinaceus, but disappears at an early stage in Talpa. The tectum cranii posterius in Erinaceus is very primitive, resembling much more that of Bos than that of Talpa. The foramen opticum is surrounded by a cartilaginous ring, of which the lower part is the ala hypochiasmata; in Talpa, owing to the rudi- mentary condition of the eye muscles, no ala hypochiasmata is present. As this communication is but part of a series, no attempt has been made to introduce the mass of literature which has made the history of the chondrocranium ; only such references are made as directly bear upon the subject. At the end of the series an attempt will be made to summarise the results, and at the same time to point out more precisely how the results confirm or differ from those hitherto obtained in mammals. All the plates, with the exception of 7a (which is by my former pupil J. L. Delicati) and those bearing my own initials, are by Mr S. A. Sewell. The figures are by myself, For the embryos represented by the plates and figures I am indebted to Prof. J. P. Hill, Prof. F. Wood Jones, Dr G. Barker, and my former pupil Dr J. H. Morgan. VOL. LU. (THIRD SER. VOL. XNI.)—JAN. 1918. Age &: 250 The Primordial Cranium of Erinaceus ewropwus CHIEF REFERENCES. Vorr, “‘ Das Primordialcranium des Kaninchens,” A natomische Hefte, 1 Abteilung, 116. Heft (Bd. xxxviii., H. 3). Fiscuer, E., ‘“ Das Primordialeranium von Talpa Europea,” Anat. Hefte, Abt. 1, H. 56/57 (Bd. xvii.), 1901. Bok, “ Entwickelungsoorginge in der occipitalen Region des Primordial- craniums beim Menschen,” Petrus Camper, Deel ii. Afi. 5, 1904. Fawcett, Various communications to Journal of Anatomy on chrondocranium, especially the reconstruction of the head of a 30-mm. (Bryce) embryo. Weser, Die Sdéugethiere. FLower and Lyppexer, Mammals, Living and Extinct. Parker, W. K., Phil. Trans. Roy. Soc. London, 1885. Jowrn. of Anat.) [PLATE I, =a ieee eens 2 CIS SUG, pe ~ ee — —-_ - — For. dorsale. 7 xg 3 » _ — — — Processus alaris sup. ~ oe ® — — Sulcus dorsalis. Be Tees oN Maxilla, le tice Mind — .- Prom, frontalis. Recess, frontalis. . .. — Comm, sph.-eth. Eth.-turb, IT. -_ Fiss, orbit. ras. —~—- Os frontale. _ Lam. trans. post. Tect. nasi post... _ — Ala orbitalis. _ For. opticum. -——_—-— Cart. pterygd. _ — Ala temporalis. Comm. orbit. pariet... + For. carotic. — Pars coch. ~ . Parietale. For. faciale.,_ . Pars chordalis. Parietal plate. _ fossa subare, int. _ Canal, hypogloss. _ For. endolymph.. . — Cart. exoce. _ Cart. supraoce, _ Erinaceus ewropeus, Chondrocranium of 19-mm. (Hill) embryo, from above. Professor EDWARD FAWCETT. — Parstrabecularis. a Te ) Journ. of Anat.) [PLate II. ---- Cupula ant. ec~--- Process. cupularis. ~~~ Fenestra narina. _2--- Process. alar. sup. Proc. Jat. vent. Bhs, ._ —— Incis. pretrans. _ Sule. naso-lac. and Duct. nas. lac. - ~ - —- lam. trans, ant, ~- Proc. trans. Incisivum. - - -- - a Proc, parasep. incis. -- — -# = ii i” Se Cart. parasep. ant. ~~ — Max. turb. Proc, pal. max. — —- Sule. ant. lat. ~~ ——— — Prom. max. ciel en em me Eth.-turb. I. _—-——— Vomer. Proc. zyg. max. ~~» —— = --——-—--- Eth.-turb. I. _-_.—-— Comm. sph.-eth. Parietale. .-~-—-— ---—— Law. trans. post. -—-— Ala orbit. ~— Pars int. orb. nas. —— Ala hypochias. Palatinum. .—— — oh Pa -~—— -Ala temp. Can, hypophys.~— — —-- —Cart. pteryg. Squamosal.-— —~—-Pars trabec, » _ _Ant. trab. coch. comm, and For. carotic. *, Malleus cart. ~ Can. facialis. — et: 4 ai "ag Crista parotica. - ~ Pars coch. For. tymp. et peri- lymphat. For. jug. oie —-— -Can. hypogl. Atlas, -—— Erinaceus ewropeus. Chondrocranium of 19-mm. (Hill) embryo, from below. Professor EDWARD FAWCETT’. Journ. of Anat.) ‘okquia ([[TF{) “MU-gT JO UNTUBIOOIPOY{O Jo qoodse [B10qR] Yor] ‘snandowna snomUpeg “plooltg ~~ sproadyy *qavo Upy ¥ : *pedy-or (UT, no | nw a *prlol {qs “001g ; of Youwaq-seg s ' “sso[30d Ay ‘fuary Lc Se A ’ Path “stes0d ws} Vly a wt ; “ “ af ! 1 2 “par *sdvo-"900 “S81 _ “B[RGIpaBy, ' “go1gored B4SLI) - “qul ‘olules “WOIg “qsod ‘oymtes "WOlg = — *BUIO}XO ‘O1VGNS BSSOY -- = *900X0 "QIBO —— [aYOIT “8D ‘tuvdudy uoulsey, ~ ; “st[BPoVJuAdNs WMO -— ‘dns — — .1e[e "001g; -—— ee _‘tndno "001d. _ ‘Burieu ‘souog ‘ = - ! i i t as | z 8 | : : ' { Fae ~ : ! ! ( ' i] i z a | 1 t * . . . ® aS g ' | ‘ } z a 5) = s o 2 a : I | I ! 8 5 Ss = on 3 : 3 s | » i} eS { { I 1 | s S = = = 25 4 bel 1 ae a ee ake $8 a. -S ; as i 23 : = o 4 : i | . 1 f on igh ‘ Fe S =] ° see Bites @ a ence Pe Ped Bees yack ea 4 “3 ik scent “Aes * a . . * . 7 ° = * : e- = S&S a = ae a} _ ape Sues RCAF Rhee ae eee oe 2 . Ze Py 3 5 S ~~] a = 7 Q = = Saye ra e ® = Ss 3 > 3 3 be 3 z= 3 z o = = = 6.09 1s BS See gE <¢ = g TK - ays fea ls os a | a = = er = AS & - a > & : 8 : a & : ° pa E Ag. te SS g@ ¢g a S £é E he. te IES Boeke = & 5 = i ee a £ n S S = E Z Professor EDWARD FAWCETT. (Puare LV. “Journ. of Anat.| “dns ‘ele ‘001g ‘que vyndng “eUulIeu “ued -ofaquia ({[UF{) “UU-gT Jo UNTTRLOOIPUOYO Jo qoodse [R1aqe] IySIy = “-snandouna snaonUry “UUNUTZRTed *plooltg s “plosdyy “JAv9 “RLV \ : “pedy-OrdLL ‘\ . *wnsowenbs jerpouvaq-iseg » \ \ ‘ ‘ "yeAy-owws9g » \ ne \ \ eayh, \ f , \ \ \ / “diay ery Be oN / ‘ \ ‘ ' \ ‘ ’ \ \ “Jur snnbi[qo “osnyy ‘Jurcoreueg ‘aT BIUSU “IO \ “ETN IPUBIT *[aYOoTV “94¥,) “elTEXe “CUMAISIOUT *SURIy ‘N0I1dg ; i] ' i ' ' ' “que savd ; ! ’ “oer "seu ONG “a[eS¥N “qel “Que ‘oT ' ‘ \ \ : . \ 1 ' \ \ ——— ‘[Apuoowred ‘901g ———— “yur ‘sd¥o-"000 ‘SSI = —— — — "900X0 “4180 -—- ‘pedeqys yy ~-—-— ‘dns ‘sdvo °000 ‘ssiq “g00Budns “"y4ABD \ 1 ' 1 ‘a[eqaleg ug . “quoay “WO ‘aTVWUOAT ® Professor Epwarp FAWCETT. ase [PuaTE V. Journ, of Anat. | —- ato A (‘qa199 -a1d sed) isvu -ydag = —- ‘sqooer uRdIQ es" que ‘dasvied'g1e9 “que iseu “400.7 ~~ ~~ ea que “SUBIy “CeT] -=--="qung-O113 ¥ ‘sIjesaop “o[ug -—--- ‘sI[esiuop "uaq ~~ y--- *yndno ‘901g inaceus ewropeus. Septum nasi of 19-mm, (Hill) embryo, from left side. lp E vi Professor EDWARD FAWCETT. (Puate V1. Journ. of Anat.) ‘dns ‘s11e[e “901d ---- “qany “OLyy B\-- “que ‘suBsg “WET “yur “ger “901g ----- “Qing "xe “que Iseu "yoat, --- 777 ~----—‘youoovidns "889097 no snqRutoun "901d *UTqaNy-Ose N - "que ‘sso0oy . *OIUIGS BISLID “T “qang-"yjo “que xIpey “sI[By UOT “ss0ay *]RUIQIN-0F UOT -— "XBU ‘sso00q pate ‘T-qany-"Wa --— (Jat xtped) “T “qan}-"q9H “yqo-"yds “uIWOD . —- “TE qdng "a “ysod “SUB1} “WRT or “qsod isvu "y00 ----- = . - *qsod seu epndng Erinaceus ewropeus. Medial aspect of left half of nasal capsule of 19-mm, (Hill) embryo. Professor EpwARD FAWCETT. = 7 I Journ. of Anat. } [Puare VII. ! 7 2s \_ Ala orbitalis. Poet ae Comm. sph.-eth } -~ Ala temp. | anal. hypophys. - --- - :--~ 7 3 ~- Parietal plate. | Pars coch, ... .. |S). | For. carotic. te --" — aii ae --- Comm. suprafacialis, antrabec.-coch. ant. f : : ie ~- al green ~- For. acustic. sup. Pars coch. ~ Crista falciformis. ~ Prom. semic. ant. \ For, acustic. inf. : Fossa subarcuata inf. (nm. chordo-coch. - For. jugulare. Pars chordalis. --- Fiss. occip.-caps. sup. -- For, endolymph. Canal. h on en OF — Cart. supraoce. Proce. alaris. Cart, exoce, Erinaceus ewropeus. Medial aspect of right auditory capsule and environs of 19-mm. (Hill) embryo. Professor EDWARD FAWCETT. Journ. of Anat.] (PLate VIIa. Parietal plate. -- -----~-- Ala temporalis. -- ---------~74e Ala hypochiasm, --~--~- > Foramen opticum. . Ala orbitalis. ----- Sept. interorb. nasale, & Bos taurus. Left lateral aspect of chondrocranium of 19-mm. (Wood-J ones) embryo. Pars posterior. Comm. sph.-eth. (ant. part). Pars anterior. -——-------- Professor Epwarp Faworr’. | z i ahs F | Journ. of Anat.| (Pnare XIII. — a ---- He -- Tect. nasi ant. ; Prom. front. ~---- Fiss. orb. nas. \ ¥---Ala orb, + --Cupula nasi post. \ »~ Pars interorb. nasalis. y , ~- For. optic. ..-~- Ala hypochiasm. ii --- Comm, orb. pariet. —— d6©6Lrl tw Cl Ne ,.-- Fiss. trabec. sept. z. --Ala temp. } Sop neaiteccaca “ ---~-— Pars trabecularis, For. carotic. .....% = _#..- —~—.¥Fontan. sph. pariet. Fen. basi-cran, ..--- .~--~-- Parietal plate. —+ - Comm. supra- Nerv. facial. --- facialis. oss. subare. int. — ict. endolymph. ~ = For. jug. Pars chordalis -~ 7 For. endolym, ---Can. hypogl. esi ne Cart, supraocce. oe ee Cart. exoce. Bos taurus. Upper aspect of chondrocranium of 19-mm. (Wood-Jones) embryo. Professor BRnwarn Rawarrr ) Journ, of Anat.) [PLaTe XIV. | Pa Cart. supraoce.-- -- ~---- pa mare oo mn Cart; Z000. Fiss. occ.-caps. sup. ------. — j ~~ —-Fiss. occ.-caps. inf. Comm. occ. pariet. -- --- - —=* ~-- Cart. exocc. Prom. semic, ant. _ Prom. semic, lat. Crista parotica. --------6 Condyle. TWMCUB <> = sew sone , F —- Proc. paracondy. e —-- For. jug. ‘w-- Nerv. facialis. ‘ : 4 : ~3 Nerv. petros. sup. maj. --~-- =~ es ~ . ~-- Proc. styloid. Nerv. trigem. ~~~ —-~—-- Cart. Meckel. Ala temporalis. ---4 4 Ala hypochias, ~ ; } +. .._. Cart. pterygoid. Nw Ala orbitalis. - For. opticum. -- } 3 ~~ Pars interorb. nas. Cupula post. - - ~ Palatinun. Frontale. ---~ Fiss. orb. nas, >~~~ 2% -- Sulc. lat. post. Comm, sph.-eth. }-—- Vomer. Prom. maxill.-.— > Prom, front... - -- Prom. ant. - -- -~- --- Maxilla. Eee Sept. nasi. Sulo, antilats cae / 8 ¥or. epiphaniale. - -- —---- ‘ \ te adh ss Mandibula. Cart. parasept. ant, ' , Memb. mucosa. - Duct. nas. lac, -- See DEES Incisivum Fen. dorsalis. _... a Proc, alaris sup, -----Sey Memb, mucosa, - ---- Fen anb;) oso . — M. obliq. inf. Ala orbitalis. -— - ?¥For. epioptic. --— For. optic. ...- __..~ Ala hypochias. ~--- Ala temp. ee —-- Can. hypophys. EE EE ” .- For. carotic. _ Comm. trabec. coch, ant. . i B-— Comm. trabec. coch. Pars coch. — Lj = = — . : ; post. : ; -Chorda dorsalis. ~Malleus. -N. facialis. Comm. orb, par. -=~~ Fiss. basi-coch. - -Pars chordalis. Foss. stapedii. - -Proc. styloideus. -- For, jugulare. Cupula post. - ; WA = == Occ. condyle. caps. aud, ; ee Tatusia novemeincta. Chondrocranium of 12-mm, (Wood-Jones) embryo, from below. The parts dotted are unchondritied, the lamina transversalis anterior and the anterior paraseptal cartilages are procartilaginous. Professor EpwAarp Fawcer’. Journ, of Anat. | {[Puate XVII. f , i —- Tect, nasi ant. Sept. int. orb. nas. Prom. frontalis. _.-—---~~ . Lam, trans. ant. --- Eth.-turb. I. — Cart. paras, ant. Comm. sph.-eth. Fiss, orbito. nasal. —- Ala orbitalis, -— ? For. epiopticum. -\— OE oe eee For. opt. ~--Cart. Meckel. —-Ala temp. «~—-Can. hypophy. . For. carotic, ym) trabec. coch. ant. _ mr trabec. coch. post. — t Pars coch. P lat ‘ar. plate. Fossa subare. ant. int. For. tymp. Prom. semic. ant. Pars chord. | Chorda dors. Prom. semic. ant. ._. For. endolym. ~--—- For. jug. Can. hypogloss. wee 4___ Cart. exocc. Tatusia novemeincta. Chondrocranium of 12-mm. (Wood-Jones) embryo, from above. Lamina transversalis anterior and anterior paraseptal cartilages are procartilaginous, All areas dotted are as yet unchondrified. } Professor Epwarp FAaworrr. Journ. of Anat. ] {Puate XVIII. 33°° Sept. nasi.--- fs Proc. alaris sup. -~— Lam. trans. ant. - Pars ant. 7 — Incisivum—pars paraseptalis. . Ductus naso-lacrimalis. Sule. ant. lat.-— Cart. parasep, ant. --~-- oS ee ~~~ Maxilla. Prom. maxillaris. - Maxillo-turb, - Cart. parasep. post. - Pars interorb. nas, - - - Frontale. Lam. trans. post. = _ __ Ala orbitalis. .._--- Palatinum. Cupula nasi post. ? For. epiopticum. Ala hypochias. Fontan, sph. par. Ala temp. Comm. trabec. coch. ant. Pars cochlearis. Malleus. _ Incus. . Foramen, tymp. Aud. caps. For, jug. Can. hypogloss. Proc, paracondyl] Tatusia novemcincta, Fiss. orb. nasal. For. opticum. Fiss. trabec. septalis. Pars trabecularis. For. caroticum. Chorda dorsalis. Fen. basi-cranialis. - Nervus facialis, ~ Fiss. basi-coch. Stapes. Pars chordalis. Proc. styloideus. Fossa stapedii. Unchondrified part of auditory caps. Fein nhs = Tect. cranii post. Chondrocranium of 17-mm. (Wood-Jones) embryo, from below. The lamina transversalis anterior is unchondrified. Professor EpwWARD FAWCETT. Journ. of Anat. | Homo. (PLatE XIX. Lo” a eetatan Hypophysis. _..------ Pars trabecularis. ah wy ro sea Chorda dorsalis. none Art. carotid. ---= Chorda dorsalis. —— ~ - Art. staped. et ae os et ee enn Pars chordalis, Chondrocranium of 13°6-mm. (Morgan) embryo, from above. Professor EDWARD FAWcRETT. L 2) a v Journ. of Anat. | (PLATE XX. —---— Pituitary body. “=--~- Hypophyseal duct. ~----- Fars trabecularis, Caps. auditoria, -------~ Pars chordalis, - ‘ ‘== - Chorda dorsalis. x Ni ess DENS. Homo. Chondrocranium of 14-mm. (Barker) embryo, from above, Professor Enpwarp Faworrr. Journ. of Anat.) (Prats XXI. 23 ~------- Cupula ant. ta . ~-~ Fenes. narina. ~~~ Sule. dorsalis. -- Nasale. Cart. parasep. ant. : : “ igh stds a ' --- Maxillare. Pars interorb, nasalis. Evh.-turb. I. - ---- Frontale. - ~ Comm. sph.-eth. ~ Tect. nasi post. --- Fiss. orb. nasal. Ala orbitalis (radix ant.). Fiss. orb, sup. a... Ala temp. Cart. Meckel. -— - For. caroticum., - Comm, orb, pariet. - Pars cochlearis. Meatus aud. int. Pars chordalis. Caps. audit, - For, endolymph. - For, jug. - Can. hypog). - Can. hypogl. a OO a Tect. cranii post. Dasyurus viverrinus. Chondrocranium of 9°3-mm. (Hill) embryo, from above. Professor Epwarp FAWoETT. Journ. of Anat. | [Puate XXII. oe. === Hypophysis. = =k ~~~ Carotid. +- Trabecula (Pars trabecularis). ~- Medial wall of cochlear cap- sule chondrified. =~-~- Cochlear duct. ~-- Pars chordalis. a === Chorda dorsalis. eo-== Dens. pe Talpa. Chondrocranium of 11-mm. embryo, viewed from below. Professor EpwArD FAWCETT. z 3, CONGENITAL STENOSIS OF THE PULMONARY ARTERIAL VALVE, WITH PATENT FORAMEN OVALE, IN AN OLD MAN: WITH REMARKS ON THE VALVULAR FACTOR IN r CARDIAC ACTION! By ALExanpDER Morison, M.D., F.R.C.P. J.. W., aged 72 years, of no occupation, was admitted into the Great Northern Central Hospital on August 31st, 1914, under the care of Dr Tidy in Dr Symes Thompson’s wards. He died on September 4th, 1914, and the necropsy took place on the same day. The clinical. notes of the case found are very meagre. The patient appears on admission to have had pain in the right side, with a temperature _ of 101° Fah., which gradually rose to 104°°6. The pain is stated to have been due to pleurisy, and the patient died, as mentioned, on the fifth day after admission. The previous history relates that he had had valvular disease of the heart for forty years, and an apoplectic seizure with affection of speech twenty years prior to his death. The diagnosis of his condition recorded is that of right hemiplegia, hemorrhage into the internal capsule; glaucoma ; and congenital disease of the heart. The cardiac diagnosis was probably made after death, and the cerebral condition was probably embolic, as will. appear on reading the details of the necropsy, although the head was not examined. There is no mention of the patient’s having been cyanotic, but it may reasonably be assumed that he was not. ; T accidentally saw the heart in the mortuary and, recognising its interest, directed it to be preserved. I have obtained Dr Symes Thompson’s per- mission to describe the specimen. The examination of the body was made and the following notes taken by Mr Ernest Shaw, M.R.C.P., the Pathologist to the Hospital. I have added some details, which are placed in brackets. “The body of an obese old man. The head was not examined. Neck: tissues very congested; larynx, trachea, and cesophagus natural. Lungs: extremely congested, almost black in colour. Pericardiwm: universally and tightly adherent to the heart and great vessels. Heart: 164 ounces; enlarged generally ; muscle very soft and a little pale; cavities filled with fluid blood and clots; auricles large and communicating through a very large double opening in the septum. [The actual measurement made in the Paper read at meeting held on June 29th, 1917. 252 Dr Alexander Morison formalinised heart, which, like other additional measurements stated, may be regarded as less than in the recent specimen, was 3x3 centimetres. The loose membranous curtain which should have closed the foramen was perforated by two large openings of about equal size, namely, 15 x 10 millimetres.] The pulmonary arterial valve’ was conical, with a small opening at the apex; its segments fused, thick, and containing much calcareous material, but without recent change. [The actual size of the stenotic opening was 6x6 millimetres. The calcareous change was in the anterior segment of the arterial opening, at a point where the twé cusps which formed the valve met. The right auricular wall was very slightly hypertrophied. The coronary sinus was normal, as was the tricuspid valve, the cusps of which were bulged towards the auricle. The wall of the right ventricle was much hypertrophied, measuring, at its thickest, 14 centimetre, and the length of the cavity from the pulmonary valve to the apex was 7 centimetres. Some shrinkage of the cavity was probably due to formalin.] The aortic and other valves were natural. No vegetations anywhere. [The left auricle was normal, and the wall of the left ventricle at its thickest rather more than 1} centimetre. That is, the muscular development of the left did not much exceed that of the right ventricle.] The aorta was dilated and slightly atheromatous. The pulmonary artery was large and its walls thin. [This thinness of an unused or comparatively little used vessel has been noted before. The condition of the ductus arteriosus is not mentioned, and the vessels in the specimen are cut off short of its site. The septum ventriculorum is entire. ] Abdomen: the stomach and intestines are natural; the liver weighed 55 ounces; there was slight thickening of its surface. It was congested and firm. The gall-bladder, pancreas, and suprarenals were natural; the spleen weighed 7 ounces, was large, and contained a large yellow infarct ;' the kidneys weighed 5 ounces each, were congested, rough, and granular, and showed several small cysts in the cortex. [That is, there was evidence of chronic interstitial nephritis. } Remarks.—Vhe interest of this case consists in the evidence it affords of a congenital stenosis of the pulmonary arterial valve in a marked degree being not incompatible with long life—the bearer having outlived the allotted span of man—and in the collateral conditions obviating stagnation and promoting circulation, which rendered such longevity possible. All degrees of closure of the pulmonary arterial orifice occur, from partial stenosis to complete occlusion, which, in ‘its higher degrees, is usually associated with imperfection of the interventricular septum, patency of the ! This was probably an old embolism of non-septic character, and argues a like origin for the uninvestigated cerebral lesion, Congenital Stenosis of Pulmonary Arterial Valve, in an Old Man 253 foramen ovale and permeability of the ductus arteriosus, or with one or other of these conditions. The nearer the heart approaches to the normal in its ultimate develop- ment, naturally the more compatible is the condition with easy circulation and with life. With any considerable degree of mechanical impediment to the outflow of blood from the venous towards the arterial heart, by way of the pulmonary circuit, unless there exists a sufficient collateral avenue, the shorter is the life of the subject. Dr T. B. Peacock, writing in his well-known treatise on Malformation of the’ Heart (2nd ed., 1866, p. 78), states that “the particular situation occupied by the obstruction [of the pulmonary arterial orifice, whether in the conus, at the valves, or in the vessel] is of much less importance . . . than the state of the inter- ventricular septum and the condition of the foetal passages—the foramen ovale and the ductus arteriosus,—as upon the persistence of one or other of these passages depends the possibility of the circulation being at all maintained. 3 “In by far the largest number of recorded cases . .. the septum of ~ the ventricles is reported to have been imperfect, so that the aorta derives its blood from both ventricles, thus indicating that the defect at the pulmonic orifice occurred at the early period of foetal life. When this is the case, the blood readily passes from the right to the left side of the heart and so enters the systemic circulation. In the comparatively few cases in- which the obliteration [of the pulmonary orifice] occurs after completion of the septum, the foramen ovale must necessarily form the channel of communication between the two sides of the heart. By this means, however, the circulation is apparently less readily maintained than in the former case, the children in whom such a condition existed having all died very early, while some of those in whom the septum of the ventricles was imperfect survived several years.” In The Practitioner for 1888 (p. 112 et seg.) I gave details of -a very marked case of this kind with imperfect septum ventriculorum, slight patency of the foramen ovale, and closure of the ductus arteriosus, which survived, but with marked cyanosis, for seven years. Given, however, sufficient patency of the foramen ovale in such cases, it is questionable whether there would be any advantage over it in a defective interven- tricular septum. Peacock mentions the case of a young man, 20 years of age, who presented much the same conditions as does the heart I have shown. ‘That is, he had stenosis in a considerable degree of the pulmonary arterial orifice, a patent foramen ovale, and closed interventricular septum. This case would probably have survived longer, had he not died of pulmonary tuberculosis (op. cit., p. 112). Finally, Peacock also mentions 254 Dr Alexander Morison the case of a girl, 16 years of age, who showed definite, but less, contraction of the pulmonary orifice than did the young man referred to, in whom both the interventricular septum and the foramen ovale were closed. She also died of pulmonary tuberculosis (op. cit., p. 122). It is manifest, therefore, that the greater the obstruction at the pulmonary arterial . orifice, the more patent must be the collateral avenue towards the left heart to secure easy entrance of the blood into the left heart and relieve that tension of the blood in the chambers of the right heart which has quite appropriately been termed “back pressure.” To the freedom of patency exhibited by the foramen ovale in the heart of the old man which I have now shown, he undoubtedly owed his long life, the condition of his left heart being normal. The unfortunate meagreness of the clinical notes of this case deprives us of information which might have been interesting. The patient appears only to have been aware of having “valvular disease ” of the heart for forty years, which would make him about thirty when that condition was discovered, and it may be assumed that prior to that date he did not suffer much cardiacally. Having apparently become hemiplegic twenty years before death, and probably compulsorily without occupation, he must, during this period, have led a quiet invalid life. The point of importance, however, is that the bearer of the malformation was a well- nourished man in advanced life, who had borne the mechanical cardiac lesion described from his cradle. The renal state was, in all probability, secondary to the circulatory conditions, but may in some measure have contributed to the hypertrophy of the left ventricle, which was, however, not great. That the site of the lesion is an important, consideration in this con- nection is proved by experience. Obstructive conditions, in themselves and apart from their degree, appear to be more important up to and including the mitral valve, than are those allowing regurgitation into the right heart and left auricle. Up to this point, blood is entering the central organ of circulation, and propulsion is a less dominant factor than attrac- tion or absorption, which is pulmonary as well as cardiac. In the pulmonary factor in the circulation, attraction is more potent than propulsion, although both forces are exercised; in the cardiac factor the reverse is the case, propulsion is more important than attraction, although here also both forces are in play. Lesions allowing regurgitation from the aorta into the left ventricle, on the other hand, which expose that chamber to increased pressure from the systemic blood column, appear to be, in themselves, more important for the efficiency of the heart than obstruction at the aortic orifice. In this matter, however, the degree or amount of the lesion, whether regurgitant or obstructive, has importance. ‘This is testified Congenital Stenosis of Pulmonary Arterial Valve, in an Old Man 255 by the amount and character of muscular hypertrophy behind the site of the lesion. Even when obstacle to the circulation is essentially vascular, and not intracardiac, this point is patent to the pathological anatomist. The ventricle lying behind an aortic aneurism may be absolutely normal in its dimensions if the aortic valve be competent, while a leak in this valve may induce a cardiac hypertrophy, making possible the comparison of the human heart with that of an ox. To gauge the amount or degree of a valvular lesion, and to determine its site, is the task of the clinical physician. The first of these duties requires experience and judgment, and is, even then, a matter of approximate estimation; the signs indicating the site of the lesion are, of course, well known, although not always so easy of inter- pretation aright as might be supposed. JOURNAL OF ANATOMY THE INTERNAL MAMMARY LYMPHATIC GLANDS. By E. Puaiuip StiBBE, Demonstrator of Anatomy, University of Durham College of Medicine. THIS investigation was undertaken at the suggestion of Mr Sampson Handley, who speculated on the possibility of removing the internal Internal mammary artery. Internal mammary vein. Cut edges of costo-sternal fascia. Cut edges of transversus thoracis muscle. Fic. 1.—Exposure of internal mammary lymphatic glands from behind. mammary glands in cases of carcinoma mamme, especially where the inner part of the breast is affected, and these glands therefore the more likely to be involved; it has long been accepted that their afferents include vessels - from the inner third of the breast. Being led by this speculation to “look up” the subject, I was unable to find full information as to numbers, VOL. LIl. (THIRD SER. VOL. XIII.)—APRIL 1918, 18 258 Mr E. Philip Stibbe position, or accessibility. I have therefore examined the glands in a series — of sixty ! subjects, with a view to ascertaining— (1) Whether they are comparatively constant (a) in number, : (6) in position. . (2) Whether they would be readily and safely accessible in the course of breast-removals. The method consists in simple dissection, without any preparation of the part. In a few subjects the dissection was done from the front; the intercostal spaces were first examined, and then the costal cartilages were removed, to see if any glands lay directly behind them. a In most subjects the dissection was done from behind, on that portion of the chest-wall which is removed from the body in post-mortem examinations; this piece was so removed as to leave one or two inches of the intercostal spaces lateral to the internal mammary vessels (fig. 1). i every subject the internal mammary vessels were examined from the level of the upper border of the first rib, down to their point of division behind the sixth costal cartilage. The majority of the glands are large, dark, and firm; where they were small, pink, and soft it was still easy to pick them out of the bright and very soft fat in which they are embedded. ; TABLE OF NUMBER OF GLANDS IN EACH SPACE. Contained no One gland. | Twoglands. | More than two, gland in Space, Total, | Per cent. | Total. | Per cent. | Total, | Per cent. | Total. | Per cent, First . ; F . 4 38 OL a 76 19 16 6 5 Second . by ie We 2 91 | 76 24 20 2 2 Third . ; : ‘ 21 18 75 62 19 16 5 4 Fourth ‘ ; . | 109 91 112 9 eee see ‘ie Fifth . : d « POR ie Be 15 12 Sixth (or behind sixth cartilage) . ee |.) 88 75 | 62 | | ' The original intention was to dissect a hundred subjects, but pressure of war work has made this impossible, The conditions found have been radislenii constant to justify fairly reliable conclusions; I do not anticipate that these will require modification when the series is completed, * Only three of these were in adults, SS — si : 7 7 h: 7 The Internal Mammary Lymphatic Glands SumMMARY OF TABLE. First space : : : ‘ . one or two glands in 110 cases=91 per cent. Second space. ; ; : : ; ss DIG SOG== 5; Third space : 2 : ; 5 ; 3 Shs Sse <5, Fourth space. : ; : . . no gland in 41090°*,,--e9l. ,, Fifth space ; , . 2 é : ¥ 10D) = 8G Sixth space : ; ; ; . one gland in 70: gy ORS is Gland at bifurcation of vessels in either fifth or sixth spacein . 90 ,, =75 ,, TotaL NUMBER OF GLANDS. 259 The average total number of glands on the two sides is 8°5; that is, the average or typical case has four glands on each side, or four on one side and five on the other (fig. 2). These are one in each of the first three Fic. 2.—Internal mammary lymphatic glands of a female, aged 47. Spaces on each side, one in the fifth or sixth space at the bifurcation of the vessels, with an extra gland in one of the upper spaces. 260 Mr E. Philip Stibbe VARIATION ACCORDING TO AGE (figs. 2, 3, 4). The number of glands diminishes with increasing years. Average total number in infancy. . . 11-4 glands »9 2 age 20 to 50 . : AG ae Mee a pee se Ue se) as EO pee meee y eas » » a OO tO. ‘ PC SN fe: pe » 3 - Vent 10s, : Far 3g Re eer 6 VARIATION ACCORDING TO SEX. About one-third of the subjects were females: they did not show any definite variation as compared with male subjects. ‘ POSITION OF GLANDS RELATIVE TO THE VESSELS. The characteristic arrangement is that shown in fig. 1. Considering individual spaces, the position of the glands is constant in over three- quarters of the cases; taking all the spaces together, however, considerably more than half the subjects show a variation. TABLE OF POSITION RELATIVE TO VESSELS IN INDIVIDUAL SPACES. First Space.—Gland or glands medial to vessels in 105 =88 per cent. Second ” ” ” ”» ” ” 91 ae 76 ” Third __,, :. lateral __,, x 95=79° ,, Below the third space no gland was found medial to the vessels. POSITION IN THE CHEST-WALL (fig. 5). The glands always lie in front of or on a plane anterior to the internal mammary vessels. The latter in the upper spaces lie at a considerable depth; they are embedded in soft fat, a thick pad of which intervenes between the vessels and the internal intercostal muscle. The lymphatic glands are embedded in this intervening fat. RELATION TO CosTAL CARTILAGES. Excluding the gland at the bifureation of the vessels, not one gland was 4 found entirely hidden behind a costal cartilage. a 4 3 ee ? _ The Internal Mammary Lymphatic Glands 261 RELATION OF GLANDS TO THE PLEURA (figs. 1 and 5). In the intercostal spaces from the second or third downwards the internal mammary vessels are separated from the pleura by the transversus thoracis muscle (triangularis sterni). In the first, or first and second, spaces the artery is usually described as lying on the pleura; if this were correct aif W v S \ Ly | . ‘ 4 ‘ 1 | j i = si \, Te a | i i =s Fig. 8.—Internal mammary lymphatic glands of a male feetus. the glands, being situated, as described, in the fat in front of the vessels, are remote from the pleura. There is, however, in the first and second spaces an equivalent of the transversus thoracis muscle in the form of a thin but very definite layer of fascia lying between the vessels and the pleura. This I have ventured to call the costo-sternal fascia. The costo- sternal fascia stretches across the anterior parts of the first and second intercostal spaces, being in the same plane as the transversus thoracis muscle, with which it is continuous below. It is attached to the backs 262 Mr E. Philip Stibbe of the first and second costal cartilages. Medially it blends with the periosteum on the back of the manubrium and upper part of the gladiolus — of the sternum. Laterally, in the intercostal spaces it has a moderately well-defined free border. It is separated from the pleura by a layer of fat of varying thickness. SURGICAL CONSIDERATIONS. Removal of these glands in breast-operations seems to have been under consideration some twenty years ago by W. S. Halstead (Annals of Fic. 4.—Internal mammary lymphatic glands of a male, aged 74. Surgery, xxviii. p. 572):—“ Dr W. H. Cushing, my house-surgeon, has in three instances cleaned out the anterior mediastinum on one side for recurrent cancer. It is likely, I think, that we shall in the near future remove the mediastinal contents at some of our primary operations.” Describing his operation at a later date (1894), however, Halstead does not seem to have followed up the idea. . ee i i ee ee ee ee a ee aT a ee Pe tt laa cacy Maa tha The Internal Mammary Lymphatic Glands 263 Mr Sampson Handley kindly allows me to quote the following from his experience :—He has explored the second and third intercostal spaces in five cases of breast-cancer. In each space the internal intercostal muscle, with the anterior intercostal membrane, was reflected as a flap with the é a Pleura. fa ¥ h*| lf ” Fed Sy Costal cartilage. F 3 F Loy ie Internal mammary artery. aa he) py 3 ‘ J Fat. f A & Pectoral muscles. °¢: . } 2 Lymphatic gland. Anterior intercostal : membrane. t ; % : Costo-sternal fascia. Internal intercostal — iv ne muscle. : Costal cartilage. —= Ne AN vAEiL __Transversus thoracis muscle. KI | Fie. 5.—Diagrammatic vertical section of anterior end of second intercostal space. base outwards. The technique was easy, though bleeding was trouble- some in one case. In four cases no lymph-glands were found, though Mr Sampson Handley thinks they were probably present but—in the absence of induration—not easy to pick out from the fat. In the remaining case two glands were found in each space (second and third), all infected with cancer. The smallest gland was almost microscopic, but showed infiltration of the fat around it. The case was 264 The Internal Mammary Lymphatic Glands a cancer of the inner part of the breast, and although operated on before it had reached an advanced stage, bse patient died within six months, of internal recurrence. Mr Handley endorses the point I have emphasised of the remoteness of the glands from the pleura, but is of opinion, after his experience, that routine exploration of the intercostal spaces is not desirable. Mr H. Blakeway, in five post-mortem records of deaths from breast- cancer, finds no mention of the internal mammary lymph-glands; in one post-mortem of his own the glands were specially examined, and found — not to be infected. He is of opinion that routine exploration of the spaces would not be justifiable. Surgically, therefore, this paper has led to somewhat of an anti- climax; it may, however, be justified as a small addition to descriptive anatomy. SUMMARY. 1. The internal mammary lymphatic glands are four or five in number on each side, one each in the first, second, third, and fifth or sixth spaces. 2. The number of glands is greater than this in infancy, and less in old age; between the ages of twenty and sixty it is nearly constant. 3. There is no characteristic variation in the female. 4. In the first and second spaces the gland is usually (88 per cent. and 76 per cent.) medial to the internal mammary vessels; in the third space it is usually (80 per cent.) lateral to the vessels. 5. The glands are remote from the pleura, being separated from this by the following, from before backwards :— (1) Fat containing internal mammary vessels. (2) Costo-sternal fascia. (3) A thin layer of fat. 6. The glands are therefore readily accessible to operation without undue risk to the pleura. Surgical opinion, however, is not in favour of this removal as a routine measure in operations for breast-cancer. I have to thank Professor R. Howden for permission to carry on this investigation in the Anatomical Department, and for suggestions and criticism. I am also indebted to Professor M‘Donald of this College for material, and to Messrs Sampson Handley and H. Blakeway for the in- — | | 7 formation already referred to. A few of the dissections were done by students of this College; the diagrams are the work of Mr S. A, Sewell. ee ee bee ee ne : 4 1 ON ABNORMAL SEXUAL CHARACTERS IN TWIN GOATS. By EstHer Rickarps and Professor F. Woop Jones, London School of Medicine for Women. THE subject of the “free-martin” is one that has attracted the attention of many anatomists and veterinarians. John Hunter (1728-1793) and Alexander Numan of Utrecht (1780-1852) were among the first and the most distinguished of those who made this question a special study. To the breeders of cattle it has always been a curious problem, and it has proved a theme around which a host of legends—some with a basis of truth and some apparently lacking this basis—have sprung up. The free-martin has been a thing of interest, of curiosity, and of legend ; but quite recently it has become a rather more real and important matter. ‘The embryologist early saw the strictly scientific importance of these cases of irregular sexual development, and appreciated them for the light they throw upon the stages of normal sexual evolution. But of late years the problem has become of practical and economic importance—not to the breeders of cattle, who knew the conditions as far back as Roman times, but to the modern small-holder who breeds that epitome of economic zoology, the goat. When it is affirmed of a “celebrated he-goat ” that “one season every kid he sired, eleven in number, was a hermaphrodite,” it will be realised that the question of the caprine free-martin ceases to be a purely academic one. It is for this reason that the present cases are described in detail; for although the literature is burdened with records of examples of the free- martin, the great bulk of these records lack just those details which make for a more thorough understanding of the anomaly. The two goats described in the present paper (bred by Mrs H. of Enfield) were twin births, and both were feet presentations; in every respect the kids were identical. No doubt whatever was entertained regarding their sex, for both appeared to be typical females; moreover, comparison was to be had with a single nanny kid born to another mother only a few days previously. One kid was kept by the breeder, and the other was given away ; but in both cases it was noticed that within a week or ten days after birth definite changes had occurred in the external genitalia which produced a more masculine appearance. a 266 Esther Rickards and Frederic Wood Jones The change was brought about by pronounced elongation of the genital tubercle—a growth which converted the “vulva” into a more slit-like orifice on the under surface of a remarkably penis-like clitoris. A veterinary surgeon was called in to examine one of the kids, and, although noting the anomalous condition of the external genitalia, he was of opinion that the sex was certainly female, since the nipples and the mammary glands were those of a normal female of that age. It was obvious, how- ever, that the malformation of the external genitalia, which increased with rather remarkable rapidity, would render the animal useless for breeding, and it was therefore decided that the animal should beslaughtered. Before it was killed several details of behaviour were noted: it conducted itself in a thoroughly masculine manner towards females; its whole bearing and instincts were definitely, almost precociously, masculine. It was useless to examine the development of such characters as horns and beards, since its mother was a horned and bearded nanny; nor was the kid old enough to have developed the characteristic smell of the billy goat. When it was killed the genitalia were removed by the butcher, and they provide the bulk of the histological material for this paper. After these specimens had been examined it was determined to trace the twin kid, which was procured alive, and its genitalia were excised and examined immediately after it was killed, and this kid provided the material from which the naked-eye anatomy is described. In every way this second kid had advanced in a masculine direction further than the one killed earlier. It must therefore be understood that in this description the histological characters are taken from the first-killed kid, and the gross anatomy from the older kid. As a matter of fact, the two specimens are identical save that in the second and older example the development of masculine external characters had proceeded rather further. The actual condition of the membranes at birth cannot be given, since no material was preserved and no scientific witness was present. : The genital tubercle of the older specimen was well developed (see fig. 1). A distinct glans, 15 em. in length, uncovered by a prepuce, was slightly dependent from the abdominal wall. A depression in the site of the urethral orifice of the male marked the ventral surface of the glans, but this depression did not have any connection with the urethra, the glans being imperforate. The body of the genital tubercle is marked by a well- developed median raphé, which terminates posteriorly in the anterior margin of the anus. The area covered by this median raphé was 3 cm. long, and in its length, and situated immediately behind the glans, a slit-like orifice marked a local failure of the inner genital folds to complete their fusion. The aperture in ee ee ee ee ee On Abnormal Sexual Characters in Twin Goats 267 the older kid was 5 mm. in length, and it was bordered upon either side by the free edges of the inner genital folds. Immediately lateral to this raphé region the body of the genital tubercle was distinguished as raised areas of skin constituting the body of penis hy, Eo nai oe genital apertute. oe Raphé formed by closed labia minora Anus . mas AZ) ZA \ Fic. 1.—External genitalia. Taken from the second kid. the penis. These lateral portions of the body of the penis were continued backwards around the anal orifice to the base of the tail, where they met and merged with the general-skin surface. These folds are homologised as the outer genital] folds. The space enclosed by these folds is homologised as the area of the ectodermal cloaca (see fig. 1). This appearance of the external genitalia of the older kid is in marked contrast with the condition present at birth. The figure (fig. 1) and the 268 Esther Rickards and Frederic Wood Jones description obviously denote the condition of a hypospadiac male with a somewhat undeveloped penis. At birth, however, the genital tubercle was no larger than the clitoris of a normal nanny kid, and the orifice in the raphé was by far the most conspicuous feature, and presented the appear- ance of a small—but otherwise normal—vulva. The nipples, two in number, and inguinal in position, were of the Vas deferens Cornu uteri - Bladder Urerus Ureter. Seminal Vesicle. Merof Postale . ‘ sinvs, Posterior end of Corpus cavernosum Reetum (thriwy back) Fic, 2. —Internal view of genital organs, seen from behind, the sacrum being removed, Taken from the second kid. normal size for a nanny kid; and the most striking feature about them was the fact that they depended from prominences which presented the typical appearance of the young udder. Dissection showed that the penis consisted of two corpora cavernosa, which presented the S-shaped bend characteristic of the normal male goat. -The posterior ends of the corpora cavernosa passed as fibrous masses upon either side of the urogenital sinus and rectum, and became lost in fibrous tissue behind the rectum. The orifice on the raphé passed directly into a large urogenital sinus On Abnormal Sexual Characters in Twin Goats 269 which, encircled by the divided posterior ends of the corpora cavernosa, was subdivided into an anterior urethral part and a posterior genital part. The former passes into a normal bladder and the latter into a typical uterus.. The urogenital sinus is 3 ems. long, and sections show it to be thin- walled, lined by epithelium of the transitional type and surrounded by a certain amount of erectile tissue—presumably the corpus spongiosum. A pair of lateral swellings in the wall of the sinus marks the site of the urogenital junction (fig. 2). Sections were made of this region, and they are found to be composed of prostatic tissue. Their ducts open into _ the urogenital sinus. Another pair of lateral swellings can be seen on either side of the uterus, above the prostatic bodies. These are partly embedded in the wall of the uterus, being free above and attached below. They have a knobbled appearance, and are situated rather to the anterior aspect of the organ (fig. 2). Before the microscopic structure of these parts is considered, the gross anatomy. of the genital apparatus will be described. The uterus is typical of a female goat of that age; the Miillerian ducts, or cornua uteri, pass laterally from each upper angle to the anterior abdominal wall, and thence pass out through the internal abdominal ring (fig. 3). Running alongside the Miillerian duct is a second thinner duct which passes proximally through the internal abdominal ring with the Miillerian duct and distally into the substance of the uterus, where it will be sub- sequently traced. This second duct is the Wolffian duct. The Miillerian and Wolffian ducts, traced through the internal abdominal ring, are found to traverse the inguinal canal and pass to the testicle, which lies in a pouch under the skin—the scrotum. This pouch is, moreover, under the site of the mammary gland, and this produces the elevation and apparent large size of the gland which seemed so like that of a normal female. The testicle consists of a testis and epididymis: the former, oval in shape and pearly white in colour, measured 1:3 em. in length and 1 cm. in breadth; the latter, almost crescentic in form and pink in colour, lies in relation to the external border and superior and internal poles of the testis, to which it is attached. Sections through the testis show that the gland is perfectly typical of the male sex, and, moreover, is absolutely normal in character. The seminiferous tubules are still solid, as one would expect in a young goat. The Wolffian duct is found to fuse with the upper part or globus 270 Esther Rickards and Frederic Wood Jones major of the epididymis, and the Miillerian duct, passing behind the testis, fuses with the lower part or globus minor of the epididymis (fig. 4). At the point where the Miillerian duct joins the epididymis the gubernaculum is attached (fig. 5). No structures resembling the stalked or sessile hydatids were observed. Reclum Genital Artery Hypogostric Artery Wolffian duct ; Uterine cornua Uterus Tnt. Abdominal Ring Bladder —— Urachus . Fic. 8.—Internal view of genital organs, bladder thrown forwards. Taken from second kid, To return to the caudal end of the Wolffian duct: its course in the anterior wall of the uterus was followed by means of serial sections of that organ. i In the upper sections the duct is seen embedded in the anterior part of the side wall of the uterus; as it is traced distally it seems to come into relation, anteriorly and externally, with the upper part of the lateral swelling in the uterine wall already referred to. This body is partly free and partly embedded in the wall of the uterus. It presents a coiled Eribieian 2. ; eee On Abnormal Sexual Characters in Twin Goats 271 ‘appearance i in section, but probably i is @ single cavity, thrown into a number of folds, which in section gives the impression of numerous cavities. Lower sections show this body—which is undoubtedly the seminal vesicle—wholly Wolffian duct Wolffian ducl Mollerian duct Mullerian duct Testis Clobus major of Epididy mis Clobus mayor Testis Cubernaculum y) Globus minor of Epididy mis Fie. 4.—Anterior view of testicle. Fic. 5.—Posterior view of testicle. Taken from the first kid. embedded in the uterine wall and still in relation- with the Wolffian duct, while yet lower the duct opens into the cavity of the seminal vesicle. In these sections the cavity of the urethra is seen anteriorly, while Oblique ly eu epithelium accor wet 0) 8) by fy « me cet a oy \ S56 hi ee u SS aft? EEE ae Ignis ‘ Fic. 6. Epithelium of uterus. Epithelium of Wolffian duct. posteriorly, and in much closer relation to the ducts and seminal] vesicles, is the lumen of the uterus (fig. 7 i.). The minute structure of these various parts shows very definitely the three groups to which they belong. The uterus is lined by thick transi- tional epithelium (fig. 6 A); the seminal vesicles and the lower end of the Wolffian ducts are lined by a single layer of elongated, almost columnar 272 Esther Rickards and Frederic Wood Jones cells (fig. 6 B); the urethra is lined by a thin layer of transitional epithelium. The tissue between the coils or folds of the seminal vesicles differs only from the epithelium lining the cavity in that it is seen obliquely; around the coils, smooth muscle fibres are seen (fig. 6 B). In the lower sections just described, another structure embedded in the M= Maolleriay Duet. D = Diverticulum. W = Wolffian Duct, - (Seminal vesicle) U = Urethra. Fic, 7.