eowate Aeris ey Cheek ee, aire a ie os oe Ss ane * ia 4 te f mee ata emer Mtoe euhiy 61 chewiree: Nee eee Le THIS BOOK MAY NOT BE PHOTOCOPIED TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. VOLUME xX. LONDON: PRINTED FOR THE SOCIETY : SOLD AT THEIR HOUSE IN HANOVER-SQUARE; AND BY MESSRS. LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. 1879. eecaes % wigs aka Pe Rae CONTENTS. I. On the Axial Skeleton of the Struthionide. By St. George Mivart, F.R.S., Sec. L.S., Professor of Biology at University College, Kensington . . . . pagel II. On the Ancient or Quaternary Fauna of Gibraltar, as exemplified in the Mammalian Remains of the Ossiferous Breccia. By Guorce Busx, F.R.S., F.Z.S., &e. . 53 Ill. Notes on the Manatee (Manatus americanus) recently living in the Society's Gar- dens. By A. H. Garron, U.A., F.Z.S., Prosector to the Society . . . . 137 IV. On Dinornis (Part XXI.): containing a Restoration of the Skeleton of Dinornis maximus, Owen. With an Appendix, on Additional Evidence of the Genus Dromornis in Australia. By Professor Owen, C.B, F.BS., P.ZS., Ge. . 147 V. On the Structure and Development of the Skull in Sharks and Skates. By W. \. Ioan, ee a ir aN ms Pye dole) VI. A Description of the Madreporaria dredged up during the Expedition of H.M.S. ‘ Porcupine’ in 1869 and 1870.—Part II. By Professor P. Martin Duncan, M.B. (Lond.), F.R.S., President of the Geological Society . . . . « . 285 VIL. On the Skull of the Afgithognathous Birds.—Part I. By W. K. Parker, F_R.S., IRL S HALOS © 0: PoE Aes. Ss A iw oo ew eet ay Pe at pe OL VIII. On the Aaial Skeleton of the Pelecanidw. By St. Grorce Mivarr, /.2.S., 2.Z.S., Professor of Biology at University College, Kensington . . . . . . . 315 IX. A Monograph of the Ostracoda of the Antwerp Crag. By GrEorGe STEWARDSON Brapy, ID., F_L.S., CMZS., Professor of Natural History in the University of Durham College of Physical Science, Newcastle-upon-Tyne . . . . . 379 X. On the Brain of the Sumatran Rhinoceros (Ceratorhinus sumatrensis). Ly A. H. Garson, McA ERS. Prosector to the Society. . . . . + . 5 . « aD XI. A further Contribution to the Knowledge of the existing Ziphioid Wales. Genus Mesoplodon. By WititaM Henry Fiower, F.RS., V.P.ZS.. . . . . 415 iv CONTENTS. XII. Notes on the Fins of Elasmobranchs, with Considerations on the Nature and Homo- logues of Vertebrate Limbs. By Sv. Groner Mivart, V.P.Z.S. . . page 439 XIII. On the Mechanism of the Odontophore in Certain Mollusca. By Patrick GEDDES.* 7 ch) sae) hae ee Se a Po ees XIV. On the Hearts of Ceratodus, Protopterus, and Chimera, with an Account of undescribed Pocket Valves in the Conus arteriosus of Ceratodus and of Protopterus. By ¥. Ray Lanxester, W.A., F.R.S., Professor oy a ry and a ative Anatomy in University College, London . . . . : ae 408} XV. Observations on the Uraniide, a Family of Lepidopterous Insects, with a Synopsis of the Family and a Monograph of Coronidia, one of the Genera of which it is composed. By J.O. Westwoon, IA, BES. @¢. 2. 5 . . 2 . =. OUT XVI. Supplementary Notes on the Curassows now or lately living in the Societys Gar- dens. By P. L. Scuater, W.A., Ph.D., F.R.S., Secretary to the Society. . 543 Liss OLE HOS letyoyerds) Commerc tn WOL 2G 5 6 56 9 6 o 0 o ate o 5 o BT Inidexiols Species (cGwk.) yeh! eee ere, Sete ee. Pee ee Oe TRANSACTIONS OF THE ZOOLOGICAL SOCIETY. I. On the Axial Skeleton of the Struthionide. By St. Groner Mivarr, F.R.S., Sec.L.S., Professor of Biology at University College, Kensington. Received August 28, 1874. Read November 17th, 1874. - Ina paper read before the Zoological Society’ in June 1872, the axial skeleton of the Ostrich was described in considerable detail, that it might serve as a type and standard for future comparisons. The present paper is offered as a first instalment of a series of such comparisons; and the genera selected are the allied ones, Rhea, Dromeus, Casua- rius, Apteryx, and Dinornis, so that a general conception of the axial skeleton as it exists in the Struthionide may be arrived at. It has not been thought desirable here to enter into the same amount of detail as in the description of the typical form, in order not to occupy an undue space in the Society’s ‘Transactions.’ The detailed description of the type, already given, may facilitate further comparisons should they be desired. thea, however, appears so peculiar a form as to merit exceptional notice. The specimens examined are all in the Museum of the Royal College of Surgeons ; and the illustrations are thence taken by kind permission of the authorities of that Institution. THE AXIAL SKELETON OF RHEA. In Rhea there are fourteen cervical, and three cervico-dorsal vertebre (fig. 1, c & cp). There are both three dorsal and three dorso-lumbar vertebre, the first two of the latter not being ankylosed to the sacrum. To these succeed about nine lumbar vertebre, all ankylosed together ; and these are followed by three sacral vertebra, the expanded rib-like 1 See Trans. Zool. Soc. vol. viii. p. 385. vou. X.—part I. No. 1.—WMarch, 1877. > B 2 MR. ST. GEORGE MIVART ON THE AXIAL SKELETON OF RHEA. RWSherwin Fig. 1. Prepelvic part. C, cervical yertebre ; CD, cervico-dorsal vertebre; D, dorsal and dorso-lumbar vertebree. Fig. 2. Pelyic and caudal parts. at, antitrochanteric process; 7, ischium; 7, ilium; Jp, ilio-pectineal process ; Pp, pubis; ps, interobturator process; s¢, supratrochanteric process. AXIAL SKELETON OF THE STRUTHIONID 2. 3 processes of which abut against the ilia just behind the acetabula (fig. 8, xxx11I, d & p, and fig. 9,p). After these vertebre there is a sort of gap, the nine anterior caudal vertebre either becoming entirely absorbed between the ilia or never being distinctly developed. Five imperfect, half-absorbed vertebree appear postaxiad of the ilia; and to them succeed six free and, in a sense, less incomplete vertebre, which terminate the series. Thus there are but twenty-three vertebre belonging to the first four categories, instead of twenty-seven as in Struthio. THE CERVICAL VERTEBRA. ATLAS AND AXIS OF RHEA (natural size). Fig. 3. Fig. 4. — = y Fig. 3. Preaxial view of atlas. Fig. 4. Lateral view of axis. ac, preaxial articular surface of centrum ; d, diapophysis; 4y, median hypapophysis ; /h, lateral hypapophysis ; az, prezygapo- physis; Ap, hyperapophysis ; zs, neural spine; 9, odontoid process; pe, postaxial articular surface of centrum; p/, pleural lamella; pz, postzygapophysis of axis; 2, postzygapophysis of atlas. The atlas, compared with that of Struthio, is relatively much more dorso-ventrally and less transversally extended. ‘The cup for the occipital condyle is nearly complete, having but a very small and narrow dorsal excavation. The lateral hypapophysial pro- cesses (fig. 3, //) are more marked, and the median hypapophysis relatively smaller. The diapophyses (fig. 3, d), though distinct, are very much, even proportionally, smaller. No hyperapophyses are distinguishable; nor is there any costal bony spiculum. Even in the very immature specimen in the Museum of the College of Surgeons (No. 1361 F), the neural lamine are entirely ankylosed to the quasi-body, and show but.a trace of their own dorsal union. The avis (fig. 4) has a greater vertical extent than that of Struthio, while its antero- posterior excess over the atlas is much less. There is a pit below the odontoid process. The postaxial central surface is much more prolonged ventrad by the relatively much larger hypapophysis (fig. 4, hy), which also extends further towards the preaxial margin of the centrum. ‘The neural spine (fig. 4, ns) is relatively shorter (pre- and postaxially) and higher. ‘The postzygapophyses are still more in excess of the prezygapophyses than in Struthio. A very large pneumatic foramen is placed above the parapophysial root BQ 4 MR. ST. GEORGE MIVART ON THE of the pleural lamella; and the interzygapophysial ridge may be very conspicuously perforated. The third vertebra is again relatively shorter and higher than in Struthio. Both articular central surfaces look more dorsad and ventrad respectively; and the lateral margins of the postaxial one are not, or but very slightly, concave. The hypapophysis extends more postaxiad; but the neural spine does not extend so much preaxiad as in Struthio. The postzygapophyses look entirely ventrad. The hyperapophyses are as large as in the axis and are grooved pre- and postaxially. The neural spine does not rise suddenly preaxially, and is hardly at all preaxially excavated. The interzygapo- physial ridge is perforated, but not so conspicuously as in the axis. FOURTH VERTEBRA OF RHEA (natural size). Dorsal view. Letters as before. The fourth vertebra is again much shorter relatively as well as absolutely than in Struthio, is more equal in length to the third, and more quadrate when viewed dorsally. its transverse diameters between the pree- and between the postzygapophyses being sub- equal. The neural spine is slightly excavated preaxially, but is deeply so postaxially. The fifth vertebra——The contrast as to length between this and the preceding ver- tebra in Rhea is something like that existing between the third and fourth vertebre in Struthio. The neural arch is not more cut away preaxially than in the fourth vertebra. There is no hypapophysis, and hardly a trace of hyperapophyses. The neural spine (like that of Struthio’s fourth vertebra) is reduced in size, and is slightly and subequally excavated at both ends. ‘The transverse diameter of the preaxial part of the vertebra (viewed dorsally) decidedly exceeds that of its postaxial end. The sivth vertebra is much longer and more slender, and differs from the fifth much more than does the fifth from the fourth, the increase in length thus taking place later in the vertebral series in Rhea than in Struthio. The styliform ribs are very much shorter; but, as in Struthio, the catapophyses are first marked in this vertebra. The AXIAL SKELETON OF THE STRUTHIONIDA. 5 neural spine is excavated, but not obliquely so, at each end. The ventral angles of the postaxial central articular surface are more drawn out than in the preceding vertebre, or than in the sixth vertebra of Struthio. The seventh vertebra.—All that has been said of the seventh vertebra of Struthio applies to that of Rhea, except that in the latter the rib is rudimentary, and the neural spine is much smaller. The eighth, ninth, and tenth vertebre also agree with those of Struthio, except that the neural spine is hardly excavated in the eighth, and in the others only behind, and that the ribs are rudimentary—although the pleurapophysial continuation of the para- pophysis projects much more preaxiad than does the preezygapophysis, which is not the case in Struthio. ELEVENTH VERTEBRA OF RHEA (natural size). Fig. 7. Fig. 6. Postaxial view. Fig. 7. Lateral view. Letters as before (see p. 3), except that c denotes the catapophysis, m the metapophysis, p the parapophysis, and ps the rudimentary rib. The eleventh vertebra—This appears more or less to replace the eleventh, twelfth, and thirteenth vertebre of Struthio. It is very like its predecessor, but is slightly more massive. Its neural spine is grooved above, the groove widening postaxially. The pre- zygapophysis projects but very slightly, if at all, preaxiad of the most preaxial part of the pleurapophysis. The ventral angles of the postaxial central surface are much drawn out (fig. 6); and thus (unlike Struthio) its ventral margin is greatly in excess of its dorsal margin, the ventral surface of the centrum being very concave transversely. The twelfth vertebra seems to answer to the thirteenth and fourteenth vertebra of Struthio, and differs from its predecessors as does the fourteenth vertebra of the Ostrich from its predecessors. Unlike the fourteenth vertebra of Struthio, however, it is still rather slender (pre- and postaxially lengthened), and the ventral angles of the postaxial surface of the centrum are greatly drawn out. 6 MR. ST. GEORGE MIVART ON THE The thirteenth vertebra may be compared with the fifteenth of Struthio. The most noticeable differences are the still rudimentary condition of the styloid rib, the forked central postaxial surface, and the decided extension of the pleurapophysis preaxiad of the prezygapophysis. This vertebra is hardly less extended pre- and postaxially than is the twelfth vertebra; and its proportions are more slender than in Struthio. The fourteenth vertebra, if compared with the sixteenth of Struthio, differs in that the ventral surface of the centrum is not of increased*width, and that the styliform rib (though still very small) is rather more instead of less developed than is its preaxial predecessor. There is no trace of a hypapophysial ridge. The pleurapophysis still extends preaxially beyond the prezygapophysis. THE CERVICO-DORSAL VERTEBRA. The fifteenth vertebra agrees with the eighteenth of Struthio, except that its preaxial central articular surface is almost divided into two lateral surfaces. The postaxial central surface widely diverges ventrally, and (as in the fourteenth vertebra) exhibits half its extent when the vertebra is laterally viewed. ‘There is a hypapophysis, which springs from a single root, but tends to bifurcate laterally. The ribs are more developed than in any of the cervical vertebre ; they still extend preaxiad of the praezygapophyses. There is no true neural spine, the neural lamine being medianly grooved pre- and post- axially. The siateenth vertebra, which may be taken as equivalent to the nineteenth of Struthio, is but very slightly larger than the fifteenth. The postaxial articular surface of the centrum differs greatly from that of its predecessor, there being no diverging ventral extensions, and its ventral margin being almost convex. The hypapophysis is relatively larger than in Struthio, extending the whole length of the ventral surface. It is much more inclined preaxiad, and both relatively and absolutely more extended yentrad. The neural spine may be said to bifurcate laterally, there being a deep median pre- and postaxial groove, thus differing greatly from the same part in Struthio. The seventeenth vertebra has a hypapophysis somewhat larger relatively than the hypapophysis of the twentieth vertebra of Struthio. There is a stumpy neural spine, the preaxial surface of which is nearly vertical. The postzygapophyses are not entirely postaxial to the centrum. The prezygapophyses are hardly less extended laterally than in the sixteenth vertebra. THE DORSAL VERTEBRZ. The eighteenth vertebra differs from its predecessor much as the twenty-first of Struthio differs from its two predecessors. The neural spine is at once higher and AXIAL SKELETON OF THE STRUTHIONIDA. 7 much longer pre- and postaxially, while its postaxial excavation is very much smaller. The central postaxial surface is medianly produced ventrad. The di- and parapophyses are as in the twenty-first vertebra of Struthio. The hypapophysis is smaller than in the seventeenth vertebra, but extends preaxiad much beyond the preaxial surface of the centrum. The nineteenth vertebra has the neural spine more pre- and postaxially, but not more dorsally, extended than in the eighteenth vertebra. There are no fosse beside it. The hypapophysis is very slightly produced. The diapophysial articular surface is placed medianly on the transverse process. The posterior zygapophyses do not extend post- axially beyond the centrum. The twentieth vertebra exhibits the trace of a hypapophysis at each end of the cen- trum. It is very like its predecessors, the zyg-, di-, and parapophyses being of about the same size and relative position. The neural spine is rather shorter pre- and post- axially and a trifle higher. Its postaxial excavation is smaller. THE DORSO-LUMBAR VERTEBRA. The twenty-first vertebra rather resembles the twenty-fourth of Struthio; but the neural spine is much less high, even relatively. There is no trace of a hypapophysis postaxially ; but there is a rudiment at the preaxial end of the centrum. The di- and parapophysial articular surfaces are about as remote from each other as in the twentieth vertebra. When the vertebra is viewed dorsally no radiating lamell are to be seen, thus differing from the twenty-fourth vertebra of Struthio. The twenty-second vertebra appears to answer to the twenty-fifth of Struthio, which it greatly resembles, except that its neural spine is less high, though it is much longer and more slender than in the preceding (twenty-first) vertebra. A vertical ridge distinctly divides, medianly, the postaxial neural excavation. The twenty-third vertebra differs from its predecessor in that it is shorter pre- and postaxially. Its neural spine is higher and more slender, the preaxial surface of its centrum is less concave, and the di- and parapophysial articular surfaces are less remote one from another. THE LUMBAR VERTEBRZ. The twenty-fourth vertebra is the first postdorsal vertebra which has no distinct rib. It appears to answer to the twenty-eighth of Struthio, but differs in that its neural spine is not higher than its predecessor’s, that its transverse processes incline preaxiad instead of postaxiad, and that the neural arch is so cancellous and imperfectly ossified. 8 MR. ST. GEORGE MIVART ON THE SACRUM OF A YOUNG RHEA. Fig. 8. Lateral view (natural size). d, diapophyses of 34th and 35th vertebra: ; i, ilium; pa, large parapophysis of 27th vertebra; P, its articular surface for ilium ; ~, parapophyses of 34th and 35th vertebre ; f, transverse process of 33rd vertebra. Ventral view (3 natural size). at, antitrochanteric process; il, ilium; pa, parapophysis of 27th vertebra; p, parapophyses of 33rd, 34th, and 35th vertebra ; 1, 2, 3, 4, and 5 indicate the 24th, 25th, 26th, 27th, and 28th vertebra respectively. The twenty-fifth vertebra (fig. 8, xxv, and fig. 9, 2) forms a part of the solid sacral mass. A single large foramen is placed between its neural arch and that of its serial successor. The twenty-sixth vertebra (fig. 8, xxvi, and fig. 9, 3) exhibits no sign (as in the imma- ture Struthio) of a union by suture of its neural arch with its centrum. Two super- imposed foramina open, from the neural canal, between the neural arch of this vertebra and that of the twenty-seventh. The ¢wenty-seventh vertebra appears to answer to the thirty-first of Struthio; but it differs from it in that it sends a larger truncated parapophysial surface (figs. 8 & 9, pa) ventrad and postaxiad to abut against the preacetabular process of the ilium. ‘This is possibly, but not probably, formed (as in Struthio) partly by an adjacent preaxial para- pophysial process of the twenty-eighth vertebra. Two small superimposed neural AXIAL SKELETON OF THE STRUTHIONID. 9 foramina are placed between this vertebra and its successor. The diapophysis has aborted or has*become fused with the neural spine. The twenty-eighth vertebra (fig. 9, 5) seems to answer to the thirty-second vertebra of Struthio, but is not concave either ventrally or laterally. Iam unable to say whether its neural arch is deplaced or not. There is no diapophysis; and the parapophysis may be absent, or it may (as in No. 1361 &) be present asa process abutting against the preacetabular process of the ilium, like the parapophysial process of the twenty-seventh vertebra, but considerably smaller than the latter in all its proportions. The vertebre from the twenty-ninth to the thirty-second inclusive’ are devoid of trans- verse processes, but have a pair of superimposed neural foramina at the postaxial ends of their neural arches. A median hypapophysial keel extends postaxially, beginning beneath the thirtieth vertebra. The last (thirty-second) vertebra may send a thin para- pophysial lamella to abut against the ilium, and so resemble the vertebre of the next category. These vertebrae without transverse processes might be distinguished as Lumspo-SacrAL VERTEBR&, a category not present in Struthio, in which genus these vertebre have parapophyses, as have also the more postaxial lumbar vertebre. THE SACRAL VERTEBRA. The thirty-third vertebra—This is the first vertebra which normally develops a transverse process (fig. 8, ¢), abutting against the postacetabular (or rather, here, swpra- acetabular) part of the ilium. This transverse process seems, in the young, rather diapophysial than parapophysial in its nature; but with age the plate descends to a lower level. There is a strong median subvertebral keel. The thirty-fourth vertebra. In this vertebra the parapophysis becomes more conspi- cuous (fig. 8, p). In the young it is seen to form (as in S¢ruthio) m conjunction with the diapophysis (d) a flattened surface for the ilium. The vertebra is generally smaller than is its serial predecessor ; but the median subvertebral keel is well developed. The thirty-fifth vertebra is like its predecessor, but smaller generally, while the united di- and parapophysial surface (on each side) is larger. With age, as the adult condition is gradually attained, the sacral vertebrae become drawn relatively preaxiad through the much less rapid rate of increase of the last five lumbar vertebre (twenty-eighth to thirty-second). Thus these vertebre become rather supraacetabular in position, as in Struthio, than postacetabular; and thus the vertebral column hardly appears, as it does in Struthio, at the bottom of the acetabulum when the pelvis is viewed laterally. THE SACRO-CAUDAL VERTEBR2. The thirty-siath vertebra.—In the young this vertebra has a tolerably developed centrum and a transverse process (formed of both di- and parapophysis) abutting 1 It may be the 28th to the 32nd, or only the 29th to the 31st, that are thus devoid of transverse processes, vou. x.—part I. No. 2.—WMarch, 1877. c 10 MR. ST. GEORGE MIVART ON THE against the ilium, though the neural arch and spine are quite rudimentary. At a more advanced stage (as seen in No. 1057") all that remains in an osseous condition is a long and narrow body with a pair of transverse processes of very cancellous texture. Post- axial to this vertebra the vertebrzee become atrophied in the adult to such a degree that merely a long narrow osseous band, of very cancellous structure, represents the bodies, (from that of the thirty-seventh to the forty-fourth inclusive) of the vertebre ; and these have diverging parapophysial processes of similar unsubstantial texture. The thirty-seventh, -eighth, and -ninth vertebre are elongated vertebra enclosed between the posterior parts of the ilia. Behind these and above the ischia are four (in No. 1057), or five (in Nos. 1361 & 1361 F) vertebra, which gradually become less imperfect as we proceed postaxially—the bodies broadening out and the neural spines getting shorter, thicker, and more stumpy,—the ultimate vertebree more or less ankylosing with the ischia. THE CAUDAL VERTEBRZ. The forty-fifth, -siath, -seventh, -eighth, -ninth, and fiftieth vertebre.—Postaxial to the ischia are six vertebre, which gradually diminish postaxially. None have transverse processes. ‘The first three, sometimes the first five, have more and more minute neural arches. The last vertebra is grooved dorsally; and sometimes the last three are so grooved. The last apparent vertebra, the pygostyle, is not dorso-ventrally expanded into an osseous plate, as it is in Struthio, but is cylindrical and short, not being twice the length of the vertebra preceding it (fig. 2). It looks as if made up of only two vertebre ankylosed together. THE PELVIS. In the adult the pelvis consists of twenty vertebra and two ossa innominata. When viewed preazially, its aspect differs greatly from that presented by the pelvis of Struthio, on account of the absence of the descending pubes and pubic symphysis, in Rhea, as also because the ischia curve inwards, converge, and unite together just post- axiad to (and, of course, on the ventral side of) the acetabula, and thence contiuue onwards, so united, postaxiad. The iliac roof of the first pelvic vertebra is much more concave on each side than in Struthio. The ilium also sends out a sharper process (the supratrochanteric process) above each trochanteric process (fig. 2, st). When viewed postaxially, the same absence of a pubic symphysis and the presence of an ischiatic one produces a very great difference of aspect from this point of view also. Again, the summit of the pentagonal mass is horizontal, owing to the crest of the ilium not rising dorsally as much as in Séruthio. Viewed laterally, the part which was, in the description of the pelvis of Struthio, ? In the Museum of the College of Surgeons. AXIAL SKELETON OF THE STRUTHIONIDZ. 11 compared with a bird’s skull, has the beak smaller and the cranium less dorsally convex, while the upper of the two anterior bars (7. ¢. the ischium) joins the beak by ossification of the ligament uniting it with the ilium. Moreover the pelvis differs from that of Struthio in that the two bars (ischium and pubis) extend postaxiad beyond the ilium to a greater degree, and do not curve ventrad distally. The supratrochanteric process is also decidedly preaxiad to the antitrochanteric process, instead of slightly postaxiad as in Struthio. ‘The ilio-pectineal process is much shorter; and the vertebral column is scarcely to be seen through the acetabulum; while the latter cavity is placed nearer the middle (pre- and postaxially) of the total length of the ilium. The os innominatum is much less deep dorso-ventrally in proportion to its pre- and postaxial extent. The pre- acetabular part of the ilium has its external surface more concave than in S¢ruthio, as is also the preaxial margin of the ium. Viewed dorsally, the pelvic ribs are relatively larger than in Struthio. The antitro- chanteric processes project at about the middle of the ilium’s length pre- and post- axially and behind the anterior third of the total length of the pelvis. Just in front of the processes the supraacetabular processes project strongly outwards, relatively as well as absolutely much more so than in Struthio. In Rhea only three processes are seen to extend postaxiad (instead of five as in Struthio), all joining together distally. The median process of these three (ilia and ischia in one) expands distally so that its lateral margins are concave. No vertebral spines are visible in the adult (as they are in Struthio), the two ilia ankylosing together dorsally in the adult, and even in the im- mature (e. g. in No. 1361 ©) meeting together, though in the very young (e.g. in No. 1361 F) the lumbo-sacral and sacral spine-tips come to the surface between the ilia from a little preaxiad to a little postaxiad of the preacetabular process. The two lateral processes above spoken of are, of course, the two pubes, each pubis forming a gentle curve (not so bowed outwardly as in Struthio) with a convex external margin and a concave internal one. The transverse diameter of the pelvis is pretty uniform. It is greatest between the ends of the pelvic ribs, and next greatest across the pubes. The posterior ends of the ilia do not diverge as in Struthio. When viewed ventrally, the most striking differences between the pelvis of Rhea and that of Struthio are, the greater size of the pelvic ribs in the former, the fact that but three bony processes extend postaxiad, and the binding of the sacro-caudal vertebrae and the postaxial parts of the ilia by the medianly united ischia. Also the iliopectineal eminences are smaller, and the centra postaxiad to the sacral vertebre are aborted. Although there are from three to five /wmbo-sacral vertebre (7. e. without parapo- physes and between the lumbar and sacral vertebra), yet they are so closely approxi- mated antero-posteriorly as to leave but a very small fossa between them on each each side and the adjacent acetabulum. 12 MR. ST. GEORGE MIVART ON THE THE Inium (figs. 2 & 8, 77). This bone extends itself over twenty-two vertebree, namely from the twenty-third to the forty-fifth inclusive. Compared with the same bone in Séruthio, its dorsal margin is more convex, its anterior margin more concave (the ventral preaxial angle being more prominent), and the postacetabular part is not so much in excess of the preacetabular portion. It would taper preaxiad but for the ossification (before mentioned) of the ilio-ischiatic ligament which causes it to expand vertically towards its distal end. The gluteal lines do not descend (ventrad) so much as in Struthio; and the stronger supra- acetabular process comes to jut out more horizontally as well as more strongly, making, with its fellow of the opposite side, a flat rhomboidal surface on the dorsum of the ilium. The ilia are flattened against the included postacetabular vertebre to a remarkable degree. Tue Pusis (fig. 2, p). This bone is like what that of Struthio would be if the latter were sharply cut off at the postaxial end of the ischium ; only it is not quite so much bowed outwards. There is, of course, no pubic symphysis. In the young it does not join the ischium distally, but quite resembles the osseous part of the pubis of the young Ostrich when the sym- physial part is all cartilaginous, TueE Iscuium (fig. 2, 7). This bone is very slightly, if at all, shorter than the pubis, and ankyloses postaxially both with that bone and with the ilium. It seems to form about the ventral third of the antitrochanteric process. It also slightly ankyloses with the pubis more proximad, so as to cut off the anterior part of the obturator foramen as a separate and much smaller foramen (fig. 2, between Jp & ps). The two ischia unite together postaxially a little behind the acetabulum, and thence expand transversely as they proceed postaxiad, forming an elongated sheet of bone (concave in both directions on its ventral surface) beneath the sacro-caudal vertebre. At its distal end it sends down a process, curving at first ventrad and then preaxiad, which ankyloses with the extreme distal end of the pubis. Thus, as it were, the outer ridge of the ischium of Struthio is drawn out, while the surface between the (here relatively approximated) dorsal and ventral ridges coalesces with the corresponding surface of its fellow of the opposite side. THE VERTEBRAL RIBS (fig. 1). There are nine vertebral ribs, the first and last becoming in the adult (as in Struthio) ankylosed transverse processes. The fourth, fifth, and sixth of these bones unite with sternal ribs. (See fig. 1.) The jirst rib is attached to the fifteenth vertebra, and ankyloses with it in the adult. It is very small, triangular, and very little longer than broad. AXIAL SKELETON OF THE STRUTHIONIDA. 13 The second rib is much.as in Struthio, but somewhat shorter relatively. There is no noticeable convexity on its postaxial margin. The third rib has its capitulum rather larger, compared with that of the second rib, than is the case in Struthio. Its shaft is more vertical. ‘The tubercular surface is but little larger than is that of the second rib. The postaxial margin is proximally concave, then slightly convex. The fourth rib is remarkably different in its curvature from that of Struthio; and its excess in length over the third rib is less. Owing to this curvature, it seems to be rather the tuberculum than the capitulum which carries on proximad the general curva- ture of the rib. This curvature is slightly sigmoid, and there is no “angle,” the pro- minence being preaxiad and below the capitulum instead of, as in Struthio, postaxiad and below the tuberculum. ‘The vertebral margin is as sharply concave as in the third rib. The articular surface of the tuberculum is not much larger than in the third rib. The preaxial margin of the shaft is proximally convex, then more or less concave. The postaxial margin is at first markedly concave, then (distad) slightly convex. Both mar- gins are sharp, the lamella of bone (described in Struthio as being on the ventral pre- axial side of a ridge running down from the capitulum) of the proximal part of the shaft projecting strongly preaxiad and scarcely at all postaxiad, thus doing away with an “angle” (such as exists in Struthio), and producing a curious ‘preaxial convexity, whence a ridge is continued on proximad, over the divergence of the tuberculum and capitulum, right up to the vertebral margin and preaxial edge of the tuberculum, as is the fifth rib of Struthio. There is an articulated uncinate process; but its distal por- tion is not bent dorsad, but continues on obliquely in one line with the proximal part, the direction being mainly postaxiad, but slightly dorsad. The jifth rib has, again, a singularly different curvature from that of Struthio. The capitulum and tuberculum diverge at a rather more acute angle than in the fourth rib ; otherwise it generally agrees with the fifth rib of Struthio. The stath rib has also a reversed curvature compared with that of Struthio. The capitulum and tuberculum diverge much as in the fifth rib; and the uncinate process is as long as that of the rib last named. ‘The proximal prominence is more marked, causing the preaxial margin distad to it to be more concave than in the fifth rib. The seventh rib is free distally, and is much shorter than the sixth, and rather so than the fourth. Its shaft also is more curved. ‘The capitulum and tuberculum also form a wider angle than in the sixth rib, namely about 75°. The pneumatic foramen is of about the same size as in the sixth rib; but the articular surface for the diapophysis is not so large. The preaxial prominence of the preaxial margin is slightly more marked still ; and the concavity of the preaxial margin is more marked also. There is no uncinate process. The eighth rib is again more curved than in Struthio. In length it is intermediate between the second and third ribs. The capitulum is about as long compared with the tuberculum as in the seventh rib. The preaxial frontal convexity is still very marked. There is no trace of an uncinate process. 14 MR. ST. GEORGE MIVART ON THE The ninth rib articulates with the twenty-third vertebra, and seems to correspond with the tenth rib of Struthio. Being more preaxially directed, it diverges more from the eighth rib than does the tenth from the ninth in Struthio. It is also more equal to its predecessors in length than is either the ninth or the tenth of the last-named genus. The preaxial prominence is again very marked, but there is no preaxial concavity distal to it. The rib is more flattened than is either the ninth or the tenth of Struthio, and, once more, it is more curved. THE STERNAL RIBS (fig. 1). Of these there are apparently but three. It may be that a minute one exists pre- axiad of these three ; but, if so, it is not to be found in the specimens examined. The frst sternal rib seems to answer to that described as the second in Struthio. It is rather longer than the second vertebral rib, but is not so slender as the corresponding one in Struthio. Its outer surface is convex ; its inner surface is flattened. It joins the fourth vertebral rib, and belongs to the eighteenth vertebra (fig. 1, 1). The second rib has its proximal end quite rounded; otherwise it resembles the third sternal rib of Struthio (fig. 1, 11). The third rib does not exceed the first in length by half the length of the latter. Its distal end is absolutely less expanded dorso-ventrally than is that of the second rib. Its postaxial surface is hardly more excavated than is that of the second rib. In other respects this bone resembles the fourth sternal rib of Struthio (fig. 1, m1). THE STERNUM. STERNUM OF RHEA (3 natural size). Fig. 10. Fig. 10. Ventral yiew. Fig. 11. Lateral view. ce, eoracoid grooves ; ca, costal angles; f, median yentral prominence; 7, a ventral articular surface; s, a dorsal articular surface ; 7x, lateral xiphoid processes, AXIAL SKELETON OF THE STRUTHIONIDA. 15 This bone is narrower in proportion to its pre- and postaxial length than in Struthio. The coracoid grooves are not nearly so approximated; and their dorsal margins project but little, if at all, preaxiad beyond their ventral margins. The costal angles are narrower in proportion to their length. There is no median preaxial projection. The lateral xiphoid processes are small, and project but little postaxiad, leaving but a moderate notch (and no median xiphoid process) between them. The sides of the sternum are less concave. The median ventral prominence (fig. 10, f) is not flattened. Each pleu- rosteon has but three articular surfaces; and each of these has a deep excavation postaxial to it. The ventral facet of each articular surface exceeds the dorsal one in size; but they are most nearly equal in the third articular surface. VERTEBRAL PARTS AND PROCESSES. The centra resemble generally those of Struthio, except as to which are the ones which became ankylosed. ‘These parts may all but abort altogether, as in the anterior caudal vertebre. The neural lamine do not show such signs of shifting as in the lumbar vertebre of Struthio. The neural arches may entirely abort, while the centyra still exist in an im- perfect manner, as in the anterior caudal vertebre. The substance of the lamine may be so cellular as to be most extremely imperfect in development, as in the lumbar and sacral vertebre. Diapophyses may be quite absent, if not blended with the spinous processes, as in the last six presacral vertebre. The neural spines are most expanded, pre- and postaxially, in the dorsal vertebre. They may be laterally bifid, as in the third cervico-dorsal vertebra. The zygapophyses are developed as in Struthio; and the metapophyses offer no great differences. Hyperapophyses are only conspicuous in the second, third, fourth, and fifth vertebra. Diapophyses and parapophyses are occasionally absent where they are present in Struthio, notably in the last five lumbar vertebre. No such processes abut against the ischium. The plewrapophyses of the cervical region differ greatly from those of Struthio by their shortness and non-styloid form. Hypapophyses are nowhere distinctly paired. Catapophyses are found preaxiad to the fifteenth vertebra. The vertebre are raised above (7. e. dorsally to) the acetabula. The vertebral ribs have a different twist from that existing in Struthio. The wncinate processes do not seem to be more than three in number. 16 MR. ST. GEORGE MIVART ON THE AXIAL SKELETON OF EMU. Fig. 13. RN Sherwin Fig. 12. Prepelvic part. Fig. 13. Pelvic and caudal parts. C, cervical verterbee ; CD, cervico-dorsal yertebree ; D, dorsal and dorso-lumbar yertebrie; ca, costal angle. Letters of pelvic part as in fig. 2. AXIAL SKELETON OF THE STRUTHIONIDA. 17 THE AXIAL SKELETON OF DROMAUS (figs. 12 & 13). In Dromeus there are seventeen or eighteen cervical and three or four cervico-dorsal vertebre; there are three or four dorsal and but two dorso-lumbar vertebre; to these succeed eight lumbar and three sacral vertebree, followed by ten or eleven sacro-caudal and eight or nine caudal vertebre. Thus there are about fifty-four vertebre in all; and of these as many as twenty-six or twenty-seven belong to the first four categories, thus differing from Rhea and agreeing with Struthio, as is also the case with the last category (the caudal vertebra), which are, within one or two, as numerous as in’ the African genus, THE CERVICAL VERTEBRA. ATLAS AND AXIS OF EMU (natural size). Fig. 14. Fig. 15. Fig. 14. Preaxial view of atlas; v, foramen enclosed by costal spiculum. Fig. 15. Lateral view of axis. Here and subsequently the letters refer to the same parts as in the corresponding illustrations of Rhea, except where differences are specified. The atlas differs from that of Struthio as does that of Rhea, except that the median dorsal notch is larger, that there is no hypapophysis, and that there is a costal spiculum on either side. The neural arch is somewhat more axially extended than in either of the other genera; and the diapophyses are even smaller than in Rhea. The dorso-ventral height of the whole bone, compared with its breadth, is intermediate between that existing in the other genera, though more like that of Rhea. The aais exceeds the atlas in axial length still less than it does in Rhea; and its relative dorso-ventral extent is yet greater. There is also no hypodontoid pit for a liga- ment. The postaxial central surface is still more prolonged ventrad by the relatively still longer hypapophysis, which is vertically grooved at its root postaxially, and extends quite to the preaxial margin of the centrum, where it extends still further ventrad and quite as suddenly. The neural spine is relatively as well as absolutely shorter (axially) and may be higher than in Rhea; but the pneumatic foramen is smaller, and vou. X.—PaRT I. No. 3.—March, 1877. x hs D 18 MR. ST. GEORGE MIVART ON THE the interzygapophysial perforation minute. The postzygapophyses with their hyperapo- physes extend more postaxiad than in either of the other genera. The third vertebra agrees in its proportions with that of Rhea, and differs from that of Struthio, except that the styloid rib is relatively longer than even in Struthio. The neural spine is excavated in front (slightly) as well as postaxially. The hyperapophyses are more grooved dorsally than in Rhea, while the interzygapophysial foramen is larger than that of the axis instead of smaller. The metapophyses are already marked. The fourth vertebra quite agrees with that of Rhea, except that the rib is longer, and that the neural spine (as in Struthio) is so grooved both pre- and postaxially that the median part is much reduced. The metapophyses are larger than in either of the other genera. The fifth vertebra contrasts, as to length, less with the fourth than in Rhea, the cervical vertebre increasing more gradually in length, and never attaining such a rela- tive development axially as in either of the other two genera. The transverse diameter of the preaxial part still more exceeds that of the postaxial part than in Rhea; the rib is relatively much longer. The interzygapophysial ridge may or may not be perforated ; but the metapophyses are more strongly marked than in the fourth vertebra. The sxth vertebra differs only slightly from the fifth in length. The neural arch and spine assume the characters of those of the thirteenth cervical vertebra of Rhea. The interzygapophysial ridge is only minutely perforated. The rib here, as throughout, is longer than even in Struthio relatively to the axial extent of its supporting vertebra, extending to, or beyond, the most preaxial part of the postaxial central articular sur- face in all the cervical vertebrae. The metapophyses are still stronger; and catapophyses begin to appear; and they diverge ventrad much less than in Rhea. The seventh vertebra presents similar characters to those of the sixth; but the meta- pophyses are again stronger, and the interzygapophysial ridge is large and perforated, the lamella forming a canal which passes ventrad and postaxiad on each side of the neural lamina. The following vertebra, from the eighth to the eleventh, present characters similar to those of the seventh ; only the neural. spine grows somewhat longer, though being still short and stumpy, even in the eleventh vertebra. In the twelfth vertebra the development of the neural spine is much like that of the sixteenth vertebra of Struthio. From this vertebra postaxiad, the interzygapophysial ridge may or may not be perforated. The vertebre from the thirteenth to the seventeenth inclusive are all much more alike than is the case in Rhea; they gradually, however, become larger and more massive, the neural spines rising very slightly. The catapophyses approximate together in the seventeenth vertebra, still, however, remaining distinct, as do those of the last cervical (fourteenth) vertebra of Rhea, and not uniting as in the last cervical (seventeenth) ver- tebra of Struthio. AXIAL SKELETON OF THE STRUTHIONID#. 19 THE CERVICO-DORSAL VERTEBRA. The eighteenth vertebra is like the first cervico-dorsal (fifteenth) of Rhea, except that the preaxial central surface is not nearly divided medianly, and that the postaxial central surface does not diverge ventrad, the ventral margin being scarcely at all concave, though more in excess of the dorsal margin than in Struthio, while the degree to which it is visible in profile is also intermediate. The hypapophysis bifurcates distally, though having a single root. It does not extend so much ventrad, relatively, as in Rhea. The rib may be rather more or rather less developed than in the preceding vertebra. It may, in the adult condition (as ¢. g. in No. 1358), be unankylosed. The prezygapo- physes exceed the preaxial extension of the parapophyses more than in Struthio, and much more than in Rhea. The neural spine is not dorsally grooved, but is excavated both pre- and postaxially. The nineteenth vertebra, compared with the sixteenth of Rhea, has the postaxial surface of its centrum less different from its predecessor ; yet its ventral does not so much exceed its dorsal margin. The hypapophysis is much more like that of Struthio than that of Rhea. The neural spine is not grooved dorsally. The diapophysis is more axially extended. The parapophysial articular surface is deeper and more dorso-ventrally extended; and that of the diapophysis is also rather more concave. The twentieth vertebra has its hypapophysis more like that of Struthio than that of Rhea, although, as in the last-mentioned genus, it extends much postaxiad. ‘The trans- verse processes are more axially extended, and the neural spine is smaller, than in Rhea’s seventeenth vertebra, which this one otherwise much resembles. The twenty-first vertebra has its hypapophysis as large as that of the twentieth ver- tebra, except when (as in No. 1358) an extra intercalated vertebra alters the relations, The transverse process is rather more, and the neural spine rather less extended axially. The postaxial excavation of the neural spine is less in defect of that of the preceding vertebra than in Rhea; otherwise the characters of this vertebra are much like those of the first dorsal vertebra in that genus. As in Struthio, there is a pit on each side of the preaxial surface of the neural spine. ‘This pit is serially homologous with the interzygapophysial canal. THE DORSAL VERTEBR#. The twenty-second vertebra has a very small hypapophysis, though a larger one than in the other two genera. The diapophysial articular surface is preaxiad to the middle of the transverse process. The neural spine is not so axially extended as in the pre- ceding vertebra, but more dorsally. There is a deep fossa on each side of the preaxial surface of the neural spine. The postzygapophyses do not extend postaxiad of the centrum. The twenty-third vertebra has no hypapophysis at either end of its centrum. The D2 20 MR. ST. GEORGE MIVART ON THE par- and diapophysial processes are more distant from one another than in the pre- ceding vertebra. The neural spine is higher, but not shorter axially. The postaxial excavations of the neural spine are much as in the twenty-second vertebra. The pre- axial fossee of that spine are deep. The twenty-fourth vertebra has no trace of a hypapophysis, even in front; otherwise it differs from the twenty-fourth vertebra of Struthio as does the twenty-first vertebra (first lumbar) of hea, except that the neural spine is more slender, the transverse process more extended axially, and the parapophysial surface larger. The preaxial fossee of the neural spine are deep; and it is more excavated postaxially than in Struthio. THE DORSO-LUMBAR VERTEBR2. The twenty-fifth vertebra is sometimes (as shown in fig. 12) a dorsal vertebra. It may have its neural spine higher than the corresponding vertebra of Rhea; but it is never so high as is that of Struthio. The neural spine’s postaxial excavation is not divided by a vertical ridge. ‘There is no sharply defined concavity beneath the pre- zygapophyses, and no excavation beneath the transverse process; otherwise this vertebra is like the twenty-second vertebra of Fhea. The twenty-sixth vertebra resembles the twenty-third of Rhea, except that the pre- axial articular surface is higher dorso-ventrally and more concave. There is no deep fossa ventrad and external to the prezygapophysis. THE LUMBAR VERTEBR. The twenty-seventh vertebra apparently answers to the twenty-fourth of Rhea and to the twenty-eighth of Struthio. It may be a dorso-lumbar yertebra. It has no con- spicuous rib, and is the first trunk-vertebra so distinguished. It becomes ankylosed to the sacral mass, as does its predecessor also. Ji differs from the twenty-fourth vertebra of Rhea in that its ventral surface is less concave. In the young the transverse pro- cess is directed less preaxiad than in the adult; and there is a rudimentary rib (as in No. 1358 a) beneath it, directed outwards and postaxiad and greatly expanded at its distal end, where it unites with the inside of the ilium (fig. 16, xxvut, p/). The twenty-eighth vertebra has its neural spine fused with that of the two preceding and the four succeeding vertebre into one great mass of cellular diploé. Its trans- verse. process is smaller than is that of the twenty-seventh vertebra; and there is no rudimentary rib even in the young. The twenty-ninth and thirtieth vertebre can only be distinguished as differing from the twenty-sixth and twenty-seventh of Rhea in that they have smaller transverse pro- cesses, and that the thirtieth sends no process to the preacetabulum. The thirty-first vertebra is quite like its preaxial predecessor, thus differing from both the twenty-eighth of Rhea and the thirty-second of Struthio (fig. 16). AXIAL SKELETON OF THE STRUTHIONID. 21 SACRUM OF EMU (natural size). Fig. 16. Ventral view. (Specimen No. 1358 4.) The thirty-second and thirty-third vertebre are about equal in size. ‘They closely blend together, and together send down a thick process, which ankyloses with the preacetabular part of the ilium, ischium, and pubis. The thirty-fourth (or eighth lumbar) vertebra has no transverse process visible on its ventral aspect. It may therefore be reckoned as a LUMBO-SACRAL vertebra. Its ventral surface is medianly somewhat keeled axially. An extra, or ninth, lumbar vertebra is intercalated in some skeletons. SACRUM AND PELVIS OF EMU ( natural size). xxvil XXVIII Vertical antero-posterior section. (Specimen No. 1387.) THE SACRAL VERTEBR’. The thirty-fifth, thirty-sivth, and thirty-seventh vertebre all send out smull transverse processes to abut against the postacetabular part of the ilium. ‘These pro- 1 The sacral vertebree are reckoned as three in number, only from analogy with Struthio. 