— Diagrams of serial sections of the uterus and urethra, with the Wolffian ducts and seminal vesicles, showing the means by which the cavities communicate, anterior wall of the uterus, in the median plane, is seen to arise from one or two diverticula, As one traces the sections more distally one sees that the seminal — vesicles fuse, forming one cavity, which is anterior to the cavity of the diverticulum, which in turn is anterior to the cavity of the uterus (fig. 7 ii). Thus in this region the structures are, from before backwards: the urethra, the fused seminal vesicles, the diverticulum, and the uterus. On Abnormal Sexual Characters in Twin Goats 273 Below, the diverticulum is seen to open into the uterus, so that only three cavities are seen in sections through this region (fig. 7 i.). Then the fused seminal vesicles are seen to open into the uterus, so that in these sections only two separated cavities are seen (fig. 7 iv.). Finally, the urethra opens into the anterior aspect of the fused seminal vesicles and so by | U = Uterus. D= Diverticulum. W = Wolffian duct. & = Seminal vesicle. Ur > Urethra. : U.S + Uro-genital Simus Fig. 8.—Diagram showing the lower ends of the Miillerian and Wolffian ducts, and the junction of uterus and urethra. - means of them and the diverticulum communicates with the uterus (fig. 7 v.). This is the site of the urogenital junction, and embedded in the walls of the cavity on either side is the upper end of the prostate gland. Fig. 8 may be supposed to represent the structures embedded in the anterior wall of the uterus, and to show the relations of the uterus and urethra, and the means by which the cavities communicate. SUMMARY AND CONCLUSIONS. This case is published since, so far as we are aware, it is the only one of its kind recorded. Mr C.J. Davis, who has devoted especial attention to caprine free-martins, concludes that they may be born “(a) singly ; (b) as twin with a normal male; (¢) as twin with a normal female; (d) as one of triplets, the normal kids being male and female; (e) as one of triplets, the normal kids being both males.” But headds: “ Breeders are unanimous in asserting that in their experience two malformed kids have never been born at the same time.” There would be but little merit in recording this case simply for the reason that so far no similar instance had been recorded ; VOL. LII. (THIRD SER. VOL, XIII.)—APRIL 1918. 19 274 Esther Rickards and Frederic Wood Jones but, as a matter of fact, far more importance than mere unusualness is attached to it. The underlying causation of the anomaly has often been probed, and, so far, two main ideas have been brought forward to account for the inception of the condition. | It has been supposed by Berry Hart that the abnormally sexed indi- vidual is the product of a single ovum which also produces the normal foetus. This theory, therefore, starts with the assumption that the normal and abnormal twins are monozygotic. On the other hand, the more recent work of Lillie assumes that the normal and abnormal foetuses originate from two distinct ova—that they are dizygotic in origin. Unfortunately, — we are unable to make any definite assertions concerning the membranes or circulatory mechanisms of the two kids in the present case; but, taking into consideration the exact likeness existing between them, it is almost impossible to disbelieve in their origin from a single ovum: every feature of the whole anatomy of the two kids compels a belief in their monozygotic origin. If the account of this case cannot furnish definite proof that the twin kids are monozygotic in origin, it at least lends no support to the sup- position that they are dizygotic. Further, this case disproves in toto Lillie’s further contention that the abnormally sexed individual is produced by the action of the sexual hormones developed by the other twin. Lillie supposes that while in the uterus a normal male derived from one ovum passes, by vascular connection, its sexual hormone with the circulation of a female derived from another ovum, and that in this way the sexual characters of the female twin become deranged. It is obvious that in this case of a identical twins no such hormonic influence can be involved. But apart altogether from this case, the whole series of facts concerning caprine free- martins brought together by Davis, and the whole experience of goat- breeders, show the unsoundness of Lillie’s theories. Were the abnormally sexed individual to be a female whose sexual characters were disturbed im utero by male hormones, one would suppose that after birth, and when freed from the influences of these hormones, the kid would tend to assert its true sex more and more, and to become more thoroughly female in character, That exactly the reverse is true is a matter of common knowledge. Caprine free-martins have won first prizes as she-kids and as she- goatlings, so thoroughly female are they in appearance and behaviour when young; but as they grow older their true sex begins to show itself very definitely. They are reported as standing “stretched out with a concave back” in a typical male position, as “rearing up and butting like a male,” as “ worrying the female goats with attentions like a male would,” as developing, “and very strongly, the characteristic smell of the male goat,” ete., ete, These characters develop with advancing age. ° o On Abnormal Sexual Characters in Twin Goats 275 Again, the histological condition of the genital gland in this case leaves no doubt that the organ is a typical testis, and so confirms the findings of Berry Hart, Spiegelberg, and others. In every way the case bears a strong resemblance to the caprine free-martin described by Keith, in which also the gonads were determined to be male. We are dealing, obviously, with male animals, the external genitalia of which are incompletely masculine at birth, and in which also the usual rudiments of the female internal genitalia are altogether unduly developed. The male gonad appears to exert its determining influence at an unusually late period, so that a disharmony results from an abnormally prolonged “indifferent ” stage of sexual differentiation. All effort to trace a precise pedigree for the parents of these kids has failed. This is to be regretted, since it was hoped that more light might be thrown upon the very interesting question of the origin of the extra- ordinary frequency of this malformation in goats. From the evidence collected by Davis and others it appears certain that the malformation is associated with the Toggenburg breed, and that, in all probability, it is transmitted through the normal males. REFERENCES IN LITERATURE. _ (1) D. Berry Hart, “ Numan, the Veterinarian and Comparative Anatomist of Utrecht: a forgotten Observer of the Free-martin,” Hdinburgh Medical Journal, June 1912. . (2) C. J. Davis, ‘Caprine Free-martins,” 7'he Veterinary Journal, February 1913, p. 62. (3) Horatio Hackerr Newman, The Biology of Twins, Chicago, 1917. STUDY OF AN OPODYMOUS KITTEN. By J. A. Pires ve Lima, M.D., Professor of Topographical Anatomy in the Faculty of Medicine of Oporto (Portugal). . In March last a servant of my laboratory brought me the dead body of a malformed kitten. He told me that the mother was three years old, and had already had four broods, each of four kittens. All were perfect except the one described here, which was dead-born. The mother is white all over, Fic. 1, except the head and tail, which are black, while the anomalous kitten is white on the ventral surface and dark grey on the dorsal one. It is of the male sex, and weighs 95 grammes. The stump which remains of its umbilical cord measures 15 em, Save the irregular conformation of the head, the anterior part of which is double, the rest of this animal has a normal external appearance, The head, single at the back, gradually divides into two, as one advances forward. The monster has two ears Study of an Opodymous Kitten 277 (fig. 1); three eyes, the medial one being constituted by two coalescent ocular globes with two cornes, united along a vertical line; two noses, each with two nostrils, the internal ones only being permeable; two mouths, each with .its own tongue. On the medial line of the anterior surface of this bifid head is a crest of hairs (fig. 1 and fig. 2, 1) which divides it into two perfectly symmetrical parts. The hairs are obliquely implanted outwards on both sides. The two heads are remarkably symmetrical, the whole being deviated to the left. There is but one medial palpebral cleft, but it is seen to be formed of two clefts that have become united at their internal angles. The medial Fie, 2. ocular globe is lined at the corners with broad and thick half-moon-shaped folds, which are united on the medial line both at the top and at the bottom (fig. 3, .s.). The conjunctiva is single, as well as the sclera; but there are two optic nerves, which penetrate the central eye at 5 mm. iseaune from each other. Each of the cornew presents a convexity of its own. The eccentric eyes have wide membranous plicas. I have taken the following measurements relating to the body of this animal :— Distance between the internal corners of the two eccentric eyes. 4 cm. Breadth of the palpebral cleft of the medial eye ‘ . 17 mm. Distance between the corner of the medial eye and the internal corner of the lateral eye. + agen From the internal angles of the lateral eyes to the tip of the nose . ae 4 em; From the angle of the medial eye to nasal tip . : A ay pe Between the ‘free tips of the two noses. ‘ ; . 35 mm. 278 Professor J. A. Pires de Lima Distance between the lines of implantation of the ears, measured over the vault of the skull : * 3 cm. From the medial point between the two noses and the free extremity of the tail . : : ; E : . 20 Length of the tail . : ‘ : : ooy ae a Sa Circumference of the neck : tae Circumference of the abdomen, taken at the umbilicus 9 The definition of an opodymous monster according to Isidore Geoffroy Saint-Hilaire is: “Body single, head single behind but separated into two distinct faces from the ocular region.” It is a diprosopus triophthalmus, according to Forster’s nomenclature. I began the dissection at the neck. The first layer being temoved, I found, enclosed under the concentric mandibular bone, a big submaxillary” gland. Next to this was a muscular stratum formed by the undifferentiated suprahyoid muscles. I continued the dissection forward, and uncovered the fore parts of the mandibular bones; the two medial ones are united at their hinder extremities, so that there are three symphyses between the four bony parts, as represented in fig. 2. I went on dissecting backwards, pre- paring the hyoid bone, the larynx, and the trachea, which, with the ceso- phagus, are single (fig. 2). I opened the thorax and abdomen, finding nothing anomalous in the respective viscera. I made an incision along the lower margins of the mandibular bones, cutting the two tongues, which were united at their bases. I entered a roomy single pharnygeal cavity, which communicated on one side with the mouths, on the other with the narrow cesophageal canal. The epiglottis was small in comparison with the wide pharynx. The single base of the combined tongues is of course very wide, and is proportionable to the extensive pharynx. On each of the palatine vaults there was a deep furrow, antero-posterior in direction, representing the lateral nasal fossa of each nose, the half which is not open to the outside, as we have already seen. As a compensa- tion, each of the furrows communicates with the respective medial nasal fossa by means of a short transverse canal. The two medial nasal fosse open to the outside, but not to the pharynx. ‘These canals have an oblique direction backwards and inwards, ending in culs-de-sac. If the monster were viable, the respiration would be carried out by the air entering through the internal nostrils, the only permeable ones; thence it would pass into the mouth through the transverse canal of communication between the internal nasal fossee and the palatine furrow, which represents the lateral nasal fossa. ‘The above-mentioned furrows were occupied by the internal halves of the corresponding tongues. In order to study the external configuration of the skull of the specimen cs 4 ee er Study of an Opodymous Kitten 279 I made a transverse incision in the soft parts from one ear to the other, and a longitudinal one from the free tips of the noses to the condyles of the occipital bone. Then with a periosteal elevator I exposed the surfaces of the skull and face bones (see figs. 3, 4). Fig. 3 exhibits the norma facialis of the skull, and fig. 4 its norma verticalis. The symmetry of the skull was remarkable (fig, 4, A, B). In some bones of the face (maxillary, fig. 3 (Mc, Me), nasal (N), inter- maxillary (I), and in the condylar portion of the occipital (O)), and the temporal bones, the process of ossification was imperfect. In the constitution of the border of the medial orbit the following bones took part: frontal and medial maxillary (fig. 3, F c, M ¢), as well A Fic. 3. as the internal malar bones (J, J), which are reduced to two small half- moon-shaped pieces, with their concavities turned upwards. The medial orbit is 14 mm. wide and 1 em. high. The four frontal bones present salient eminences; the internal ones are 138 mm. and the external ones 16 mm. wide. Above and behind the frontal bones we see the parietal, three in number; a medial (figs. 3 and 4, Pe), perfectly regular and symmetrical, 22 mm. in its greatest width and 20 mm. in length, with a longitudinal groove in its axis, but without any vestige of suture. Of the two lateral parietals, the left is constituted by two independent pieces (Pe, Pe’). It was apparently developed from two points of ossification, the independence of the two pieces being due to the position of the heads | of the monster, which is, as we have said, strongly bent to the left. Between the lateral parietal, the medial, and the occipital bones (tig. 4, O) _ there is a supraoccipital (S) shaped like a regular trapezium, the greatest width of which is 8 mm. and the least 3 nm. 280 Professor J. A. Pires de Lima After having studied the skull, I separated the parietal and the supra- | occipital bones in a single piece, a somewhat difficult operation, because the ~ internal surface of the bones was closely adherent to the meninges, at the sutures. Beneath the medial parietal the dura mater was very thick, with a large falx attached to it. I removed the encephalon, which on its upper surface is composed (fig. 5) of two concentric hemispheres (H ¢) corresponding to the medial parietal bone, and of two medial hemispheres (H e) united to the lateral ones. Each of the medial hemispheres presents two sulci, antero-posterior in direction, bounding three convolutions. The lateral present a few sulci, irregular and slightly marked. Behind the four hemispheres a single cerebellum .is seen. Fig. 6 represents the inferior aspect of the encephalon. Between the lateral hemispheres, where one notices a sulcus antero-posterior in direction, the mesial hemispheres are enclosed like a wedge. Each nose was supplied with a pair of well-developed olfactory bulbs. Although this monster is not a rare one, the specimens that have been dissected are not numerous. As we have already said, it was Isidore Geoffroy Saint-Hilaire (1) who gave this genus of monsters the designation of opodymus, rejecting the term polyopse which his father had coined for them. The creator of scientific teratology duly noted the remarkable sym- metry of both faces of the opodymi. He had observed that their tongues — were united at their bases; they had but one cerebellum ; and he mentions several cases in the human species and in other mammifers, specially in the ee ee ee a Te cat, as well as three birds and one fish. He refers to an opodymous — child of seven months which was publicly exhibited in Spain in 1775. — ‘Taruffi (2) quotes several cases gathered from medical literature, relating to our species, the cat, the ox, the sheep, etc. Gadeau de Kerville (8) _ summarily described an opodymous chickling, Guinard (4), together with Study of an Opodymous Kitten 281 Professor Lesbre, studied an opodymous kitten which lived five days. It had two concentric palpebral clefts, but only one medial orbit, which lodged an ocular globe, constituted by two coalescent ones, but with two optic nerves. _ Lesbre and Forgeot (5) published in 1906 a work dealing with this group of monsters. These authors had an opportunity of studying five opodymous monsters—cats, a calf, and a lamb, one of the cats being the one mentioned in Guinard’s book. From a study of these five cases, Lesbre and Forgeot gave a description of the opodymous genus. In the cases of this genus there may be two inedial parietal bones or a single one, as in my specimen. Finally, Lesbre and Pécherot (6), in 1913, studied the head of an opodymous calf which lived six days, having been put to death because it was unable to suck. The internal nasal fossz of this calf ended in a cul- de-sac, and the external nasal fosse communicated with the mouth and pharynx. It had only three cerebral hemispheres, and the medial sub- maxillary glands. were small—separate. It looked diophthalmic from without, but it really possessed a medial orbitless eye. BIBLIOGRAPHY. , (1) IstporE Grorrroy Sarnt-HiLairE, Histoire naturelle et particuliére des anomalies, Paris, 1832-1837. (2) Tarurri, Storia della teratologia, Bologna, 1881-1894. (3) GapEau DE KERVILLE, “Description de quatre monstres doubles,” Jowrnal de _ PAnatomie et de la Physiologie, 21™° année, 1885, p. 304. (4) Guinarp, Précis de tératologie, Paris, 1893. (5) Lespre et Foreror, “Contribution a |’étude anatomique des monstres hypsiloides et des monstres xioides,” Journal de U Anatomie et de la Physiologie, 42Qme année, 1906, p. 357. (6) LESBRE et P&cnErRot, cdem, 49™° année, 1913, p. 555. THE EXAMINATION OF - SKELETON OF KNOWN AGE, RACE, AND SEX. By Miss KaTHieen F. Lanper, B.Se., The London School of Medicine for Women. PROFESSOR COLLIMORE, whose skeleton is described in this paper, was an Englishman, who was born on Ist December 1878, and died on 17th Fic, 1.—Portrait and Skull. March 1907, aged 28. He held the post of Professor of Literature in the University of Fribourg, Switzerland. On the maternal side he was partly Jewish. The sternum of the skeleton is a substituted one, the original having been destroyed in the post-mortem examination; no observations on this bone could therefore be recorded. This fact possibly deprives the thoracic measurements of some of their value. His photograph, side by side with one of his skull, orientated to the The Examination of a Skeleton of known Age, Race, and Sex 283 same position, is shown on fig. 1. The contrast between the narrow appearance of the latter and the broad fleshy face is striking. It seems improbable that anyone examining the skull would postulate a type of face similar to that seen in the photograph. The measurements were taken with a Flower’s craniometer and Parsons’ rigid bar. CRANIAL MEASUREMENTS. mm. Length (maximum length, inclusive of glabella) ~ . ; 4100 Breadth( ,, breadth). . . 140 Minimum frontal breadth (actual minimum) . Soe xf taken on temporal lines . ; . 99°5 Height (basion tobregma) . . 136 Auricular height. ; ‘ . 125 Total facial height (pogonion to nasion) . ; ; ; . 124 Upper * (prosthion to ,, ). ; ; Sige Cranial base (basion to nasion) , , . 106 Facial ,, ( ,, to prosthion) F ; . 192 Bizygomatic (maximum) . : : . 135 Bigonal (maximum) , ; ‘ ; : ; . 120 Right orbit, length : ; ; . 36 » 9 breadth ; ‘ : : ‘ A a ae Left ,, length : : : 2 . 3 dhe ssi ty breadth ; x ; ‘ ; . 740 Measured inside the orbital margins aod edgidied the lachry- mal fossa :— Nasal length (nasion to nasal spine) ; ; ey} » breadth of aperture : ; ' : ; . A. 3B » length of a ; ; : 3 : : Say Measured inside the nasal margins :— Palate, length (maximum) ; - : Sea 2 J », breadth, neeene M? . : ; ‘ : ; a 22 Biangular- . : ‘ } ; ; . 102 Bicondylar . ; : ‘ ; . 122 Symphysial height ; : : ; ‘ . a, Sigmoid height 48 Dental length (combined length of crowns of premolars and molars) 45 Circumference of cranium, inclusive of glabella ; ; . 523 Cranial capacity (rape seed) . k ‘ . 1520 c.c. Me —- Fig. 2.—Temporal bone, showing bony canal for ninth nerve. no means as simple as the text-books would lead one to suppose. From its superficial origin the nerve passes antero-laterally, lying beneath that portion of the choroid plexus of the fourth ventricle which appears beneath the foramen of Luschka. It is also covered by the flocculus. It lies upon the lateral portion of the occipital bone known as the jugular eminence, or eminentia innominata, upon which, as is well known, it usually produces a groove. Travelling directly towards the aqueductus cochleariformis, it is directed to the base of the pyramidal depression on * _ the posterior surface of the petrous portion of the temporal bone. At the bottom of this depression the nerve bends at an acute angle, and 334 Relations of Glossopharyngeal Nerve at Exit from Cranial Cavity descends against the posterior surface of the petrous bone in a deep groove leading downwards from the pyramidal depression; and it is while in this groove that the nerve is separated from the vagus and the jugular vein by the inferior petrosal sinus, which is running downwards, backwards, and lateralwards, to empty itself into the jugular vein. While in the same groove, the nerve is hemmed in by a little bridge of fibrous tissue stretching across posterior to it; this little bridge of tissue is in about 25 per cent. of cases osseous, making a bony canal opening above just mesial to the atueductus cochleariformis, and below at the inner end of the plate of bone which lies between the jugular fossa and the carotid canal. This bony canal for the exit of the ninth nerve is, in some degree of development, a very noteworthy feature of the base of the skull; but, so far as I am aware, it has escaped description, except for a brief mention in the tenth edition of Quain’s Text-book of Anatomy, the later edition and all other text-books completely ignoring the point. : Fig. 2 shows the canal as it frequently appears upon the temporal bone. MULTIPLE ANOMALIES IN A HUMAN HEART. By HeErnNANI Bastos MonrTeErRo, Assistant in the Anatomical Institute of the Faculty of Medicine of the University of Oporto. THE heart which I propose to describe presents a series of anomalies. The foramen ovale is open; there is an interventricular foramen; a malforma- Fie. 1, tion of the pulmonary valves is present; there are also remnants of the valves between the sinus venosus and right auricle. . The specimen—a heart of an adult—was one of a number received for dissection. Its clinical history is unknown. As seen in fig. 1, the arteria pulmonalis has only two sigmoid valves : 336 Hernani Bastos Monteiro one of them, the larger, is situated before and somewhat to the right; and the second, of smaller dimensions, is to be found behind and a little towards the left. After having spread out the arteria pulmonalis, which measured 7-1 cm. in circumference, I found that the anterior sigmoid valve was 4 em. long on its free border, and its greatest height—the distance between the free border and the adherent one—being 1:8 em. Examination shows that this sigmoid valve is formed by the fusion of the anterior and the right posterior (fig. 2). Thus the free border presents, as the illustration indi- cates, four lunule instead of two, and two Morgagni’s ‘nodules. This border is thickened at the point corresponding to the union of the two sigmoid valves. From the adherent border and the parietal aspect of this. valve arise two fibrous cords, which reunite and then continue in a sort of Fie. 2. raphé along the wall of the arteria pulmonalis, thus separating two sinuses _ of Valsalva, which are very distinct in this specimen. This arrangement is illustrated in fig. 2. i The part of the anomalous valve which corresponds to the anterior sigmoid valve is larger than the one corresponding to the right posterior. — Thus, of the free border only 1:7 em. belongs to the latter, whereas 2°3 em, 4 belongs to the former. The aforesaid greatest height of 1:8 em. corresponds — to the medial part of the anterior sigmoid valve; the posterior measures but 15 em. in height at its medial part. The second valve corresponds to the left posterior sigmoid valve. Its lunule are rather indistinct. It — measures 3'1 em. along its free border; its greatest height is 18em. The — height is greatest at the medial part of the valve. The aortic valves — are normal, g In the right ventricular cavity there is a tendinous cord, which arises * from the “innominate fasciculus” and becomes attached to the free border ef of the anterior valve of the tricuspid (fig. 1). ; The interventricular orifice is situated in the pars membranacea (fig. 3). ¢ Multiple Anomalies in a Human Heart 337 This orifice measures 8 mm. in height. It is continued backwards by a sort of furrow, which is bounded on the inferior part by the pars muscu- losa of the septum. The orifice on the right side is concealed by the internal valve of the tricuspid. In this heart both the fossa and annulus ovalis are very distinct. The orifice in the left auricle is seen in fig. 4; it has the form of a vertical cleft, measuring, on the side of the left auricle, 1-4 em. in height. On this side the membrane of the fossa ovalis has the form of a crescent, with its concavity turned forwards. The membrane forms a complete valve to the orifice, and thus, during life, prevents passage of blood from the left to the right auricle. The reticular formation of the right auricular cavity is shown in fig. 5. It spreads like a net from one wall of the auricle to the other. It becomes attached on the right to the crista terminalis. On the septal side the reticulum is attached to the orifice of the inferior vena cava and the sinus of the great coronary vein, corresponding at these points to the Eustachian and Thebesian valves. This anomalous formation, which extends downward and from the right to the left, divides the right auricle into two portions : the one supero-internal, into which the venz cavee and the coronaria flow ; the other infero-external, which communicates with the right ventricle through the auriculo-ventricular orifice. We have to deal here with the persistence in an adult of an embryonic disposition of parts. The reticulum represents vestiges of the separation between the sinus venosus and the primitive auricle. The reticulum is formed by three different portions. At the medial part there springs VOL. LII. (THIRD SER. VOL. XIII.)—APRIL 1918, 23 338 Hernani Bastos Monteiro from the external to the internal wall of the auricle a sort of musculo- tendinous ribbon. This ribbon, which is 4°3 cm. long, measures 0°6 cm. in width at its medial part.. From its anterior border arises a tendinous net, and from the posterior another, the latter being more developed than the former. These nets are inserted on the crista terminalis. On account of their irregularity it is impossible to give a minute description of them One can form a better idea of it by examining fig. 5. Professor Pires de Lima (2) has described a pulmonary artery haviadl but two sigmoid valves, as well as an aorta with only two valves. Fia. 4, There are many records of observations of numerical anomalies of the aortic and pulmonary sigmoid valves (3 to 11); however, according to q Martinotti’s and Sperino’s opinions (12), “non sono di comune frequenza,” therefore such cases as this one “sono sempre interessanti per lo studioso,” M. Lucien and A. Harter (13) say that the anomalies of deficiency have — especially been noticed at the aortic orifice, and very exceptionally in the — arteria pulmonalis. 4s In the Museum of Pathological Anatomy of the University of Coimbra _ (14) four specimens with persistence of the interauricular foramen are to be found; and Henrique Parreira (15) has described a specimen “ which, however, would not have permitted the mixture of the blood.” 4 According to published observations (16 to 25), the interauricular com- _ ——————— Multiple Anomalies in a Human Heart 339 munication may be made through a single orifice or through several, and may be found on a level with the fossa ovalis or, which is much rarer, quite independent of it. The communication may be owing to absence of the interauricular septum (26). The communication between the two ventricles may be due to the want of the interventricular septum or to the presence of a single open orifice either in the pars membranacea or in the pars musculosa (27 to 30). Generally, other anomalies of the heart or of the great vessels are associated with it. Henrique Parreira and Geraldino Brites (31) have described two cases - Fie. 5. of perforation of the interventricular septum. In one of them (a child one year old) there was a cleft in the muscular tissue of the septum; in the other (a dead subject, 1:1 metre high) there was a communication in Alvarenga’s space. Professor Thiago d’Almeida (32) has published clinical observations on cases of perforation of the interventricular septum diagnosed by him. Among the books I could consult, I only met with references to reticular formations in the right auricular cavity in two numbers of the Journal of Anatomy and Physiology and in the Bibliographie anatomique. W. Turner (33) has described a specimen in which there was a fenestrated valve at the mouth of the vena cava superior, and another in the vena cava inferior, united by a fibrous cord. And he adds: “Two other human 340 Hernani Bastos Monteiro hearts also are referred to in which a fibrous cord extended from the Eustachian valve, along the right posterior wall of the auricle, almost as far as the mouth of the superior cava.” Is this fibrous cord His’s crista terminalis? At that date this formation had not yet been described, His’s works being later (1885). In Turner’s opinion the reticular formation would represent the right segment of the large double muscular valve found at the opening of the sinus venosus in the heart of the bird. Robert B. Thomson (34) has described two large, cribriform Eustachian valves found in adult hearts. These formations were well developed, especially in one of the specimens, in which they might be described as Fic. 6. “like a spider’s web stretched out so as to occupy a position in rather more than the lower half of the right auricle.” These webs were attached in such a manner that “their attachment corresponded accurately to a large part of the line of demarcation between the right part of the sinus venosus which is absorbed into the auricle and that part of the auricle which arises from the auricle proper.” In each of the specimens described by R. Thomson there was a foramen ovale. A. Weber (35) describes two reticular formations he met with in the heart of a woman 70 years old. One of them was extensive, whereas the other was very small, The former extended from the superior and posterior part of the auricle as far as the inferior and anterior part of the interauricular septum, its insertions corresponding to the crista terminalis. The latter Multiple Anomalies in a Human Heart 341 reticular formation was a little net situated on the posterior border of the fossa ovalis and which became lost in Vicussens’ annulus. In this heart the membrane of the fossa ovalis had several perforations, which established the communication between the two auricles. This author says: “ Les formations réticulées d’aprés Chiari ne seraient pas trés rares; malgré cela, trés peu d’auteurs les ont signalées” ; and he quotes Rokitansky’s, Przewosky’s, and Chiari’s observations. Weber thinks this formation may be derived either from the venous valve of the sinus reuniens alone, or, as in the case he describes, from these valves and also from the septa of the sinus. Sometimes in the right auricle reticular formations are to be found, of slight development, such as the small net shown in fig. 6, which represents a specimen in the Museum of the Anatomical Institute of the University of Porto. It is probably to a formation like this that Greenfield (25) refers; but the engraving which accompanies the article of this author being rather indistinct, one cannot obtain an exact idea. BIBLIOGRAPHY. _ (1) HernAnt Bastos Monrerro, Catdlogo do Museu de Anatomia normal, coordenado sob a direccéo do Prof. J. A. Pires de Lima, Oporto, 1917. (2) Pires pe Lima, “ Notas de Anatomia: I. Sobre anomalias numericas das valvulas sigmoideas,” Gazeta dos Hospitues do Oporto, n. 5, 1911. . (3) ArtHur Keitu, The Lancet, 1909, vol. ii. pp. 359, 433, and 519. . (4) B. Taompson Lowne and Dr Maxweru T. Masrsrs, Descriptive Catalogue of the Teratological Series in the Museum of the Royal College of Surgeons of England, London, 1893. (5) W. S. Greenrietp, “A Case of Malformation of the Heart,” Jowrnal of Anatomy and Physiology, 1890, vol. xxiv. p. 423. (6) Kerra Camppett, “An Accessory Segment in the Pulmonary Valve,” Journ. of Anat. and Phys., 1896, vol. xxx. p. 347. (7) Francis O. Simpson, “Congenital Abnormalities of the Heart in. the Insane,” Journ. of Anat. and Phys., 1898, vol. xxxii. p. 679. (8) H. K. Assorr, “Case of Abnormal Arrangement of Aortic Valves,” Journ. of Anat, and Phys., 1904, vol. xxxviii. p. 103. (9) A. H. Youna, ‘Rare Anomaly of the Human Heart,” Journ. of Anat. and Phys., 1907, vol. xli. p. 190. (10) D. G. Rem, A Case of Multiple Heart Anomalies, 1912, vol. xlvi. p. 86. (11) Idem, Journ. of Anat. and Phys., 1914, vol. xlviii. p. 174. (12) Martinortr and SpErino, Sulle anomalie numeriche delle semilunari aortiche e pulmonari, Turino, 1884. (13) M. Lucien and A. Harrer, “Deux anomalies des valvules sigmoides de Vartere pulmonaire,” Bibliographie anatomique, 1908, vol. xvii. (14) J. Marques pos Santos and Axperta Pxssoa, Catdlogo do Museu de Anatomia pathologica da Universidade de Coimbra, 1915. 342 Multiple Anomalies in a Human Heart (15) Henrique Parrerra, “ Situs viscerum inversus completus,” Bulletin de la Société portugaise des Sciences naturelles, Lisbonne, 1916. (16) Davrp Hepsurn, “ Double Superior Vena Cava, Right Pulmonary Veins opening into the Right Auricle, and a se gate Interauricular Foramen,” Journ. of — Anat. and Phys., 1887, vol. xxi. p. 438, (17) ‘Seventh Report of the Conimittee of Collective Investigation of the Anatomical Society of Great Britain and Ireland for the year 1896-1897,” reported by F. G. Parsons and Arthur Keith, Journ. of Anat. and Phys., 1898, vol. xxxii. 164. 2 (18) AmprosE BrrMincHam, “ Extreme Anomaly of the Heart and cut Vessels,” Journ. of Anat. and Phys., 1893, vol. xxvii. p. 139. (19) Aprian Stokes, “ Abnormal Position of the Heart and -Great Blood- Vessels, associated with pie ate, of the Viscera,” Journ. of Anat. and Phys., 1909, vol. xliii. p. 301. (20) R. J. Guapstone and C. H. Rerssmann, “Two Examples of Cardiac Malformation,” Journ. of Anat. and Phys., 1916, vol. 1. p. 228. (21) T. Warpror Grirrira, “Example of a large Opening between the two, Auricles of the Heart, unconnected with the Fossa ovalis,” Jowrn. of Anat. and Phys., 1899, vol. xxxiii. p. 261. (22) Journ. of Anat. and Phys., 1899, vol. xxxiii. p. 502. (23) E. Fawcerr and J. V. BLACHFORD, “The Frequency of an Opening between the Right and Left Auricles at the Seat of the Foetal Foramen ovale,” Journ. of Anat. and Phys., 1901, vol. xxxv. p. 67. (24) T. W. P. Lawrence, ‘A ‘Case of Congenital Malformation of the Heart, with Abnormalities of Abdominal Viscera,” Jowrn. of Anat. and Phys., 1902, vol. xxxvi. p. 63. (25) W. S. GREENFIELD, “A Case of Malformation of the Heart, with large Deficiency in the Interauricular Septum, Patomy of the Foramen ovale, and —- Stenosis of the Aortic Orifice,” Journ. of Anat. and Phys., 1890, vol. xxiv. p. 423. (26) R. J. Prosyy Wittams, “ Unusual Malformation of the Heart,” Journ. = of Anat. and Phys., 1894, vol. xxviii. p. 305. (27) Gorpon SanpErs, “A Case of Congenital Malformation of the Heart, with Transposition of the Aorta and Pulmonary Artery,” Journ, of Anat, and Phys., 1893, vol. xxvii. p. 464. (28) Davinson Buack, “Two Cases of Cardiac Malformation, more especially of the Infundibular Region,” Journ, of Anat, and Phys., 1914, vol. xlviii. p. 274. (29) W. E. Carngcte Dicxson-and Joun Fraser, “A Congenital Abnormality of the Heart and Blood-Vessels,” Jowrn. of Anat. and Phys., 1914, vol. xlviii. p. 210. (30) A. W. Meyer, “Spolia anatomica,” Jowrn, of Anat. and Phys., 1914, vol. xlviii. p. 107, (31) Henrique Parrema and Gerratpino Brires, “Deux cas d’anomalies cardiaques avec cyanose congénitale,” Bulletin de la Société portugaise des Sciences naturelles, Lisbonne, 1916. “ 32) Tutaco p’Atmetpa, “A doenga de Roger,” Medicina Contemporanea, 1914. 33) Journ. of Anat, and Phys., 1869, vol. iii. p. 452. (34) Ropert B, THomson, “Note on Two Large and Cribriform Eustachian Valves,” Journ. of Anat, and Phys., 1911, vol. xlv. p. 443. (35) A. Weber, “ Formations réticulés de Voreillette droite et fosse ovale 4 EJ anormale d’un cour humain adulte,” Bibliographie anatomique, 1898, vi. p. 17. MUSCULAR ACTION. By W. Cotin Mackenzie, M.D., F.R.C.S.Ed., Melbourne. In army medical circles the view is held that, of the wounded men return- ing from the French battlefields in the Great War, some 65 per cent. are of an orthopoedic nature. Whether the injury be one of muscle, bone, joint, nerve, or central nervous system, the lesion is of a nature in which the question of muscular function becomes of prime importance for purposes of treatment. The war has demonstrated that much more attention should be paid to the teaching of myology than heretofore. Of one thing we can be certain, the teaching of muscle function—the all-important factor— cannot be satisfactorily carried out in the dissecting-room alone. In the dissecting-room the greatest attention should be paid by the student to the insertion of muscles. Without an accurate knowledge of muscle insertion there cannot be an accurate knowledge of muscle function. A muscle fibre has the double physiological property of contraction and relaxation, 7.¢. the property of increasing or diminishing the tension within itself ; but the fibre has not the property of voluntary elongation. A relaxed muscle depends on the contraction of its opponent for its elongation. Thus each muscle is a reciprocal elongator of its opponent, and when we use the term “muscular action” we are not speaking of the single action of muscle contraction but of a double action, viz. contraction and shortening of the muscle producing the desired effect on the centre of motion, and relaxation and elongation of its opponent. Just as important as a knowledge of the action of a muscle itself is a knowledge of the action of its opponent. In text-books, opponent muscles should be specifically stated. The statement that fifteen external rotators of the hip, including such powerful engines as the psoas and great gluteus, are balanced by one muscle and part of another can be no longer tolerated. The student, when studying the action of a muscle, should study the action of the opponent as well, because of the importance of the opponent in the treatment of lesions of nerves, muscles, and joints, Thus, in cases of weakness or paralysis of the deltoid he should associate the opposing ad- ductor, viz. the pectoralis major, with the deltoid. He should recognise that there can be no recovery if the opposing pectoral is allowed to contract, and should immediately guard against its occurrence. 344 Muscular Action A muscle cannot be in a state of relaxation and contraction at one and the same time. Contraction and relaxation are two opposite physiological states. A muscle cannot be at one time an extender and a flexor as is described in the case of the interossei and lumbricales of the hand. If we flex the two interphalangeal joints, the interossei (the extenders) are relaxed to allow of the contraction of their opponents and reciprocal elongators, viz. the sublimis and profundus. Yet, holding these joints flexed, 7.c. with the interossei out of action as contractors, we can flex readily at the meta- carpo-phalangeal joint. This is perfurmed by the lumbricales owing to the relaxation of the physiological antagonist, viz. the extensor communis. The lumbricales flex at the metacarpo-phalangeal articulation, the interossel extend at the interphalangeal articulations. A muscle is rested when its opponent is in a state of relaxation and elongation beyond the state normally regarded as necessary to produce a condition of equilibrium with its opponent. Thus, if we regard the state of equilibrium between the pronators and supinators of the forearm as that of mid position, then the forearm must be oversupinated to rest the supinators and overpronated to rest the pronators. Too much stress cannot be laid on the right method of giving rest to muscle. It is essential to the recovery of nerve injury or nerve disease. In a median nerve injury the position of rest would be flexion of the elbow—overpronation of the fore- arm—flexion of the wrist, the thumb flexed across the palm, and flexion of the fingers. The latter is most convenient although the inner lumbricales and profundus tendons are supplied by the ulnar nerve. In a case of weakened deltoid, whether from direct injury, injury of nerve, or polio- myelitis, such a condition as a hanging upper extremity should never be seen. ‘The arm should be immediately supported in an abduction splint. al JOURNAL OF ANATOMY THE SUBLINGUA AND THE PLICA FIMBRIATA. By FREpERIC Woop Jones, Professor of Anatomy in the University, London, School of Medicine for Women. THAT the curious structure known as the sublingua in the Lemurs should be the expression of any very primitive condition is an assumption which, though commonly made by comparative anatomists, is not likely to impress anyone who is familiar with the habits of Lemurs as being at all probable. If one were asked to point to a peculiarly highly-specialised structure, one that had been definitely elaborated in response to a functional need, it would be difficult to select a feature more satisfying than the lemurine sublingua. ‘This curious structure is one that cannot rightly be understood by studying it merely as a fold upon the under surface of the tongue; it must be correlated with other details of the anatomy of the mouth and must be observed in the living animal. The incisor teeth of the typical ’ Lemurs are peculiar; the lower ones being strikingly elongated and set at such an angle in the jaw that their “biting” edges face directly forwards instead of being directed upwards towards the upper incisors. Moreover, it is evident that these peculiar lower incisors have some function which is not shared by the upper incisors, for the incisor series of the upper jaw is composed of somewhat rudimentary peg-shaped teeth which are evidently not exercising their full functional réle. They are not engaging the lower incisors in any action of biting, and in adult skulls they are very commonly in an extremely reduced condition. More than this, the lower incisors are reinforced by the lower canines which, assuming the same elongated form, become procumbent with the incisors, and are at times mistaken for members of the incisor series. Following upon this change in the lower canines, the first lower pre- molars have assumed the rédle and form, but not the position, of true eanines, and have added to the confusion produced in the minds of some who have dealt with this question (see fig. 1). This curious arrangement of the lower incisors and canines is a very striking one, and there can be no question as to its purpose, since these teeth play little or no part in the VOL. LII. (THIRD SER. VOL. XIII.)—JULY 1918. 24 346 Professor Frederic: Wood Jones | process of alimentation, but are specialised purely as a hair comb. So far as I know, this function was first detected by Cuvier, who, writing of Lemur catta, said that the lower incisor teeth “sont de véritables peignes ” (Hist. Nat. Mammif., 1829, p. 218). Hair combs and feather preeners— have been evolved de novo time and again, and from a series of very different structures. There is no need to turn aside to the curious serrated claws of such birds as Fregeta or Sula; for a hair comb, which is also derived from the front_teeth, is present so much nearer to hand _ in Galeopithecus. The dental hair combs of the true Lemurs and the so-called Flying Lemurs afford striking examples of parallel evolutionsin response to functional demands, for though the end attained is the same Fic. 1.—Anterior teeth of the lower jaw of Lemur catta, seen from below. in the two cases, the method of making the comb is strangely different. — In Lemur the incisors and the canines are ranged parallel and procumbent — like the “teeth” of a comb; in Galeopithecus the biting edges of the — individual incisor teeth have ‘become finely serrated, and each tooth thera fore furnishes many “teeth” for the comb; this method may be said to — constitute an economy of incisors as hair combs, and the canines therefail » do not join the incisor series, but on the other hand become modified in — the direction of the molar series. - These two animal types have developed these dental hair combs probably. 4 for the reason that special adaptations have rendered it difficult to perform the toilet of the woolly hair by scratching—the Lemurs having no free claws — upon the fingers and only one specialised one'on the toes, and the limb of & Galeopithecus being hampered by the flying membrane. In the case of the | 4 i mee EE The Sublingua and the Plica Fimbriata 347 Lemur the application of the hair comb is an oft-repeated business, and one that is very easy to observe in all its details. With the development of this very highly-specialised dental structure there has arisen a need for specialised toothbrushes, and again it is of interest to note that the two forms have evolved entirely different forms of organs for this purpose. In Galeopithecus the anterior edge of the tongue is finely serrated in harmony with the serrated incisors, a very efficient toothbrush being provided for the hair comb. Lemur, on the other hand, has developed some structure beneath its tongue into a horny leaf-like organ which is finely serrated along its edges and sharp-pointed at its tip (see fig. 2). The use of this structure as a toothbrush has, I believe, been Fie 2.—Tongue of Lemur catta, seen from below, to show the plica mediana and plice fimbriate. long familiar to Mr R. I. Pocock, and there can be no doubt whatever that the “sublingua” of Lemur is a purely functional organ, evolved from some part of the tongue for the purpose of cleaning the functionally specialised lower incisor teeth. Such facts as these show the lemurine sublingua in rather a different light from that by which it is regarded as an inherited and functionless rudiment of some premammalian ancestral structure. But even though it is in Lemur a functional, highly-developed, and specialised structure, it is of necessity made from some basis which other mammals must possess in some degree. This basis is a curious one, and one that is so variable in its manifestations that its most primitive expression in the mammals must remain somewhat uncertain. Nevertheless, one particular basal type appears to embrace all the 348 Professor Frederic Wood Jones modifications so far met with, and that wal be described provisionally as the ideal form. This form finds expression in certain Marsupials, a few Eutherian ~ adults, and several Eutherian embryos. In this type, the tongue bears no papille whatever upon its under surface. There may or may not be a median strengthening elevation extending from base to tip; but upon either side of the middle line, starting near the middle line at the free tip and diverging as they proceed back- wards, are folds with projecting edges which are covered by thickened horny epithelium. The free edges of these folds may be variously serrated. At the attached base of the tongue they turn forwards and downwards towards the floor of the mouth as two crenated folds that terminate Fic. 3.—Tongue of Petawrus breviceps. The continuity of the plica fimbriata with the plica sublingualis is well seen. i near the middle line of the floor of the mouth not far behind the symphysis menti. At the point where the fold upon the side of the tongue curves forwards to join the fold on the floor of the mouth a common fold runs backwards — along the attached margin of the tongue and here becomes gradually lost (see fig. 3). In many eases it would probably be more correct to describe this last fold as running along the side of the posterior portion of the tongue and floor of the mouth, subsequently bifureating into one fold — which runs along the side of the tongue, and another which runs along — the floor of the mouth. Any or all of these folds may be finely serrated, — more coarsely dentated, merely crenated, or practically plain. Since the — lemurine toothbrush is an elaboration of the fold upon the free portion — of the tongue, it is natural to inquire if the whole basal structure may not~ have some such function, and that this is the true explanation of the — presence of these folds appears to be extremely probable. Such a view is strengthened by the fact that their development is certainly not an — isolated thing, for upon the outer side of the dental series an exactly similar serrated fold is, in many animals, produced from the lower lips — The Sublingua and the Plica Fimbriata 349 ~ and cheek folds. This curious fold is well seen in the dog, and, so far as I know, no purpose has ever been ascribed to it; but I believe its true role is a cleanser of the teeth, and that the same service is performed within the mouth by the folds which lie below the tongue. It must be remembered that, although in Homo the tongue may explore and cleanse the whole dental series, the relation of the tongue, teeth, and floor of the mouth forbids a lingual exploration of the molar series in many animals. In the new-born Palearctic wolf figured (fig. 4) the condition is very remarkable, for here the labial folds outside the dental series are associated with serrations of the side of the tongue within the dental series. This is another lingual toothbrush, which, as in Galeopithecus, is not formed from Fie. 4.—New-born cub of Palearctic wolf with cheek removed, to show the lip fold and the dentated side of the tongue. a fold beneath the tongue but by the papillated surface of the tongue itself. Although I gather from Professor Wright that the use of the tongue as an organ for cleansing the teeth had impressed Leonardo da Vinci and had prompted him to some speculations upon this very point, it must be confessed that this aspect of the tongue as a functional structure has subsequently become rather overlooked. In order to follow the fate of the specialised folds which lie below the tongue it is necessary to adhere to a precise nomenclature, and reference _ will be made again to a tongue such as that of Petawrus breviceps, which appears to show all the folds in fairly full development (see fig. 3). The median strengthening rod is known as the plica mediana, and it is an expression of a sublingual development of the septum of the tongue. It is an extremely variable structure. The midline of the under surface of the tongue may be elevated, flattened, or depressed, for the 350 Professor Frederic Wood Jones thickening of the septum as a supporting structure may take place within —- the substance of the tongue or below its inferior surface. In Petawrus it is a well-marked free ridge upon the under surface of the tongue. It is this structure which finds one form of expression in the “lyssa” or Fic. 5.—Tongue of Lemur catta, The plica fimbriata forms the - : well-developed sublingua. The plica sublingualis has prac- tically ceased to exist. “tollwurm ” of the dog’s tongue. In Lemur it constitutes the median ridge of the “sublingua.” The fold which runs from the tip of the tongue towards its attached base is the plica fimbriata, and the fold which runs from the posterior portion of the tongue to the floor of the mouth beneath the tongue is Fic. 6.—Tongue of Tarsius spectrum. Plica mediana, plica fimbriata and plica sublingualis are all well developed, the plica sublingualis. This fold at its anterior limit in the floor ‘of 4 ; the mouth marks the site of the openings of the submaxillary salivary gland, and may be reduced to a structure no. more prominent than the — so-called Wharton’s papilla or caruncula. a The present paper is concerned simply with the comparative develop- _ ment of these folds within the limits of the order Primates, as that order is at present constituted. 4 In the Lemurs the characteristic features are determined by a great a The Sublingua and the Plica Fimbriata 351 development of the plice fimbriate, which constitute a highly-specialised organ adherent to the tongue over the great extent of its length and supported in the middle line by the plica mediana, ‘his structure composes the “sublingua” of the Lemurs. The tip of the sublingua is Fic. 7,—Tongue of an adult Chimpanzee. Plica fimbriata and plica sublingualis both well developed. free of the under surface of the tongue for a short distance. In all the Lemurs the plica sublingualis is an extremely rudimentary structure, entirely wanting upon the side of the tongue, and represented merely by minute caruncule on the floor of the mouth. Fie. 8.—Tongue of a human feetus, Plica fimbriata and plica sublingualis are both well developed. ; The development of the plica mediana and a conspicuous “sublingua” derived from the plice fimbriatee upon a somewhat lemurine model is seen in the Didelphia in Trichoswrus vulpecula, but in this type the plica timbriata is by no means so highly specialised as it is in Lemur. The tongue of Yarsiws stands in very marked contrast to anything seen in the Lemurs, for here all those parts which we have pictured as being components of the generalised condition are fully developed (see fig. 6). The plica mediana is strongly marked. The plice fimbriate occupy a relatively large area of the under side of the tongue, and they 352 Professor Frederic Wood Jones are adherent to it in the whole of their length. The plica sublingualis is conspicuous, and shows a well-marked crenated edge. Histologically these folds show a well-marked thickening of the surface corneous layers such as is present in the almost horny “sublingua” of the Lemurs. Fic, 9.—Tongue of Macacus nemestrinus. ‘The plica sublingualis is well developed. ‘The plica fimbriata has disappeared, In the study of the interrelations of the different members of the order Primates the tongue of Yarsius is of the utmost importance, since no other members of the order save Man and the Anthropoids show this 3 q Fic. 10,—Tongue of a Marmoset (Callithria penicillata). A, from the side ; q B, from below, The plica fimbriata is well developed ; the plica sub- f. lingualis, though small, is large for a New- World monkey, E generalised condition of these folds. In the first place, it must be noted j that the tongue of Tarsius differs entirely from that of any of the Lemurs % in the possession of the plice sublinguales in a condition of full develop- - ment, that it is of the generalised form, not widely different from that of ‘f Petuwrus, a condition from-which presumably the Lemurs have specialised 7 2 The Sublingua and the Plica Fimbriata 353 just as T'richoswrus has departed in a somewhat lesser degree. In the train of Tarsius the tongues of the Chimpanzee (fig. 7) and of Homo (fig. 8) follow very closely, for in both types the two folds are well developed. But among the Primates they stand alone in this generalised condition, for all the monkeys of the Old World and the New are specialised in one direction or the other from this generalised form. In Fic. 11.—Tongue of Cebus albifrons from below, showing presi soles ey plice fim- briate and minute plice sublinguales. the Old-World monkeys the plica fimbriata is absent and the side of the tongue is entirely smooth, but the plica sublingualis is fairly well developed (see fig. 9). In the New-World monkeys the plica fimbriata is usually well developed and the plica sublingualis though present is not conspicuous (see fig. 11). In the Marmosets it attains a fair degree of development, but in other genera it is reduced to the merest rudiment (see fig. 10). In none of the New-World monkeys, however, is the plica fimbriata developed into such a highly-specialised structure as the “sublingua” of the true Lemurs. A HITHERTO UNDESCRIBED MALFORMATION OF THE HEART. 3 By H. Buakxeway, MS., F.R.C.S., Surgical Registrar, formerly Ss Demonstrator of Anatomy, St Bartholomew’s Hospital. (From the Anatomy Department, St Bartholomew's Hospital.) THE specimen described below was sent to me in 1912 by my friend Dr 3 J. A. Bell, M.C., Captain R.A.M.C. (T.), who was at that time Medical Officer = at Camberwell Workhouse; I am greatly indebted to him for giving me the opportunity of examining and describing it. ¥ The heart (which is now in the Museum of the Royal College of Surgeons of England) is that of a male infant, born in the workhouse; the child, which weighed 7 lbs., survived only thirty-eight hours. Other clinical sl details are lacking. a ‘3 Unfortunately the specimen suffered considerably in the process of removal from the body—the more unfortunately, because it shows a Be condition which, so far as I am aware, has not up to the present been ee described, namely, that the aorta, though of normal size, has no communica- __ tion with either ventricle, but was filled from near the apex of the left _ ventricle by way of the anterior interventricular branch of the left coronary = artery. if In fig. 1 the heart has been drawn so as to show most of its abnor- — malities. There is a very obvious disparity in the size of the cavities of 4 2 the right and left sides of the heart: this affects especially the auricles, — the right auricle being large, while the left is remarkably smaller than _ normal, 3 The right awricle shows the normal division of its interior. into two. : 4 parts by a prominent crista terminalis, behind which is a large opening of the inferior vena cava; that of the superior vena cava appears to have been, in part at least, cut away; no Eustachian valve is present. At its a left lower extremity the right auricle receives the wide opening (unguarded by any valve) of the left horn of the sinus venosus, which becomes con- tinuous, round the left side of the left auricle, with a persistent left superior vena cava; a great cardiac vein is present in the left auriculo-ventricular furrow, and also opens into the sinus. Enough of the interauricular septum remains to show that the foramen ovale was widely open, s — ee A Hitherto Undescribed Malformation of the Heart 355 The right auriculo-ventricular orifice and the tricuspid valve (tig. 2) show no gross abnormalities, though the cusps of the valve are unusually large. The opening measures 15 mm. in its longest diameter. 2 6 Fic, 1.—General view of the specimen. 1, right auricle; 2, interauricular septum, with patent foramen ovale; 3, opening of inferior vena cava; 4, left horn of sinus venosus and left superior vena cava, from which arrow emerges; 5, cavity of left auricle; a rod passes through the mitral orifice; 6, the abnormal communication between the left ventricle and the anterior interventricular coronary artery; a bent arrow passes along the latter into the aorta ; 7, the point of origin of the right coronary artery from the aorta; 8, com- mencement of the pulmonary artery, in two pieces; 9, left side of interventricular septum ; above is seen the oblique ridge which divides the cavity of the left ventricle into two parts. ‘ The right ventricle shows considerable enlargement, which is confined to the body of the ventricle, the infundibulum, although it does not appear to be smaller than normal, being disproportionately small as compared with the rest, The cavity is seen to extend behind the left ventricle as far as 356 Mr H. Blakeway — the left margin of the heart, the septum between the right and left cavities being very obliquely placed. Only the very commencement of the pulmonary artery is preserved, and that in pieces, but it is clear that the orifice was approximately of normal size; and it has a normal valve with three cusps. The left awricle (fig. 1) is many times smaller than the right; no pulmonary veins are seen opening into it;.and although they may have Fic. 2.—The right ventricle has been widely opened, 1, the two parts of the pulmonary artery ; 2, infundibulum of ventricle ; 3, tricuspid valve; 4, right auricle; 5, ascending aorta; 6, right side of interventricular septum ; the septal cusp of the tricuspid valve is applied to its upper part. Note the cavity of the ventricle, in shadow, carried far to the left behind it, been cut away, yet the incised wall of the auricle can be brought together in such a way as so nearly to reconstruct the chamber that it is difficult to see how the veins could have reached it. There is a small auricular appendix, with miniature musculi pectinati in its interior. * From the auricle a smooth, rounded mitral orifice, 4 mm. in diameter, leads into the left ventricle; this latter is of approximately the normal size, and has a well-developed muscular wall; but in almost all other respects it is a highly abnormal chamber. Its upper sixth part is partially separated from the rest by a prominent oblique muscular ridge (see figs. 1 and 3), A Hitherto Undescribed Malformation of the Heart 357 The two parts are very different in appearance and connexions; column carnee, though small, are prominent features of the ridge and of the upper subdivision of the cavity: certain of them are developed into musculi papillares, which have attached to them the chord tendinex belonging to the two tiny cusps of the mitral valve. But in the lower subdivision muscular columns are absent: the endocardium .is smooth, becoming thicker and opaque towards the apex, where also it is somewhat puckered. eS ‘ 4 c. \ . Rd Y M'' i M\\ \ | Wf, Yj t : \ i re \ hy y Yb Ke f,' SN ~~ ® i\ \\\\ \ SN : Fic. 3, —Enlarged view of part of the left ventricle, opened widely to show the upper subdivision of the cavity, bounded below by a thick muscular ridge. 1, left auricle, turned bs 2, left.superior vena cava; 3, anterior cusp of Y mitral valve; 4, the id angle between the anterior cusp of the mitral valve and the uppermost part of the interventricular septum. The interventricular septum is entire, and of good thickness throughout its extent; the part of it that is related to the upper subdivision of the left ventricle has applied to its right side the septal cusp of the tricuspid valve; on the left side (see fig. 3), where it should lead up to the aortic orifice, no such opening is to be seen, the narrow space between the septum and the anterior cusp of the mitral valve ending in a blind angle above. The question as to how the ventricle emptied itself is solved by an examination of its lower portion; here, near the apex, the cavity bends forward and to the right, and narrows to a funnel-shaped opening, 2 mm. in diameter, which leads directly into an anterior interventricular channel on the surface of the heart; at the junction of this vessel with the cavity 358 Mr H. Blakeway of the ventricle the channel is slightly narrowed by an irregularly circular thickening of the endocardium (fig. 4). This remarkable vessel is the only communication between the left yentricle and the aorta; it passes upwards, accompanied by a microscopic cardiac vein, in the anterior interventricular groove, displaying in its interior the origins of several branches which pass obliquely upwards and laterally to the walls of the ventricles: it then opens in a funnel-shaped - fashion into the commencement of-a well-developed ascending aorta. This interventricular vessel is, beyond doubt, the anterior inter- Fic. 4.—Enlarged view of the communication between the apex of the left ventricle and the anterior inter- ventricular coronary artery. The point of communication is hooked open (at 1) to show the slight transverse constriction ; 2, cavity of left ventricle; 3, cavity of right ventricle. ventricular branch of the left coronary artery; this, not only from its position and connexion with the aorta, its supply of vessels to the heart wall, or its microscopic structure (see fig. 6), but still more from the fact that the main trunk of the left coronary is seen arising from within the funnel-shaped upper end of the vessel, and then running to the left in the auriculo-ventricular furrow (fig. 5). Examination of the beginning of the ascending aorta places the identity of the coronary vessel further beyond question, The aorta, as has been said, does not communicate with the base of the left ventricle; but its interior, in the situation where the communication should be, shows a shallow depression, divided into two lateral parts by a very low ridge: from the right-hand portion arises a normal right coronary artery. It is impossible to doubt that these shallow depressions represent two of the sinuses of Valsalva; and a less distinct indication of the third is seen where A Hitherto Undescribed Malformation of the Heart 359 the left coronary artery communicates with the aorta. The relation of these shallow depressions to one another and to the pulmonary artery will be recognised as being normal. When this specimen was demonstrated to the Anatomical Society information as to certain points which have since been made clear by further dissection was lacking, and considerable doubt was expressed as to the correctness of the interpretation of the specimen given above. It Fic. 5.—Enlarged view of the interior of the commencement of the aorta, 1, origin of right coronary artery from one of the rvdimentary sinuses of Valsalva; 2, anterior ventricular branch of left coronary artery ; within it is seen the opening of one of its branches to the right ventricle ; 3, cavity of left ventricle ; 4, left auricular appendix, turned aside; 5, left superior vena cava; 6, main trunk of left coronary artery; 7, part of pulmonary artery. was suggested that the vessel seen in the anterior interventricular groove might not be a coronary artery, but that it might rather represent a part of the bulbus cordis, the aperture near the apex of the left ventricle being the ostium bulbi. Such a suggestion was at the time a tempting one; it gave prospect of a reasonable explanation, based on embryological knowledge, for an otherwise inexplicable condition. Indeed, the striking difference between the two portions of the left ventricle suggested that possibly only the small upper subdivision was really the ventricle proper, 360 . Mr H. Blakeway the ostium bulbi being at the level of the muscular ridge marking the communication between the two portions of the cavity (figs. 1 and 3). Microscopic examination of the ventricular wall from each portion of the chamber does not lend support to this view. Sections were made at the two points marked with crosses on the cut surface of the ventricle in fig. 1; they show no essential difference in structure: near the apex the muscle is much thinner, and, as the naked-eye examination suggests, the endocardium is thick (fig. 7), and-beneath it bundles of cellular fibrous Fic. 6.—Transverse section through middle of anterior interventricular coronary artery and accompanying cardiac vein. x 19, tissue, with a structure similar to that of the surface endocardium, pass in places between the muscular fasciculi—an appearance which is not seen in the upper section of the ventricle. But the section of the wall of the left ventricle near its apex is very different from that of the interventricular channel conveying blood to the aorta ; the microscope confirms the conviction formed by dissection that this vessel is an artery, the interventricular branch of the left coronary (see fig. 6). ! While the connexion of this artery with the left ventricle and the total obliteration of the normal aortic orifice seem at present impossible of satis- factory explanation, certain conjectures may be hazarded. A Hitherto Undescribed Malformation of the Heart 361 The two conditions are no doubt interdependent. The aorta did origi- nally arise widely and directly from the ventricle ; all that we know of its development, the saccular depression at its commencement, and the rudi- mentary sinuses of Valsalva, all make this certain. The obliteration, then, was a late process, occurring after the truncus arteriosus had been separated Into its pulmonary and systemic parts. Was it the result of the abnormal _ communication between the coronary artery and the ventricle, or was the aortic orifice first obliterated, and the coronary communication opened up Fic. 7.—Section through wall of left ventricle in the position of the lower of the two crosses seen on the cut surface of the ventricle in fig. 1. x 35. as a result? If the obliteration of the aortic orifice was the primary condition, we may suppose a pathological inflammatory process to have been probably at work; but perhaps not necessarily so: in the young embryo, the facility with which parts fuse together which before were distinct is only matched by the ease with which solid tissues may become pervious. But our wonder is constantly excited, not only by the rarity with which complex developmental processes go awry, but nearly as much by the fact that when they do so the very malformations are generally subject to laws, and are readily recognised, though not explained, as arrests of development and their results. No arrest of development will explain the obliteration of the aortic VOL. LII. (THIRD SER. VOL. XIII.)—JULY 1918. 25 362 A Hitherto Undescribed Malformation of the Heart orifice seen in this heart, for from the earliest stage of its development the aorta and the ventricle are in free communication. The stenosis is an irregular abnormality ; but since the causes, whatever they may be, which lead to abnormal development do not act entirely capriciously, it would be unreasonable to suppose that this irregular aortic stenosis was in fact capricious. The possibility of the process being syphilitic has to be con- sidered. The mother cannot be traced; but the sections of the coronary artery and of the heart wall do not suggest a syphilitic lesion, and such an explanation would fail to cover the other anomalies seen in the specimen. Less unlikely, then, than an explanation depending on a merely capri- cious (because causeless) closure, or upon an inflammatory stenosis of the aorta, is the view that the aperture into the coronary artery from near the apex of the ventricle may have been the primary anomaly, the aortic closure resulting from this. While no details as to the development of the coronary arteries are yet known, it seems not impossible that during the absorption of the muscular sponge-work in the formation of the ventricular cavities such an abnormal communication between ventricle and coronary arteries might arise: the sharp difference between the upper and lower subdivisions of the ven- tricular chamber appears susceptible of some similar explanation, 7.e. that the smaller upper subdivision alone is the ventricle proper, matching the auricle in size, while the lower part is an abnormal cavity formed by the opening up of the muscular sponge-work belonging rather to the right ventricle than the left. The meaning of the obliquity of what has been called above the interventricular septum, and of the extent of the right ventricle behind it to the left border of the heart, would thus become apparent. If then, as in this heart, extreme under-development of the left auricle and ventricle already existed, and the blood-flow through the left side of the heart were consequently much diminished, even such an apparently small cause as the opening up of the ventricle into the coro- nary artery might produce so great a result as the closure of the orifice of the aorta. ee ae 7? ts ee ee es i SOME /INDICES AND MEASUREMENTS OF THE MODERN FEMUR. By J. R. D. Horry, M.D., Se.B., Chief Demonstrator of Anatomy, Trinity College, Dublin. THIS investigation was originally undertaken in order to ascertain the average degree of pilastring in modern femora, but gradually the work was extended to include other points of interest. The specimens here dealt with were taken from the bones ordinarily in use for teaching purposes in the Anatomical Department of Trinity College, Dublin. Bones with obvious irregularities due to disease were rejected, but with this exception no attempt was made at selection, the object being to obtain as representative a series of average conditions as possible. It is to be regretted that no opportunity has so far offered itself of obtaining femora definitely known to be in pairs, and herein lies, perhaps, the greatest objection to some of the conclusions which have been arrived at. An endeavour is now being made to get together a series in which great care has been taken to make the pairing and sexing absolutely definite and known. The measurements were all made and the various indices calculated before it was attempted to subdivide the collection into groups according to sex and side. This resulted in some irregularity in the numbers in the various sections, but is not a serious disadvantage. The bones were sexed by Parsons’ method, and as all the measurements, etc., are given in the various portions of Table I., it is comparatively easy to see how many possible mistakes have arisen. Only a small proportion presented any great difficulty as regards sexing, and these were carefully considered several times before being finally allocated to their special group. The measurements were all checked, in some cases several times, and may be taken as accurate. Parsons advocates consideration of the following points in endeavouring to ascertain the sex of a femur :— 1, Diameter of head— estimated with Sect as held parallel to long axis of neck - 449 + 48 ¢ mm. 2. Bicondylar width : - 709+ 75¢ ,, 3. Least transverse diameter of shaft —- 269+ 29¢ ,, 4. Head length index -— 1049 +109 ¢ 5. Platymeric index - 759 + 90¢ 6. Maximal length . - 409 + 45 g cm. 364 Dr J. R. D. Holtby : The collection comprised 56 male bones, 35 from the left side and 21 from the right; of the remainder, 20 were of the right side and 24 of the left from female subjects. Table I. in its four subdivisions gives the various measurements and indices of each of these groups, whilst Table II. shows the averages in each. It will now be convenient to consider the various points in order. LENGTH. Both oblique and maximum measurements were taken, The former is obtained by use of a measuring board, the two condyles lying flat on the foot plate, the caput against the upper plate. The maximum length is the greatest length between the head of the | femur and the under aspect of the medial condyle. According to Pearson, the maximum length exceeds the Saati oh by 32 mm. in males.and 3°3 mm. in females. In my series, account was not taken of fractions of a millimetre in measur- ing the length, so I cannot refer to the decimal. In males it was sufficient to add 3 mm. to the average oblique length to get the average maximum on — either side of the body; but among the female groups, whilst an addition of 3 mm. was required on the right side, 2 mm. was sufficient on the left. It — must be remembered that the addition of these figures to the oblique measurement will only give the maximum length in the average; individual examples vary widely—in some cases as much as 6 mm. will have to be added, and in others the two measurements may be identical. Occasionally, indeed, the oblique length may even exceed the maximal. There is no marked difference between the oblique and maximal lengths as regards sex, and a comparison of these measurements is thus of no utility as regards sex grouping. Maximum length (average), cm. Oblique length (average), cm. Male, oe Female, } re | Male, Female, | Right. | Left. | Right. | Left. | Right. | Left. | Right. | Left. ie iB | | 1 | 453 | 46°0 | 414 | 41°6 | 45°0 | 45°7 | 41‘1 41°4 | It will he seen that there is a difference in the length of the bones from the two sides, which is more marked in males than in females. Of course ee EES oo a ee rs Some Indices and Measurements of the Modern Femur 365 the bones in this collection were not in pairs, but Parsons, who noticed a similar condition in his Rothwell series, says that it is borne out by measure- ments of known modern pairs. In his series, however, the sexual element is not shown, the left bone in both sexes being 3 mm. longer than the right one. In my collection the left male bones were 7 mm. longer than the right ones, the left female specimens only 3 mm. longer than their fellows. THE RANGE OF VARIATION IN OBLIQUE AND MAXIMAL LENGTHS. | Maximum length, Oblique length. . Males. Right . : > ‘ 40°4-51°6 cm. 40°3-50°5 cm. Left , ‘ ; ‘ 41-52°3 ,, 41-52°1 ,, Females, Right . , ° ° 37°4-45°2 ,, 37°3-44'6 ,, Left . «| 862-464 ,, 35°8-46'1 ,, The curves on fig. 1 show the distribution of the variations in oblique length. It will be seen that a bone with an oblique length of less than 40 is almost certainly female, whilst one with a measurement of over 46 is almost invariably male. DIAMETER OF THE HEAD, On reference to Table II. it will be seen that the average diameter of the head was almost the same on the two sides in female specimens, but that among males there was a considerable difference between the two sides. In Parsons’ series the diameter of the head was not uniform in paired bones, but varied from 0°5 to 1 mm. The range of variation is shown as under :— Males, ” Females, | Right. Left, Right. | Left, Average . . -| 48mm, 49°6 mm. | 42 mm. 41°5 mm. | Range : . - 43-51 - 44-56 37 °5-45°5 37-46 The curves on fig. 2 show the distribution of the range. It may, of course, be objected that no deductions can be drawn from the figures in 366 Dr J. R. D. Holtby this connection, as the diameter of the head was one of the factors relied upon for determining the sex to which a given specimen belonged. It will, however, be seen from the curves that it has not been allowed to act unless other points bore out its evidence, for there is a considerable amount of FEMALES RIGHT | LEFT. a 36 38 40 42 94 46 48 50 52 54 MALES Pa RIGHT LEFT N Ftv, 1,—Curves to show variations in oblique length. overlapping between the two sexes, especially between the 44 and 46 mm. limits. Bones under 44 were almost always females, and those over 46 were practically invariably males. It will be seen also that the height of the curve on both sides in female specimens corresponds to those with a diameter of 40 to 42 mm.; among male bones, the height of the curve is reached at 46 on the right side and at 48 on the left, Ss - Some Indices and Measurements of the Modern Femur 367 Relationship between Diameter of Head and Bicondylar Width. The average relationship is shown below, the figures indicating that © the bicondylar width was the given multiple of the diameter of head, Males. Females. Right. Left. Right. | Left. 1°6 15 16 1°6 so that the large diameter of the femoral head on the left side was not accompanied by a corresponding increase in the bicondylar width. 38 40 42 44°46 48 50 - FEMALES RIGHT - LEFT 38 9 42 44 45 48 S50. 52 5S# S56 S58 60 MALES RIGHT LEFT Fie. 2.—Curves to show the distribution of variations in the diameter of the femoral head. 368 Dr J. R. D. Holtby The latter, as will be seen from Table IIL. is almost identical on the two sides. Roughly, the average relationship for bones of both sides and sex is that the bicondylar width is 1-6. times the diameter of the head. The range of variation among individual cases is comparatively small (from 1:5 to 18), but is quite sufficient to allow of mistakes being made in estimating one measurement from the other; usually the error would not exceed 3 mm. in calculating the diameter of the head from the bicondylar width, but it may be as much as 8 mm. if the estimation of the bicondylar width be based on a known diameter of the head. Thus it might be difficult to accurately sex a bone with one damaged ex- tremity unless the other indications were well defined. In Hepburn’s series the bicondylar width was in the average 1:'7 times the diameter _ of the head. Heap LENGTH INDEX. This was introduced by Parsons as a means of ascertaining the pro- portions of the diameter of the head to the length of the bone. If the head of the female be, as is stated, smaller than that of the male, not only absolutely but also relatively to the length of the femur, the index should show sexual differences. The maximal length of the bone is divided into the diameter of the head multiplied by 1000, and an index is thus obtained. The multiplication by the figures 1000 is only done in order to show the result as a whole number to allow of ready comparison. Males, Females, Right. Left. Right. Left, Averages . : ‘ 105°5 106°4 102°5 100°2 Range . . , | 96°6-118°2 | 96°3-121'7 | 98°6-107'4 | 87-1087 Thus there appears to be on both sides too much overlapping, as Parsons pointed out, to render the index useful for sexing except at the extremes; bones with indices of over 108 are usually males, whilst those with indices below 96 ave almost invariably females. Generally speaking, the head length index is smaller in long bones than in short ones, the latter having relatively large heads. See curves on fig. 3 (a). Some Indices and Measurements of the Modern Femur 369 Bicondylar Width. - Parsons recommends that this be measured with a sliding scale, the condyles lying flat on the long stem, the limbs of the scale touching 6 RIGHT SIDE ‘ ° . . * * . . . . . . . ° e . Bm a 93 I5 I7 99 at 103 105 a7 109 “ “3 “us “Pr “ge / 88 90 92 94 96 98 10a 102 104 106 /08 “0 “ue “eg “6 “8a 120 s . . . . . . . u, . . . . . . . . . ‘ . . . LEFT S/DE : : . . . . . . . * . . . * . . . . . . . . . . ag W 93 95 97 99 +0t 103 105 107 109 mt "3 ug “7 “gg v2 88 90 92 9F 96 98 400 102 104 106 +08 “a “2 “ge “sé “Ee 12a . e . . . e . . . . * . . . . . ’ . . * . + Fic. 3.—Chart to show the distribution of the variations in the head length index. the condyles close to their articular margins. As can be seen from Table II., the measurement on the average is practically the same on FEMALES MALES 399 5B 102 106 HO 4 94 398 102 106 110 14 U8 122 Longest Shortest. Fic. 3 (a).—Curves to show head length indices in longest and shortest bones. either side. There is an important sexual difference corresponding to that of the diameter of the head, as is well seen in the curves on fig. 4, 370 Dr J. R. D. Holtby The maximum point for female bones was at 70, irrespective of side; among the male specimens, however, the maximum distribution point is reached at 76 on the right side, whilst on the left it is between 76 and 80. Bones with a bicondylar width of under 72 are, almost certainly, female, those with a measurement of over 74 being usually male (Parsons stated 62 69 66 68 70 72 74 76 FEMALES ~ 62 64 G6 68 70 72 74 76 78 80 82 8&4 86 88 MALES Fic, 4.—Curves to show variations in bicondylar width. The upper curves represent specimens from the right sides, the lower from the left. that between 70 and 75 the sexual factor was indefinite). It will be noted that left-sided bones more especially seem to show specimens in or near the sex limit points. Relationship of Bicondylar Width to Oblique Length. Although averse to the multiplication of indices, I have found it convenient to use one to show this relationship. It is based on the one Some Indices and Measurements of the Modern Femur 371 introduced by Parsons as the head length index, and is obtained by dividing the oblique length in millimetres into the bicondylar width multiplied by 1000. If one selects the longest and shortest bones from any series, it will be seen that there is, in the great majority of cases, a striking variation in the index, short bones having relatively broad lower extremities. The following averages show this :— Average index 163 Longest bones (12). Shortest ,, (17). Average ,, (12). Females. Longest bones (8). . | Average index 158 174 | Shortest (9). : =f » 169) 171| Average ,, (12). . pe » 165) The figures as to the index in bones of average length are given to complete the comparison, but among them the range of variation is very great. The curves on fig. 5 show the variation for the longest and shortest specimens, and indicate that although there is a considerable (50 '§5 160 185 FEMALES MALES nS Longest Shortest. Longest Shortest LErT 165 170. 175 180 185. Fic. 5.—Curves to show variations in the bicondylar length index in longest and shortest bones. degree of overlapping, yet in the main the index is higher in short bones than in long ones. It will be seen that there is less overlapping in males than in females, and that the lower indices are not seen in short bones, and vice versa. 372 Dr J. R. D. Holtby LEAST TRANSVERSE DIAMETER OF SHAFT. This is usually a little below the middle of the bone; the measurement should not be taken near the extremities. In this series the averages and range were as under :— Males. Females. Right. | Left, Right. Left. Averages . 27:5 mm, 28 mm. 24°5 mm, 25 mm. , Range 24-32 24-32 23-27 23-27 It will be noticed that the left bones are somewhat thicker than the right ones. Parsons found a similar condition somewhat more marked in his Rothwell series. The measurements from male specimens. above quoted are a little lower than those given by Parsons as obtained from modern bones in St Thomas’ and Guy’s Hospital collections. The range shown in the above table is not so extreme as in the Rothwell series. PILASTRING. The condition of femur @ pilastre is best shown by means of an index, obtained by taking the greatest antero-posterior measurement (which may ~ or may not be at the level of the middle of the shaft); at the same level take the transverse diameter in a plane parallel to the posterior surfaces of the condyles. Antero-posterior measurement x 100 Transverse measurement = pilastric index. The averages in this series were— Males, Females, Right, Left. Right. Left, 110°5 107°2 109°7 107 The remarkable correspondence of the indices on the left side in the two sexes might not be borne out in another series, but in Hepburn’s small collection of modern Scotch femora a similar condition was found; he did not, however, draw attention to it. When should a bone be regarded as definitely pilastred? It will be found that the condition is fairly obvious when the index exceeds 114 on the left side and 116 on the right. The curves on fig, 6 show the distribution of the variation, and also the number of markedly pilastred am ea ial oo ee eee, ae ee, a f Some Indices and Measurements of the Modern Femur 373 bones in each group. Hepburn, not differentiating between side or sex, gives 109°3 as the average for his series; using a similar method, the average here would be 1083, As this is somewhat misleading, the question of grouping should always be considered. It can be seen from the curves that the left groups show a steady drop above the 114 limit, whilst the right ones develop a secondary rise above 90 92 96 100 10% 108 2 HA WE WB 120 124 128 132 136 FEMALES He RIGHT N+ ocean LEFT MALES » RIGHT | LEFT Fic. 6.—Curves to show the range of variation in the pilastric index. the 118 level, indicating that not only is the average index higher on the latter side, but also that marked degrees of pilastring are more frequently seen there. There is no definite selatiocialis between the degree of pilastring and the length of the bone, high and low degrees of the former being associated with both relatively long and short specimens. Causation of Pilastring. , Manouvrier supposed that the pilastric ridge was caused by the ‘vasti (lateralis and medialis), just as the cranial crests in the gorilla are formed by the temporal muscles. It may here be remarked that a specimen may show a fairly prominent linea aspera and yet not be pilastred; the linea aspera must be raised out of proportion to the width of the shaft in order to show a distinct degree - of pilastring. I will refer later to the association of pilastring and platymery, remark- ing only now that the two are probably not due to the same cause, as there is no definite correspondence in their distribution. 374 Dr J. R. D. Holtby Now, if pilastring be dependent on, as Manouvrier stated, the degree of development of the vasti, one would expect to find a fairly definite condition of the pilastric fosse with high and low degrees of the condition. In this series the nature of these fosse was carefully noted in each case, and it was found whilst the medial “fossa” was usually flat or slightly rounded, the lateral one varied much in its conformation. Most frequently it was concave, but it was as often so in bones with average or relatively low indices as in those with high indices, and in the latter cases it was frequently flat, or even somewhat convex. It seems likely that the degree of prominence of the ridge may be influenced not only by the muscles at its margins but also by those attached to its lips, and here it may be said that the lateral lip is usually more proniinent in well pilastred bones than it the medial one. PLATYMERIA. To estimate the degree of flattening at the upper end of the shaft it is necessary to take the lowest antero-posterior measurement, usually near the small trochanter, and the transverse width at the same level. Antero-posterior x 100 Transverse =platymeric index. Parsons has suggested that a bone should be regarded as distinctly platymeric if the index be 75 or under. In this series the averages were :— Males. Females. Right. Left. Right. Left. 85'2 82°5 | 81°6 80'8 There thus appears to be some difference in the degree to which the condition is marked on the two sides, the left bones being flatter than the right ones. The curves on fig. 7 show that “distinct” platymeria is more common among females than among males. It is indeed one of the features recommended by Parsons for consideration in endeavouring to sex a femur, yet it is evident from the curves that a bone had not been assigned to a definite sex group on account of platymery alone. This is also shown by the comparatively small difference in the average indices of the two sexes; relatively high indices were found in both. It will be noticed that a distinet degree of platymeria is more common on the left side than on the right, and that a larger number of left-sided specimens show also a moderate degree of flattening (index between 75 and 80). ae ? ‘eee Ee pe Oe A 5 St . - i BR Sint a Oe et (ne Some Indices and Measurements of the Modern Femur 375 The averages in this series for the male specimens are almost exactly identical with those of Hepburn’s modern Scotch femora. His female RIGHT LEFT 70 75 80 8 90 95 100 70 75 80 85 30 95 100 _ FEMALES MALES Fic. 7.—Curves to show the range of variation in the platymeric index. averages are higher than those here given, but as he only included four female bones, his figures are scarcely reliable. Neither Hepburn or Scott referred to the sexual difference, which was first indicated by Parsons. RELATIONSHIP BETWEEN PLATYMERIA AND PILASTRING. If platymery be solely due to the influence of the vasti (lateralis and medialis) one would expect to see some definite relationship between it and pilastring, as was pointed out by Manouvrier, who stated that there is a close relationship between the amount of flattening in an antero-posterior direction and the flattening in a transverse direction (platymeria and pilastring): “The highest degree of the one is not, however, necessarily associated with the highest degree of the other.” Hepburn agreed in the main with Manouvrier’s conclusions. In this connection one must clearly understand that a high degree of platymeria means a low index, whilst a high degree of pilastring is associated with and indicated by a high pilastric index. The results shown in my series are not at all in conformance with the above views. There appears to be no fixed relationship between the two conditions. In each group I have selected the bones with (a) the most marked degree of pilastring, (b) the highest degree of platymery. Females. Right Side.—Of seven specimens with distinct pilastring (116 and over), not one showed “distinct ” platymery—in the average specimens one bone 376 Dr J. R. D. Holtby in seven should display the latter. Of the six most platymeric bones, only two showed distinct pilastring—little more than the average distribution of the condition in the group. Left Side.—Of four bones with distinct pilastring, two showed distinct platymery ; the other two had indices between 75 and 80. Ofseven distinctly — flattened bones, only two had evident pilastring, three others having very low indices (below 100). Males. Right Side-—Seven specimens showed distinct pilastring—not one had a distinct degree of platymery. Of six bones with platymeric indices of 80 and under, none showed high pilastric indices. Left Side—Of six distinctly platymeric pilastred bahies, only two presented a platymeric index of below 80. Of twelve specimens with a platymeric index of 80 and under, only two had distinet pilastring; three others had pilastric indices of 100 and under. A similar want of correspondence between the two conditions is seen if ~ one takes sub-groups with low degrees of platymeria and pilastring. The sex factor has a marked influence on the distribution and extent of platymeria, but is in abeyance in the case of pilastring. This does not — support the idea of a homogeneous cause. THE AMOUNT OF BowINa. This was estimated by Parsons’ method. “'The femur is laid close to the edge of a flat table in such a way that the posterior surfaces of both condyles are in contact with the table. The sliding limb of the scale is then removed, and the scale placed vertically against the edge of the table in such a manner that the fixed limb touches the point of greatest convexity of the femur about the middle of the shaft. The distance of the most convex point from the surface of the table is then shown on the scale. By keeping the scale pressed against the table edge any deviation from the vertical is prevented.” As the mere amount of bowing would mean little unless taken in conjunction with the length of the femur, he advocates the use of an index, “obtained by dividing the oblique length into the amount of bowing multiplied by 100; in this way the greater the index the greater is the bowing compared with the length of the femur.” To bring this index into line with the head length and bicondylar length indices it is more convenient to multiply the amount of bowing by 1000 in place of 100, and so obtain a whole number. Si Se ee ee Od Eh gee ° a : Pr, ‘ i Lee Le eee ‘see , aS toe is Se aie ee es os tas a — 7 a ee Te es , . tae a Some Indices and Measurements of the Modern Femur 377 The amount of bowing itself showed a close parallelism on the two sides. en | Females. Males. Right. | Left. | Right. Left. | 56 mm. | 56 mm. | 60 mm. 58 mm. The sexual difference is small, and may be obliterated if the relative lengths be considered as in the index. INDEX OF BOWING—AVERAGES. } ) Females. Males, Right. Left. Right. Left. | } 138 134° 136 I need only quote two examples to show how misleading the mere amount of bowing would be if it were taken as indicating the degree of bowing, and how necessary it is that the length of the specimen should be con- sidered, as it is in the index of bowing. Two male bones, each with an index of 141, showed an amount of bowing of 72 mm. and 57 mm. respectively; two other bones with an amount of bowing of 61 mm. showed indices of 144 and 163 respectively. From fig. 8 it will be seen that, on the whole, short bones are more curved than long ones. Sex and side appear to have little influence on the degree of curvature, the slightly higher index in females shown above being due to the shorter length of these groups. There is no definite relationship between the degree of bowing, degree of pilastring, and degree of platymery, marked conditions of the latter two being found with either high or low indices of bowing. 3 The question sometimes arises as to the number of specimens necessary to enable one to form an opinion as to the average characteristics. In order to ascertain the size of the group necessary I divided these bones into divisions of 10, 20, and 50 as they had come to hand when first measured. Only the indices of platymery, pilastring, and bowing were considered, as measurements of head—length, width, etc.—depend so much on the relative proportions of the two sexes in each group. The results set out below VOL. LII. (THIRD SER. VOL. XIII.)—JULY 1918. 