22 MR. ST. GEORGE MIVART ON THE cesses are almost subequal in size; but the most postaxial is axially broadest. There is hardly a trace of a median keel. These vertebra are not raised up above the aceta- bulum as they are in Rhea. THE SACRO-CAUDAL VERTEBRZ. These vertebree, which are those from the thirty-eighth to the forty-seventh (or forty-eighth) inclusive, differ much from those of Rhea by their more substantial ossification. From Struthio’s vertebre (thirty-ninth to forty-sixth) they differ in that their transverse processes are more like the transverse processes of the sacral vertebre. Thus, when the pelvis is viewed ventrally, there is no break behind the sacral transverse processes; but the openings between the transverse processes become smaller and more distant postaxiad, owing to the gradually increasing axial breadth of the successive trans- verse processes. ‘The bodies of these vertebra are well ankylosed; and their spines be- come rapidly shorter as we proceed postaxially. They never ankylose with the ischia. THE CAUDAL VERTEBRZ. The caudal vertebrae (that is to say, those from the forty-eighth to the jifty-fifth inclusive) differ greatly from those of Rhea, and even exceed those of Struthio, in the development of their transverse processes, which project outwards and postaxiad. Their development, however, is irregular both as regards size and direction. Each neural spine is flattened dorsally, and vertically grooved postaxially. The first and, sometimes, also the next two caudals have strongly projecting postzygapophyses, which, however, do not attain the vertebra towards which they tend. The pygostyle is irregular and subcylindrical, and very unlike that of Struthio. It is about twice the length of the vertebra preceding it, and looks like three small and diminishing vertebre ankylosed together. THE PELVIS. In the adult the pelvis consists of twenty-one or twenty-two vertebra, as well as of the ossa innominata, Viewed preaxially and postazially it differs from the pelvis of Rhea in the non-union medianly of the ischia, and the non-descent of the postaxial part of the ilium, as also in the less concavity between the supra- and antitrochanteric processes. When viewed Jaterally the whole ilium is more convex dorsally than in Rhea, and the postacetabular part of the ilium more so than even in Struthio. The ischium and ilium almost join at their distal ends postaxially. The supratrochanteric process is very slightly postaxiad to the antitrochanteric one. The acetabular opening is much smaller AXIAL SKELETON OF THE STRUTHIONID. 23 than in the preceding genera; but the vertebral column is seen behind it as in Struthio. The preaxial margin of the ilium is still more concave than in Rhea. When the pelvis is viewed dorsally the supraacetabular processes nearly hide the antitrochanteric processes. Hight serial openings are visible in the middle of the post- acetabular part of the ilium. The surfaces between these apertures are composed of the flattened summits of the sacro-caudal neural spines interposed between the more postaxial parts of the two ilia. The pubis appears, thus viewed, very slightly, if at all, concave inwards; but the ischium is slightly concave outwards. ‘The posterior ends of the ilia diverge as in Struthio, thus differing from Rhea. Viewed ventrally, this complex bone presents five pieces proceeding postaxially as in PELVIS OF EMU (? natural size). Fig. 18, Ventral view. az, antitrochanteric process; 7, ischium; 2/, ilium; /p, ilio-pectineal process; p, pubis; sp, interobturator process. Struthio; but, as in Rhea, the ilio-pectineal eminences are smaller than in the African genus; but the ischia are more visible medianly, thus differing from Rhea. The pre- acetabular part of the pelvis is relatively greater than in either of the other genera; and the number of transverse processes there is greater. Also the transverse processes are broader generally, and fill up the middle of the more postaxial half of the ventral surface of the pelvis with almost continuous ossifications. The sacral transverse pro- cesses are much smaller than in Struthio, but are much better ossified than those of Rhea, as are also the more postaxial transverse processes. The fossa left between the one (34th) or two lumbo-sacral vertebrae and the adjacent acetabulum on each side is very small and inconspicuous. Tue Inivm (figs. 13 & 18, i/). The ilium extends over the vertebre from the twenty-sixth to the forty-fifth inclusive. It is like that of Rhea, except in the points already noticed, and that it is much less flattened against the postacetabular vertebre than in hea and less so than in Struthio. 24 MR. ST. GEORGE MIVART ON THE Tue Pusis (figs. 13 & 18, p). This bone does not join the ischium, but ends freely distad, as it does in the young Rhea. It projects least postaxiad of the three pelvic bones. Tue Iscuium (figs. 15 & 18, 7). This bone also ends freely distad, as it does in the young Rhea. Its ridges are con- ditioned as in Struthio. It expands slightly distad in both directions, as a hammer. It is the pelvic bone which extends furthest postaxially, yet very little beyond the ilium. THE VERTEBRAL RIBS (fig. 12). There are nine vertebral ribs, the last of which becomes ankylosed with the pelvis in the adult. The fourth, fifth, sixth, and seventh vertebral ribs join sternal ones. The jirst rib is much longer than broad. It remains unankylosed in the adult, and is attached to the nineteenth vertebra. The second rib is broader than in Rhea or Struthio. It has a convexity on its post- axial margin, a little below the tuberculum. The third rib is broader and more convex than in Rhea or Struthio. The tubercular articular surface is a good deal larger than in the second rib. The postaxial margin is convex dorsally, and ventrally concave. The fourth rib is like that of Rhea; but the uncinate process may be absent or pre- sent, when it is short and broad. The fifth rib is very like its predecessor, but is slightly shorter and more curved, and the angle formed by the capitulum and tuberculum is rather more obtuse. There may or may not be a short, thick uncinate process. The siath is like that of Rhea, and is shorter and more curved than its predecessor. There is no uncinate process. The seventh rib differs from that of Rhea in that it is not free distally. It is rather slighter, shorter, and more curved than its serial predecessor, while its diapophysial surface is quite as large. Its preaxial prominence is not more marked. The eighth rib is but very slightly, if at all, longer than the second rib. The pre- axial convexity does not project much; but the ridge which crosses that part where the head and tubercle diverge is very strong. The ninth rib, which may be articulated with or ankylosed to its vertebra, is shorter, more slender, and less curved than its predecessor. It is more curved than in Struthio, and less so than in hea. THE STERNAL RIBS (fig. 12). Of these there are four; and they increase in length postaxiad with very much greater rapidity than in Rhea, the third being twice the length of the first. AXIAL SKELETON OF THE STRUTHIONIDA. 25 The first sternal rib is rather shorter than the second vertebral rib, and is rather more slender and curved than in Séruthio or Rhea. The two distal articular surfaces are blended together to form one long articular groove; and the same is the case in all the other sternal ribs. It joins the fourth vertebral rib, and belongs to the twenty- second vertebra. The second sternal rib is also longer, more slender, and more curved than in either Struthio or Rhea. The same greater elongation and slenderness may be also predicated of the third and fourth sternal ribs. The fourth rib does not always attain the sternum, but may (as in No. 1358) be applied to the postaxial surface of the third rib, ending distally in a point. THE STERNUM. STERNUM OF EMU (¢ natural size). Fig. 19. Fig. 20. Fig. 19. Ventral view. Fig. 20. Lateral view. The sternum much resembles that of Rhea. The coracoid grooves, however, are approximated and very differently conditioned. Their dorsal margins extend so far preaxiad of their ventral margins that their dorsal boundary thus forms a large part of the sternum on each side. The two ventral margins together produce a prominence on the preaxial part of the sternum; and at the preaxial end of this prominence is a very slight and irregularly shaped notch. ‘There are no lateral xiphoid processes or median postaxial notch. There is an elevated, but not flattened, tract of bone at about the middle of the ventral surface. There are on each side four continuous articular surfaces, with a deep excavation behind and in front of each. The costal angles (fig. 12, ca) are VOL. X.—PART I. No. 4.—Warch, 1877. E 26 MR. ST. GEORGE MIVART ON THE extremely long, and project inwards and postaxially in a way not to be found in Struthio or Rhea. ‘The inner surface of the sternum is not so concave in either direction as in the last-mentioned genus. VERTEBRAL PARTS AND PROCESSES. These portions of the skeleton generally resemble those of Rhea, except in the fol- lowing points :— The centra never abort; nor do the newral lamina, as far as can be determined; but it is possible that they may do so in some part of the pelvis. The neural spines of the cervical vertebre are better-developed, but are not laterally bifid, though they are nearly so in some caudal vertebre. It is doubtful whether diapophyses are ever absent, save in the last presacral vertebree. Hyperapophyses. The interzygapophysial canals form conspicuous structures. The cervical styloid rids are well developed, and resemble those of Struthio and not those of Rhea. Catapophyses are developed from the sixth to the seventeenth vertebra inclusive. The hypapophysis is paired in the first cervico-dorsal vertebra. The vertebre are not raised above the acetabula. The vertebral ribs are stouter than in Rhea, but with the same twist. The wneinate processes do not seem to be ever more than two in number. THE AXIAL SKELETON OF CASUARIUS. The axial skeleton of Caswarius so much resembles that of Dromeus, that a much less detailed notice of the former than of the latter is alone needed. In Casuarius there are fourteen, fifteen, or sixteen cervical and four or three cervico- dorsal vertebre; there are five dorsal and two or three dorso-lumbar vertebree; to these succeed nine or ten lumbar and three sacral vertebrx, followed by nine, ten, or eleven sacro-caudal and eight or nine caudal vertebre. Thus there may be from fifty-five to fifty-nine vertebra in all; and of these as many as from twenty-five to twenty-seven belong to the first four categories, thus differing from Rhea and agreeing with Struthio and Dromeus. It also agrees with the last-named genus in the number of its caudal vertebre., Fig. 2 vit vin f, p/ Y Jp mn 2) AXIAL SKELETON OF THE STRUTHIONIDA. AXIAL SKELETON OF CASSOWARY. 1. Fig. 22. XX XVI ox XX xxd Fig. 21. Prepelvic part. Fig. 22. Pelvic and caudal parts. Description as in figs. 1 & 2, p. 2. D 4 Ail i pl, pelvic rib. cD 09 em Z a NN 4 fess E2 27 28 MR. ST. GEORGE MIVART ON THE The number of vertebre in the different species appears to be as follows :— galeatus. australis. bennettii. Obras guclye) Goeoucoouauccdgno0050enoUD 15 16 14 (@ervico=dorsalls conten ccc eel 4 3 4 | Cervical and cervico-dorsal together .......... 19 19 18 I DYosct-) Ul tentne Aes Cea cra romero OO <. cro oicaaedio Hain. c 5 5 5 Dorsoslumbar> Satis ase ce toto eenecantatreane 2 3 2 Vertebre of first four categories together ...... 26 27 25 Gum barsiecs cmc cteiee ots ie entree eaten 10 9 9 Sore sodn dup doo ouHdesoU Und ooosobagenen 3 3 3 Sacro-caudall fever -nerciieeietrrerseneret rere terereee 11 10 9 or 10 Gadel) Arye Cuvier (J. c. p. 841) suggests that more probably these might have been portions of the skulls of a Ruminant. But I see no a priori reason against Major Imrie’s suggestion. On this point a few remarks will be found in the sequel. ? Phil. Trans, vol. lx. p. 412, 1794. * Op. cit. p. 339. QUATERNARY FAUNA OF GIBRALTAR. 59 feline species, one of which was probably the Felis caligata of Egypt and Abyssinia, together with a large species of Bear, apparently intermediate between U. ferox or priscus and U. arctos, which there is reason to believe at that period also abounded in the opposite regions of Africa *. 2. That while some of these species, as the Elephant and Rhinoceros, are now wholly extinct, and known only by fossilized remains of the Pleistocene epoch, others, as the Spotted Hyena, now only exist in the widely distant regions of Southern and Western Africa, being wholly unknown in the northern parts of the continent. 3. That some, again, as the Leopard and Booted Cat, are now no longer found in the European area; whilst others, as the Southern Lynx and Spanish Ibex, are still frequent in the mountainous regions of the Iberian peninsula. 4, That the entire fauna exhibits purely African affinities. 5. That the occurrence of Hlephas antiquus*, Rhinoceros hemitechus, Hyena crocuta, Felis pardus, F. caligata seu caffra, &c., at this remote southern point of Europe, unmixed with species of northern origin, serves strongly to indicate that those species made their way (in part at least) to the more northern regions, where their remains occur in Pleistocene deposits, through the isthmus at one time connecting Europe and Africa at the present Straits of Gibraltar. II. GenERAL REMARKS ON THE MAMMALIAN REMAINS. The Mammals whose remains alone constitute the subject of this paper belong to the following genera :— I. Carnivora. 1. Ursus. 2. Hyena. 3. Felis. 4. Canis. II. PERIssopAcTYLA. 5. Equus. 6, Rhinoceros. III. Arriopacty.a. A. Ruminantia. 7. Cervus. 8. Capra. 9. Bos. B. Non-Ruminantia. 10. Sus. IV. RoDENTIA, 11. Lepus. V. PROBOSCIDEA. 12. Elephas. 2 On this point, vide Cuy. 1. c. tom. vii. p. 200. ? Can this be the Elephant that, according to Pliny, at one time abounded on the Atlas range ? 60 MR. G. BUSK ON THE ANCIENT OR III. Ursus. The Genista Cave has yielded a considerable number of bones of a species of Bear, all of which were procured from the deeper passages. They are completely fossilized, and thickly incrusted with hard ferruginous stalagmite. The principal specimens are :— 1. A portion of the right maxilla (Pl. V. fig. 3), extending from the anterior border of the first true molar to a short distance beyond the second, and containing both teeth quite entire though a good deal worn. ‘Their dimensions are, m. 1 ‘93°70, m.2 1"-5x°8. All the finer sculpturing is worn away ; and the third or posterior cusp on the outer border of m. 2 is entirely obliterated. What remains of the sides of the tooth indicates that the crown-surface must originally have been much compressed. 2. The right mandible (Pl. IV. figs. 1, 2). This specimen comprises the whole of the horizontal ramus as far back as the angle, and showing the small inferior wncinus!. The ascending ramus, including the condyle, coronoid process, and angle with the ‘“‘ crochet” *, are wanting, having evidently been detached before the bone was invested with stalagmite. ‘The remainder of the jaw is complete. It includes the following teeth or empty alveoli :— The empty alveoli of the three incisors; the fang of the canine, broken off level with the border of the alveolus ; the empty alveolus of the first premolar (pm. 1) ; the absorbed alveolus of the third premolar (pm. 3), which was probably shed from violence ; the fourth premolar (pm. 4) and first molar (m. 1) entire, and corresponding in amount of wear with the maxillary teeth above described ; and the empty alveolus of the last molar (m. 2). The length of the ramus from the incisive border to the extremity of the uncinus is 8:35, its vertical diameter immediately behind the last molar 2”°5, and at the diasteme 1-9; the length of the diasteme is 1-7, and of the symphysis 3’"0. The teeth or sockets afford the followimg measurements :— in. Gamine ssf) 205), 52, wou) ue ee ae ee ean OST: Pts hy Pd ee lS Me gram cin E 8 ee eee Oper) M1 &, 03? ey ote ee Gy Se ee ICO peco.U) ie i ee oun ae en go Hb cor) m. 3 Se otal ie ek es bags a a eT 0.0) Molarisenies).. <9.) ure eta Nace 0) The lower border of the bone is thick and rounded but nearly straight. 3. The third specimen is about half the right ramus of a young animal, in which the canine is only partly protruded from the alveolus. It contains that tooth entire, * « Crook-process” of Falconer. * « Angular process” of Falconer. QUATERNARY FAUNA OF GIBRALTAR. 61 the open alveoli of the first and second premolars (pm. 1 and pm. 2), the latter much smaller than the former; the fourth premolar (pm. 4) recently broken off, but the fangs remaining in the alveolus; the first molar perfect and unworn, and about two thirds of the second molar. ‘lhe vertical diameter of the ramus at the second molar tooth is about 1-4, and the same at the diasteme. The diasteme measures 1”-0, and the symphysis 2-0. The teeth measure :— in PMP ahs earns sie, Tn Feu intiuas «jz eOORIee Pt mers 2h ees oe trylelbep e “LO AO [Ong Cah er ee) Men See Ror ieee Supe vay RAO SG sil Fenn a a er ah ah a 8 it ete acto LEO SEO The Speier tune Mirtle vee ad yt isl Od The teeth in other respects correspond exactly in pattern with those in the other mandible above described, the only difference at all worthy of note being that the first molar is slightly narrower. 4. The next specimen is the anterior part, also of a right mandible (Pl. IV. figs. 3, 4), of a fully mature animal, containing the canine with the apex recently broken off, the open alveolus of the first premolar (pm. 1) close to the canine, the fourth premolar (pm. 4) fortunately quite entire though somewhat worn, with a portion of the alveolus of the first molar (m. 1) filled with stalagmite. There are no traces of the alveoli of the second and third premolars. ‘The fracture by which this portion was detached from the remainder of the jaw is ancient, the surface being covered with hard stalagmite ; and it may not very improbably have been the work of the Hyena. The diasteme measures 1”-7; and the vertical height of the ramus at that part is 1-7; the length of the symphysis 2’°5. The teeth measure :-— in. ames Mh SAM eo ea ah A auleoe OG FREY oy yee RANE aed Hist ele 2099220 FRA ic ig de Ue RNa ee ee, Jee Gin Pa TZ This specimen is particularly valuable in a diagnostic point of view, from its presenting the fourth premolar in a tolerably perfect condition. The dimensions of this tooth show its small size as compared with the other teeth; and in its pattern it also corresponds exactly with the usual form of the same tooth in U. arctos, as will after- wards be more particularly pointed out. 5. The posterior half, or nearly so, of the left mandible of a young animal, but apparently not the same individual as that to which the immature mandible above described belonged, since the teeth appear to be rather larger. The specimen has been much comminuted; and different portions were found widely apart; so that at the time when the figures were prepared the several fragments had not been recognized as belonging to the same bone, and consequently one portion only of it is represented VoL. X.—Part 11. No. 2.—Augqust 1st, 1877. K » 62 MR. G. BUSK ON THE ANCIENT OR in Pl. V. figs. 4, 5, whilst that containing the second molar is shown separately in fig. 7. The specimen when put together exhibits the small angular crochet and a mere rudimentary wncinus, about half of the condyle, and a portion of the coronoid process, together with a part of the horizontal ramus containing the two hinder molars m. 2 and m. 3 in beautiful condition. These teeth measure :— in. Ti pe Or ye oe peel ee Ae ee ky epee etn Oem 9° x6 E Oo) The third molar is oblong with parallel sides, but constricted behind, chiefly on the inner side. There is no trace of a sulcus on the outer border (vide infra); and the surface is very faintly sculptured. 6. The only other specimens connected with the dentition are two detached canines, both of comparatively small size, and resembling in general characters those of U. arctos. They measure *9 x‘55 and ‘9x ‘G5 respectively. One is figured in Pl. V. fig. 6. Of the other parts of the skeleton those chiefly worth noticing are :— 7. A nearly entire axis vertebra, wanting only the posterior epiphysis of the centrum (Pl. VI. figs. 7, 8, 9). The body, including the odontoid process, and allowing 0-1 for the missing epiphysis, measures 2”°7 in antero-posterior length. The width from side to side across the anterior articular processes is 2-6, and the length of the spine is also 2-6; the neural canal is 1’-1 in diameter. The width from side to side at the posterior articular facets is nearly 2”°3. The width of the body at the narrowest part is 2-0, and the entire height of the bone 21. 8. The proximal half, or rather more, of the left ulna (Pl. V. fig. 1) of an old and apparently very muscular individual. The olecranon, measured in the antero-posterior direction from the upper point of the greater sigmoid cavity, is 2-6 wide, and the dis- tance from the same point to the point of the coronoid process 2-0. Thelength of the lesser sigmoid cavity is 1’°8. The corresponding measurements in an ulna of U. horribilis (ferox), Baird, from California, are 2-7, 2-0, 1’°8, or as nearly as possible the same. In another speci- men of the smaller variety of the Grizzly Bear (U. richardsoni, Baird), the dimensions are 2-5, 1-9, 1’5. 9. The corresponding part of a second right ulna of a younger and less robust animal, but otherwise agreeing with the above. 10. The proximal end of the right radius, which fits to the last-mentioned ulna, and no doubt belonged to the same skeleton. Its proximal articular head measures 1-7 x 1-25, and the least circumference of the shaft is 2’°6. , 11. One of the most remarkable among the ursine remains consists of the nearly entire right tibia, with the upper end of the fibula attached above, by bony ankylosis, QUATERNARY FAUNA OF GIBRALTAR. 63 to the side of the head of the tibia, and at about the middle of its length to the shaft of that bone by a very thick mass of callus. The bones, which are comparatively of small size (the distal end of the tibia measuring 1”-5 x 2’-7), appear to have suffered compound fracture some time before the death of the animal, and to have become re- united in a very irregular though firm fashion. The consequent distortion is so great, that the transverse axis of the distal articular surface is twisted round about a quarter of a circle; and as the lower fragments of both bones override the upper to a consider- able extent, the limb must have been shortened as well as distorted. It is not easy to understand how a wild predaceous animal could have contrived to maintain the struggle for existence during the long period required to effect even such an imperfect cure, or even afterwards, in such a mutilated condition. The instance would certainly seem to show, at any rate, that sometimes “ swvis inter se convenit ursis.” 12. Besides the above, there are several metacarpal and one metatarsal bone, belong- ing to perhaps four individuals, differing a good deal, as it would seem, in size and age. These are represented in Pl. VI. figs. 1, 2, 3,4, and 5 (fig 10 is the fifth metacarpal of Felis pardus). These bones will be afterwards more particularly referred to. Notwithstanding the number of these ursine remains, they hardly supply sufficient materials for the determination, with certainty, of the species to which they belong, which must, I fear, in the absence of further evidence, be left in considerable doubt. Some of the more important characters by which the existing and fossil species of Bears are distinguished are afforded by the skull and face. No portion of either, with the exception of a fragmentary maxillary, is contained in the Gibraltar collection; and as the remaining parts of most diagnostic value, as the lower jaw and teeth, exhibit characters intermediate, as it were, between U. fossilis sive ferox and U. arctos, it appears to me impossible to decide whether the Gibraltar Bear should be referred to the one or the other of those species or, it may be, to a third distinct from either. That these remains have no relation whatever to U. spelwus is sufficiently obvious from the dimensions and other characters of the teeth, and notably of the fourth lower premolar and last molar (pm. 4 and m. 3), together with the presence of an open alveolus of the first premolar in all three specimens of the mandibie, and also from the size and proportions of the other bones, more especially of the metacarpals, metatarsal, and phalanges, which, of all the bones of the skeleton, are perhaps most characteristic of U. speleus. ‘The Great Cave-Bear may therefore at once be dismissed from considera- tion ; and we may proceed to inquire which of the other known existing or fossilized specimens afford the nearest points of resemblance. These species are but few in number, and may, I think, at any rate for paleontolo- gical purposes, be included under the following specific types :— 1. Ursus Frossiis, Goldfuss. U. priscus, Cuy. U. ferox fossilis, mihi. U. bourguignati, Lartet. K 2 64 MR. G. BUSK ON THE ANCIENT OR 2. Ursus FEROX, Richardson. U. horribilis, horridus, richardsoni, Baird. U. cinereus, Gray. U, piscator. 3. Ursus arcros, Linn. U. fuscus, niger, Alb. Magn.; Goldf. U. norvegicus, pyrenaicus, collaris, &c., F, Cuv. U. isabellinus, Horsfield. U. syriacus, Hempr. & Ehr. U. cadaverinus, formicarius, longirostris, Eversm. 4, URsvs LARTETIANUS. 5. URsus FAIDHERBIANUS. 6. Ursus LETOURNEUXIANUS. 7. URsus ROUVIERI. Bourguignat. 1. Ursvs rossiuis, Goldf. From the time of Goldfuss' all paleontogists, except Blumenbach and De Blain- ville, have recognized at least two distinct specific forms amongst the Ursine remains found in caverns. To one of these, basing his description upon a perfect cranium, with the lower jaw, found in the deepest part of the Gailenreuth cavern, Goldfuss applied the term U. fossilis°. This form has appeared to me to coincide so very closely with the existing U. ferox, or horribilis, of North America, that I was induced some years since to suggest that they might be regarded as specifically the same, so far as cranial and dental characters are concerned. Regarding, therefore, this second species of Cave-Bear as undistinguishable by dental and osteological characters from the Grizzly Bear and its varieties, what is here said of the one, in comparing it with U. arctos, will apply to the other. I have already observed that some of the most important distinctive characters between these very closely allied forms are found in the cranium and face, parts which are not afforded in the the Gibraltar collection; the comparison, therefore, can be only very incomplete and inconclusive. The parts upon which I have been compelled principally to rely for the means of diagnosis are the horizontal ramus of the lower jaw, the dentition, and to some extent the axis vertebra. With respect to the other bones of the skeleton, my own observation leads me quite to agree with A. Wagner *, who remarks that after twenty years’ study of bears, fossil and recent, he considers that no characters can be drawn from any of the bones of the skeleton except the skulland teeth. The only differences, he says, may be regarded as individual, except as respects the metacarpals and metatarsals, and, he might have added, the phalanges. + Goldfuss, Ac. Cres. Leop. Nova Acta, x. 1821, p. 449. * Cuvier (op. cit. vii. p. 242) says that Goldfuss had given the name of U. priscus to this skull, but upon what authority I am not aware. Tho term employed by Goldfuss is U. fossilis. * Wiegm. Archiv, 1843, i. pp. 24-42, QUATERNARY FAUNA OF GIBRALTAR. 65 As regards the lower jaw, of which we have little more than the horizontal ramus, the only difference that I can perceive between it and the corresponding portion in three mandibles of U. fossilis from the Gower caves consists in the greater thickness of the interior border, which is very much thinner in the Welsh specimens, the jaws otherwise corresponding exactly in dimensions as to vertical diameter and thickness. The important differential character afforded by the angular “ crochet,’ which is thicker and more incurved in U. fossilis and ferox than in U. arctos, is unfortunately absent in the Gibraltar bone, as are also those derivable from the ascending ramus, coronoid process!, &c. The two remaining teeth in this mandible are rather smaller than in the Gower specimens, but very slightly so; and the most important of them in a comparative point of view, the pm. 4, is wanting, having been shed during life, and the alveolus wholly obliterated. This tooth, however, exists quite entire in one of the other specimens, and corresponds in all respects much more closely with that tooth in U. arctos than with that of U. fossilis. The important characters afforded by the pm. 4 in U. speleus are very well known; and, to a less decided extent, it seems to me to afford one of the most important means of diagnosing U. arctos from its most closely allied congener. In studying the dental characteristics of different species of Bears, we have to regard, first, their size and proportions to each other, and, secondly, any differences of pattern they may present. With regard to the former of these points, as will be obvious from the odontograms (Pl. X XVII. nos. 6-10), no conclusions could in the present case be drawn, except that the teeth in the Gibraltar specimens are, with perhaps one ex- ception, fully as large as those of the existing U. ferox (No.7), or even of U. fossilis (No. 6), and, on the whole, somewhat larger than the largest known to me of U. arctos (No. 10). As regards the second point, it may be observed that, unfortunately, the general resemblance of most of the teeth in all the larger carnivorous Bears, as regards form and pattern, is so close, and moreover so liable, within certain limits, to vary very considerably, that but few of them are practically available for the purpose of drawing specific distinctions. From the circumstance, also, that a very moderate amount of wear destroys the more minute particularities of sculpturing, characters of that kind can seldom be employed with any advantage. For these reasons, it will generally be found sufficient to advert only to those teeth which afford the most marked and least readily effaceable features. These appear to be the second or last upper molar (m. 2), and the first and last molars of the lower jaw (pm. 4 and m. 8). The last upper molar in U. speleus, besides its much greater size, is usually distin- guished by its more or less oblong form, the sides being nearly parallel, and the hinder Professor Owen (Brit. Fossil Mamm. p. 83) adduces the greater breadth and height of the coronoid process as a point of difference between U. priscus and. U. arctos. 66 MR. G. BUSK ON THE ANCIENT OR end not much narrower than the middle, and never, or scarcely ever, at all pointed. The grinding-surface also, in the perfectly unworn condition, is characterized by its com- parative flatness or expansion transversely. On the outer border there are three cusps, the hindmost of which, however, is very low and soon worn off. Of the two larger conical cusps the anterior is stronger than the posterior, which latter presents no trace of an accessory intermediate tubercle in front. In U. ferox and U. fossilis the tooth is more contracted behind than in U. speleus, but less so than is usually the case in U. arctos, and the attenuation may be said to com- mence further back. The grinding-surface in the unworn tooth is also much less com- pressed from side to side than it isin U. arctos, though more so than in U. speleus. Of the three outer cusps, the anterior two are more nearly of equal size than they are in U. speleus, and there is no appearance of an accessory intermediate tubercle on the hinder of the two; the third or posterior cusp is sometimes entirely wanting, and always of insignificant size, and soon worn away. In U. arctos the crown in the perfectly unworn or germ condition is much more com- pressed, the two sides falling in, as it were, so as to leave merely a wide furrow between them. ‘There are only two cusps on the outer border, of which the anterior is con- siderably the larger, and the hinder in most cases has a small portion in front con- stricted off so as to form a small accessory tubercle intermediate between the two cusps ; and in this species the internal basal cingulum would appear to be usually less developed than in the others. The third lower molar (m. 3) in U. speleus is in most cases readily distinguishable by its peculiar form, which is also, from what I have observed, tolerably constant. The form of the tooth corresponds, in fact, as might be expected, with that of the expanded upper molar to which it is opposed. The form of the tooth may be described as oblong or subquadrangular, usually with an angle posteriorly. The outer border is divided into two distinct though low cusps by a deep sulcus a little behind the middle of its length. The grinding-surface is very minutely and richly tuberculated or, as it may be said, granulated. The anterior and internal cusp is comparatively little developed. In U. fossilis the much smaller tooth is usually of a subtriangular form and generally rounded behind. Sometimes it is more elongated; but even then it is readily distin- guishable from that of U. speleus by its greater attenuation posteriorly (corresponding with the upper molar) and other characters. In the typical triangular form there is no sulcus on the outer border, or but a very faint one close to the hinder end; and this is soon destroyed, if it ever existed, by slight wear. When the tooth is more elongated the outer side presents a shallow sinus, subdividing the border, as in U. speleus, into two unequal low cusps. The anterior and inner angle of the tooth is raised into a strong conical cusp. The grinding-surface is coarsely plicated or ridged, and rarely tuberculated or granulated in the interstices of the ridges }. ‘In the figure of the mandibular dentition of U. priscus, in Brit, Foss. Mamm. (fig. 35 B) the m. 3 QUATERNARY FAUNA OF GIBRALTAR. 67 In U. arctos the tooth is also triangular in form, but usually more angular behind, in accordance with the more pointed shape of the corresponding upper molar. It has usually no constriction on the outer border; and the anterior and inner cusp is generally but little developed. The grinding-surface presents only a few coarse folds or ruge, and is never, so far as I have seen, tuberculated or granulated in the slightest degree. ‘The fourth lower premolar (pm. 4) is in some respects the most characteristic of all the teeth; and were its characters rather less liable to vary by defect in U. fossilis and U. ferox, it might almost by itself be considered sufficient to give assurance of the species. Its distinctive characters in U. spele@us are well known; and when fully developed and unworn it seems to me to afford excellent distinctive characters also between U. ferox and U. arctos. In Ursus speleus, besides the principal cone, there are usually on the inner side two, and always one, smaller cusps, of which one is anterior in position to the principal cusp. In all other Bears this tooth has either a single conical cusp, or at most a single small internal tubercle posterior in position to the principal cusp, and corresponding to the hinder of the two internal cusps in U. speleus. In all the large carnivorous Bears this tooth presents séveral common characters, the differences exhibited in various species depending solely, as it would seem, upon the degree of development or suppression of minute parts. In all cases the tooth presents a large conical cusp, which is placed nearer the anterior than the posterior border of the crown. An acute ridge or keel, in perfectly unworn teeth, descends from the point of the cusp in front to the anterior end of the crown, where it terminates after making a slight curve inwards, in a more or less distinct though always very minute tubercle. Two similar but more strongly marked ridges descend from the apex of the cone towards the hinder border of the tooth. In fully developed and perfectly unworn teeth in U. fossilis, ferox, and maritimus (and, I have no doubt, also in U. arctos, though I have seen no tooth of that species young enough to show it), these hinder ridges are more or less distinctly serrated, especially the outer one, which is always continued to the hinder border of the tooth, whilst the internal ridge rarely reaches more than halfway between the base of the cone and the hinder border. Now the differencesin the fourth lower premolar, as between U. fossilis, JSerox, and arctos, consist solely in the varying degree of development of these minute parts. In well-marked teeth of U. fossilis taken in the germ-condition, the anterior carina and the tubercle at its base are strongly developed. ‘The outer of the two hinder carine, which is deeply serrated, terminates at the posterior border of the crown in a small tubercle, on the inner side of which is placed a second tubercle of equal size ; so that the hinder extremity of the crown (or the talon, as it may be termed) might in appears to be deeply sinuated on the outer border ; but this appears to have arisen from the circumstance that the figure has been taken from Goldfuss’s specimen in the British Museum, in which the tooth has a piece chipped off at that spot. 68 MR. G. BUSK ON THE ANCIENT OR well-marked instances be described as bifid. Sometimes, instead of two distinct tubercles, there is a row of several smaller ones; and it not unfrequently happens that, owing to pressure against the next tooth, the appearance of any tubercles in that situation is destroyed. The inner of the two posterior ridges, after descending a short way, ends in a distinct and sometimes considerable-sized somewhat pointed denticle or accessory cusp, which is that above referred to, and is that noticed by Prof. Owen in speaking of U. priscus!. In a specimen of U. ferox (No. 1137 6, B.M.), although the two divari- cating posterior ridges are quite distinct, there is no appearance whatever of the inner tubercle or cusp; nor is there any indication of its having been worn off; and the hinder tubercles, if ever present, have been removed by pressure. But in a second specimen of an older animal (No. 11374, B.M.), in which the ‘tooth is much worn, the site of the internal tubercle is still quite visible, as well as the presence of the bifid pos- terior talon; and in a specimen, also a good deal worn, of the same species in the College of Surgeons, the same appearances are distinctly shown. In the typical speci- men of U. priscus ( fosstlis) in the British Museum, whose teeth are also much worn, the inner tubercle is very faintly indicated, nor is the bifid talon distinctly discernible. In general characters, therefore, the tooth exactly resembles that of U. ferox (No. 1137 4, B.M.). In U. arctos the pm. 4 is distinguished, besides its usually much smaller dimensions, by the extremely minute size or total absence of the inner tubercle, which in this species, so far as I have seen, rarely exceeds a large pin’s head in size, and is often wholly wanting. ‘he posterior talon sometimes exhibits a very minute tubercle, but is more usually quite smooth, and never distinctly bifid asin U. ferow. On applying these observations on the teeth to the diagnosis of the Gibraltar Bear, it appears to me that that form occupies a position intermediate, as it were, between U. arctos and U. ferox. 1. As regards the second upper molar. In the Gibraltar jaw this tooth resembles that of U. ferox, besides its size, in its comparatively slight attenuation behind, and that of U. arctos in the apparent falling inwards of the side, though to what extent this reached cannot be determined in the present worn condition of the tooth. 2. The third lower molar also resembles that of U. ferox much more nearly than that of U. arctos in its greater dimensions and more oblong form, corresponding with the larger and longer upper molar; but there is no appearance whatever of a sulcus in the outer border, such as would most probably have been seen in a tooth of the same size in U. fossilis or U. ferou. 3. The fourth lower premolar, on the other hand, in its comparatively small dimen- sions, the diminutive size of the inner denticle, which is not bigger than a pin’s head, and the perfectly simple non-tuberculated inner border, more nearly resembles, in fact quite agrees with, the same tooth in U. arctos. ’ Brit. Foss. Mammals, p. 81. QUATERNARY FAUNA OF GIBRALTAR. 69 The evidence, therefore, as to species afforded by the dentition of the Gibraltar Bear, is not very satisfactory, and leaves it doubtful whether it should be referred to U. fossilis or to a large form of U. arctos. Another specimen in which we may seek for further evidence is the axis vertebra, which seems to present some distinctive characters from that of U. ferox; but these are perhaps not very marked, nor probably of much intrinsic importance. And it must be remembered that the Gibraltar bone is of a young animal, and that with which it was compared of fully mature age. The comparative dimensions of these bones are given in the following table. Comparative Dimensions of Axis of Ursus ferox and U. ——, Gibraltar. : * | Width across Total length. ee Mpcitiacros Eats Length of | Diameter of | EL cular process. articular spine. neural canal. | process. | n in. in. in. in, 1. Gibraltar specimen ...... 27 2-0 2°7 2°6 1:0 hy TOE tanh ACN as Sticoee | 2:8 1:9 2-4 33 0-9 3. U. arctos, 218c, B.M. .... ia *. ae a 2350 ot 4, U.isabellinus ..........- oe 127) 2:3 pe 2°65 0-6 | From these figures it will be seen that the main differences as regards dimensions are in the greater width across the anterior articular processes, which would indicate of course a longer atlas and skull, and probably, therefore, a larger animal, which is quite in accordance with the comparative dimensions of the teeth in the two cases. The neural canal also is larger in the Gibraltar bone, which is an indication in the same direction, whilst, on the other hand, the total length is slightly less, and that of the spine considerably so—both of which differences may reasonably be assigned to the different ages of the individuals. In other respects it is to be observed that in the axis of U. feror the posterior articular facets are larger and rounder, the transverse processes much larger, the neural lamine thicker, and the odontoid process smaller—all of which differences may also be fairly referred to age and stature. The spine in U. ferox is produced backwards some way behind the level of the posterior articular processes, whilst in the Gibraltar bone it is on a level with them ; and in the former it is much hollowed posteriorly: but these characters may also perhaps be referred to difference of age. I have also compared the axis with that of U. arctos (var. isabellinus*). The latter is of about the same general size, but differs from the former in several particulars, some of which, but not all, can also be regarded as due to age. 1 T may mention that in this specimen of U. isabellinus in the British Museum there are six sacral yertebre. VOL. X.—PART Ul. No. 3.—August 1st, 1877. L 70 MR. G. BUSK ON THE ANCIENT OR 1. The spine, which is of the same length, is 1-3 instead of 1-0, and much more rounded at both ends on the upper edge, and much deeper in front (exactly the reverse of U. speleus). Bnt viewed from above it is thinner and much more constricted above the posterior articulating processes. It is also deeply hollowed behind (as in U. ferox). 2. The body is much narrower in the middle. 3. The anterior articular processes are smaller and more rounded or convex. 4. The odontoid process is smaller. 5. But the most remarkable difference is in the size of the neural canal, which in U. isabellinus is little more than half the greatest diameter of that in the Gibraltar bone. 6. The form of the posterior surface of the centrum is quite different; but as the epiphysis is wanting in the Gibraltar specimen, the difference may perhaps be thus in some measure accounted for. The other, as I have observed, can hardly be assigned to age, nor, as it would seem, to the circumstance that the Isabelline Bear was of smaller size than the Gibraltar one. Hence, again, then, we are brought to the conclusion that the Gibraltar Bear ap- proached more nearly to U. fossilis or U. ferox than to U. arctos (var. isabellinus). As regards the metacarpals, I need only remark that, having carefully compared them with those of U. ferox, of which a good specimen has lately been acquired by the Royal College of Surgeons, I can perceive scarcely any difference worthy of note between them. So far as dimensions are concerned, this will be obvious from the accompanying Table (see Appendix, opposite). It will be seen, from this Table, that the Gibraltar metacarpals must have belonged to four individuals, and that the largest of these most nearly coincides with the speci- men of U. ferox in the College Museum, which, it is to be remarked, is of small size for that species, and, to judge from the wear of the teeth, rather old. In the fifth metacarpal of U. ferox the crescentic tuberosity on the outer side of the head is more pronounced, whilst in the Gibraltar bone it is wider just above the lower capitulum. But the muscular impression}, or tuberosity, on the outer side below is more pronounced in U. ferov. The perimetral index is slightly greater in the Gibraltar bone; and its shaft very slightly more compressed. In the second metacarpal, of which there are two specimens in the Gibraltar col- lection, one considerably smaller than the other, though quite mature, rather more difference is perceptible. In the first place, the proximal end is more produced be- hind than in U. ferox, and the shaft in the smaller specimen very much more com- pressed ; but in the larger of the two this compression is not observable. The distal articular head in both is smaller, and less rounded than in U. ferox. ’ For insertion of the peroneus brevis? | To face page 70.) Me. 5 i | ee | | | to | | | oi 1) E | . é [Pees a | 4 es Pe ee dil i | | 500 | 3-4 100x95 65x80 | 1-7 | -500 | 35 | 120x90'70x100 1:9 | -545 } | | | | | | | | | 400 || 29 | 90x80] 60x75) 12 | -413 ‘756 | 37 145x130 82x85 | 28 | -756 | 36 | -491 | 3-3 (100x100 70x90 18 | -545 | | 38 |100x x 100 | | -419 | 31 | 75x75 | 50x70 | 1:3 | -419 | -483 | 32 |100x90 | 65x80 16 | -500 447 || 3-9 120x130 70x 90 1-9 “487 | | i.=perimetral index. | older. APPENDIX, Table showing Comparative Dimensions of Metacarpals of specimens of Ursus. {To face page 70.) Me. 1. Me. 2. Me. 3. h Me. 4 | ae 4 a a 4 4 a ae SLED aes Wes ety) eS Le Le lect lag Il 2 | Gibraltar 1...... Sengienieng Seams ere 3 82 | 95x70 34 /100x95/65x80 | 1-7 | -500 Senate ls panands gob: Secannpsan = Bei wl eosulllaee tere re ae . Poe “ - - os i] as = ze 7) | + | 2 | 35 |120x90)70%100) 1:0) | +545 | olaer. Sr ponerceeeer dooce IG “8 = - |. |.» |) 20) |90x53/60%70) 14 | +489 : 3 | eee Pecocrircee cen Ggepe a oa . | .. | « | S12 |00x60) 70x75) 16 | -516 53 38 ce | ove | ef) 285 190x565 | 60x70} 116) 408 | 285 | 85x55) 6Ox70)), 11 | +386) 3:0) | 90x60 | 6x70) 12) 4001 29 | 90x80) eox75| 12 | 119 U. spelows, 43740, BM, .....-...- ..| 285 [85x12074x90) 18 | -6al | 3:5 |130x 9390x115 2-35 | B71 | 36 |110x96|100x102) 25 | 694 |) 3:7 |140x 106/100 x14) 28 145x180) 82x85! 28 | «759 iTAgpe mealies 525 ace es AonY 35254 | rece essa ace eerste S| gee so | eee estemet af bo se [Pape | PBS 58 > mp, AERUSS cece coscuseseenoeers| 29 |65x90/60x70) 14 | -482 |) 29 |100x6070x75| 16 | 851) 3:05 | 96x65 | 65x70) 15 | -491 | 3:05 | 95x70 | 75x85| 15 100x100) 70x90) 1:8 | -545 U. bourguignati .... 6-66. cee eee es) ve rig no 56 ee i ws “a “ as a0 als np on oe 26 aA S43 pia -. || 88 [00x x 100 p e ji UT, arctos, Sweden, 218 cor 1182u, BM.) 262 |51x60}40x50] ,. | .. | 29 |75x40/52x56) 11 | -345 |) 30 | 80x50 | 50x55) 14 | -B66 | 31 | 80x50] 50x60| 13 | 410) 31 | 75x75 |50x70| 13 | ag : U, isabellinus, 1010 9, BAL. -. 27 |80x90/60x70) 1-2 | 444) 29 |1-0x8060x80) 14 | 483 |) BO | 90x70) 65x70) 14 | 466) 31 | 90x70 | 70x00| 15 | 493) 32 lt00x90| 65x80) 10 | -500 U. arctos, 2181, BM. .. 2... -... 202. 32 pe x110(60 x70) 14 | 437 | 36 noxs0 65x80) 17 | 472) 87 [10x70 | 70x70) 155| 418 | 3:8 110x70| 70x00) 17 | 447) 39 |120x130| 70x90) 1-9 | 487 ‘ EXPLANATION oF Sysinors. px=proximal extremity. dx=distal extremity. 1. c,=least circumference. p-i.=perimetral index. QUATERNARY FAUNA OF GIBRALTAR. 71 The only difference in the fourth metacarpal is the greater compression of the shaft in the Gibraltar bone. The first metatarsal is of exactly the same length as that of U. ferox; and the only difference between the two is that in the Gibraltar bone the proximal articular surface is broader and Jess produced at the inner and posterior angle. As regards U. arctos, as the only metacarpals belonging to that species that I have been able to compare with the Gibraltar bones and those of U. ferox are from a young individual, and in which the epiphyses are not fully completely united, though nearly so, 1am not able to say more, making allowance for their much smaller size, than that they appear in the compression of the shaft of the fourth and fifth, and in the comparatively small size of the distal capitulum, to bear a very close resemblance to the Gibraltar bones 1. Doubt being thus left as to which, if either, of the two generally recognized forms above noticed the Gibraltar species should be referred to, it will be interesting to recall the circumstance of the discovery, about the year 1866 or 1867, by M. Bourguignat, in a cavern at Djebel-Thaya, in the province of Constantine, in Algeria, of abundant ursine remains, which were considered by him to belong to four distinct species, dif- fering considerably, as it would appear, not only in size, but also in the relative pro- portions of the bones of the extremities, the teeth, &c. The first published notice of this discovery appeared in 1867 2, in which a brief description is given of a form upon which M. Bourguignat has bestowed the name of U. faidherbianus, founding his diagnosis, however, solely upon the lower teeth. In the next year he published the discovery, in the same cavern, of three more forms, to which he assigns the rank of species, viz. U. lartetianus, letowrneuxianus, and rou- vieri®. M. Bourguignat was led to conclude, upon evidence which he has not, so far as I am aware, yet published, that these different forms belonged to different epochs, which nevertheless appear to have overlapped each other. The oldest form, to which he assigns as its latest date 8500 B.c., is U. lartetianus ; the next in point of antiquity is U. letowrneuxianus, which came down from about 8000 to 3500 B.c.; whilst the other two (U. rowvieri and U. faidherbianus) are traced to quite a recent epoch, and even, according to M. Bourguignat, may be still existing or have but very recently become extinct. It is much to be regretted that M. Bourguignat has not as yet given more detailed ! It is much to be regretted that neither in the British Museum nor in the Royal College of Surgeons are there any satisfactory materials for studying the osteology of the Common European Brown Bear in the wild state. The bones of long-caged animals are so generally deformed, and especially in the Bear, which seems to peculiarly liable to chronic rheumatic arthritis, as to’ be wholly useless for any purpose of palontological comparison. * Notice sur un Ursus nouveau. Paris, 1867. 