26 378 Dr J. R. D. Holtby indicate that no group of less than twenty can be used to draw conclusions from, and that a group of fifty would be more desirable.. The maximum Wa 120 125 130 135 190 150 160 170 1/80 FEMALES =a Longest RIGHT = =a. SP Shortest. ~OP- gh SP oan Longest. LEFT Shortest MALES = Longest. RIGHT = , AVA y, Z WIT) ‘ SINE St i _ Cav. f. (CVs). =b be, ITI. Cmd, — ‘ . Ss \ Was \ » "(a J Carc. — SS SSS . — Way Cav. Vy (Csv,). The Muscles Related to the Branchial Arches in Raia clavata 387 of fibres forming a single continuous muscle which is partly separated into two parts by a furrow on its external surface. The fibres that are inserted on the palato-quadrate I do not find described by either , Tiesing, Marion (1905), or Luther, but they form an important part of the muscle in all my specimens, and they, together with the musculus superioris II. and parts, described immediately below, of the fascia on levator labii the ventral surface of the adductor, quite certainly re- present the muscle Addy of Vetter’s descriptions of the Selachii; as comparison with Luther’s descriptions of this latter muscle in Heptanchus will make evident. The fascia that covers the ventral surface of this division of the adductor is a tough membrane which has posterior, anterior, and antero- lateral extensions. Posteriorly the fascia extends over the posterior edge of those fibres that are inserted on the mandibula, but penetrates the muscle between those fibres and the fibres that are inserted on the palato-quadrate. That part of the fascia that extends posteriorly over the posterior edge of the muscle in part vanishes on its posterior surface, but is in part continued as a fibrous band, of variable form, that has its insertion on the anterior surface of the muscle Cs, of Tiesing’s descriptions, near the middle of the dorso-ventral length of that muscle and not far from its proximal, deeper edge. The mesial edge of the fascia runs into the subdermal tissues at the angle of the gape of the mouth, and is there continuous with the fascia related to the labial cartilages. At about the posterior third of its mesial edge the fascia gives origin to a stout tendon which runs mesially across the external surfaces of the musculi adductor mandibule lateralis I. and depressor mandibularis, and then internal to the musculus coraco-mandibularis, there lying in the membrane that separates the latter muscle from the musculus coraco-hyoideus. In that membrane it continues mesially, and is continuous, in the median line, with its fellow of the opposite side of the head. Anteriorly, the fascia separates into two parts. The lateral portion turns dorso-anteriorly over the lateral edge of the muscle, crosses the lateral surface of that muscle and then that of the levator labii superioris II., and reaches the dorsal surface of the latter muscle. There it turns antero-mesially and is inserted on the dorso-posterior surface of the nasal capsule, this part of the fascia being the levator labii superioris V. of Tiesing’s fig. 13, pl. 7. The mesial portion of the fascia is slightly thickened along its lateral edge, and from that thickened portion, which lies along the antero-lateral edge of the ventral surface of the adductor, numerous tendinous strings, or a tendinous band, extend to the ventro-posterior edge of the lateral ethmoidal process (cartilago-parethmoidalis, Luther). These tendons are 388 Mr Edward Phelps Allis shown by Tiesing in one of his figures of this fish (fig. 11, pl. 6), and are there called by him the levator labii superioris V.; but the tendinous band shown in that figure is not the same as the one shown in his fig. 13, pl. 7, which, as just above explained, is also called the levator labii superioris V. _ Beyond the line of origin of these tendons the main fascia turns upward over the rounded anterior edge of the adductor mandibule lateralis IL, internal to the levator labii superioris IV., and then becomes in part a broad band which, running dorso-mesially, is inserted on the dorso- posterior surface of the nasal capsule. This band gives insertion, along the dorsal portion of its anterior surface, to certain fibres of the musculus levator labii superioris III., and it lies closely against a stout membrane | that here covers the posterior surface of the nasal capsule, this latter membrane giving insertion to the musculus levator labii superioris LV. The m. levator labii superioris I. is as described by Tiesing; and Luther (1909, p. 46) considers this muscle -alone, of the several levatores described by Tiesing, to represent the levator labii superioris of the Selachii (m. preorbitalis, Luther). The m. levator labii superioris II. is a thick muscle which arises on the mandibula by a large tendinous end which lies internal (dorsal) both to the adductor mandibule lateralis I. and to that part of the adductor mandi- bul lateralis II. that is inserted on the mandibula, the point of attachment of the tendon lying anterior (oral) to the points of insertion of the adductores. The muscle runs at first dorsally, or dorso-antero-laterally, internal and mesial to the adductor mandibule lateralis II., this being approximately as this muscle is shown in Tiesing’s fig. 15, pl. 7, of Rhinobatus, but not at all as it is shown in his figure of Raia. A few of the fibres of the muscle are inserted on the internal surface of that portion of the large fascia related to the adductor that gives origin, on its external surface, to those fibres of the adductor mandibule lateralis II. that have their insertions on the palato-quadrate. The levator then receives a bundle of fibres, of variable importance, from that part of the adductor mandibule lateralis IT. that is inserted on the mandibula, and, turning anteriorly, is inserted on the internal surface of what I have above described as the antero-lateral extension of the large fascia on the ventral surface of the adductor mandibule lateralis IL, its fibres crossing the dorso-lateral surface of the bulging anterior portion of the latter muscle. Most of the fibres of the levator cross, at a considerable angle, that terminal portion of the antero-lateral extension of the large fascia that, as above explained, has its insertion on the posterior surface of the nasal capsule, but a part of them are inserted on the base of: that portion of the fascia, that part of the fascia thus being, in part, the tendon of insertion of the levator. Tiesing and Marion both give this as The Muscles Related to the Branchial Aedes in Raia clavata 389 the sole insertion (origin) of the levator, and it apparently has so definitely become in Rhinobatus, as shown in Tiesing’s fig. 15, pl. 7. The m. levator labii superioris II. is considered by Luther to represent that part of the adductor mandibule of the Selachii that Vetter described as the muscle Addy, with which conclusion I fully agree, excepting that, as already stated, I would include in the muscle Addy those fibres of the adductor mandibule lateralis II. that have their insertions on the palato- quadrate. The long tendinous anterior portions of the muscle Addy of the Selachii are represented in portions of the fascia on which, in the Batoidei, the two muscles above referred to are inserted. This muscle Addy, so constituted, I consider to represent the musculus interbranchialis of the mandibular arch, the musculus adductor mandibulee lateralis II. being that part of the constrictor superficialis of the mandibular arch that lies distal (external) to the interbranchialis and is represented, in the Selachii, by the so-called dorso-ventral fibres of the constrictores superficiales. This distal portion of the primitive constrictor has here slipped over the distal edge of the musculus interbranchialis on to the anterior surface of that muscle, and also on to that surface of the cartilaginous bar of the mandibular arch, and, _ by its contraction, it compresses the mandibles and so increases the effect of the primary adductor of the arch, the latter muscle having been cut out of the middle of the dorso-ventral length of the proximal fibres of the primitive constrictor. The interbranchiales of the branchial arches of this fish have their origins both on the epal and the ceratal elements of their respective arches, and they in part run outward perpendicularly to those elements, as will be later described, this accounting’ for the course of the fibres in the mandibular arch. The fact that it is in this fish, and hence probably also inthe Selachii, only those fibres of the constrictor that lie distal to the interbranchialis that thus slip over on to the anterior surface of the arch, to there be added to the primary adductor, would then account for the varying relations of the definitive adductor of the Selachii to the nervus trigeminus, that nerve sometimes lying on the external surface of that muscle and sometimes between its superficial and deeper portions. The m. levator labii superioris III. is as described by Tiesing, and its relations to the levator labii superioris IV. would seem to indicate that it is a differentiated portion of that muscle. The m. levator labii superioris IV. is a thin band of muscle fibres which arises from the external (ventral) surface of the large fascia on the ventral surface of the adductor mandibule lateralis II. and, running antero-dorso- - mesially in a curved course, is inserted on the dorso-posterior surface of the nasal capsule. The so-called m. levator labii superioris V. of Tiesing’s descriptions is, 390 Mr Edward Phelps Allis as already stated, simply part of the large fascia that covers the ventral surface of the adductor mandibule lateralis II. The muscle Csd, is as described by Tiesing. The m. levator maxillee superioris arises by two tendons, as deseribed by Tiesing, but I find the lateral edge of the deeper, inferior one of these two parts of the muscle attached by ligamentous tissues to the lateral end of the mandibula. In one of the two specimens examined, but not in the other, this ligament formed the lateral prolongation of an aponeurosis that partly crossed the muscle transversely. In this same specimen the ventral surface of the inferior muscle was crossed transversely by muscle fibres and also by numerous fibrous strings, these strings forming, mesial to the muscle, a fascia which was attached to the lateral edge of the trabecular region of the cranium. The m. depressor mandibularis was, in one of my two specimens of Raia clavata, a single muscle such as is described by Tiesing and also by Marion. In the other specimen the muscle was double on each side of the head, one muscle lying immediately antero-mesial to the other, and, on one side of the head, partly internal to it. Both muscles arise from the lateral edge, or ventral surface, of the large median fascia that ensheaths the musculus coraco-mandibularis, and both are inserted on the mandibula, the anterior muscle by a tendinous end, but the posterior muscle directly on the cartilage, the muscle bundles of this latter muscle being directly continuous with those of the musculus adductor mandibule lateralis I. Tiesing says that this muscle is innervated by the nervus facialis; Marion calls it the posterior part of the constrictor of the mandibular arch, which would indicate that he found it innervated by the nervus trigeminus. I find both parts of the muscle innervated in Raia clavata by a branch of the nervus trigeminus, and two strictly similar muscles in Rhynchobatus are included by Luther among those innervated by the same nerve. A small musculus intermandibularis, similar to the muscle described by Marion in Raia erinacea as the muscle Csy,, was found in one of my two specimens but not in the other. Luther describes, in Raia clavata, a few feeble muscle fibres which he considers to represent the intermand!baiaaa but these fibres did not reach the median line. The muscle Csd, and the levator maxilla superioris have certainly both — been differentiated from the dorsal end of the primitive constrictor of the mandibular arch, and it would seem as if the levator maxilla superioris must be formed from those dorsal portions of the proximal fibres of the primitive constrictor that were left after the primitive adductor was cut out of them, the muscle Csd, representing corresponding portions of the more distal fibres of the constrictor. The levator maxille superioris would The Muscles Related to the Branchial Arches in Raia clavata 391 then be the serial homologue of the arcuales and interarcuales of the branchial arches, which Dohrn (1884) says are differentiated from the proximal fibres of their respective arches. The depressor mandibularis and the intermandibularis have been differentiated from the ventral end of the primitive constrictor, the depressor mandibularis, because of the continuity of its fibres with those of the adductor mandibule lateralis I, quite certainly from the proximal fibres of this part of the constrictor. The m. levator. hyomandibularis, levator rostri, and depressor rostri are as described by Tiesing, the levator hyomandibularis arising from the cranial wall immediately dorsal to the muscle Csd,. In each of my two specimens of Raia clavata the levator hyomandibularis is connected by a small bundle, or a few fibres, with a part of the muscle Csd, of Tiesing’s and Marion’s descriptions. On both sides of the one specimen of Raia radiata that was examined, three little bands of muscle fibres had separated from the muscle and, running together to a point, were attached to tissues related to the thymus. These little bands were not found in Raia clavata. The m. depressor hyomandibularis, excepting only a small anterior bundle, arises from the lateral edge of the large median ventral fascia, internal to the depressor mandibularis. The one bundle of the muscle that does not have this origin arose, in one of my two specimens of Raia clavata, by along tendon from the ventral surface of the mandibula of the opposite side of the head, not far from the symphysis, the tendon of the muscle of one side of the head traversing, as it crossed the median line, a perforation of the tendon of the other side. In the other specimen the tendons of the muscles of opposite sides were directly continuous with each other, and there was no insertion on the mandibule. As in the case of the levator hyomandibularis, this depressor muscle is connected by a small bundle, or by a few fibres, with a part of the muscle Csy, of Tiesing’s and Marion’s descriptions. The muscles related to the branchial basket may now be considered, and these muscles form a group that is wholly separate and distinct from those above described, excepting only in that, as just above explained, the muscle Cs, of Tiesing’s and Marion’s descriptions is connected by a few fibres both with the levator and the depressor hyomandibularis. The gill-pouches on either side of the head of Raia clavata have practically vertical anterior and posterior walls and horizontal dorsal and ventral walls, this being simply an exaggeration of the conditions described by me in Scyllium and Mustelus (Allis, 1917), in which latter fishes the anterior wall of each gill-pouch runs, from within, at first postero- externally and then curves posteriorly parallel to the external surface of the body. The lateral walls of the pouches of Raia are vertical in position, 392 Mr Edward Phelps Allis but slightly concave because of the pressure against them of the tissues related to the lateral fin. The anterior pouch of the series lies between the facialis and glossopharyngeus arches and is nearly transverse in position, inclining slightly anteriorly. The four remaining pouches lie be- tween each succeeding pair of branchial arches, from the glossopharyngeus to the fourth vagus, and incline successively more and more posteriorly. The five pouches, together, thus form a series of box-like structures placed side by side and wider at their lateral than at their mesial ends, each ~ box straddling the space between two adjacent branchial bars and being separated from its neighbour on either side by the related branchial diaphragm. The internal, or pharyngeal opening of each pouch is large. | The external opening is small, and lies at the lateral end of the ventral wall of the pouch, close against its posterior wall. The four branchial diaphragms that lie between the five branchial pouches, excluding the anterior and posterior walls of the branchial basket, are wedge-shaped, being thick at their internal edges and becoming rapidly thinner externally. The branchial rays form the posterior surface of the wedge, and lie against the anterior wall of the next posterior gill- pouch, all of them extending to the outer edge of that pouch. There, all of them, excepting the middle ray of the series, turn abruptly posteriorly, practically at right angles, and, becoming more or less flattened and — enlarged, fuse in part with each other, as is well shownin Foote’s (1897) figure of Raia radiata. These flattened outer ends of the rays lie in the dorsal and ventral walls of the next posterior gill-pouch, and extend nearly to the posterior edges of those walls, the bent outer end of the lateral one of the dorsal rays lying over the angle between the dorsal and lateral walls of the pouch, and the corresponding end of the lateral one of the ventral series lying over the angle between the ventral and lateral — walls, the bent and flattened outer ends of the rays thus forming a strong support to the dorsal and ventral walls of the pouch. Between the lateral ones of these two series of rays there is but a single ray, the middle one of the entire series, the base of this ray being contiguous with that of the adjacent ray on either side, but its outer end being separated from the ° outer end of each of those rays by a considerable interval. This middle ray of the series in each arch is not bent at its outer end, as the other rays all are, being straight throughout its entire length. It is directed laterally and slightly ventrally, and extends slightly beyond the level of the lateral wall of the gill-pouch next posterior to it, there being imbedded in the anterior edge of that wall, and, pushing it outward, forming a slight rounded protuberance in the wall, Its distal half has, in each arch, cut through the musculus interbranchialis of the arch, and is exposed on the The Muscles Related to the Branchial Arches in Raia clavata 393 anterior surface of that muscle, its proximal portion being covered by an aponeurotic line; and it and this aponeurotic line together have the position of the black line shown in Marion’s fig. 12 of Raia erimacea, and there apparently giving insertion to the fibres of the interbranchialis. What this black line is intended to represent I am unable to determine. It is shown by Marion as a branch of a main black line that follows the curve of the branchial bar of the arch, that lies along the anterior surface of the proximal ends of the fibres of the musculus interbranchialis, and that sends several smaller branches outward obliquely across the fibres of that muscle. The main black line cannot represent either of the main blood-vessels of the arch, for both those vessels lie posterior to the interbranchialis, and if Fie, 3. xii. it is intended to represent the nerve of the arch, the line that gives insertion to the fibres of the interbranchialis does not exist as a branch of the nerve. | The lateral fin of either side lies along the lateral surface of the branchial basket, and it and the related tissues are easily removed, leaving a clean surface of separation. When it has been removed there is exposed, on the concave lateral surface of the branchial basket (fig. 3), a series of three horizontal and five vertical vessels lying in the connective tissues that cover the lateral surface of the basket, these vessels quite certainly being venous and not lymphatic ones, for they can be injected from the brachial vein of T. J. Parker’s (1894) descriptions. The vertical vessels overlie the lines between the outer edges of each two adjacent gill-pouches. The horizontal vessels lie one at the dorsal edge of the lateral surface of the branchial basket, one at the ventral edge of that surface, and the third one midway between the other two. When these venous vessels and the VOL. LI. (THIRD SER. VOL. XIIL.)—JULY 1918. 27 394 Mr Edward Phelps Allis enclosing tissues are removed, it is seen (figs. 4. and 5) that the middle horizontal vessel overlies a line formed by the ventral edges of the dorsal constrictores superficiales, the fibres of those constrictores running antero- Csd. vs (Csdz) Csd. f. (Csd,). ie Ibr, (Csdq). | Tbr, (Csv,), ' Csv. f.(Csvs). © Fic. 4. x14. posteriorly and overlapping externally the fibres of the ventral constrictores, which here run ventro-dorsally. Directly internal to the ventral edge of each dorsal constrictor the corresponding ventral constrictor is crossed by bp. ITI. COsd.v,(Csds). Csyv. f. (Csv;). ini aT ee. ie oe a ie Csd, f. (Csdy). 7 Csv.tv, (Csv,). : , Fie, 5, x1}. le a horizontal aponeurosis, and this aponeurotic line must be the horizontal tendon described by both Tiesing and Marion in these fishes, unless it be that that tendon is simply the ventral edge of the continuous sheet formed by the dorsal constrictores. 'Tiesing says of this so-called tendon (Sehne) — > J = The Muscles Related to the Branchial Arches in Raia clavata 395 that it forms the boundary between the dorsal and ventral constrictores. Marion says that it starts anteriorly “from the lateral end of the hyo- mandibular cartilage and, running along the anterior surface of the gill region, connects with the propterygium and continues back in the lateral region, affording insertion to both the dorsal and ventral constrictores.” No such tendon was found in either of my three specimens, and it is evidently represented in the aponeurotic line above referred to, and to be later more particularly described. Marion says that the middle branchial ray in each arch runs outward to this tendon: In my three specimens the outer end of that ray lies at a relatively considerable distance ventral to the aponeurotic line, and there is no tendon related to it. The dorsal constrictores superficiales, as identified by both Tiesing and Marion, form a continuous muscle-sheet. The ventral constrictores form such a sheet on the ventral surface of the branchial basket, but on the lateral surface of that basket they are separate and distinct muscles. Each of these muscle-sheets is crossed, transversely to the body, by five aponeu- rotic lines. The dorsal lines start at or near the mesial edge of the sheet — and, running laterally in nearly straight lines that diverge slightly antero- posteriorly, reach the lateral edge of the branchial basket and there turn “ventrally along its lateral surface. The posterior four lines overlie the lines between each pair of the five gill-pouches, the anterior line lying along the dorso-anterior edge of the anterior pouch. That part of each of these five lines that lies on the dorsal surface of the branchial basket thus overlies the line where the branchial rays of the related arch bend abruptly | posteriorly in the outer wall of the next posterior gill-pouch. On the lateral surface of the branchial basket, the posterior four aponeurotic lines all extend ventrally to the ventral edge of the muscle-sheet in each of my three specimens, the anterior line also so extending in both specimens of Raia clavata. In the one specimen of Raia radiata (fig. 6) this anterior line extended ventrally to the same horizontal level as the posterior lines, but it did not there reach the ventral edge of the muscle-sheet, this ap- parently being a specific characteristic. In all three specimens the ventro- lateral ends of the dorsal aponeuroses of the second and third vagus arches -have each become a wide fascia instead of a simple aponeurotic line. In their course along the lateral surface of the branchial basket the several aponeurotic lines have no relations to the branchial rays, those rays not extending downward on to this surface of the basket. The ventral aponeurotic lines start at or near the mesial edge of the ventral muscle-sheet, and have a lateral and diverging course similar to that of the dorsal lines, but all of them are somewhat curved. When each of the posterior four lines reaches the ventro-mesial edge of the gill-opening 396 Mr Edward Phelps Allis of the next anterior gill-pouch, it curves posteriorly along that edge and then continues onward almost to the anterior edge of the next posterior gill-opening, this part of each aponeurotic line thus having no relation either to the line between two gill-pouches or to the line of the bend in the branchial rays. The anterior line of the series has a strictly similar course, but as there is no gill- pouch anterior to it, it does not cross the ventro- mesial edge of a giil-opening. Overlying each of these aponeurotic lines, both the dorsal and the ventral ones, there is a vein and an artery, the veins being branches of the dorsal and ventral horizontal veins on the lateral surface of the branchial basket, and the arteries being branches of certain of the branchial arteries. The muscles related to the glossopharyngeus and first two vagus arches & = * Fic. 6. x14. are practically similar, and are, as described by both Tiesing and Marion, the musculi constrictores superficiales, interbranchiales, interarcuales II. (arcuales), and adductores. The constrictores superficiales of these three arches are the muscles Cs,, Cs,, and Cs, of Tiesing’s and Marion's descrip- tions, and they are assigned respectively, by both those authors, to the first, second, and third vagus arches, The muscles in these three arches being practically similar, it will suffice to describe them in the glossopharyngeus arch alone, and it will be best to begin with the musculus interbranchialis. The musculus interbranchialis of the glossopharyngeus arch (fig. 9) lies upon the anterior surfaces of the branchial rays of that arch, between those rays and the posterior wall of the first gill-pouch, and it is separated into dorsal and ventral portions in part by the straight middle ray of the arch and in part by the aponeurotic line that lies upon the anterior surface of the proximal portion of that ray. The fibres of the dorsal half of the muscle have their origins on the ray, on the related aponeurotic line, and ce Cane ovo [ae ct ; zs ee a _ er Pe = bs A) ees oe, oe! Bae es eee The Muscles Related to the Branchial Arches in Raia clavata 397 on the epibranchial of the arch, and, running dorsally, but spreading both mesially and laterally, they all reach the line of the outer surfaces of the two gill-pouches between which they lie. There they are all, excepting a few lateral bundles, inserted on the dorsal linear aponeurosis related to their arch, that aponeurosis being the second one of the series of five, and the fibres of the interbranchialis being inserted on it throughout its entire length. Ventral to the ventro-lateral end of this aponeurosis, along the line between it and the outer end of the middle branchial ray of the arch, the few remaining bundles of this dorsal portion of the interbranchialis have their insertions in connective tissues of the region. The fibres of the ventral half of the musculus interbranchialis of this arch have their origins on the middle branchial ray of the arch, on the related aponeurotic line, and on the ceratobranchial of the arch, and they run ventrally, spreading mesially and laterally. A few bundles of the mesial fibres of the muscle, when they reach the level of the ventral surfaces of the two gill-pouches between which they lie, turn, in preserved specimens, sharply laterally for a short distance, and then sharply postero-mesially, and are inserted on the lateral edge of the large median fascia on the ventral surface of the head, immediately internal to the musculus depressor hyomandibularis. The double bend in this muscle in preserved specimens gives to it, in superficial view and as shown in fig. 2, the appearance of arising from the external surface of the ventral linear aponeurosis of the arch; but this appearance is misleading, for the fibres actually lie in the plane of, and mesial to, the other fibres of the muscle. They represent those portions of the proximal fibres of the primitive constrictor of the arch that lie ventral to the triangular piece cut out to form the adductor of the arch, these fibres not having been later subjected to a separation into inter- branchialis and constrictor superticialis muscles. They correspond to those fibres of the primitive constrictor that are described by Vetter, in the Selachii, as the deeper fibres of that constrictor, and they were correctly so considered to be by Tiesing. Marion, on the contrary, considers them to be overdeveloped fibres of the interbranchialis (1895, p. 17). Similar muscles are found in the first, second, and third vagus arches, and there are corresponding dorsal muscles in the second and third vagus arches; the ventral muscles apparently being serial homologues of the depressor hyo- mandibularis, and the dorsal muscles serial homologues of the levator hyomandibularis. The remaining fibres of the ventral portion of the glossopharyngeus muscle belong definitely to the interbranchialis, and are inserted on the ventral linear‘aponeurosis of the arch up to the point where that aponeurosis - turns posteriorly across the ventral edge of the gill-opening of the first gill- 398 _ Mr Edward Phelps Allis pouch. Beyond that point the fibres are inserted in connective tissues that lie in part upon the external surface of the anterior (proximal) edge of the ventral constrictor superficialis of the arch, and in part along that edge of that muscle, the line of insertion lying along the line between the first and second gill-pouches, and the conditions here being somewhat complicated, as shown in fig. 6. In the first, second, and third vagus arches strictly similar conditions are found, the musculus interbranchialis of each of these branchial arches of Raia thus extending between the branchial bar of its arch and the dorsal and ventral linear aponeuroses that overlie the lines where the branchial rays of the arch bend posteriorly in the outer wall of the next posterior gill-pouch. In Mustelus the fibres of the dorsal half of each musculus interbranchialis are, when they reach the outer edge of the next anterior gill-pouch, there in part inserted on the extrabranchial of their own arch and in part inserted on the linear aponeurosis that overlies that extra- branchial (Allis, 1917), that aponeurosis extending dorso-anteriorly across the dorsal end of the next anterior gill-opening. Where there are both dorsal and ventral aponeuroses, as is said by Vetter (1874) to be the case in Acanthias, the conditions are doubtless strictly similar, the musculus inter- branchialis of each arch of that fish thus undoubtedly lying, as in Raia, between the branchial bar of its arch and the extrabranchial, or related aponeurosis, of the arch. The interbranchiales of Raia and the Selachii are thus quite certainly strictly homologous muscles, notwithstanding the marked difference in direction of their fibres. There is, as already stated, no interbranchialis in the fourth vagus arch in any of these fishes, and neither Tiesing nor Marion describes it in the facialis arch. It will, however, later be shown that the so-called musculus Cs, of these authors’ descriptions is the interbranchialis of the facialis arch and not the constrictor superficialis. Anterior to the distal (external) end of the ventral half of the inter- branchialis of each branchial arch of Raia, in the angle between that muscle and the next anterior ventral constrictor superficialis, there is a small muscle (fig. 9), with one to three muscle bellies, which arises in connection with the linear aponeurosis of the arch, and, running laterally, is inserted on the ventral edge of the gill-opening related to that apo-— neurosis. These muscles are innervated by branches of the nerves that innervate the interbranchialis against which they lie, those branches, in each arch, passing over the posterior edge of the little muscle to penetrate it on its ventral (external) surface. These little muscles have evidently been specialised to act on the gill-openings, The dorsal and ventral constrictores superficiales of the glosso: The Muscles Related to the Branchial Arches in Raia clavata 3899 pharyngeus arch are represented in the muscle fibres that lie upon the external surface of the second gill-pouch, and hence between the second and third dorsal and ventral linear aponeuroses, those aponeuroses being related, respectively, to the first and second gill-openings and also to the bends in the branchial rays of the glossopharyngeus and first vagus arches. The fibres of the dorsal constrictor, the muscle Csd, of Tiesing’s and Marion’s descriptions, all run antero-posteriorly, and, excepting a few ventro-lateral bundles, all extend from one aponeurosis to the other. The few ventro-lateral bundles referred to incline somewhat ventro-posteriorly, and do not reach the first vagus aponeurosis, ending, without evident insertion, along the free ventral edge of the muscle. The fibres of the ventral constrictor superficialis, the muscle Csv, of Tiesing’s and Marion’s descriptions, run, from in front, postero-mesially, extending between the two related ventral linear aponeuroses. These two aponeuroses are, as already stated, curved, eacl® crossing the ventro- mesial edge of the gill-opening next anterior to it, and then continuing posteriorly toward the middle of the length of the next posterior gill- opening. The fibres of the constrictor, extending between these two aponeuroses, are accordingly inclined, progressively, more and more mesially the nearer they lie to the lateral edge of the branchial basket, and finally acquire a course parallel to and along the anterior edge of the first vagus gill-opening. These lateral, and hence distal, fibres of this portion of the constrictor are directly contiguous with fibres that arise from the antero-lateral edge of that terminal part of the glosso- pharyngeus aponeurosis that lies between the glossopharyngeus and first vagus gill-openings, and these fibres together form a broad band which, running laterally and slightly anteriorly, turns over the ventro- lateral edge of the branchial basket and then upward along its lateral surface until it reaches the ventral edge of the dorsal constrictor of the arch. There the muscle band is crossed by the horizontal aponeurosis, already described, that lies directly beneath the ventral edge of the dorsal constrictor superficialis of the arch. Dorsal to this horizontal aponeurosis, the muscle band continues upward along the lateral surface of the branchial basket and then a short distance mesially along its dorsal surface, but the several bundles of fibres that form the band here become successively shorter, from the anterior to the posterior edge of the muscle, the muscle thus presenting the appearance of a band that has been cut across diagonally. The dorsal ends of most of these bundles have no special insertions, excepting in the connective tissues that cover the underlying first vagus branchial pouch, but the longer, posterior ones are inserted on the internal surface of the dorsal constrictor superficialis 400 Mr Edward Phelps Allis of the arch, in the angle between that constrictor and the interbranchialis of the first vagus arch. The ventral constrictor superficialis of this fish thus corresponds to the dorsal constrictores of Mustelus (Allis, 1917) in that the proximal fibres of the muscle extend between the linear aponeurosis of their own arch and that of the next posterior arch, while the distal fibres have a ventro-dorsal course between the gill-openings related to those two aponeuroses. Along the line where these ventro-dorsal fibres of the muscle of Raia pass over the ventro-lateral edge of the branchial’ basket, they are crossed by a little bar of cartilage which lies upon their external surface, loosely attached to them by connective tissues.- Mesial to this bar, the gill-opening of the first branchial pouch overlaps externally the _ anterior (proximal) edge of the ventro-dorsal fibres of the constrictor, and that edge is there also overlapped externally by the outer ends of certain of the fibres of the ventral portion of the interbranchialis of the arch, as already described. Anterior to the anterior end of the little bar of cartilage, and for a short distance dorsal to it, there is, on the external surface of the anterior edge of the constrictor muscle, a line of tough connective tissue which gives insertion to certain fibres of the muscle and also to certain of the fibres of the interbranchialis of the arch, as shown in fig. 6 and as already referred to when describing the musculus interbranchialis. Dorsal to this line of tough connective tissue, the constrictor muscle passes between the outer ends of the middle branchial rays of the glossopharyngeus and first vagus arches, completely filling the space between them. There are five of the little bars of cartilage above referred to, one related to each of the ventral constrictores, and they have already been described by Max Fiibringer (1903) in this fish and others of the Batoidei, but he says that they each lie on the outer surface of the related musculus interbranchialis, this being due to a misconception of the homologies of these muscles. In the first and second vagus arches the constrictores superficiales are strictly similar to those above described in the glossopharyngeus arch. In the third vagus arch the conditions are somewhat different. The dorsal and ventral constrictores superficiales of this arch are the muscles Csd, and Csv, of Tiesing’s and Marion’s descriptions, and they are considered by them to belong to the seventh visceral, or fourth vagus, arch. The mesial fibres of both the dorsal and the ventral con- strictor separate, as already explained, as the deeper fibres of Vetter’s descriptions of the Selachii. The remaining fibres of that part of the dorsal constrictor that lies on the dorsal surface of the branchial basket run posteriorly and somewhat mesially from the linear aponeurosis of The Muscles Related to the Branchial Arches in Raia clavata 401 their arch to the hind edge of the fifth gill-pouch. There the lateral fibres of the muscle are inserted by ligamentous tissues on the shoulder girdle, the remaining fibres passing over the hind edge of the pouch and then turning downward along its posterior wall to have their insertions on it, this wall of the pouch lying against the musculus trapezius and inclining strongly ventro-anteriorly. On the lateral surface of the branchial basket the dorsal and larger portion of the fibres of the dorsal constrictor all arise from the dorsal linear aponeurosis of their arch, but the ventral fibres arise from an aponeurotic line that is evidently the homologue of the horizontal aponeuroses of the more anterior arches. This horizontal aponeurosis of the third vagus arch always begins at the point where the horizontal aponeurosis of the - second vagus arch meets the dorsal linear aponeurosis of the third vagus arch, and it has a curved, or angular, ventro-posterior course. The ventral fibres of the lateral portion of the dorsal constrictor of the arch always arise from the dorsal edge of this aponeurosis, those fibres of the ventral constrictor that have a ventro-dorsal course being inserted on its ventral edge. The posterior fibres of that part of the ventral constrictor of the _ second vagus arch that has a ventro-dorsal course are also inserted on this aponeurosis, in each of my three specimens, there overlapping externally the fibres of the third vagus arch. From this surface of origin the fibres of the dorsal constrictor of the third vagus arch run dorso-posteriorly, the ventro-posterior ones inclining more and more dorsally, and they all- pass over the hind edge of the fifth gill-pouch and are inserted on the posterior wall of that pouch. The ventral constrictor of this arch, the muscle Csv, of Tiesing’s and Marion’s descriptions, closely resembles the muscles of the more anterior arches excepting in that its ventro-dorsal fibres end at the horizontal aponeurosis of the arch, above described, and do not there pass upward internal to the dorsal constrictor of the arch. On the ventral surface of the branchial basket, the fibres of this constrictor, running postero-mesially, turn dorso-postero-mesially over the hind edge of the fifth gill-pouch and there separate into four or five parts which all pass between the musculi coraco-branchiales, or between parts of those muscles, and are inserted in part on the posterior wall of the fifth gill-pouch and in part on the membranous wall of the pericardial cavity. Each musculus interbranchialis is innervated by branches of the nerve of its arch, those branches running outward on the anterior surface of the muscle. Most of these branches, when they reach the outer edge of the muscle, penetrate it, and traversing it and the related linear aponeurosis issue, at the hind edge of the latter aponeurosis, on the dorsal surface of ) 402 Mr Edward Phelps Allis the next posterior constrictor superficialis. Such branches are found in each arch, going to all portions of the next posterior constrictor superficialis, excepting only the dorso-posterior corner of the ventral constrictor, and they unquestionably innervate these muscles; the constrictor superficialis that lies next posterior to a given aponeurosis thus belonging to the arch — to which that aponeurosis is related. The dorso-posterior corner of the ventral constrictor of each arch is apparently innervated by branches of one or two of the large branches related to the next posterior interbranchi- alis. This branch (or branches), when it reaches the outer edge of the interbranchialis to which it is related, turns anteriorly on to the dorso- posterior edge of the dorso-posterior end of the ventro-dorsal fibres of the constrictor superficialis of the next anterior arch, and sends branches into it, as shown in figs. 7 and 8. In four of eight cases that were examined, : Fic, 7, x14. Fic. 8. x14. a portion of the fibres so innervated formed a muscle bundle that was somewhat separate from the remainder of the muscle, this suggesting a muscle similar to the one found in corresponding position, and with similar innervation, in the ventral half of the arch, and there differentiated in reference to some action on the gill-opening. This ventral muscle I con- sider to be derived from some part of the muscle of the arch to which the nerve that innervates it belongs, for there is no apparent, reason whatever to assume that it has any other derivation. The corresponding dorsal muscle would then seem to be a similar muscle, but there in process of dis- placement and disappearance; for I do not believe that the conditions there presented represent a stage in a change of innervation of certain fibres of the constrictor superficialis of a given arch from the nerve of its own arch to that of the next posterior arch. If it be assumed that they do represent a stage in such a change of innervation, then it must be explained why these little muscles are not differentiated in relation to the gill-openings of the fifth branchial pouches as well as to those of the more anterior pouches, for no such muscles, either dorsal or ventral, are found The Muscles Related to the Branchial Arches in Raia clavata 408 related to those openings. If, on the contrary, these little muscles are derived from some part of the primitive constrictor of the arch to which the nerve that innervates them belongs, then their absence in relation to the fifth gill-openings is fully explained by the absence of musculi inter- branchialis and constrictores superficiales in the fourth vagus arch of either side. In the third vagus arch the muscles are all innervated exactly as in the more anterior arches, no branch of the nerve of the fourth vagus arch having, so far as could be determined, any relations to any of them. A musculus adductor branchialis is found in each of the four branchial arches above considered, and also a small musculus interarcualis II. (arcualis), this being as both Tiesing and Marion have stated. From the above descriptions of the muscles related to the glosso- pharyngeus and first three vagus arches of Raia, it is evident that the primitive constrictor muscle of each of those arches, when it reached the related dorsal and ventral linear aponeuroses, each turned posteriorly above the next posterior gill-pouch and then extended onward to the linear aponeurosis that lies along the hind edge of that pouch. Whether or no the fibres of the muscle then still continued onward, overlapping the fibres of the next posterior constrictor, cannot be determined from my dissections. Dohrn’s (1884) figures of longitudinal vertical sections of embryos of Torpedo, however, strongly favour the assumption that the fibres of each dorsal constrictor were so continued posteriorly; for these constrictores are shown by him as a continuous line of muscle fibres running posteriorly above the three vagus gill-pouches, but partially interrupted between the glossopharyngeus and first vagus pouches. The ventral constrictores are, of the contrary, shown by Dohrn each turning posteriorly above the next posterior gill-pouch, and definitely ending when it reaches the constrictor of the next posterior arch. That the fibres of the ventral constrictores should there definitely end, while the fibres of the dorsal constrictores continue onward beyond the corresponding points, seems to me improbable ; but, however it may be, the conditions, as shown by Dohrn, definitely confirm my conclusion that the interbranchialis muscle of each arch was primarily continuous with the constrictor superficialis that overlies the gill-pouch next posterior to it. The musculi constrictores superficiales of Raia are thus strictly com- parable to those in the Selachii, instead of being markedly different from them, as they would be if, as Tiesing and Marion both maintain, the _ muscle of each arch lay anterior to the gill-opening of the gill-pouch next anterior to that arch. Marion probably here simply accepted Tiesing’s earlier and uncontested statement regarding them, and Tiesing evidently 404 | _ Mr Edward Phelps Allis based his conclusion wholly on what he considered to be the innervation of the muscles, without giving any consideration to the resulting impos- sible position of the ventro-dorsal fibres of each ventral constrictor. According to him, the muscle Cs, of his descriptions is innervated by “ feinen Seitenisten des N. Glossopharyngeus,” and the muscles Cs,—Cs, each, respec- tively, by corresponding branches of the first to the fourth vagus nerves. No such branches could be found, in any of my specimens, going to the muscles indicated. es ; The constrictores superficiales of the Batoidei and Selachii thus being homologous, it is practically certain that the linear aponeuroses related to those muscles also are. The aponeuroses in each arch of Raia must then be, as in the Selachii, structures developed, during the life of the individual, as the result of the passage of a primitively continuous muscle over an underlying skeletal structure; and if these underlying structures are not simply the angles formed by the bends in the dorsal and ventral branchial rays, there must have been, primarily, independent extrabranchials in this fish that later fused with the branchial rays to form the flattened and partly fused outer ends of those rays, that are actually found in the aduit. But, however this may have been, the under-— lying structures must have had, in embryonic stages, a greater distal extent, relative to the muscles, than they actually have in the adult, for each of the dorsal aponeuroses extends distally considerably beyond the branchial rays of the related arch. The conditions in Raia can then be explained by assuming a descent from a Selachian such as Acanthias, in which there were both dorsal and ventral: linear aponeuroses (Vetter, 1874), each aponeurosis extending posteriorly slightly beyond the related edge of the gill-opening of the gill-pouch next anterior to it, as the dorsal aponeuroses actually do in the adult Mustelus (Allis, 1917). The lateral fin of the Batoidei then gradually developing, with the associated flatten- ing of the head and the gill-pouches, the gill-openings would necessarily be forced from the lateral surface of the head, and they have actually been | forced upon its ventral surface. The dorsal and ventral branchial rays have been actually, at the same time, forced on to the dorsal and ventral surfaces, respectively, of the branchial basket. This change in position of the gill-openings and the branchial rays would probably first give rise to conditions such as are actually found in the third vagus arch of Raia, the distal end of the ventral linear aponeurosis being simply forced into a curved course, while the corresponding end of the dorsal aponeurosis was left considerably dorsal to the gill-opening and pulled into a curved course, the hollow of the curve being actually directed dorsally. That part of the constrictor superficialis of the arch that primarily lay between the fourth , The Muscles Related to the Branchial Arches in Raia clavata 405 and fifth gill-openings and had a dorso-ventral course would retain that course, and would still lie between the two related aponeuroses, but it would lie largely dorsal to the gill-openings. In the more anterior arches those fibres of the dorsal constrictor of each arch that lay proximal to the curved distal end of the dorsal aponeurosis of the arch then apparently sheered ventrally over the more distal fibres, and so gave rise to the conditions actually found in those arches. The horizontal aponeurosis across the ventro-dorsal fibres of gach ventral constrictor would then be simply the persisting distal portion of the primitive dorsal aponeurosis of the muscle of the arch. Thus considered, there would be nothing in the muscles of these four branchial arches of Raia that is not strictly comparable with the muscles in the Selachii, excepting only the little muscles that lie, in the dorsal and ventral halves of each of these arches, in the angle between the outer end of the interbranchialis of the arch and the constrictor superficialis of the next anterior arch. These little muscles have no apparent homologues in the Selachii. That they are the homologues of the little bands of fibres that I described in Scyllium (Allis, 1917) as there being in process of differ- entiation in relation to the gill-openings, I consider wholly improbable, for in that case they would have lost their primitive innervation by the nerve of their arch and have acquired a secondary innervation ‘by the nerve of the next posterior arch. The muscles Cs, and Cs, of Tiesing’s and Marion’s descriptions now remain to be considered, and these muscles are, respectively, related to the first gill-pouch and the facialis arch as the interbranchiales and con- strictores superficiales of the more posterior arches are each related to the corresponding gill-pouch and branchial arch. This position, and also their innervation, shows that they are, respectively, the interbranchialis and constrictor superficialis of the facialis arch, and they are so designated by index letters in my figures. They were considered by Tiesing and Marion to be, respectively, the constrictores superficiales of the facialis and glossopharyngeus arches, and as this identification of them has long been accepted, it will simplify the descriptions to still refer to them by the letters given to them by both those authors. The posterior edge of the first gill-pouch is nearly transverse in position, but its anterior edge is directed quite strongly antero-laterally, the pouch being considerably wider at its lateral than at its mesial edge, and strgd- dling the space between the glossopharyngeus arch and the epibranchial and ceratobranchial of the facialis arch. The branchial rays related to the latter arch all bend posteriorly, at their outer ends, in the outer wall of the first gill-pouch, the middle ray of the series included, and there is 406 Mr Edward Phelps Allis no space between this middle ray and the next adjacent one on either side, as there is in the more posterior arches. A dorsal linear aponeurosis overlies the bend in the dorsal rays and separates the muscles Csd, and Csd,, this linear aponeurosis being strictly similar to those in the more posterior arches but, in Raia radiata (fig. 6), differing from those aponeu- roses in that it does not extend ventrally to the ventral edge of the muscle Csd,. As far as the ventro-lateral end of this aponeurosis of Raia radiata, the fibres of the muscle Csd, extend between it and the aponeurosis * related to the glossopharyngeus arch. Ventral to that point the fibres of Csd. g. (Csd,). Csd. f. (Csd,). Csy. f. (Csvs3). Smet ttf) fe Ibr,. ut emeee MDI. g. bmp Csv. g. (Csy,). Csv. f. (Csv,). Fie. 9. x1%. the muscle are directly continuous with those of the muscle Csd,, and they do not extend posteriorly to the same extent that the dorsal fibres do, simply overlapping the ventro-dorsal fibres of the muscle Csv, and there ending, attached by connective tissues to its external surface. ‘The ventral linear aponeurosis overlies, in all these fishes, the bend in the ventral branchial rays of the facialis arch and lies between the muscles Csy, and Csv,, but this aponeurosis is usually less developed than the corresponding ones in the more posterior arches; this being in accord with the position and the less strongly developed condition of the ceratobranchial of the facialis arch. The lateral end of the aponeurosis may even turn toward the glossopharyngeus aponeurosis but vanish before reaching it. Lateral to this lateral end of the ventral aponeurosis, there is a little superficial bar’ The Muscles Related to the Branchial Arches in Raia clavata 407 of cartilage strictly similar to those in the more posterior arches. The distal fibres of the ventral constrictor, Csv,, run dorsally internal to this little cartilage and then internal to the dorsal constrictor, Csd,, exactly as in the more posterior arches, but these fibres form a wider and more strongly developed band than in the latter arches. These ventro-dorsal fibres of this constrictor are crossed by a horizontal aponeurosis which is strictly comparable to those in the posterior arches, but it does not, as in those arches, follow the ventral edge of the overlapping portion of the dorsal constrictor, lying, in its anterior portion, dorsal to that edge. a Ibr, (Csd,). Csy. f, (Csvs). = C, we br, (Csv2). Csy. f. (Csy,). Fie. 10. x14. The muscle Cs,, which is the interbranchialis of the facialis arch, ditfers in certain important respects from those muscles in the more posterior arches, and is shown, in anterior view, in fig. 10. In this arch the articulat- ing ends of the epibranchial and ceratobranchial are directed laterally, or even antero-laterally, instead of postero-laterally as in the more posterior arches, and because of this no adductor muscle has been cut out of the primitive constrictor of the arch. The fibres that correspond to that adductor have, however, here slipped over on to the anterior surface of those two elements of the branchial bar of the arch, exactly as in the more posterior arches, and certain of the most proximal fibres have acquired insertion on the ventral end of the pharyngeal element of the arch, the so-called hyomandibula, and so given rise to the levator and the depressor 408 ae Mr Edward Phelps Allis hyomandibularis. The next distal fibres of the primitive constrictor form, in certain specimens, a somewhat separate bundle which passes dorso- ventrally over the lateral edge of the pharyngeal cavity, extending approxi- mately from the level of the dorsal end of the epibranchial to the ventral end of the ceratobranchial, these fibres thus having approximately the position of an adductor of the arch. A few fibres connect this bundle, at the middle of its length, both with the levator and with the depressor hyomandibularis, the conditions varying in different specimens but clearly showing the primitive continuity of these several muscles. The next distal fibres of the primitive constrictor form the musculus interbranchialis, and practically all of them have their insertions on an aponeurosis that overlies the middle one of the branchial rays of the arch—few, if any, of them being inserted either on that ray or on the epibranchial or ceratobranchial of the arch. The aponeurosis does not extend to the outer end of the middle branchial ray, certain of the muscle fibres radiating, fan-shaped, from its outer end, and the dorsal and ventral halves of the interbranchialis there being continuous. Certain of these radiating fibres always pass over the antero-lateral edge of the branchial basket, and they there form, in Raia radiata, the ventral portion of the dorsal constrictor Csd, of Tiesing’s and Marion’s descriptions, as shown in fig. 6. Ventral to tied fibres, a few bundles of the muscle are inserted, as in the more posterior arches, along the anterior edge of the ventro-dorsal fibres of the ventral constrictor of the arch, the muscle Csy, of Tiesing’s and Marion’s descriptions, all the remaining fibres of the muscle being inserted on the dorsal and ventral aponeuroses that separate this muscle from the muscles Csd, and Csv, respectively. No musculus arcualis has been differentiated in this arch. Branches of the nervus facialis go to all parts of the muscles Cs, and Cs,, exactly as in the more posterior arches. There is thus no possible question as to these muscles representing, respectively, the interbranchialis and the constrictor superficialis of the facialis arch. . The m. coraco-mandibularis is an azygous muscle which arises from the membranous ventral wall of the pericardial chamber, and, running forward in the median line, is inserted on the mandibule internal to the musculus intermandibularis, It is enclosed in a membranous sheath which arises posteriorly from the shoulder girdle and the ventral wall of the pericardial chamber, and it forms the median ventral fascia of the fish, one layer of the fascia lying external and the other internal to the coraco-mandibularis, The lateral edges of this fascia give origin, on either side, to the musculi depressor mandibularis, depressor hyomandibularis, and depressor rostvi, and insertion to the four muscles formed by the deeper proximal fibres The Muscles Related to the Branchial Arches in Raia clavata 409 of the constrictores superficiales of the glossopharyngeus and first three vagus arches. _ The m. coraco-hyoideus of either side arises largely from the internal surface of the internal one of the two layers of the median ventral fascia above described, but partly also from the ventral (external) surface of the coraco-hyomandibularis, and, running forward between the so-called afferent mandibular artery (arteria thyreo-spiracularis, Dohrn) and the common trunk of the afferent arteries to the hyal and first branchial arches, is inserted on the cartilage called by Gegenbaur (1872) the copula of the facialis arch (“ Zungenbein copula”), but by Parker (1876) the hypo- branchial of the first branchial arch. This cartilage, which is evidently the hypobranchial of the first branchial arch, was, in one or two specimens that were examined, formed of two pieces, one on either side, the two pieces articulating with each other in the median line. The m. coraco-hyomandibularis lies internal to the coraco-hyoideus, and arises, posterior to the latter muscle, from the internal surface of the internal one of the two layers of the median ventral fascia, the muscles of opposite sides there being contiguous in the median line. It runs antero- laterally, internal to the common trunk of the afferent arteries to the hyal and first branchial arches, and was inserted, in one of my two specimens, on the hyomandibula alone. In the other specimen it was also so inserted on one side of the head, but on the other side it was inserted by two tendinous ends, one on the hyomandibula and the other on the mandibula, This muscle has, to the afferent arteries, the relations that the coraco- branchialis of the first branchial arch of the Selachii has to those same arteries, and it is apparently the homologue of that muscle. The mm. coraco-branchiales all arise from the shoulder girdle, partly tendinous and partly fleshy, and separate into three parts. The anterior one of these three parts is inserted by two ends, one on the long anterior end of the fifth hypobranchial of Parker’s descriptions (the large posterior copula of Gegenbaur’s descriptions), and the other on the third hypo- branchial. The next division of the muscle is inserted on the fourth and fifth hypobranchials, and the third division on the fifth ceratobranchial. _ ‘The mm. coraco-arcuales communes arise from the shoulder girdle and are inserted mainly on the ventral wall of the pericardial chamber, a small lateral bundle on either side being inserted on the ventral surface of the internal one of the two layers of the median ventral fascia. The muscle marked x in fig. 1 arises in part from the fascia that covers the dorsal surface of the trunk muscles and in part from the anterior vertebra or vertebra, and it is inserted, by two tendinous ends, on the ventral surface of the hind end of the chondrocranium, the two tendinous VOL. LIL. (THIRD SER. VOL. XIII.)—JULY 1918. 28 410 Mr Edward Phelps Allis ends passing one on either side of the nervus vagus. Its innervation was not determined, but it would seem as if it might be a modification of the musculus sub-spinalis of the Selachii. LITERATURE. Aus, E. P., jr. (1917), “The Homologies of the Wliisclsk related to the Visceral Arches of the Gnathostome Fishes,” Q./.4Z.8 , vol. lxii., London. Donen, A. (1884), “Studien zur Urgeschichte des Wirbeltierkérpers, IV,” Mitteil. a. d. Stat. Neapel, Bd. v., Leipzig. Foorr, E. (1897), “ Extrabranchial Cartilages of Elasmobranchs,” Anat. Anz., Bd. xiii., Jena, Foirsrincer, M. (1903), ‘* Notiz tiber oberflachliche Knorpelelemente im Kiemen- skelet der Rochen (Extraseptalia),” Morph. Jahrbuch, Bd. xxxi., Leipzig. GEGENBAUR, C. (1872), Untersuchungen zur vergleichenden Anatomie der Wirbel- thiere: Heft 3, Das Kopfskelet der Selachier, ein Beitrag zur Erkenntnis der Genese. des Kopfskelets ‘der Wirbelthier ‘e, Leipzig. Lutuer, Avex. (1909), “ Beitriige zur Kenntnis von Muskulatur und Skelett des Kopfes des Haies Stegostoma tigrimum Gm. und der Holocephalen, mit einem Anhang iiber die Nasenrinne,” Acta Soc, Sc, Fennice, Helsingfors. Marton, G. E, (1905), “ Mandibular and Pharyngeal Muscles of Acanthias and Raia,” Tuft’s College Studies, vol. ii., No. 1, Mass. PARKER, T. J. (1884), A Course of Instruction in Zootomy, London. Parker, W. K. (1876), “On the Structure and the Development of the Skull in Sharks and Skates,” 7rans, Zool. Soc., vol. x., London. Tresine, B, (1896), ‘ Ein Beitrag zur Kenntnis der Augen-, Kiefer- und Kiemen- muskulatur der Haie und Rochen,” Jenaische Zeitschr. fiir Naturwiss., Bd, xxx., Jena. Verrer, B. (1874), “ Untersuchungen zur vergleichenden Anatomie der Kiemen- und Kiefer-musculatur der Fische,” Jena. Zeit. f. Naturwiss., Bd. viii., Jena, EXPLANATION OF FIGURES. Fig. 1. Dorsal view of the muscles related to the branchial basket on the left- hand side of the head of Raia radiata. The m., levator rostri cut and turned back. Sections of the constrictores superficiales dorsales of the glossopharyngeus and first vagus arches also cut and turned back, exposing the underlying gill-pouches and the dorsal ends of the constrictores superficiales ventrales, x 1}, Fig, 2, Ventral view of the same. M., depressor rostri cut and turned back. x 1h. i 3. Lateral view of the branchial basket of Raia clavata, showing the venous vessels and connective tissues that cover the branchial muscles. x 1}. Fig. 4. The same, the venous vessels and connective tissues removed, exposing the branchial muscles. x 1}. Fig. 5. The same, the constrictores superficiales dorsales of the facialis and first The Muscles Related to the Branchial Arches in Raia clavata 411 vagus arches cut and turned back, exposing the dorsal portions of the constrictores superficiales ventrales. x 13. Fig. 6. Ventro-lateral view of the anterior portion of the branchial basket of ey radiata, x 1. Fig. 7. Anterior view of a branchial diaphragm of Raia clavata, showing the dorsal end of the constrictor superficialis ventralis of the next anterior arch. x 14. Fig. 8. The same, the constrictor superficialis ventralis pulled down so as to show the nerve that innervates it. x 14. Fig. 9. Anterior view of the m. interbranchialis of the glossopharyngeus arch of Raia clavata, the dorsal and ventral constrictor superficialis of the arch forced outward so as to be seen in the figure. x bh. Fig. 10. Similar view of the m, interbranchialis of the facialis arch of Raza . clavata, »* 1h, INDEX LETTERS. Abrs. M. adductor branchialis of glossopharyngeus arch, ap. f. Linear aponeurosis related to the facialis arch. ap. g. ~ ‘ x », glossopharyngeus arch. ap. V,- »» aponeuroses ,, » first to the third vagus arches. be. I-V. First to fifth branchial clefts. bp. I-V. 2 pouches. C. Little bars of cartilage related to each m. constrictor super- ficialis ventralis, Care. M. coraco-arcuales. Csd. f,-Csd. v,. Mm. constrictores superficiales dorsales of the facialis, glosso- pharyngeus, and first three vagus arches. Csd,-Csd The same, as identified and lettered by Tiesing and Marion. md. M. coraco-mandibularis. Csp;, . Proximal bundle of fibres of second vagus (fifth visceral) arch. Csp,- os third vagus (sixth visceral) arch. Csv. f.—-Csy. v5. Mm. constrictores superficiales ventrales of the facialis, glosso- pharyngeus, and first three vagus arches. Csv,—Csv,. The same, as identified and lettered by Tiesing and Marion. Dhym. M. depressor hyomandibularis. Dmd. ‘ ‘i ie mandibularis. Dr. Ete 7 Me rostri. Ibr,. », interbranchialis of facialis arch. Tbr. of glossopharyngeus arch. mbr. g. Middle branchial ray of glossopharyngeus arch. mbr. vy). ‘* », Of first vagus arch, Lr. . M. levator rostri. Tr, ,», trapezius. 583. M. subspinalis (?). THE PRIMORDIAL CRANIUM OF PCECILOPHOCA WEDDELLI (WEDDELL’S SEAL), AT THE 27-MM. C.R. LENGTH. By Professor Epwarp Fawcert, M.D. Edin., University of Bristol. (Illustrated by drawings of models.) ? THE embryo was kindly supplied to me by Prof. D’Arcy Thompson of St Andrews University; it had been preserved in formalin since 1892, and was in as fair condition as could be expected after such a long stay in formalin. It was embedded in paraffin, and tut into sections of 20 microns in thickness. Its total crown rump length was 27 mm. The sections were experimentally stained, some of them with Delatield’s hematoxylin and eosin, some with thionin, and others with Mallory’s triple connective- tissue stain. Of these stains the last proved to be the most satisfactory, but the result was far behind what one usually gets with this stain, no doubt on account of the long stay in formalin and the want of a mercury fixative. It is my experience that Mallory’s stain is not so effective in those cases in which there has been prolonged soaking in formalin, and the chief failure is in regard to the fuchsin part of the stain. The preliminary stain in fuchsin seems to take readily enough, but is only indifferently fixed by the phospho-molybdie acid, so that it tends to wash out after the sections are passed through the anilin-blue. However, the latter does give a really satisfactory stain for the cartilage and a magnificent stain for bone, so, apart ~ from the general somewhat blue appearance of the sections, it is pretty easy to determine what is cartilage and what is what is called procartilage. Neither hematoxylin nor thionin gave anything like such good results, — and the sections so stained were re-stained in Mallory’s stain. Being a somewhat rare animal, and possessing most of the peculiarities of the seal group, it was considered important that a model should be made of the chondrocranium as a whole, to be supplemented by additional models of certain parts which cannot be studied in any complete model. It was further hoped that perhaps the affinities of the seal to other groups of carnivora might be determined, more particularly as to whether it lay nearer the bear group or the cat group, as the position is by no means clear at the present time; but there are many difficulties. ' The expenses of this research were defrayed by a grant from the University of Bristol Colston Society. . Sie pe . a ee eee The Primordial Cranium of Pewcilophoca Weddelli 413 In the first place, there is no sufficient description of the chondro- cranium of the bear, nor can one replace that by the dog; for although dogs are common enough, they are practically sacred animals, and the removal of dead embryos from the body of a dog killed in the lethal chamber of a dog’s home is looked upon as vivisection, and one cannot obtain embryos in this way!! As I have models of the cat’s chondro- cranium as well as models of that of the ferret, a comparison will be drawn between that of this seal and those just mentioned, and so far as possible with the known facts in regard to the bear and dog. Owing to its aquatic habits the seal has naturally undergone some modification of structure—and these largely concern the parts in the nasal region. Thus the lacrimal conducting apparatus is awanting, together with the lacrimal bone ; there are no sinuses in the cranial bones, although their representatives are well developed in the chondrocranium. The olfactory region of the nasal capsule is of very small size, the organ of Jacobson is absent, and the auditory ossicles are of large size, and as a consequence the muscles which are attached to them are enormous. The thyroid cartilages show a very primitive form at the stage modelled, and it will be shown that the tectum cranii anterius is of special interest. Taken as a whole the chondrocranium is very simple, and might be used as a standard from which one might work forwards 6r backwards. Method of Description.—In accordance with the plan adopted in the ease of ‘Microtus and Erinaceus, the primordial cranium of this seal will be treated as consisting of a neural and a visceral part, and the bones will then be described separately. THE NEURAL CHONDROCRANIUM. The neural chondrocranium will be considered as consisting of the following parts, viz. :— 1. A central stem. 2. Appendages to the central stem. 3. Commissures connecting the central stem with the aforesaid " appendages. 4. Lateral structures. 5. Lateral commissures. , 6. Dorsal structures. 1. THE CENTRAL STEM. The central stem is divisible into three parts, which are from behind forwards the pars chordalis, the pars trabecularis, and the pars interorbito- nasalis (Pls. 1, 2). 414 Professor Edward Faweett In the stage now described these parts are all fused together, but there is no difficulty in marking their limits, as will be done later. The central stem (Pls. 1, 2) asa whole stretches from the anterior margin of the foramen magnum to the tip of the nose. It is wide transversely behind, but narrows as traced forwards, until, when it becomes pars inter- orbito-nasalis, it becomes higher than wide. ' The Pars chordalis forms posteriorly the anterior margin of the foramen magnum, and at the same time part of each of the occipital condyles. Its posterior margin between these condyles is somewhat concave, the concavity being directed backwards. Its anterior margin is narrow transversely, and only separable from the hinder part of the pars trabecularis by the site of emergence of the notochord and by an imaginary line drawn transversely through the anterior margin of the basi-cochlear fissure. The lateral margins of the pars chordalis are divisible into two segments, viz. a longer posterior and a shorter anterior segment. The longer posterior segment is in part bounded by an imaginary antero-posterior line drawn through the inner margin of the foramen hypoglossi from condyle to jugular foramen. Along this line the lateral margin is directly continued into the exoccipital cartilage of its side; for a short distance, however, in front of this continuity the posterior segment of the lateral margin is free, and it here forms the medial boundary of the foramen jugulare. The anterior end of the posterior segment of the lateral boundary of the pars chordalis is indicated by a laterally projected process, the chordo-cochlear convmisswré, which, running out at the junction of posterior and anterior segments of the lateral border of the pars chordalis, fuses by procartilaginous tissue with the cochlear segment of the pars cochlearis of the auditory capsule. At the anterior margin of the base of this commissure the anterior segment of the lateral border of the pars chordalis commences. This parallel with its fellow forms the medial boundary of the basi-cochlear fissure. The anterior segments of the lateral boundaries are placed much nearer together than the two posterior segments, and on that account the anterior part of the pars chordalis, viz. that part in front of the chordo-cochlear commissures, is narrower than that behind them, ‘The surfaces of the pars chordalis are superior or caval and inferior. The superior surface of that part which lies behind the chordo-cochlear commissures is concave, that part in front of these commissures is flat. The inferior surface is convex posteriorly, as might perhaps be expected froin the condition of the superior surface of the corresponding region; whilst the inferior surface of that part in front of the chordo-cochlear commissures is concave from behind forwards, the concavity being bounded by a lateral ridge on each side. The chorda dorsalis (Pl. 1) has a somewhat interesting relation to the The Primordial Cranium of Pawcilophoca Weddelli 415 pars chordalis. As far forwards as the chordo-cochlear commissures it lies on its caval aspect, then for a short distance it sinks under the perichondrium. It again emerges at about the junction of the pars chordalis with the pars - trabecularis, on whose posterior surface it is placed as far as the summit of the crista transversa. Practically the part which lies under the peri- chondrium corresponds with that part of the pars chordalis which lies in front of the chordo-cochlear commissure, the part which many authors have described under the name of pars otica. The Pars trabecularis (Pls. 1, 2).—This is the region associated with the pituitary body. Its boundaries are determined by imaginary lines, both anteriorly and posteriorly. The posterior boundary is determined by an imaginary line drawn transversely through the anterior wall of the basi-cochlear fissures. The notochord too, emerging, as it does, at about the middle of this transverse line, separates in part the pars trabecularis from the pars chordalis (Pls. 1, 2). The anterior limit is not so easy to determine, but when careful examination is made with the microscope it will be found that the chondri- fied pars trabecularis can be traced forwards as a somewhat thin projec- tion under that region to be subsequently described as the region of the middle clinoid processes and there to end as cartilage. A transverse line drawn through the hinder roots of the ale orbitales will sufficiently delimit this part in front (Pls. 1,2). The lateral margins of the pars trabecularis extend from behind at the anterior wall of the basi-cochlear fissure to the hinder (post-optic) limb of the ala orbitalis. Each lateral margin at once, at its posterior extremity, gives outwards to the cochlear capsule a cartila- ginous bar—the posterior trabeculo-cochlear commissure. This commissure separates the basi-cochlear fissure behind from the foramen caroticwm in front, and it is of considerable antero-posterior width. From the middle of the lateral margin of the pars trabecularis another bar of cartilage—the anterior trabeculo-cochlear commisswre—arises, which, passing outwards and backwards in front of the foramen caroticum, completes this foramen by joining with the cochlear capsule some distance lateral to the posterior trabeculo-cochlear commissure. The remainder of the lateral border is free and forms the medial border of the spheno-parietal fontanelle, but a broad, thin lamina of fibrous tissue is attached which seems to serve as an attachment for part of some of the ocular muscles. The surfaces of the pars trabecularis are two in number, viz. a superior or caval and an inferior. The superior or caval surface is concave about its middle to form the pituitary fossa. This fossa is bounded behind by a somewhat conical crista transversa, on the middle of whose posterior aspect the chorda dorsalis lies. The fossa is bounded in front by a large 416 Professor Edward Fawcett transverse eminence which corresponds, so far as I can see, with the olivary eminence of the human skull, and this eminence at each lateral extremity is prolonged backwards in the form of a conical projection, which appar- ently is a middle clinoid process. The pituitary fossa is prolonged under the middle of the olivary eminence as a small deep pit. ‘There is no trace of a pharyngeal canal, and both olivary eminence and middle clinoid ‘ processes are at this time in a procartilagium condition resting on the upper aspect of the chondrified part of the pars trabecularis. Whether they really belong to the pars trabecularis or are to be regarded as a backwardly thrust part of the pars interorbito-nasalis I must leave an open question, as younger stages are necessary in order to settle the point. For convenience we may at present regard them as belonging to the pars trabecularis, and regard this structure as terminating at the imaginary line above alluded to as drawn transversely through the hinder margin of each post-optic limb of the alz orbitales. The inferior surface of the pars trabecularis is marked in the median antero-posterior line by a slight keel- like projection, which is continuous anteriorly with a similar projection on the inferior aspect of the interorbital part of the pars interorbito-nasalis, The Pars interorbito-nasalis (Pls. 1, 2) may, with the reservation above’ mentioned, be considered as commencing at the imaginary transverse line drawn through the hinder margin of the post-optic limbs of the ale orbitales. It is at first of considerable width, because it sends outwards on each side a wing-like extension which is the so-called ala hypochiasmata (Pl. 2). This ala at this stage is for the most part in a procartilaginous con- dition, and it is joined on the upper aspect of its lateral terminal part by the already well chondrified post-optic limb of the ala orbitalis. This limb in individual sections presents the appearance of a separately chondrified nucleus, but successive sections modelled show that it is really part of the post-optic limb of the ala orbitalis, and it is from this that the rectus muscles of the eyeball take their origin. In front of the alee hypochiasmate the pars interorbito-nasalis is fused with the posterior branch of the pre- optie limb of the ala orbitalis (Pls, 1, 2), and so completely fused is it at this stage, that it is quite impossible to say where the line of fusion was. Between the two optic limbs of the ala orbitalis lies the foramen opticum (Pls. 1, 2), which is of considerable size, and its antero-median margin is projected outward for some distance to form a muscular process from which the superior oblique muscle of the eyeball arises, Anterior to the posterior branch of the pre-optic limb of the ala orbitalis the lateral margin of the interorbital part of the pars interorbito-nasalis becomes free, and it bounds medially a fissure, the foramen prwchiasmaticum (Pl. 1) of Macklin, a foramen which in the seal is a fissure inclined obliquely from behind The Primordial Cranium of Pecilophoca Weddelli 417 forwards and inwards until it nearly meets its fellow anteriorly. Macklin’s view is that this foramen is formed from the foramen opticum, and in a sense that may be true, but it is certainly more immediately formed here by thé development of a bar of cartilage, which, growing forwards from the front margin of the pre-optie limb of the ala orbitalis, reaches and fuses with the hinder part of the nasal septal part of the pars interorbito- nasalis. (In the cat and ferret this bar of cartilage does not reach the septum nasi, but fuses with the dorsal edge of the cupula nasi posterior.) The part of the pars interorbito-nasalis which has just been described is that part usually alluded to as the interorbital septum, but with the addition of the hinder: element, to which are appended on each side the ale orbitales. Its upper surface between the foramina optica is slightly concave from side to side, and was termed by Voit the lamina hypochiasmata, and it is prolonged outwards into the posterio-median part of the foramen opticum as the ala hypochiasmata. It is slightly concave from before backwards. In front of the lamina hypochiasmata the upper surface is of obscure extent laterally, seeing that it is fused with the anterior limb (pre-optic) of the ola orbitalis; but in front of this, where the foramina prechiasmata exist, it is of triangular form, whose apex is forwards. The inferior surface of the interorbital part of the pars interorbito-nasalis is marked by an antero-posterior keel-like ridge, which is continuous posteriorly with the ridge already mentioned on the pars trabecularis, and anteriorly ‘with the median part of the lower edge of the nasal septal part of the pars interorbito-nasalis, to which it ‘is bent at an angle of something like 130°, open below. Intrinsically the interorbital septum is comparatively narrow, as is determined by the outline of its chondrified part, which, some- what triangular in coronal section behind, changes anteriorly to a pear- shaped outline with large end downwards. The pars nasalis of the pars interorbito-nasalis is the septum nasi, and at the same time that part of the pars interorbito-nasalis which is included within the nasal capsules. Its detailed description may therefore be with advantage deferred until the nasal capsule as a whole is dealt with. No more need be said here than that it may be divided into two parts, viz. a subcerebral and a precerebral part. 2. STRUCTURES APPENDED TO THE Cures. STEM. These are from behind forwards :— (a) The exoccipital cartilages ; (b) The auditory capsules ; (c) The alee temporales ; (d) The ale orbitales ; (¢) The lateral parts of the nasal capsule, 418 Professor Edward Fawcett (a) The Exoccipital Cartilages. Each of these projects laterally from opposite sides of the hinder part of the lateral border (as defined above) of the pars chordalis of the central stem. Each is developed by direct outgrowth from the central stem, and, as said above, may only be artificially distinguished from that central stem by an imaginary line drawn antero-posteriorly through the medial margin of the hypoglossal canal. Each then starts by two roots, one in front of and the other behind the hypoglossal canal, and the hinder root bears the | condyle on its lower and hinder aspect. The two roots blend lateral to the hypoglossal canal and form a common mass, which for a short distance passes directly outwards as the so-called laming alaris. This lamina alaris separates the auditory capsule from the foramen magnum. Its anterior border near its commencement shoots forwards under the infero- medial aspect of the pars canalicularis of the auditory capsule to form a jugular process, which shows a conical, somewhat downwardly directed, projection, the paracondyloid process. The jugular process is attached to the pars canalicularis by dense cellular tissue, but there is no cartilaginous union. Its upper surface is grooved by the lower end of the sinus venosus lateralis. Lateral to the jugular process the anterior border of the lamina alaris is bevelled from behind downwards and forwards and placed under the infero-medial surface of the. pars canalicularis, from which it is separated by a comparatively wide but gradually diminishing interval, the recessus supra-alaris, This interval gradually diminishes when traced from cavum cranii to the exterior, and is ultimately reduced to a mere fissure, the fisswra supra-alaris, or better, the fisswra exoccipito-capsularis. The posterior border of the lamina alaris is sharp, and forms a part of the lateral boundary of the foramen magnum. The upper surface of the lamina alaris is grooved from behind forwards and inwards by the sinus venosus lateralis, but the inferior surface shows a general broad concavity which serves for the attachment of the rectus capitis lateralis muscle. The lamina alaris is succeeded by the terminal part of the exoccipital cartilage, which is directed backwards in a curve whose convexity is outwards and concavity directed inwards to bound laterally the middle third or so of the foramen magnum. It terminates posteriorly opposite a line drawn transversely through a point just behind the middle of the foramen magnum in a pointed extremity. The outer convex margin of this segment of the exoccipital cartilage is overlapped by the supraoccipital cartilage, and a fissure filled with con- nective tissue separates the two. It is only near the extreme posterior end of this border that any cartilaginous fusion exists between exoccipital and le eed i The Primordial Cranium of Pacilophoca Weddelli 419 supraoccipital cartilages, so that the exoccipital preserves its independence to a remarkable degree at a comparatively late period. The medial border, as has already been stated, forms part of the lateral boundary of the foramen magnum. The upper surface of this segment is crossed in its antero-lateral part by the sinus venosus lateralis. A few words may now be said with regard to the primary foramen magnum. It is of large size, and bounded completely by cartilage. In general shape it may be described as that of a lozenge; its general direction is backwards, but its anterior half has a slight downward inclina- tion. Being lozenge shaped, the foramen may be said to possess four angles, viz. an anterior, a posterior, and two lateral. The anterior angle is obtuse, and occupied in its middle by the chorda dorsalis; it corresponds with the incisura anterior, which is so well marked in some other animals. This anterior angle is formed at the expense of the pars chordalis of the central stem. The posterior angle is very obtuse, and formed at the expense of the conjoined supraoccipital cartilages (tectum cranii posterius). The lateral angles are the most pronounced, and are acute; they are formed between the posterior pointed extremity of the exoccipital cartilage and the supraoccipital cartilage of the corresponding side. The boundaries of the foramen are antero-lateral and postero-lateral. Each antero-lateral boundary is formed from before backwards by, first, the pars chordalis _ and, second, the exoccipital cartilage. The postero-lateral boundary is formed by the supraoccipital cartilage alone. (b) The Auditory Capsules (Pls. 1, 2, 3, 4, 5, 6). _ Each auditory capsule is at this stage in a somewhat imperfect con- dition owing to the youth of the stage examined, but in general features resembles that found generally in mammals. Kach is roughly pyramidal in form, with base outwards on the lateral surface of the chondrocranium, and with apex directed towards the central stem. The long axis of the whole is placed very nearly at right angles with that of the central stem. Each consists of two main subdivisions, viz. a pars canalicularis, the more lateral, and a paz's cochleuris, the more medial, and the latter (following Voit) is capable of further subdivision into a “vestibular” and a “cochlear ” segment. The pars canalicularis has but slight attachment to neighbouring parts of the chondrocranium, and that to the lateral wall by means of the 420 Professor. Edward Fawcett lamina parietalis. The cochlear segment of the pars cochlearis, on the other hand, is moored to the central stem by three distinct commissures, two of which connect it with the pars trabecularis, and are the anterior and posterior trabeculo-cochlear commissures, and the third, which it receives ‘from the pars chordalis of the central stem, is the chordo-cochlear commissure. The Pars canalicularis is that part which more especially lodges the semicircular canals. It represents the general basal part of the pyra- midal auditory capsule, and appears on the lateral aspect of the chondro- cranium. . It may be regarded as having four surfaces, viz. a lateral or basal, which is seen on the lateral aspect of the chondrocranium; an anterior, which forms the hinder wall of the primary tympanic cavity; a swpero- medial, which forms a part of the floor of the cavum cranii, and is characterised by the presence of a large hollow in its middle—the jfossa subarcuata posterior interna (Pl. 5); an infero-medial, which is directed downwards and backwards towards the lamina alaris and processus jugularis of the exoccipital cartilage (Pls. 1, 5). These surfaces may be examined in the order above named. The lateral surface is roughly triangular in form, with base forwards and apex backwards, the latter, the so-called cwpula, being hidden from external view by the opercular process of the supraoccipital cartilage. This surface has the following borders: an anterior or basal, a superior, and an inferior; the two latter converge posteriorly, and meet at the cupula. The anterior or basal border (Pls. 3, 4, 6) is almost vertical, and shows | from above downwards, first, a slight convexity, which I take to be the early condition of a teymen tympani; below this is a hollow, which, because it lodges the crus breve of the incus cartilage, is the fossa incudis; this is succeeded at a somewhat medial plane by a slight projection, which, since it gives attachment to the processus styloideus, is identified as the erista parotica, The lower end of the anterior border merges into the anterior end of the inferior border at a rounded angle, which from the muscles it gives attachment to must be regarded as the processus mastoideus. The inferior border (Pls. 3, 4) commences in front at the processus mastoideus, and terminates behind at the cupula, In its anterior half it is almost straight and horizontal, but in its posterior half it rises somewhat rapidly, to terminate at the cupula. It is semicylindrical in form, and corresponds in the interior with the posterior semicircular canal, hence it is named the prominentia semicircularis posterior, The medial side of this segment of the inferior border to within a short distance of the The Primordial Cranium of Pwcilophoca Weddelli 421 ‘ cupula is in relation with the jugular process and lamina alaris of the ex- occipital cartilage from before backwards, but is actually separated from them by the fissura supra-alaris, or fissura exoccipito- -capsularis ; the hinder part of this segment of the inferior border is overlapped from without by the processus opercularis of the supraoccipital cartilage. The superior border (Pls.*3, 4) commences in front at the tegmen tympani, and terminates behind at the cupula. It, like the inferior border, may be divided into an anterior and a posterior segment. The former ’ gives attachment in its whole length to the lamina parietal is; the latter, which slopes downwards and backwards to the cupula, is free, forms the lower boundary of the supraoccipito-capsular fissure, is semicylindrical, and caused by the anterior semicircular canal. It is therefore the pro- minentia semicircularis anterior. The Lateral surface (Pls. 3, 4), which is enclosed by these boundaries, is on the whole convex, but about its centre shows a. shallow depression, which is bounded above and behind by the prominentia semicircularis anterior, below and behind by the prominentia semicircularis posterior, below and in front by a prominence which, rising up from about the junction of the anterior and posterior segments of the lower border, reaches the basal border just lateral to the crista parotica. This prominence is caused by the lateral semicircular canal, and is termed the prominentia semicircularis lateralis. The shallow depression enclosed by these prominences may be termed the fossa subarcuatu lateralis. The Anterior surface of the pars canalicularis (P1. 6) forms the posterior wall of the primary tympanic cavity. It is narrow from side to side, but deep from above downwards. Its upper part is slightly hollowed to receive the body of the incus cartilage, and is wider than the lower part. The lower part is grooved, more laterally to lodge the facial nerve, and more medially to lodge and give origin to the stapedius muscle, which is of large size. The groove for the stapedius is deeper than that for the facial nerve. The Supero- medial surface of the pars canalicularis (Pls. 1, 5) is of large size, is inclined mainly upwards, but with a slight tilt medialwards and backwards. It is easily recognised by the large central depression, called from being arched over above by the prominentia semicircularis anterior the fossa subarcuata anterior, and from its internal position relative to the pars canalicularis, and to distinguish it from the one on the lateral aspect above described, the fossa subarcwata anterior interna. The supero-medial surface is bounded above in front and behind by the prominentia semicircularis anterior, which anteriorly swells out to form a prominentia utricula-ampullaris anterior. In this prominence a well- 422 Professor Edward Fawcett marked transverse ridge is present, which recalls a similar condition in the cat. This in the cat is the outward prolongation of the commissura facialis, but as no such commissure exists in this specimen at the stage modelled, perhaps the ridge may represent an early stage in its formation (if it ever do be formed). Owing to incompleteness of chondrification, the promi- nentia utriculo-ampullaris superior is invaded by an angular extension of the meatus auditorius internus, through which the ampulla of the mem- branous anterior semicircular canal can be seen. Below and behind, the supero-medial surface of the pars canalicularis is bounded by a well-marked prominence which lodges in its interior the crus commune of the anterior and posterior semicircular canals, and is called the prominentia cruris, communis. This prominence posteriorly ends at the cupula, and anteriorly swells out medial to and below the fossa subarcuata anterior interna to form a well-marked medial boundary to the said fossa. From the lower basal angle of the meatus auditorius internus a slit-like extension of that vacuity extends into the anterior terminal part of the prominentia cruris communis. This slit-like extension is separated on the right side from a foramen—the foramen endolym- phaticwm, which transmits the ductus endolymphaticus—by a small bar of cartilage, but on the left side this bar has not yet chondrified. The ductus endolymphaticus, which emerges from the foramen endolymphaticum, is a long, at first cylindrical, narrow tube, which after emergence runs alongside the prominentia cruris communis, at first lodged in a groove thereon, but soon it loses immediate contact with the cartilage, and comes to be medial to the sinus venosus lateralis, and is continued backwards as a wide sac-like structure which terminates in a narrow-pointed extremity medial to the supraoccipital cartilage. No pressure vacuity exists in the pars canalicularis, which is rather exceptional, and doubtless to be explained by the want of close relationship between the hinder part of the duct and the cartilage. The Infero-medial surface of the pars canalicularis is triangular in form, and is directed downwards and backwards towards the jugular process and lamina alaris of the exoccipital cartilage, from which it is separated by the recessus supra-alaris. The apical region of this surface reaches outwards almost as far as the cupula, its basal region reaches medialwards as far as the foramen jugulare. It is bounded above by the prominentia cruris communis, below by the prominentia semicircularis posterior. About its middle there is a deep pit, which burrows into the pars canalicularis medial to the lower end of the prominentia semicircularis posterior; this pit is the fossa swbarcuata posterior (P1. 