3 Notice prodromique sur quelques Urside d’Algérie. Paris, 1868. bo L 72 MR. G. BUSK ON THE ANCIENT OR particulars of these supposed species, since those contained in the short notices above cited are insufficient to allow us fairly to judge of the correctness or not of his determinations, and especially in a case in which we are required to accept the extraordinary discovery of four entirely new specific forms in such a limited loca- lity. And I am obliged to confess that, having been allowed, through M. Bour- guignat’s extreme kindness and liberality, to study his collection of Ursine remains from the Thaya cavern, they hardly appeared to me to present characters which I should have thought sufficient (considering the extreme variability of all Bears) to justify the distinctions he has set up. As an instance, I may mention that M. Bourguignat divides his four species into two groups, one of which, consisting of the two ancient forms U. lartetianus and U. rouvieri, is characterized by the presence of a perforation at the bottom of the olecranon-fossa, which he regards as a special character distinctive of certain African Urside, as contradistinguished from those with- out a perforation, which he considers to belong to a European type. Leaving on one side the evidence upon which M. Bourguignat may rest in attributing this or any other character to African Bears, I would merely remark that more extended ob- servation has perhaps since convinced him that such a character is of no value whatever, as it may occasionally be observed, certainly in U. spelwus, and probably in all Bears, as it is also in Man and many other mammals. But as regards the Gibraltar Bear, it is matter of considerable interest to inquire whether it may not have an intimate relationship with one or other of these ancient Algerian forms. Probability is highly in favour of such a supposition. And the question then arises, What is or what are the known species to which M. Bourguignat’s Bears most closely assimilate ? Our means of judging with respect to this are at present very limited; but, to judge ~ from the lower dentition of U. faidherbianus (Odontogram 9 a, Pl. XXVII.), there is nothing opposed to the supposition that it represents U. arctos, or a small form of the ferox type, from the comparative width of the fourth premolar, which is greater in U. fossilis and U. ferox than it usually is in U. arctos’. ; The only other of M. Bourguignat’s forms of which I have any data is U. letowrneux- ianus, of which the maxillary dentition is shown in Odontogram No. 9, Pl. XXVII.; and from this it would seem to have been a Bear with teeth in size fully above the mean of U. speleus, and with a second upper molar much larger than in any U. fossilis or U. ferox that has come under my observation. The presence, however, of the first and third premolars shows that it did not belong to U. speleus. But one of the most interesting points connected with M. Bourguignat’s discovery of Ursine remains in the Algerian cavern is the establishment, beyond doubt, of the * In U. arctos I have not as yet met with an instance in which the thickness of pm. 4 reached 0'"3, it being usually 0'-25 ; whilst in U. ferow it is always at least 0'"3 thick, and sometimes 0'"35, QUATERNARY FAUNA OF GIBRALTAR, 73 existence of a large Bear on the African continent from the period when it was still continuous with Europe down to a comparatively recent period, if not to the pre- sent day, although zoologists seem disposed to dispute its present existence in the north of Africa. M. Bourguignat, in his Notice &c. (1867, p. 4), gives a very interesting summary of the evidence afforded by various writers on the subject of the former and present exist- ence of the Bear in North Africa, which appears to have been disputed as far back as by Pliny, who nevertheless cites historical evidence to the contrary. With respect to the present or very recent existence of a Bear resembling the European Brown Bear, the authority of L’Abbé Poiret? is quoted, who asserts that in the Atlas Moun- tains the brown Bear is found, and is very carnivorous. During his stay at Mazoule an Arab brought him the skin of a Bear which he had killed. In 1841 Mr. Edward Blyth? related the capture in 1834 of a couple of Bears near Yetuan, from particulars furnished to him by Mr. Crowther, an officer of the 63rd Regiment. On these particulars was founded by Schinz* a new species, U. crowtheri, which was adopted by Pucheran*, and by Gray’. Lastly, according to Capt. Loche’, author of several works on the mammalogy of Algeria, the Brown Bear (U. arctos) would appear to exist in the Atlantic chain of mountains in Morocco, whence it often invaded the French provinces. And the same writer states that he had seen at Mar- seilles, seven or eight years before, a Brown Bear which had been sent by the Emperor of Morocco. In addition, however, to this evidence, M. Bourguignat adduces that of M. Letour- neux, a councillor of Algiers, who states that whilst he was “ procureur impérial” at Bone he learned from the Arabs of the Edough that formerly, according to tradition, Bears had abounded in that country, which committed great ravages in the vineyards. On another occasion he was informed by the Arabs of Ouled-sidi-Bekri that Bears had infested their mountains within fifty years; and a man related that one of the last Bears had been killed by his father. According to these people, the Bears in question were small, thick-set, of a brown colour, with a white spot under the throat, very fond of honey and fruit, and when fighting raised themselves on the hind legs. The Caid Boa-Roabi of Zardeza, whose district reached almost to Thaya, assured M. Letourneux that he had often seen traces and heard cries of the Bear in the moun- ’ Voyage en Barbarie, ou Lettres écrites de l’ancienne Numidie pendant les années 1755 et 1786. Tom. i. p. 238 (1789). 2 Proc. Zool. Soc. 1841, pp. 64, 65, + Synops. Mammal. p. 302. 1842. 4 « Esquisse sur la mammalogie du continent Africain,” Rey. et Mag. de Zool. 1855, p, 499. 5 Proc. Zool. Soc. 1864, p. 698, ® Catalogue des Mamm. &e. de l’Algérie, p. 30. 1858. These characters, M. Bourguignat observes, closely correspond with those assigned by himself and M. Lartet to U. fazdherbianus, 74 MR. G. BUSK ON THE ANCIENT OR tains of Gherara Dhebhar; and another Scheik, living close to Heliopolis, stated to him that he had often seen the Bear, and followed it in the evening into the very mountain of Thaya. In further confirmation of the very recent, if not present, existence of a Bear in the immediate neighbourhood of the cavern explored by M. Bourguignat, he relates that he himself, during his exploration of the cavern, noticed in the soft soil large foot-prints, as sharp and fresh as if they had not been made more than an hour or two. They were the footprints of a heavy animal, and excited great emotion amongst the Arabs who accompanied him, whose exclamations of Deb! Deb! the Arab word for Bear’, showed, at any rate, that they were not only familiar with its name, but also not un- prepared to witness its sudden appearance. The existence of a fossil Bear in Algeria has, however, been long well known, a considerable portion of a cranium having been discovered so far back as 1835 by M. Milne-Edwards’ in an ossiferous breccia fifty metres above the level of the sea, in a red calcareous tufa. M. Milne-Edwards, from what he was able to make out with regard to the size and shape of the cranium, was induced to think that, although of very large size, it presented more resemblance to that of U. labiatus than of any other living species. I may now state the conclusions which, as it appears to me, may be drawn from the above evidence, about the Ursine remains from Genista cave. 1. That they belong exclusively to the more ancient fauna. 2. That they afford evidence of at least four individuals, varying in size and age very considerably, one of which has suffered compound fracture of the hind leg, from which it had recovered with great deformity of the limb. 3. That it was a species of large size, and probably equal to the largest existing Brown or Grizzly Bears, but not equal to U. speleus. 4. That it differs essentially in dental and other osteological characters from U. speleus. 5. That the preponderance of its characters is in favour of its being closely related to U. fossilis sive priscus, or to a form intermediate between that and U. arctos, var. isabellinus. 6. That it may have been also closely related to one or other of the fossilized Bears whose remains were discovered by M. Bourguignat in the Cavern of Thaya in Algeria. " [have been lately informed, however, that by “ Deb” the Arabs understand, not the Bear, but the Hyena. * Ann. d. Se. Naturelles 2™¢ sér, Zoologie, tom. vii. p. 216 (1837): “Note sur une bréche osseuse situce entre Oran et Mers-el-Kebir.” QUATERNARY FAUNA OF GIBALTAR. 75 IV. Hyana. The principal specimen belonging to the Hyena yielded by the Genista cave is a fine cranium, in several pieces, but which, when placed together, constitute two portions, which respectively comprise the facial and occipital regions (Plates I. and II.), the intermediate connecting part having been broken or lost in the extraction from the hard matrix. The facial segment presents half of the right orbit, part of the zygomatic arch, and the whole of the maxillary of the same side, together with all the molars in situ, with the exception of the small tubercular. It shows also part of the alveolus of the canine, the incisive border being wanting. On the left side only a small part of the maxillary is left, containing the three anterior premolars. The palate connects the two lines of teeth, and is complete on the right side almost to the extreme posterior verge, exhibiting the large digital fossa within the carnassial, and an indistinct indica- tion of the alveolus of the apparently uniradicular molar. The posterior portion of the cranium includes the whole of the occipital region, a part of both parietals, and the greater part of both temporals, with the auditory foramina and bulle, together with the glenoid fosse and occipital condyles, which are quite perfect. The right glenoid fossa is in a diseased condition; and what remains of the temporal zygomatic process appears to indicate that that process had been dis- eased, or perhaps the seat of an old ununited fracture. The sagittal crest is wanting ; but the occipital area and spine are tolerably entire (fig. 4). The teeth are all much worn; and the general condition of the bones is also such as to prove that the animal was aged. The specimen was found at a depth of 36 feet, in the upper chamber of the Genista Cave; but it is thoroughly fossilized, dense, and heavy. The cerebral cavity and all the hollows are occupied by a thick deposit of ochreous stalagmite, in a mass of which the whole was imbedded. In the subjoined Table will be found the principal dimensions that the condition of the specimen will allow to be taken, contrasted with those obtained in H. spelea (one example), H. crocuta fera (mean), H. brunnea (mean), H. striata (mean) :— 76 MR. G. BUSK ON THE ANCIENT OR TABLE showing Dimensions of the Cranium and Teeth in Hyena. H. crocuta! aan | Gibraltar H, brunnea.| H. striata. paced Tipper. ee. | (Mean.) | (Mean.) inches. inches. inches. inches. inches. 1. Length from occipital crest to incisive border ...... 10°6 10:8 10°7 10-1 9°0 2. », condyles to incisive border .......... 9°9 oles 9°8 9-1 8:2 3. Width between outsides of CHIE sa nconoisaes de 2-1 2:2 2-0 2:0 16 4, Width of skull over auditory foramina............ 3-9 ee 39 3-4 31 5. Width of base of occipital triangle .............. 4:0 6: eighties er Sere et ete erccet es ctrecraitie: 3:8 ie Foramen magnum, vd. and trd. Shae aes «yee 1:0x°:9 Sl Auditory bully erp ew era weer acces al) o2igeail Nao HOERMSE ET odio nua ag Odibd Caqomor 6oo000000 3-4 37 3°34 3:12 2:7 aK) Bess el Meer oreo mere d octacke rc onGroonsers OOo teC ‘80 x -25 | 33 x32 | 30x °27| 27x -25| -23x -22 UMS IPT ER Des eecusreperenctuesteyer si ne Xoy foksooielsptseete ice tapi ae 72x51 |-68 x -43 | 67 x 50 | 68 x -46| -62 x 40 LG hancecn tlonmanto KoUTOoA Sones Cas Ame See 1-05 x °72/1-00 x °73) 94 x +70 | ‘93 x 85 | -80 x 52 WSseRmne A aioe Maleate aes Sie ss blbete toes Steeda telat 1-50 x 80 |1-60 x 87 |1-46 x *84/1-42 x -63)1-17 x -70 DS OND ieecantey cvakeystaictanataperaVat Meet iei aisha ye a hcbstisckeret ss ‘2x —? 0 ‘2x1 |°51x-21) 50x21 The mere inspection of these numbers will be sufficient to show that the Gibraltar fossil Hyena is in all essential particulars, as regards cranial and dental measurements, very closely in accord with Hyena crocuta and H. spelewa, whilst at the same time it offers considerable differences from both H. drunnea and H. striata. Further examina- tion only tends to prove that this similarity is real, and that the Hyena found in the Genista Cave is, in fact, identical with the Spotted Hyena of Southern and Western Africa, and quite distinct from the Striped Hyena of Western Asia and Northern Africa. This conclusion is so contrary to what might have been expected, that when it was forced upon us it could not fail to excite the greatest surprise and interest; and Dr. Falconer went into the question of the specific identification of the specimen with his well-known acuteness and zeal. Unfortunately at that time there were no known materials in London for studying the cranial and dental characters of Hyena brunnea, although there were two skulls erroneously assigned to that species in the British Museum}. The consequence was that Dr. Falconer had no means of determining differences between H. crocuta and H. brunnea, and was misled into the impression that the Gibraltar Hyena was of the latter species. For the same reason he was induced to regard the Hyena fusca of G. St.-Hilaire as distinct from H. brunnea, Thunb., and to adopt the notion that H. maculata, Kaup, was distinct from H. crocuta, Erxl. (sp.). * A full account of the cause of this confusion will be found in my paper “* On the Cranial and Dental Cha- racters of the existing Species of Hyena,” published in the Linnean Society’s Journal, Zoology, yol. ix. p. 59, 1866, QUATERNARY FAUNA OF GIBRALTAR. Hf Since then, however, we have been furnished with abundant materials for the clearing up of what was to him necessarily obscure, and no difficulty now exists in distinguishing between the three existing species of Hyena from their cranial and dental characters alone ; and I may remark that any difficulty even at that time would have beén removed, had Dr. Falconer been acquainted with Dr. Wagner’s excellent paper “on the Specific Differences by which H. brunnea is distinguished from ZH. striata and crocuta, as manifested in the skull and dentition”. In this memoir, with which I was wholly unacquainted at the time when I communicated my paper to the Linnean Society, nearly all that I have remarked was anticipated more than twenty-three years ago. The principal cranial and dental characters by which the three existing species of Hyena are distinguished may be briefly stated as under? :— H. striata and H. brunnea, so far as regards cranial and dental characters, agree in so many particulars, as upon superficial inspection to be readily confounded. The chief points in which they agree are also those in which they both differ from H. crocuta and its fossil congeners. 1. In both species the upper tubercular molar is triradicular and tricuspid, and rarely less than 0’-5 in length by 0’"2 in its shortest diameter; while in H. crocuta and its allies this tooth is normally biradicular and bicuspid, though not unfrequently by abortion or fusion uniradicular or entirely absent, and it is never more than 0:2 or 0”-21 in length by about 0”1 in the shorter diameter. 2. In having the three lobes of the upper carnassial tooth (pm. 4) subequal in the antero-posterior direction. 3. In having a more or less distinct accessory point on the inner side of the hinder cusp of the lower carnassial(m. 1). It is true that a minute tubercle (or rudiment, rather, of a similar point) is not unfrequently seen in nearly the same situation in H. crocuta, and perhaps still more frequently in H. spelea*; but in those species it never assumes any thing like the size it presents in H. striata and H. brunnea, though it is considerably less in the latter species than in the former. Some difference also may be noticed in the exact situation of the accessory point in H. crocuta and spelea, in both which species it is usually situated as it were in a hollow beneath the base, at the inner and hinder border of the posterior cusp, whilst in H. striata and brunnea it rises distinctly on the inner surfaee of the cusp. Other points of agreement between these two species may be noticed, as, for instance, the presence in both of a distinct anterior talon to the second premolar, and of a 1 « Auseinandersetzung der specifischen Differenzen durch welche sich die H. brunnea von der H. striata und crocuta in der Beschaffenheit des Schiidels und Gebisses unterscheidet, yom Prof. Dr. A. Wagner,” Miinch, Abhandl. iii. p. 609, 1843. * Linn. Soc. Journal, ix. p. 65. > In twelve lower carnassials of H. spelwa, from Kent’s Cavern and Kirkdale, in the national collection, a small accessory point was noticed in five, whilst in seven there was merely a trace of one. VoL. X.—PakT Il. No. 4.—August 1st, 1877. M 78 MR. G. BUSK ON THE ANCIENT OR well-defined anterior talon to the first, second, and third premolar, which is larger, however, as, in fact, are all the talons, in H. striata. In H. striata and H. brumnea the second and third premolars are placed with their long axis oblique to the line of the alveolar border, and the third premolar is obliquely truncated behind, whilst in H. crocuta this tooth is square behind. The opening of the nares is rounded in H. crocuta, and more or less pyriform in H. striata and brunnea, in which also the anterior palatine foramina are very much larger in proportion. Dr. Wagner further adduces the form and size of the auricular bulla as distinctive marks. In H. crocuta this part is much more developed than in H. striata (and to a less extent in H. brunnea); and its posterior wall is convex or flat, whilst in H. striata and brunnec it is more or less concave. The extremity also of the mastoid process is compressed in both those species and rounded in H. crocuta. According to Dr. Wagner the occipital crest curves more backwards in H. striata and brunnea than it does in H. crocuta, in which it does not project beyond the level of the condyles. In my paper above referred to, and also in that of Dr. Wagner, several other points in the conformation of the skull and face in which differences are observable are adverted to; but as these for the most part refer to parts that are deficient in the Gibraltar speci- men, there is no occasion here to repeat them. It will perhaps be simply necessary to indicate that in all the main points above noticed the Gibraltar cranium and teeth exhibit the characters of H. crocuta as distinguished from H. striata and brunnea. 1. The form of the upper carnassial, in which the posterior cusp forms about half the length of the tooth !. 2. The minuteness, if not absence, of a tubercular molar. 3. The squareness of the hinder border of pm. 3. 4, The expansion of the auricular bulla, and the convexity of its hinder wall, and the roundness of the mastoid process. 5. The uprightness of the occipital plane and the wavy outline of the lateral ridges by which it is bounded. 6. And, as distinguishing it more particularly from H. brunnea, the rotundity and fulness of the parietal region of the skull. These considerations, together with the comparative dimensions given in the Table, and those of the teeth, as shown in the odontograms (Pl. XXVII. Nos. 1-5) of the maxillary molar dentition of the Gibraltar Hyena contrasted with those of H. spelea, H. crocuta, H. brunnea, and H. striata, will sufficiently demonstrate the identity of the Gibraltar Hyena with H. crocuta and H. spelea, which may be regarded, perhaps, as specifically the same. * 'I'his will be better seen in the figure, Plate I., than in the specimen itself in its present state, the tooth haying suffered injury since the drawing was made, by which the greater part of the enamel has been detached. QUATERNARY FAUNA OF GIBRALTAR. 79 This identification of the Gibraltar Hywna with H. crocuta is perhaps one of the most interesting results of the exploration, as affording a strong confirmation of the pro- bability that the Cave-Hyena found its way into Europe from Africa, at least in part, by way of Spain, through which country its track may, in fact, almost be followed, since a jaw, pronounced by Dr. Falconer to be that of H. spelwa, is described by Don Casiano de Prado as having been discovered in the Cavern of Pedraza, near Segovia, associated with very ancient objects’. It has also occurred in Sicily’, where the specimen is described by Dr. Falconer as “ certainly not of the Indian Striped Hyena, but of the Hyena crocuta, or Spotted Hyena of the Cape,” and at Mentone, where it seems to have been coeval with man. It is also of interest here to remark that, so far as is at present known, no fossil remains referrible to H. striata have been discovered in Spain. ‘he most southern known locality for that species appears to be the cavern of Lunel-Viel, in which the H. intermedia of Marcel de Serres undoubtedly represents H. striata. Nor does it seem to have occurred in the fossil state in Italy; so that in the present state of our knowledge it would appear to be not at all improbable that at the time when Europe and Africa were continuous by land the fauna of the latter continent did not include the Striped Hyena, whose centre of distribution, we may conclude, was probably in Asia. The only other remains of Hyena are numerous coprolites, to one of which is closely adherent a fragment of the atlas of an Ibex. ‘hese objects, of course, show that the animal must have lived at no great distance from the spot at which the remains were entombed. V. FELIS. The fossil remains from the Genista Cave establish by very distinct evidence the existence in the ancient fauna of several species of Felis, varying in size from the Leopard to nearly that of the Wild Cat. Of these the largest was a form undistinguishable from the existing F. PpaRDUS, or Panther, of the opposite African coast, The specimens referrible to this animal, all of which, it is highly probable, belonged to one and the same individual, are :— 1. A portion of the left maxilla, containing the alveolus of the outer incisor, the canine broken off recently to the level of the alveolus, the empty alveolus occupied by stalagmite of the first premolar (pm. 2) and the entire second premolar (pm. 3). (Pl. III. fig. 2.) 2. The nearly entire right mandible, retaining the canine and three molars (pm. 3, pm. 4, m. 1), all perfect, and the socket of the outer incisor. (PI. III. fig. 1.) 3. Half the left mandible pairing with the above, although they were found at some 1 Descripcion fisica y geologica de la Provincia de Madrid, 1864, p. 216. * Falconer, Pal. Mem, ii. p. 465. M2 80 MR. G. BUSK ON THE ANCIENT OR distance apart, retaining the outer incisor, canine, and two anterior molars (pm. 3, pm. 4), oe 4, The proximal extremities of the right and left ulnas, each with the olecranon broken, and probably gnawed off; all the fractured surfaces are covered with stalagmite. 5. An entire right tibia. 6. A portion of the shaft of the corresponding left tibia. 1. The mandible measures from the incisive border to the angular process 6”. Its height at the middle of its length about 1’"3; the length of the diasteme is ‘7, of the symphysis 2”:0, and of the molar series 2’-1. The incisor series or the space between the canines inside measures about 0:8. The canine is °6 <°5, pm. 3 6x3, pm. 4 78x -40, and m. I (carnassial) ‘8040. The jaw is very thick and massive, its width immediately behind the canines being 1’"6 nearly. In the mandible of an African F. pardus (No. 4540, C. S.), the length from the incisor border to the angle is 63, the height at the middle 1’"3, the length of the diasteme ‘8, of symphysis 2-0, of the molar series 2”°1, of incisor series ‘8, the canine ‘65 x ‘50, pm. 3 ‘D0 x 25, pm. 4 ‘80°35 m. 1°75 x ‘37. 2. The tibia measures 9”°9 in length, the proximal end 2”:1x2’:25, the distal 1”-01”:6; the least circumference is 2”°9, and the perimetral index -291. The tibia of an African Leopard in the Royal College of Surgeons is 9’°8 long, the proximal end 2”°4%2"°8, the distal 1-01-55, the least circumference 2”:8, and the perimetral index ‘285. The same close resemblance is exhibited in the dentition, as will be seen by inspection of the odontograms Nos. 21 and 22 (Pl. XXVII.), of which the former represents that of the Gibraltar Leopard, and the latter that of F. pardus (No. 4544, R.C.5.). Under the appellation of / antigua, Cuvier! notices the occurrence of a very similar form in the ossiferous breccia at Nice, associated with the Lion and several ceryine ruminants. It was also met with in the cavern of Gailenreuth; and it is also recorded by M. Marcel de Serres as occurring in the cavern of Lunel-Viel. M. Lartet? met with it in the Cavern of Mars, in the maritime Alps, and M. Gervais® in the Cave of Mialet. Lastly, Messrs. Boyd Dawkins and Sanford* enumerate it amongst the species met with at Banwell, and at Bleadon Hill and Hutton Cave in the Mendip Hills, The latter writers remark (p. 179) “ that the remains from the pliocene beds of Mont Perrier, in Auvergne, ascribed by MM. Croizet and Jobert to Felis antiqua, are too large to have belonged to the largest Leopard, though M. de Blainville believes that F. pardinensis and F. arvernensis are identical with the Panther.” They also justly remark that the /. pardoides of Prof. Owen® differs from the Panther in the lowness of the crown of the last [lower] molar,” to which might be added that the Crag tooth also 1 Op. cit. t. vi. p. 333. 2 Ann. d. Sc. Nat. 5™¢ sér. viii. p. 170. * Anim. Vertéb. vivants et fossiles, 1867-1869, p. 68, pl. xv. ‘ Brit. Pleistocene Mamm. part iv. p. 177, 1872. * Brit. Foss, Mamm. p. 169, QUATERNARY FAUNA OF GIBRALTAR. 81 differs from that of the existing Leopard, if we may judge from the figure, in the excavation of the anterior part of the anterior cusp on the inner face, on which aspect, in the Gibraltar specimen and the existing Panther, the surface is uniformly convex. In conclnsion, it may be remarked that the size of the lower true molar in the Bleadon-Hill Leopard, ‘8 x4, exactly corresponds with that of the same tooth in the Gibraltar specimen; and this is a character in which these two specimens appear to exceed any recent Leopard with which I have had an opportunity of making comparison. In the figure of the lower jaw of Felis pardus, however, given by M. de Blainville 1, the antero-posterior length of the tooth appears to be the same as in the two fossil instances. I would observe, however, that, although in the Gibraltar Leopard the lower car- nassial is so unusually long, the penultimate is of exactly the same length, and the next but very slightly longer. 2. FELIS PARDINA, Oken. The true Lynxes (excluding the Caracal) constitute a peculiar well-defined subgeneric group of Cats, characterized, so far as external features are concerned, by long legs, short tails, and usually tufted ears, to which, as more intrinsic characters, may be added the almost invariable absence at all ages of the foremost small upper premolar (pm. 2), which is generally present in almost all other felines—and, according to Keyserling and Blasius 2, by the circumstance that the nasals are separated from the maxillaries by the intervention of the descending process of the frontal meeting the premaxillary) Blasius adds, as another character of the Lynx Cats, that the lower carnassial (m. 1). is tricuspid. But in this he is manifestly in error, since that condition obtains only in one of the four or five species constituting the group; it is in fact confined, so far as my observation extends, exclusively to the northern Lynx of Europe and Asia. Of the group thus characterized, several European forms have been described under different specific names ; but at present I believe zoologists are tolerably unanimous in considering that there are in the Old World only two specifically distinct forms. The larger, best-known, and more widely distributed of these is Melis lynx, Linneus and Pallas, under which are included :— F. cervaria, Temminck, Nilsson, Cuv. F. lupulina, Thunb. F. lyncula, virgata, Nilss. F. lynx, Schreber, Temminck, Bechst., Keyserling and Blasius, Blasius, Schinz, Blainville, &c. 1 Qstéographie, pl. xxxvi. (Felis, pl. viii.). Vid. also jaw of the fossil Leopard from Lunel-Viel, pl. xhy. (Felis, xvi.). Dr. Gray made the same observation, and applies it to “ all the species of Lynx both from the Eastern and Western Hemispheres,” apparently unaware that he had been anticipated by Keyserling and Blasius. -Proc. Zool. Society, 1867, p. 259. 82 MR. G. BUSK ON THE ANCIENT OR F. borealis (pars), Temminck, Thunberg. Lynceus borealis, Gray. L. lupulinus, Gray. These various forms appear to differ from each other merely in certain details of coloration and size; and I have been unable to find any mention of more important dental or osteological characters, which, in a palontological point of view, at any rate, are alone available. The range of this Lynx appears to be very extensive, reaching from the northern shores of Siberia, throughout the whole of Europe, to the southernmost part of Italy, and from the Caucasus to the extreme west, where its remains show that it was also a member of the quaternary fauna’. But, with this very extensive distribution, no evidence at present exists of the occur- rence of F. lynx south of the Pyrenees, although it is said to have been killed in that range of mountains in 1833. In the south of Spain, at any rate, andin Portugal north of Lisbon, it is replaced by a distinct and somewhat smaller form, which in some parts even appears to be tolerably abundant. There is furthermore evidence of its having inhabited the peninsula at a very remote period, its remains, to judge from the figure of a lower jaw, having been discovered by M. Delgado? in the “ Casa de Moura.” It is to this second European Lynx that the remains of a species of Felis considerably less than the Leopard found in the Genista Cave appear to belong. They consist of a considerable part of the maxillaries of one individual and a portion of the right maxillary of a second, together with a large part of the lower jaw, the distal extremity of the left humerus, the proximal end of a corresponding ulna, and the distal end of a tibia. One of the maxillary specimens is represented in Plate HI. fig. 3, a, 6, ¢, d,e. It consists of the nearly entire left maxillary with all the teeth and the lower or malar portion of the orbital border. The two molar teeth (pm. 3 and pm. 4) are quite perfect ; but the canine is broken off close to the alveolus. On the right side only a small portion of the maxillary remains, containing the perfect canine; of the incisors four remain entire 3, and the two central alveoli are filled with stalagmite. There is no vestige whatever in either specimen of the anterior premolar (pm. 2). The canine (fig. 3, ¢) shows two deeply defined grooves on the outer and hinder 1 A jaw undistinguishable from that of F. lyna was discoyered by Dr. Ransome in a fissure of the Magnesian Limestone at Pleasley Vale in Derbyshire, associated with bones of the Wolf, Fox, Roedeer, Vole, &c. (Brit. Assoc. Report of Sections, 1866, p.16). It has been described and figured by Messrs. Boyd Dawkins and Sanford (Brit. Pleistocene Mammals, part iii. p. 172, 1868), who remark that the geological age of this relic cannot be determined with absolute certainty, though they think it may probably belong to the Post- glacial period. 2 Commissiio geologica de Portugal. Estudos Geologicos—Primeiro Opusculo—Noticia acerca das Grutas da Césaréda, p. 92, pl. ii. figs. 4,5, 6, 8, 9, 1867. > Two have unfortunately been recently broken off. QUATERNARY FAUNA OF GIBRALTAR. 83 aspect ; and there is none on the inner face, which shows a very indistinct trace ‘of a short ridge. The tooth is almost cylindrical. The length of the diasteme is not more than 0-2. The characters presented by the other maxillary specimen, which is of an older animal, so far as they go, exactly agree with those of the former. The mandibular specimen consists of the greater part of the left ramus, and about half of the right (Pl. III. fig. 4). A portion of the left ramus, corresponding to the second tooth, is wanting; but the lower carnassial on that side remains entire, and the fangs of the anterior premolar (pm. 4), recently broken off, are left in the alveoli. On the right side the canine is quite entire, as are the two anterior premolars (pm. 3 Sate ay If the dimensions of this jaw are compared with those afforded by the specimens above referred to, and with those given by Messrs. Boyd-Dawkins and Sanford, the dif- ference between the Gibraltar specimen and F. pardina and F. lynx will be seen to be but slight. The chief points to be remarked are the greater width of the diasteme in the Gibraltar specimen, and the less length of the molar series in both it and F. par- dina, as compared with F. lynx. This is commensurate with the greater antero-posterior diameter of the carnassial in the latter species, as shown in the odontograms. In both these particulars, viz. the width of the diasteme and the comparative shortness of the molar series, the Gibraltar jaw agrees with that figured by M. Delgado. As the condition of the Gibraltar specimens affords no means of judging of the cranial characters, it would be needless on this occasion to discuss them. It may be stated, however, that they are amply sufficient to prevent any confusion between the Northern and Southern Lynx, and with the Caracal, even in the fossil state. The same may be said with respect to the dental characters, which are also alone sufficient for a satisfactory diagnosis. As, fortunately, the Gibraltar remains afford the complete dentition, I will proceed to point out the more salient characteristic differ- ences, 1. between the true Lynx and the Caracal, and 2. between the Northern and the Southern European Lynx. It does not appear to be necessary to compare any other feline species. The two American Lynxes are so much smaller as to be out of the question altogether. 1. As to the dental differences between the Lynxes (proper) and the Caracal. (a) The Caracal almost invariably possesses an upper anterior molar, though usually of small size, and sometimes apparently deciduous at an early period. (0) Its canines are in almost all cases smooth or ungrooved; its premolars are much more compressed ; and the inner anterior tubercle of pm. 4 is smaller.’ 1 In a skull of F. caracal (No. 981 a, B. M.) the canines have, to use Dr. Falconer’s expression, “ the enamel 5 smooth and ungrooved.” And the same is markedly the case in a skull from Tangiers, 4—23 B.M. The il canines are also smooth in two Caracal’s skulls in the Royal College of Surgeons, one of which, No.4587, from 84 MR. G. BUSK ON THE ANCIENT OR 2, As regards the dental distinctions between the Northern and Southern Lynx, we have to remark (a) that in F. lynx the teeth generally are larger, though of the same proportionate width, and (/) that the upper tubercular molar is very much larger, and triradiate (the middle root very small). 3. The lower carnassial is furnished with an additional small cusp behind ', which is wanting, not only in /. pardina, but, so far as I am aware, in almost all other felines. The only cases in which I have noticed any thing of the kind is in F. picta (Leopardus pictus, Gray) (No. 14954 B. M.), from South Africa, and in F. viverrinus, Gray. 4, The pm. 3 is broader or squarer behind in F. /ynx, whilst it tapers almost to a point in F. pardina, and the teeth generally are wider posteriorly. 3. Fruis catieata, Temminck, A third species of Felis, of smaller size than the Lynx, is represented by :— 1. An entire left mandible, with all the teeth except the incisors. A fragment of a second left mandible retaining the carnassial tooth only. An entire right humerus. The distal end of the left humerus. The proximal end of apparently the corresponding ulna. The distal half of the right tibia. Two metatarsal bones. a Se ewe es Cod) 1. The entire mandible (PI. III. fig. 6) measures 2-8 from the incisor border to the condyle, and about the same to the extremity of the crochet. ‘The coronoid process rises to the height of 1-2; and the height of the ramus at the second tooth is 05. The condyle is 0-6 in length, and is rather slenderer than it appears in the figure, its greatest thickness being about 0-15. The length of the three molars is 095, and of the diasteme about 0°’26. The dimensions and proportional sizes of the teeth are shown in the odontogram No. 16. This beautiful specimen was extracted from a very hard ferruginous matrix. ‘The bone is very dense, and almost black from manganesic infiltration, so that there can be no doubt that the animal belonged to the most ancient fauna of the rock. The teeth, with the exception of the carnassial (m. 1), which has lost the hinder cusp (by recent fracture), are nearly entire, the small anterior and posterior cusps in pm. 3 and pm. 4 haying been either worn or broken off. The canine, which is greatly worn behind, has a simple deep groove on the outer side ; the Hunterian Collection, is named /’, lyna in the Catalogue, though there can be no doubt of its belonging to “the Persian Lynx,” /. caracal, and not to the true Lynx. But in a skull of a large Caracal from the Zoolo- gical Society’s Collection, No. 981 ¢, B. M., a single fissure is obscurely evident on the outer side of the canines. ' The absence of this cusp in the Canadian Lynx is referred to by Messrs. Dawkins and Sanford, QUATERNARY FAUNA OF GIBRALTAR. 85 and perhaps a second may have existed until worn away by the usure of the tooth. There is no distinct appearance of a groove on the inner aspect. The coronoid process is as much reclined as in the Common Cat, and projects back- wards to a vertical line, which would touch the condyle and angular crochet. The masseteric fossa is very deep, and its anterior margin much raised. 2. The fragment of the second mandible, so far as it goes, precisely corresponds with the former, except that it is scarcely stained with manganese, although it was imbedded in the same kind of matrix. 3 3. The entire humerus is 4’"3 long, the proximal end ‘95 x ‘70, the distal -50 x -85, the least circumference 1-0, perimetral index ‘232. This bone has a very recent aspect, and is obviously of much more recent age than the rest. 4. The fragment of a humerus has an ancient look, but corresponds in all respects with the above. 5. The portion of ulna, from the condition of the bone, would seem to belong to the opposite side of the same individual as No. 4; whilst 6. The tibia and metatarsals have the more recent aspect of the entire humerus. There can be no doubt, however, that all these bones belong to the same species. Of the numerous existing species of Cats smaller than the Lynx, and varying in size from the Serval to F. maniculata, and in which the upper carnassial tooth never exceeds 0:5 or 0°55, and the lower 0-4 or 0:45 in length, the only ones with which it appears worth while to compare the Gibraltar form are :— 1. FE.is sERVAL, Schreber, Gray. Felis capensis, Forst. galeopardus, Desmarest. senegalensis, Lesson. Chaus servalina, Gerrard, Blainv. serval, Buffon. 2. Feuis caus, Giildenst , Temminck, Blainv. Felis catolynz, Pallas. — libyca, Olivier. —— affinis, Gray. dongolensis, Hempr. and Ehr. jacquemontii, I. Geoff. St.-Hil. — riippelli, Brandt (non Schinz). servalis, Blainv. Chaus libycus, Gray. jacquemontii, Gerrard. VoL. X.—ParT 11. No. 5.—August 1st, 1877. N 86 MR. G. BUSK ON THE ANCIENT OR 3. FELis cALIGATA, Temminck. Felis caligata sive bubastes, Blamv. bubastes, Ebr. & Hemp. (Symb. Phys.). cafra sive caffra, Desmarest, Gray. —— “ Caracal de Libye,”’ Buffon. “ Booted Lynx,’ Bruce. Chaus caffer, Gray, B. M. Cat. 4, Frevis MANtcuLata, Riippell. Felis riippelli, Schinz. To which might perhaps be added :— 5. Feuis catus (fera), Linn. 6. FEuis catus Macna, Schmerling. As my comparison has been necessarily limited to the lower jaw and teeth, I will confine what I have to remark to these parts alone, which appear amply sufficient for the purpose of diagnosis. Of the species above enumerated, that whose lower jaw most closely resembles the Gibraltar specimen is F. caligata s. bubastes, as represented in M. de Blainville’s figure , ' which has been taken from a mummified specimen. Comparison of this figure with that of the Gibraltar specimen (PI. III. fig. 6) will at once demonstrate their almost exact resemblance. It will be useful also to compare M. de Blainville’s figures of the lower jaw of F. servalis, F. chaus (mummy) (which seem to me to belong to the same species), F. maniculata fera, and F. catus fera, together with Dr. Schmerling’s figure of the jaw of F\ catus magna, which, although the teeth are, or appear in the figure to be, rather smaller, seems to me to be identical with the Serval. The several Odontograms of these Cats (Pl. XX VII.) will further serve to show the differences and resemblances amongst them as regards the dentition; whilst those of the recent and mummified F. caligata, compared with that of the Gibraltar specimen, will further demonstrate so far the identity of these three forms. Besides this comparison with published figures, I have carefully contrasted the Gibraltar jaw with that of the so-termed Chaus caffer, Gray (857 A, B. M.) from the Cape of Good Hope. The jaw and teeth are exactly of the same size. In fact there is no perceptible difference between them; and it is clear that whatever may be the species to which the Museum specimen (procured from M. Verreaux) belongs, the Gibraltar one is the same. ‘The only difference worth noticing is the circumstance that the coronoid process is narrower at the bottom in the latter; but both have the same slope of the coronoid * Ostéographie, Pl, xlyii. (Fulis, xix.). QUATERNARY FAUNA OF GIBRALTAR. 87 process in front, or very nearly so. In the pattern of the teeth the median cusp of pm. 2 is thicker in the Gibraltar specimen, though the tooth is of exactly the same length at the base. In neither is there any distinct trace of a hinder cusp or even of the anterior, though there has probably been a small one in both. They both differ altogether from Chaus servalinus (133.a, B. M.) from Senegal, in which the teeth are not only much larger, but the pm. 3 and pm. 4 have a large anterior and two small hinder cusps, as in most of the Felide. In Chaus libycus (11726, B. M.), whose mandible is of the same size, the coronoid process is less reclined, and the teeth, except pm. 3, longer, with a much more strongly marked cingulum behind. Another strong point of resemblance between the Cape and the Gibraltar specimens consists in the configuration of the masseteric fossa, which is very deep, and has sn abrupt narrow elevated ridge bounding it below; whilst in Chaus libycus the fossa is much shallower, and the ridge less elevated. In the Cape and Gibraltar jaws the “crochet” is much incurved, but scarcely at all so in Chaus libycus. The distance also from. the lower border of the “crochet” to the under surface of the condyle is the same in the Cape and Gibraltar specimens, and considerably greater than in Ch. libycus. There can be no doubt, therefore, that the Gibraltar and Cape species are one and the same; nor can there, I think, be any doubt that the Indian form named Chaus libycus in the British-Museum Catalogue is distinct. The parietal width in C. libycus is 1-9; in C. caffer 1”-95, or nearly the same; so that the animals are probably of nearly equal bulk. But the orbit in C. caffer is 2"°3 in its vertical diameter, and that of C. libycus only 1”-9. The bony orbit is almost complete in C. caffer, and much less complete in C. libycus, in which also the infra- orbital foramen is smaller and more elliptical, and larger and rounder in C. caffer. In Chaus caffer, again, the nasals are equal with the maxillaries, whilst in C. libycus they exceed the maxillaries, as in the Lion. In C. libycus the lacrymo-maxillary suture is in front of the edge of the orbit, and in @. caffer coincident with or rather behind it. In the narrowness and reclination of the coronoid process the Gibraltar jaw re- sembles (among the species above named), besides F. bubastes, F. maniculata and Felis catus ferus, and, it may be added, the Domestic Cat also; but it is distinguished from the three latter, not only by its greater size, but also by the far greater thinness of the inferior boundary of the masseteric fossa; and from F. maniculata by the less abrupt or defined termination of the fossa anteriorly. In F. maniculata also the “ crochet” does not project so far backwards, though this may probably be an uncertain character. From these considerations there appears to be every reason for believing that the smaller fossil Cat of Gibraltar is F. caligata, a species which appears to have a very extensive range from one end of Africa to the other, and to have formed one of the three feline species which were regarded as sacred by the ancient Egyptians, and were N 2 88 MR. G. BUSK ON THE ANCIENT OR frequently converted into mummies, the other two being F. chaus and, in all proba- bility, F. maniculata, which was apparently the domestic cat of the period. 4, FELIS DOMESTICA. Numerous remains of the common domestic Cat were met with in the more recent deposits in the caves and fissures, and may be often picked up on the surface. But I have not yet met with any really fossilized Cat smaller than F. caligata. VI. Canis, Herpestes, MELEs. No other remains of Carnivora have been met with in the Gibraltar collection belonging to the more ancient fauna. The jaw of a Fox (PI. III. fig. 7), incrusted with red stalagmite from the upper part of the Genista Cave, is the only instance of any great apparent antiquity. Numerous specimens belonging to the same species, undistinguishable in the bones and teeth from the common Fox, have occurred in other fissures and caverns, mixed with recent bones and human rubbish, as magi be expected, seeing that the Fox is at present a living inhabitant of the Rock. In the same category of more recent specimens, it may be mentioned that in Genista Cave No 3 the skull, without the lower jaw, of a Mangoose (Herpestes ichnewmon, Tllig.) was found, of the present existence of which on the Rock I am not aware, but which occurs in the south of Spain. Under the same circumstances the skull, jaw, and bones of the Badger (Meles taxus) of different ages, have also occurred, although, what is perhaps rather remarkable, the remains of this most ancient, perhaps, of all existing quaternary mammals except the Lion, have not as yet been discovered in the true ossiferous breccia. Bones of the Dog are frequently found on the surface; but no trace of the Wolf has been noticed. VII. Equus. Of the genus Equus, the materials, although not numerous, are amply sufficient to establish the identity of the species with E. caballus. The principal specimens are :-— 1. A large fragment of the cranium (PI. VIII. fig. 1), comprising the greater part of the left side of the face, from the posterior border of the orbit to the incisive extremity of the diasteme. It shows the whole of the maxillary and lachrymal bones, together with part of the frontal; and about two thirds of the circuit of the orbit are left. The nasals are wanting. ‘The specimen is slightly cracked behind; and a crack runs across the frontal and through the orbit. It would appear that the specimen had originally been much more perfect, and that a considerable portion was lost in its extraction from the matrix. ‘The dental series QUATERNARY FAUNA OF GIBRALTAR. $9 consists of the empty pit of the small caducous first, and the three following milk- molars in full wear, together with the first true molar in germ imbedded in the jaw. In the dimensions and pattern of the teeth, and the other characters of this specimen, it will be seen to exhibit nothing to distinguish it from the corresponding parts of an ordinary colt or filly. The specimen, which bears no marks of gnawing, was procured from beneath two thick floors of stalagmite ; and it was imbedded in a hard calcareous ochreous matrix, from which it had to be chiselled out. Two other maxillary fragments, each containing two milk-molars, were found at a greater depth. And besides these, two other detached milk-molars, upper and lower, were also met with. It may be presumed that all these specimens, though found at different depths and widely apart, may in all probability have belonged to the same individual. The crowns of the teeth are slightly sun-cracked, indicating exposure to the atmosphere before their entombment in the cave. 2. Another specimen, also possibly belonging to the same animal, is a proximal phalanx (Pl. VII. fig. 3), from which the proximal epiphysis has been detached. It is rather slender in form. 3. A large portion of a scapula of a mature animal (Pl. VIII.). The glenoid cavity measures 3°25 x 25, which denotes a horse of considerable size. 4. Corresponding with the scapula is the entire proximal extremity of the right humerus (Pl. IX. fig. 1). 5. An entire left caleaneum (fig. 3). 6. The distal epiphysis of the left femur (Pl. IX. fig. 2). This portion of the femur appears to have been detached at the epiphysial junction before interment, as the entire surface was covered with a thick layer of ochreous crystalline deposit. 