5). Basalwards of this fossa the infero-medial surface swells out to form the promimentia utriculo-ampullaris posterior (Pl. 5). The Primordial Cranium of Pacilophoca Weddelli 423 The Pars cochlearis (Pls. 1, 2, 5, 6), which is the antero-medial con- tinuation of the auditory capsule, consists as before said of a “ vestibular” and a “cochlear” segment. The “ Vestibular” segment is that part especially which shows on its caval aspect the meatus auditorius internus (Pls. 1, 5), and on its lateral aspect the foramina vestibuli and cochlee. It is interposed between the pars canalicularis and the cochlear segment of the pars cochlearis. It presents to view four surfaces, of which one is superior, one medial, one inferior, and the remaining one lateral. The swperior surface is small, and grooved to form the upper part of the sulcus facialis (Pls. 1, 5), on which the germ of the facial nerve rests, Whether it is ever bridged over by a cartilaginous suprafacial commissure I do not know, as no embryo old enough to determine that point is at my disposal. . The medial surface (Pls. 1, 5) is almost wholly occupied by the meatus auditorius internus, which, no doubt owing to the young stage modelled, is very large, and not divided into what may be presumed to be its subsequent compartments. The opening is triangular in form, and if the base of the triangle be regarded as. its outer side, then each basal angle is projected backwards and outwards in the form of a narrow fissure which, in the case of the upper basal angle, leads to the prominentia utriculo-ampullaris anterior, and in the case of the lower basal angle to the prominentia cruris communis, and at the same time transmits the ductus endolymphaticus on the left side. In Plate 5 the relations of the membranous to the meatus auditorius internus are shown, and it will be noticed that the saccule, the commencement and the terminal coil of the cochlear duct, are visible, also that the ductus endolymphaticus leaves the posterior inferior basal angular fissure above mentioned. The inferior surface (Pls. 2, 5) is small, and directed towards the foramen jugulare. It is almost wholly occupied by the conjoined foramina cochlee and perilymphaticum. The lateral surface (Pl. 6) forms the medial wall of the primitive tympanic cavity, and from above downwards shows first the sulcus facialis next the foramen vestibuli (fenestra ovalis) which is occupied by the foot of the stapes, finally the promontorium caused by the initial part of the basal coil of the cochlea. Above and in front of the foramen vestibuli a flattened area is caused by the very large tensor tympani muscle. The “ Cochlear” segment of the pars cochlearis (Pls. 1, 2, 5, 6) is some- what small in size; this is markedly the case in comparison with that of the cat (Pl. 11), but it is on the whole perhaps relatively larger than 424 Professor Edward Faweett that of the ferret (Pl. 10). It is moored to neighbouring parts in the following way: viz. postero-medially to the central stem by the small chordo-cochlear commissure, to the pars trabecularis by the anterior and posterior trabeculo-cochlear commissures. It is separated from the central stem by the large basi-cochlear fissure, and from the pars trabecularis by the foramen caroticum. Its surfaces may be described as medial or caval, inferior and antero-inferior. The medial surface is roughly triangular in shape, and at the apex is directly continuous with the caval surface of the posterior trabeculo-cochlear commissure. This surface is comparatively flat, in which respect it contrasts markedly with the condition in the cat, in which it bulges into the cavum cranii to come into pretty close relation with that of the opposite side. Me The antero-inferior surface is the largest, and is convex in all directions. It is crossed from below upwards in an antero-posterior direction by the internal carotid artery, but possibly on account of the small size of that artery no sulcus caroticus is visible upon it. To the upper part of this surface the anterior trabeculo-cochlear commissure is attached. The inferior surface is small, and indented by the common foramen cochlee and foramen perilymphaticum. A commissura suprafacialis. does not exist at this stage, but there seems to be a small spur on the medial side of the sulcus facialis of the vestibular segment of the pars cochlearis which may represent it. (c) The Ala temporalis (Pls. 1, 2). This, which is developed separately, is of comparatively small size, is thin and flat, and placed in front of the middle of the anterior trabeculo- cochlear commissure, from which it is separated by dense cellular tissue, which in appearance is very much like that intervening between the cartilages and long bones at a joint. The lateral border of the ala temporalis is notched for the downward passage of the mandibular division of the fifth cranial nerve, and by this notch one may divide the ala into an anterior and a posterior segment. On the caval surface of the ala the Gasserian ganglion, the maxillary and ophthalmic divisions of the fifth cranial nerve rest. The inferior surface, which is not quite so flat as the superior, shows signs of a median antero-posterior rounded ridge, which is the representative of the processus pterygoideus of the ordinary ala temporalis, The inferior surface is subtended by a large wedge-shaped mass of connective tissue, from whose apex the pterygoideus internus muscle arises, and from whose infero-lateral surface the lower (sole) head of the pterygoideus externo arises. } | The Primordial Cranium of Pecilophoca Weddelli 425 (d) The ‘Ala orbitalis (Pls. 1, 2, 3, 4). This structure, which belongs also to the lateral structures, is of large size, and forms a very important constituent in the floor of the cavum cranii. It may be best described as triangular in form, with its apex directed towards the central stem. Its basal angles are placed in the lateral wall of the cavum cranii. The body or main mass forms at once a part of the floor of the cavum cranii and the roof of the orbital cavity. Its anterior border forms the posterior boundary of the orbito-nasal fissure, and also, after fusion with the hinder part of the nasal capsule, forms a projection over the nasal cavity, of which more will be said later. The posterior border of the ala forms the anterior boundary of the spheno-parietal fontanelle, and near the central stem develops a backwardly directed process which has all the appearance of an anterior clinoid process. The basal border forms part of the brim or rim of the cavum cranii. The angles of the ala, which are an apical, an anterior, and a posterior basal angle, are of importance and interest. The apical angle is curious, inas- much as it divides primarily into a pre- and a post-optic limb, of which the former again divides into two. The post-optic limb from the processus clinoideus anterior turns suddenly downwards and inwards towards the ala hypochiasmata of the central stem and fuses with its outer extremity, which at this time consists of procartilage. It is from this narrow, downwardly directed part of the post-optic limb that the eye muscles, save the superior oblique, take origin, and from its advanced state of chondrification as compared with that of the ala hypochiasmata one can easily distinguish the two. The pre-optic limb forms the anterior wall of the foramen opticum and then divides into two branches, of which the more posterior is fused with the central stem (interorbital part of it), whilst the more anterior is attached to the hinder part of the subcerebral segment of the septum nasi. Between these two branches a long oblique fissure bounded medially by the upper edge of the interorbital septum is found. A somewhat similar fissure is described by Macklin in man under the name foramen prechiasmaticwm, and he regards it as being a cut-off part of the foramen opticum. From the condition here and that of the ferret (Pl. 10) and cat (Pl. 11) I prefer to think of it as an isolated part of the orbito-nasal fissure, which isolation has been brought about by the fusion of the anterior branch of the pre- optic limb of the ala orbitalis with the nasal capsule, and the further exten- sion of that branch to the septum nasi. That in a simpler condition is seen in the ferret, in which the anterior branch of the pre-optic limb only reaches the cupula nasi, thus separating the orbito-nasal fissure into a median and VOL. LI. (THIRD SER. VOL. XIII.)—JULY 1918. 29 426 Professor Edward Fawcett _ a lateral part, but it does not extend so far inwards as the septum nasi. Not only does this anterior branch of the pre-optic limb of the ala orbitalis cut off this fissure, but it at the same time forms a roof for the hinder part of the nasal cavity, a roof which, however, is perhaps but a temporary one. The anterior basal angle of the ala orbitalis fuses with the frontal prominence of the pars intermedia of the paries nasi to form the spheno- ethmoidal commisswre and so bound the orbito-nasal fissure laterally. It is covered on its outer aspect by the frontal bone. The posterior basal angle is fused with the forward extension of the lamina parietalis, forming in this way the orbito-parietal commissure. Not far from the middle of the ala orbitalis two foramina are found; a few small blood-vessels_ pass through one of them, the other does not seem to transmit anything: A similar condition is to be observed in Tatusia; may they be representatives of the reptilian foramen epiopticum ? (e) The Lateral Wall of the Nasal Capsule. This structure may be conveniently left until the nasal region as a whole is described. 3. COMMISSURES CONNECTING THE CENTRAL STEM WITH THE APPENDAGES (Pls. 1, 2). : These, which are on each side from behind forwards, the chordo-cochlear, the posterior and anterior trabeculo-cochlear commissures, have been sufficiently alluded to already, and do not require further description. 4, LATERAL STRUCTURES (Pls. 1, 2, 3, 4). The structures which may be grouped under this heading are :— 1, The supraoccipital cartilages, 2. The laminz parietales (parietal plates), 3. The ale orbitales. — |. The Swpraoccipital cartilage (Pls. 1,3, 4) is primarily a separate entity, which rapidly assumes a triangular form. The angles are an anterior, an inferior, and a postero-median. The anterior angle is fused with the lamina parietalis, forming what may be called a commisswra swpra- occipito-parietalis, between whose inferior margin and’ the pars canali- cularis of the auditory capsule is the supraoccipito capsular fissure. Its inferior angle forms a processus opercularis, which overlaps on its outer side both the hinder end of the pars canalicularis and the outer edge of the exoccipital cartilage. It is quite independent of the former, and is : 7 ee ee The Primordial Cranium of Pacilophoca Weddelli 427 only very slightly fused with the apical region of the latter. The postero- median angle projects over the back part of the cavum cranii and fuses with its fellow to form the tectum cranii proterius, a structure which at the same time forms the upper margin of the primary foramen magnum. 2. The Lamina parietalis (Pls. 3, 4, 5)—This, otherwise known as the parietal plate, springs from the anterior half of the upper margin of the pars canalicularis of the auditory capsule. Its hinder margin is at first for a short distance free where it forms the anterior margin of the supraoccipito- capsular fissure, but the greater part of this border is blended with the anterior angle of the supraoccipital cartilage to form the supraoccipito- parietal commissure. Anteriorly the lamina parietalis projects forwards as a more or less sagittally directed plate forming the upper border of the spheno-parietal fontanelle, and ends by joining the posterior basal angle of the ala orbitalis to form the orbito-parietal commissure. From the upper margin of the lamina parietalis an angular plate projects, which has attached to it a somewhat narrow coronally directed bar of cartilage, which would seem to be the lower part of a tectum cranii anterius, seeing that at the ' vertex, and still in the same coronal plane as this, are to be found two coronally placed cartilages of about the same antero-posterior length as the lower bar just mentioned, which must be the upper parts of a tectum eranii anterius. It is possible that these cartilages occupy the neighbour- hood of the future bregma, although the state of ossification is not nearly far enough advanced here to enable one to more than surmise that such is the case; whatever their position with reference to the roof of the older cranium may be, their very forward position in this animal is very striking, and is new to me. 3. The Ale orbitales.—These having been already described with the appendages to the central stem need not be alluded to further. 5. THE LATERAL CoMMISSURES (Pls. 1, 3, 4). These from behind forwards are the swpraoccipito-parietal, the orbito- parietal, and the spheno-ethmoidal. The position and mode of formation of each of these has perhaps been sufficiently indicated already above, and, taken together with the structures which they connect, they form the greater part of what may be termed the “rim” of the cranial cavity. This rim is formed as to one-half from before backwards by the anterior margin of the subcerebral vacuity of _ the roof of the nasal capsule by the upper margin of the frontal prominence of the pars intermedia of the nasal capsule, by the spheno-ethmoidal com- missure, the upper edges of the ala orbitalis, the orbito-parietal commissure, 428 Professor Edward: Faweett . the lamina parietalis, the supraoccipito-parietal commissure, the upper . edge of the supraoccipital cartilage, and, finally, the upper edge of the corre- sponding half of the supraoccipital cartilage. By the ensemble of these, what I have termed the “rim” of the cavum cranii is formed. 6. DorsAL STRUCTURES. These are the structures which lie dorsal to the cavum cranii, and may be termed tectum cranii anterius and tectum cranii posterius. The Tectwm cranii anterius (Pls. 3, 4) seems to be composed of four cartilages, two of which lie laterally and project up from that part of the “rim” of the cavum cranii which is composed of lamina parietalis and orbito- parietal commissure. Each lateral segment is only here and there attached to the lamina parietalis by continuous cartilage; for the most part it is separated therefrom by fissures filled with connective tissue, and probably has an independent origin. The remaining cartilages which enter into the composition of the tectum cranii anterius are two, placed one on each side of the middle line in the same coronal plane as those just described, and - showing slight signs of fusion with one another. Each is of irregularly quadrilateral form, and its long axis is directed coronally. What the ultimate fate of this tectum is I do not know, nor do I know if the several > parts of which it is composed remain separate or fuse together, Its far forward position, situated as it is anteriorly to the coronal plane of the fossa hypophyseos, is very striking. The Tectum cranii posterius (Pls. 1, 3, 4)i is situated at the back of the cavum cranii, and forms by its lower margin the dorsal edge of the primary foramen magnum. It is formed by the union. of the prolonged postero- median angles of the supraoccipital cartilages. The Nasal Capsule. The nasal capsule consists of a septum and lateral parts appended to it. Like that of the other carnivora which I have modelled, viz. the cat and ferret, it is comparatively short and squat, approaching more nearly the human type than that of other quadruped animals, but it consists essentially of the same divisions, The septwm nasi (Pls. 2,7), which is the forward continuation of the central stem, consists of two parts, viz. a subcerebral and a precerebral. The antero-posterior length of the former is about one-third that of the latter. Its height at its commencement behind is very little, but it rapidly increases in height when traced forwards, until at its anterior end this part of the septum is the highest of all. As the lower border of this segment act a * ee : 7 a a a a The Primordial Cranium of Pwcilophoca Weddelli 429 remains practically constant in level, it follows that the increase in height is due to a gradual rising of the upper border, and this rise is culminated at the crista galli, which means the anterior limit of the upper border of the subcerebral part of the nasal septum. The lower border of the sub- cerebral part of the septum is wider than the upper border. It is convex from side to side, and thicker behind than in front; the median part of its general convexity is continuous with the keel-like ridge on the under aspect of the interorbital part of the interorbito-nasal septum, and that in turn with the similar ridge on the under surface of the pars trabecularis. This part of the under border is bent at an angle of about 130 degrees with that part of the central stem which lies posterior to it. The upper margin of the subeerebral segment of the nasal septum is thinner than the lower one, and, as before said relative to the lower border, rises from it as it is traced forwards to the crista galli. Actually the plane of this upper border is in continuity with that of the remainder of the caval surface of the central stem behind it; it therefore forms an angle with the lower border, and this angle amounts to about 70 degrees. ‘To the hinder part of this subcerebral segment of the upper border the anterior secondary branch of the pre-optic limb of the apical angle of the ala orbitalis is attached. ‘ The precerebral part of the septum nasi is at its hinder end almost of the same height as the subcerebral segment, but its upper border somewhat suddenly descends for a short distance, after which it runs forwards practically parallel with the lower border. The upper border of this segment gives off on each side that part of the capsule which forms the roof of the corresponding nasal passage, and as the roof arches outwards it follows that the upper margin of this segment of the septum is placed at the floor of a sulcus dorsalis nasi, which extends almost from the front of the crista galli to the tip of the nose. The lower margin runs forwards, continuing the direction of the lower border of the subcerebral part of the septum and maintaining the thickness of the latter. Opposite its middle third is placed on each side an anterior paraseptal cartilage, which is separated from the septum itself by a septo-paraseptal fissure; anteriorly, however, this paraseptal cartilage, which will be described more fully later, is continuous by its antero-superior limb with that lateral offshoot of the inferior border of the septum called the processus lateralis ventralis. The anterior third of the lower border of the precerebral segment appears to bifurcate into two laterally running lamine (each of which is a processus lateralis ventralis) almost until its most anterior limit is reached, when it becomes single and undivided, projecting suddenly downwards before its termination in the form ofa peg-like process. There is no internarial 430 Professor Edward Fawcett foramen in the septum, nor is there any organ of Jacobson in which the Phocide agree with the Cetacea. The Lateral part of the capsule (Pls. 3, 4, 8) consists of a roof, lateral wall, and a floor. Of these the last is very incomplete. The Roof i is narrow from side to side, and is attached medially to the upper margin of the precerebral part of the septum nasi, and it is to be clearly understood that what is now being described is the roof of the precerebral part of the nasal capsule, in other words, the tectum nasi anterius, for there is at this time no intrinsic cartilaginous roof to the subcerebral part. The roof of one narial passage is separated from that of the other by a sulcus dorsalis nasi, whilst laterally it almost insensibly continues over into the lateral wall; not far from its anterior end it is perforated by a small foramen dorsale. The anterior end forms the upper margin of the fenestra narina, and there is developed at its outer end a small processus cwpularis. Postero-laterally the roof is separated from the remainder of the lateral part of the nasal tapsule by the sulews antero- lateralis, at whose upper end a large foramen epiphaniale is developed for transmission of the lateral branch of the nasal nerve. The Lateral wall of the nasal capsule as seen from the outside (Pls. 3, 4, 8) is divisible into three parts, viz. a pars anterior, & pars — intermedia, and a pars posterior, The pars anterior is comparatively short, being about the equal in antero- -posterior length of the pars intermedia and the pars posterior taken together. It is deeper behind than in front, has an upper rounded border by which it is continuous with the roof, an inferior border which is free in its whole length, and an anterior border which is short and lies almost at right angles to the upper and lower borders. Posteriorly the pars anterior diverges somewhat from the middle line before merging with the pars intermedia at the sulcus antero-lateralis. The in- ferior border commences anteriorly at the angle of junction with the anterior border, which angle projects very slightly downwards to form a processus alaris swperior ; behind this, at a short distance, a slight notch is seen, which perhaps may be regarded as an inciswra pre-transversalis, behind which a downward projection, which is probably the outer end of an incomplete lamina transversalis anterior, is found ; this is sueceeded by a slight inciswra pro-transversalis. The remainder of the inferior border is practically straight, and is the maxillo-turbinal. The whole of the lateral wall of the pars anterior is lower in height than the septum, hence that structure is exposed to view when the nasal sac is removed; if that be left in position, then its lateral wall is visible below the lower border of the pars anterior (Pl. 8), The pars intermedia is of large size, lateral projection, and depth, is The Primordial Cranium of Pwcilophoca Weddelli 431 easily separable from the pars’ anterior by the sulcus antero-lateralis, but its separation at first sight is not so obvious from the pars posterior, because the sulcus lateralis posterior is placed so high up and runs a course almost parallel with the outer margin of the subcerebral vacuity of the nasal capsule. . The sulcus antero-lateralis is easily found, its commencement above being sufficiently indicated by the large foramen epiphaniale, below which it extends downwards under cover of a projecting hood of cartilage as a deep groove; below this it broadens out into a broad triangular hollow, whose deepest part reaches the inferior margin of the lateral wall of the nasal capsule. The sulcus postero-lateralis commences above and in front at the root of the spheno-ethmoidal commissure, and passes backwards parallel with the outer margin of the subcerebral vacuity and only a short distance below it. Each of these sulci has a corresponding ridge on the medial aspect of the lateral wall; thus the upper deep part of the sulcus antero- lateralis corresponds precisely with the crista semicircularis, whilst the sulcus postero-lateralis corresponds with the anterior root of the first ethmo-turbinal and with a ridge common to the spring of the second ethmo-turbinal and the lamina transversalis posterior. The Pars intermedia which lies between these sulci is deep, prominent, and large, wider from before backwards below than above. It may be divided into an upper prominence and a lower one. ‘The upper prominence is the prominentia frontalis, so called because it is covered by the frontal bone (Pl. 4). It projects forwards as a sort of hood over the foramen epi- phaniale and the deep upper part of the sulcus antero-lateralis. Posteriorly it is connected by the spheno-ethmoidal commissure with the anterior basal angle of the ala orbitalis, inferiorly it is separated by a slight sulcus from the prominentia mavillaris. The reverse side of the prominentia frontalis is the recessus frontalis. The prominentia maxillaris is of very large size, and divisible into two surfaces, viz. an anterior and a posterior, of which the latter forms the medial wall of the orbit as the planum antorbitale; at the lower part of the ridge separating these surfaces, the inferior oblique muscle of the eyeball takes origin. The inferior margin of the pars inter- media, in its arfterior half at least, forms the hinder part of the maxillo- turbinal. The prominentia maxillaris of the exterior corresponds with the recessus maxillaris of the interior. The Pars posterior (Pls. 2, 8) is, as appears to be common to the seal tribe, of very small size. It is certainly long from before backwards, but narrow from above downwards. It is marked off from the pars intermedia below and in front of it by the sulcus lateralis posterior, _and it ends posteriorly at the cupula nasi posterior, which is fused 432 Professor Edward Fawcett with the under aspect of the anterior branch of the pre-optic limb of the ala orbitalis. : When the nasal capsule as a whole is viewed from above it is seen to consist of two very distinct parts, which are a subcerebral posterior part and a precerebral anterior part. ‘The subcerebral part is divided mesially by the upper edge of the septum nasi, and embraces the interior of the _ pars posterior as well as the recessus frontalis of the pars intermedia. There is no tectum to this part save that formed by the anterior branch of the pre-optic limb of the ala orbitalis, as the lamina cribrosa has as yet not — developed. The plane of the upper aspect of the subcerebral part of the nasal - capsule is practically that of the rest of the floor of the cavum cranii. Its great width as compared with the greater part of the precerebral segment is very striking; moreover, its antero-posterior length too in comparison with that of the latter is at once noticeable. The precerebral part is wide behind, but narrow in front. It is inclined downwards at a considerable angle with the subcerebral part. This angle or downward inclination is at first somewhat sudden, but later much less. A median dorsal nasal sulcus runs from the front almost the whole length of the dorsal aspect of the precerebral part of the nasal capsule, and at the bottom of this sulcus lies the septum nasi. Towards the anterior part of the precerebral segment one notices on each side of the middle line, and at a little distance lateral to it, the foramen dorsale. The Solum nasi (Pl. 2).—This is very imperfect, but extremely iiteraaed ing, because it, at this stage at any rate, seems to show a condition closely approximating to that of man. The lamina tranversalis posterior fails to reach the immediate neighbourhood of the septum. There are no posterior paraseptal cartilages. These, however, may possibly develop at a later stage. : The lamina transversalis anterior is wanting in its middle segment, being present laterally and medially only. Laterally it exists in the form of a downward projection from the lower margin of the lateral wall of the nasal capsule, as has already been mentioned. Medially it forms that part of the processus lateralis ventralis which lies immediately in front of the anterior paraseptal cartilage. On account of the absence of the middle part of the lamina transversalis anterior the fenestra narina is continuous with the fenestra basalis as in man, a rostro-ventral fisswre being formed. The anterior paraseptal cartilage is moderately well developed, but consists of a single lamella only as in man, because there is no organ of Jacobson to be lodged in and supported by it. In man the organ of Jacobson, as is well known, lies against the septum nasi at some distance The Primordial Cranium of Pecilophoca Weddelli 433 above the level of the anterior paraseptal cartilage. In its whole length the anterior paraseptal cartilage is separated from the inferior border of the septum nasi by a septo-paraseptal fissure (Pls. 1, 2, 7).. Cartilage and fissure lie opposite the middle third of the lower border of the septum. Anteriorly the cartilage is deepest, and it here divides into two somewhat rod-like branches. The upper of these two branches is fused with the hinder end of the processus lateralis ventralis; the lower branch projects freely forward below and almost parallel with the upper, and is apparently the cartilage of the naso-palatine duct. The processus lateralis ventralis passes laterally from the lower part of the anterior third of the septum nasi, and in coronal section has the appear- ance of being the result of bifurcation of the septum. For the most part it represents the medial segment of the lamina transversalis anterior. THe MepriaAL ASPECT OF THE LATERAL WALL OF THE NASAL CaPsuLeE (PI. 9). This is of especial interest on account of the peculiarities which it presents, which seem to be characteristic of the group to which the seal belongs. It is divisible into the same three parts as the lateral aspect of this wall, viz. from before backwards, pars anterior, pars intermedia, and pars posterior. Of these parts the pars anterior is intermediate in size, the pars posterior the smallest. The Pars anterior is concave from above downwards, narrow anteriorly at the fenestra narina, but very deep posteriorly where it is confluent with the pars intermedia, from which it is separated in its upper half by the - erista semicircularis, and below that only imperfectly separated by a slight ridge corresponding with the lower part of the sulcus antero-lateralis. Its lower margin is only slightly inclined towards the septum nasi, but the mucous membrane covering it is prolonged for a considerable distance towards the septum, and forms the soft part of the maxillo-turbinal. A lateral nasal gland and duct—the gland at this stage very small—lie in the lateral wall some little distance above the maxillo-turbinal. No trace of a cartilaginous naso-turbinal can be observed. The pars intermedia is very deep from above downwards, and much wider below than above. It is divided into two parts by the union of the crista semicircularis with the anterior root of the first primary ethmo-turbinal. The part above this root is the recessus frontalis, which is of comparatively small size but of considerable depth, so deep anteriorly, that it causes the prominentia frontalis to bulge considerably over the foramen epiphaniale. The deepest part of the floor of the recessus frontalis is perforated for the exit of the 434 Professor Edward Fawcett lateral branch of the nasal nerve, which appears on the external surface of the nasal capsule. The perforation is the foramen epiphaniale. Below the connection of the anterior root of the first primary ethmo- -turbinal with the crista semicircularis is found that large recess which is the recessus maxillaris. This recess, which is at its upper and anterior part partly occupied by the lateral nasal gland, is bounded above and behind from before backwards by the. first -ethmo-turbinal and by the lamina transversalis posterior. Externally these structures are indicated by a sulcus, which is the sulcus postero-lateralis, which commences above just to the outer side of the anterior end of the spheno-ethmoidal com- missure and terminates after an obliquely downward and backward course below the cupula posterior. Its course is nearly parallel with that of the upper margin of the pars posterior. I do not make out any real recessus anterior. The Pars posterior is of very curious form, and characterised by con- taining the ethmo-turbinals, of which there are two. Its floor is not so complete as usual, since the lamina transversalis posterior at this stage is not sufficiently chondrified to come into close relation with the septum nasi. Its anterior boundary is formed by the anterior margin of the first primary ethmo-turbinal, and its upper wall is the same as the lateral border of the subcerebral vacuity in the nasal capsule above. The concavity of this region is a long narrow one, commencing at the anterior border of the first primary ethmo-turbinal and ending posteriorly at the cupula. It is interrupted about its middle by the second primary ethmo- turbinal, which, lying parallel with the first primary ethmo-turbinal, is in the main attached to the upper surface of the lamina transversalis posterior. It is only when careful stock is taken of the pars posterior that one is able to understand how the lamina transversalis and cupula posterior have been formed by the backward thrust of the nasal sac. The conditions presented to view, not only in the model but in the microscopic sections from which this model was made, are strikinglylike those found in Dasywrus viverrinus at the 7-mm. and 9°5-mm, ¢.r. stages, the only real difference consisting in the direct continuation of the hinder wall of the nasal capsule with the septum in Dasyurus. Both primary ethmo-turbinals are at this stage of simple form, and the first, which is the largest, shows no sign of bifurcation at its free extremity. THe VISCERAL SKELETON (Pls. 1, 2, 3, 4). This consists of the usual parts, and presents little of interest. The cartilage of the first arch (Meckel’s cartilage) ends posteriorly as usual in the malleus cartilage, whose appearance is of the usual kind, a a ee | The Primordial Cranium of Pecilophoca Weddelli 435 Meckel’s cartilage itself passes at first downwards and forwards, and soon comes to be covered laterally by the mandibula; at about the junction of the middle third with the anterior third of that bone it turns suddenly upwards, and becoming narrow blends with its fellow medial and slightly anterior to the bony mandible. At this stage it shows no sign of ossification. As regards the Malleus, all that may be added is that it has inserted in the usual place an unusually large tensor tympani muscle, under whose tendon the chorda tympani nerve runs forwards. The Jncus cartilage seems to be somewhat large in proportion to the malleus, and it projects on that account more laterally than usual. Its body bears the usual facet for the malleus, and seems to give off four processes, one which passes forwards along the outer side of the head of the malleus for a short distance, another which passes backwards behind and somewhat medial to the head of the malleus. These two processes taken together may be regarded as affording an unusual amount of articular surface to the head of the malleus. ‘The remaining processes are the crus longum, which descends behind and parallel with the manubrium mallei, and terminates where the latter makes its final bend inwards. It is thus about one-third shorter than the manubrium mallei; the angle between the lower end of the crus longum and the manubrium mallei is occupied by the chorda tympani nerve. The crus breve is very thick, and directed downwards and outwards. It terminates near the upper end of the very small crista parotica in a sharp point. On the lateral side of its root lies the extremely small and just ossifying squamosum.! The Cartilages of the second arch (Pls. 2, 3,4) are represented by the stapes, the processus styloideus, and the cerato-hyal. The Stapes cartilage (Pl. 6), which is perforated by an extremely small atrophied stapedial artery, is connected at its head by dense connective tissue with the medial side of the crus longum of the incus cartilage. Its foot-piece is loosely fitted into the foramen vestibuli of the vestibulat segment of the pars cochlearis. There is no trace of any connection between the stapes and the remainder of the hyoid arch. The processus styloideus is attached by dense connective tissue at its proximal end to the lower part of the crista parotica of the pars canali- cularis of the auditory capsule. It soon increases enormously in size, but soon again, after being crossed by the facial nerve on its infero-lateral 1 According to Weber, Die Stiugetiere, 1904, pp. 544 and 545, “Wie bei diesen (Cetaceen), werden auch die Gehérknéchelchen massig und schwer.” This statement certainly agrees with what I find here so far as concerns the incus cartilage, but I do not note anything extraordinary about the size of the malleus and stapes cartilages. However, the general statement made by Weber may be correct, and will perhaps account for the large size of the tensor tympani and stapedius muscles, 436 Professor Edward Fawcett aspect, becomes narrow and cylindrical, when it is crossed on its dorsal aspect from without inwards by the chorda tympani nerve. It retains a | cylindrical form for some distance, and in coronal section is quite circular. Below the level of the lower end of the manubrium mallei it gives origin to the musculus stylo-pharyngeus from its inner side, and during this time its shape in coronal section is transversely oval owing to the fact that the cartilage has at this point turned sharply inwards. On arriving near the back lateral wall of the buccal cavity an interruption in the cartilage takes place, and the succeeding cartilage rapidly assumes a greater calibre, soon itself breaking up into two, of which the more distal bends downwards as the cerato-hyal to articulate with the anterior end of the thyro-hyal and the supero-lateral end of the basibranchial. For a considerable part of its length the processus styloideus-lies in close association with the posterior belly of the digastric muscle, being on its dorsal aspect, but at this stage neither the stylo-glossus nor the stylo-hyoid muscle has any direct attach- ment to it. A well-named cerato-branchial muscle connected not only the cerato-hyal but the segment proximal to it with the thyro-hyal or first branchial cartilage. The Cartilage of the first branchial arch, viz. the thyro-hyal, is dorsally directly continuous with the superior cornu of the corresponding ala of the thyroid cartilage, and in front of this union the internal laryngeal nerve (which possesses here a ganglion, like that pointed out by Nicholas in man) enters the larynx. The anterior end of the thyro-hyal is connected with the inferior aspect of the outer end of the basibranchial (corpus hyale) by dense connective tissue. The basibranchial cartilage or corpus hyale is a large transversely disposed cartilage, whose lateral extremities are connected by dense con- nective tissue with the cerato-hyal above and the thyro-hyal below. Above and below it may be seen the remnants of the thyro-glossal duct; the upper segment consisted of little more than débris of its original structure, the lower one apparently being in complete development. The Cartilage of the second branchial arch (4th visceral) forms the thyroid cartilage. It is deep posteriorly, the deepness being accentuated by the upper and lower cornua, which are directly continuous respectively with the thyro-hyal and the cricoid cartilage. In the former case the continuity is cartilaginous, in the latter by dense cellular tissue. When traced forward the ala of this cartilage very considerably diminishes in — height, until it is no higher or deeper than the thyro-hyal. Having reached the plane of the lateral end of the basibranchial (corpus hyale) it suddenly increases in depth, mainly at the expense of its upper border, which rises up to form a considerable projection, which almost reaches the hinder . ee a a a | oe “ ee a ee a 4 The Primordial Cranium of Pecilophoca Weddelli 437 margin of the basibranchial. The anterior border of the ala of one side is not fused with that of the other; a comparatively wide interval in fact separates the two. The whole appearance of the ala at this time very strongly resembles that of Tatusia at the 12-mm. stage. Whether a foramen will appear in it at a later period in the same way as it does in the older stages of Tatusia I do not know. ‘The condition is interesting. The Cartilages of the third branchial arch, viz. the cricoid and arytenoids, are moderately well developed. The former is practically complete, and articulates by dense cellular tissue with the inferior cornua of the thyroid cartilage. The arytenoids are only moderately chondrified, and connected basally by dense connective tissue with the cricoid at the usual place. THE OssEouS SKELETON (PI. 4). ”? This at the stage modelled consists entirely of membrane or “covering bones, which aré the frontale, parietale, incisivum, maxillare, zygomaticum (malar), squaamosum, mandibula, and pterygoid. The Os frontale is in a more advanced state of ossification than the os parietale. It lies lateral to the basal part of the ala orbitalis, the spheno- ethmoidal commissure, and the prominentia frontalis of the pars intermedia of the nasal capsule. It projects for some distance above these cartilages, and is deepest at its middle, tapering off to a point anteriorly, where it overlies the prominentia frontalis. The Os parietale is smaller in size than the os frontale, and is only ossified to a very small extent, forming a quite thin lamella lateral to the lamina parietalis where that is bending forwards over the spheno-parietal fontanelle. It at this stage scarcely projects above the level of the lamina parietalis. The Os incisivum is very small, consisting of a transversely laid plate which closed below that part of the rostro-ventral fissure in the floor of the nasal capsule, which has lost the lamina transversalis anterior. It reaches the processus lateralis ventralis of the solum nasi, but does not send backwards along the medial side of the anterior paraseptal car- tilage any paraseptal process; perhaps the stage is too early for this to be ossified. The Os mawillare is of comparatively large size, but very primitive in form at this stage. It consists of a body from which there projects upwards in front of the prominentia maxillaris of the pars intermedia of the nasal capsule a massive ascending or frontal process. This body vertically below the frontal process bifurcates into a lateral vertical and an approximately horizontal lamella. The lateral lamella is the outer 438 Professor Edward Fawcett alveolar wall, the median one is the palatine process, which projects but a short distance under the rostro-ventral fissure of the solum nasi towards the septum nasi. From the postero-inferior angle of the body of the maxilla an outward projection arises, which is grooved above by the very large infraorbital nerve, and which lateral to this groove sends upwards a short ascending process; beyond this infraorbital sulcus the maxilla terminates in a sharply pointed zygomatic process, which articulates for a considerable distance with the zygomaticum (malar). The Os zygomaticum is a small rod-like bone, which lies along the lateral aspect of the zygomatic process of the maxiilare and projects some- what beyond it. It is separated by a long interval from the squamosum. The Os squamosum is just in the very beginning of ossification, and forms a small scale of bone on the lateral aspect of the body of the incus cartilage. This is its usual position at the commencement of ossification. The Os palatinum is well formed, somewhat triangular in appearance when seen from the outer side at this stage, with apex upwards. Its basal border is slightly inrolled to form the commencement of its palatine process. The antero-posterior length of the bone slightly exceeds its height. The Os pterygoidewm 1 is found almost immediately behind the palatinum, is of some size, and more or less perched on the upper lateral aspect of the cartilaginous humulus, which is.of great size. The Mandibula is of large size, and stretches from within a short distance of the malleus cartilage behind to the anterior end of Meckel’s cartilage, whose general course it very closely follows. The condyle is unossified, but there is a small ossified processus coronoideus. Near its anterior end, that is, at the junction of the hinder two-thirds with the anterior third, the bone takes a sudden bend upwards, just as Meckel’s cartilage does, and in the concavity of the bend an incisure is developed to allow the outward passage of the nervus mentalis. Fronyr the medial side of this part there arises an internal alveolar wall, which arches upwards over the outer side of Meckel’s cartilage. At this stage no accessory cartilages are formed anywhere in connection with the mandibula. GENERAL CONCLUSIONS. One of the most striking facts about the primordial cranium of Weddell’s seal at the. stage above described is its extreme simplicity, and, with the exception of a somewhat specialised nasal capsule, it might very well serve as a standard from which to work backwards or forwards in phylogeny. The great preponderance of the cerebral segment over the nasal region is 2 , . , ~ en a Ee eee eS Se a The Primordial Cranium of Pecilophoca Weddelli 439 very noticeable, and its almost circular appearance as viewed from above gives it an almost human appearance. The connections which the chordal part of the central stem form with the cochlear capsule (one on each side) are simple in the extreme, and really are outgrowths of the chordal plate itself. Then too the trabeculo- cochlear commissures are of a very simple type. The central connections, however, of the ala orbitalis with the central stem are essentially, so far as I know, signs of a comparatively high stage of evolution. They appear first in carnivora, and are repeated in man. That the post-optic limb con- -nects through the ala hypochiasmata is clear, and similar to what happens elsewhere, but the original mode of union of the two pre-optic limbs is not possible of observation at this stage. The connection of the more anterior of these pre- optic limbs with the hinder part of the nasal capsule seems to be a carnivorous feature. The exoccipital cartilage, projecting as it does backwards with little or no connection with the neighbouring cartilages, is a great help to the proper understanding of that structure. The supraoccipital cartilage has too a very simple and typical form, and its share in the production of the tectum cranii posterius is very interesting. The existence of a tectum cranii anterius, placed so far forward as it is, and consisting of four parts arranged in the same coronal plane, is likewise a point of very great interest. “The small size of the cochlear capsules in comparison with those of the cat is another point of note. The extremely simple form of the ala temporalis too is another interesting point, but it is possible that at a later stage it may become more complicated. As regards the nasal capsule, the small size of that part which contains the ethmo-turbinals is a feature which is retained up to adult life and almost characteristic of the seal group, in which it is known that the ethmo-turbinals are small and occupy but a small part of the interior of the osseous nose. On the other hand, the large size of the maxillary prominence and corresponding maxillary recess seem to be in harmony with what obtains in the adult condition, but at this stage the maxillo- turbinal cartilage is very small and not even inrolled, whereas in the adult it is of very great size and complexity. There is at this stage no sign of a cartilaginous naso-turbinal. This too is usually a large structure in the Phocide in adult life. The condition of the lamina transversalis _ posterior, the absence of an intermediate segment to the lamina trans- versalis anterior, are to be regarded as indicating a high stage in phylogeny. The small size of the anterior paraseptal cartilage is clearly due to the absence of an organ of Jacobson. As to the ancestral history of the animal and the group to whith it 440 Professor Edward Fawcett belongs, and further, as to its affinities, this communication helps but little. As regards the former, not at all perhaps; as regards the latter, it is impossible to say more than that in certain particulars, e.g. the relation of the ala orbitalis to the nasal capsule and the form of the pars trabecu- laris, together with the presence of well-marked and separate trabeculo- cochlear commissures, a well-marked chordo-cochlear commissure with a large basi-cochlear fissure, the seal resembles the cats, but I do not know how it compares with the Arctoidea in these respects. The following, taken directly from Weber, 1904, p. 551, is of interest :— “Vorgeschichte. Die Paliontologie wirft bisher keinerlei Licht auf die- Vorgeschichte der Pinnipedia. Es ist zwar eine Anzahl derselben bereits aus dem Miociin bekannt; diese meist unvollstiindigen Reste schliessen sich aber, insoweit sie sich beurteilen lassen, in erster Linie eng an unsere heutigen Phocidae an. Friiher bereits wurde die Ansicht geiiussert, dass die Pinnipedia direkt von Creodonta abzuleiten wiiren. Wenn man dabei auf die geringe Zahl der 1. wies, da ja auch bei Creodonta I.—fehlen kann, so vergass man wohl, dass dies bei Pinnipedia dichtbar, ein Verlust ist, der erst seit Jiingerer zeit, seit Anpassung an das Leben im Wasser datiert. Neuerdings fiihrt Wortman die Pinnipedia auf+ Patriofelis und damit auf die+Oxyaenidae zuriick, welche fiir den einen noch Creodonta sind, fiir andere bereits Carnivora, die den Katzen sich nihern. Dieser Ansicht Wortmans ist aber sowohl Winge als auch Osborn entgegengetreten, Ein zusammenhang der Pinnepedia mit den Ursidae ist wohl die gesichertste Annahme. Vergleichung der recenten Formen lehrt eine menge auffallender Uebereinstimmungen kennen, die auf Blutverwandtschaft deuten. Ich nenne den Bau der Trommelhohle, das Verhalten des Maxilloturbinale, das in beiden astig ist (Hypomycteri); den langen Darmkanal ohne Flexura duodeno-jejunalis, der an einfachem Mesenterium commune aufgehingt ist ; die gelappten Nieren, das Fehlen der Cowperschen Driisen. Damit kommen wir zum Schluss, dass die Pinnipedia mit den Ursidae zusammen primitiven Amphicyon-artigen Carnivora entsprangen und allmiichlich auffillige Um- fornung durch ihre Lebensweise erfuhren,” PRINCIPAL REFERENCES, Voir, * Das Primordialcranium des Kaninchens,” Anat. /efte, 1 Abt., 116 Heft (38 Bd., H. 3). Weer, Die Stdugetiere, 1904. Fawoerr, “The Primordial Cranium of Microtus amphibius,” Journal of Anatomy, vol. lii. Journ. of Anat. | ‘aaoqe wo ‘okiquie (uosdmoyy Aory,q]) ‘ww-2Z Jo WNTURIDOIpUOYH § “*([ves SJ[eppeA\) 2272ppa4y wo0YdopIWmMT ; {980d yuBI0_ "9005, --~----_-------- “HUNUDeI “LO ++ --- *plopApioo “901g =~ *200X9 “QIBQ —------ ‘ssopsod Ay "yeawg _.. “SLIv[e “UTUTBT 2... *sdvo ‘ooovldns ‘ssi ----- *Slivpe-waidns “sso0eq - herepnznl 107 “ST[RPLOYO Sled “UNUIUIOD SIINID “WOIg * “*qUI "QUB “OUNQNS BSSOT "SUBI} “UA “UIg *ydurffopue "yond “que “olmes “WOLT *Y909-18Bq “SSI *I909-OpAOY *UIUOD “QUr "pne “qeoyy *yooo ‘sdeg *[yooo “gong *SI[BSLOp BpPAOYH - : 4 ‘ . 7 *snqnooeg “qsod bpeaatiae oe those ; ROU. oe bs ; ; A =: *[eBlovy “olng ‘190180 “10g --“ ii f.. —-- "qUB *YOOO-"daqBaAy “MIWOD * f “Aydoddy vssog any ----------- | 7. ~ ‘srTeyoqaed “Urey “pom ‘alps "901g ten tem es nh el > a om t SUBIY BISHD “STjerodu94 BLY * ws : F *[Bloey "AION ‘orndo “10g. --—- ‘avd-"yds ‘uv}U07 c---- eee --== "19001 JO UIZIIO [eIqAeg “sBryooaid “10g ----- 7 *orydoord araay ‘que xIpey ---._.B ‘aydorde “10g 4 ------3 w-=-- ‘red "qio *wWULOD “SI[BJIGIO BLY ------ ‘orydoaad atcaz) Que xipey ‘qsod ‘seu vendng --- . “(yur sued) -s-----~. "Que UBIO “4997, a i “SIyeseu BsOOnyT ‘I ‘qang-owyyg ‘ : “““- ue ‘qdosvied “4189 *apTeqUoay SO j f ; [---- "yja-"yds “mutog “OUTOIWUDS BISHIO ‘a se ; "SBU-OUIQIO “Sst -— “IQUAA “9RT ‘901g ‘a[equoly SQ --—-----— ‘ueydide “iog ---------- Shekel ~------ WCF ¥zS1I9 Professor EDWARD FAWCETT. (Prats la. Journ. of Anat. | 9 a> a *sade(ty1v0 peyidioo0xe oy Ayjeroedse Zutmoys ‘oXiquia (aosduroyy, Aory,q]) “utut-2Z Jo ULNLUBLOOIpUOYo jo J[ey tapary jo yoodse [wan *([¥as s. (appa) 2279pp2.4 vooydopsoq “qsod 1ue10 9997, ‘sso[S0ddy ‘;wueng —----------. *dpooowsdns "yaeQ + -----= *d1900x0 “4aeyy -------- *SLIVTV “Wey eines *ploj{puod o01g *sdvo ‘dpooovidns ‘ssi = ~=---- ‘sdvo ‘drooovidns ‘ssi “=== “UMD SLINIO “UOIg ° “JUL ‘OIVqns -BssoT --="sI[Vsiop Bpi0yy “QUI “GIpne “4QveTy - === *"17900-Opi10Yyo *UIUIOD ‘gavnSnffsog os esee ee ““ ee bg): s r . * - Beh. ETS “que|dute ‘olaqn “wWo1g sows ‘dns ‘uur0o wMIplowig --- = ~~~ ~*~ : iil ae Si oe “___. ‘SHrepouy “ong ‘ : = we. “yooo-1seq “S81 << ~., *e+9s—7.=* *q000 “‘Sdgs *sUBI} BASIE) “qsod ‘oaqea} “tUt0D ‘OOIv) “10,7 “que "Y09-"daqeay “MUMOD “sLIBNdaqed} sIV_ = - Professor EpwARD FAWCETT, [Pirate II. Journ. of Anat. } ‘mojeq mo ‘okiquio (uosduroyy, Aory,q]) “wu-2Z Jo WNTURLOOIpUOYO “2yappe4, vooydopomy *ysod t1uedtd “Q00q, ------=*. “sI[esiop Bployg ~--- “(SLIB[R “UB[) “D00X ----- ‘sso[zodAy “10g *|Apuoovsed “001g -- *orepndnl ‘10g --- “sI[BpIOYO SIvq .-- *yooo yo ‘ydurdyIed “10g - *podeys ‘ssoy - *yoOo-18¥q "SSI *QIQSOA “10g - hate “SNORT .- *yooo *sd¥g __ “PHJOIv) “YIy .. *snouy *OIJOIBO “IO - “gue "yO “NaqBIy “ULTIOD *(snjnurey) “SA104d “qavQ -- -"4yo0o ‘sdvg qsod “909 *99qery “WUOD *99qQBIy SIVA --"duie1 Bly -=~"qolred "ueyT “dw94 BlV -~ =~ "ied "qo “UlUIOD = ‘g[eqoled “vorndo ‘ysod eruay, ~~ 001 JO USO -j . o19do “107 “""""-dns ‘brqo jo UIsLIO 4 ST TE "STLVIIGIO BLY — B® PEE ; ‘orjdooad vue yy, y “4ul good ‘osnyy- ~*qsod ‘suvig “Wey ehie * *Og 3 Ee igor === “ISB “dag =< TT “Qan4-"Us -- "Seu “qJo ‘SSIq =---o[UqUOL “pour *yoor “osnyy __._ "yur “byqo ‘osnyy -------- *XBU “WOIg ~-" ; , “Hi DyQO: “Ost--——==-—— bias yp j ; - pmmmn eT “Qing-"Use “que “ydosvivd “4189 .-------------- 4 ; Bs, mae ae=- “XBU WO1d “gered “seu “gonp "Wu)) ------------------- quod] ‘W101 g ----"-------------- “JRE “JuB snorng .--= *UIMAISIOUT ----- “QUOA “4R[ ‘OOIg “7-7 “sI[WSBU BSOON]Y -=------ Professor EpwARD FAWCETT. {Piate ILI. Journ. of Anat. | eyo! ‘apis yor wo ‘okiquia (uosdmoyy, Aory,({) “utur-7Z Jo wNIUBIOOIpUOYD *([P98 8,[[@PP2 MA) 2219PP2AL w20ydopL90q “qsod ‘uwid "yooy, -----+ *900B1dNS *41N9 *019d0 ‘001d - —-— — - DWARD FAWCETT. {PLATE V. Journ. of Anat. | “que ‘olules “MOIg = qu “oreqns Bssog -~—-~-, *WIUUOO “Ando * “qU] “oytes “uvo vyndury ---> <5 ~ ++ “ysod “oreqns Vg --- *> > —~+---~- Comm. sph.-eth, Crista semic. ~ ~~-- Pars interorb. nagalis. \ -~~-- Eth.-turb. I. ‘ =+—-- Tect. nasi post. ->- Oculus. ; , : > an orbit. Muse. rect. lat. Ala temp. Nervus, maxill. Fontan. sph.-par. Comm. trabec.-coch. ant. For. caroticum. Comm. trabec.-coch. post. Comm. suprafaci- alis. Canal. facialis. For. acustic, sup. For. acustic. inf. Fiss. basi-coch. => For. endolymph. : Chorda dorsalis. - Canal. hypogloss. ==» Cart. exocc. For. magnum. see ~-- Tect. cranii post. Felis catus. Chondrocranium of 32-mm. (Hill) embryo, from above. Professor EDWARD FAWCETT. The Primordial Cranium of Pwecilophoca Weddelli 441 Fawcett, “The Primordial Cranium of Zrinaceus ewropceus,” Journal of Anatomy, vol. lii. Mackin, “The Skull of a Human Fetus of 40 mm.,” American Journal of Anatomy, vol. xvi., No. 4, Sept. 1914; vol. xvi., No. 3, July 1914. CORRIGENDA TO PAPER ON ERINACEUS EUROP.®US, VOL. LI. P. 211, 1. 3, read “19” for ‘* 25.” P, 242, 1. 33, read “lamina transversalis posterior” for “posterior paraseptal cartilage,” P. 247, 1. 4, read “19” for “25.” a me nl ae e VOL. Lil. (THIRD SER. VOL. XII.)—JULY 1918. 30 ON THE STRUCTURE OF THE HUMAN SOLEUS MUSCLE. By Dovue.as G. Rerp, M.B., Ch.B., Edin., M.A. Trin. Coll. Cam. QUAIN, Poirier, and others give fairly good accounts of the general structure of soleus. Frohse and Frankel in Bardeleben’s Anatomy, 1913, deal in some detail with the innervation of this complex muscle, and have even ventured to enumerate the anastomoses which are to be found in its substance between the branches of the two nerves of supply. Le Double and others deal with the grosser variations of the muscle in man and other vertebrates. ; But nowhere can I find mention of the points to which I here especially refer. ‘ One muscle, which I noted whilst superintending dissection, is pro- foundly modified (see C, fig. 2). In fig. 11 the anterior (deep) surface of a practically normal soleus is shown, and the student is recommended to reflect the muscle as is here indicated, viz. by dividing its tibial head.’ Indeed this head is the (relatively) recently acquired portion of the muscle, and is almost a human characteristic. It is the head incised by the surgeon in the high operation of ligature of the posterior tibial artery. In this way the nerve which arises from the upper part of the posterior tibial (see fig. 1) and supplies the “anterior” or “fishbone-like” portion of the muscle is not liable to be injured. A, in fig. 2, also shows the structure of the normal soleus. It is the appearance seen on transverse section at the level of the line in fig. 1, and in B, fig. 2, indicating the central tendon of the fishbone- like part. In the anomalous muscle C (see fig. 2) a tendinous lamina ((q) in figure) lies in the substance of the muscle, reaching practically to its tibial border. ' In the original the foot is also shown, for, as in other of my anatomical drawings, I have tried here to get connection between parts in different regions, It has well been said that the illustrations in the books too often fail to show the continuity between different parts of the same structure. In the original the relations of the plantar vessels and nerves to one another are shown, The external plantar artery nowhere directly “lies upon the caleaneum.” It is separated from its inner surface by the flexor hallucis longus tendon, the external plantar nerve, and flexor accessorius, Fig. 1 was prepared from a drawing made upon a varnished glass plate held over the specimen. * This is the method which I find is now advised in Cunningham's Practical Anatomy, as revised by Robinson, aM ae ie 4 i i" vical . i Ee ll era ah 54 ee a Pe Soe ee ee eee ee ee ee core On the Structure of the Human Soleus Muscle 443 4 ~ It is formed largely of longitudinal fibres, and becomes continuous with a tendinous sheet which extends on the anterior (deep) surface of the muscle to the upper part of its fibular border (see (b) fig. 2). From the anterior -—An anomalous muscle (popliteus minor) Inner and outer heads of gastrocnemius (cut) Tendon of - \.. Biceps Semitendinosus \ Ay Plantaris : Bursa between \\\\ Wy 2 ph and Aw ce | \s\ endon of semimemb. at QS -Inferior internal articular arkery Expansions of MURS SK Internal popliteal (tibial) nerve tendon of semimemb. WAR = Lower end of popliteal artery _ ——— Fibula (upper 3 of shaft) Nerve to fishbone-like part of soleus >. Aponeurotic part of soleus Popliteus — he Tibial head of soleus Cut) Part of tibialis — SW posticus (posterior) ly ——Peroneal artery Flex. dig. longus -Central tendinous septum Nerve to flex hall. long. Posterior peroneal septum Flex. hall. long. = Peroneus longus pulled somewhat outwards Posterior tibial \ - Peroneus brevis artery and Nerve Tendon of flex. dig. long. Part of shaft of tibia Tendon of tibialis posticus eater ion) Tendo Achillis (cut) GR 4 ge ee Wi = —_ S _— me | a$ a2 os La oy 4 a ps muscular sheet — \/ Tendon of / \) Yy fishbone-like part fibular border of soleus { Tibial border of soleus ~~ = — Overlapping fasciculi Muscular fasciculi Tendo Achillis ge oC Fic. 2,—A, Transverse section (semi-diagrammatic) of the normal right soleus muscle, B. Slightly abnormal right soleus, ©, Anterior surface of right soleus. The figures 1, 2 indicate corresponding parts, The appearances of the muscle suggest that the fibular lamina 2 (see A and B, fig. 2) has become greatly enlarged, that the tibial lamina 1 is not represented except superficially (see (c) in ©, fig. 2), and that there is nothing to represent the middle tendinous sheet of the fishbone-like part of the musele, the fasciculi which arise from lamina 2 extending to be inserted into lamina 1. On the other hand, the relative size, position, and a ‘ x : . o EE nt net DLO ea ane ke en ee’ On the Structure of the Human Soleus Muscle 44.5 relations of parts do not suggest this modification; and a muscle which is so greatly altered is, I think, most noteworthy. The “anterior,” “deep,” or “ fishbone-like” part presents several other variations. It is sometimes distinctly asymmetrical, and varies consider- ably in its relative width. In B (fig. 2) it is modified by the presence of a triangular muscular sheet which conceals the upper part of its “central” tendon. On reflecting this sheet a typical bipenniform part was exposed. ~The lower part of the central tendon is also hidden, more or less, in different specimens (see figs. 1, 2) by fasciculi which pass obliquely over it from the tibial towards the fibular side, being inserted into the outer aspect of the “central” tendon as well as into the outer part of the anterior surface of the tendo Achillis. To see the mode of termination of the bipenniform “ muscle ” these fasciculi must be dissected away. In other specimens they are absent, and the central tendon is then exposed downwards to its lower extremity. The tendinous fibres of the fibular lamina (2 in A, B, fig. 2), which in ‘its upper part arises from the posterior surface of the shaft of the fibula (see fig. 1), vary considerably in the extent to which they appear, and extend downwards upon the surface of the muscle to the outer side of the “fishbone-like ” part. (compare, e.g., figs. 1 and 3). In three-fourths of the specimens examined the lower part of the fish- bone-like portion (“muscle”) approaches distinctly closer to the outer than to the inner (tibial) border of the muscle. It may extend actually on to this border (see, ¢.g., E, in fig. 3). Indeed its “central” (middle) tendon may pass below into continuity with the fibular border of the tendo Achillis, and the muscle becomes penniform in its lower part (see E, fig. 3). In one-fourth of cases, as in muscle D (fig. 3), the “fishbone-like” part ‘extends below on to the tibial border, and in this case it becomes penniform below on the reverse side. The penniform part, together with the oblique fasciculi which spring from the same tendinous sheet (2, see fig. 3), then give rise to a sort of bipenniform process, and the muscle may appear as if it had been sliced down on the lower part of its tibial aspect (see D, fig. 3). In the soleus shown in fig. 1 the central tendon passes first towards the fibular side of the muscle and then becomes bent, so that its lower part, covered by the overlapping fasciculi referred to, passed to end near the tibial border of the tendo Achillis. Only in one in sixteen specimens 1 In one specimen the inner half of the “fishbone-like” part is covered by longitudinal tendinous fibres, and, without care, what is really a bipenniform muscle might have been described as Peeks These tendinous fibres are a prolongation of the posterior border of the central tendon. Compare this with E, fig. 3. 446 | Mr Douglas G. Reid does the central tendon end exactly on the middle line of the tendo Achillis, and the muscle is remarkable for its great symmetry. In many cases the “central” tendon is practically an antero-posterior septum (as is seen in the section of the normal soleus A, in fig. 2) in con- nection with the posterior aponeurosis and the tendo Achillis, and coming to the anterior surface of the muscle by its anterior border. In other cases (1) Tendinous sheet Posterior aponeurosis Middle tendinous sheet =- de Y,' | faruetes Yi 3 she D et y LW Tibia! border of soleus | aes border of Ye 1} \ Vp soleus ‘(NY Prolongation of \V; central tendon \\ Tendo Achillis Fic, 3.—Two right soleus muscles. D simulates a left soleus muscle. (These drawings were made first upon a varnished glass plate held over the specimens.) it is distinctly oblique, being directed backwards towards the fibular or sometimes towards the tibial side. It may be distinctly curved. In case E (fig. 8) this middle tendinous lamina lies in a distinctly frontal plane, and comes to the anterior surface of the muscle, not as a mere border, but in the form of a relatively broad strip separating the two parts of the bipenniform muscle from one another. The tendinous sheet after bending on itself is attached to the anterior surface of the posterior aponeurosis, but its lower part becomes continuous with the outer border of the tendo Achillis. Specimen D (see fig. 3) presents an arrangement which is just the | ee an ae a a a ae i 1 oe Ope es, Ce ae ee > ae eo ee A VERO On the Structure of the Human Soleus Muscle 44.7 reverse of that presented by specimen E. In both there is the bipenniform process already mentioned. In E the muscle appears as if it had been quite abruptly sliced down on its fibular border. This appearance of slicing down is seen also in D, but is not so marked. It is correlated with the abrupt encroachment of the “fishbone-like” part of the muscle on the border of soleus. The muscle looks distinctly lop-sided in E; and it is noteworthy that the outer head of gastrocnemius related to E, and which is as broad but not quite as thick as the inner head, descends somewhat lower than the inner head. This is possibly a correlated variation, and should be looked for in specimens in which there is a marked “slicing away” or notching. Anyone would quite readily have mistaken the gastrocnemius as belonging to the left leg. But marked cutting away is not generally to be found. Indeed the other muscles specially examined present little or no indentation of their borders (see also figs. 1, 2). The tibial tendinous lamina (1) is better developed in E than is usual, whilst, apparently as a variation correlated with the form and direction of the central or middle tendinous sheet (see fig. 3), the fibular tendinous lamina (2) does not extend into the substance of the muscle. ‘The muscular fasciculi take origin from its inner border and posterior surface only. It is most noteworthy that a soleus muscle may present an arrangement which is just the reverse of what is normal, and unless careful one could readily mistake the side to which the muscle belongs. In this paper, then, I have merely touched upon the question of the architecture of soleus and a related muscle and their correlated variations. This question, however, opens up a very large field for research to anyone whois observant in the course of the dissection of the human body and who keeps in mind the truly memorable words of Cuvier—‘ single parts cannot change without bringing about modifications in the other parts.” I have indicated, what seems obvious, that there are variations which may be compensatory—a “cutting away” supplied by some addition else- where; and one must try ever to render variations of any kind intelligible by determining, as far as possible, the mechanical causes which really operate in producing them. Comparative anatomy also may help us to see the mechanical causes at work in the case of “atavistic” variations. 1 Haughton’s Principles of Animal Mechanics (Longmans, Green & Co.) will be found useful to those interested in the mechanical advantages of such arrangements of muscular fasciculi as those referred to in this paper. 448 On the Structure of the Human Soleus Muscle It will be seen that the examples of the soleus cee I have described may be arranged into the following groups :— (1) Symmetrical forms. (2) Forms in which the bipenniform part ends on the fibular side of the muscle. (3) Forms in which it ends on the tibial side. (4) Forms in which its tendon is concealed below by overlapping muscular fasciculi, or (5) Above by a more or less triangular muscular sheet. (6) Forms in which its central tendon appears as an edge on the anterior surface of the muscle. (7) Forms in which the tendon, which may be distinctly curved, appears on the surface as a lamella which occasionally is prolonged over. one-half of the “ muscle.” (8) Forms in which there is a penniform muscle. a ee OEE AONE A EH SEMI IY SAE THE HOMOLOGY OF THE MAMMALIAN ALISPHENOID AND OF THE ECHIDNA-PTERYGOID. By H. Leiguton KEsTEven, D.Se., M.B., Ch.M. I. THE HomMoLoGy OF THE MAMMALIAN ALISPHENOID. The Argument in Brief. (1) The alisphenoid of the Crocodile can be shown to be developed from a true otosphenoidal plate. (2) In Sphenodon both otosphenoidal plate and epipterygoid are present. (8) The Crocodilian alisphenoid is therefore not an epipterygoid. (4) It is therefore clearly erroneous to quote the Crocodilian condition in support of the alisphenoid-epipterygoid homology. (5) That homology having definitely been disproven in this case, the theory is thereby weakened. (6) There is reason to believe that the Cynognathid epipterygoid lying lateral to the Gasserian ganglion was not a true cranio-mural element, but, as in the Chelonians, was a pseudo-mural element, the true wall being membranous and medial to the ganglion. (7) There is little doubt that the alisphenoid of the Ophidians is abso- lutely homologous with that of the Crocodile. (8) There is every reason to regard the relation of the Gasserian ganglion to the cranial bones and the situation of the first branch of the fifth nerve as features whose importance has been overstated. (9) In one case at least (Python spilotes) the Gasserian ganglion lies medial to the alisphenoid. (10) Among the Theria the Gasserian ganglion lies as often medial to the otocrane as medial to the alisphenoid. (11) There is no doubt about the homology of the otocranial elements of Reptiles and Therians, and that the ganglion has somehow come to lie medial in the Therians whereas it was lateral in the Reptiles. (12) Why doubt the homology of the alisphenoids in the two groups ? _ (18) Other nerves besides the first branch of the fifth have intracranial] courses in one or more cases and extracranial in others :-— 450 Dr H. Leighton Kesteven | (a) The sixth nerve in most if ‘not all Lacertilia, Ophidia, and Chelonia runs an extracranial course from posterior to the sella. (b) In the birds this nerve runs an intraosseous course lateral to the sella. (c) In the Therians it is intracranial to well in front of the sella. (14) Is not this an illustration of how a nerve may become intracranial as the shape of the brain case alters ? (15) In Perameles obesula the second and third branches of the fifth nerve have an intracranial course almost as lengthy as that of the first branch, yet there can be no suggestion that it is not the alisphenoid they lie.on. (16) The embryological evidence adduced in support of the alisphenoid epipterygoid homology, when viewed correctly, is found to remove what might have been a vital objection to regarding as homo- logous the alisphenoids of Reptiles and Therians. (17) The Reptilian alisphenoid is developed from the otosphenoidal plate. (18) The otosphenoidal plate is a parachordal derivative. (19) The basipterygoid process is a trabecular derivative. (20) The ala temporalis has an origin independent of the processus alaris. (21) The processus alaris is the Mammalian homologue of the Reptilian basipterygoid process. ; (22) The cartilaginous ala temporalis may be regarded as a remnant of the disintegrated otosphenoidal plate. (23) In one case at least (Trichosurus) the ala temporalis has primitively complete independence of the processus alaris and trabecule. There are then no valid arguments against the correctness of the older view that the alisphenoids of Reptiles and Mammals are homologous, and it follows that the alisphenoid of Birds is homologous with both. The Discussion. The history of the development of the alisphenoid in Crocod ilus johnsoni is as follows :— Stage | (length 25 mm.).—The chondrocranium has at this stage reached its full development, none of its elements shows any ossification, whilst the covering bones are more or less ossified, as also are the membrane bones of the craniovisceral skeleton (fig. 1). Immediately in front of the otic capsule is the oval foramen prooticum, which transmits the roots of the fifth nerve; in front of this is a rod_ Homology of Mammalian Alisphenoid and Echidna-Pterygoid 451 of cartilage, continuous above and below the foramen with the capsule. Below, this rod is also continuous with the cartilaginous basis cranii lateral to the posterior end of the sella; anteriorly, it is attached to the membrana spheno-obturatoria. The pterygoid articulates with it both above and below the foramen prooticum. ES «> aa 0 Nv. Vil and vill Fie. 1,—Crocodilus johnsoni, Chondrocranium seen from within. Stage 2 (length 40 mm.).—The ossification of the whole skull is nearing completion (figs. 2 and 3). The bar in front of the foramen prooticum, like all the other cartilages, Parietal. Alisphenoid, Foramen prooticum. eo ec alee va fissure. Supraoce, Epuotie. Opisthotic, Palatine. Exoccipital, - Prootic. =n Basioce, Re UNS Basisphenoid. 3 \ ee Pterygoid, Sella turcica. Fic, 2.—Side wall of cranium of young crocodile to show the extent and relations of the alisphenoid. The sphenoptic fissure transmits the structures usually transmitted through the sphenoidal and optic fissures, except the first branch of Nerve V., which arises in the foramen prooticum in this case. is ossified, and there is an extensive ossification of the membrana spheno- obturatoria spreading forward from it. This with its superadded membrane bone is the alisphenoid; its relations are as follows: Below, it articulates medially with the postclinoid eminence of the basisphenoid medially and 3 with the pterygoid externally. Viewed from within it is found to articu- late posteriorly with the prootic below, above this to form the anterior 452 Dr H. Leighton Kesteven boundary of the foramen prooticum, above this it articulates with the prootic a second time, along the roof it articulates with parietal behind and frontal in front. Viewed from without, immediately behind the articulation with the pterygoid, the alisphenoid articulates with the quadrate; above this articulation the prootic and the foramen prooticum separate quadrate and alisphenoid, which meet again above the foramen; along the roof are the articulations-with parietal and frontal. The Gasserian ganglion is found lying against the cartilaginous rod and membrana spheno-obturatoria in Stage 1, and in a fossa on the outer aspect of the alisphenoid just in front of the foramen in Stage 2. A comparison of my Stage 1 with stages Q and R of Sphenodon, as Post frontal. Frontal. Parietal. Alisphenoid. Nye vas Squamosal, —-——_——— Sphenoptic fissure. f » Int. orb. sept. Quadrate Prefrontal. Lacrimal. — Nasal, Uy ge ra —~ Premax. Exoccipital, Quadrate jugal. Jugal. ~~ Quadrate. —~. Basioccipital. Bee kes Basisph. ~~ Palatine. Jugal. Maxilla. Pterygoid. Basisphenoid. Prootic. | Foramen prooticum. Fie. 3..—Young Crocodilus johnsoni. figured by Howes and Swinnerton (1), pl. iii. fig. 4 a pl. iv. fig. 10, can leave no doubt that the cartilaginous rod which has been described above is represented by the body and process 4 of the otosphenoidal plate ; that is, that the alisphenoid of the Crocodile is developed from an otosphenoidal plate. On pl. iv. fig. 3 in the same work the otosphenoidal plate is shown in transverse section, with the epipterygoid lateral to it and separated by nerve elements below, which are probably Gasserian ganglion; the first branch of the nerve is definitely indicated medial to the epipterygoid and lateral to the otosphenoidal plate. Quite apart from the situation of the ganglion and nerve, however, a comparison of the two leaves one convinced that the alisphenoid of the Crocodile, being developed from an otosphenoidal plate, cannot be regarded as a modified epipterygoid. argue ih 4 xa Homology of Mammalian Alisphenoid and Echidna-Pterygoid 453 Broom, who in 1907 (2) expressed the opinion that we are justified in assuming that the alisphenoid of Mammals is homologous with part of the cartilaginous pterygoid arch of the Reptiles—thereby giving birth to the epipterygoid-alisphenoid homology theory—at that time proposed to regard the alisphenoid of Crocodiles as homologous with the columella cranii of other Reptiles. He had previously shown that in several diverse orders of the Lacertilia there was actual or presumptive embryological evidence of that cartilaginous continuity of the columella cranii and quadrate which had already been demonstrated by Howes and Swinnerton in Sphenodon. The alisphenoid in Stage 2, described and figured above, might appear to lend strong support to the idea that it is really homologous with the columella. We have here a bar of bone extending from the pterygoid and basisphenoid below to the parietal above, situated immediately in front of the foramen prooticum, and showing actual contact with the quadrate below the foramen (thus apparently maintaining a condition surely primi- tive, since it is found in earlier stages in several different Reptiles); finally, this bar is derived from cartilage, and the expanded condition of the whole bone is due to secondarily added membrane bone. On the other hand, this bar has been shown to be developed from a portion of the cartilaginous neural cranium derived from the parachordal cartilages, is absolutely in the same anatomical plane as the inner wall of the otic capsule and the membrana spheno-obturatoria, is continuous below with the cartilaginous basis cranii, is in contact only with the quadrate, and ossifies quite in- differently thereof at a later date. Clearly then, though the epipterygoid homology of the alisphenoid in Crocodiles is an inviting idea, we have to decide against it in the face of the evidence. Broom is not alone in quoting the Crocodilian condition in support of the epipterygoid-alisphenoid homology, for the same error was made by W. K. Gregory (3) in 1913. Since the Crocodilian condition has been quoted .by two such capable observers as Broom and Gregory in support of the theory, it may now be cited as evidence against, and to that extent the theory is weakened. Further, inasmuch as the error arose from the study and comparison of adult features, we are justified in being slow to accept evidence of like kind from the paleontological record as weighty, more especially when it is remembered that in the case of the fossils it is impossible to check by dissection the assumed relation of the bones to soft structures. Watson (4) offers, an hypothetical outline of the evolution of the epipterygoids in the Therapsids, this evolution culminating in the Cyno- gnathids, in which there is a “posterior process occupying the position 454 Dr H. Leighton Kesteven of a quadrate ramus of the pterygoid, which reaches the quadrate, but passes in front and outside that bone” and “forms the side wall of the anterior part of the cranial cavity.” From this last he draws the con- clusion that the anterior portion of the cranial cavity “is therefore not homologous with that of other reptiles, but increased by the addition of epipterygoid cavities, is analogous to but not completely homologous with those of mammals.” He further says, “that this is so is shown by the position of the process which I have at various times described as a pro- cessus anterior inferior of the prootic in the Therapsids. This process forms the lower border of the incisura prootica. It always lies medial to the epipterygoid, leaving a space in which the semilunar ganglion must have lain... . The anterior superior process of the prootic of a Cyno- gnathid, whicli passes inside the upper end of the epipterygoid, forms part of the side wall of the anterior part of the brain cavity.” Now, an examination of the skull of the Crocodile checked se dissec- tion reveals the fact that in all these features the alisphenoid agrees with the Cynognathid epipterygoid, except that the processus anterior inferior prootici, which lies medial to the Gasserian ganglion, is flush with the alisphenoid, there being no space between the two in which the ganglion night lie. Watson rightly decided that the two bones are not homologous in the presence of this one feature. It will be shown later that though Watson was correct in his conclusion, the evidence on which he relied was untrustworthy. There can be no doubt that the bone termed epipterygoid in the Cynognathids and Therapsids generally is correctly named; it remains to be proven that it is never a true cranio-mural element. Watson correctly assumes that the ganglion lay in the interval he describes between the processus anterior inferior prootici and the epiptery- goid; he is, however, wrong in assuming that it is therefore intracranial. In all the recent Reptiles in which an epipterygoid is present the ganglion lies lateral to the inner wall of the prootic and its anterior inferior process when present and medial to the epipterygoid, but outside the lateral cranial wall which, membranous in this region, rises from the basis cranii along the basisphenoid, and prootic anterior inferior process when present toward the sphenoidal flanges of the parietal and frontal bones, the posterior and inferior boundaries of the prootic foramen being bony, the anterior and superior membranous, Were we unable to check by dissection our observations of and dedue- tions from the Chelonian skulls we would assuredly decide that the Gasserian ganglion and first branch of the fifth nerve are intracranial. In a es Homology of Mammalian Alisphenoid and Echidna-Pterygoid 455 Chelone midas-the ganglion would be correctly assumed to have lain between the post-clinoid process and the processus anterior inferior prootici on the inner side and the epipterygoid process of the pterygoid on the outer, just below the suture with the alisphenoidal lamina of the parietal, and therefore medial to the vestigial epipterygoid, apparently within the cranial cavity. A dissection shows, however, that the true cranial wall is membranous, being attached below to the processus anterior inferior prootici and to the post-clinoid process, extending thence upward and outward to be attached to the alisphenoidal lamina of the parietal well above the ganglion, which thus lies medial to the epipterygoid and lateral to the anterior inferior process of the prootic but outside the cranial cavity. In Chelodina longicollis the skull features would lead to similar conclusions, not only the ganglion but also the first division of the nerve apparently lying within the cavity, the latter for the greater part of its length. Here, again, a membranous cranial wall intervenes between the brain and the ganglion. The pseudo-cranial wall is in this case formed by a sphenoidal lamina of parietal above and a much smaller sphenoidal flange of the pterygoid below. A comparative study of the Reptiles reveals the fact that except in certain specialised forms the true bony brain-box is composed of the elements of the occipital segment, the elements of the otocrane, one or more dorsal membrane bones in front of the supraoccipital, and a basisphenoid below in front of the basioccipital. Laterally in front of the otocrane the wall is or was membranous or cartilaginous, except dorsally, where more or less constantly a sphenoidal flange from the parietal and (or) frontals enters into the cranial boundary. In all these cases we find: the seventh nerve traverses the otocrane, the fifth leaves the cranium immediately in front of the otocrane, the sixth, fourth, third, and second pass forward with longer or shorter intracranial course to reach the orbit. It is to be concluded, then, that the primitive brain-case was devoid of components other than those derived from the occipital vertebra, the basi- cranial axis, the auditory capsule, and the covering bones dorsally. In certain specialised forms—Cynognathids, Crocodiles, Ophidians, and Chelonians—we find expanded bony plates actually or apparently bounding the cranial cavity laterally in front of the otocrane. Dissection of the recent forms discovers to us that these plates are of two quite different kinds, so that we can recognise on the one hand pseudo- mural elements (Chelonians) and true cranio-mural elements (Crocodiles and Ophidians). The Cynognathid epipterygoid is doubtless a pseudo-mural plate. The 456 Dr H. Leighton Kesteven true cranio-mural elements—alisphenoids of Crocodiles and Ophidians—must surely be correctly homologised with the alisphenoid of Mammals. The Ophidian alisphenoid lies in the side wall of the cranium, flush with the inner wall of the otocrane; like that of the Crocodiles, it articulates with the true basicranial bones below, the prootic behind, and the parietal and frontal above. Together with the prootic it bounds a Gasserian fossa in which the ganglion lies. At the outer end of this fossa are the foramen ovale and foramen rotundum. Though the ganglion here lies definitely more within than without the cranial cavity, there can be no doubt that the two alisphenoids are homologous. res Believing that the Cynognathid epipterygoid was a true cranio-mural element, Watson concluded that the anterior portion of the cranial cavity was not homologous with that of other reptiles, but had been added to by epipterygoid cavities. This conclusion rests on the assumed intracranial situation of the ganglion and first branch of the fifth nerve, and is in accord with Gaupp’s previously expressed ideas. It has, however, been just shown that the Cynognathid epipterygoid was in all probability not a true cranio-mural element and the ganglion not intracranial, and it is further contended that the true Reptilian ali- sphenoids are homologous with the Therian alisphenoids. As against this idea, the only features to which importance is attributed are the intracranial situation of the ganglion and nerve in the Theria. But it is surely an error of judgment to lose sight of the fact that how- ever or wherever the cartilaginous precursor of the ventral portion of this bone may be found in mammals, the bone itself in its cranio-mural portion is always developed in the sphenoidal membranous side wall of the brain cavity exactly as in the Crocodiles and (presumably) Ophidians. When, further, it is realised that the relation of the ganglion to the cranial wall is distinctly variable; one must conclude that this is a feature whose value in the present problem has been overestimated. In the recent Lacertilia and Chelonia the ganglion lies always upon the outer side of the cranial wall. In the Crocodilia it lies embedded in the side wall, the first branch of the nerve passing forward through, not outside, the bone. In the Ophidia (Python spilotes) it lies inside the cranial cavity, the first branch leaving the cranium by a veritable foramen rotundum, which is distinct from the foramen prooticum, here a true foramen ovale. In the birds (Anas, Podargus, Dromeus, Gollus, Meleagris, and Corvus) the ganglion appears always to lie on the floor of the temporal fossa on — Homology of Mammalian Alisphenoid and Echidna-Pterygoid 457 the alisphenoid in front of the petrosal, with the foramen rotundum very close to the foramen ovale, and the first branch passing forward through the bone, not upon it. Among the Mammalia there is a good deal of variation in its relation to the bones in the side wall of the cranium. In Echidna it lies on the ala temporalis, the Echidna-pterygoid, and the prootic. In Ornithorhynchus, Watson’s sections show it to lie on the ala temporalis and the anterior process of the prootic. In some Marsupials (Phascolomys and Phascolarctus) it is situated entirely in front of the petrosal; in others (Dasyurus, Tricho- surus, Perameles, and Macropus) it is situated farther back, so as to give J), LACERTILIA 2. CHELONIA 3. Grocopitus 4. Py THON 5. Aves ey 5 Fle: 2m aM. Epi. Pile 6 PHAScoLoMYS 7. DASYURUS 8 PERAMELES 9. CANIS OBESULA St), Pro-o : Pr H ‘Ati, Ay Al Ali. Ali. % Fic, 4,—A series of diagrams illustrating the varying relation of the Gasserjian ganglion and the branches of the fifth nerve to the cranial side wall, Of particular interest are numbers 3, 4, 5, 6, and 7. Epi ptery., epipterygoid; pter., pterygoid ; proot., prootic ; sph. mem., sphenoidal membrane. ‘rise to a shallow furrow on the anterior end of the petrosal. In Tatusia and Edentate the ganglion is situated in a furrow on the medial aspect of the petrosal. In Pteropus I find the ganglion in front of the petrosal. In all other mammals which I-have been able to dissect I find the ganglion partially or wholly medial to the petrosal except in the Primates; in these it lies on the upper aspect of the petrosal. These varying situations are shown in the diagrams fig. 4. In all the Reptiles (except Crocodiles and the Ophidians) the Gasserian ganglion lies lateral to the alisphenoid and sphenoidal membrane, also lateral to the otocrane and in front of it. Among the mammals the ganglion lies commonly medial to the otocrane, and not in front of it. There can be no doubt about the homology of the otocranial elements, therefore it is admitted that somehow the ganglion has come to lie VOL. LI. (THIRD SER. VOL. XIII.)—JULY 1918. 31 458 Dr H. Leighton Kesteven medial to the periotic bones. Why not to the alisphenoid also? The sixth nerve like the first branch of the fifth has changing relation to the cranial floor in reptiles, birds, and mammals. In Chelone midas I have shown (5) that this nerve perforates the post-clinoid process; that is to say, it leaves the cranial cavity behind the pituitary fossa, and therefore runs an extracranial course almost from its origin. Such is the condition in most Reptiles owing to the marked upward tilt of the brain-case forward of the fossa. In birds the sixth nerve very generally runs an intraosseous course lateral to the fossa. It does so in all the forms mentioned above. In the mammals the nerve runs an intracranial course for a considerable distance, usually leaving the cranium in front of the fossa. It is unthinkable that anyone would claim for the mammalian cranium “subclinoid cavities” which are not present in the reptilian, as evidenced by the extracranial situation of the sixth nerve, yet such a claim would be on all fours with a theory that refuses to recognise as homologous the alisphenoids of Ophidians and Mammals on account of the extracranial situation of the first branch of the fifth nerve in the Ophidians. In most mammals the second and third branches of the fifth nerve have little or no intracranial course, but in Perameles .obesula these two pass forward alongside the first branch on the inner side of the alisphenoid to a foramen ovale almost level with the foramen rotundum. This is, however, without significance, for, as in P. naswta, the foramen ovale is, as in other mammals, immediately below the fore end of the ganglion. Being without significance in connection with the homology of the alisphenoids in the two forms, this variation is of interest in showing that a given nerve may be intracranial in one form and extracranial in another, without any such profound difference in the related bones, as is demanded by Gaupp’s theory. It would appear much more in conformity with probability to assume that as the continually widening brain pushed its side wall out, the first branch of the fifth nerve has come to lie internal to the bone by maintain- ing the shortest course to its destination. In the Crocodile, indeed, one seems to be able to recognise the first stage in this process already accomplished. Here the first branch of the nerve is found not outside the bone but tunnelling it from the ganglion in the prootic fossa, directly forward for more than half the width of the bone, In the young (Stage 2, ante) this tunnel is represented by a very obvious groove on the outer aspect of the bone. Once a lamina has been formed on the outer side, it is not difficult to imagine that the bone internal Homology of Mammalian Alisphenoid and Echidna-Pterygoid 459 to the nerve should sooner or later become absorbed under the pressure of the expanding brain. The delicate lamina of bone so constantly present on the two sides of the first branch of the nerve as it lies on the cranial floor in the marsupial cranium, and commonly forming a tunnel for the nerve as it approaches the foramen rotundum, may well be regarded as the remnant of the inner lamina of the alisphenoid. These lamine are certainly evidence of the osteoblastic tissue internal to the nerve, and the most obvious source of such tissue is certainly the primitive side wall. Far from lending support to the epipterygoid-alisphenoid theory, embryology definitely disproves it within the Reptiles, and indicates that whereas the Reptilian alisphenoid is developed ad initio within the sphenoidal membrane from the otosphenoidal plate, the epipterygoid owes any participation in the cranial boundary entirely to secondary acquisition. If, then, it is demonstrable that in the Reptiles, in which class the epipterygoid reaches its maximum development, an alisphenoid has been developed quite independently of that bone, one cannot but ask: Why tax our imagination by supposing that the alisphenoid of Mammals is derived therefrom ? Broom advances the condition observed by himself in Trichosurus, and Watson adds the conditions found by Wineza for the cat, dog, and bear, by Levi, Gaupp, and Fawcett for man, and Noordenboos for the mole, in support of the alisphenoid-epipterygoid homology. These investigators have shown that the extremity of the ala temporalis is distinct from or ossifies independently of the processus alaris. Gaupp has advanced reasons which justify us in regarding the processus alaris as homologous with the Reptilian basipterygoid process. (The restriction of Gaupp’s conclusions to the processus alaris is correctly maintained by Watson (loc. cit.).) In Trichosurus, Broom has shown that the extremities of the alee temporales first appear as small cartilaginous rods lateral to the trabecule and not united in any way with the rest of the chondrocranium ; later these rods become attached to the trabecule, and are continued upward between the two main branches of the fifth nerve, to be later added to by intra-membranous ossification. The condition is compared with that described by himself in Chameleo ; but unless Chameleo differs fundamentally from Lacerta, the two are not comparable, for whereas in Trichosurus there are no cranio-mural anlagen medial to the ala temporalis rod above the basicranial plane, in Chameleo there is the whole of the otosphenoidal fenestrated cartilaginous plate between the epipterygoid and the brain. 460 Dr H. Leighton Kesteven It is in no way surprising that the alisphenoid in Mammals grows upward from the basicranial plane, for it is thus that apparently all cranial elements take origin throughout the vertebrate series above the fish. There is no reason to doubt that the otosphenoidal plate which ossifies to form the alisphenoid in the Crocodile had an origin similar to the oto- sphenoidal plate in Sphenodon— sts is, as an outward and upward growth from the basicranial axis. The independent origin, and separation from the trabecular cartilage, of the ala temporalis may be regarded as resulting from the fact that it is a derivative of the parachordal cartilage which has undergone profound and more or less disintegrative modification as it appears in the Therians. Contrariwise, were it possible to show that the alisphenoid of Mammals originates from the processus alaris, then indeed would it be proven other- wise than homologous with the alisphenoid of Reptiles, for, as already stated, there is reason to regard the processus alaris as homologous with the Rep- tilian basipterygoid process. This latter is developed from the trabecular cartilage, while the Reptilian alisphenoid, ossifying in the otosphenoidal plate, is a parachordal derivative. | It has been stated that the tenia clino-orbitalis of Echidna and Ornithorhynchus are really homologous with and remnants of the rep- tilian cranial side wall. Whether this be so hardly touches the present discussion ; their true homology can only be discussed when we are in a position to decide whether they are ethmosphenoidal (as seems probable), trabecular, or otosphenoidal in origin. The conclusion the foregoing discussion leads us to is that the alis- phenoid bones of the Crocodile, the Ophidia, and the Mammalia are homologous bones, and it follows without argument that the alisphenoid bone of the Birds is homologous with both. Il. THe Homo.ocy or THE ECHIDNA-PTERYGOID. Watson (oc. cit.) remarks of the prototherian skull: “The absence of the alisphenoid in the side wall of the cranial cavity is a very remarkable difference which alone makes the monotreme skull differ far more greatly from that of any Therian than the members of the latter group do among themselves.” He was, however, struck with the similarity of the Echidna- pterygoid and the tympanic wing of the alisphenoid of Marsupials, and he proposed to regard them as homologous structures. In this proposal Watson is undoubtedly correct, and in support thereof descriptions and drawings of the bones are here offered. In Macropus (figs. 5 and 6) the tympanic wing may be described as — Ee Homology of Mammalian Alisphenoid and Echidna-Pterygoid 461 —— Sph. ot, fiss, For. rot. Art. car. int. Ba, sph. ° F. ovale. Nys. VII., VIII. Nvs, IX., X. Nvs, XI., XII. F. ov. Ven. cap. lat. Ty. bone. Ty. cavity. Petr. Par. occ. pr. Ex, oc. Petr. Nvs. XI., XII. Fic. 6.—Macropus. On the right of the figure the tympanic wing has been removed to show the relation of the bones above it. . 462 Dr H. Leighton Kesteven that portion of the bone lying behind the anterior boundary of the foramen ovale. Anteriorly it articulates with the basisphenoid medially, and with the squamosal laterally. Lateral to the foramen it forms a portion of the cranial floor. The inner boundary of the foramen is formed by a spur of the bone. Extending backwards it passes beneath the petrosal, sending dorsally on its inner side a flange which articulates with the inner descending wall of the tympanum. In this region it forms the floor of the tympanum, and articulates with the tympanic bone laterally. Behind _ Pre. sph. Ju. For. ovale, iS Ptery. - Art. car. int. I Bust. Squa. Petrosal, Petrosal. Ex. oce. Nvs. XI., XII. Nvs. IX., X. Fic. 7.