7. Besides the above, there are in the collection several portions of upper and lower jaws, and a few detached teeth. With the exception of one or two old and much-worn teeth of the permanent series and of large size, nearly all the specimens would appear to belong to young or immature animals. It is to be remarked that none of the equine remains exhibit any evidence of gnawing, or of human agency—although, from the circumstance that by far the greater part of the collection consists of the remains of young or.quite immature individuals, one might be almost inclined to suppose that they were the relics of animals that had been used for food. 1. Head. All the specimens described or referred to above are strongly mineralized, and, as has been already said, are in that respect in exactly the same condition as the bones of Rhinoceros, Cervus, Ibex, &c., which are indubitably fossil. There can be little doubt that the Horse was a member of the contemporary fauna. In Genista Cave No. 2 a couple of distal phalanges were met Son which appear to 90 MR. G. BUSK ON THE ANCIENT OR belong to the ass and its foal. But as their condition plainly shows that they belong to a very recent period, they do not come within the scope of this account!. Another equine specimen, also found in the east fissure, presents some very curious characters. It is the greater part of the horizontal ramus of the right mandible, con- taining five teeth in an uninterrupted series, but apparently in a very irregular mode of growth. The two anterior ones are first and second deciduous molars, with a level disk of wear, exhibiting the enamel-flexures clearly: but the teeth, so far as can be judged, are apparently as long as the true molars; for the second projects in a curious manner straight out of the alveolus to the height of rather more than two inches, without showing any indication of division into fangs, and it is very deeply sulcated on the dorsum or outer surface. Its enamel-flexures are shown in Pl. IX. fig. 4, plainly indi- cating that it, and the one in front of it of similar pattern, are truly deciduous teeth. The third tooth is in germ and just emerging from the alveolus; so that its pattern cannot be ascertained; the two hinder teeth, however, are distinctly permanent molars. VIII. RuiNoceros. The genus Rhinoceros is represented by a considerable number of specimens, which were procured at various depths in the Genista Cave and east fissure. Though some few are stated to have been met with in the dark-coloured cave-earth in which the human relics &c. occurred, yet as these specimens, as regards mineral condition (that is to say, infiltration with calcareous matter, and incrustation with the ferruginous crystalline stalagmitic deposit, &c.), differ in no respect from those which occurred at the greatest depths, and since there is every reason to believe that the bones found in the highest level formed part of the same skeleton as that to which some of the deeper ones belonged, it must be concluded that their presence in the black earth was in some way accidental. The principal specimens to which it is needful to call attention may be arranged as belonging to:—I. Head; II. Trunk; III. Anterior extremity; IV. Posterior extremity. I. Specimens belonging to the Head are :— 1. A right upper fourth premolar (pm. 4, d). Pl. X. figs. 1-3. 2. A right upper molar(m.1,d). Pl. X. figs. 4, 6. ‘ In the excavation of the east fissure the entire skeleton of a Horse was met with, at a few feet only below the surface. In general condition the bones presented very much the same character as many of the fossil bones from a greater depth, and had been deprived of the greater part of the animal matter. At first Captain Brome thought he had come upon the remains of a fossil Horse; but, to his surprise, when the foot-bones were exhumed, the shoes with which the animal had been shod were found in situ; and it was ascertained that the bones, much altered as they were, had belonged to a fayourite Arab charger, which had been buried at the spot about twenty-five years before. The instance is a very striking one, in showing the fallacious nature of eyi- dence derived merely from the mineral condition of buried bones when exposed to free percolation of water in a calcareous bed, QUATERNARY FAUNA OF GIBRALTAR. 91 3. A left upper molar (m. 1, s). Pl. X. figs. 6, 7. 4. A left upper second premolar (pm. 2,s). Pl. X. fig. 8. II. To the Trunk :— 5. A nearly entire atlas. Pl. XVIII. figs. 1, 2. III. To the Anterior Extremity :— 6. An entire naturally detached proximal epiphysis of the left humerus. Pl. XI. fig. 1. 7. The head and a considerable part of the shaft of the right humerus of a fully mature animal. 8. A fragment of the shaft of the opposite humerus, fully adult. 9, A nearly entire left radius. Pl. XIV. figs. 1, 2. 10. A perfect right os lunare. Pl. XV. figs. 4-8. 11. The distal extremity of the third metacarpal. Pl. XVII. figs. 6-8. 12-15. The distal extremities of four metacarpals differing somewhat in size. 16, A second phalanx. Pl. XVI. fig. 9. IV. To the Hinder Extremity :— 17. The nearly entire head and upper portion of the shaft of the right femur. Pls. XII., XIII. 18. A detached right third trochanter, probably of the same femur. 19. A nearly perfect right tibia of an immature animal, with the epiphyses ununited. Pl. XIV. figs. 3, 4. 20. A crushed fragment of the middle of the shaft of a tibia. 21, 22. Two right astragali, one of which is figured in Pl. XV. figs. 1-3. 23, 24. An entire third right metatarsal with the proximal end of the corresponding fourth metatarsal in natural apposition. Pl. XVI. figs. 1-4. 25, 26. The proximal half, or nearly so, of another third right metatarsal, and the corresponding entire fourth metatarsal apparently fitting to it; the latter is figured in Pl. VII. figs. 1-5. 1. Head. As most of the more important of these specimens are represented of the natural size in the Plates accompanying the paper, it will be unnecessary to enter at any length into detailed descriptions of them. For the diagnosis of the species, however, to which they belong, it will be requisite to notice more particularly those which may be regarded as affording the best characters. Amongst these are, in the first place, the teeth. 1. The most perfect is the right upper fourth premolar. (Pl. X. figs. 1-3.) The tooth, which is very little worn, measures 1’-8 in the antero-posterior direction, 92 MR. G. BUSK ON THE ANCIENT OR and 2’-2 in the transverse (1’°8 x 2'"2) taken at the base of the crown. The extreme height of the crown in its present state is 2"-5; and it probably never exceeded 2"-6 when quite unworn. On the dorsal or outer surface the second, fourth, and fifth coste are very prominent, the second, as usual, being the most so. The first or anterior costa is rounded off; and the third is faintly indicated, though distinct enough throughout the whole length of the crown ; at the base there is a rather deep sulcus between the second and fourth coste. The anterior vallum, “bourrelet,” or “ guard,” which is well developed, is nearly horizontal when the tooth is held in its natural position. It terminates at the inner and anterior angle, and does not extend at all on the inner face of the anterior colline. The posterior vallum (“ posterior collis,” B. Dawk.) is deeply notched in the middle and without a median cusp. It is prolonged into a strongly developed ridge, or “ bourrelet,” which crosses the inner face of the posterior colline obliquely from above downwards, nearly to the base of the crown, and terminates in the median sulcus in a very minute cusp, as shown in fig. 2. The collines are slender and tapering. The anterior valley is wide, and traversed by a strong crista (combing-plate), which arises from the angle of junction of the hinder colline with the dorsal lamina. The uncus (“crochet”) is represented by a very thin projecting plate, which appears, in the deeper part of the valley, to meet the crista so as to circumscribe a small pit. The posterior valley is much expanded, and, owing to the depth and width of the notch in the posterior vallum, very open. A small posterior crista or “ combing-plate” projects into this valley from the inner wall of the dorsal lamina. The surface of the enamel, which is about 0”:06 thick where it is unworn, is marked with very fine parallel sulci, and, in parts, with equally delicate transverse ruge. 2. The tooth represented in figs. 4 and 5 appears to be the first right upper molar. It is much more worn than the preceding ; and, in addition to this, the posterior vallum is entirely broken away. It measures about 2” in either direction; and what remains of the crown is about 2” in height. The second dorsal costa is very prominent, whilst the others are represented only by low rounded elevations. The uncus (“ crochet”’) comes off at a right angle with the hinder colline, and curves slightly outwards, reaching nearly across the anterior valley. In thickness and sculpture of the surface the enamel of this tooth corresponds with that of the premolar above described. 3. The tooth represented in figs. 6 and 7 appears to be the corresponding molar of the opposite side. It is more imperfect than the last; but from what remains of it its character would seem to be identically the same. In both these teeth the posterior vallum is too much broken to afford any evidence either for against the existence of a median cusp. 4. The small molar represented in fig. 8 has lost nearly the oe of the dorsal lamina. From the dimensions of the remaining portion, however, it may be concluded QUATERNARY FAUNA OF GIBRALTAR. 93 that it is a second premolar of the left side (pm. 2,8). ‘That it is not a milk-molar is shown by the circumstance that the opening of the median sulcus does not extend nearly to the bottom of the crown. The existence of an oblique descending ridge on the inner face of the hinder colline seems clearly to indicate its relation with the pm. 4d described above. falc Besides these more perfect teeth, there is (1) a broken fragment of a much-worn upper molar of small size, which, so far as can be judged, probably belonged to the deciduous series ; (2) amere chip from the inner face of a small much-worn premolar. Judging from the apparently different ages or states of wear of these teeth, it is highly probable that they must have belonged to at least two individuals, one of which was much older than the other. With respect to the specific characters afforded by the teeth, it may be stated that they seem to correspond in every particular with Dr. Falconer’s description of those of Rhinoceros hemitechus, as will be seen on reference to his account of that species in ‘ Palzontological Memoirs,’ vol. ii. Without going into needless detail, I may quote what he says on the distinctive characters of the premolars and molar teeth of this species }. i He remarks that the premolars of R. hemitechus may be characterized :— 1. By the absence of an internal basal “ bourrelet ;” 2. By there being only two fossettes in the worn crown ; 3. By the middle valley being traversed by the processes of a bifid crochet emitted from the posterior barrel, and by a parallel combing-plate given off by the outer or longitudinal ridge ; 4, By being invested like the true molars with a very thick coat of cement. He proceeds to remark that the absence of a basal “ bourrelet,” besides other cha- racters, distinguishes the premolars of R. hemitwchus from R. leptorhinus and R. megarhinus. Since R. leptorhinus of Cuvier is synonymous with 2. megarhinus of Christol, and R. leptorhinus of Owen with R. hemitechus, Falc., it is not quite clear what his meaning is in the expression just quoted. But, as partly explanatory of it, I may cite a note of his, made in the British Museum in June 1864, with reference to a tooth numbered 36770, which runs thus :—“ A true right from Peckham, exactly in the same stage of wear of crochet and outer ridge as the Gibraltar molar; and the ter- mination of the transverse valley, as in it, is a triangular fissure without complication. It hasno basal ‘ bourrelet.’ It is probably m. 1, like the Gibraltar tooth; and the two are of nearly the same size. It has no combing-plate.” Dr. Falconer, in a sidé-note, says that this tooth “ ought to be figured with the Gibraltar bone,” thus marking emphatically, what I know was his opinion, that the two teeth were specifically identical; and his recommendation as to the giving of a figure of the Peckham tooth should have been obeyed, had the necessity for it not have been obviated by the 1 Op. cit. p. 328, pls. xvi., xvii. ke. VOL. X.—PART 11. No. 6.—August Ist, 1877. ) 94 MR. G. BUSK ON THE ANCIENT OR publication of an excellent figure of the tooth in question by Professor Boyd Dawkins, in his paper on R. leptorhinus (R. hemitechus) '. The above citation may suffice to show that in Dr. Falconer’s opinion the Gibraltar tooth belonged to the same species as that figured by Prof. Boyd Dawkins, which is undoubtedly R. hemitachus. Again, with respect to the molars, Dr. Falconer observes (op. cit. p. 329) that the character which best distinguishes them from those of all other species lies in the peculiar form of the “crochet” or promontory, projected forwards from the posterior colline into the transverse valley. ‘‘ In all species, fossil or recent,” he says, ‘“ except R. hemitechus, the ‘ crochet’ forms a plate, which is emitted at a very open angle with the posterior colline, and directed more or less diagonally towards the anterior outer corner of the crown.” Again (p. 331), “ if the penultimate true molar in R. hemitechus be examined, the crochet presents a thick massive body thrown straight forward, and forming an acute angle with the anterior margin of the posterior barrel. It is flat or concave above [externally] and convex below |internally], narrow at the base, and thickening to a blunt margin. In mass it bears a much larger proportion to the dise of the hind barrel than in most of the other species. In the corresponding molars of R. megarhinus, Christ. (pl. iii. fig. 5, of Christol’s Mem., and pl. ii. fig. 5, Gervais’s Palé- ontol. Frang.), besides the difference of alignment in its offset from the hind barrel the section of the crochet is wedge-shaped, thinning from a broad base to a sharp edge ” 2. But it is needless to say more with respect to the hemitcechine characters of the Gibraltar teeth, so far as Dr. Falconer’s opinion respecting them is concerned. Besides the reference he himself gives to the specimen (B. M. No. 36770) figured by Prof. Boyd Dawkins, the exact correspondence between the figure of m. 2 of R. hemitechus given in Paleontographical Memoirs, vol. ii. pl. xvi. fig. 1, with fig. 4, Pl. X. of this communication, cannot fail to satisfy us of the identity of the two forms, and of their distinction from that presented by R. megarhinus, Christ. I have not thought it necessary to say any thing with respect to the points by which the Gibraltar teeth are distinguished from the corresponding ones in &. tichorhinus, the differences in all respects being too marked to require comment. Nor, having shown their apparent identity with those of R. hemitechus, is it requisite to say much respecting their distinction from the molars of J. efruscus, which species would other- wise naturally have suggested itself as a very likely subject of comparison, nor respecting their relation to the teeth of R. bicornis, which, again, might have suggested itself as not unlikely to be found in company with H. crocuta. As regards 2. etruscus, the figures and descriptions of the Gibraltar teeth already * Journal of the Geological Society, vol. xxiii. pl. x. fig. 5. * T am inclined, from my own observations, to think that Dr. Falconer placed, perhaps, too much importance upon the characters afforded by the “ crochet,” which appear to be very variable. QUATERNARY FAUNA OF GIBRALTAR. 95 given will be quite sufficient to show that they are toto celo distinct; and the same may be said with respect to 2. bicornis, and, I may add, to any other of the existing species with which I am acquainted. 2. Trunk. The only bone belonging to this division of the skeleton, beyond a small fragment of a rib, in all probability rhinocerine, is the at/as, represented in Pl. XVIII. figs. 1 & 2. The remains of this vertebra consist of nearly the entire left and a considerable portion of the right half of the bone. When placed in their proper positions the two fragments have evidently formed parts of the same vertebra, which has been broken in the process of exhumation, or extraction from the hard stalagmite deposit in which it was imbedded. The bone is that of a mature animal. In its present condition it measures about 11-0 in extreme breadth ; and if it had any unossified epiphyses, it might, when entire, have measured perhaps 12'-0. The antero-posterior diameter of the transverse processes is about 4”, or a little more ; and the extreme antero-posterior length of the bone, measured between the summits of the anterior and posterior articular processes, is about the same. The dorsal portion of the rfng is absent; so that the configuration of that part, which would seem to differ a good deal in different species (at any rate it does so in R. wnicornis and R. bicornis), cannot be ascertained. The posterior articular surfaces look obliquely inwards, their planes meeting at an angle of about 100° or 110°. The anterior articular cup, when the fragments were fixed in plaster of Paris in their proper position, fitted exactly upon the occipital condyles of a skull of 2. hemitechus in the British Museum. ‘The only existing species with whose atlas I have had an opportunity of comparing the Gibraltar specimen are R. wnicornis and R. bicornis, from both of which it differs so widely in many respects, that it appears to me needless to enter into any particular comparison. I have not been able to compare the atlas with that of any fossil species; but Cuvier! notices and figures a fossil atlas of Rhinoceros, which was found in 1750 near Schartfels (Schwarzfels?), and first described by Hollman?, which presents many characters in common with it, amongst which are:—(1) the comparatively small size, Hollman’s specimen not being more than 13”-7 broad and about 5'-0 in the antero-posterior width of the transverse processes; (2) the incompleteness of the anterior arterial foramina, which in the Gibraltar bone are represented by wide notches; (3) the obliquity of the posterior articular surfaces, whose planes in Hollman’s specimen, according to Cuvier, met at an angle of about 90°. With respect to the last two particulars, Cuvier remarks that in a recent atlas (probably 2. wnicornis ?) with which he instituted a comparison the arterial foramina were complete, and the posterior articular surfaces formed a right angle with the longi- tudinal plane of the bone. 1 Ossemens Fossiles, ed. 4, t. iii. p. 143, pl. 46. figs. 6-8. + Comment. Soc. Gotting. 1751, p. 251. 96 MR. G. BUSK ON THE ANCIENT OR In the absence of other direct means of comparison, it is difficult to arrive at any positive determination; but, from the circumstance of the Gibraltar atlas fitting so exactly as it does upon the skull of R. hemitechus, Falc., it is allowable to assume that it must have belonged to a species closely allied to, if not identical with that form. ze 3. Anterior Extremity. One of the best-marked specimens belonging to the anterior extremity is a perfect, naturally detached proximal epiphysis of the left humerus (Pl. XI. fig. 1). It includes the entire articular surface, the great tuberosity, and the occipital groove and ridges complete. The bone is dense and heavy, and is uniformly incrusted on all its surfaces with a thin layer of crystalline ferruginous stalagmite. The articular surface measures 4”-2 in the transverse, and about 5:0 in the antero-posterior direction. The radius of the curve taken tranysersely is about 2'-9, and in the antero-posterior direction about 2'-0. The occipital groove is three inches wide and about 1-1 deep at the outer side. I have not as yet met with the corresponding part of the humerus of any fossil form except that of 2. megarhinus from Grays (No. 23111, British Museum), from which the Gibraltar specimen differs so widely, more especially in the form of the bicipital groove, that there can be no doubt of their specific distinction. I have also compared it with the same part in &. dicornis (KR. keitloa), with the result that, as regards the form of the bicipital groove, it approaches that species more nearly than any other with which a comparison has been instituted. In order to illustrate the different and, as it seems to me, important characters that are afforded by the conformation of this groove in various species of Rhinoceros, I have subjoined ideal sections across it in 1. The Gibraltar Rhinoceros, fig. 2, 2. R. megarhinus, Grays, fig. 3, 3. R. bicornis (R. keitloa), fig. 4, in all of which the letter (a) is placed on the outer side. Rhinoceros of Gibraltar. The specimen No. 7 consists of about the upper half of the left humerus of a fully mature, probably aged animal. It is unfortunately much mutilated, the bici- QUATERNARY FAUNA OF GIBRALTAR. 97 pital groove and ridges and the great tuberosity being lost. The part of the bone at which they were attached is crushed into numerous fragments, probably by the fall Fig. 3. Fig. 4. Zi ij N c wa | upon the bone of a heavy fragment of rock; the fragments, however, although much displaced, are solidly united by red calcareous stalagmite. The shaft is also broken across transversely about five inches below the summit of the articular head; and the two portions of the bone were found at a considerable distance apart. At the lower end the shaft is fractured very irregularly ; but there is no appearance anywhere of the bone having been gnawed by the Hyena. There is a small incised mark close to the lower end; but this appears to be quite recent. The fragment, as it is, is about 11 inches long; and the least circumference of the shaft, at the point where it is usually smallest, is about 8’"7. The proximal articular surface is very nearly entire. It measures about 4"-2 in the transverse, and about the same in the antero-posterior direction !. The radius of its curve is 2'"0. The injured condition of the bone precludes the possibility of ascertaining very precisely any of its characters. It may be remarked, however, in comparison with a humerus of R. bicornis (R. ‘In R. bicornis the antero-posterior measure is 4"-6, and the transverse 3-9 or 4-0. 98 MR. G. BUSK ON THE ANCIENT OR keitloa) :—(1) that the two bones would appear to be of as nearly as possible the same size; (2) that on the anterior aspect of the shaft there is no appearance of the rounded bulging observable in the middle part of that aspect in the humerus of 2. bicornis ; (3) on the posterior aspect in 2. bicornis there is a strongly marked semicircular ridge, or linea aspera, running upwards and inwards from the deltoid crest to the posterior border of the articular surface of the head, no trace of any thing like which is visible in the Gibraltar specimen; (4) the rough muscular impression on the inner side of the shaft is situated an inch lower down, and is larger in size in the Gibraltar bone. With respect to R. megarhinus, comparison with a magnificent perfect specimen of the humerus of that species from Ilford (No. 23111, Brit. Mus.) shows that it was of larger size, or, at any rate, thicker, having a least circumference of 9-9. But the arti- cular surface of the head is of pretty nearly the same dimensions, viz. 5'°3 x 4':2 in the antero-posterior and transverse directions. These dimensions, it should be remarked, are identical, or nearly so, with those of the detached epiphysis, and but very little different in the antero-posterior extent from those in the present specimen. I have not as yet met with a humerus of the smaller Thames-valley species retaining the upper extremity. There is none, therefore, with which the present specimen could be compared, except as regards the least circumference of the shaft, which in R. hemitechus is about 8-8, or an inch less than in the larger species. Another specimen is a small portion from about the middle of the shaft of a right humerus, and corresponding in all respects as to contour, condition, &c. with the same part in specimen No. 7, and doubtless belonging to the same individual. Like the bone just described, it appears to have been crushed and then recemented by cal- careous infiltration. Another important specimen is a nearly entire left radius, Pl. XIV. figs. 1, 2. This bone is 15-0 long; and the proximal end measures 2!-4 x 5'"8 in antero-posterior and transverse diameters, and the distal end 2!"6 x 3-9, whilst the least circumference of the shaft (5 inches below the summit) is 55. The perimetral index, therefore, of the bone is about *560. The principal, in fact the only difference of any importance between this radius and that of R. hemitachus from Ilford, as represented in a perfect specimen of the latter in Sir Antonio Brady’s magnificent collection, happily now in the British Museum (No. 45245), is that the former, with the same length, is rather the slenderer (in the proportion of 12 to 13), the least transverse diameter of the shaft being in it 1-95 and in the Ilford specimen 2’"1. But the antero-posterior diameter is the same in both, viz. 1-9, The least circumference in the Ilford specimen is 5""7, and its length 14!5, showing a perimetral index of -393. The proximal end of the Gibraltar bone measures 375, and of the Ilford 38. The distal end in the former is 25 x 3-9, whilst in the Ilford bone it measures 2’°6 x 42. This difference, at first sight, is considerable, and might, justly perhaps, be regarded as QUATERNARY FAUNA OF GIBRALTAR. 99 of importance, were it not explained by the circumstance that the lower end of the Ilford radius is diseased, presenting numerous large vascular openings and a general spongy condition, evidently due to what in surgical pathology would be termed chronic osteitis. This diseased condition is strongly manifested also in the radio-ulnar arti- cular surface; and it is a curious circumstance that Sir Antonio Brady’s collection of rhiuocerine remains from the Ilford brickfields contains a right and a left tibia in precisely the same mineral condition, and corresponding in proportional size with the radius, one of which tibiz is very extensively diseased at the distal end, exhibiting the consequences of chronic osteitis in a very advanced stage, being much enlarged and extremely spongy, whilst its fellow presents a similar disease at a less advanced degree. But even in this bone the affection has advanced further than in the radius. It may therefore be fairly concluded that all these bones in the Ilford collection belonged to one and the same individual, and therefore that the enlarged condition of the distal extremity of the radius is abnormal. ‘The shaft of the Gibraltar radius, besides its being a little slenderer than that of the other, is rather more curved in front, and the groove for the extensor tendons at the lower end is deeper and more pronounced than in the Ilford specimen. It is, however, well shown in the latter; and its depth is apparently diminished merely in consequence of the diseased condition of that end of the bone. Now, as this extensor groove appears to be one of the most distinctive characteristics of the radius of R. hemitechus, as compared with that of any other species, recent or fossil, with which I have had an opportunity of contrasting it, its existence and depth in the Gibraltar specimen appear to afford strong evidence in favour of that specimen belonging to the same species. On the dorsal aspect of the two bones I am unable to perceive any difference worthy of note, beyond the fact that all the muscular impressions are more strongly marked in the Ilford specimen. But this is a question of age or development; and in all other respects the bones appear to be identical. The os lunare (Pl. XV. figs. 4-8) is absolutely perfect. In mineral condition it is extremely dense and heavy, and highly infiltrated with manganesic oxide. The only fossil bones with which I have been able to compare it are a right and left from Grays, Nos. 22038 and 22038 4, in the British Museum, and belonging in all probability to the same individual. These bones are considerably larger than the Gibraltar specimen, measuring :— Grays. Gibraltar. lm. in. Meno thee wir ict ace cee Hie: oh Sento, 2s90 2°6 INV LG Lies ea Sipe Dee ee Nice oath eo 4 2230. IR, Antero-posterior diameter . . . . . 2°3 21 The scaphoid facet (a, fig. 4) in the Gibraltar bone is 1”°8 long, and in that from Grays 2-12; and the facet (a, fig. 6), which is triangular in the Gibraltar, is circular 100 MR. G. BUSK ON THE ANCIENT OR in the Grays specimen. The os dunare from Grays, from its size, I should refer, at any rate provisionally, to the larger of the two species found there, viz. R. leptorhinus (R. megarhinus); but as I have not seen any specimen of the same bone belonging with- out doubt to R. hemitechus, I can only employ the smaller dimensions of the Gibraltar bone as a doubtful argument in favour of its belonging to that species. The other specimens appertaining to the anterior extremity are portions of five metacarpal bones. Unfortunately these are all fragments of the lower end of the bones only. They include the distal articulation, and in some nearly half of the shaft of two right second and two left second metacarpals, together with a similar fragment of the third right. To judge from their dimensions, these bones would appear to have belonged to two individuals, one considerably larger than the other; but, except in size, the corresponding bones do not appear to present any appreciable differences whatever. The following dimensions, giving the antero-posterior and transverse dia- meters of the distal articular end, will suffice to show their not inconsiderable difference in size :— in, IAN EOC) iy sana Use aoe cena a aS ale Dia NICEO' (5) ial gtd SOE ee he A wens ey hnu nee MO BNE (2) (la, 4 AI Wl ta Nai arth eT Cal ANTICO A(G) peak, tinea al Shoat bei waddle weit daw LOLS SSpINTUGIRS (A) nis Melamed Se aca ada ace copa ILD In comparison with other species, it may be noted that the distal articular trochlea in R. bicornis (femur of same size) is 17 x 21, in R. hemitachus (Uford) (mean) 16 x 20, and in R. megarhinus, Grays, 2°12°5. These numbers afford additional evidence of the close similarity between the Gibraltar species and R. hemitachus, and at the same time of its distinctness from 2. megarhinus. 4, Hinder Extremity. The only clearly recognizable specimens belonging to the femur are (1) a fragment consisting of about the upper third of the bone of the right side, and (2) a detached third trochanter, with a portion of the shaft from which it springs, and, without doubt, belonging to the same bone. (1) The larger portion (Pls. XII. & XIII.) presents the head considerably abraded on the upper surface, the entire trochanter major and a portion of the shaft rather more than six inches long below the head; and on the inner border is the prominent tro- chanter minor, the lower point of which is at a distance of 5’6 below the lower border of the articular head. On the anterior aspect (Pl. XII.) the bone is much excavated below the head and trochanter, which arches over it on the upper and outer part. A slightly elevated ridge runs directly downwards from the outer border of the arti- QUATERNARY FAUNA OF GIBRALTAR. 101 cular surface, nearly in the middle. Besides this there are no particular marks. The posterior surface (Pl. XIII.) is very slightly excavated behind the trochanter major ; and in that situation it is perforated by several large vascular foramina, and has a coarse reticulated sculpture, but no digital pit. Elsewhere it is smooth, and evenly and very slightly convex. The outer border of the bone, below the overhanging end of the trochanter major, is square or flattened, the flat surface about an inch below the projecting border of the trochanter being 1:8 across; and the angles bounding it both before and behind are right angles, and both, but especially the anterior, quite sharp. The articular head, which measures about 3”:4 in antero-posterior diameter, has lost its upper half, apparently by abrasion, the abraded surface being originally covered, like the rest of the bone, with a thick hard crystalline stalagmitic incrustation. The specimen was found broken into several fragments, some of which were met with at 11-14, and others at 26 feet below the stalagmite floor of the Genista cave. And it is memorable as being the first fossil bone discovered by Capt. Brome beneath the stalagmite. It is obvious, from its appearance, that the bone has been crushed. (2) The detached third trochanter projects about 2”-6 from the border of the shaft, and is 2-6 wide. It curves gently forwards. This fragment, incrusted all over with stalagmitic deposit, was found at a depth of 11 feet; but that it belongs to the same femur as the portion above described is manifest from its colour and condition, and also from the circumstance that it shows evidence of having suffered from the same crushing influence. As compared with the femur of #. bicornis of the same dimensions, the prin- cipal differences observable are :— 1. That the trochanter major stands out at a right angle, or nearly so, whilst in R. bicornis its upper border slopes gradually downwards. 2. That the outer border of the shaft below the trochanter major, in R. bicornis, slopes obliquely backwards, the anterior angle being very acute instead of rectangular. 3. That the raised ridge on the anterior surface is placed more internally. 4, That the third trochanter is more curved forwards. 5. That there is a deep circumscribed digital fossa behind the trochanter major. With respect to the further determination of the species to which this interesting specimen should be referred, the only ones with which it is at all necessary to compare it are Rk. hemitechus and R. etruscus. I may mention that Dr. Falconer, who devoted much time and trouble to the com- parison of the Gibraltar bone, was inclined to refer it to 2. etruscus, considering it “the only one we cannot reconcile with the character of R. /eptorhinus.” He remarks that “‘the Gibraltar femur agrees so closely in form and proportions with a femur of [assigned to] R. etruscus presented to the national collection by Mr. Pentland, that had they been found in the same deposit they would have been referred without hesita- VOL. X.—PaRT I. No. 7.—August 1st, 1877. P 102 MR. G. BUSK ON THE ANCIENT OR tion to the same species. The correspondence holds in the contour of the articular head and large trochanter, and in the amount of uncination and expansion of the blade of the third trochanter, which yield distinctive characters of the femur in different species of Rhinoceros. The complete synostosis of the articular head, and strong tendinous ridges on the surface of the bone, prove that the Gibraltar femur is of an adult and even old animal. We have seen,” he says, “no adult femur of R. megarhinus that would correspond with it in size.” The following Table shows the comparative dimensions of the Gibraltar bone and of A. etruscus, as taken by Dr. Falconer and myself, together with those of 2. hemitechus and £t. bicornis &c, :— = || 4 a a 2 wals |-a z E a-p| & Ne} » 3 3 s so Di 4 BR a Be F 3 iloFla Sg |e 5 imensions of Kemur in various species of S 5/2 a8 qe| 3 Rhinoceros, a |/se\s aa|@ Cay oT eae. >) og CO /SR/SOR| SB] oe] - 3 J |SBloels3\33/s8 : 2=/ 65/5 a, S| ie aja |< |E lq4 |= Gibraltartfemunescemaeeieeieci tice ne 3:3 | 7-4 | 3-2 | 4:3 | 2:5 | 2-7 PLUM ELEUSCUS: ID a Mantas Sey tetMeraiet ele tonere iets 35 | 7-9 | 3:1 | 4:8 | 1-7 | 3-0 A. bicornis (Be, Kevtlod)) ..c. <5 vs et s~ cee 3:3 | 7-4 | 3-2 | 3-5 | 1-7 | 2-8 on », haltron Walden... ..<...0:....- 4:0 | 7-9 R. hemitechus (Brady) ................ 3°6 | At the time when this study of the Gibraltar femur was made by Dr. Falconer and myself we were unable to find a fossil specimen that came so near to it in size and general characters as the femur of FR. etruscus above referred to; but since then the specimens of the smaller Thames-valley femur in Sir A. Brady’s collection afford suffi- cient ground for considering it not at all improbable that it may belong to the same species, or one closely allied to it. It may also be remarked that, if the propor- tions between the femur and humerus were the same in the extinct form as they are in the existing R. bicornis, the diameter of the articular head in the Gibraltar bone quite accords with that of the head of the adult humerus already described. They may therefore be safely regarded as belonging to the same species, and, as I should suppose from their both being crushed in the same way, in all probability to the same animal, which all the other evidence seems to show was undistinguishable from R. hemitechus. A third specimen is a nearly entire tibia, represented in Plate XIV. fig. 4. This bone was found broken into numerous fragments, several of which were met with many feet apart; but I have been able to put them together in such a manner as to give a very fair representation of the bone in its entirety. It belongs to an immature animal, as both the proximal and distal epiphyses are naturally detached. It has therefore probably not reached its full size. Dr. Falconer and I compared it with QUATERNARY FAUNA OF GIBRALTAR. 103 numerous fossil tibiz in the British Museum, and especially with one from the Thames valley (No. 21884), termed &. leptorhinus, Owen. With this the Gibraltar bone agrees very closely in size, form, proportions, section of shaft, and contours of articular surfaces, whilst in all these respects it differs greatly from both Rhinoceros etruscus and R. tichorhinus. The tibia of the former, a Val-d’Arno specimen (No. 28805), while quite as long, is much slenderer in the shaft, with smaller articular surfaces ; whilst in R. tichorhinus the tibia is thicker, shorter, and more massive in all its pro- portions. The principal dimensions of the Gibraltar and Thames-valley specimens, together with those of the Val-d’Arno tibia are as under :— S S g s a 3 i BE 5 € | & Dimensions of Tibia in Rhinoceros. 33 Zz 3 E a 4 a5 Bar| 3 = cho ss ca 4 8 5 5, as g 3 =| 4 < 4 on Gibraltarjspecimen! (7... =- 2-22 150 | 51x49 | 30x38 “70 | -466 Tford. Brady’s collection ............ 146 | 52x49 | 30x37 69 | -472 Ditto dittowt s Tipit es sera 15:5 x51 x37 INS eellstey EEN Ie Seehciy bck OO eUOABBO Ge 145 | 51x48 | 27x39 R. etruscus, Val d@ Arno. No. 28805...... 141 | 43x38 | 24x34 | 62 439 R. tichorhinus (Wirksworth) .......... 15-9 R. incisivus (Santan)....7...........06 151 | 46x44 | 25x37 | 6:2 410 MEER URUCOTNES) .rateteret tsevads ele Celeste ees 145 | 55x55 | 33x43 | 75 | 517 From these figures will be seen the close correspondence, except perhaps in length, between the three Thames-valley specimens and that from Gibraltar, and the distinc- tion between them and the Etruscan form. 5. Astragalus. Of this important bone, the collection contains two specimens in excellent preserva- tion; one, in fact, is quite perfect (Pl. XV. fig. 3), and the other very nearly so. They are both of the right side, and, as regards mineral condition and in all other respects, precisely alike. The more perfect one was found cemented into the same mass of breccia as the distal epiphysis of the tibia above described, into which, when cleaned, it fits exactly.- There can be no doubt, therefore, that it belongs to the same individual. The second astragalus was found in a different situation, viz. at 21 feet in the black earth of the Genista cave, or at the same level very nearly as a metatarsal bone to be presently described. But as all the rhinocerine bones are in the same state of mine- ralization, they all doubtless belong to the same period. The principal dimensions of these astragali are given in the subjoined Table, together with those of the same bone in some other extinct and recent species :— P2 104 MR. G. BUSK ON THE ANCIENT OR | eval ale dali allan eee C) = 2 =} bina! 2 ia {3 [22 | # | 3 er ie Oe | = |8¢ 255 4s 42 Dimensions of astragalus in Rhinoceros. | & Re \ cane Mee sy Coes) ae Porn rss aamleelg 2 aa 38 BS ee ad] 35 [388] 23 a8 =e | £ | 8 | 38 |Ses|. bg Ee ss data fe ketal) Ce || ENTE Va ice 5° =e | wi ap eg 5 4 4 5 Gibraltar:vNo lio yst ne atte octet dacs 3°20 | 2:70 | 3:10 | 2:30 \3-1x2:1 |2:5x1-6 |2-7x1-6 | 5 Not) se aecetataa syomsmeceen | 3°20 | 2°65 | 3:15 | 2°37 x21 x1°6 | 2:7 x 1-7 | R. hemitechus, Uford, No. 22010, B.M. ..| 3°20 | 2°50 | 3:00 | 2°37 x 2:3 x16 ely Ws - No. 20815; BM. .. 2... -| 3:20 soe 3°30 cat Xigeso| 9 aea 5, megarhinus, Grays, No. 21617, B.M. ..| 3:65 | 2-9 3°60 | 3:00 x 2:5 x19 | 34x26 1-99 55 NON ZL Se «aye Goat wees ces 3°50 bs 3°50 Ed Xx 2-5 x17 |3:3x 2-1 | Ree DME Geese ocensanoodns0¢ 2-95 a 2°90 ao ||, S322 x15 | 2:8 x 1-7 | 5 Ss oMalacarray.ca-cupeehureiaetmesic te 3:00 | 25 | 3.00 | 2:25 x 2-0 x16 | 26x | , tichorhinus, Brixham collection ...... 2:90) 2:5 | 3:1 25 |2:9x 2:0 Apa pepds< ike) aires Cin alo) po necodedceaLaoe 2-70 | 2:5 3:15 | 2:5 |2°7x 2:2 |2-5x 1:65) 2-8 x 1:8 (wae as Sattron: Waldron... eerste 2-7 be 31 ae asia anes | 2:9x1-9 The mere inspection of the figures in these columns will show how very closely the Gibraltar bones correspond in size and proportions with that of R. hemitechus, or the smaller of the Thames-valley species, and at the same time how widely in the main they differ from any other of the species noticed. Had all the other characters been equally in accordance, there would have been no occasion for any further comparison ; but since, notwithstanding the remarkable resemblance in dimensions, the Gibraltar astragalus offers some characters by which it would appear to differ not inconsiderably from either R. hemitechus or R. megarhinus (the only two with which it is at all worth while to bring it into comparison), it will be necessary to add a few remarks. The astragalus of R. megarhinus, besides its much larger size, is distinguished from that of R. hemitechus (from the Thames valley) by several peculiarities. 1. The anterior or scapho-cuboid facet is more convex, and the proportional widths transversely of the scaphoid and the cuboid segments different, the former being to the latter as 1-000 to -405, whilst in 2. hemitechus the proportion is as 1:000 to -437. 2. In R. megarhinus the outer calcaneal facet extends, with an even outline, quite to the posterior and outer edge of the bone, whilst in R. hemitechus the posterior border of the facet is sinuous, a rough surface for the attachment of a strong ligament being left. between the margin of the facet and the postero-external border. 3. In the astragalus of R. megarhinus the outer ridge of the trochlea is more rounded or thicker than in R. hemitechus. 4. In R. hemitechus the crescentic internal malleolar facet is continued close up to the anterior border of the bone; and there is little or no constriction between its anterior termination and the scaphoid facet, whilst in R. megarhinus the crescentic facet terminates at a distance of nearly an inch behind the border of the bone; and, corresponding with this, the upper arched border of the facet measured along the are QUATERNARY FAUNA OF GIBRALTAR. 105 in R. megarhinus is 4-2, whilst in the far smaller astragalus of &. hemitechus the cor- responding arc measures 44, The Gibraltar astragalus agrees with that of R. hemitechus (1) almost exactly in size and proportions, (2) in the comparative flatness of the scapho-cuboid facet; and with that of R. megarhinus it agrees (1) in the form of the external calcaneal facet (vide Pl. XV. fig. 2), (2) in the termination of the crescentic internal malleolar facet at some distance behind the anterior border, as shown in the accompanying woodcut. But it differs from both in the greater proportional width of the cuboid Fig. 5. Astragalus of Rhinoceros from Gibraltar. facet, which stands to that of the scaphoid as 660 to 1000, and in having a marked con- striction (2 in woodcut) in front of the anterior termination of the crescentic facet, in which respect, though to a much less extent, it bears some resemblance to R. ticho- rhinus, though differing very widely in all others. It does not appear necessary to con- sider the relations of the Gibraltar bone with the astragalus of R. etruscus, nor with that of R. bicornis, its difference from both being too obvious to require comment. 6. Metatarsus. The bones belonging to this part of the skeleton are :—two right third metatarsals, one entire, and the other represented by the proximal half; and two right fourth meta- tarsals, one of which also is entire, and of the other only the proximal half remains. The entire third and the broken fourth metatarsals fit so exactly to each other, and correspond so precisely in colour and condition generally, that there can be no doubt of their belonging to one and the same individual. The entire fourth metatarsal, and the broken third do not fit very exactly; but this may arise in part perhaps from the circumstance that the third is a good deal worn. The bones, however, differ somewhat in colour, and were apparently found a good ‘way apart; so that the pre- sumption would appear to be in favour of their belonging to distinct individuals. 106 MR. G. BUSK ON THE ANCIENT OR As it was found in the dark cave-earth above the stalagmite floor of the Genista cave, in which one of the astragali was also discovered, it may be probable that the entire fourth metatarsal belonged to the same foot with that bone, although the depths at which they occurred were about 7 feet apart. 1. The entire third metatarsal (Pl. XVI. figs. 1-5) is 6”-7 long; the proximal end measures 1/6 x 1’"9, and the distal the same. The transverse diameter of the shaft at the middle is 1’-8, and at the smallest part 1-6; the least circumference is 4-2, and the perimetral index :626. The corresponding dimensions in a perfect third metatarsal in the Brady collection (of which an outline figure is given in Pl. XI. fig. 2) are almost identically the same—viz. length 6-7, proximal end 1-7 x1'"9, and the distal 1"-6x1"-9. The width of the shaft at the middle is 1"-8, and at the smallest part 16, the least circumference 4'"3, and the perimetral index *642. Nor do the bones present any appreciable difference whatever in conformation. The corresponding measurements in R. megarhinus, No. 19842, B. M., from Grays, are—length 8"-0, proximal end 2-5 x —?, distal 1-9 x 28, transverse diameter of shaft at middle 2-3, and at the smallest part 2"-1, the least circumference 5'-2, and perimetral index °650. The following Table shows the relative dimensions of the third metatarsal in other instances of fossil and recent Rhinoceros; but the above will suffice to show the close resemblance between the Gibraltar metatarsal and that of R. hemitechus, and at the same time its divergence from the megarhine type :— nS 3 ao) Ss é oS rd 3 3 5 Z Ret as 5 Piss! & es) od Dimensions of third metatarsal in various a ie = z = I g ca es a species of Rhinoceros. gg FI & a3 2 S Gus 3 58 :$ 3s 2 & | be Bal Gost loi ee el see wil weet ee alee =| <4 <4 < A Ay Ay (Gibraltarsispis-wse ek ee emer reeteres 66 |1-6 x1-9)155x1-9'8 x1-75) 42 | -6386 | -218 R. hemiteechus, Uford, No. 20245, B.M. ..| 6:8 |1:6 x2:2) x 2-0/85x1:8 | 4:4 | 647 | :211 3 33 Grays, No. 20267, B.M...} 6:4 /1:7 x1-9/1'6 x1:98 x16] 3:9 | -609 | -200 a 33 Sir A. Brady’s collection ..| 6°7 [16 x1-9/1'55x1:9'9 x18 4:2 | -621 | -200 BS 5 of Pleapptnerere 63 |16 x1-9)1-4 x1-8 » megarhinus, Grays, No. 19842, B.M. ..| 8-0 x 2:5)1:°9 x 2:39 x21 5-2 | -650 | -233 3 5 No. 23761, g. B.M....... 7-75 |1:9 x 2-4/1-85 x 2°39 x2:0 | 5:0 | -657 | -222 ,, tichorhinus, Kent's Cavern .......... 6:6 |1°75 x 2-2 x 2:2/1-0x 2:0 | 5:0 | -757 | -200 | ,, etruscus, Val d’Arno, No. 28808, B.M...| 6°85 |1:8 x2:0)1:5 x17/9 x16 | 3:9 | 567 | -180 | » ineisivus, No. 29646, B.M. .......... 6-2 {1:7 x2:0/1:8 x2:0/9 x1-6 41 | -668 | -180 5) Otcorns\(Et. Mettlod)| 05. eee es ee ne 5:9 {1-6 x1:9/1'5 x1-9/-8 x1-65) 4:2 | -712 | -206 = SUMAUCRSISS) RAGASe limes «ud eieeiaren a ae 6-3 x 2-0 x 2-0 | » untcornisy R:CiWasewiseseciocs Aisa a> © 79 |2°2 x~2-511:9 x 2°812-2 x25 2. The fourth metatarsal, represented in Pl. XVII. (figs. 1-5), is quite perfect, and of a fully mature animal, inasmuch as there is no trace whatever of the epiphysial suture. QUATERNARY FAUNA OF GIBRALTAR. 107 The bone is 6-2 long, the proximal end 1-6 x1'-75, the distal 1"6 x 1""4 the least circumference of the shaft 3-4, and the perimetral index ‘548. In the same bone from Ilford, No. 20816, B.M., the corresponding numbers are 6':0, 16x15, 1"°8 x 1'"'8, 3-4, and -566. Taking this specimen as the type of the fourth metatarsal in BR. hemitechus, it would, so far as the above dimensions go, appear to be more robust in the shaft, and to have a thicker distal extremity than the Gibraltar specimen. Subjoined is a Table showing the relative dimensions of the bone in other species :— Dimensions of the fourth metatarsal in z = z Rhinoceros. 2 =} a =) ™ a i=] a # a Z = 2 2 4 a A 4 a Gibraltar specimen ..............-0-: 62 | 16x1-75 | 16x1-4 3-4 | 548 R. hemitechus, Grays, No. 20816 ........ 60 | 16x15 18x18 | 3:4 | -566 KR. buornis (KR. keitloa) .......-....--.% 56 | 1:6x1°6 13x13 31 | -553 From the foregoing account of the Rhinocerine remains it may be concluded :— 1. That they belong to at least three individuals, varying somewhat in size and,” more particularly, in age. 2. That notwithstanding these differences, there is no reason for supposing that they represent more than one species, which was of about the same stature as, though somewhat slenderer in the extremities than the existing R. bicornis. 3. That in the dental and most of the osteological characters the Gibraltar Rhino- ceros, if not identical with, more closely resembled the smaller of the two Thames- valley species (R. hemitechus', Falc., R. leptorhinus, Owen, R. merckii, Lartet) than any other known extinct or recent form. * With respect to the proper appellation of the smaller Thames-valley Rhinoceros, which was regarded by Dr. Falconer as identical with the species first distinguished by him in the Gower Caves and elsewhere under the name of Ff. hemitechus, and also with that previously described by Prof. Owen from Clacton, notwithstanding all that has been written, some difference of opinion may well be entertained. Although in this account of the Gibraltar Rhinoceros I have employed the familiar term proposed by Dr. Falconer, I am by no means sure that it would not be better to retain the name given to the Clacton species by Prof Owen. It is quite true that R. leptorhinus, Owen, is not 2. leptorhinus of Cuvier, which, though it appears to have included R. megarhinus, Christol, and #. etruscus, Fale., does not seem to have embraced R. hemitechus; but, for the reason that Cuyier’s term is of uncertain application, has been pretty generally superseded by R. megarhinus, and that the 108 MR. G. BUSK ON THE ANCIENT OR 4, That, although in general the osteological characters are nearly identical with those of R. hemitechus, there are some differences, more especially in the astragalus, which it appears difficult to account for. 5. That, whether identical or not with R. hemitechus, it cannot possibly be referred to R. megarhinus, Christol, R. etruscus, Fale., or R. bicornis, the only other species about which there could be any question. TX. CErvus. Cervine remains abound in the breccia of Gibraltar. The predominance of rumi- nants was observed by John Hunter! in such of the specimens as were examined by him; and among the few which passed under his hand Cuvier identified two species of Cervus—one resembling the Fallow Deer, the other a larger form, which he regarded as being unknown in the existing European fauna’. Some of the paleontologists who have described the contents of the Bone-caves of the south of France have extravagantly multiplied the species of extinct Deer upon the most trifling grounds. Amongst these M. Marcel de Serres was preeminent ; of the cervine remains occurring in the caves of Lunel-Viel and Bize he has contrived to distinguish no less than ten extinct forms, to eight of which he gave specific names that merely serve to cumber the records of paleontology, since the majority of them appear to belong merely to varieties of the common or the Barbary Stag, or the Rein- deer °*. The Genista cave and fissure have yielded a considerable number of cervine bones derived from nearly all parts of the skeleton. But of the antlers there are only a few name of leptorhinus, Ow., has gained very extensive adoption, is employed in the British Museum, and has, moreover, been accepted, as I conceive, with fair reason, by Prof. Boyd Dawkins (Quart. Journ. Geol. Soc. xxiii. 1867, p. 217), it seems to me that it would be convenient and proper if paleontologists could agree upon its definitive adoption. A further reason might also now be given for the rejection of Dr. Falconer’s appellation, in the circumstance that it is anatomically incorrect, since it has been fully shown by Mr. W. Davies, in his excellent description of the Brady Collection (Catalogue of Pleistocene Vertebrata from the neighbourhood of Ilford, 1874, p. 30), that the smaller Rhinoceros of the Thames-valley had in all probability as complete an osseous septum as R. tichorhinus itself. Nowithstanding my great respect, or even veneration, for any opinion of the lamented M. Lartet, I am unable to accept R. merckii, Kaup, as the equivalent of R. hemitechus, seeing, if for no other reason, that the typical specimens of the teeth of R. merckit procured from Dr. Kaup himself are in the British Museum, and are indisputably those of R. megarhinus, Christol. Dr. Kaup, in fact, must haye confounded more than one species under the term R. merckit. 1 Phil. Trans. 1794, p. 408. * L.c. vol. iv. p. 173. * = a 7 9 ff DY? Spans ool Soc. Crt. JOT¢ VY EM B del Dinkel lith URS tise. Ak Cares: W.West &C° imp TEED Se eu beeen dir 49 398eM M LOU {OF FOOL . _ 15 mati dary @ 389M WHT PMC PP awa Lh Rlet 7%b °F PSD Zz » Vol 10° PL, Vil Z) IC Trans Doolt « to Q | WWest & C° imp E MB. de) et lith. CABALLUS. EQUUS Y TIX. (ZS bel foe oh 0 Lo Sram W.West & C°imp EMB del et lth CABALLUS. OUUS Be rant Loot Joe. & ol 10 IL X E.M.B.del.G West lith RHINOCEROS HEMIT@CHUS. WWest & C°imp DEEDES DEL 10F%, XI. EMB acl Dinkellith RHINOCEROS HEMIT@®CHUS. WWest 2 C2imp rant Loot foc. Vol 10-PL HI. E.M B.del G West lith RHINOCEROS HEMIT@CHUS. W.West & C° amp Trans Fook toc: Vol 10: L¢. AML. E.M.B.del. G-West lith RHINOCEROS HEMIT@®CHUS W West & C° imp Cy ta = ; i as n ° fe i 0 O° a ise a 1 Danikel lth el EMBd STM X W West & C° imp MT@CHUS : 4 5S gaa ai a0) Tes sient yy ete 5 Peden a : be Soke he (Fa OY ff 4 eran. Loot. FOC E MB del. G West. lith RHINOCEROS HEMIT@CHUS W West & C? imp LXV Py (Ss 2 10. de LOC: 7 Zool Soc. 4 os W.West &C° imp HEMIT@CHUS. RHINOCEROS EM. Bel. Geo West lith. -9 nee + iene E MB del Geo-Weat lith RHINOCEROS lope : (O-4 “A 7 eFramd Loot. SC HEMIT®CHUS Vel /OFY. XVI WoWest & C? imp Ye XVIIL Yi LM: Ze 00 Tot: a Gg If FAAS. XK c SPOILS =A 5 Sos j—2. RHINOCEROS [B del, Dinkel!it> — W West &C° amp E.M.B del. Dinkel lith DAMA CERVUS | eed WWest % C2 imp BIE, 470 8. BARBARUS FRY US Cr 3 Dinkel del et hth VU. XX. ( ec Za Lo FOC Dot/ @ » WD, x Tra WWest & C° imp BARBARUS CERVUS E M.B. del Dinkel hth ~~ aA duis 99 282M IA xXaai WaT 49198 PAC / \ \ ; & ee ee : oT a p bur yo uoyms \ (0G9 oN ; ‘ : xaqgr woeiauhy \ | smu gg \ | a 6 | \ sag? eundp,* EEE WEES *G) bel % amy) 19S) ; 3 } ve S e NS ve, : > te BG ra \ an 10482794) } aes nae \ \ | dan § wbopoy \ a —— \ om ames dn_ = ~vbopoyy \ 2 / j \ f \ / | | Vs ee Ore | \ 2 6 | / | Acland Sikes wilh x : \ } x / \ Nt \ } * te / \ / - — he = — DERG IV YF 7° POD? ‘ wanes - . x > he 3 4 <7_.+ — a an tl EMB del Dinkel lith Trans Look Foc. Vol, 10°F XX WWost & C%imp Furr 49 8 Sey MW WT AU PP a Wea NX PEOL PU, POP; E MB del. Dinkel lith LES Bs EXE Trans Lo0l Soe. Vol 10 L0 XV WWest £C° imp 2 S m t i ae . r * 2 : = ‘ — y ee 3 - * ’ = ’ a ? J 7 ; “ ae. ao Ra ayer Co Am : - 4 ~ a 4 lramd L00€ Soc. Col 10. PF? XX W.West & 0° imp IS ehexGs 3 | g a ad ao] Q = : é \ . =a ‘ } ‘ AKU. y) L 10-Lt L00U. At TUNE XA iS Tt t i a Hb Bera itarita os 1 0 i CI SNNNNE N Boe eee etc ANS 2S fo? re ee ee : ana THEN H SANS LC oo a H Beeaeaeee HH “ tt HH HEHE \ : \ CCN nee aueeaue a Begaa HHS AAS PEELE EEE PREECE EEE ESE ESSE HEE : t a oe a Hl an Tot 2 a | Be SN | ne ee ee : 5 EEE SEE EEE EEE 4 a tH i \ PRR EEE EERE . : HH i aon = t Coe Fe | rt par neaeee x sae Pag ale im 2 t a means i HAH ‘ F tH H H ct N ate s a 5 ®. S oe N sone C a ri } Baa 4 N x ene a 4 S 4 roe EE REEEH BE BE “a (a fa ta 4 ie | 8 H + aa ms LC 4 fl i am 4 Li Uy ‘ ESEEECH SEE PEEE EEE BEEEEEEE EE SEE HE | A EHR | peeatett bss EH Ht E at Li a aa - ase a0 i y ti oH suse BH sail lime aaa a ar OI 7 pila as Odontograms. [ 137 ] Ill. Notes on the Manatee (Manatus americanus) recently living in the Society's Gardens. By A. H. Garrop, I.A., F.Z.S., Prosector to the Society. Received October 11th, 1875. Read November 16th, 1875. [Puates XX VIII.-XXX.] UNTIL the arrival, on August the 6th of this year, of a living female Manatee in the Society’s Gardens, very few Europeans had had the opportunity of seeing a live Sirenian ; and those who had previously been so fortunate, had not, in most cases, been able to observe it under the favourable circumstances of position afforded by a small shallow pond approachable on all sides. In this pond it was not the least difficult to watch the creature minutely whilst it was feeding, as well as to make other notes of its habits which would necessarily be overlooked by those who meet it in its native haunts. Dr. Murie, in his valuable monograph on the Manatee !, justly remarks, that “ in museums it is customary to see such bloated over-stuffed specimens, that from them, as well as figures extant, an unfair idea of the configuration is obtained.” Dr. Murie’s own illustrations, with their beautiful textural details, represent the animal much more accurately. The only fault in them, if it is a fault, is rather in the opposite direction to that above indicated by him. His specimens were dead and preserved ; they had evidently shrunk slightly, the living animal appearing a little more rounded and presenting less well-marked skin-folds. The following are the measurements of the animal which I have had the opportunity of dissecting, taken almost immediately after death :— inches. otallencthe ews eee Meat ume os ek ss ge wee OU Circumference (8 inches in front of umbilicus). . . . . 50 Greatest breadth of tail(8 inches fromend) ... . . 19 VUITELE UKE SISA POE: VAIS pte aut met i ieee a a pememnmmgsea-tah Peeadee | MESSRS ECTS ea aman | pene i Rly i Rime RS i URE” 7) Circumference at root of tail (7 inchesbeyond anus) . . . 28°75 eapimattalllapper = Pre. el ele es tes LOS Wereth-Or LOLewIND es ae ee ee et ey ae Reanueve to trout OL SHOUL ij eee tle} EEO Breadth ai saout haliway dow... : . - - + + 2... oo Meiwnt of snautat muddle tine...) es le Oe rom ear to-earoyer head)... st et, OB ibnomveye te eye apoye heady.) =<. ee OE Promvear toveyeou onelside, .°. (2) 2A ob) BO 1 Trans. Zool. Soe. viii. p. 127. VOL. X.—Part 111. No. 1.—October 1st, 1877. U 138 MR. A. H. GARROD ON THE MANATEE. Whether in Dr. Murie’s figured specimen the snout and front part of the head had been swollen during life, whether the enlargement was the result of post mortem change, or whether it was on account of the youth of the individual, that part of the body is represented of considerably larger proportionate size with reference to the rest of the animal than it was in the Society’s living example, in which the head was more distinctly like a clean-cut truncated cone, without any well-marked transverse folds or appearance of puffiness. It is, however, in the oral margin of the upper lip that the deadness of Dr. Murie’s specimen is most manifest; and the fact that the peculiarity in the mechanism of that organ has not yet (so far as I am aware) been described, indicates how impossible it sometimes is to predict special functions from anatomical structure alone. The upper lip is prehensile; in other words, the animal is able, by its unaided means, to introduce food placed before it into the mouth without the assistance of the com- paratively insignificant lower lip. To understand how this is accomplished it is neces- sary that the structure of the labial margin should be described. A front view of the head(Pl. XXVIII. fig. 2) shows that on each side of the narrow median portion of the superior labial margin of the oral orifice depends a rounded lateral lip-pad of considerable size, which gives a deeply notched appearance to the upper lip. Of these lip-pads, Dr. Murie tells us’ that “at the dependent angle on each side of the muzzle is a circumscribed oval prominence, half an inch in diameter, where the ridges, furrows, and bristles [found elsewhere] are specially pronounced. This spot would seem to possess most tactile delicacy; for twigs of the infraorbital and facial nerves are abundant thereto. .... The above disposition strongly reminds one of the moustachial apparatus of the Walrus; but their shortness and rigidity render them unequal to perform the office of a sieve, as is the case in the Pinniped: they therefore incline to the hirsute covering of the muzzle of the Hippopotamus.” In another place? the same author, speaking of the lifting muscle of the upper lip (the levator labit supe- rioris proprius), remarks, ‘ Many vessels penetrate the root and origin of this levator ; this, no doubt, led Vrolik® to regard ‘the structure of the upper lip as plainly an erectile tissue.’” Observations on the living animal verify the correctness of Vrolik’s surmise. It is a post mortem change which causes the thickly bristled erectile lip-pads to be directed downwards. In the living animal their erectile tissue is distended with blood on most occasions, especially whilst feeding, and the pads are by this means directed inwards, towards one another, in such a way that the deep median notch which they go to form is even deeper and the bristles meet across the middle line. Further, these pads have the power of transversely approaching towards and receding from one another simultaneously (Pl. XXVIII. figs. 1 & 2). When the animal is on the point of seizing, say a leaf of lettuce, the pads are diverged transversely in such a * Trans. Zool. Soe. viii. p. 133. 2 Loc. cit. p. 149. * Mem. Zoolog. Soc. Amsterdam, 1852, p. 59. MR. A. H. GARROD ON THE MANATER. 139 way as to make the median gap of considerable breadth. Directly the leaf is within grasp the lip-pads are approximated, the leaf is firmly seized between their contiguous bristly surfaces, and then drawn inwards by a backward movement of the lower margin of the lip as a whole. The appearance produced by the movements of this peculiar organ is very much the same as that of the mouth in the silkworm and other caterpillars whilst devouring a leaf, the-mandibles in these insect larve diverging and converging laterally in a very similar manner during mastication. As to the mechanism of the process, the erectile nature of the lip-pads no doubt assists in the manner above indicated. These organs, when dilated, form more satis- factory muscular origins than when flaccid. Innumerable transverse muscular fibres connect their basal portions, forming the bridge above the median lip-notch, and approximate them. They are evidently separated and diverged by the levator labii superiors proprius, which is almost entirely distributed, in its insertion, to the con- siderable amount of skin-covered fibro-elastic tissue forming the interval between the nares and the lip-margin. With reference to the valvular mechanism for closing the nostrils during submersion, it may be mentioned that these circular orifices have each a flap valve, which forms the floor or inferior wall of the nasal tubes when the animal is breathing, but which rises and completely occludes it when closed, as represented in both the figures on Pl. XXVIII. Looking at the living animal generally, the most striking peculiarity was the slug- gishness of its movements. When crossing its pond there was none of the lateral movement of the body so characteristic of the Seals. All flexions were up and down, the whole trunk bending a little in that direction, the base of the tail doing so freely at a clearly marked transverse fold-line in that region. An opportunity occurred to me for seeing it out of water, when its pond was drained dry for a short time. From my observation on this occasion it is perfectly evident to me that the Manatee is purely aquatic in habits, and that it never willingly quits the water. When on land it seemed perfectly unable to advance or recede, the only move- ments it performed being that from its belly to its back, and vice versd. In these it made use of its limbs, flexing the body and the tail at the same time. When resting on its belly it seemed extremely uncomfortable, apparently on account of difficulty in breathing, which it is easy to account for on the assumption that the weight of the body being supported on the almost ribless anterior walls of the thorax and abdomen, compressed the abdominal viscera against the lungs, and so greatly diminished the respiratory movements. It would not remain in this position for any length of time, but seemed comparatively comfortable whilst lying quietly on its back, as it did until the inflowing water had refilled the pond sufficiently to allow of its again supporting itself in the water. u2 140 MR. A. H. GARROD ON THE MANATEE, I may mention that the power of moving the slightly exserted elbow was consider- able, whilst that of the wrist was small but apparent. It used its limbs much more freely than do the Seals, sometimes employing the extreme margins of the paddles to assist in introducing food into the mouth, at others employing them in progression along the bottom of the pond, during which time the swimming-tail could not be brought into play to any extent. The Manatee came into my hands within an hour or so of its death, and one of the earliest things that seemed desirable to do was to obtain some of its blood for examina- tion. The first cut through the skin was sufficient to prove how different an animal it is from any of the Pinniped Carnivora; for instead of the muscles being of a deep almost black-red hue, as they are in the Seals, they were of a pale pink, more like veal or pork than any other flesh known to me, much lighter than beef. The blood-disks are circular and non-nucleated, as it was certainly known they would be. Their size, however, is their peculiarity. From the valuable investigations of Mr. Gulliver, which are incorporated in their entirety in the Society’s ‘ Proceedings’ for 1875 (p. 474 et seqg.), it is known that the largest mammalian blood-disks are found in the Elephants (z7;5 of an inch in diameter), Great Anteater (3755), Sloths (zse5), Aard-vark (z55), and Walrus (g755). In the Manatee the diameter of the largest reaches 3455 of an inch, others being considerably smaller. If there is any stress to be laid on the size of the blood-disks in the classification of animals, as it seems almost impossible that there should not be from the comparative constancy in their size in closely allied species and genera, then the relationships of the Manatee to the Artiodactylate Ungulates must be most distant, as small size of blood- disks is a special peculiarity of those latter animals, the largest being z)/,5 of an inch in diameter, namely in the European Bison (Bison bonassus). In the Elephant and Edentates, on the contrary, the blood-disks are particularly large. With reference to the digestive organs there is not much for me to add to previous descriptions. In the stomach of our specimen the plications of the mucous membrane were slightly different from the figures given by Dr. Murie. Several well-marked, though not large longitudinal folds run along the lesser curvature of the first cavity from the cardiac orifice to the entrance of the second. These are bounded on the vertebral and ventral faces of the organ by a large, similarly directed fold, which imper- fectly separates off the irregularly plicated portion in connexion with the greater cur- vature from that in the region of the lesser. The mucous membrane of the second stomach is raised into rounded anfractuous folds, much like those of the human cerebral surface. The muscular parietes of the whole organ are very thick, extra- ordinarily so at the cardiac end. There is no pyloric dilatation of the duodenum. The intestines in their muscularity are very cat-like. A large number of vessels, forming quite a rete mirabdile, is to be found at the ileo-cxecal valve, in the angle between the large and small intestines. As Dr. Murie states, the ceca and the MR. A. H. GARROD ON THE MANATEE. 141 commencement of the colon are, when undisturbed, situated in the left diaphragmatic corner of the umbilical region of the abdomen. The following are the lengths of the intestinal viscera :— Smallbintestine:.0 25a: Malt, ROG. ee See ole 2 iharpesmiestinie gy! al. 0s eh se aetiek parken20b 838 Ceca from apices to ileo-cecal valve BR Each cecum externally . 0 1; The stomach was small in comparison with the size of the animal. The same may be also more certainly said of the bifid heart and of the lozenge-shaped spleen. The liver has a very peculiar shape, resulting from the very aberrant position of the lungs, which much curtail the transverse space usually occupied by that organ. Dr. Murie remarks, “ Jn situ, but still more so when removed, the entire liver has great resemblance in shape to the inflated lungs of an ordinary mammal.” This accurate simile would be even more so if the organ were compared to the lungs of a mammal distended with, say, solid jelly, and then cut down by a transverse slice which removed about one fourth their bulk from their apices. ‘The liver may be also said to form a cylinder, flattened from before backwards, transversely truncated in front, and irregularly excavated behind, or on its abdominal surface. The heart rests, with the intervention of a fibrous expansion from the diaphragm, on the truncated anterior end of the organ, which corresponds to the diaphragmatic surface as-usually described. Its dorsal surface is separated entirely from the spinal column by the interpolation of the lungs between the two. The bulk of the liver is formed’ by the two lateral lobes, between which, at the anterior end, are wedged the central lobes, the right of which is considerably the larger. The drawings (Pl. XXIX. figs. 1 & 2) will explain this better than any amount of description. The suspensory ligament is strong, and the umbilical notch small. The right lateral fissure is not deep, and does not extend up to the truncate superior surface, whilst the left lateral is considerable and does so, going quite to its vertebral border. The caudate lobe is only a slight extension of hepatic tissue along the vena cava: the Spigelian is elongate, conical, and directed backwards; it is well seen in the dorsal view. : Another peculiarity is a considerable bridge of hepatic tissue, extending, on the concave abdominal surface of the liver, from the vertebral portion of the right central lobe to the middle of the left lateral lobe. This bridge is not quite half as broad as it is long, and it is bent into a semicircle, the convexity of which is directed abdominally. With reference to the parts connected with generation, my observations entirely agree with those of Dr. Murie so far as the mammary development is concerned. No teats were to be found, nor any decided indications of their whereabouts. Just internal 142 MR. A. H. GARROD ON THE MANATEE. to the ventral insertion of the limbs there were one or two slight mamilliform thickenings of the integument, which had no deeper-seated glands apparently con- nected with them. The vagina is partitioned into two parts by the transverse hymen, in which two small apertures exist, each not more than 3 inch in diameter. The one, the common uroge- nital portion, is 24 inches long, with the median urethral orifice just in front of the hymenal openings. Between it and the short prepuce-like clitoris is a blind pocket, just large enough to admit the ungual phalanx of the thumb. The true vagina is two inches in length, with slight irregular foldings; its walls are dense and fibro-cartila- ginous in texture, as is the well-developed os uteri. The uterus closely resembles that of the Dugong, its cornua being 53 inches long, the body being only 24 inches. Its lining membrane is longitudinally plicated. The rudiment of the urachus is more preserved than usual; and from the anterior extremity of the bladder a fine tube extends along its axis for a short distance. Dr. Murie has described and figured the brain. He, however, mentioned that the specimen he examined was not in the most fit state for investigation. I was able to remove the organ within twenty-four hours of the animal’s death, when it was not in the least injured; my opportunities have therefore been more favourable, and the differences I have detected make me think it better to give fresh illustrations of its conformation. The brain does not present convolutions properly so called. The Sylvian fissure is large and bifurcates high up ; it ceases some way externally to the main longitudinal fissure of the brain; it also separates sufficiently at its origin to show the situation of the lobus opertus. A hippocampal sulcus also gives origin to the hippo- campus major. There are distinct indications of a superior frontal sulcus. The cal- loso-marginal sulcus is well marked, continuous, and extends far back, ceasing in front opposite the genu of the corpus callosum. Below this point the frontal lobe extends a great distance, as is clearly shown by Dr. Murie, and it presents slight indications of sulci on its antero-inferior angle (vide Pl. XXX.). The corpus callosum is short from before backwards, and in a median longitudinal section only covers the anterior and a little of the superior border of the thalamus opticus. Between it and the fornix no septum lucidum is developed, the precom- missural fibres forming a thick median longitudinal plane, extending from the corpus callosum,to the fornix. The anterior white commissure is small, as is the pineal gland. The corpora quadrigemina are not large. There is no trace of a posterior horn to the lateral ventricle. The hippocampal sulcus, however, extends so far up above the posterior portion of the thalamus opticus as to develop the basal part of the hippocampus major into a triangular lobule, partly obli- terating the back of the lumen of the cavity of the lateral ventricles, and bounded along its base-line, which runs from behind and externally forwards and inwards, by the ample fringe of the choroid plexus. The anterior horn of each lateral ventricle is MR. A. H. GARROD ON THE MANATEE. 143 deep, in correlation with the great extent of the lobe it excavates. The walls of the cerebral hemispheres are not at all thick, the lateral ventricles being capacious. Prof. E. R. Lankester being specially interested in the question as to the cervical nerves, dissected them out in the specimen under consideration. The following are his notes on the subject :— *« Dr. Murie, in his paper on the Manatee, states that in a specimen dissected by him there were eight pairs of cervical nerves, of which two pairs issued between the second and third cervical vertebre. Hence he is led to infer that the presence of only six cervical vertebra in the Manatee’s neck is to be explained by the supposition that the third vertebra is suppressed, leaving the two pairs of cervical nerves before and behind it, respectively third and fourth, unseparated bya vertebra. My attention was called to this statement by Prof. Edouard Van Beneden of Liége, who two years since dissected a Manatee at Brazil, and having especially searched for Dr. Murie’s third and fourth pairs of cervical nerves, had succeeded in finding but one in the position indicated. Prof. Van Beneden urged me, if opportunity should occur, to examine this point care- fully in the Manatee then living in Regent’s Park. I was kindly allowed to make this investigation by my friend Prof. Garrod; and I have to state that the following result was obtained by making first of all a dissection of the external course of the cervical nerves, and subsequently removing the skull and first three cervical vertebre one by one, so as to trace the nerves to their origin in the medulla. A single pair of nerves issues between the occiput and atlas, a single pair between the atlas and axis, a single pair between the axis and third cervical vertebra, a single pair between the third and fourth cervical, and a similar arrangement obtains for the fourth and fifth, fifth and sixth, sixth and first dorsal. The piece of the spinal cord, with the pair of nerves attached to it, which issues between the axis and third cervical vertebra (hence the third pair of cervical nerves) was removed by Prof. Garrod and preserved in spirit. The third pair of cervical nerves is considerably larger near its origin than is either the first or second nerve; but the most marked increase of bulk is observed in the fourth, which issues between the third and fourth cervical vertebrae. There was no trace in the specimen dissected of any additional nerve, or of any structure which could be taken for such a nerve, in the interspace between the second and third cervical vertebre. The third cervical nerve is connected by a slip to the fourth, and furnishes through this slip the highest origin for the phrenic nerve. “Dr. Murie does not expressly state that he followed his supposed two pairs of cervical nerves issuing between the second and third vertebre to their origin in the medulla; and from his drawing one is justified in supposing that he drew an infer- ence as to the existence of these two pairs of nerves from the position of two nerve- trunks, which it is possible he would have found united to form a single trunk if he had pursued them to their origin. “The fact which I am anxious to put on record is, that in two Manatees dissected 144 MR. A. H. GARROD ON THE MANATEE. since the publication of Dr. Murie’s statement as to these cervical nerves, careful search was made with the view of confirming his statement, and the condition described by him was not found to obtain; but the usual and normal arrangement of a single pair of cervical nerves between each successive pair of cervical vertebrae was in both cases ascertained with precision.” Mr. C. 8. Tomes has also kindly examined the minute structure of the teeth, with the following results :— “The teeth of the Manatee present several peculiarities of structure, and the resemblance in external form which they bear to those of the Tapir is not fully carried out in their histological characters. I am not acquainted with any one tooth which combines the characters of the tooth of the Manatee; and I believe that an examina- tion of a microscopical section would serve, with certainty, to identify a tooth as belonging to this creature. « eo = —— a P =e. > ie ates JSmit lith Hanhart imp. BRAIN OF MANATUS AMERICANUS yp fom IV. On Dinornis (Part XXI.): containing a Restoration of the Skeleton of Dinornis maximus, Owen. With an Appendix, on Additional Evidence of the Genus Dromornis in Australia. By Professor Owrn, C.B, F.RS., P.Z.S., &e. tead December 7th, 1875. [ PLates XXXI.-XXXIII.] THE indications, communicated in 1869, of an established form of Dinornis surpassing in size those to which the names Dinornis giganteus and Dinornis robustus had been applied have been strengthened by later discoveries, which equally justify the term maximus, and have contributed to an almost complete restoration of that huge species or propagable variety’. Osseous remains, agreeing in their dimensions or proportions with those figured in the undercited memoir, have been disinterred, chiefly from that notable locality “ Glenmark Swamp,” in the province of Canterbury, New Zealand, which have enabled the indefati- gable Director of the Canterbury Museum, Dr. Julius von Haast, F.R.S., to set up a skeleton, of which he has favoured me with photographs, bearing the name of Dinornis maximus, under which it is exhibited in that Museum. A side view of this skeleton is prefixed to Buller’s admirable work on the existing ‘ Birds of New Zealand’ (4to, 1872); but the angle of the trunk to the hind limbs somewhat exaggerates the height of the bird in comparison with the figure of the native Chief in the same plate. The abundance of the evidences of this hugest of known birds (recent or extinct) afforded Dr. Haast the materials for a proposition of exchange, which I had pleasure in submitting to the Trustees of the British Museum; and a series of casts, prepared for articulation, of the Megatherium americanum was, with their sanction, transmitted to the “ Direction of the Museum” of the capital of the Canterbury Province, in ex- change for a series of bones of the Dinornis maximus. In this series so few parts of the skeleton were deficient as to permit of its articula- tion. These parts were the atlas and axis vertebre, portions of the sternum and of the sternal ribs, also the scapula and coracoid of the right side, those rare bones of the left side being transmitted. Some of these missing or mutilated parts, from other specimens of D. maximus, I have since had the opportunity of describing through the liberality and kindness of other contributors, and especially of the accomplished keeper of the Museum of Science and Art in Edinburgh, William Archer, Esq., F.R.S. In now submitting a summary of the osteology of the species I have to supplement former papers chiefly by details of the vertebral structures. These I propose to combine ? Trans. Zool. Soe. vol. vi. p. 497, pls. Ixxxix. & xe. VOL. X.—PART 111. No. 3.—October 1st, 1877. Y 148 PROFESSOR OWEN ON THE GENUS DINORNIS. with comparisons of the homologous bones in the largest living wingless bird, Struthio camelus—to which end the admirable and usefully illustrated monograph by Professor Mivart, F.R.S., “ On the Axial Skeleton of the Ostrich”', lends peculiar facilities. In these comparisons I adopt most of the technical terms of aspect and position proposed by Prof. Mivart, in addition to my own, and I subjoin, for the convenience of students, their vernacular equivalents :— preaxial = fore, anterior ; postaxial = back, hind, posterior ; dorsal or neural = upper; ventral or hemal = lower, under ; neurad = upward ; hemad = downward ; antero-posterior or prepostial = longitudinal, fore-and-att ; dorso-ventral or neuro-hemal = vertical, high, deep ; lateral = side. medial = relating to the middle line or mid-vertical longitudinal plane of the body. ATLAS, or FIRST, VERTEBRA (natural size). Fig. 2. Aspects. Fig. 1, neural (or dorsal); 2, heemal (or ventral); 3, preaxial. The atlas vertebra of Dinornis robustus is described and figured in Zool. Trans. vol. vy. p. 395, pl. 53. To the three figures there given, answering to the 2nd, 3rd, and 4th in Mivart’s Memoir?, I here add views of the neural or dorsal, fig. 1, of the hemal or ventral, fig. 2, and of the ‘ preaxial,’ fig. 3, surfaces of the atlas of D. mazi- mus, to complete the comparative illustrations of the bone in the genus Dinornis. The preaxial articular cup (fig. 3, ac) for the occipital condyle is formed in great part by the hypapophysis (simulating the centrum), the neural vacuity being supplied by the true centrum of the atlas (‘odontoid process,’ fig. 4, ca): the sides of this vacuity are formed by a pair of articular surfaces developed on the atlantal neur- * Trans. Zool. Soc, vol. viii. p. 385. > Tom. cit. p. 388. PROFESSOR OWEN ON THE GENUS DINORNIS. 149 apophyses, homotypal with the prezygapophyses in the succeeding vertebre. The vacuity is progressively encroached upon by the growth of the prezygapophyses, and, in the atlas of the aged individval of Dinornis maximus (fig. 3), it is reduced to the chink a. The neurapophyses have met and coalesced above the neural canal, which was not the case in the atlas of the younger, but full-grown, subject of D. robustus (doc. cit.). Assuming the atlas of Struthio camelus! to have been from a full-grown and mature individual, a similar confluence of the neurapophyses having taken place, the following differences are chiefly notable between it and the corresponding vertebra of Dinornis maazimus. In the Ostrich the antarticular vacuity (not marked in Mivart’s figures), answering to a in fig. 3, remains much more widely open; the hypapophysial surface (ac) is less deep in proportion to its breadth; its lower border has not the pair of low tubercles (fig. 3, ¢, t); the hemal surface of the hypapophysis is produced downward and backward into the guasi-hemal spine, hy (Mivart, figs. 2-7); this is not present in Dinornis, but is replaced by a pair of low tuberosities, fig. 2, hy (which productions served for the attachment of the ‘ longus colli’), as in Apterya, but are variable. The difference between Dinornis and Struthio in the relative size of the vertebrarterial foramina v is well marked ; the larger size of the canal in Dinornis relates to its better- developed brain: the roof of the neural canal is relatively less extended from before backward in Struthio; it is convex, rough or irregular in surface, with a feeble indi- cation of a medial ridge at the fore part in Dinornis (fig. 1, 1), and with a hyper- apophysis (ib. hp) as a low tuberosity above each postzygapophysis. The postaxial articular surface presents, in Dinornis, a subquadrate convexity, and is not flat transversely in the present species or individual: the upper shortest border is moderately concave ; the lower longest border is framed, as it were, by a backwardly extended ridge, of which the pair of tubercles (fig. 5, ¢, ¢) form part. The postzyga- pophysial facets (z’, z’, fig. 2) look more obliquely backward than in Struthio, where their aspect is almost wholly inward or ‘ mediad.’ In the general though slight convexity of the postaxial articular surface of the atlantal hypapophysis, in the slenderness of the pardiapophysial bar (fig. 3, pd), defining outwardly the vertebrarterial canal, and in the parial disposition of the hypapophysial tubercles, the atlas of Dinornis elephantopus*® in the main agrees with that of D. robustus and D. maximus. In Struthio the neural arch has a less relative antero-posterior breadth, and the same proportional difference prevails in the guasi-centrum ; the processes, f, ¢, in fig. 2, are not developed; the preaxial cup has a wider upper emargination. The length of the axis in Dinornis maximus (fig. 4) is about four times that of the 1 Trans. Zool. Soe. yol. viii. p. 388, figs. 2-6. 2 Tbid. vol. iii. (1842), pls. xxxiv. and xxxv. fig. 2, a**. 3 Ibid. vol. iv. (1856), p. 162. Y¥2 150 PROFESSOR OWEN ON THE GENUS DINORNIS. atlas. and equals about 23 inches. The preaxial surface of the centrum (fig. 5, ca) is twice as broad as high, and is concave transversely, but less deeply than it is neuro- hemally. It is not prolonged neurally, as in the Ostrich, “ on to each side of the AXIS, or SECOND, VERTEBRA (natural size). Aspects. Fig. 4, lateral: 5, preaxial; 6, postaxial ; 7, hemal (or ventral). base of the odontoid process”; a slight non-articular concavity separates it on each side from the convex articular surface on the under part of the odontoid (figs. 4 and 7, : Bee! ca, 0). This surface is convex, narrower, and more produced than in the Ostrich”. * Mivart, loc. cit. p. 391. * Tbid. fig. 12, 0. PROFESSOR OWEN ON THE GENUS DINORNIS. 151 The postaxial surface (figs. 4 and 6, pc) has reverse proportions to the preaxial one, the longest diameter being vertical. It is divided into a pair of narrow vertically concave facets by a still narrower medial tract, the transverse contour being thus rather angular than, as in Struthéo and most birds, convex. The under border is nearly straight, and the transverse extent of the neural margin slightly exceeds that of the hmal one. The articulation slopes downward and backward. The hypapophysial process (figs. 4, 5, 6, 7, hy) is relatively more produced heemad than in Struthio, and descends vertically and slightly backward from the postaxial surface, thus adding to the length of the vertebra. The neural canal (figs. 5 & 6, 7) is half the length of the entire vertebra, and its width is nearly one third of the breadth. Its area is a full ellipse with the long axis vertical, not transverse as in Struthio. The ridges on the fore part of the hemal surface of the centrum in Struthio! are not present in Dinornis maaimus or in D. robustus. The parapophysis is represented by the short obtuse ridge, p, figs. 4 & 7; the pleurapophysis (ib. pl) by a similar one above, near the middle of the outer wall of the vertebrarterial canal, v. From the diapophysial plate (fig. 6, d), completing that wall above, a ridge (fig. 4, r) extends back- ward to the postzygapophysis, z’; this ridge circumscribes externally the vertical canal (figs. 4 & 7, s). On the medial side of this canal is the pneumatic orifice (fig. 7, 4), leading to the cancellous part of the neurapophysis. The canal s is not noted in Struthio, and in place of one large hole there are irregular pneumatic foramina2. The postzygapophysis is thrice the size of the prezygapophysis, but the antero-posterior hardly if at all exceeds the transverse diameter ; its aspect is as in Struthio and birds gene- rally. The hyperapophysis (figs. 4 & 6, hp) is relatively more prominent than in Struthio. The neural spine gains thickness as its base extends backward; its summit is broken off in my specimen, exposing the wide-celled pneumatic texture. In Dinornis elephantopus the hyperapophyses (fig. 5, hp) are relatively larger and higher ; the prezyapophyses are relatively less. The odontoid process (figs. 4 & 7, ca) is less than half the size of the atlantal hyp- apophysis; its free extremity is obtusely rounded. The third cervical vertebra, as in Struthio and birds generally, gains a transverse breadth of the neural arch, anteriorly, with concomitant size of the prezygapophyses (fig. 11, z), fitting the postzygapophyses of the axis. From this gain results a quadrate form of the roof of the vertebra; but, from the less relative length, or greater breadth, of this part in Dinornis, the roof (fig. 11) is a transverse quadrilateral, not so oblong as in Struthio: in both birds the angles are rounded off 3. The centrum is a horizontal wedge, with the edge anterior, concave, and formed by the neural border of the preaxial articular surface (fig. 9, ac), which, broad and con- cave transversely, is short and almost flat vertically, but here slopes from the vertical so much backward that it is on a plane with the contour of the hemal surface of the * Mivart, loc. cit. p. 391, fig. 12, ¢. * Ib. loc. cit. p. 393. Ib. loc. cit. p. 395, fig. 15, 152 PROFESSOR OWEN ON THE GENUS DINORNIS. centrum, as carried back by the hypapophysis (Ay), and forms a very open angle, with the base-line extended from its hemal border to the same border of the postaxial THIRD VERTEBRA (natural size). Fig. 9. Fig. 10. Fig. 11. Aspects. Fig. 8, lateral; 9, preaxial ; 10, hemal (ventral); 11, neural (dorsal). Oo PROFESSOR OWEN ON THE GENUS DINORNIS. lay: surface. Hence, in a direct front view (fig. 9), little more than the concave fore border of the preaxial surface (ac) is seen: while the whole surface is fully in view in the under view of the vertebra (fig. 10, ac). It looks almost wholly downward (hemad) in Dinornis, not obliquely downward and forward (hemo-preaxiad) as in Struthio". The postaxial surface (fig. 8, pc) much resembles that of the axis vertebra: its trans- verse contour is sinuous, a medial convexity dividing two concavities; the aspect is more upward than backward; the vertical exceeds the transverse diameter, but in a minor degree than do the reverse proportions of the preaxial surface. The pleurapophysis (fig. 8, pl) is more prominently marked than in Struthio; a low tuberosity represents the diapophysis (fig. 10,d); the interzygapophysial bar (fig. 11, 7) has gained breadth; the interzygapophysial foramen (fig. 11, s) is well defined. On the under part of the centrum (fig. 10) the transverse preaxial articular concave tract is followed by a broad depression beyond. From this begins the medial ridge, which expands into the tuberous hypapophysis (hy). This is less produced than in the axis. The chief differences from the Ostrich, besides the shorter or broader and deeper proportions of the entire vertebra, are seen in the more distinct hypapophysis, the better- developed hyperapophyses (fig. 8, Ap), and the more distinctly bifid character of the neural spine (figs. 9 & 11, ns). This spine rises a short way before it divides; a ridge extends from the fore and hind margins of each division, and defines the depression (for the insertion of elastic liga- ments) in front and behind the undivided base: this, at its summit, is not more than FOURTH VERTEBRA (3 natural size). Aspects. Fig. 12, lateral; 13, hemal. one fifth of the antero-posterior extent of the neural platform. (The specimen figured is from a higher and larger individual than the articulated skeleton.) 1 Mivart, loc. cit. p. 394. 154 PROFESSOR OWEN ON THE GENUS DINORNIS. The fourth cervical vertebra of Dinornis maximus (figs. 12, 13), representing, with increase of size, the proportional characteristics of the third, resembles that of the Ostrich in the greater production of the pleurapophysis, p/, and in the absence of the hypapophysis. The interzygapophysial plate, 7, not extending to the postzygapophysis, z, does not circumscribe the space forming the foramen, s, in figs. 8 & 11; and the pneumatic foramen (fig. 12, pn) comes into view. The zygapophysial articulations, z, z, become elongated; the posterior outlets of the vertebrarterial canals expand. There is little, if any, modification of the pre- or post-axial articular surfaces (ac, pc) of the centrum. The hemal depression (fig. 13, ¢) behind the preaxial surface (ac) is deeper than in the third cervical; and the angle between this and the hind half of the centrum, ¢’, owing to the non-development of the hypapophysis, is more marked. The spine (fig. 12, ns) repeats its small basal extent and bifid character. The diapophysial plate (fig. 12, d) extends its origin from the outer side of the prezygapophysis, z, halfway towards that of the postzygapophysis, before it bends down to coalesce with the pleurapophysis, p/; the broad outer wall of the vertebrarterial canal (fig. 13, v) thus formed is the ‘pleurapophysial band’ of Mivart'. It sends forward from its lower anterior angle a short obtuse parapophysis (fig. 12, p). The riblet, p/, extends backward from the opposite or hinder angle. ‘he vertical hind border of the ‘band’ has two semilunar insertional impressions, the angle (fig. 12, @) between which is less produced than in Struthio. The pleurapophysial band has a relatively greater vertical extent than in Struthio; and this relates to the corresponding excess of vertical over longitudinal dimensions in the entire vertebra of Dinornis as compared with Struthio. In the direct under view (fig. 13) the pleurapophysis extends almost to the vertical level of the postzygapophysis, 2’ (compare with fig. 24, Mivart, Joc. cit.); a more marked difference from Struthio is in the bifid neural spine of Dinornis. There is no medial hypapophysial ridge in D. maximus. In D. elephantopus the fourth cervical has the hinder half of the lower surface of the centrum relatively wider than in fig. 13, ¢; the prezygapophyses are less produced forward than in fig. 12, z. Glancing along the cervical region, in the articulated skeleton of Dinornis maximus, one sees, as in that of D. elephantopus, that the two (parial) neural spines continue to be developed throughout that series of vertebra, the uniting basal band subsiding some- what in the fifth cervical, and each spine being then represented by a ridge continued forward from the hyperapophysis, converging toward its fellow as it rises; but it attains no great height in any vertebra. In the fourteenth cervical, where the parial neural spines are most marked in this respect, the uniting base gains in vertical extent. * Loe, cit. p. 398, Or PROFESSOR OWEN ON THE GENUS DINORNIS. 15: The parial hypapophyses (‘ catapophyses,’ Mivart) commence at the fifth cervical as low tubercular ridges. They come nearest to each other at the fourteenth, but do not, in Dinornis, circumscribe a hemal canal in any vertebra’. In the fifteenth cervical the parials combine to form a single medial hypapophysis near the middle of the length of the under surface. In one skeleton of D. elephantopus this coalescence takes place at a sixteenth cervical, the antecedent series having one more vertebra than in the skeleton of D. maximus here described. The pleurapophysial plate is sculptured outwardly by longitudinal ridges and channels; the riblet loses relative length after the sixth or seventh cervical. The pre- and post-axial articular surfaces retain their essential character throughout, being con- cayo-convex in opposite directions; the fore surface is always superior in breadth, and this dimension, though less in the hind surface, is greater than the vertical diameter. A larger proportion of the neural surface of the fore end of the centrum is uncovered by the neural arch after the third and fourth cervicals. From the neck series are selected vertebrae for views corresponding to some of those given by Mivart of the Ostrich, which best illustrate the modifications of such vertebrae in the larger flightless bird. SIXTH VERTEBRA (4 nat. size). Fig. 14. ee hy GZ pe Fig. 14, hemal (ventral) aspect. The hypapophyses in the sixth cervical (fig. 14, hy) are oblong, smoothly obtuse tuberosities. The exterior of the parapophysial part (p) of the pleurapophysial plate Comp. with this modification the cervical vertebra in the Flamingo (‘ Anat. of Vertebrates,’ vol. i. p. 29, fig. 20, h). VOL. X.—PART 111. No. 4.—October 1st, 1877. z 156 PROFESSOR OWEN ON THE GENUS DINORNIS. (p, ps) is longitudinally channelled and ridged ; the riblet is shortened, as in Struthio !, but in a greater degree. The interzygapophysial bar, though short, leaves a foramen before it islostin the base of the postzygapophysis. ‘The metapophysis (fig. 14, m, m) is a mere slight outswelling of the diapophysial mass. The anterior depression at the hemal surface of the centrum is no longer defined; it is the beginning of the longitudinal channel hf, banked by the hypapophyses. Behind these the surface is smooth and flat; then again becomes slightly concave transversely at the expanded hind part of the centrum, pe. The neural spine is represented by a pointed ligamentous surface above the fore border of the arch; behind this rises the pair of low obtuse processes subsiding into the hyperapophyses above the postzygapophyses. As the neural roof subsides behind the part between the bases of the parial, guasi- neural, spines, their connecting-bar is so indicated. ‘The hind part of each spine extends, subsiding to the hyperapophyses. TWELFTH VERTEBRA (i natural size). Fig. 15. Aspects. Fig. 15, lateral; 16, postaxial. The twelfth cervical is chiefly distinguished by the nearer proximity to one another of the hypapophyses, the extremities of which, in a direct hind view (fig. 16, hy), appear below the divisions of the postaxial surface, pe, not at its sides, as in Struthio’. The transverse diameter of the postaxial part of the neural arch, taken outside the pedicles, is one fourth less than the same diameter of the preaxial part. The riblet (fig. 15, p/), though longer than in the six or seven preceding cervicals, is relatively ? Mivart, loc. cit. fig. 25, p. 400. ? Ib. loc. cit. p. 493, fig. 30, c. PROFESSOR OWEN ON THE GENUS DINORNIS. 157 shorter than in the fourth (fig. 12, pl). The neural arch attains its greatest length in the twelfth vertebra. The connecting bar (fig. 16, 2) of the parial neural spines is slightly raised, and is better defined before and behind by the rough pits for the elastic ligaments than in some of the antecedent cervicals, The breadth of the neural platform across the postzygapophyses is relatively greater, in a small degree, than in the antecedent cervicals; yet their articular surfaces (fig. 15, pz) remain longer in proportion to their breadth. The perforation of the interzygapophysial bar continues. The longitudinal ridges of the pleurapophysial plate (fig. 15, pl) are more pro- minent ; the upper one assumes more the character of a metapophysis (ib. m). The cervical vertebra of Dinornis giganteus, figured of the natural size in plate 40, vol. iii. of the ‘ Zoological Transactions,’ is the answerable one to the twelfth in the huger representative species of the South Island of New Zealand. It is less broad in proportion to its length, and thus conforms to the more slender metatarsals cha- racteristic of D. giganteus of the North Island. The transverse connecting bar of the neural spines, marked s in fig. 5 of the plate 40, rises nearer to the summits of the parial divisions; the ridges continued from these, forward, converging to the fore margin of the neural arch, are longer and broader than in Dinornis maximus; the hollow behind the neural spines is also broader. The thirteenth and fourteenth vertebre in Dinornis are most nearly matched by the sixteenth and seventeenth in Struthio camelus, in which species the eighteenth vertebra FOURTEENTH VERTEBRA (4 natural size). Fig. 17. Aspect. Fig. 16, heemal (ventral). (fig. 41 of Mivart), like the fifteenth in Dinornis (fig. 19), first changes its parial or double for a single hypapophysis, Ay. My figure 17 may therefore be contrasted with Z 9 “a 158 PROFESSOR OWEN ON THE GENUS DINORNIS. figures 34 and 39 of Mivart, which show the same aspect (hemal or ventral) of the yertebree compared. And here I may note that the processes seen in profile in Mivart’s figures 33 and 34 are indicated by the symbol ¢ in the sixteenth vertebra, and by the symbol fy in the seventeenth; similarly, they are described as ‘ catapophyses’ at p. 405, and ‘ hypapophyses’ at p. 406. I note them, under the latter denomination, in the thirteenth as in the fourteenth vertebra of Dinornis maximus, and the similarity is such between these vertebra that I proceed to the description of the fourteenth. In the fourteenth cervical (fig. 17) the more approximated hypapophyses, hy, arise from a low common prominence further back from the preaxial surface, and the lon- gitudinal channel in the lower surface, in the twelfth and thirteenth vertebre, is here somewhat interrupted by such prominence. ‘The transition of these into a single medial hypapophysis is thus indicated. The present vertebra in Dinornis maximus approaches the character of the seventeenth vertebra in the Ostrich}, especially in the above modification of the hemal surface 2, to which view of the vertebra of Dinornis maximus, corresponding to the seventeenth of Struthio, I here restrict my illustrations of such vertebra. The processes (¢c in Mivart’s fig. 34, hy in his figure 59) are serial homotypes. The recognition of this fact led me to speak of Mivart’s ‘ catapophyses’ as “ parial hypa- pophyses ” inthe Memoir on Cnemiornis (Trans. Zool. Soc. ix. p. 260), and again, under a sense of the convenience of a substantive term, as ‘ prehypapophyses’ (ib. ib.), in contradistinction with the ‘hypapophyses’ at the hind part of the centrum in the axis and third cervical. The antero-posterior extent of the pleurapophysial plate is shortened in the fourteenth vertebra of Dinornis, as in the seventeenth of Struthio; but the pleurapophysis itself is less produced in Dinornis. The neural spines have not approximated and coalesced as*in Séruthio. The section of the supporting column of the parial neural spines is transversely quadrate; both fore and hind surfaces are impressed by a definite rough tract for the elastic ligaments. The preaxial surface retains a greater relative breadth to the postaxial than in Séruthio; the vertebrarterial canals are relatively wider. The next step in the transmutation of Mivart’s ‘catapophysis’ into the normally situated single hypapophysis in birds is presented by the fifteenth cervical of Dinornis maximus (figs. 18-21), which is the last of that series in the present skeleton. A single obtuse process descends from a low base coextensive nearly with the hzmal surface of the centrum (fig. 19, hy) ; but the base of this process in one example is connected by a ridge continued from each side to the hind border of the pleur- apophysis (ib. p/), and there is a slight swelling (the final trace of the parial character) at the beginning of each ridge. A pair of low tuberosities, connected by a ridge, mark the hind border of the lower surface of the centrum. With the vertical extension of bone, hy, from this surface for muscular attachments, 1 Mivart, loc. cit. p. 406, figs. 35-39. ? Tb. ib. fig. 39, PROFESSOR OWEN ON THE GENUS DINORNIS. 