—Perameles obesula, The tympanic bulla has been opened in fig. 7, and in fig. 8 its limit is shown by a dotted line. the tympanic bone there is an articulation with the petrosal, and medial to this with the paroccipital process of the exoccipital. In Perameles nasuta, lateral to the foramen ovale the wing forms the whole of the front half of the tympanic cavity including the roof, the temporal tegmen being lateral to this and behind it, a small spur from the petrosal intervening. The tympanic bone is to its outer side. The wing is bounded behind by the petrosal, and there is no articulation with the exoccipital, The condition in Dasywrus maculatus is so similar to P. nasuta as to need no separate description. In Perameles obesula (figs. 7 and 8) the very large size of the ali- sphenoidal bulla tympani has thrown the foramen ovale forward almost —_ a ee ee ee rey MIS AGRE PINES ERO NRE kes : Homology of Mammalian Alisphenoid and Echidna-Pterygoid 463 level with the foramen rotundum. The roof of the bulla supplies an extensive area of the cranial floor, extending from the basioccipital and basisphenoid right to the side wall. On the cranial floor the alisphenoid and squamosal do not meet; they are separated by a spur of the petrosal, which also separates the alisphenoidal bulla from the squamosal tegmen tympani, and at the same time forms the inner half of the roof of the true tympanic cavity, to which the bulla supplies a large antero-medial accessory chamber, the two communicating by a large opening. The pterygoid by a posterior splint comes to articulate with the inner side of the bulla. In Phascolarctus one finds a tympanic bulla equally as extensive as the last, and again the external aperture of the foramen ovale is removed from the petrosal. The “foramen” is here a canal, whose inner aperture, close to the petrosal, is above the bulla, and whose outer aperture is anterior Orb, sph. F. opt. Sph. fiss. — B. sph, -7 Art. car. int. ~ 3 Br, art, car. int. —=— B. occ, Nvs. VIL, VIII. ——— Nvs, XI., XII. Fic, 8.—Perameles obesula, Right half of the cranial cavity. to the bulla. The bulla is laterally flattened, and anteriorly appears as though formed in an enlarged posterior end of the pterygoid wing. The pterygoid bone is applied to the inner side of the bulla. In the Wombat (Phascolomys) a very imperfect bulla tympani is formed by the squamosal, and the tympanic wing of the alisphenoid is represented by a spur which lies behind the foramen ovale. The condition in Phascolomys approximates closely to what is found in all other mammals except some Insectivores which resemble more closely other Marsupials. In Echidna (figs. 9 and 10) the Echidna-pterygoid has the following relations. Anteriorly it articulates with the ala temporalis (which in all probability is really processus alaris only); medial to the articulation with this last, the bone is in contact with the pterygoid, and with the palatine below this along its own inner and forward limit. Lateral to the ala temporalis it forms the median and posterior boundary of the foramen 464 Dr H. Leighton Kesteven ovale, and behind that an appreciable extent of the cranial floor, being limited here by the sphenoidal processus anterior prootici laterally, the For. ovale, Al. temp. Prootic. — B. sph. ae . Tymp. . Nv. VII. Ven. cap. lat. Nvs, IX.-XII. Fen. ro. . Ex, oc Art. car, int. - Bust. Fic, 9.—The alisphenoid has been partly removed on the right (left of figure) to show the articulation with the petrosal in the roof of the tympanum, and articulation with the same bone on the cranial aspect is indicated by a dotted line. ; Eth. Ne: a For. ovale. Ali. Ven. cap. lat. Petr. —_— Nvs, VII., VIII. Nvs, IX., X., xi. 32), * B, occ, ~~ Oce, cond, Fie, 10,—TInner aspect of right half cranial cavity of Echidna, prootie posteriorly, and the basisphenoid medially. It is now continued back beneath the petrosal to articulate with the basioccipital. Its outer — edge is free, and articulates with the tympanic. Medial to the tympanic the bone is excavated to form the inner portion of the floor, the inner wall, Homology of Mammalian Alisphenoid and Echidna-Pterygoid 465 medial portion of the front wall, and a small antero-medial portion of the roof of the tympanic cavity. The rest of the roof is formed by the petrosal, there being no squamosal component of the tegmen. There is no articula- tion between the Echidna-pterygoid and the squamosal owing to the sphenoidal flange of the prootie. The Eustachian tube in the Marsupials issues from the tympanum, with the alisphenoid wing in front to the outer side and below, the petrosal behind to the inner side and above, and the basioccipital or basioccipito- basisphenoid suture to the inner side of the petrosal at the point of departure of the tube. The Eustachian tube in Echidna comes through a notch, with the Echidna- pterygoid in front and below, the petrosal behind and above, and the basi- occipital to the inner side at the point where the tube leaves the two bones. There is no doubt that the palatine in Echidna has acquired its marked extension backward to well behind the basioccipito-basisphenoid suture as an individual and purely secondary variation, so that the Echidna-pterygoid- palatine articulation is quite unimportant in the present connection. That being so, one cannot but be convinced that the similarity of situa- tion and relations between the marsupial tympanic wing and prototherian Echidna-pterygoid is either a most remarkable series of coincidences, or else arises phylogenetically, and the two structures are homologous. The latter is the more reasonable view to take, and no cogent arguments can be advanced against it. The Echidna-pterygoid, then, is an alisphenoid tympanic wing. If, as Watson has contended, such a wing is a primitive feature in the Therian skull—and its presence in Marsupials and Insectivores certainly lends support to such a contention—it was well to impress that fact on our morphology by discarding the meaningless and misleading term “ Echidna- pterygoid” in favour of alisphenoid. To the writer it appears as probable that the prototherian alisphenoid is vestigial of an alisphenoid developed in their common ancestor to the extent to which the Marsupials retain it. Why in the Prototheria the sphenoidal portion became aborted no explanation is offered. I wish to acknowledge my indebtedness to Professor W. A. Haswell and Professor J. T. Wilson, who by lending me from their private libraries have placed the literature within my reach. To Professor S. J. Johnston my thanks are due, not only for the loan of books, but also for the loan of material. Gin Grin, QUEENSLAND, 18th June 1917, 466 Homology of Mammalian Alisphenoid and Echidna-Pterygoid KEY TO LETTERING OF THE FIGURES. Ali. . . . .,Alisphenoid. B.oc.. . . . Basioccipital. Art. car. int. . Internal carotid artery. | Hust.. . . . Groove for cartilagin- B. sph. . . . Basisphenoid. ous Eustachian tube. Eth. . . . . Ethmoid. Fen. ro. . . « Fenestra rotunda. Ex. oce. . . . Exoccipital. Floc. fossa . . Floccular fossa. ge Sarre Foramen opticum. F. ov. and For. \ 0 .amen ogi: For. proot. . . Foramen prooticum. ovale J Front. . . . Frontal bone. For. rot... . Foramen rotundum, Me. 232 3 eee. DUG se ee Orb. sphen. . . Orbito-sphenoid. Oce. cond. . . Occipital condyle. Pal.-. . . ~.' Palatine. Ot. cap. . . . Otic capsule, Petr.. . . . Petrosal. Par. occ. proc. . Paroccipital process. Pre. sph... . Presphenoid. Pl, bas. . . -. Planum basale. Sph. fiss. . . Sphenoidal fissure. Ptery.and Pter, Pterygoid. Sph. memb. . . Sphenoidal membrane. | Sph. ot. iss. . Sphenoptic fissure.! Ty. and Tymp. Tympanic. Squ. and Squa. Squamosal. V.dip. . . . Diploic vein. Ven. cap. lat. . Vena capitis lateralis. Al. temp. . . Ala temporalis. 1 This term is used to designate the composite sphenoidal fissure and optic foramen present in quite a few Marsupials and many Reptiles. REFERENCES. 1. Howes AND Swinnerton, 1901, “The Development of the Skeleton of Sphenodon punctatus,” Trans. Zool. Soc. London, xvi. 2. Broom, 1907, “On the Homology of the Mammalian Alisphenoid Bone,” Rep. South African Assn, Adv, Set. 3. Grecory, 1913, “Critique of Recent Work on the Morphology of the Vertebrate Skull,” Jowrn. of Morph., xxiv. 4, Watson, 1916, “The Monotreme Skull,” Phil. Trans. Roy. Soc. London, Ser, B, vol. cevii. 5. Kesteven, 1910, “The Anatomy of the Head of Chelone midas,” Roy. Soc. New South Wales, xliv. 6. Gaupr, E., 1902, “Ueber die ala temporalis des Siiugerschiidels, ete.” Anat. Hefte, Bd. 19. 1905, “Neue Deutungen auf dem Gebiete der dehre vom Siiugetrinschidel,” Anat. Anz., Bd. 27. 1908, “Zur Entwickelungsgeschichte und _ vergleichenden Morphologie des Schiidels von Echidna aculeata var. typica,” Jenaische Denkschriften, Bd. 6, Teil 2. 1910, “Séugerpterygoid und Echidnapterygoid, ete.,” Anat. Hefte, Bd. 42, ” ” ” —— i ¥ = . a =~ ¥ —— a ee ee ee en THE HEART OF THE LEATHERY TURTLE, DERMOCHELYS (SPHARGIS) CORIACEA. WITH A NOTE ON THE SEPTUM VENTRICULORUM IN THE ,REPTILIA. By Cnas. H. O’DonoGHuE, D.Sc., F.Z.S., Senior Assistant in the Zoology Depart- ment, University College, London. THE Luth or Leathery Turtle is the largest of living Chelonians, and is of considerable interest in comparative anatomy, inasmuch as it is the sole representative of the order Atheca, an undoubtedly primitive group with a number of peculiar characters, the most remarkable being probably the shell, which is quite unlike that of any other form. Its interest is enhanced by its rarity, and>owing to its enormous size when adult, not much is known of the anatomy of its soft parts, the skeleton and muscular system alone being thoroughly known. This being the case, it is desirable that any additions that can be made to our comparatively slight knowledge of this isolated species should be recorded. Furthermore, since the observations here made differ in certain important particulars from the only others on the heart known to me, namely, those of Burne (2), it would appear useful to give another and more detailed account of this important organ. No figures of the heart have been given previously, and the accompanying ones were kindly drawn for me by Miss G. M. Woodward. I have to tender my best thanks to Dr S. F. Harmer, Keeper of Zoology in the British Museum of Natural History, for his courtesy in granting me facilities for examining the heart of a specimen recently acquired for the national collection. The animal in question was a large female measuring 7 feet 1 inch from snout to tip of tail, and weighed 93 cwt. It was caught somewhere off the Scilly Islands on June 7, 1916. References to certain points in the anatomy of Dermochelys are not frequent in zoological literature, and the first appear to be some on the alimentary canal and related viscera by Temminck (14) in 1836. These were added to considerably by Rathke (13) ten years later, who gave a description of the trachea, cesophagus, and stomach. The same author later (12) describes some of the muscles and also certain veins in a very young specimen, and this is, so far as I know, the first account of any part of the circulatory system, but does not bear on the present observations, The 468 Dr Chas. H. O'Donoghue muscles, more particularly those of the fore limb and girdle, were dealt with by Fiirbringer (3), and again Vaillant (15) contributed a more stig 5 os account of the abtaenbity: canal and method of feeding. The last worker was Burne (2), who has described with accuracy and some detail a number of supplementary points in the general visceral anatomy. The muscles in particular are fully dealt with, so that the muscular system is now quite well known and a number of useful additions made to the previous accounts of;the alimentary canal. What concerns us more especially is that he has described certain veins and for the first time ‘given a description of the heart. Before passing on to the detailed description of that organ it will be as well to indicate the points in which the present observations differ from those made by Burne. 1. This author describes the heart as being “somewhat long and narrow,” whereas the present specimen was decidedly short and broad, and indeed in general proportions is similar to that of other Chelonians, e.g. that of Chelydra serpentina, as figured by Fritsch (4). 2. The apex of the ventricle is stated to taper to form a long, stout gubernaculum cordis, while I found the apex of the ventricle very bluntly rounded and the gubernaculum short, as is clearly shown in the accompany- ing text-figure. This may be due to the difference in age of the two specimens, since that employed by Burne was considerably younger than mine. It was undoubtedly a quite immature specimen, for its total length was only 135 cm. as against 7 feet 1 inch, and also the oviducts were imperforate; a similar condition of the heart apparently occurs in the young crocodile (C. acutus). Then, too, the diameter of the aortic arch on each side when flattened is given as 2 ems. instead of 3°5-4cm. A similar difference is recorded in Chelone by Rathke (12), who found that in the embryo the heart is long as it is in most other Reptilia, while in the adult it is very broad as in the Chelonia in general. 3. The walls of the auricles are said to be “ very thin in comparison with Chelone midas, and show very little trabecular structure.” In the present specimen the trabecular structure is quite strongly marked and the walls appear actually thicker, 7.¢., taking into account the difference in size, relatively quite as thick as in C. midas. 4, In both specimens the actual cavity in the ventricle is astonishingly small, but the ventricular wall forming the whole of the lower half or two- thirds of the ventricle was found to be extremely spongy, and blood coagula were present in its lacune all through it right down to the apex. Burne, on the other hand, found this part of the heart to consist “practically of solid muscle,” ee de ee eT Pe ._ The Heart of the Leathery Turtle 469 These two points (3 and 4) again may perhaps be due to the one animal being so much younger than the other. 5. The left auricle is noticeably smaller than the right, but the difference is not so great as to describe it as being “relatively very small—not more than a quarter the size of the right.” The state of distension of the auricles at the time of death probably influences their apparent size to a very appreciable extent. 6. A further point to be noted, for which it is more difficult to find an explanation, is the statement that “the carotids, 7 em. above their origin from the innominate artery, suddenly dilate to at least twice their original diameter and then very gradually narrow again towards the head.” Unfortunately, when the heart was removed, before I had an opportunity of examining it, the great arterial and venous trunks were all cut off fairly close to the heart, so that I have not been able to confirm the presence of this decidedly unusual condition. In the portions of the carotids present, 85 cm. in length, there is no indication whatever of any dilatation. In other particulars I am able to confirm Burne’s description. The heart of the present specimen of Dermochelys (vide text-fig. 1) is a very large massive organ whose general form is undoubtedly Chelonian and corresponds most nearly to that of Chelydra serpentina, with which it agrees fairly closely, as a comparison between the figures given here and those of Fritsch (4) will show. A very well-developed sinus venosus is present situated to the right of the middle line, and opening, as always in the Reptilia, into the right auricle. The sinus was actually cut into in removing the heart, with the result that it is not quite complete. The two pre-caval veins, the right and the left, are intact, but the post-caval is missing, together with a portion of the dorsal wall of the sinus. The pre-caval veins are very large, the left being 6 em. in diameter when flattened, and the right about 1 em. larger. The con- formation of the remaining parts of the sinus, however, are such that they indicate quite clearly the position of the post-caval vein, which has there- ‘fore been restored in the drawing for the sake of completeness. It opened quite close to the left pre-caval vein at the left postero-lateral corner of the sinus. In Chelydra serpentina, Fritsch (l.c.) figures two hepatic veins opening into the base of the post-caval, and in addition two quite large independent hepatic veins opening directly into the sinus venosus at its right antero-lateral border only a short distance from the base of the right pre-caval vein. This part of the sinus is absent in my specimen, so that it is not possible to say definitely whether they are present in Dermochelys or not. If they are present, however, it is evident that they were relatively very much smaller than in Chelydra, and for this reason they have been Si, Fic, 14.—Ventral view of the heart and main vessels of Dermochelys. Ad nat. Drawn by Miss G. M. Woodward. L.C. iL: Su. bs: LP.— L.A. LiP.V. V.C.S. RPV. Via, 1p.—Dorsal view of same, Ad nat. Drawn by Miss G, M. Woodward, C.8,, coronary sulcus; C,V., coronary vein; G.C., gubernaculnm cordis; L,A., left auricle; L.C,, left carotid artery ; L.P., left pulmonary artery; L.P.V., left pulmonary vein; L.8,, left systemic arch ; ete left subclavian artery ; f Ug ig arch; P.V., pulmonary vein; R.A., right auricle ; , right pulmonary artery; R.P.V., right pulmonary vein; K.S., right systemic artery; B.S.C,, right systemico-carotid trunk; R.Su., right sub- clavian arte 8.V., sinus venosus; V., ventricle; V.C.D., right pre-caval vein ; V.0, +) Dost. caval vein; V.C.8., left pre-caval vein, ., Tight carotid artery; rig The Heart of the Leathery Turtle 471 omitted from the drawing. Into the sinus venosus, to the right of the left pre-caval vein and posterior to the post-caval, opens the coronary vein, which appears as a very tough vessel about 1°5 em. in diameter, passing across the coronary sulcus. | The sinus venosus opens into the right auricle by a sinu-auricular aperture about 12 cm. in length starting at the right posterior corner of the auricle and running forwards and outwards at an angle of about 45° to the long axis of the heart. It is guarded by two valves along its sides which are continued on along the auricular wall for a short distance beyond the opening as flaps. As pointed out above, there is a decided but not great difference in size between the auricles. They are separated by a well-developed septum auriculorum, which is complete in Dermochelys and not perforated as it is stated to be by Huxley (8, p. 264) in some Chelonia. The left auricle is a large sac whose actual size varies with its distension, but in the specimen measures about 1410°5 cm. It is thin-walled as compared with the ventricle, but considerably stouter than the sinus venosus, and in its postero-dorsal corner reaches a thickness of about 1°5 em. Inside, particu- larly on its dorsal aspect, it has a number of well-marked muscle strands, the musculi pectinati. The ,two pulmonary veins open into the middle of its median dorsal border by a single small aperture not guarded by any well-marked valves. The right auricle, as already noted, is considerably larger, measuring in the specimen. 175 x15 em., and is also thicker walled, Its walls in the latero-dorsal region attain a thickness of about 2°5 cm., and are plentifully supplied with muscle strands. The position of the sinu- auricular aperture with its large valves has been described. The ventricle is an extremely stout sac of a blunt, rounded triangular shape, measuring along its base by the coronary sulcus about 24 cm. and from this border to the apex about 17 cm. The walls are enormously thick and muscular, reaching a thickness in the apical region of as much as 10-11 em. They are entirely composed of a number of muscular trabecule which are woven together to form a loose, spongy tissue in whose spaces are blood-clots. No traces of definite column carnez or musculi papillares were found, and if present at all must have been small and removed in clearing the large solid clot with which it was filled. To the apex were attached numerous tough, stout muscular bands, some separate and others united by a binding tissue constituting a very well-developed gubernaculum cordis such as has been described by Beddard (1) in Tupinambis, and stated by him to be general in Lacertilia, and also noted by other writers in other species of reptiles. This as it were ties the ventricle to the pericardium, by a reflexion of which it is covered. 472 Dr Chas. H. O'Donoghue The two auricles open independently into the ventricles by quite large slits at their postero-mesial corners; and although the apertures are fairly close together, they are absolutely separated from one another by enormously strong moderately complex valves. Guarding the left auriculo-ventricular aperture is a strong oval valve about 1 em. thick, of almost cartilaginous consistency. It does probably contain cartilage, but I did not wish to damage the specimen more than necessary. The valve is attached to the latero-ventral wall, its free edge hanging down into the ventricle. This causes the aerated pulmonary blood entering the ventricle to be sent across to its left side. Parallel with this is a similar but stronger and much larger valve guarding the right auriculo- ventricular orifice. Its position agrees closely with the partly muscular, partly cartilaginous septum in the turtle Chelone midas as described by Huxley (8), and so we can recognise a space on either side of it which that author has termed the cavum arteriosum on the left and the cavum venosum on the right. This description is quite accurate, as can be seen at once by the dissection of a heart of Chelone. This valve in Dermochelys is unlike that on the left or that in the turtle in that it is curled upon itself to form a deep trough whose concavity is directed anteriorly. ‘lhe non-aerated venous blood enters in front and to the left of this valve, and is consequently forced out to the left side of the ventricle. Again as in the turtle, a strong thick band.of the ventricular muscle, constituting a distinct: though incomplete ventricular septum, is also present, and, running more transversely than the valves, is so arranged that when the ventricle is contracted it practically cuts off a portion of the cavum venosum, and this chamber has been termed by Huxley (i.c.) the cavum pulmonale. ‘The cavity of the ventricle is thus divisible into three portions: a right ventro-lateral cavum venosum, a left dorso- lateral cavum arteriosum, and an almost entirely separate division of ' the former, the cavum pulmonale. The’ anterior end of the incomplete septum is also tough like cartilage, and under it, 7.c. ventral to it, is situated the opening of the pulmonary trunk, so that the venous blood is poured into the heart quite close to the aperture of this vessel. Clear of the septum, i.e. lateral to it, lies the opening of the left systemic arch. This is to the side where the septum is not complete, and therefore when the ventricle is distended it is near the region where the two bloods mingle, and so may receive aerated as well as non-aerated blood. When the ventricle is contracted, however, it opens from the small right ventro-lateral chamber, as does the pulmonary arch. The large common trunk giving rise to both the right systemic and ‘The Heart of the Leathery Turtle 473 the carotid arches, on the other hand, opens from the ventricle completely on the dorsal side of the septum, and so from the remaining part of the cavum pulmonale, but still in that cavity and on the right side of the right auriculo-ventricular valve. Thus it follows that no arterial trunk comes off from the cavum arteriosum, and the aerated blood must pass through the corner of the cavum pulmonale on its way out. The right systemic and carotid trunk is, then, more or less completely separated from the opening of the other arterial trunks during ventricular systole. During diastole, in spite of its opening lying in the cavum venosum, the Dorsac. VENTRAL. Fic. 2.—Diagram of the valves, septum, and ventricular apertures of : Dermochelys, viewed from the posterior aspect. I.8., incomplete septum; L.A.V., left auriculo-ventricular aperture under L.A.V.L., valve of same; L.S., opening of left systemic trunk; P., opening of pulmonary trunk; R,A.V., right auriculo-ventricular aperture under R.A.V.I., valve of same ; R.S.C., opening of right systemico-carotid trunk. trough-like right auriculo-ventricular valve directs the non-aerated blood away from the systemico-carotid trunk, while there is nothing to prevent it receiving blood from the left auriculo-ventricular opening. There is little doubt, then, that the non-aerated blood from the cavum venosum leaves mainly by the pulmonary arch, and the aerated blood practically entirely by the right systemico-carotid trunk: The left systemic appears to convey in the main non-aerated blood, but there is no reason to suppose that it does not also have a slight admixture of aerated blood. It is therefore obvious that functionally there is a fairly -complete separation of the two blood-streams in the ventricle. In his excellent description of the turtle heart, which I have followed to a certain extent above, Huxley makes the main division of the ventricle, VOL. LII. (THIRD SER. VOL. XIII.)—JULY 1918. 32 474, Dr Chas. H. O'Donoghue that by the auriculo-ventricular valve, into cavum arteriosum and cavum venosum. It would, I think, be better, for reasons that will appear obvious in what follows, and also on account of its functional importance, to throw the stress more on the division brought about by the ventricular septum. Thus we have a cavum pulmonale, from which only the non- aerated blood leaves the heart, and a cavum systemico-caroticum, from which practically only aerated blood leaves. The other division into cava arteriosum et venosum is to be looked on as a functional complica- tion of the auriculo-ventricular valve necessitated by the incompleteness of the true ventricular septum, and not of importance in the future complete subdivision of the ventricle. The bases of the arterial trunks are guarded in a typical manner by semilunar valves. The two subclavian arteries arise from the corresponding carotids quite near their base. The heart of Dermochelys, then, is typically Chelonian, and agrees fairly well externally with that of Chelydra serpentina, and internally i is much like Chelone midas. , THE SEPTUM VENTRICULORUM. During the foregoing investigation of the heart of Dermochelys, the hearts of a number of other Reptilia, in my own possession, at University College, the Natural History Museum, and the Royal College of Surgeons, were examined for purposes of comparison. Furthermore, the heart of the Tuatara (Sphenodon punctatus), of which animal I hope shortly to publish a full account of the circulatory system, was also examined in a series of specimens kindly placed at my disposal by Professor A. Dendy, F.R.S., to whom I wish to tender my thanks. Asa result of these inquiries, several interesting facts came to light, and it soon became evident that the state- ment of Goodrich (6), who lays some stress on the position of the septum ventriculorum in his scheme of classification of the Reptilia, was incomplete, and indeed, in respect of certain members of that class, somewhat misleading. This author states (p. 271) that “. . . in the Reptilia the interventricular septum tends to divide the chamber into a left cavity leading to the base of the systemic arch, and a right cavity leading to the base not only of the pulmonary but also of the left systemic arch”; and further indicates the same idea graphically in a very clear diagram on p. 273, which he states applies to Reptilia in general, excluding the Crocodilia. In order to clear up this point, if possible, the relation of the openings of the aortic arches to the septum ventriculorum was examined in the following animals ;— The Heart of the Leathery Turtle 475 RHYNCHOCEPHALIA: Sphenodon punctatus. LAcerTILIA: Varanus salvator, Heloderma horridum, Lacerta viridis, L. agilis, Phrynosoma ‘cornuta, Seincus tridactylus, Chameleo sp. ? Tiliqua scincoides. OpuipiA: Python tigris, Boa constrictor, B. murina, Crotalus horridus, Tropidonotus natria. CHELONIA: Chelone midas, Testudo greca, Dermochelys coriacea. CrocopiLia : Crocodilus americanus, Crocodilus sp.? (a smaller speci- men), Caiman sp. ? It soon became evident that each group possessed its own peculiar and more or less characteristic arrangement of the arches and septum. They will therefore be treated briefly in groups from the most to the least specialised. CrocopiL1A.—In the first place, it is only in the Crocodilia that a com- plete septum ventriculorum is present. Here, taking Crocodilus americanus as a typical representative, we find the septum is quite complete, and it runs from the right latero-dorsal to the left latero-ventral wall of the ventricle, although it is almost in the dorso-ventral plane. The two ventricular cavities are thus a right lateral, slightly ventral in position, and a left lateral, slightly dorsal. The common trunk of the right systemic and carotid arches, which for convenience may be termed the right systemico- carotid, comes off the left chamber, into which the left auricle pours the aerated blood; while the left systemic and the pulmonary arches come off from the right chamber, and so convey nothing but non-aerated blood derived from the right auricle. This condition, as pointed out by Goodrich, will lead directly to that in birds by the loss of the left systemic arch. Similar relations were found in Crocodilus sp.? and Caiman sp. ?, and also in Alligator Lucius, as described by Gegenbaur (5, p. 387). CHELONIA.—The condition in Dermochelys has been fully described above, and it will be seen that, although the septum is incomplete, yet it bears the same relation to the pulmonary and right systemico-carotid arches as in Crocodilia. The difference, due to the incompleteness of the septum, is that, as the opening of the left systemic arch lies right opposite to the free end of the septum, there is always the possibility that a certain amount of aerated blood may enter it during diastole. In the crocodile, of course, this is quite impossible. The ventricle of Testudo and of the turtle, as clearly stated by Huxley (l.c.), is similarly divided, and the relations of the septa and vessels are indicated diagrammatically in the accompanying figure. OpHIDIA.— Python tigris may be taken as a representative of this group. 476 Dr Chas. H. O'Donoghue The interventricular septum is very well developed and attached to the left dorso-lateral side of the ventricle, running from base to apex and directed towards the right ventro-lateral wall. Thus it runs in a direction entirely different from that of the septum in either Crocodilia or Chelonia— in fact, nearly at right angles to it,—and divides the ventricular cavity, at any rate in systole, into a right dorso-lateral and a left ventro-lateral L.3, R.S3.¢. Fic, 3,+-Diagrammatic representation of the arterial trunks (I.) close to the ventricle, (I1.) near the top of the auricles, in (A) Crocodilus americanus, (B) Chelone midas, (C) Python tigris, (D) Varanus salvator, and (E) Sphenodon, LS8,, Incomplete septum; L.C., left carotid artery; L.P., left pulmonary artery ; L.S., left systemic arch; P., pulmonary arch; R.C., right ton artery ; K.P., right pulmonary artery; R.S., right systemic artery; R.S.C., right ' systemico-carotid trunk; 8 , complete septum, chamber. It is tough and fibrous at the anterior end, but becomes somewhat spongy like the ventricular wall near the apex, and it is possible that the ~ two chambers may be in communication one with another even during systole through these pores. Another great difference between this and the foregoing two groups is the position of the arterial apertures. From the right dorso-lateral chamber come off not only the right systemico- carotid but also the left systemic arch, Only the pulmonary arch, on the od The Heart of the Leathery Turtle 477 other hand, comes off from the left ventro-lateral chamber. ‘The openings of the right and left systemic arches lie close together, only separated by the union of their walls to form a common band; whereas the pulmonary aperture is quite distinct, and situated on the other side of the incom- plete septum. Both auricles open into the right latero-dorsal chamber, the left auricle opening fairly dorsally about the middle of its anterior end, and so it lies a short distance from the openings of both the right systemico-carotid and the left systemic arches. Thus the aerated blood can pass straight from auricle to these trunks. The right auriculo-ventricular aperture lies in the most ventral corner of the right lateral chamber, quite close to the openings of’ the right systemico-carotid and left systemic, and also to the top of the incomplete septum. It is, however, guarded by a large bi-lobed valve running somewhat transversely. The anterior flap of the valve, when open during auricular systole, partly or perhaps completely closes the systemic arches ; while the posterior valve is of a somewhat curious spout shape that directs the non-aerated blood across the septum to the left lateral chamber. Although the exact modus operandi of the right auriculo-ventricular valve cannot be determined in dead hearts, there seems little doubt that during auricular systole the major part of the non-aerated blood is sent into the left lateral chamber, from which the pulmonary arch arises. The right lateral chamber, on the other hand, is filled mainly with aerated blood, and may perhaps receive any overflow of non-aerated blood from the smaller left. chamber. Similar conditions also obtain in the other snakes examined, and the external twisting of the arterial trunks in 7’ropidonotus natrix has already been pointed out (9). | LacertitiA.—A large heart of Varanus naleaion in the collection of . the Royal College of Surgeons was examined, and can be regarded as more or less typical of the group. ‘The septum here, as in the Ophidia, runs from the left latero-dorsal wall of the ventricle across towards the right latero- ventral wall, but is decidedly more dorso-ventral in position than in the Ophidia. As in that group, however, it is incomplete, so that it is only during systole that the ventricular would appear to be completely divided into a right lateral, slightly dorsally situated chamber and a smaller left lateral chamber lying slightly ventrally. The left chamber again gives off only the pulmonary artery, while the right systemico-carotid and left systemic arches both come off from the larger right chamber. Into this chamber also, as in Ophidia, both the auricles open. Their valves are much as in Python—i.e. the left opens centro-dorsally, while the right opens in the ventral corner, and by means of its valves discharges its non- * 478 Dr Chas. H. O'Donoghue aerated blood past the openings of the right systemico-carotid and left systemic arteries across the septum into the left chamber, and so to the pulmonary arch. Parker (11, p-. 174) states that in the lizard, Lacerta viridis, the two aortic openings lie on the right, and are separated from the pulmonary opening on the left by a muscular partition. This has been verified, and is also true of L. agilis. A similar condition is shown by Wiedersheim (16), in a figure on p. 407, in L. muralis. In these Lacertide, however, the partition is in a less developed condition than in Varanus. For the information regarding Tiliqua scincoides I am indebted to Professor J. P. ° Hill, F.R.S.: it is typically Lacertilian. The other species also conform to this general plan, but there appears to be a greater variation in the degree of development of septum ventriculorum among the Lacertilia than in other Reptilia—not an unexpected thing in view of the fact that they are un- doubtedly a less specialised group than any other of the reptiles, with the exception, of course, of the Rhynchocephalia. Thus the conditions in Ophidia and Lacertilia are quite different from what is implied by Goodrich both in the extract given above and in the diagram. The ventricle in these two groups, containing by far the greater number of living species of reptiles, is indeed partially divided into a right and left chamber, but the two systemic arches come off from the right side and the pulmonary arch alone comes off from the left. There is thus a considerable difference in the relation of the septum to the arterial trunks between the Ophidia and Lacertilia on the one hand and the Crocodilia and Chelonia on the other. Not only is there a greater twist on the arterial trunks, which leave the top of the heart in relatively the same position, while the pulmonary leaves the ventricle more to the right in the latter, but the septum is actually situated on opposite sides of the pulmonary artery. In Ophidia and Lacertilia it lies between the pulmonary and left systemic, while in Crocodilia and Chelonia it lies between the pulmonary and right systemico-carotid. Another striking and important difference which has not been empha- sised formerly is that, whereas in the Crocodilia and Chelonia, as in birds and mammals, the aerated blood is poured into the left side of the ventricle, in Ophidia and Lacertilia the reverse is the case and the aerated blood passes into the right ventricular chamber. RHYNCHOCEPHALIA.—In view of the fact that the aortic arches in Sphenodon are, as has been shown recently (10), in a more primitive condition than in other reptiles, it was hoped that an examination of the ventricle might throw some light on the question of the septum, Four hearts were dissected, and two were examined in serial sections. The first The Heart of the Leathery Turtle 479 series had been cut at 104 by Professor Dendy, and the second was cut for this purpose at 20%. They have indeed shown a simple condition. There is no septum ventriculorum comparable with that in the other Reptilia; although the ventricular wall is irregular, no fold appears to correspond with this septum. Moreover, though the exact structure has not yet been satisfactorily determined, it appears as if the three arterial arches come off from a sort of extension of the ventricle in which extensions of their ends form a sort of valve. If this should prove to be the case, it may be that we are here dealing with a rudimentary conus, a structure well represented in the amphibian heart, but which Greil (7) has shown is absorbed into the ventricle in the Reptilia. The state of preservation of the interior of none of the hearts has allowed me to speak definitely on this point. The pulmonary arch lies on the left of the other two openings which appear to come off from a short common stem situated to the right. Thus we meet with a condition more simple than in other Reptilia, but which, nevertheless, is distinctly reptilian. It also resembles the Lacertilia and Ophidia more closely than the other two groups, since, although the means are not clear, it is obvious that the non-aerated blood leaves the right side of the heart and the aerated hlood the left side. Greil (/.c., p. 227), in an instructive paper on the development of the heart in the Vertebrata, concludes: “ Wie bereits in der Einleitung bemerkt wurde, zeigen die Herzen der untersuchten Reptilien nur unterschiede graduelle Natur. Diese unterschiede betreffen vor Allen den Bau und die Ausbildung der Scheidewiinde im Inneren der Kammer, und es lassen sich die Herzen der Reptilien in dieser Hinsicht gewissermassen zu einer Entwicklungsreihe zusammenstellen, an deren Aufang das Herz von Hatteria und der niederen Saurier, und deren Ende das Herz der Crocodilier zu stehen kommt.” This passage is correct to a certain extent, but it certainly does not indicate sufficiently clearly the primitive- ness of the heart in Sphenodon, in which the septum does not appear to have started; nor does it notice the fact that the position of the septum when present varies considerably in different groups. In the Reptilia, then, we have two distinct types of heart. The one, found in Chelonia and Crocodilia, in which the aerated blood is found on the left side of the septum whether it is complete or not, can easily be regarded as leading to the condition in birds. The other is character- istic of Ophidia and Lacertilia, and has the non-aerated blood on the left side of the septum, and thus, as Goodrich points out—although his state- ment of the relation of septum to the arches is open to objection,—cannot in any way lead to the condition in a mammal, where of course the non- aerated blood is in the right ventricle. 480 The Heart of the Leathery Turtle . The Rhynchocephalia, in as far as the position of the aerated blood leaving the heart, are as it were definitely committed to a Lacertilian as opposed to a Chelonian or Crocodilian line of development, yet retain certain primitive features. The absence of a septum is one point, and the fact that the two systemic arches appear to come off from a short common trunk; for, as Goodrich remarks (p. 271), “the Theropsidan and Sauropsidan types must have evolved from more symmetrical primitive type, in which the ventricle and aortic arches were both single.” Here, then, in Sphenodon, although still removed from this ideal condition, we find undoubted hints as to its existence and a certain amount.of approxi- mation to it. LITERATURE. (1) Bepparp, ‘Contributions to the Anatomy of the Lacertilia: (2) Some Points in the Structure of the Tupinambis,” Proc. Zool. Soc., vol. i., 1904. (2) Burne, “ Notes on the Muscular and Visceral Anatomy of the Leathery Turtle (Dermochelys coriacea),” Proc. Zool. Soc., vol. i., Aug. 1905. (3) Firprinerr, “Zur vergleichenden Anatomie der Schultermuskeln,” Jena, Zeitschr., Bd. viii., 1874. _ (4) Frrrscn, “Zur vergleichenden Anatomie der Amphibienherzen,” Arch. f. Anat., Physiol., u. Wiss. Med., Reichert and Du Bois-Reymond, 1869. (5) Geeensaur, Vergleichenden Anatomie der Wirbelthiere, Leipzig, 1901. (6) Gooprica, “On the Classification of the Reptilia,” Proc. Roy. Soc., B, vol. 89, 1916. i : (7) Gre, “ Beitriige zur vergleichenden Anatomie und Entwicklungsgeschichte des Herzens und des Truncus arteriosus der Wirbelthiere,” Morph. Jahrb., xxi., 1903. . (8) Huxey, The Anatomy of the Vertebrated Animals, London, 1872. (9) O'Donocuur, “The Circulatory System of the Common Grass-snake (Tropidonotus -natrix),” Proc. Zool. Soc., Sept. 1912. (10) O’Doxocuusg, “A Note on the Ductus caroticus and Ductus arteriosus and their Distribution in the Reptilia,” Jowrn. of Anat., vol. li., pt. ii., 1917. (11) Parker, Comparative Anatomy of Vertebrates, London, 1884. (12) Raruxe, Veber die Entwicklung der Schildkroten, 1848, (13) Raruxe, “Ueber die Luftréhe, die Speiserréhe und den Magen der Sphargis coriacea,” Archiv f. Anat. u. Physiol., 1846. (14) Temminck, Fauna Japonica (Reptilia), 1836, (15) Vatnant, “Remarques sur l’appareil digestif et le mode d’alimentation de la Tortue luth,” Comptes Rendus Acad. Set., t. exxiii., 1896. (16) Wiepersuem, “Comparative Anatomy of the Vertebrates,” English edition, 1902. ae INDEX Alisphenoid, homology of the mammalian, and of the echidna-pterygoid, H. Leighton Kesteven, D.Se., M.B., Ch.M., 449. Allis, E, Phelps, junr., the muscles related to the branchial arches in Raia elavata, 383. Arterial valve, congenital stenosis of the pul- monary, with patent foramen ovale, in an old man: with remarks on the valvular factor in cardiac action, Alexander Morison, M.D., F.R.O.P., 241, Blakeway, H., M.S., F.R.C.S., a hitherto un- described malformation of the heart, 354. Blood-vessels, the earliest stages of development of the, and of the heart, in ferret embryos, Chung-Ching Wang, M.D., Ch.B. (Edin.), 107. Cetacea, reproductive organs of, Professor Alexander Meek, 186. Cockayne, E. A., D.M., F.R.C.P., and R, J. Gladstone, M.D., F.R.C.S., F.R.S, Ed., case of accessory lungs associated with hernia through a congenital defect of the dia- phragm, 64. Cranial cavity, relations of the glossopharyngeal nerve at its exit from the, E. Joyce Partridge, M.R.C.8., L.R.C.P., 332. Cranium, primordial, of Hrinaceus europeus, Professor Edward Fawcett, 211. — of Pecilophoca Weddelli (Weddell’s seal) at the 27-mm. C.R. length, Professor Edward Faweett, M.D. Edin., 412. Fawcett, Professor Edward, primordial cranium . of Hrinaceus ewropeeus, 211. —— primordial cranium of Pecilophoca Wed- delli (Weddell’s seal), at the 27-mm. C.R. length, 412. Femur, some indices and measurements of the modern, J. R. D. Holtby, M.D., Se.B., 363. Fibula, reduction of the mammalian, Thomas Walmsley, M.D., 326. Gladstone, R. J., M.D., F.R.C.S., F.R.S. Ed., and E. A. Cockayne, D.M., F.R.C.P., case of aecessory lungs associated with hernia through - a congenital defect of the diaphragm, 64. Glands, internal mammary lymphatic, E. Philip Stibbe, 257. Heart, a hitherto undescribed malformation of the, H. Blakeway, M.S., F.R.C.S., 354. —— multiple anomalies in a human, Hernani Bastos Monteiro, 335. —— of the leathery turtle, Dermochelys (Sphargis) coriacea. With a note on the septum ventriculorum in the Reptilia, Chas. H. O’Donoghue, D.Se., F.Z.S., 467. Holtby, J. R. D., M.D., Se.b., some indices and measurements of the modern femur, 363. Jones, Professor F. Wood, sublingua and the plica fimbriata, 345. and Miss Esther Rickards, sexual characters in twin goats, 265. abnormal Kesteven, H. Leighton, D.Sc., M.B., Ch.M., homology of the mammalian alisphenoid and of the echidna-pterygoid, 449. Kitten, study of an opodymous, J. A. Pires de Lima, M.D., 276. Lander, Miss Kathleen F., B.Sc., examination of a skeleton of known age, race, and sex, 282. the pectoralis minor: study, 292, Lima, J. A. Pires de, M.D., study of an opodymous kitten, 276. Lungs, accessory, associated with hernia through a congenital defect of the diaphragm, E. A. Cockayne, D.M., F.R.C.P., and R, J. Glad- stone, M.D., F.R.C.S,, F.R.S, Ed., 64. Mackenzie, W. Colin, M.D., F.R.C.S. Ed., muscular action, 343. Meek, Professor Alexander, reproductive organs of Cetacea, 186. Monteiro, Hernani Bastos, multiple anomalies in a human heart, 335.» Morison, Alexander, M.D., F.R.C.P., congenital stenosis of the pulmonary arterial valve, with patent foramen ovale, in an old man: with remarks on the valvular factor in cardiac action, 251. Muscle, human soleus, structure of the, Douglas G. Reid, M.B., Ch.B. Edin., M.A. Trin. Coll. Camb., 442. observations on the omohyoid, Thomas Walmsley, M.D., 319. a morphological VOL. LII. (THIRD SER. VOL. XIII.)—JULY 1918. 33 ~ 482 Muscles related to the branchial arches in Raia clavata, E. Phelps Allis, jun., 383. Muscular action, W. Colin Mackenzie, M.D., F.R.C.S, Ed., 348. Nerve, relations of the glossopharyngeal, at its exit from the cranial cavity, E. Joyce Partridge, M.R.C.S., L.R.C.P., 332. O’Donoghue, Chas. H., D.Sc., F.Z.S., heart of the leathery turtle, Dermochelys (Sphargis) coriacea. With a note on the septum ventri- culorum in the Reptilia, 467. Opodymous kitten, study of, J. A. Pires de Lima, M.D., 276. Partridge, E. Joyce, M.R.C.S., L.R.C.P., rela- tions of the glossopharyngeal nerve at its exit from the cranial cavity, 332. Pectoralis minor: a morphological study, Miss Kathleen F. Lander, B.Sc. , 292. Reid, Douglas G., M.B., Ch.B. Edin., M.A. Trin. Coll. Camb., structure of the human soleus-muscle, 442. Rickards, Miss Esther, and Professor F. Wood Jones, abnormal sexual characters in twin goats, 265. Index Sexual characters in twin goats, abnormal, Miss Esther Rickards and Professor F. Wood Jones, 265. Shaw, D. Mackintosh, form and function of teeth : a theory of ‘maximum shear,” 97. Skeleton of known age, race, and sex, examina- tion of, Miss Kathleen F, Lander, B.Sc., 282. Stibbe, E. Philip, internal mammary lymphatic glands, 257. eae Sublingua and the plica fimbriata, Professor F. Wood Jones, 345. Teeth, form and function of: a theory of ‘*maximum shear,” D. Mackintosh Shaw, 97. Tetrapod shoulder girdle and. fore-limb, Lieut. D. M. S. Watson, M.Sc., R.N.V.R., 1. Turtle, heart of the leathery, Dermochelys (Sphargis) coriacea, With a note on the septum ventriculorum in the Reptilia, Chas, H. O’Donoghue, D.Sc., F.Z.S., 467. Walmsley, Thomas, M.D., observations on the omohyoid muscle, 319, —— reduction of the mammalian fibula, 326. Wang, Chung-Ching, M.D., Ch.B. (Edin.), earliest stages of Sevaene of the blood- vessels and of the heart in ferret embryos. 107. Watson, Lieut. D. M. S., M.Se., R.N.Y.R., evolution of the tetrapod shoulder girdle and fore-limb, 1. D 2850 4 VRINTED IN GREAT BRITAIN BY NEILL AND CO,, LTD , EDINBURGH, t ¥ —~ :~ a = Me ae ee ae Se is J sf 4 He tify te ~ ae! ut 1S fd . tet hie Baa e RA $ pte ty Nigh Pateach ath ity ah tee “bid ite Tes ap tes AR A Ss oui DOs ny i heh talee . a et a doe tl an pee Ee ely ie A ‘fret 44) it Lk a ath te ti i yates ‘1 ures ei pepht pre +s - Af WE, S nists tele ed 5 Nya 7) . +fie | we D ath, ; ii iat? yry } ftw Ayyhh rare ort > lias BP's big wit a P <>