159 a corresponding but greater one marks the opposite or neural surface, one process (hy, fig. 18) descending, the other (ns) ascending. The neural spine gives off a pair of low tuberosities, one on each side, near its summit: from each there is continued the usual FIFTEENTH VERTEBRA (3 nat. size). Fig. 18. oF Aspects. Fig. 18, lateral; 19, heemal (ventral), ridge curving back to the hyperapophysis (fig. 18, hp), which still overtops the post- zygapophysis, pz. A mere rudiment of the interzygapophysial band now remains, but behind it is a small foramen leading to the cancelli of the neurapophysis; a corresponding foramen is noticed in Struthio!. The pneumatic foramen is, as usual, beneath the base of the diapophysis, which process shows its tuberous outstanding metapophysis (fig. 18, m), well marked above the pleurapophysial band, p/. This, as in the seventeenth (last cervical) vertebra in Struthio, is short antero-posteriorly, and each margin is concave, with a blunt production of its hinder and lower angle still representing the cervical riblet. Each vertebrarterial canal (figs. 20, 21, v), asin Struthio, exceeds the neural canal in capacity. If the transverse expansion of the fore part of the centrum be reckoned as due to the * Mivart, loc. cit. p. 409, 160 PROFESSOR OWEN ON THE GENUS DINORNIS, FIFTEENTH VERTEBRA (2 nat. size). Fig. 20. Sherwin N hy Aspects. Fig. 20, preaxial; 21, postaxial. PROFESSOR OWEN ON THE GENUS DINORNIS. 161 bases of ‘parapophyses,’ the fore or preaxial articular surface (ac, fig. 20) may be said to extend thereupon; but the parapophysis, ps, in Dinornis maximus, may better be held to spring out external to the preaxial surface than in Struthio. The outer border of the prezygapophysis (fig. 20, az) now begins to rise, and gives a more inward or medial aspect to the oblique articular surface. The postzygapophyses (fig. 21, pz, pz) show a corresponding change in the contrary sense, but they do not extend postaxially beyond the centrum in so great a degree as in Struthio!. A marked difference between the vertebree here compared is in the greater height and greater breadth of the neural spine in Dinornis; but the chief distinction is shown by the coexistence in Struthio of an independent or movable pleurapophysis with the first appearance of the single and simple hypapophysis. In Dénornis such condition of the hypapophysis is associated with a continued confluence of the riblet, p/. In other words, the single hypapophysis marks the first dorsal vertebra in Struthio? and the last cervical in Dinornis (fig. 18, hy). The character of the fifteenth cervical in the series of the skeleton of D. maximus is that of the sixteenth cervical in the neck-series of Hutton’s skeleton of D. ele- phantopus, and this is followed by a seventeenth cervical, or one with anchylosed pleur- apophyses, beyond which there are seven vertebre for a dorsal series. But in both the sixteenth and seventeenth cervicals the neural spine is bifid; the ridges from the hyperapophyses converge to the base ofa single neural spine only in the first of the series of vertebra in which the pleurapophyses retain their independence and mobility. If my series of cervicals in D. maximus be, as it seems by characters of juxtaposition, the correct number, Hutton’s specimen of the skeleton of D. elephantopus has two additional cervicals, in all seventeen, instead of fifteen as in Apteryx. These remarks are based on a photograph of the skeleton in the Otago Museum. In a cervical vertebra (figures 22, 25, 24) of D. giganteus, which I regard as homo- logous with the fifteenth or last cervical of D. maximus, the neural spine (fig. 23, ns) retains, as in D. elephantopus, its bifid character, but the parial portions are relatively less developed, and their connecting bar (ib. 4) has begun to rise, indicating, as it were, a rudiment of the single and longer neural spine in D. maximus. The hypapophysis (figs. 22, 24, hy) is single, and its base is supported by the ridges from the pleur- apophyses representing the ‘catapophyses’ of Mivart. The posterior hypapophysial tubercles (ib. hy’) are better marked than in figure 19 (D. maximus). The neck-vertebree in every species of Dinornis in which I have been able to deter- mine them correspond, with unimportant modifications, with those above described and figured, and in like degree differ from their homologues in Struthio. In the well-marked class of Vertebrates characterized by the many cervical vertebre, these, as a rule, are small; but in Apteryx, and especially in Dinornis, they are excep- tionally large. Some of those in Dinornis maximus almost equal in size the neck- vertebre of the horse. ’ Comp. fig. 18, pz, with Mivart, loc. cit. p. 406, fig. 35. 2 Mivart, loc. cit. p. 408, fig. 40, hy. 162 PROFESSOR OWEN ON THE GENUS DINORNIS. The muscular system, as has been shown in Apteryz, is correspondingly developed ! ; and when, to the proportionably still more powerful neck-muscles in Dinornis, were added their thick integument and covering of feathers, the neck must have been a feature of strength very different from the slender character of that lengthy part in an Ostrich and the like living birds. FIFTEENTH VERTEBRA, Dinornis giganteus (3 nat. size). Fig. 22 Fig. 23. g. <=. Aspects. Fig. 22, lateral; 23, neural; 24, hemal. This consideration adds significance to the record of one of the oldest living colo- nists in New Zealand, recently published. * Trans. Zool. Soc. vol. iii. p. 288 et seq. pls. 31-35. PROFESSOR OWEN ON THE GENUS DINORNIS. 165 “In 1844, at Wellington,” writes Mr. Hamilton, “I was present, as Governor Fitzroy’s private secretary, at a conversation held with a very old Maori, who asserted that he had seen Captain Cook. Major Richmond, then Superintendent of Wel- lington, was, I think, also present. I cannot recollect who was the Governor’s inter- preter. ‘This Maori (Haumatangi), so far as my memory now serves me, I should guess was 70 years old; at all events he was brought forward as one of the oldest of his people then residing about Port Nicholson. Being asked ‘Had he ever seen a Moa?’ he replied, ‘ Yes, he had seen the last one that had been heard of.” When questioned as to what it was like, he described it as a very large tall bird, with a neck like a horse’s neck. Atthe same time he made a long upward stroke in the air with his right hand, raising it far above his head, and so as to suggest a very fair idea of the shape of a Moa’s neck and head, such as I have since seen them in the skeleton birds of the magnificent collection which Dr. Julius Haast has gathered together in the Canterbury Museum. There is no bird or animal of large size indigenous to New Zealand to which an old Maori could liken the Moa. The horse was probably the only creature imported by us in 1844 in which he could possibly find any kind of likeness calculated to give ws a fair general idea of the shape and height of the bird’s neck and head. If he had never himself seen a Moa, how—unless he had received its description, handed down from Maoris, who had seen one—could he possibly have hit upon such an idea as to refer us to the tall arched neck of the horse for a likeness ? The gesture which he made with his hand remains impressed upon my memory as freshly as if seen only yesterday, as one that was singularly descriptive. It was like a sketch being made, as it were, in the air” }. Reckoning, by a convenient, though somewhat artificial character, as a first dorsal the vertebra which first retains its pleurapophyses as independent movable elements, such vertebra (the sixteenth), in Dinornis maximus, answers to the eighteenth in Struthio, of which Prof. Mivart gives two instructive views (3 natural size) 2. I subjoin a corresponding figure (figs. 25, 26), similarly reduced, of the first dorsal in Dinornis maxinus. If the hypapophysis (fig. 26, hy) be taken as a guide, the present vertebra in Dinornis would answer to the nineteenth in Struthio, which is the second vertebra in that genus showing the single medial hypapophysis at this region of the spine, and asso- ciated with the articular facet, p, for the movable pleurapophysis. In Dinornis the parapophysis, p, is less produced forward or outward; the neural spine, 7s, is more elongated and inclines forward; it is also thicker, more quadrate in section. In another vertebra it is less elongate than in the figure and less inclined forward ; the costal surface, also on p, is likewise deeper and is subcircular in shape. 1 “ Notes on Maori Traditions of the Moa,” by J. W. Hamilton, Esq., ‘ Transactions of the New-Zealand Institute,’ vol. vii. 1875, p. 121. 2 Loe. cit. p. 408, figs. 40, 41. VOL. X.—PaRT 11. No. 5.—October 1st, 1877. 2A 164 PROFESSOR OWEN ON THE GENUS DINORNIS. ‘The preaxial surface has its vertical extent not so inferior to the transverse as in Struthio. The pleurapophysis articulates with a small shallow pit on a very short parapophysis ; the ‘head’ is supported on a neck 14 inch long, and slender in proportion to the body and tubercular process, which is sent off at an angle of 45’ with the neck; it terminates by a smooth round tubercle, fitting a corresponding pit on the lower surface SIXTEENTH VERTEBRA (‘ 1st dorsal,’ 3 nat. size). Fig.26, Fig. 25, neural aspect. Fig. 26, lateral aspect. of the diapophysis, which it thus underprops. The body of the rib is flattened, 1 inch 3 lines broad at the divergence of the cervical and tubercular branches; it is slightly curved inward and forward, and gradually terminates in a point. No hemapophysis (sternal rib) is developed in the sixteenth (1st dorsal) vertebra of Dinornis maximus. In the first dorsal vertebra of D. elephantopus the hypapophysis is more central in position, more tuberous, less compressed, with a shorter base; in other words, retaining more of the character of that process in the last cervical. The seventeenth vertebra, answering to the twentieth or third dorsal in the Ostrich, and repeating the character of the hypapophysis in the first dorsal, exemplifies also the difference of being the first of the vertebral series, traced from the skull, in which the segment, or osteocomma, is completed by a perfect hzmal arch. PROFESSOR OWEN ON THE GENUS DINORNIS. 165 The centrum and neural arch show no notable differences from those of the sixteenth vertebra. The pleurapophysis, with a similar double articulation, has increased in size and conspicuously in length; it also supports an ‘ epipleural appendage,’ which is broad and slightly curved upward, where it overlaps the succeeding rib. The hemapophysis is ossified and synovially articulated with the pleurapophysis above and with the hemal spine (‘sternum’) below. Towards its hemal end it expands and develops a tubercle. The size of the dorsals in D. maximus increases slightly as they recede, and chiefly in breadth, by the outgrowth of the diapophyses, accompanied by a greater size of the rib and greater length and divergence of its capitular and tubercular processes. In the present vertebra (third dorsal) the base of the hypapophysis, though shorter than in the second dorsal, occupies a greater extent of the hemal surface of tlie centrum than jn the third dorsal of Struthio. Not more of the fore part of the neural surface of the centrum is exposed than in the antecedent dorsals and terminal cervical vertebra. The postaxial surface continues to be narrow in proportion to its height ; but its transverse convexity increases, and is relatively greater than in Struthio. The transverse concavity of the preaxial surface has also increased; it is still convex ver- tically along its middle third. In a homologous vertebra of the present species of Dinornis I have noted a variety in the hypapophysis in the interruption of its basal extent producing a small guasi second hypapophysis near the postaxial surface. The pleurapophysis, with a slight increase of length, and of that of its appendage, is as in the second dorsal. The hemapophysis (‘sternal rib’) articulates by a trans- versely extended bitubercular end with the sternum. The nineteenth vertebra (fourth dorsal, figs. 27-29), corresponding with the first of those having their pleurapophyses free and articulating with their haemapophyses in Struthio (figures 47, 48, ‘ Mivart,’ p. 413), has the centrum less cuneiform in transverse section, the sides converging, with a certain convexity, hemad to a low and short ridge or keel, produced and thickened anteriorly, near the preaxial surface (fig. 27, hy). Prof. Mivart reckons the dorsal series as commencing with the vertebra thus typi- cally complete in regard to its hemal arch. I prefer to retain the character of a free pleurapophysis as denoting the present class of axial segments. Thus the nineteenth vertebra in Dinornis, or fourth of the dorsal series, answers to the twenty-first in Struthio, which is the fourth supporting a free pleurapophysis (vertebral rib), and the first in which this element articulates with its hemapophysis (sternal rib). The hemapophysis of the twentieth vertebra in Struthio is developed, but is articulated only with its spine (sternum) and does not join by its opposite end the pleurapophysis. Such condition I have not yet seen in any species of Dinornis. With respect to the twenty-first vertebra in Struthio, Mivart remarks, that “it is so much like the twentieth that little need be said in its description” (p. 413). My figure 28 may therefore be contrasted with figure 46 in Mivart’s monograph (p. 411, loc. cit.) for illustrations of the differential characters in question. 2A2 166 PROFESSOR OWEN ON THE GENUS DINORNIS. NINETEENTH VERTEBRA (‘4th dorsal,’ 1 nat. size). Fig. 28. gee Aspects. Fig. 27, lateral; 28, preaxial; 29, neural. Q PROFESSOR OWEN ON THE GENUS DINORNIS. 167 I subjoin three views (figs. 27, 28, 29) of the nineteenth, or fourth dorsal, vertebra of Dinornis maximus, corresponding with those views of the twenty-first vertebra of Struthio given by Mivart in figs. 47, 48 (loc. cit.). In this comparison, among the distinguishing characteristics of the Dinornithic vertebra, are, first, the greater relative height of the neural arch and spine, whereby the distance of the parapophysial, p, from the prezygapophysial, z, surfaces is relatively greater. The diapophysis, d, is less ex- tended lengthwise, but more produced transversely and vertically ; it also terminates with a tuberosity which might be reckoned as a low metapophysis, fig. 28, m, overtopping the small articulation, d, which now looks outwardly for the attachment of the rib’s tubercle. A reticulo-pneumatic fossa (72, fig. 27) intervenes, as in Struthio, between the par- and diapophysis. A larger pneumatic foramen (ib. pn) opens behind the diapophysis (d). The neural spine (7s) retains its characteristically greater height and thickness, with minor relative antero-posterior extent than in Struthio. Its fore and hind surfaces are occupied by well-defined rough tracts for the elastic ligaments (fig. 28, /). The preaxial surface (ib. ac) less transversely and more yertically extended than in the antecedent vertebra, retains something of its bilobed character by the emargination of its upper and lower articular borders. The parapophysis (p) projects with its rib- surface distinct from the articular facet (ac) of the centrum. The postaxial surface shows an unsymmetrical form in two examples of this vertebra, encroaching further upon the left side of the centrum in one, and upon the right side in the other. Both are individual varieties. The hypapophysis (hy, fig. 29), reduced vertically, is coextensive with the under surface of the centrum, and slightly produced and expanded at both ends (fig. 27, hy, hy’). In the direct view from beneath (fig. 29) the more advanced position of the diapophyses (m) and the less produced hind part of the neural platform and its postzygapophysial sur- faces (pz) are well shown, in contrast with Mivart’s figure 48, in Struthio (Joc. cit. p. 413). The hemapophysis of the fourth dorsal did not, in the specimen under description, articulate directly with the costal border of the sternum, but through the medium of the hemapophysis of the third dorsal. In the fifth dorsal (twentieth) vertebra the fore and hind productions of the hemal keel of the centrum assume the character of distinct hypapophyses, of which the anterior (iy) is unciform, being produced forward with the end upcurved, so as to receive the tuberous hind part of the hypapophysial ridge of the fourth dorsal into its concavity ; the posterior one (fy’) is low and simple. Figure 30 is a side view of the centrum of this vertebra, showing this singular and, as far as I have observed, unique development of pre- and post-hypapophyses. The parapophysial rib-cup (p) is rather larger, and the neural spine has greater fore- and-aft breadth than in the preceding (fourth dorsal) vertebra. This spine greatly exceeds in both height and thickness that in Struthio. A pair of depressions, answering 168 PROFESSOR OWEN ON THE GENUS DINORNIS. TWENTIETH VERTEBRA (‘5th dorsal,’ 3 nat. size). Fig. 30. Fig. 30, lateral aspect. TWENTY-FIRST VERTEBRA (‘ 6th dorsal,’ 3 nat. size). Fig. 31. Fig. 32. ig. 31, lateral; 32, preaxial aspect. PROFESSOR OWEN ON THE GENUS DINORNIS. 169 to those marked f1 in Mivart’s figure 57, of the last dorsal vertebra, are here well marked. In the sixth dorsal (twenty-first vertebra, figs. 31, 32) the hypapophysis is sup- pressed, as in the twenty-second (fifth dorsal) vertebra of Struthio (‘ Mivart,’ figs. 49-51)}, with which the present will be compared. The articular surface (fig. 32, ac) thus almost ‘‘ entirely occupies the preaxial end of the centrum” ?: only a few lines breadth on each side of the neural half of that surface is non-articular in Dinornis, and may be ascribed to the fore part of the parapophysis (ib. p). The vertical as compared with the transverse diameter of the preaxial surface is greater than in Struthio. The cha- racteristic height of the neural spine in Dinornis (figs. 31, 32, ns) is still more marked in this comparison. ‘The pneumatic orifice (fig. 31, pn) between the par- and di-apo- physes is the chief one for admission of air into the vertebral substance; but a small homologue (ib. pn’) of the posterior pneumatic orifice remains. The postaxial surface (fig. 31, pc) is absolutely and much more relatively approxi- mated to the postzygapophysis (ib. pz) than in Struthio. The neural canal (fig. 32, 2) is transversely, not vertically, elliptical (comp. Mivart’s fig. 51). The sides of the preaxial surface are much produced, and the transverse concavity of that surface is proportionally deepened. The lower border of the postaxial surface is more produced than in Struthio, rendering the lower contour of the centrum in Dinornis more concave (comp. fig. 31 with fig. 49, Mivart, loc. cit.). The zygapophysial surfaces are relatively more ex- tensive in Dinornis, the dorsal vertebre being more securely interlocked in the larger terrestrial bird. The characteristically broad and massive proportions of these vertebre in Dinornis are well brought out in comparing figs. 27-32 with figs. 47-51 of Mivart, loc. cit. The minor length and greater thickness of the diapophyses, d, and the much greater development of the neural spine are exemplified in fig. 32 as contrasted with fig. 51 (Mivart, loc. cit.). The vertebra in Dinornis, which answers, in rib-character, to that in Struthio sup- porting the eighth pair of movable pleurapophyses, is that which supports the seventh pair. In both genera it is the hindmost rib-vertebra not confluent with the sacrum. In the present skeleton of Dinornis it is the twenty-second vertebra, counting from the occiput ; in Struthio it is the twenty-fifth. Of this Prof. Mivart gives four figures *. The chief differential characters of its homologue in Dinornis maximus will be exem- plified in the two subjoined cuts from the lateral (fig. 33) and postaxial (fig. 54) aspects. In the comparison of figure 33 with figure 54 (‘ Mivart’), the deep longitudinal concavity of the under surface (¢c) of the centrum may be first remarked, due in Dinornis to a downward production of the border of the preaxial articular surface (ac) and a still greater production in the same direction of the postaxial surface (pc), augmented by the development of a pair of hypapophyses (fig. 34, hy). These are not developed in ? Loc. cit. p. 414, > Ib. p. 414, 3 Loe. cit. p. 419, figs. 54-57, 170 PROFESSOR OWEN ON THE GENUS DINORNIS. TWENTY-SECOND VERTEBRA (‘7th dorsal,’ } nat. size). Fig. 33. Fig. 34. Aspects. Fig. 33, lateral; 34, postaxial. Struthio. Rudiments may be indicated by the letters hy in Mivart’s fig. 54; but they are not noticed in the text. In the transverse direction the hemal surface of the centrum is convex, as in Struthio. The preaxial surface in Dinornis is subquadrate, through the production of its inferior angles, which, like the upper ones, are rounded off. Its transverse concavity is less than in the sixth dorsal: it is relatively larger in proportion to the centrum than PROFESSOR OWEN ON THE GENUS DINORNIS. 171 in that vertebra. The postaxial surface is more definitely subquadrate, with the angles rounded off and the upper and lower borders emarginate (fig. 34, pe). The transverse convexity is not greater than the vertical concavity; both are feeble, so that the entire surface approaches to flatness; and in a duplicate homologue the flattening is greater than in the specimen figured. In both the surface has lost its synovial smoothness, through suppression of motion upon the first sacral vertebra. ‘The trans- verse dimension does not exceed, as it does in Struthio, the vertical one. The neural canal (fig. 34, m) is more depressed than in the twenty-first vertebra, and still more deviates from the form shown by the hinder outlet in Mivart’s figure 56 of the Ostrich. The parapophysis in Dinornis (fig. 35, p) is represented merely by the raised margin of the capitular concavity. The diapophysis is less massive in proportion to the rest of the vertebra, and especially the neural spine, than in the antecedent dorsal. The neural spine is not carinate along either the fore or the hind border; both present a flat rough surface, about two thirds the breadth of each smooth lateral surface. A transverse section of the spine thus gives an oblong quadrate figure. «Archetype and Homologies of the Vertebrate Skeleton,’ 8vo, 1848, p. 159, fig. 27. % * Descriptive Catalogue of the Osteological Series contained in the Museum of the Royal College of Surgeons of England,’ 4to, 1853, p. 266, no. 1885. * «The Museum of the College of Surgeons fortunately possesses a preparation of the saeral vertebre (figs. 58, 59, 60, and 61) of a young Ostrich in an unanchylosed condition, which enables the serial description of PROFESSOR OWEN ON THE GENUS DINORNIS. (‘dorsal’ or upper) view (fig. 61, loc. cit.). The modifications of sacral structure here ex- hibited, which have proved most instructive in their application to the vertebre of extinct animals, are the alternate disposition of certain centrums and neural arches and of a few other centrums and pleurapophyses. In Mammalia such disposition of the heads of ribs across the articular intervals of the centrums is the rule in dorsal vertebre, and a like disposition of the neural arches occurs in the dorsal vertebre of Chelonia; but the concurrence of the alter- nating positions of centrums with both ele- ments appears not to have been observed in the sacral region of any vertebrate until the task of determining the singular detached cen- trums in the /guanodon and other large extinct Reptilia led me to a series of researches into the sacral structures and their development in existing Vertebrates. These researches led, among other results, to the detection, in the long sacrum of birds, of “a shifting of the neural arch from the middle of the body to the interspace of two adjoining centrums, each neural arch being there supported by two con- tiguous vertebre, the interspace of which is opposite the middle of the base of the arch above, and the nervous foramen is opposite the middle of each centrum” !. By this modifica- tion, “that part of the spine subject to greatest pressure is more securely locked to- gether;” and I further remarked that, “this structure is beautifully exemplified in the sacrum of the young Ostrich” 2. The detached centrums of such vertebra ) 175 D. crassus. D. elephantopus. D. robustus. D. maximus. D. grganteus. D. rheides. yielded the key to the characters of the Soom tie modiieationa cf instars num in the genus Dinornis. individual vertebra to be completed” (Zool. Trans. viii. p.420). Probably my preparation, No. 1885, may be here alluded to. * «Reports of the British Association,’ 8yo, 1841, “On British Fossil Reptiles,” p. 106. 2elipaaips 176 PROFESSOR OWEN ON THE GENUS DINORNIS. Dinosaurian vertebre figured in plates xii. et seg. of my ‘History of British Fossil Reptiles’1. Every subsequent discovery of a true Dinosaur has confirmed the appli- cability and value of this character of the extinct order. The sacrum in Dinornis I continue to characterize, as in other birds, by the anchy- losis of the vertebra, through which that single mass of the ‘spinal column’ results. The concomitant anchylosis of the iliac, ischial, and pubic bones constitutes the ‘pelvis. In birds ‘‘ anchylosis converts a large proportion of the vertebral column into a sacrum” 2. When it is said, and legitimately in its taxonomic application, that “ the Cetacea have no sacrum,” it is to be understood that vertebree homologous with the sacrals of the bird have not coalesced: when a mammal is said to have but two “ sacral vertebre,” the homologues of two of the sacral vertebree of the bird have coalesced. In Dinornis maximus seventeen vertebre have so coalesced, and include the homologues of vertebrae which in mammals retain their primitive freedom, and may be characterized as ‘dorsal,’ ‘lumbar,’ and ‘ caudal.’ When it is said that birds have no lumbar vertebre, a similar remark applies to that which has been offered respecting the absence of sacral vertebre in the Cetacea. The first or foremost sacral vertebra in Dinornis maximus (twenty-third of the entire series) offers to its pleurapophysis, which retains its mobility, a parapophysial cup near the upper and fore part of the centrum, and a small rough facet on its diapophysis. The pleurapophyses of the second and third sacrals are anchylosed each to its parapo- physis, and thence, by a bony plate, continued from the upper part of the ‘cervix’ to the lower part of the diapophysis. These two last ribs are progressively shortened, but still project beyond the iliac roof. Their more reduced serial homologues form the transverse osseous bars abutting against the outswelling antacetabular part of the ilium, with which the pubis has coalesced (Trans. Zool. Soc. vol. iii. pl. xix. fig. 2). After the second sacral centrum that element, in succeeding vertebre, quickly loses length. In the section of the sacrum of Dinornis maximus (Pl. XXXI. fig. 2, 4 nat. size) the interval between the first (1) and second nerve-outlets, which is 2 inches 7 lines, is reduced to 6 lines between the sixth (6) and seventh outlets, and to a less extent in the four succeeding outlets. Between the twelfth (12) and thirteenth outlets the interval is 9 lines, and it increases to 1 inch 6 lines between the fifteenth and sixteenth (16) outlets. . Of these nerve-outlets, the separation of the motory from the sensory division is well marked at their commencement from the neural canal in the fourth sacral, and so continues to the twelfth. In the thirteenth the size of the outlet is much reduced: from this part the myelon is restricted to the supply of the terminal contracted part of the spinal column called ‘tail;’ and here we have the sign of the beginning of the 1 Quarto, part vi. (1855). * «Anat. of Vertebrates,’ S8yo, 1866, 11. p. 29. The homologies of certain of these with the vertebrae called ‘ dorsal,’ ‘lumbar,’ and ‘caudal,’ in other vertebrate classes, are given in my ‘ Archetype,’ &., Syo, 1848, p- 90 et seq. PROFESSOR OWEN ON THE GENUS DINORNIS. 177 caudal series of vertebra, of which five have been enlisted or conscripted into the service of the sacrum. The sacro-neural canal (Plate XXXI. fig. 2) retains a vertical diameter of 8 lines along the first three vertebre; it then expands gradually to the sixth and rapidly to the ninth sacral, where the vertical diameter reaches to 1 inch 6 lines. The anterior or ‘hemal’ myelonal columns would seem to have made a bulge between the eighth and tenth vertebre; and the neural canal, again contracting, shows its diameter of 8 lines between the thirteenth and fourteenth outlets, and is reduced to 4 lines in the last or seventeenth sacral vertebra. The motory and sensory divisions of the nerve-outlets continue distinct to the outer surface of the vertebra as far as the twelfth, the neural (dorsal, upper) or sensory division being the smallest, and diminishing more rapidly than the hemal (lower, motory) outlet after the ninth of these. The canal gains in transverse as in vertical expanse, but in a rather less degree. Parapophysial abutments cease after the eighth sacral, and are resumed at the twelfth. The diapophysial ones increase in length from the fifth sacral, but with much diminished breadth, to the ninth sacral, when they increase in breadth as well as length, and curve upward, backward, and slightly outward to buttress up the expanded postacetabular part of the ilium. Between the smooth compact inner layer of bone forming the neural canal and the somewhat thicker outer layer, the osseous substance of the sacrum is coarsely reticulate and pneumatic. Larger subserial vacuities mark, in vertical section (Plate XXXI. fig. 2), some of the anterior obliterated vertebral interspaces; and the longest or chief lamine, rising from the roof of the neural canal, indicate the neural spines at distances corresponding to the nerve-outlets, answering to the fourth sacral vertebra. The spine-plate curves gently forward; while those of the sixth, seventh, and eighth sacrals rise vertically, and the succeeding ones curve gently backward. In the comparison of the sacrum of Dinornis, as exemplified by the present spines, with that of Struthio, as illustrated in Prof. Mivart’s paper1, I may premise that the first three (anchylosed) vertebre are reckoned, by its author, as ‘dorso-lumbar’ (26th and 27th) and lumbar (28th) vertebre. It will be understood, therefore, that in my description of the specimen “in the Museum of the College of Surgeons” ’*, figured in Mivart’s cut 59, “the neural arch of the fifth sacral vertebra has advanced, and rests over the interspace between its own and the preceding centrum; at the eleventh vertebra it has resumed its normal position and connexions.” My ‘fifth sacral’ is Mivart’s ‘second’ (s 2), and my ‘eleventh’ sacral is Mivart’s eighth; the last five sacrals in the twenty anchylosed vertebre of the mature Ostrich * are reckoned by him as the first five caudals in that bird. 1 Loe. cit. pp. 420-427, figs. 58-62. 2 « Catalogue’ ut supra, 4to, 1853, p. 266. 3 Trans. Zool. Soc. vol. iii. pl. 19. fig. 4. 178 PROFESSOR OWEN ON THE GENUS DINORNIS. In Dinornis the twenty-fourth vertebra (3rd sacral), answering to the twenty-eighth vertebra of Struthio (Mivart’s ‘ lumbar vertebra’), differs in presenting an unmistakably rib-like pleurapophysis, although unanchylosed. The fourth sacral in Dinornis is the first which may be said to “ present no indication of a rib,” and which would be entitled to the term “lumbar,” according to such character. I view, however, the parapophysial element of this transverse process as more probably the serial homologue of the cervical part of the preceding pleurapophysis. With this explanation the neural arch of the fifth sacral vertebra in Dinornis, as in Struthio, advances and crosses the interspace between its own and the preceding cen- trum; and the thirteenth vertebra is that in which the arch resumes its normal con- nexions. Thus the interlocked part of the sacrum in Dinornis is more extensive than in Struthio, and relates to the heavier mass which the pelvis had to transmit upon the femora. The antacetabular part of the sacrum (lst to 6th vertebra in Dinornis) is relatively shorter and broader than in Struthio; the postacetabular part is still broader in pro- portion to its length; and this part is shorter than the antacetabular part, instead of being, as in Struthio, longer. More striking differences are presented by the pelvis as a whole. The antacetabular part of the ischium is relatively longer; the postacetabular part is shorter, but much broader in Dinornis than in Struthio: the greater relative breadth of the entire pelvis would seem to relate to the larger proportional size of the egg in Dinornis. The ischium is shorter and deeper than in Struthio: it unites with the ilium anteriorly to bound there the ischiadic notch, which remains open posteriorly, as in Struthio, and is not circumscribed by a second terminal union of the ischium with the ilium, as in Dromaius. The obturator interspace, closed behind, as in Struthio, by ischial confluence with the pubis, and having its fore part defined by the descending process of the ischium, is much narrower in Dinornis, as in Apteryx. The pubis does not send off so long and well-defined a ‘ pectineal process’, as in Struthio; its body extends backward parallel with the ischium, slightly concave downward, and ter- minates in the vertical expansion joining the ischium without being continued down- ward and forward to meet its fellow at the symphysis, a structure which is peculiar, among birds, to the genus Struthio. The type of the pelvis in Dinornis is that of the Apteryx, not of the Emu or Cas- sowary; it differs therefrom in less marked modifications than from the pelvis in Struthio and Rhea. The number of terminal sacral vertebre in Dinornis maximus, answering to those defined as ‘ sacro-caudals’*, is four. The last of these in Dinornis is the thirty-ninth of the vertebral series; in Struthio it is the forty-sixth. F * For this process in Apterya australis, see Trans. Zool. Soc. vol. ii. p. 291. In the two skeletons of the smaller Kivi (Apteryx Owenii, Gd.) I have found ossification extending along the ligament attaching the pectineal process to the last sacral rib. * Mivart, ut supra, p. 426, fig. 62. PROFESSOR OWEN ON THE GENUS DINORNIS. 179 In Apteryx australis there are nine caudal vertebre, the anterior ones of greater relative vertical extent than in Struthionide ; but as they recede they gain in transverse and lose in vertical diameter. The last two coalesce to form the ‘ ploughshare’ bone. FORTY-FIRST, or SECOND CAUDAL VERTEBRA (+ nat. size). Fig. 36. Fig. 37. Aspects. Fig. 36, preaxial; 37, lateral. Iam not certain that I possess the fortieth vertebra or ‘ first free caudal’ in Dinornis. The second, if it be not the first (figs. 36 & 57), has the centrum broader in proportion to its length and height than in Struthio. The contour of the preaxial surface (ac) is subhexagonal, with the upper line short and emarginate, forming the lower boundary of the neural canal (7). The surface of ac is irregular, indicative of syndesmotic union with the sacrum (or first caudal), deviating on the whole from flatness by a slight convexity: the opposite articular surface is undulate, slightly concave at the middle third, convex to the periphery: the angles of the hexagon are rounded off. The under surface is lon- gitudinally concave, a mid channel being bounded by a pair of longitudinal ridges. A thick, short, obtuse, subbifid parapophysial ridge (fig. 37, p) extends from the middle of the antero-lateral part of the centrum obliquely backward to near the upper and outer angle of the hinder articular surface. ‘The neural canal (fig. 36, m) is small and subcircular ; in Struthio its section gives a vertical ellipse. The diapophysis is repre- sented by the upper division (fig. 36, d) of the tuberous diparapophysis. In Séruthio the diapophysis is a distinct process from the parapophysis, and is the longer and larger of the two. The neural canal in Struthio is surmounted by a thick subquadrate mass with its enlarged tuberous extremity subbifid posteriorly. In Dinornis the character of the double neural spine, which distinguishes, in the present comparison, several of the neck-vertebra, is resumed in those of the tail. A pair of low, thick, short, tuberous processes (fig. 36, ms) diverge from the roof of the neural canal and simulate a ‘ spina bifida.’ This character is continued through the caudal series to the foremost of the three vertebre (figs. 38, 39), which coalesce to form the homologue of the terminal ‘os VOL. X.—PART 1. No. 7.—October 1st, 1877. 2¢ 180 PROFESSOR OWEN ON THE GENUS DINORNMIS. en charrue,’ or ‘ ploughshare bone,’ in most other birds, and in all those that fly and possess the ‘ rectrices’ or ‘rudder-feathers,’ as the tail-quills are termed. In the description of the skeleton of Dinornis elephantopus! nine caudal vertebre are noted, as in the Apterya, reckoning the terminal bone as one of the series, and its leading distinction of shape from that in birds of flight is pointed out. The character of the bifid neural spine is indicated as “‘ a pair of tubercles supported by a low trans- »2 versely extended neural arch” ’. FORTY-SIXTH, FORTY-SEVENTH, and FORTY-EIGHTH VERTEBRA, or Terminal Caudals (3 nat. size). Fig. 38. Fig. 39. Aspects. Fig. 38, lateral ; 39, hemal. In a species of Dinornis, which Capt. Hutton thinks may be D. crassus*, the last three caudal vertebre coalesce into the ‘ ploughshare bone’ (figs. 38 & 39); but this, as in D. elephantopus, D. maximus, and doubtless in the rest of the genus, has no claim to the shape, common in birds, which suggested the vernacular name +. The neural spine is suppressed in the last two of these caudals (fig. 38, 8, 9), which are reduced to the central element with, perhaps, a neural ridge imperforate ; and this ridge forms the uppermost of the three ridges which characterize the three-sided cone constituted by these two terminal vertebra. Of the three sides the lower is the broadest (fig. 39, 9). In the penultimate (8th caudal) vertebra the lower surface (ib. 8) presents a tri- angular excavation, the base being turned forward and the sides formed by the last rudiments of parapophyses (ib. s, 9, p, p); the apex of the cavity extends to the anchylosis with the last vertebra. The sides of both vertebree are subconcave, the centrum expanding at both ends. The quasi-parapophysial expansions of the fore end of the last centrum (ib. 9, p) extend beyond the hinder expansions of the penultimate vertebra. The centrum of the last vertebra contracts to an obtuse point, grooved below. 1 Zool. Trans. vol.iv. p. 163. > Ib. * « The box also contains a complete set of caudal vertebra of D. crassus (?) from Shag Point: these are from one bird.” Letter dated “ Dunedin, N.Z., 13th Dec. 1875.” These yertebre were six in number, reckoning the soldered three as one. I doubt their including the entire series. 4 < Os en charrue,’ Fr. PROFESSOR OWEN ON THE GENUS DINORNIS. 181 The antepenultimate vertebra retains its neural arch, canal, and bifid spine (fig. 38, 7, 2), differing only in size and the stunted character of the processes from the antecedent free caudals. The lower and fore part of the sides of the centrum extend as short, broad, obtuse parapophyses. These render the under surface of the centrum transversely concave. A similar low obtuse diapophysis projects from the base of the neural arch and con- tiguous part of the centrum with which it has coalesced. The gradual diminution of the terminal vertebre of the tail to an obtuse point indicates that such an appendage was as little indicated by the plumage in Dinornis as in Apterya. In Struthio the transverse processes of the caudal vertebre have assumed, in the third of the series (fig. 65, p. 429 of ‘ Mivart’), much of the coalesced characters of the first in Dinornis (fig. 36). ‘The antero-posteriorly compressed and transversely extended mass representing the neural spine begins to shoot out its upper angles in the third caudal of Struthio, and in the sixth (fig. 66, p. 430 of ‘ Mivart’) they more nearly repeat the parial divergent spines in Dinornis (fig. 36, ns). In the eighth caudal of Struthio (fig. 67, p. 430 of ‘ Mivart’) a third low spine rises between them. The ninth caudal in Struthio (fig. 68, ib.), which is commonly found anchylosed at the neural and hemal borders of its postaxial surface with the terminal ‘ ploughshare,’ is the homologue of the foremost of the three terminal coalesced caudals in Dinornis (fig. 38,7); but it has lost its transverse processes, and a terminally trifid lofty neural arch and spine represent the low arch and pair of tuberous neural spines in Dinornis. Prof. Mivart! rightly notices the indications of the two terminal vertebra which have coalesced to form the vertically extended laterally compressed plate of bone, with its irregular more or less rounded margin, so markedly distinguishing the termination of the vertebral column in S¢ruthio from that in Dinornis. The retention of the ploughshare character in the Ostrich relates to the large size of the feathers which it supports, and which represent the ‘ rudder-quills’ (‘rectrices’) of normal birds of flight. Such caudal plumes, with the similar alar plumes, the better developed bones of the unavailable pair of wings and concomitantly developed sternum and scapular arch, concur in showing that the great existing flightless bird of Africa has receded in a less degree from the volant type than have the extinct wingless birds of New Zealand. The terminal segments of the ‘ axial skeleton’ in Dinornis differ from those in Apteryx mainly in the minor modifications of the elements and apophyses constituting the palate and beak. As these segments are omitted in Prof. Mivart’s analysis of the axial skeleton of the Ostrich, the comparison of the individual vertebre in advance of the atlas will not be here entered upon. The skull of Dinornis maximus differs chiefly in size from that of D. robustus and D. ingens. It presents the same type of beak and mouth-bones, the same low broad " Loe. cit. p. 481, fig. 69. 182 PROFESSOR OWEN ON THE GENUS DINORNIS. form of cranium. In the smaller species of Dinornis, through the minor difference in the size of brain, its case is in them relatively more convex and raised, a character which is most marked in the comparatively diminutive Apteryr. As the parts furthest from the centre are most subject to modifying influences, the bony framework of the beak, of which the palate forms part, departs in Apteryx still further than the cranium from the character of the skull in Dinornis. The palatal generic characters of Apteryx are detailed in the second volume of the Zoological Transactions, 4to, p. 285, and illustrated in plate vii. fig. 2. The repeated pressure to which the beak is subject in perforating the soil for food being transferred to the hind buttress-bone formed by the tympanics, all the beak-bones articulated therewith have coalesced—the maxillaries laterally with the malo-squamosal styles, and mesially with the palatines, these carrying on the coalescence with the yomer and pterygoids; so that the upper beak, as a single bone, articulates with the tympanics by the diverging columns of its quadrifid base, the two outer and more slender ones with the outer cups, the two inner and thicker ones with the imner cups, the latter being strongly wedged, moreover, before reaching those latter cups, between the orbital plates of the tympanic and the pterapophyses or ‘ transverse processes of the sphenoid.’ The advantage of a certain yielding movement of the tympanics under extreme pressure cannot fail to be noticed. As the dinornithic modifications of the palate are more perfectly demonstrated in the skull of a Dinornis crassus, recently transmitted to me, than in that of D. maaimus, I subjoin a figure of the base of the skull in the smaller Moa (Plate XXXI. fig. 1) ; it closely repeats the characters shown in D. ingens }. In Dinornis crassus the palatal plates of the palatines (ib. 20) are anterior horizontal expansions of those bones which coalesce with the corresponding palatal plates of the maxillaries, not passing beneath them. ‘The bony palate behind the premaxillary part of the mouth-roof may be truly termed the maxillo-palatine part of that roof, including parts of both bones. In the skull of Dinornis ingens, figured in pl. xv. fig. 5 of my fourteenth part, a portion of the suture still remains, and a smaller portion is traceable in the present specimen of D. crassus. Birds have the maxillary, 21, and palatine bones, 20, ossified, as in mammals, from separate centres, but have no maxillo-palatine bone, save by the accident of partial confluence. ‘The specific palatal distinction from D. ingens appears in the course of the suture of the maxillo-palatine plate with the premaxillary. In D. ingens the suture runs across in an irregular wavy line; in J. crassus it presents an angular form, the maxillo-palatine plate being notched to receive the angular palatine process, 22’, of the premaxillary. In D. crassus, as in D. ingens and D. maximus, the slender part of the palatine, con- tinued backward from the palato-maxillary plate, is twisted so as to bring the inner edge * Trans. Zool. Soc. vol. vii. plate xy. fig. 3. 2 Ib. p. 123. PROFESSOR OWEN ON THE GENUS DINORNIS. 183 of that plate upward, and to turn the horizontal under surface, 20, into a vertical outer surface of the bone, which rapidly gains in depth, and has its upper part bent inward, to complete with the vomer, 13, the hind wall of the palato-narial canal. At the outer and back part of the canal the palatine is thickened at its lower part to arti- culate with the pterygoid, 24. The vomer is bifid, as in D. ingens and as in the first-described skull of D. crassus (tom. cit. pl. xi. fig. 3, 13). The parial plates of the vomer overlap the sides of the presphenoids, 9, of which the anterior apex, 9'. coalesced with the narial septum, projects beyond the vomer, and partially divides the prepalatine vacuity. ‘The anterior ends of the halves are overlapped by the vomerine processes of the premaxillaries. Each half of the vomer consists of a deep vertical bony plate, almost meeting below the presphenoidal rostrum, expanding at both ends anteriorly to join the premaxillary and the palato-maxillary plates, and there bounding the palato-nares anteriorly; pos- teriorly expanding in a greater degree, and curving outward and forward to join the palatines, and form the posterior boundaries of the palato-nares. These apertures are each | inch 7 lines in length, 53 lines in breadth ; the breadth across both apertures is 1 inch 114 lines, the additional half line giving the interval between the halves of the vomer. ; The suture between the vomer and palatine, as one looks down upon the skull’s base, runs along the bottom of the vomero-palatine or postnarial fossa, along a shallow channel there ; it seems obliterated near the postero-external rather thickened border of that fossa. From this border the pterygoid process of the palatine is divided by a triangular shallow depression. The pterygoid is short, three-sided, with the sharp angle between the inner and outer facets of the under surtace of the bone turned down- ward, and continuing backward a similar ridge on the under part of the palatine. The pterygoid has an extent of articulation with the tympanic of more than half an inch in D. maximus. In this, as in smaller species of Dinornis, well-ossified sesamoid bones added to the leverage of the muscles of the foot by their interposition at the back part of both the proximal and distal joints of the metatarsal segment. ‘The tibio-tarsal sesamoid (Plate XXXI. figs. 3-6) works upon the shallow rounded surface at the back part of the ectocondylar fossa of the metatarsus. It is an elongate, trihedral, conical bone, with a slight sigmoid flexure. The end representing the base (fig. 3) is external or looks fibulad ; the opposite or inner end, or apex, is obtuse. The base is triangular, almost flat, with rough ligamentous marks. The two articular sides are half concave half convex length- wise; on the narrower side the convexity is next the base, and the reverse on the broader side. The non-articular side is almost flat and has faint linear and irregular impressions. In one specimen the ligamentous or tendinous attachments to the base were partially ossified. Here I propose to conclude the task of restoration of Dinornis maximus. The liberal 184 PROFESSOR OWEN ON THE GENUS DINORNIS. illustrations, natural size, of previous descriptions of the bones of the hind limb, in the ‘ Transactions of the Zoological Society,’ vol. vi., preclude the necessity of further trespass. The femora, tibie, and metatarsi in the skeleton of Dinornis maximus in the Canterbury Museum, and in that of the British Museum, closely accord with the first received detached bones indicative of the species. The side view of this skeleton prefixed by Dr. Buller to his excellent work on the existing Birds of New Zealand precludes the necessity of repetition, as my drawing differs in little else than showing the trunk of the bird at an angle with the hind limbs which somewhat diminishes the height of this largest of known birds as compared with that indicated by the figure of the Maori Chief which Dr. Buller has introduced into his Plate. The height of the skeleton of Dinornis maximus, as articulated in an easy standing position, in the British Museum, is 11 feet; the length of the trunk (dorsal and sacral series of vertebree) is 4 feet 4 inches; the length of the hind limb, in the same position, following the angle of the segments, is 9 feet; the total length of the skeleton, from the point of the beak to the end of the tail, following the curves of the spine, is 11 feet 4 inches. Reviewing the osteological facts in the present and preceding memoirs on Dinornis, the following characters seem to be common to the genus :— 1. Skull with a rather short, broad, moderately arched bill, not attaining the height of the cranium ; occipital condyle not projecting so far back as the upper border of the occipital foramen. 2. Horizontal palatal plates of the palatines and maxillaries more or less confluent, not uniting solidly, but suturally, with the premaxillary and the vomer. 3. An Apterygian, not Dromeine, pelvis. 4. A short, broad sternum, with small, ill-defined coracoid pits, and with three posterior notches. 5. Scapula and coracoid small and feeble, forming no angle, not developing a glenoid cavity at their bony confluence. 6. Four toes; the hallux small and high-placed. 7. Terminal confluent caudals of less vertical extent than the antecedent free caudals. In every example of associated or connected parts of a foot the small back toe has been found. Its absence in the earlier transmitted foot-bones I have since had reason to regard as accidental; and should so small and seemingly functionless a toe have been the subject of congenital defect, the true generic characters of Dinornis will be given by the species demonstrating the Apterygian structure of the foot rather than by a propagable variety exemplifying the ‘ monstrum per defectum.’ As the present may, probably, be the last of the series of papers “On Dinornis” which I have to communicate to ‘ The Transactions of the Zoological Society,’ I append to it illustrations of the most authentic evidences of the plumage of the extinct species PROFESSOR OWEN ON THE GENUS DINORNIS. 185 which have come under my observation. The subjects of figures 8, 9,10, in Plate XX XI. are copied from the illustrations of Mr. Dallas’s excellent description of the débris of feathers attached to the dried skin of the specimen of Dinornis robustus in the Museum of the Yorkshire Philosophical Society!. Figures 7 and 11 are from Dr. Hector’s and Captain Hutton’s interesting papers in vol. iv. of the ‘ Transactions of the New- Zealand Institute,’ (vol. iv. p. 114, pl. v., and p. 166, pl. ix.). With the loose character of the plumage of the Kivi the feathers of the Moa differ in having, as a rule, an ‘after-shaft’ (fig. 11, 6) of half the length or more of the main feather. As to the geological relations of the bones of the Moas, reviewing the whole evidence, I concur with the learned Professor Igino Cocchi? in referring Dinornis crassus, D. ele- phantopus, D. giganteus (var. robustus), and D. ingens to the “ Periodo attuale,” which is equivalent to the ‘neolithic’ or ‘recent period’ of ‘ Ethno-archeology.’ At the same time I think that certain remains from the fluviatile deposits in the North Island, representing the species Dinornis giganteus, D. ingens, D. struthioides, and D. didi- formis, of a heavier and less recent character than the bones from the South Island, have come from birds which lived in ‘ postpliocene,’ or quaternary, or even earlier times. But all the species seem to have existed and abounded when the present race of Maories set foot on New Zealand, and the final extirpation to have been of com- paratively recent date. DESCRIPTION OF THE PLATES. PLATE XXXI. Fig. 1. Base of the skull of Dinornis crassus. Fig. 2. Vertical longitudinal section of the sacrum of Dinornis maximus, 4 nat. size. Fig. 3. End view of metatarsal sesamoid of the same, 4 nat. size. Fig. 4. Inner side view of ditto, ditto. Fig. 5. Outer side view of ditto, ditto. Fig. 6. Tibial facet of ditto, ditto. Fig. 7. Portion of skin of neck and of feathers of Dinornis ingens (after Hector). Fig. 8. Base of a feather, with main and accessory shafts, of the same (after Dallas). Fig. 9. Base of feather, with main and accessory shafts, of Dinornis robustus, nat. size (after Dallas). Fig. 10. Part of a barb, with the barbules of the same, magn. 15 diameters (after Dallas). Fig. 11. A feather of a species of Dinornis (after Hutton). (The figures are of the natural size when not otherwise noted.) * Proceedings of the Zoological Society of London, March 14, 1865, p. 266. * *Cataloghi della collezione centrale Italiana di Paleontologia,’ Syo, Firenze, 1872, p. 63, 186 PROFESSOR OWEN ON THE GENUS DROMORNIS. PLATE XXXII. Outer or under view of sternum of Dinornis maximus. Inner or upper view of ditto. Costal border of the right side of ditto. Costal border of the left side of ditto. Anterior border of the sternum of ditto. (All the figures are of the natural size.) io! ie Rm de dad oR oe be APPENDIX. Additional Evidence of the Genus Dromornis in Australia. By Prof. Owen, C.B., F.RS., F.Z.S., &e. [Puate XXXITI.] Received January 25th, 1877. Read March 6th, 1877. A FEW weeks ago I was favoured with a letter from an esteemed correspondent in Australia,. the Rev. W. B. Clarke, F.G.S., dated “ Branthwaite, North Shore, New South Wales, 11th October, 1876,” inclosing a photograph of a portion of the pelvis of a huge bird, which bird had been found at a depth of from 150 feet to 200 feet in what is called the ‘ Canadian lead, “the bed-rock of which, at this place, is a paleozoic lime- stone, much waterworn and cavernous.” The locality is “in the county of Phillip, not far from Mudgee, on the road to Gulgong.” Mr. Clarke promises to send a cast of this specimen; but I will not delay to notice the discovery, because the photograph and the given dimensions show the specimen to have formed part of a bird’s pelvis as large as that of the Dinornis elephantopus; and I have received, from another and distant locality, a fossil bone which enables a more instructive and decisive comparison to be instituted between it and the corresponding part of the skeleton of the Dinornis most nearly corresponding with it in size. This bone was sent me from the province of South Australia, and was found in the ‘Mount Gambier range.” It is the lower portion, with the articular end a little mutilated, of a left tibia of a flightless bird (Plate XX XIII.), and corresponding in size with the same part in Dinornis elephantopus 1, and rather larger than that of Gastornis parisiensis?. As the modifications of the distal end of the tibia in birds are more salient and characteristic than those of the femur, the present specimen is valuable as a test of the conclusions drawn from the fossil femur of the large bird from Australian drift de- ’ Trans. Zool. Soc. vol. iv. pl. 43. fig. 4. 2 Quart. Journ. Geol. Soc., Aug. 1856, pl. iii. p. 204. Smit del.et lith Hanhart, imp DINORNIS CRASSUS. D.MAXIMUS “SOND dur: zo1g uzayuyy UIXXK PA-UGYG °F UE Pau CL Griesbach lth PROFESSOR OWEN ON THE GENUS DROMORNIS. 187 scribed, under the name of Dromornis australis, in volume viii. of the Transactions of the Zoological Society, p. 381, pls. 62, 63. In the first difference which I note in the Australian fossil tibia the bone resembles that of Gastornis and differs from that of Dinornis, viz. in the medial position of the ‘precondylar groove’ (Pl. XX XIII. fig. 1, p). In every species of Dinornis this groove is near the inner (tibial) margin of the fore part of the bone (see plate cit. note 1, and Trans. Zool. Soc. vol. viii. pl. 59. fig. 1, p, Dinornis gravis). In both Dinornis and Gastornis the groove is crossed by a bridge of bone. Of this bridge there is no trace in the present Australian fossil, and there is no evidence of fracture of the piers of such a bridge. The margins of the groove whence the bridge springs in Dinornis are, in Dromornis, broadly convex and entire. Dromaius and Casuarius? have the precondylar groove, but not the bridge. In both the groove is not medial, as in Dromornis, but is nearer the inner border of the tibia, less near, however, than in Dinornis. In Struthio there is neither groove nor bridge ; but in place of the groove there is a transverse rising of the bone. Apteryx offers a miniature resemblance to Dinornis in this tibial character. The distal expansion is relatively less, in comparison with the shaft of the tibia, in Dromornis than in Dinornis elephantopus (the species which Dromornis most resembles in the size of the shaft). The inner border of the distal end of the shaft (Pl. XX XIII. fig. 3, a@)is broader than in Dinornis, in which it contracts almost to a ridge as it passes to the beginning of the posterior production of the inner (tibial) condyle. InDromornis the corresponding part of the shaft, a, maintains a smooth transverse convexity to the condyle s. ‘The anterior production of the inner boundary of the rotular part of the intercondylar space (ib. fig. 2, 4) is more prominent in Dromornis than in Dinornis. The hind part of the inner condyle (ib. fig. 1,s) is less produced than in Dinornis ; and the corresponding part of the outer condyle, ¢, is less convex. There is no definite cavity below the precondylar groove for the antentocondylar® prominence of the metatarse. There are other minor differences; but the above-defined patent ones sufficiently establish the fact of a nearer resemblance in the tibia, as in the femur, of the gigantic wingless bird of Australia to the genera still there represented (Dromaius and Casua- rius), than to Dinornis, Apteryx, or Struthio. The following are comparative admeasurements :— Dinornis Struthio. fe § (elephantopus). Transverse breadth of the shaft of the tibia at in. lines. in. lines. in. lines. the commencement of the distalexpansion. 2 2 aS IBS) Ditto ditto distal condyles . 3 5 4 0 A OY Dromornis. The fossil above described is in a more mineralized condition, consequently of greater 1 Anat. of Vertebrates, ii. p. 78. 2 Osteol. Catal. Mus. Coll. Surg. 4to, 1853, vol. i. p. 250. 3 See Trans. Zool. Soc. vol. vii. fig. 3, c. VOL. X.—PART 111. No. 8.—October 1st, 1877. 2D 188 PROFESSOR OWEN ON THE GENUS DROMORNIS. specific gravity, than any bone of Dinornis which I have hitherto received. It is sup- posed to have come from a cave in Mount Gambier, South Australia; but I can only speak with certainty as to the locality, not as to the circumstances of its discovery. One cannot, of course, state confidently that it is a bone of the same species of bird as the mutilated femur from the Cave of Wellington Valley +, or of that from the drift at Peak Downs, in Queensland 2. But the relation of size to these bones, and the difference of proportion to the tibia of Dinornis exemplified in the above-given admeasurements, oppose no obstacle to the reference, rather support it, and bear out the inference hazarded in a former paper °. I believe, therefore, that Ornithology may confidently add another genus of gigantic birds to the unwinged group—a genus which existed and has become extinct in the Australian continent, and which had closer kinship with the still existing struthious genera of that continent than with the extinct Moas of New Zealand. DESCRIPTION OF THE PLATE. PLATE XXXIII Distal portion of tibia of Dromornis australis. Fig. 1. Back view. Fig. 2. Front view. Fig. 3. Inner side view. Fig. 4. Broken end of shaft, showing thickness of wall and size of medullary cavity. ? Mitchell’s ‘ Three Expeditions into the Interior of Eastern Australia,’ 8vo, ‘‘ Paleontological Appendix,” pl. 32. figs. 12, 13 (1838). ? Trans. Zool. Soe. vol. viii. p. 384 (1872). * «From the proportions of the femur of Dromornis I infer also that those of the tibia would be longer and more slender than those of Dinornis elephantopus.” From nat. on stone by JErxleben S T Als AUSTRE a DROMORNI (g¥BOi] V. On the Structure and Development of the Skull in Sharks and Skates. By W. XK. Parker, FBS. Received November 6th, 1876. Read November 7th, 1876. [Pirates XXXIV.—XLIT.] IntropucToRY REMARKS. ‘THROUGH the kindness of my friends Henry Lee, Esq., F.L.S., and F. M. Balfour, Esq., B A., I am able to give illustrations and descriptions of this type of skull in a degree of detail beyond what I had ever hoped to accomplish. Embryos and well-grown individuals of Dog-fish and Skates have been liberally supplied me by Mr. Henry Lee; these have been from the Brighton Aquarium; the embryos of Pristiurus and of Scyllium, kindly given to me by Mr. Balfour, were obtained from the Aquarium at Naples. As the youngest of these long, delicate, vermiform larve were not more than two thirds of an inch in length, the head forming a small knob, as it were, to a long and highly flexible staff, it may be imagined how minute these objects were for operating upon by dissecting-instruments. By zeal and patience the task has been mastered; these small heads have been made to disclose most important morphological secrets. If any biologist lives who is ready to deride this minute work, I would ask him to close his eyes to what is here shown him, and then give a full explanation of what is presented to him in the skull of an adult Skate or Shark. As the Osseous Fishes have undergone a large amount of metamorphosis or speciali- zation, as compared with these Selachians, more than is seen in the Ganoids, it seems natural that the Selachians should stand in a low place, zoologically. This, in many respects, would be, for them, a false position; for, embryonic as they are, in certain respects, in their adult state, yet they are, on the whole, a very high kind of Fish. They are sharply separated from all fishes, except the Marsipobranchii, by the non- related condition of their exo- and endoskeleton. Even in the Ganoids the exoskeletal scutes are brought under the influence of the endoskeleton in the head; and this inner framework draws, as it were, any and every convenient dermosteal patch into harmony with itself, enshielding itself with the enamelled pieces, which take on an outline that makes them apt for any such defensive service. vou. X.—Parr Iv. No. 1.—WMarch 1st, 1878. 25 190 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT It would seem therefore that where this morphological affinity has not exerted its force, the study of the uncombined endoskeleton would be an easy task. This is not the case; for there is another simple and peculiar character about the Selachian skeleton and skull, namely the absence of the ossified territories found in the cartilage; thus Nature has removed her usual landmarks. There is a calcareous deposit; but it is generally distributed over the cartilage as tessellated ‘“ superficial endostosis,” a mere calcification of regular groups of cells, these groups having no more morphological meaning than the “ placoid” grains that form the shagreen in which the fish is enclothed. In researches so manifestly difficult and slow as this it is well to do one thing only at a time, and that in one part mainly. It would be out of place here to enter into a discussion as to the general anatomy of the Selachians: what follows will show the value (or the writer is much mistaken) of the study of their cranial morphology. The present contribution to this part of biological science is, if the whole vertebrate series be considered, just simply the spelling-out of a word or two that may help, with the addition of many such essays and spellings, towards a true reading and interpreta- tion of the skull, and of its relation to the rest of the skeleton. But the great Selachian group may well have much labour spent upon it for its own sake; and an attempt is here made to unite and knit into one whole the labour of the gradationalist on one hand, and of the embryologist on the other. The works whose titles now follow are for the most part gradational. The labours of the embryologist are quoted also; but those at hand are by one young and talented worker, Mr. Balfour. The works and papers that have been of most value to me as containing descriptions of adult skulls of Ichthyopsida are the following, namely :— Jou. Mutter. “ Vergleichende Anatomie der Myxinoiden, oder Cyclostomen, mit durchbohrtem Gaumen. Erster Theil. Osteologie und Myologie.” Abh. Berl. Akad. 1835, p. 65 ; Wiegm. Arch, 1836, ii. p. 245. T. H. Huxtzry— 1. Elem. Comp. Anat. 1864, pp. 162-218. 2. “On the Representatives of the Malleus and Incus of the Mammalia in the other Verte- brates.” Proc. Zool. Soc. May 27, 1869. 3. “On the Structure of the Skull and of the Heart of Menobranchus lateralis.” Proc. Zool. Soc. March 17, 1874. 4, “Contributions to Morphology.—Ichthyopsida. No.1. On Ceratodus forsteri, with Obser- vations on the Classification of Fishes.””? Proc. Zool. Soc. Jan. 4, 1876. 5. “On the Nature of the Cranio-facial Apparatus of Petromyzon.” Journ. Anat. & Phys. vol. x. pp. 412-429, pl. 17, 18. Ramsay H. Traquair. “On the Cranial Osteology of Polypterus.” Journ, Anat. & Phys. vol. v. pp. 167-183, pl. 6. OF THE SKULL IN SHARKS AND SKATES. 191 Apert GontHer. “ Description of the Ceratodus, a Genus of Ganoid Fishes recently discovered in Rivers of Queensland, Australia.” Phil. Trans. 1871, part ii. pp. 511-571, pl. 80-42. Cart Gecenpavr. ‘ Untersuchungen zur Vergleichenden Anatomie der Wirbelthiere’ (part 3. Selachians). Leipzig, 1872. Then follow in natural order, the writer's own papers in the ‘ Philosophical Trans- actions’ on the structure and development of the skull in various types. The present paper is one of the same kind, but especially intermediate between M. Gegenbaur’s splendid work and the invaluable researches now to be noticed. F. M. Barrour— 2 1, “A Preliminary Account of the Development of the Elasmobranch Fishes.” Quart. Journ. of Mier. Se. Oct. 1874. 2. ‘The Development of Elasmobranch Fishes.”’ Journ. of Anat. & Phys. vol. x. pp. 517-570, pl. 21-26. 3. “The Development of Elasmobranch Fishes.” Ibid. pp. 677-688, pl. 29. 4. ‘A Comparison of the Early Stages in the Development of Vertebrates: Studies from the Phy- siological Laboratory in the University of Cambridge.’ Part 11. 1876, pp. 1-20, pl. 1. Although out of order, I must mention two important papers recently sent me by the author. Burt G. WitpeEr. 1. “Notes on the North-American Ganoids Amia, Lepidosteus, Acipenser, and Polyodon.” Proc. Am. Assoc. for Adv. of Se.: Detroit Meeting, Aug. 1875. Salem, Mass., 1876, pp. 151-194, pl. 1-3. 2. “Note on the Development and Homologies of the Anterior Brain-mass of Sharks and Skates.” Am. Journ. of Se. & Arts, vol. xii. Aug. 1876}. I mention these last papers because of the necessity of studying the nervous and skeletal systems together, and also because of the intimate relation of the Selachians with the Ganoids. In the following description allusion will be made to the condition of the stages of the skull in other types, especially in the Ichthyopsida. But the intimate relation of the skull of the Amphibia with that of the Selachians is best seen in the outlying forms of the latter group, namely Cestracion, Notidanus (Heptanchus, and Hexanchus), as these come nearest to the Chimeroids on one hand, and to the karval Batrachia and Urodela on the other. Professor Huxley’s paper on Menobranchus, a Perennibranchiate Urodele, has been of the utmost service to me; and by the time the present communication is published, I hope to have also in print my second paper on the Batrachian skull, and my first on that of the Urodeles. These, in addition to remarks on the growth and changes of the 1 Several fresh papers by my friends Balfour and Wilder have reached me since the above list was written. 252 192 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT skull in the Common Frog’, will show what is seen in Bufo vulgaris, Dactylethra capensis, Pipa monstrosa, Siredon (with Amblystoma), Seironota, and the adult skull of the lowest Amphibian, namely that of Proteus anquinus. The reader, if he compare this with my former papers on the same subject, will find that my views as to the morphological interpretation of the parts are not fixed, but oscillate, tentatively. I hold that this vacillation is safest at present ; in each succeeding paper I express the views that seem to me to be true at the time; and I would rather waver in doubt— working upwards towards the light—than become fixed in the belief of some favourite view that might turn out to be a mere verisimilitude, and essentially erroneous. No man at present is able to say whether all or part, or, if part, how much of the *trabecule cranii” are ventral, or belong to the visceral-arch series. It was not known until lately whether there were any true visceral arches between the great mandibular arch and the “horns” of the trabecule in the frontal wall of the face. The interpretation of the cranial nerves is extremely difficult when it is sought to arrange them as the serial homologues of the spinal nerves—and this not merely in the nerves of special sense, but also in the common motor and sensory nerves, such as the “trigeminal,” “ facial,” &c. My own opinion was that the facial part only of the trabecule belonged to the visceral series—its terminal arch; then I yielded to Professor Huxley’s view of the visceral nature of the rods throughout; now I sway back again, and think that their subcranial part is axial in nature. Also as to the relation of the “visceral” to the “costal” arches, here is another heavy difficulty: in morphological “ habit” they are diverse exceedingly ; and whilst the latter are developed in a continuous “somatopleure,” the former are solidifications of the cloven oro-faucial wall. ON THE DEVELOPMENT OF THE SKULL AND FACE IN THE LESSER SPOTTED DoG-risH (Scyllium canicula). First Stage: Embryo of Dog-fish, 8 lines to 11 lines in length. Keeping the development of the Frog’s skull in view, it may be remarked that, on the whole, the youngest embryo of the Dog-fish (Pl. XXXIV. fig. 1) is intermediate between the first and second stages described in the Frog (‘‘ Frog’s Skull,” pls. 3 & 4); and with these they may be compared. The “ mesocephalic flexure” was complete, the middle vesicle (C’) projecting for- wards, and the anterior vesicle (C1) loking downwards. Everywhere very translucent, * «Batrachia,” Part 2, is now in print. See Phil. Trans. 1876, part 2, pp. 601-669, plates 54-62. OF THE SKULL IN SHARKS AND SKATES. 195 the skin was of extreme thinness, and composed of very delicate cells, over the greater part of the long posterior vesicle (C*). Each sense-capsule was seen, externally, to be formed by an infolding of the skin of the embryo, the “epiblast” and adjacent layer of “ mesoblast.” Behind the great gaping oral opening there were six clefts, equidistant from each other, and not meeting below. This embryo, being treated with carmine, and examined in glycerine as a transparent object, showed well the cartilaginous pith inside the thick ridges that intervene between the clefts. The surface of these ridges had budded into a series of rounded papille, as in the unhatched Tadpole (‘ Frog’s Skull,” pl. 3, figs. 2 & 3, dr 1& 2); these were the beginnings of the external gills ; they were found on all the. “ postoral” arches except the last. In the Frog-embryo they were not present on the first and second postorals. ‘The object (fig. 1) has been drawn obliquely, exactly as it appeared to the eye under the microscope; the other figure (2) shows the exact lateral view of the parts. All but the first two postoral ridges turn directly inwards (fig. 6); the enclosed pith is a stout sigmoid rod of young hyaline carti- lage. Such a pith exists in the first two postorals (mn, hy); but they send forwards from their point of attachment a large pedate process; they are subbifurcate above. This anterior fork, in the case of the second postoral, or hyoid arch, applies itself along the side of the auditory capsule. The first postoral, or rudimentary mandible, is still more produced beyond its proper suspensorial point; its foot or fork grows forwards over the mouth, and meets its fellow of the opposite side below the eye and behind the nasal sacs. Here we have the first rudiment of the “ quadrato-pterygoid” arcade ; it is found in the “ maxillo-palatine ” rudiment of the embryo. ‘This is different from what is seen in the larval Frog at a similar stage (“ Frog’s Skull,” pl. 3. figs. 3, 2, mn, & pl. 4. figs. 1, 2, mn), where the first postoral clings close to the trabecula. Afterwards, in the third stage, these bars, in the Frog, diverge ; and then (fig. 7, p.pg) they become united by a conjugational band, the first rudiment of the large palato-pterygoid bar. Lying deeper within the tissues, in reality in the oral roof, we see the edge of a bar in the embryo Shark (Pl. XXXIV. figs. 1 & 6, ér); it is the trabecula. The seven pairs of free visceral rods undergo a large amount of metamorphosis by segmentation, so as to form a most flexible oral and branchial apparatus. Already, over the infoldings of the young eye-ball, a ridge is seen ; this is an important part of the skull when developed, namely the “ superorbital band ” (see ‘ Salmon’s Skull,” pl. 4. figs. 1-3, s.ob). The ear-sac is still on the same morphological level as the nasal sac and the outworks of the eye-ball; it is a rounded fold, which soon will nearly close, however, and chondrify beneath the skin. ‘The very rudimentary nasal sacs cleave close to the inferior surface of the depending fore brain. The mouth, notwithstanding the palatal foot-like process that passes over it, is very open and gaping; altogether the postoral bars and clefts make an “ open-work” of the whole of the mouth and throat. In the more advanced specimens (Pl. XXXIV. figs. 2-5), delicate thread-like 194 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT embryos nearly an inch in length, the skeletal tissues were acquiring a greater density, and the various folds of the skin were much more perfect. Yet over the third vesicle (C*) the integument was quite transparent, and the contents of the head visible through it. A lateral view (fig. 2), seen by reflected light in a specimen hardened both by alcohol and chromic acid, shows how the skin is acquiring its proper characters in its growth from below upwards. ‘The sense-capsule folds are closing; and the posterior edge of each postoral visceral arch has developed an opercular “vallance,” the fringe from the second being the counterpart of that which is so largely extended backwards over the hinder arches (branchials) of the Osseous Fish. Here the free labiate edge grows from the mandible and gill-arches, as well as from the hyoid arch. Arrested at this stage, they would leave the gills much exposed; but they close in to a great degree, leaving only the well-known branchial slits. Morphologically considered, they also are the rudiments of such a growth of the skin as in an early stage covers over the visceral clefts in the ‘“‘ Abranchiata.” No cartilage is developed at present in the substance of the mandibular arch at this part, merely a strong ligament which attaches the mandible to the skull between the trigeminal nerve and the apex of the hyoid arch. This liga- ment is the primary apex; the pterygo-quadrate is the secondary fork. A deep fissure is seen between the inturned end of the pterygo-quadrate bar and the olfactory sac (na). During growth the arcuate cleft between the first and second postorals has become a large triangular space, with the base above and the apex below. From the posterior edge of the upper part of the mandibular bar four clubbed filaments proceed ; they look upwards and outwards: these are the free external transitory gills of the mandi- buiar arch, the precursors of the pseudo-branchia. ‘The counterparts of these, growing out of the succeeding arches, all but the last, are four or five times as long as in the more immature specimen. They are about ten on each bar, both on the right and left side. ‘The lower and upper views of this specimen (figs. 3 & 4) are very instructive ; and if the actual form of the enclosed bars of cartilage be held in mind (see fig. 1), it will be easy to understand their structure. The first cerebral vesicle (C’) is completely beneath the second (C?); and beneath the first the curious nasal sacs are seen, with their sigmoid valvular opening. The trabecular plate shows its form even in the opaque object. In front of the mouth are seen three lobes: the paired lobes contain the soft bulbous ends of the trabecular bars; and the azygous elevation between the nasal sacs contains the prenasal or basitrabecular cartilage, an unpaired commissural bar uniting the distal ends of the trabecular cornua. ‘The solid side walls of the mouth contain not only the pedate pterygo-quadrate bar, but also the fourth upper labial cartilage ; this will be shown better in a more advanced stage. The lower jaw, seen from beneath, is a quadrilobate mass fixed behind and below by a broad short pedicle. Its external lobes are the angular and articular regions; and the submesial swellings contain the short Meckelian regions of this peculiar mandibular arch. The following arches are much more bowed out; and between them the visceral OF THE SKULL IN SHARKS AND SKATES. 195 clefts are more extended, both above and below; their edges also develop a more distinct opercular flap. The huge pharyngeal bag is now seen to be slit on both sides; six of these slits are developed, and they are partly filled in by a very beautiful growth of plicated skin and of long clavate filaments. These filaments, the external branchie, are now seen to arise from the hinder edge of the bars, and to escape from the clefts like the contents of dehiscing carpels. The plice are arranged like the cogs of a wheel; they occupy both sides of the first four branchial arches, the hinder side of the hyoid arch, and the fore edge of the last branchial. This open condition of the respiratory region of the cesophagus is temporary, but shows what is possible in a low vertebrate form ; long before the embryo escapes from the horny tendrilled pillow-case the branchial slits are reduced to much smaller dimensions, relatively. The cog-shaped plice on these semicircular bars are the rudiments of the permanent or internal gills; they are hidden; and the bowed railings are filled in by the extension of the retral opercular folds. In the under view the umbilicus is shown, and on each side of it the rudiment of the pectoral fin; the heart is in the angular space between and below the posterior branchial arches, in front of the umbilicus. In the upper view, the ear-balls (au) are seen to be about the size of the eye-balls (e), and to be ovoidal in shape; they are beginning to acquire their own cartilaginous covering. The brain-sac is at present almost entirely membranous; and both the skin (cuticles and cutis) and the stratum of cells beneath that splits into dura mater and cartilaginous skull (roof and wall) are, in the upper region, exceedingly thin and diaphanous. One of the most important views of the structure of this early stage of the Shark’s skull is obtained by making a solid vertical section to be viewed as an opaque object (see fig. 5). Now the thinness of the integument over the third brain-vesicle (C?) can be demonstrated ; and this vesicle is largely filled by a thin fluid to two thirds its depth ; the interior of the other vesicles is very soft and diffluent. The middle vesicle (C?) is very bulbous; and the anterior (C!) is now developing into the hemispheres. Above and behind the fore brain is the pituitary body (py); and it helps to enclose a space formed by the curvature of the neural axis at its cephalic end (mesocephalic flexure); this cavity in the hook of the crozier is filled with delicate gelatinous tissue ; it is the transitory “middle trabecula” of Rathke. The notochord (nc) follows the elegant curves of the neural axis where it passes into the hind brain; it reaches to the pituitary body. On each side of the notochord is a stibcartilaginous plate, the two halves of which form the “investing mass” (iv); beneath the investing mass lies the pharyngeal portion of the first branchial arch (47.1). This section well shows how the pharynx is railed in by the visceral bars, and that the mucous membrane is folded into a saw-like series, the ¢eeth lying on the inner face of the bars above. The triangular open- ing (“ spiracle ”), corresponding to our tympano-Eustachian passage, is seen to be high and short, unlike its successors. The way in which the mandibular stem has been, as it were, trained forwards, like an espalier, to the front of the mouth, is also clearly shown. 196 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT To elucidate the meaning of the oral visceral arches more thoroughly, I have given a figure (Pl. XXXIV. fig. 6) of them in their earliest stage, separated from the rest, but shown in relation to the three infolded sense-capsules. Looked at unreflectingly, the mandibular and hyoid arches might be supposed to have their apex turned completely forward, so that the proximal part of the first might be brought to the distal end of the trabecular arch or plate; whilst the other, the hyoid, should have its apex close behind the foramen ovale (5). A careful observation of these, and comparison of them with those of the Skate (Pl. XX XIX. fig. 3) has satisfied me that the fore-turned hook is in reality a secondary fork, growing in these two arches from the primary fixing-point. Thus we still have the first position of the mandibular apex below the trigeminal nerve, and that of the hyoid below the auditory sac (aw). Now, if the “third postoral” or first branchial be compared with these two, it will at once be seen that it has, like its successors, no fork, and that its apex corresponds with the heel of the foot-shaped top of the first and second bars. We shall see, anon, what becomes of the pedate process, and of the proper apex from which it proceeds; the forked part is a ‘‘conjugational spur,” having the same me- chanical meaning as the slanting bars of a simple rural gate. This stage is still further illustrated by a section which shows the floor of the rudi- mentary skull; it passes horizontally through the nasal sacs; but the eye-balls and ear- sacs are untouched (Pl. XX XV. fig. 5). The embryo which was thus prepared mea- sured 1,1; inch in length—one sixth of an inch longer than the last. The mesocephalic flexure was still perfect, the mid brain (C?) being in front, and the fore brain (C1) below; the highest part of this object is at the fore part of the notochord (nc), where it is embraced by the hind part of the trabecule. Such a preparation, then, must be supposed to “dip” both in front and behind, from the postpituitary region; thus the trabecule (#7) and the investing mass (7v) meet at a considerable angle. The skeletal parts here displayed are much less solid than the visceral rods. This is also the case in Stredon at a similar stage ; but as the granular trabecule took up the carmine much more freely than the rest, and could be mistaken for no other tissue than young hyaline cartilage, I haye coloured it as such. But the cells forming the investing mass (¢ v) and the shell of the ear-labyrinth were much less coherent than those of the trabecule; and the internasal tract was still more behindhand in growth. The huge swelling brain-sacs, especially the middle one (C?), project far forwards ; and traces of growth from the axis can be found as far forwards as the front of the elegantly plaited nasal sacs (nq). If the subcircinate series of structures in the head of this Selachian embryo be con- sidered, it will be seen that there are three pairs of sense-capsules, and their inter- capsular regions, the auditory, the optic, and the olfactory interspaces. OF THE SKULL IN SHARKS AND SKATES. 197 But the hindermost of these is not segmented off at present from the rest of the axis ; this part is composed throughout of two tapes of young cartilage, closely applied to the sides of a median rod—the notochord (nc)—whose diameter is one third of that of either lateral band (7 v). I have traced these structures back behind the inter-auditory region more than twice as far as that region extends, without finding any transverse segmentation answering to vertebral division: hence we are perfectly safe in assuming that the “basilar plate,” or investing mass, is a continuation of the substance which in the spine makes itself into vertebre 1. The inter-auditory part of the investing mass has its sides bevelled and crescentically notched or concave; and the outer edges pass to some extent beneath the capsules. They do not reach further forwards than to the first third of the capsules, but are larger in the middle than at the sides, being inwedged between the ends of the trabe- cule and the notochord. ‘They also pass a little beneath the trabecular plates in front ; for, contrary to my earlier belief, I now find that the trabecule form the ‘“ posterior clinoid” region: in the Salmon (“ Salmon’s Skull,” pl. 2. fig. 5, and pl. 4. figs. 2 & 3, tr, iv), the ends of the delicate trabecular band lie over the fore ends of the basilar plates. The cephalic part of the notochord has not yet lost the bend downwards which is so conspicuous in early stages (Balfour, No. 2, pl. 24. figs. G, H, 1); but it is much straighter than when first distinguishable. At present, instead of ending in a hooked down-turned point, it ends in a beaded manner against the back of the pituitary body (py), which gets somewhat under the notochord which grows obliquely downwards and backwards (P]. XXXIV. fig. 5, py). The end of the notochord, where it pushes against the pituitary body, is vesicular ; and behind this terminal swelling there are stv more similar moniliform enlargements ; the second and third are small, and lie in a twisted manner in front of the remaining four, which become as large in diameter as the even part of the rod (Pl. XXXYV. fig. 5, and Pl. XX XIX. fig. 6, ne). This beaded condition of the fore end of the notochord appears to me to be open to two interpretations: it may be a temporary subsegmented condition, corresponding to undeveloped or suppressed segments in the head ; or it may simply be a puckering or folding of the sheath in a vegetative attempt to grow further forwards, the result of an effort to push away the pituitary barrier. The interocular plates, or trabecule, are the parts hardest to be understood. They may be precociously solidified tracts of the same nature as the investing mass, having the notochord between them only behind, and their separateness due to a dislocation, as it were, the result of the mesocephalic flexure. 1 Professor Huxley’s term “ parachordal” (“On Menobranchus,” p. 198), for these paired bands, is not distinctive of them ; for the terminal plates of the trabecule are also parachordal. VoL. X.—PArT Iv. No. 2.—WMarch 1st, 1878. 2F 198 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT On afterthought, it is an anomaly to me that the foremost pair of visceral or “ pleural” rods should grow straight forwards as the head straightens, and then from their upper edge develop three fourths of the chondro-cranium—namely the posterior and anterior sphenoidal regions, the ethmoidal or proper olfactory region, the internasal region in front of the tract supplied by the olfactory nerves,—and then finish off this exuberant and varied skull-growth by sending three preenasal ‘“‘ suckers” (the cornua and prenasal rod) into the intermaxillary region. There must be here some suppression of originally distinct parts, or elements ; and it seems to me now to be safer to give to the avis things that are axial, and to the face things that are facial. The aais appears to me to pass insensibly into the face in the internasal region, although some may argue that even the trabecular cornua and prenasal rod are then productions of the fore end of the axial elements. The somewhat lyriform trabecul are of great breadth in the Selachians (compare Pl. XX XV. figs. 3, 5, 6, ¢, with those of the Salmon, /. ¢. pls. 1-4). In front they are shaped like pruning-hooks; the blunt hook looking towards its fellow behind the internasal tract, but not meeting it. The back of the blade looks forwards and outwards, lying close behind the olfactory sacs. Their interspace, which is largely occupied by the infundibulum and pituitary body (inf, py), is equal to their width. Behind, also, they do not meet, but apply their inner edge to the three foremost notochordal ‘“ beads” (Pl. XX XIX. fig. 6). They send a right-angled wedge between the front of the investing mass and inner face of the ear-sac, to which they cling, and in front of which they form a rounded elbow: hence the outer edge of each plate is deeply notched in a semioval manner. Externally, the front projection is the rudiment of the lateral ethmoid, the part to which the antorbital or ethmo-palatine cartilage is attached in some Selachians, in Teleostei, in Urodeles, and in Anura. The hinder elbow is the part to which the “ pedicle of the suspensorium ” is attached in the Amphibia (“ Frog’s Skull,” and Huxley on Menobranchus). 'The pterygo-quadrate ends in front immediately below the fore end of the trabeculz ; in front of the trabecule, between the granular nasal sacs, the internasal tract grows broad behind and pointed in front, the pointed tract being the rudiment of the azygous prenasal cartilage. The trabecular cornua are not at present solidified sufficiently to show their distinctness from the contiguous parts of the nasal capsules. Second Stage. Embryos of Dog-fish 14-16 lines in length. At this stage the embryo of Scylliwm canicula still retains the ‘‘ mesocephalic flexure,” but the brain (Pl. XXXVI. fig. 2) has become very complex. The pituitary body (py) lies behind the fore brain (C1, @), and the “ middle trabecula” (m.tr) is not absorbed. The true trabecula (¢7) is very much enlarged, and flattening out above and behind has begun to form the large flat floor on which the fore part of the brain-sac rests; the OF THE SKULL IN SHARKS AND SKATES. 199 edge of the outspread apex can be seen just above the pituitary body (py). At present the azygous basitrabecular rod is beneath the hemispheres (C1,a); but afterwards it forms the axis of the beak or cutwater. On the whole, the sectional view of the second stage is much like that of the first ; but a dissected head shows that changes of the utmost importance have taken place. The segmentation of the proper branchial arches will be shown better in the next stage; they all, save the last, break up into four pieces on each side—a “pharyngo-,” “epi-,” “cerato-,” and “hypo-branchial” element. But the mandibular and hyoid arches (Pl. XXXVI. fig. 1, mn, hy) undergo only one transverse segmentation. This is but a step above what occurs in the Lamprey (Huxley, “Elem.” p. 193, fig. 76, g), where the mandible does not become subdivided at all, and the hyoid arch is only severed across to form two pieces. Our next subject, the Skate, is far in advance of this; and on this point the ordinary Shark is not in advance of the Chimera. It has recently been shown that in Osseous Fishes the hyoid arch is much more complex than its successors, the proper branchials (“Salmon’s Skull,” pls. 2, 3, 4, 5, 6 & 8), being composed of the same number of cartilages; but three of these are partly segmented, each into two, by having an additional bony centre. In the Sturgeon (‘Monthly Microsc. Journ.’ May 1873, pl. 20. fig. 1) there are five cartilages on each side. Here, then, the “ pharyngohyal” and the “epihyal” are in one piece, and the “ceratohyal” and “hypohyal” are in one. Here also the mode of segmentation is different and altogether simpler: in the bony fish it is from top to bottom, the second postoral bar being longitudinally segmented (“ Salmon’s Skull,” pl. 2. fig. 3); but in the Shark it is transverse segmentation, exactly like that of the bar in front, namely the mandibular. Whilst the pedate process of the mandible runs and fastens itself to the trabecula by its apex, the hyoid applies its pedate process to the whole side of the auditory capsule on which it is hinged; but the distal part of the upper piece is strongly tied to the arch in front of it. This simplicity of the oral arches occurs in the “ Dipnoi,” and also is the pattern on which the “Urodela” are constructed, although in both these cases fibrous bones complicate the structure; but in the Teleostei, as a rule, and in Ganoids, both “ Chondrostei” and “ Holostei,” the oral arches have to be traced up from those of the Rays, which are complex; these are to be described presently. Whilst the apex of the hyoid arch as well as the pedate process keeps attached above, the apex of the mandible becomes a mere fibrous band in front of the “spiracle,” or first postoral cleft. This band is attached to the skull behind and below the fifth nerve (fig. 1, c/. 1). Here is the beginning of that peculiar modification of the Fish’s skull by which the mouth becomes so mobile, the mandibular and hyoid arches hanging from the head by a single suspensorium—the “ hyomandibular,’”—which may be either the whole of the upper part of the hyoid arch, as in the present instance, or the larger half of it, obtained by longitudinal segmentation, as in the Salmon (‘Salmon’s Skull,” pl. 2. fig. 3, h.m). 2F 2 200 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT The most advanced specimens at this stage (1 inch 4 lines long) show much that is instructive, and fairly bridge over the space between the first and third stages. The head is fast becoming straight—see the nearness of the fore margin of the nasal sacs to the front of the head (Pl. XX XV. figs. 6, 7, as compared with fig. 5); and the trabecule, investing plates, notochordal sheath, auditory capsules, and visceral arches are all now chondrified. Ina horizontal view of a preparation in which the nasal and optic capsules were cut through, and the brain all removed, except in front (Pl. XX XV. fig. 6), we get a good view of the foundations of the Selachian chondrocranium. The notochord, which with the investing mass that has been cut through at some dis- tance behind the auditory capsules, is enclosed in a strong sheath of hyaline cartilage, has lost its beaded character in front, and now has pressed its end flat against the back of the pituitary body. The halves of the investing mass are scooped along their inner edges, where they cling to the sides of the notochord. Each plate passes some distance below the auditory capsule, but much more at both ends than in the middle. These cartilaginous bands have coalesced with the trabecule in front, growing into the lower edge of the thick transverse postpituitary wall (p.cl, py, 7v, tr). The auditory capsule inside the anterior ampullar enlargement has coalesced with the thick outer end of the posterior clinoid wall. In front of the wall the trabecule dip, and are somewhat concave; behind, directly in front of the ear-capsules, each trabecula is growing upwards into an alisphenoidal crest, which runs forwards to the optic nerve (2). This is their narrow part ; further forward they expand bepindl the nasal sacs in a pedate manner, but do not yet meet at the mid line. The intertrabecular space is larger than its enclosing cartilages, and is only occupied at its end by the neck of the pituitary body. At the mid line, between the trabecule and the olfactory sacs, the granular semi- cartilaginous internasal tract is seen. I cannot discover that the tract is ever divided into two distinct bands of cartilage, although its counterpart always is so divided, delow, in the Amphibia; in front it widens, curves right and-left round each nasal sac for some distance, and in the middle sends forward an azygous rod. This latter is the prenasal rostrum, the axis of the “ cutwater;” and the lateral growths are the cornua trabecule. Each of these is bilobate ; and in the next stage we shall see what a curious modification these two projecting masses of cells have undergone. At present it should be noted that the olfactory sacs, whose dome- like roof is now fast changing into cartilage, are very close together, only leaving a narrow valley between them, and leaving scant room for the septum or internasal region of the trabecule. When the head is examined from below we see the free forward growth of the OF THE SKULL IN SHARKS AND SKATES. 201 suspensorium ; a complete separation of the upper from the lower part of the strongly bowed mandibular rod has taken place; but the pier, instead of growing upwards as a cartilaginous “ pedicle” to be attached to the trabecular elbow, as in the Amphibia, is merely fibrous upward; and the cartilage grows a far way forwards, even to the olfactory sac, and then turns almost directly inwards to form a junction with its fellow, as in the free jaw below (Pl. XXXV. fig. 7, gpg, mh). At present the free jaw-pieces are thick where they are scooped in their articular region for the quadrate condyle, but they lessen gradually to the chin; their direction is forwards and somewhat downwards. The next or hyoid arch has its sides subdivided in the same manner as the mandibular ; there is one transverse cleft a little above the middle, dividing the bar into an epi- and a cerato-hyal; here, however, the scooped face is on the upper and the rounded head on lower segment. The upper piece does not quite correspond with the hyomandibular of the Osseous Fish, where the primary bar is split down from top to bottom; here it is merely divided bya sinuous cleft. A shield-like plate, the basihyal (d.y), unites the two lower pieces under the throat. The upper piece, called, ichthyotomically, the hyomandibular (/.m), has a broad top, which is applied closely to the inferolateral region of the ear-sac (aw). At its middle in front it becomes angular, and is strongly attached to the arch in front. The proper branchial arches have a part that runs under the hind part of the head and neck, that is separately chondrified ; this is the pharyngo-branchial piece (p.dr. 1) ; it is turned backwards and inwards. ‘There is also a division of the bar into an epi- and ceratobranchial (¢.47, ¢.br) ; and afterwards there will be a hypobranchial piece below. These things will be best seen in the next stage. At present the four external branchial filaments growing out of the “spiracular opening” are one third the length of the exceedingly long clubbed threads that break from the second cleft (fig. 6, ¢.67). They are all clubbed at the free end, where the single capillary loop turns back again to the main vascular arch. After the time of this stage the embryo grows much faster in bulk than in length ; the snout becomes like that of the adult (Balfour, /.c. pl. 25. figs. P,Q); the carti- laginous side walls and roof of the cranium chondrify at a rapid rate; also the facial rods become fully segmented and their metamorphosis fairly completed. Third Stage. Embryos of Dog-fish 13-2 inches long. The Dog-fish has fairly undergone its metamorphosis at this stage, although there are many important points to be noticed, in which the parts differ from the state of things in the adult ; this is largely a matter of relative size. The mesocephaiic flexure is lost (Pl. XX XVL. figs. 3-6), and the ‘‘ median trabecula” (fig. 6) is now a mere fissure between the medulla oblongata and the mid brain. ‘The 202 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT large and complex brain well fills the flat oblong brain-cavity, in front of which we now see the exquisite folds of the nasal sac (0/). Looking at the basis cranii (Pl. XXXVI. fig. 6, and Pl. XX XVII. fig. 1), we find that the notochordal region is now a broad subquadrate tract of cartilage, the two moieties of which have coalesced; there is a very small cone of gelatinous tissue left. The investing mass (7 v) behind is developed into two lobes for articulation with the “‘atlas;” and in front its extremities are rounded, the whole basilar palate being emar- ginate in front. This emargination and the space between the prepituitary cartilages (tr) make together a thin subpituitary space of cartilage of a lozenge-shape. The “internal carotids”’ (7. ¢) pierce this space at its broadest part. Opposite these pas- sages the anterior pair of cartilages (#7) curve outwards into a short flat cornu. This cornu is the “ elbow” of the trabecular bar (#7); it is largely separated from the side of the basis cranii, being bowed out: this is well seen in the Porbeagle (Lamna cornubica), and also in Carcharias glaucus (see ‘‘ Huxley and Hawkins’s Atlas,” pl. 5. fig. 4). In the basal figure of this skull of the embryo Dog-fish (Pl. XX XVII. fig. 1, #7") the trabeculee are seen to be of great size and remarkably flattened out; they expand beneath the eye- balls, and send out an antorbital process on each side. In front of these processes the trabecule are suddenly narrowed, and end between the nasal sacs (na) in a pair of short horns (cornua trabecul, c.tr). These are the distal extremities of the trabecule, which here have a basal or azygous piece, the prenasal rostrum or ‘“ basitrabecula ” (d.tr). This is exactly like what has been described in the Bird (“ Fowl’s Skull,” pls. 81— 84, pn); its direction is a gentle curve upwards (Pl. XX XVII. fig. 4, 0.ér). The trabecule at their extremities are 4-winged in section; for the narrowed thick lower part sends upwards, but more outwards, a thin broad lamina (Pl. XXXVI. figs. 3 & 4; Pl. XX XVII. fig. 1, ¢.tr); this is the “trabecular crest,” and is a very important struc- ture. In vertebrates with high skulls and a well-developed meso-ethmoid (perpendicular plate and nasal septum), the nasal sacs come close together, and their inner plates not only coalesce with each other, but also with the ascending trabecular crests, to form the single solid septal plate. Here the distance of the olfactory sacs from each other leads to the correlated diver- gence of the trabecular crests, which coalesce with the inner walls of the nasal dome. Each outspread crest gives off a small sigmoid cornu; and these two horns curl inwards towards the basitrabecular rostrum. To help in the interpretation of these parts the nasal domes have been emptied of their olfactory folds, and the valvular “ labials ” turned a little aside (Pl. XXXVI. fig. 4, and Pl. XXXVII. fig. 1, na, 7, 1, 2, 3). The dome-shaped olfactory cartilages are not only joined by coalescence to the trabecular crests; they are also confluent with an important pair of cartilaginous bands, namely the superorbital tracts of cartilage; these have already been described in the Salmon (“Salmon’s Skull,” pl. 4, s.ob), and are well shown by Dr. Traquair in the Polypterus (Journ. of Anat. and Phys. vol. v. pl. 6. figs. 2 & 3). They are early seen OF THE SKULL IN SHARKS AND SKATES, 203 as thickenings over the eye (Pl. XXXIV. figs. 1 & 2, s.ob, and also in the Salmon, “‘Salmon’s Skull,” pl. 1. figs. 1, 3,6 & 7). We thus come to a proper understanding of the nature of the ecto-ethmoidal masses of the higher vertebrata; they are composed of the proper olfactory domes, and the superorbital bands, which reappear in front beneath the nasals and frontals at their contiguous edges (‘“ Fowl’s Skull,’ pl. 83. figs. 2 & 5, al.e). The extreme simplicity of these primordial olfactory capsules (they send no outgrowths between the folds of the Schneiderian membrane) makes them of great use for unlocking the difficulties in the higher types, where several regions are specialized and many outgrowths formed. In the Shark there is no “ fontanelle” proper; but the roof of the cranium (“ tegmen cranii”) ceases in front, close to the nasal sacs (Pl. XXXVI. fig. 3, t.cr, na), so that the cranial cavity is open in front, and the long rhimencephalic crura diverge and pass into the nasal region beneath the inner side of the dome (fig. 4). These nervous masses lie on the trabeculz, and escape into the nasal sac through a space formed between the inter- nasal region of the trabecule and their olfactory cornua (c¢.tr); this “fenestra” answers to the chink in the Bird and to the cribriform plate in the Mammal. The diverging crests of the trabecular cornua (figs. 3, 4) being so outspread, the “ olfactory fenestra ” are in this stage nearly horizontal; but in the adult (Pl. XX XVII. fig. 3. 1) they are slanting, their direction being upwards and forwards from the narrowed trabecular bar to the postero-superior margin of the olfactory dome inside. Looking at the beaked face of a large shark, such as the Porbeagle (Lamna cornubica), it does not at first seem evident what the two clumsy-looking rods of cartilage are that converge towards the end of the basitrabecular rostrum. In the present subject (Scyl/ium) they are flat bands burrowed by slime glands (Pl. XX XVII. fig. 5, 71). In this early stage they are easily interpreted (Pl. XXXVI. figs. 3, 4, 5, and Pl. XXXVILI. fig. 1, 7, i?, 1°); they are two of the labial (“‘ extravisceral”) cartilages that cluster round the large nasal opening, to which they are related as valvular folds. The second and third of these are applied closely to the edges of the dome. The second is in front; and this is of a crescentic shape; it enlarges the nasal cavity, and partly floors it. The third is a thick ear-shaped cartilage on the outer edge of the nasal dome; this is the “‘ appendix ale nasi” of human anatomy. The first of these cartilages is in front of the second; it is somewhat heart-shaped, and slightly overlaps the second ; already it is the recipient of the two or three slime- glands which have burrowed its upper surface (Pl. XXXVI. fig. 3,/!). Afterwards it reaches over the nasal dome, and contains many glands (Pl. XX XVII. fig. 3, /'). The fourth labial (Pl. XXXVI. fig. 4, /*) is lanceolate with its stout end forwards ; it under- lies the quadrato-pterygoid bar (q.pg), and lies on the upper edge of the angle of the mouth. A similar cartilage, converging towards the last, lies on the lower edge of the angle of the mouth, attached to the infero-lateral surface of the mandible. ‘This is the Jifth labial (Pl. XXXVI. fig. 4, 7°). Similar cartilages, “ extrabranchials” (see 204 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT Pl. XXXVI. fig. 4, ex.b7), appear outside each of the gill-bearing branchial arches, four on each side, making in all nine pairs of these external cartilages. Three fourths of the space between the superorbital bands is filled up by the gently convex cranial roof, ‘‘ tegmen cranii,” which reaches as far as to the superoccipital (s.0), covering in the whole of this flat-bottomed boat as with a slightly shelving deck. Between the tra- becule below and the superorbital above, there is a cartilaginous wall, reaching from the ear-sac to the nasal sac; this is the sphenoidal wall, and answers to the orbitosphenoid and the alisphenoid of the higher vertebrata. A sectional view (Pl. XLII. fig. 5, 0.s) shows how the cranial walls are chondrified, thickened above by the superorbital band, and below by the trabecule; each of these regions of cartilage gives off a wing, the upper wing roofing the eye-ball, and the lower forming a partial floor for it. The optic nerve (2) divides the orbital ala from the “ala major ”—that is, regionally. A primordial fissure between the alisphenoidal cartilage and the auditory sac (aw) is converted into a crescentic foramen, the “foramen ovale” (Pl. XXXVI. fig. 5. 5). Where the superorbital cartilage, grafted on to the nasal sacs, becomes prworbital (Pl. XXXVI. fig. 5), we get the meaning of the lower part of the ecto-ethmoid, or pars plana, of human anatomy. The nasal branch of the ophthalmic nerve (51) passes between this downgrowth and the cranial wall. The hinder part of the flattened cranial floor (Pl. XX XVII. fig. 1, 7 v) is formed by the investing mass, and is the notochordal region. It reaches from the foramen for the internal carotid to the occipital condyles (0¢.c). Sectional views, the one (Pl. XLII. fig. 6) in front, and the other (fig. 7) behind, show the degree of its downward convexity, and its thickness at either end. The notochoid (nc) is still present in the hinder section ; it is very small, and imbedded in the very substance of the cartilage ; its cartilaginous sheath has coalesced with the lateral plates. The originally oval auditory sacs have taken the form of their bulging contents, especially of the large semicircular canals, which are very evident to sight on the upper surface (fig. 3, a.sc, h.sc, p.sc). The cartilage is complete both on the inner and outer sides, save where nerves and vessels pass (Pl. XLII. fig. 6) through the anterior part of the capsule and (fig. 7) through the posterior part. The capsule has coalesced with the superorbital band in front and above, with the superoccipital and tegmen cranii supero-mesially, and with the investing mass below. The hyomandibular is articulated below the horizontal (external) semicircular canal. The changes that have taken place in the first postoral (mandibular and pterygoid) are remarkable. The true apex, or metapterygoid (fig. 3, sp.c), is a fibrous band with a grain of cartilage in the anterior lip of the first cleft, or “spiracle.” The descending bar sends forwards an enormous foot, the pterygo-quadrate bar; and the descending part is cut off from this, as the free mandible or articulo-Mecklian rod (9¢.pg, g, ar, mk). Thus the secondary arch, or pterygo-quadrate, becomes practically as strong a bar in front of the mouth as the lower part of the arch behind it. That nearly the whole OF THE SKULL IN SHARKS AND SKATES. 205 apparatus of the upper jaw and palate of a Selachian should be developed out of a pair of processes growing forwards from the primordial mandible, seems at first sight a most unlikely thing; and yet no fact in morphology has been better established. The clue to its discovery was the peculiar segmentation of the primary hyoid arch in most vertebrates, either very obliquely through the upper third or half, or even, as in the Salmon, fairly from top to bottom longitudinally, the upper segment in this case generally growing forwards as a conjugational bar, yoking the hyoid on to the man- dibular arch. Here, in the Shark, the primary and the secondary arches of the man- dible (Pl. XXXVI. fig. 4) are extremely alike both in form and size, and are specialized to form this kind of mouth very perfectly. The halves of both meet at the middle, and are conjoined by a strong ligament; sinuous, flat, and selvedged for the greater part of their extent, they become terete and incurved distally. The articular region is gently scooped for the quadrate condyle; this is better seen in the side view of the adult skull (Pl. XX XVIII. fig. 2). The very simple hyoid arch (Pl. XXXVI. figs. 3 & 4, h.m, c.hy, b.hy) is a solid struc- ture; the suspensorial part is morphologically a whole “epihyal” piece, with no pha- ryngohyal segment above, and articulating with a “ ceratohyal ” from which no “ hypo- hyal” element has been cloven; the right and left bars meet by the intervention of a very elegant flat broad keystone piece, the “ basihyal ” (b.hy), which has a convex anterior and concave posterior margin. In this extremely mobile face only one more keystone element is found, namely the “‘basibranchial ;” and this belongs to the last two arches. The branchial arches, five in number, are very uniform on the whole; and the great development and singular uniformity of these arches in the “Elasmobranchs” makes them fairly typical as to the segmentation of this class of arches throughout the Ichthyopsida. Although the basal piece is only developed in relation to the two hindermost arches, yet the two in front of these have their distal symmetrical segment attached to that piece; thus, functionally, it serves for four arches. Each arch has three fibrous joints on each side, developed in the original substance of the joint by a limited production of con- nective fibrous tissue instead of hyaline cartilage. Each bar, in its primary form, is very elegant: it is bowed out laterally ; it descends in a sloping manner forwards; and both its apex and its distal end are hooked backwards, these hooks becoming the “ pharyngobranchial” and the “hypobranchial” elements (Plate XXXVI. figs. 3 & 4, p.br, h.br). The upper third of the main bar, answering to the hyoid suspensorial piece, becomes the “epibranchial” (e.br); and the remainder is the proper ceratobranchial (c.br). In the first branchial the hypobranchial is a small nucleus cloven from the anterior lobe of the foot-shaped extremity (fig. 4, h.r, 1); but in the rest of the functional arches the hypobranchial is a typical segment—large, free, and retrally attached to the base of the ceratobranchial. Vou. x.—Part Iv. No. 3.—WMarch 1st, 1878. 24 206 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT The pharyngobranchial of the fourth arch belongs also to the fifth (fig. 3); forking below, it is attached to the apex of both the fourth and fifth. This latter arch has its epibranchial continuous with the pharyngobranchial of the fourth; it has a flat notched ceratobranchial piece (fig. 4, c.br, 5), and has no hypobranchial. The sectional views further illustrate this stage of nearly ripe embryos of the Dog-fish. A vertically longitudinal section (Pl. XXXVI. fig. 6) shows how completely the cranial cavity is filled with the brain, and that the mesocephalic flexure is obliterated. The tegmen cranii is continued over part of the fore brain; and the floor of the cranium is one continuous sheet of cartilage, formed by the investing mass behind, and by the united trabecule in front. The nasal sac (0/7) now lies in front of, as well as some- what beneath the fore brain. Beneath and behind the olfactory folds is seen the distal end of the pterygoquadrate bar (q.pg), and below the mouth the distal end of the man- dible (mn). The hyoid crus and base (c.hy, b.hy) and the distal parts of the branchial arches are shown, as also the basihyal and basibranchial, in section (b.hy, 6.67). There is but little of the notochord (nc) left; and a posterior clinoid ridge shows the rudi- ment of the “sella turcica.” The pituitary body (py) is very small; beneath it the internal carotid is seen entering the cranium. In the first of the transverse sections (Pl. XLII. fig, 5) the eyeballs are cut through, and a view is gained of the height and width of the cranium, built upon the foundation of the trabecule and their “‘ commissure.” The orbitosphenoidal side walls connect the trabecular crest with the superorbital band and the “ tegmen cranii.” Below the mouth (m), the oral “labials,” the pterygoquadrate bands, the mandibles, and the fore part of the basihyal are cut through (q.pg, mn, b.hy). In another section further backwards, and somewhat oblique (fig. 6), the hinder part of the eyeball (¢) and the fore part of the auditory sacs are shown. The basal cartilage is cut through where the trabecule have coalesced with the investing mass. The third section (fig. 7) is through the posterior and horizontal semicircular canals (p.sc, h.sc), the upper part of the oblique foramen magnum (f:m), the investing mass, and remnant of the notochord (iv, nc), and the head of the hyomandibular, or “ epihyal” (/.m). The thick auditory capsule is seen to be still distinct above from the superoccipital cartilage (so). The passage for the glosso-pharyngeal and vagus (9, 10) is seen below. The remarkable extravisceral cartilages (Pl. XXXVI. fig. 4, ex.br) are shown outside the branchial arches: they are sharp above, and dilated below ; there are four pairs of them; and the last (¢.vs, 9) is very small. These cartilages await proper classification; at present they may be bundled up with the “ labials.” Fourth Stage. Adult Dog-fish (Scyllium canicula). At first sight it might be supposed that a skull without any proper ossification in the adult state would present the greatest difficulties to the morphologist; for the yarious elements of the cranium are here thoroughly soldered together, making a com- OF THE SKULL IN SHARKS AND SKATES. 207 plex cartilaginous box, as complete as the bony box seen in the skull of the higher kinds of birds. Here, if anywhere, the cranial segments ought to be found. But, mor- phologically, these types are not at a very low level; certainly, if the cranium, even behind the pituitary body, be the result of slow secular consolidation of a series of vertebree, these fishes would seem to be a very late product of evolution. The truth of the matter is, that the Sharks and Rays are yery enigmatical as to their position in the vertebrate series; below the Teleosteans as to their skull, by two important steps or degrees (as was shown in the paper on the Salmon’s skull), they yet come much closer to the tail-bearing and tailless Amphibia than any other fishes with the exception of the “ Dipnoi.” These things must be borne in mind whilst studying the conditions of the adult skull of a type which undergoes no bony metamorphosis, yet has so very perfect a cranium and a large and perfect basket-work of visceral arches. The cranium itself (Pl. XX XVII. figs. 2,3, Pl. XX XVIII. figs. 1, 2) is a very elegant structure; it is a flat-bottomed barge, like that of the frog (“ Frog’s Skull,” pl. 9. figs. 6 & 7), but having a cartilaginous deck. On each side, in front, there is an elegant dome-shaped “ awning "—the roof of the nasal sac; the deck, or tegmen cranii, is wide open in front; beyond this opening a small ‘“‘ prow” projects, the ‘‘ preenasal or basi- trabecular” cartilage; and this is spliced obliquely by a pair of bars, the foremost extra- viscerals, which were in front, simply, and now overlie the nasal roofs (Pl. XX XVII. fig. 2, 7. 1). The basal view (Pl. XX XVII. fig. 3) shows the almost uniform breadth of the whole of the occipital and sphenoidal regions, the centre of the post-sphenoidal territory being shown by the entrance of the internal carotid (7.c). The occipital condyles (oc.c) project but little; the ridges of the otic capsule square the skull behind; and these ridges form a “ tegmen tympani,” under which there is no tympanic cavity, but a condyloid subconcave facet for the huge representative of the incus, the “ epihyal ” or hyomandibular. The lower edge of that facet is formed by the investing mass (7 v); between this and the flat, outspread trabecular elbow (>p.tr), is a notch; and a lesser notch separates the elbow from the rest of the trabecular plate, which further forwards narrows again, and then sends out the antorbital spur(a.o). The trabecule then suddenly contract, and grow upwards (Pl. XX XVII. fig. 3); and their ascending part becomes now coalesced with the inner edges of the nasal-roof cartilages, thus forming a primordial mesethmoid or septum between the nasal sacs, which is normally composed of four carti- laginous growths. On each side of this middle wall a membranous space, open in the middle, forms a sort of trap-door down into the nasal sacs, through which the olfactory fibres pass to the nasal plice, which are pinnately arranged, and entirely membranous. These spaces answer to the moieties of the cribriform plate of the mammal. The second labial (/. 2) has partly coalesced with the anterior edges of the nasal dome, and with the corresponding cornu trabecule (c.tr). The third labial (/. 3) is precisely like 2G 2 208 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT that of the pig (“appendix ale nasi”), and protects the outer edge of the lip of the nasal sac (see “ Pig’s Skull,” plate 36. fig. 1, ap. a.n). Seen from above (fig. 2) the skull is a very elegant structure, with its approximated nasal roofs, its prow-like trabecular rostrum (d.tr), its oval anterior remnant of the fontanelle, and its convex tegmen cranil. The superorbital ridges (s.ob), so early seen in the embryo, now give character to the skull; they are grafted upon the otic capsules behind, and upon the nasal cap- sules in front, and even down in this type complicate the morphology of the ecteth- moid. A superorbital foramen opens out in the front part of the groove between the superorbital and the “tegmen;” and behind the tegmen are seen the right and left “ aqueducts of the vestibule” (ag.v). On each side of these the elevation caused by the semicircular canals is clearly seen. A lateral view (Pl. XX XVIII. fig. 2) shows the relation of the parts very clearly, and especially lights up the wing of the “antorbital,” or “pars plana.” Although in the bird and mammal the antorbital cartilage is absent over the eye, yet it appears behind in the “sphenotic” region of the otic cartilage, and in front on the lateral ethmoid, running down in front of the eye as the free outstanding edge of the ect- ethmoid (“ Fowl’s Skull,” plate 81. fig. 5, p.p), whilst the “ tegmen cranii” reappears as the retral spike growing from the coalesced nasal sacs (ibid. plate 83. figs. 2, 4, 5 (eth). Here there is no distinction of orbital and larger wings of the sphenoid; the space between the superorbital and trabecular cartilages is filled in by a continuous growth of the same nature—a vertical sphenoidal wall, which is riddled with larger and smaller holes, the more important of which serve as landmarks for morphological territories (Pl. XX XVIII. figs. 2 & 4. 2, 5, 5’). Besides the olfactory fenestra, there is another between the partially fused trabecular cornua and nasal roofs (Pl. XX XVIII. fig. 4, ¢r,f). The mammals do not show this; but in birds it is very common—a re-separation of the trabecule from the nasal inner walls; it is not seen so low down as among the true “ Struthionid ” (“ Ostrich’s Skull,” plate 7. fig. 2, s.n)—but appears in the next group, the Tinamous (ibid. plate 16, fig. 8, s.), and in the Fowl in the Tinamine stage. (‘‘ Fowl’s Skull,” plate 83. fig. 4, ¢,fc). The Mammalian skull is markedly prefigured by that of the Shark in the pituitary region; there is an anterior and a posterior “ clinoid wall,” and a floor to the sella turcica perforated on each side by the internal carotid (Pl. XX XVII. fig. 5, @.cl, p.cl, py,tc). This is quite unlike what is seen in Osseous Fishes, Lizards, and Birds, which have no cartilaginous floor to their “ sella.” The notochord is replaced by cartilage ; and the articulation of the skull to the “atlas” is by two condyles. Between the superoccipital cartilage and the proper tegmen cranii above there is a crescentic fossa, looking forwards; on each side of this is the aqueductus vestibuli (Pl. XXXVII. fig. 2, ag.v). On each side of these burrowings are seen the anterior and posterior semicircular canals, the latter ending in the epiotic eminence. Outside the ampulla of the OF THE SKULL IN SHARKS AND SKATES. 209 anterior canal is seen a short thick process, the sphenotic process (sp.c): it is formed by the grafting of the superorbital arc (s.0d) on to the auditory mass. Behind the notch is the large “ pterotic” eminence ( pto), containing the horizontal canal (/.s.c). Whatever the superorbital arc may be morphologically, it is a structure of the greatest importance ; here, at its fullest growth, it shows that the “lateral ethmoid” and the periotic capsule are made compound by coalescence with it; and this original composition of these parts must never be forgotten in an ascending survey. Thus the nasal capsules are mixed up, or confluent, with the trabecular and basi- trabecular bars, with the superorbital arcs, the “‘tegmen cranii,” and one pair at least of the labials. The periotic capsule is fused with the “ parachordal” bands, or investing mass, with the arch growing upwards from that mass (occipital arch), with the tegmen cranii, and with the superorbital are. Afterwards, when we come to study such skulls as have the chondrocranium ossified into certain definite (interneural) bony territories, we shall often see a single bone formed in what was a very complex part originally; and therefore such bones must be considered as the products of metamorphosis, and not primordial elements of the cranium. A section taken longitudinally (Plate XX XVII. fig. 4) shows how the brain (€1, Cla, C2, C3) fills the cranium, and how the occipital ring and tegmen cranii cover the greater part of the brain. Below, the notochord only persists between the atlas and the basis cranii; further forwards the internal carotid (é.c) is seen entering the skull-base below the sella turcica (see also fig. 5), which has an anterior and a posterior clinoid wall (a.cl, p.cl) and contains the tear-shaped pituitary body, above and in front of which is the hollow infundibulum (inf). In front the mesethmoidal fenestra (tr,f) is seen to be a mere membranous space where the nasal sacs have not thoroughly coalesced with the trabecule; in birds this fenestra is formed by the reopening of the cartilaginous wall after complete fusion has taken place. When the brain has been removed, the various openings can be seen for the exit of the nerves: these lie low down; and the chiefest of them can be easily determined (see Pl. XXXVIII. fig. 4. 1, 2, 5, 7, 8, 9,10). That for the olfactory (1) is a large obliquely tilted window—the membranous “ cribriform plate.” The chondrocranium is not more massive and complete than the facial or visceral arches, which here attain their utmost size, and undergo no further histological meta- mophosis than the calcification, in tesseree, of the superficial cells of the hyaline car- tilage. Arrest in metamorphosis (as to bony deposit) is here the correlate of large development of the parts as to size; in number this and many other sharks agree with the Osseous Fishes, there being seven postoral arches; there are more in Heran- chus and Heptanchus, which have respectively six and seven persistent branchial slits, besides the “ spiracle.” 210 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT The pterygo-mandibular arch (Pl. XX XVII. fig. 2, and Pl, XXXVIII. figs. 1 & 2, q-P9; 7, mn, mk) is curiously swung from the outside of the basis cranii by two short ropes of fibrous tissue: the hindermost of these is the true apex of the arch, the meta- pterygoid, or “ pedicle” (J, sp.c); and the foremost is the palato-trabecular conjugation (p.tr). This latter is attached to a process on the pterygoid, too short to reach the trabecula; and the metapterygoid band in front of the spiracle (¢/. 1) is attached above to the skull, close in front-of the hyomandibular (hm), and behind and below the fifth nerve (5). Where this band is attached to the quadrate region (g), the cartilage is bevelled down, only sufficient substance being left to form the condyle for the mandible. There is, however, but little difference in size between the upper and lower jaw-plates ; for the upper broadens as it arches upwards to the dentigerous part, and then it gently curves downwards to its pterygoid or distal end, which is strongly tied by a ligament to its fellow of the opposite side. This bar is bowed out at its upper margin, as is the mandible at its lower edge ; the scooped surface thus formed gives attachment to the oral muscles. This primary mandible has a very similar form to the secondary mandible of a mammal, save in not possessing a coronoid process. Its narrowing dentigerous part is equal in size and extent to that of the upper bar; the right and left halves are similarly bound together by ligamentous fibres. The hyoid arch is as simple as the mandibulo-pterygoid, being composed of two pieces only; the whole arch cleaves close to the one in front, and is locked within it below. This is a foreshadowing of what takes place in more metamorphosed types, as is also the suspension of both the free crura from the pier of the hyoid, which now can fairly be called, from its function, the hyomandibular (hm). The metapterygoid ligament is attached below equally to the two contiguous condyles—namely, that of the quadrate and of the hyomandibular; thus the weight of the large mandible is transferred largely to the hyoid pier. The hyomandibular has a joint-cavity above and below it; and its articulation with the free cornu below is like that between the human finger-joints. The thick, ribbed, outturned edge of the mandible is attached at its rounded angle, near the joint, by a strong “ mandibulo-hyal ” ligament (m.h./), as in the Urodeles; in Osseous Fishes the angular ligament becomes bony above, as the “ interopercular.” The stout phalangiform ceratohyal (c.hy) by this ligament is kept close to and within the mandible; it is bilobate distally ; and the anterior lobe articulates with the angle of the elegant heart-shaped basihyal (Pl. XX XVIII. fig. 1, c.hy, 6.hy); there is the merest rudiment of a joint-cavity at this part, but, as in the arches behind, a profusion of fibrous tissue. As the hyoid arch is functionally branchial, it carries branchial rays (d7.r), that give strength to the pouches. ‘There are three of these on the hyomandibular, the upper of which is trifurcate, the middle bifurcate, and the lower subdivided into five or six long leaflets. OF THE SKULL IN SHARKS ADD SKATES. 211 On the ceratohyal (Pl. XX XVIII. figs. 1 & 2, c.hy, br.r) there are seven, the upper of these being quinquefid, and the rest bifid. The branchial rays on the next four arches (the principal branchials) are simple in form, and range in number from ten on the first to five on the last. ‘Two or three of these are on the “epibranchial,” the counterpart of the hyomandibular, and the rest on the ceratobranchial.” The hyoid arch meeting with the chondrocranium, and articulating with it, has no ‘‘ pharyngo-pleural ” element, a part developed in all the succeeding bars, which float, as it were, over the large pharyngeal cavity, and are not attached to the axis of the animal. The branchial arches (Pl. XX XVII. fig. 2, d7, and Pl. XX XVIII. figs. 1 & 2,-b7r) are, on the whole, very uniform; but the fifth, or last, is abortively developed ; on it, as on the hyoid, there is but one series of branchial plicee. These arches have great mobility, and form complete girdles to the huge pharynx; their apices meet together above, and their distal parts below ; behind, there is a large keystone piece to the last two of the arches (Pl. XX XVII. fig. 2, and Pl. XX XVIII. figs. 1 & 2, 6.6r). Having great mo- bility, the dorsal segments (pharyngo-branchials, p.dr) may turn forward (Pl. XXXVIL. fig. 3, and Pl. XX XVIII. fig. 2), or backwards (Pl. XX XVIII. fig. 1); normally they turn backwards, the opposite direction to that taken by the hyoid and mandibular arches. Nevertheless the segment that corresponds with the upper part of those arches, namely the epibranchial (e.br), sends forward a pedate process, the exact coun- terpart of the fore part of the head of the hyomandibular, and of the quadrato- pterygoid lobe of the mandibular arch. The pharyngobranchials are thus seen to be attached behind the pedate process, which is what occurs in the hyoid of the frog, where the pharyngohyal element (“interstapedial ”) is developed on the end of the hinder fork of the bar. Primarily the direction of the dorsal ends of the visceral arches is inwards and backwards, and the foreturned part is a secondary spur or fork ; the ventral or lower ends have also a retral habit of growth ; so that normally the hypo- pleural element is always more or less hooked backwards. In the first branchial of the Dog-fish the hypobranchial is merely a bud, segmented off from the anterior lobe of the pedate lower end of the arch. In the rest of the principal arches (h.dr 2, 3, 4) this ventral element is a long terete rod, sharply retral, and forming an acute angle with its fellow of the opposite side: this segment does not exist on the last arch; and its pharyngobranchial is not distinct from that of the fourth branchial (Pl. XXXVII. fig. 2, and Pl. XX XVIII. fig. 2). The epibranchial and ceratobranchial series are closely like the “epi- and cerato- hyals,” save that they are smaller; they are similarly palisaded with branchial rays. In the lower view (Pl. XX XVIII. fig. 1) the seriality of these arches is well seen. The hyoid arch, with its multifid branchial rays, that take the place of the bony opercular and subopercular of the Osseous Fish, is the only arch, besides the last branchial, that 1 T haye figured them in both ways for illustration. 212 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT is not supplemented with an “ extravisceral” cartilage. There are three pairs of labials in the trabeculo-nasal region—one pair on the pterygo-quadrate arcade, one pair in the mandible itself, and a pair of similar cartilages to each branchial arch save the last. The first of these external cartilages, which was a small lobe in front of the nasal sac in my third stage (Pl. XXXVI. figs. 3, 4, 5, 7.1), and had only one or two slime-glands in it, has now been carried over the great nasal dome behind,.and over the basitrabe- cular in front; it contains a row of about ten slime-glands, and forms the side of the triangular “ cutwater.” The second labial (/. 2) is a thin lunate shell of cartilage, having an irregularly crenate free margin; it has partly coalesced with the trabecular cornu and anterior rim of the inverted nasal cup; it lessens the size of the nasal opening. The third labial (/. 3) is thick, phalangiform, and twisted; it is applied as a valve to the outer rim of the nasal cup, and exactly corresponds with the “ appendix alee nasi” of the mammal. The fourth labial (/. 4) is a small pointed spatula of cartilage, applied obliquely to the middle part of the outer face of the pterygo-quadrate; its broad end is in front.! The fifth labial (2. 5) is very similar, but is smaller, and is applied to the mandible in such a manner as to meet, by its pointed posterior end, the fourth labial; the angle at which they meet is very acute in the closed mouth. The remainder of the “ extravisceral” cartilages may be called “‘ extrabranchiales ;” they are four in number, and lie inside the skin of the branchial pouches at the ex- tremity of the branchial rays; thus, in the figures (Pl. XX XVIII. figs. 1 & 2, evs they are shown opposite the arch next behind. Each bar is sigmoid, pointed above and pedate below. The second has the largest “ foot ;” the last is rudimentary, and is attached to the fifth branchial, although it belongs to the fourth. The reason for the absence of an extrabranchial cartilage from the gill-bearing hyoid is not evident ; they appear to belong to the same category as the labial carti- lages; hence I propose to call them all “ extraviscerals.” These extrabranchials are a source of strength to the gill-pouches; they lie in their lower part principally, their pointed end being attached to the free extremities of the branchial rays, between the slits (clefts), and close in front of the slit which intervenes between the arch to which that particular extrabranchial belongs and the one next following. The extrabranchials cease as such in front of the last branchial arch; but they are followed by a pair of precisely similar cartilages, which early coalesce together below by their broad pedate end. These are “ extracostal” cartilages ; and from them proceed the rays that form the skeleton of the pectoral fins. This pair of coalesced cartilages form the “ shoulder-girdle,” or scapulo-coracoid belt. * In Scymnus, Squatina, Centrophorus, and others, there is another labial in front of this, which corresponds to the fourth of the Skate——See Gegenbaur, pls. 11 & 12, and Pl. XLII. fig. 4, 2. 4, of this paper. OF THE SKULL IN SHARKS AND SKATES. 213 ON THE STRUCTURE AND DEVELOPMENT OF THE SKATE’S SKULL. First Stage: Embryo of Raia maculata, 13 inch in entire length, seven weeks after deposit of egg; and Embryos of Pristiurus, 3 and } of an inch in length. The first embryo, taken for me from the egg-pouch of the Spotted Skate in the Brighton Aquarium by Henry Lee, Esq., was, in development, intermediate between the less and more mature embryos of the Dog-fish (from the same friend) already described (Pl. XXXIV.); the two others, the gift of Mr. Balfour, were from the Naples Aquarium. The length of the embryo Rays from Brighton was much greater than those of the Dog-fish, owing to the extreme development of the tail, the anterior part being no bulkier than the smallest embryos of the Dog-fish which I have just described. The pectoral fins (Pl. XX XV. fig. 1, and Pl. XXXIX. fig. 5) are simple lunate folds on each side of the umbilicus (w); and these embryos, if they had been found detached, could not easily have been distinguished from those of a Shark, the peaked rostrum and the fan-like shape of the fins not being developed as yet. At present the skeleton of the embryo is quite granular and transparent, so that by careful management most of its structure can be made out without any dissection. These embryos show, on a good scale, the structure of a vertebrate embryo in its first or simple morphological stage. Many embryological processes have been gone through ; but now its primordial skeletal parts have been fairly differentiated. The sacs of the special sense-organs are at present horseshoe-shaped folds of the embryonic cuticle and cutis, the large closed brain-vesicles (C1, C?, C*) are full of watery fluid, and the third of these (C#) is covered very thinly by large soft mother cells. The mouth and pharynx are covered above by the axial structures, and floored below by a con- tinuous throat-skin, above which, behind, is the heart; but the sides are an open grating, hedged in by bowed bars. The mesocephalic flexure is perfect, and the mouth ccmplete. None of the visceral arches meet, right and left ; but the pterygo-mandibular bars are coming near each other both in front of and behind the oval mouth. Behind the mouth the visceral bars are yet further and further apart, and the arches themselves * gradually lessen in size; above, the hinder arches are set on to the infero-lateral edge of the vertebral structures in the cervical region; half the postoral arches are behind the third cerebral vesicle (Pl. XX XIX. figs. 1, 2, 5, 6r). Behind the azygous oral cleft the visceral openings are variable in size, the first post- oral cleft (the “spiracle,” or “ tympano-Eustachian,” c/ 1) is less than the next; it has a pear-shaped outline, and soon fills up below, so that, when developed, this opening, the “spiracle,” is seen in the dorsal region, close behind the eye, whilst the others are on the ventral aspect. The second cleft (c/ 2) is larger, and retains its lower slit- like part. The remainder, between the proper branchial arches, are tolerably even in size, but have less vertical extent behind because of the shortening of the bars. From VoL. X.—part Iv. No. 4.—WMarch 1st, 1878. 2H 214 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT all the bars, except the first or spiracle, large, spatulate, external branchial filaments grow}, six or seven on an average to each bar on each side; they are now, in the longest embryo, the largest of them, as long as the head; each contains, as in the embryo of the Dog-fish, a single capillary loop. Four short buds appear in the “‘spiracle,” or first postoral cleft (Pl. XX XV. fig. 1, c/1): in embryos that are evident Skates a large pseudo-branchia is seen inside the front wall of the spiracle, and arising from the back of the proper apex of the mandible (Pl. XL. fig. 3, ed 1, ps.dr). The branchials (67 1-5) are simple rods; not so the rudimentary mandibular and hyoid arches; these are bifoliate above, more markedly d¢furcate than in the Dog-fish, in which they were pedate (Pl. XXXIV.). Here the bifurcations are filled with granular substance, which becomes eouedh cartilage in each fork; and in this the Skate comes nearer to the ganoid and teleostean fishes. In each case the hinder spur is slenderer than the front; and in the mandibular arch it is seen that the hinder hooked snag is but loosely connected with the main part. This is in front of the ‘“‘spiracle” (cl 1); but, behind this opening, the front part of the next bar is severed from the long hinder stalk. The front fork of the mandibular arch turns downwards and forwards to become the upper jaw; the small backwardly curved spike, the hinder fork, becomes the spiracular cartilage. In the second arch (hyoid) the large front fork becomes the hyo-mandibular, and the rest of the bar the proper hyoid. A reference to the figures will show that the suspensorium of the mandible, the hinder fork, although it embraces the anterior spiracular lip, is yet attached to the head below and behind the exit of the trigeminal nerve, antero-inferiorly to the auditory sac; it is a free ‘‘ otic process.” The anterior fork of the hyoid arch is in inferior relation to the auditory sac; it articulates broadly with it afterwards as the hyo-mandibular, whilst the hinder fork, or “ stylo-ceratohyal,” is loosely connected with that sac behind. In the various types, and even in the various stages of one and the same type, the hinder division of the second postoral is very variously articulated to the surrounding parts. It is worthy of remark that that which distinguishes the Rays from the Sharks most completely, namely the mode of segmentation of the hyoid arch, is already evident in the embryos at this stage. The Skate breaks up this arch in the same manner as the Sauropsida and the Mammalia; whilst the Shark shows it in a simple and low form. I have worked out the basis cranii in the youngest Pristiurus, 2 of an inch long, and the visceral arches in the larger specimen of Pristiwrus, } of an inch long (Pl. XXXV. figs. 3 & 4). When the base of the skull is seen from above (fig. 3) there are three intercapsular regions displayed, although the bend of the head throws the eye-ball into the same vertical line as the nasal sac (see fig. 1). If we suppose a curved line passing con- 1 They are small in the first cleft. OF THE SKULL IN SHARKS AND SKATES. 215 tinuously through the middle of the sense-capsules in the first figure, that will represent the edge of the preparation figured in fig. 3. Such a line would bend strongly down- wards (at about 120°) in front of the ear; and that is where the middle and hinder regions meet; these are the trabecular and parachordal tracts. The parachordal tracts (7 v), or investing mass, are cut away, as it were, to fit to the convex edge of the ear-capsule ; and being bevelled, they pass under each sac somewhat. They then broaden, where the postaural nerves (9, 10) pass out, and then, narrowing a little again, are equal parallel bands, running without transverse segmentation some distance into the cervical region; and the numerous roots of the vagus and the following spinal nerves are seen at their outer edge (fig. 3. 10, sp.n). In front the parachordals thin out a little, and pass under the ends of the next pair of bands—the trabecule; these are wider, shorter, and more solid tracts of crowded cells, taking up the carmine more perfectly, and rapidly becoming hyaline cartilage. They form afterwards the flattest of skulls; and now are wide flat bands, angular behind, where they are wedged in between the parachordals and the ear-sac; they then narrow at their waist, and broaden into a rounded spatula in front, close behind the olfactory capsules. They are not merely prechordal tracts; for the notochord reaches to their middle (Pl. XXXV. fig. 3, Pl. XL. fig. 7, tr, nc), where it is twisted, recurved, and constricted in a moniliform manner, like the floral hairs of Tradescantia virginica. The rest of the notochord is unconstricted, and passes, of an even size, into the uncleft spinal region. The pituitary space is just equal in size to the notochordal tract of the trabecule; it is oval, and is only finished in front by a newer and softer tract of tissue—the rudimentary internasal band. The infundibulum passes obliquely into the pituitary sac (if, py), the latter getting somewhat under the apex of the notochord. Behind, the posterior clinoid edge of the trabecule runs inwards and forwards and externally ; the trabecule then widen to their “elbow” or most projecting postero- external point. In the angle between this point and the fore margin of the ear-sac, the preauditory nerves (5, 7) are seen as large leashes of young fibres in four main divi- sions, the fourth (facial nerve, 7) hooking round the front of the capsule, Antero- internally to that capsule a mass of cells can be seen, the rudiment of the Gasserian ganglion, and postero-internally another smaller mass, the rudiment of the ganglion of the ninth and tenth nerves. The oval olfactory sacs (na) have between them a dagger- like internasal tract, which, like the covering of those sacs, is very soft at present ; this is the rudiment of the internasal part of the trabeculz, the cornua, and the preenasal rostrum (7.7¢). The pointed end of this tract now lies behind the mid brain, truly a remarkable posi- tion for the rudiment of that long beak which afterwards forms the axis of the “ cut- water” of the Skate, and even of the Saw-fish, whose structure corresponds to that of the Skate (Gegenbaur, pl. 14. fig. 2). 2H 2 216 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT In an embryo of Pristiwrus 73 of an inch longer than the last (? of an inch), all the postnasal structures of the skull-base were clean removed, and the upper wall of the fauces and the pharynx shown (Pl. XX XV. fig. 4). The parts last described are shown 7 situ, as in fig. 5; and the whole series of post- oral visceral arches are shown: they were seen as a transparent object, and the prepara- tion flattened a little. The pith of the facial rods took the carmine well, and had become true hyaline cartilage. If this figure be compared with the side and lower views of the younger embryo, drawn in its undissected condition (figs. 1, 2), the following description will be easily understood :— The first cleft or spiracle (c/ 1) is wide at the top and a mere chink below ; its direc- tion during the mesocephalic bend is backwards and downwards, its opercular lip, or “ vallance,” looking upwards, and, from lack of breadth, exposing the four short external branchial filaments. Within this opercular fold an oval cartilage is found (fig. 4, sp.c); this is the ‘ spiracular cartilage,” and corresponds to the apex of the mandibular bar of an embryo Salmon (Salmon’s Skull,” pl. 1. fig. 7, mn). Compared with the Urodele or Batrachian, this rudiment must be considered to correspond to the “ otic process,” and not to the * pedicle” (Huxley, 4, p. 42)1. On each side, below this spiracle, its operculum and its contained cartilage, the wall of the mouth (m) is thick, and strongly bent upon itself; this thick part extends from the chin behind to the front of the mouth and the nasal sacs. ach lateral cartilage is a thick half-link. Were the two confluent at the mid line, they would make a transversely placed link or oval ring; but they lie askant, the forebent part being higher than the part behind the mouth. The fore part is flatter, and the hind part thicker than the rest; but they both are widened at the end, and turn, in a pedate manner, outwards. Behind their middle the cartilage is becoming somewhat loose in texture; a transverse cleft has begun in its substance ; and this evident break in the cell mass has its concave margin looking forwards. Behind this line we have the rudiment of the lower jaw, in front of it the upper; and the latter will have a convex end, to fit into the concavity of the former. The upper jaw is called the “ pterygo-quadrate ” bar ; for the quadrate region has its apex indepen- dently developed, and the large foregrowth of the bar is built into the whole side of the palatal ceiling. The mouth, thus encircled and fringed in front by the slightly bilobate fronto-nasal process (fig. 2, f:7.p), is ear-shaped, a long transverse oval with a short anterior expansion. Considered as the first branchial arch, which in reality it is, we thus have an inde- pendently developed upper segment and a commencing division of the main bar into an “ epibranchial ” piece, which grows forwards and inwards, and a “ ceratobranchial,”’ which grows downwards and inwards. The next arch is similar to the last; it is the hyoid, and also has an oval cartilage above, and a long bar lower down, also strongly bent upon itself. * See also the adult Skate’s skull (Pl. XLI. fig. 4, mt.pg). OF THE SKULL IN SHARKS AND SKATES. 217 A difference is to be noted at once; for the spiracular cartilage is developed in the posterior fork of the first visceral arch, and this nucleus in the anterior fork (Pl. XXXV. figs. 1 & 4). The mandibular nucleus looks outwards and backwards (sp.c); but this is turned directly forwards at right angles to the main bar (fig. 4, h.m, hy. Compare also here the adult skull, Pl. XLI. fig. 4, mtpg, hm); the “metapterygoid” is the same as the spiracular cartilage in the Skate). The main hyoid bar is a slender tape of cartilage, pointed finely at each end, and far from the mid line at both ends; it is undivided at present. The next five arches are the ordinary branchials: the first is much larger than the hyoid; and then they gradually lessen, the last being very small. These arches are strongly bent upon themselves, the pharynx of the embryo Skate being a depressed pouch with extensive lateral fissures or clefts (figs. 1, 2 & 4). The main part of each arch is pointed above, the point looking slightly forwards, and blunt below, this rounded lobe looking backwards; both ends come much nearer to the mid line than in the hyoid arch. In the roof of the pharynx, where the front clefts close, there is, just above each main branchial, an independently chondrified piece—the “ pharyngo-branchial ” ( p.br); each piece is lunate, pointed outside, blunt within, and having its point turned more backwards than outwards, although their general direction is transverse. These free cartilages cannot very safely be considered as the serial homologues of those above the two first arches; they are not at the side of the face, but right beneath the edges of the parachordals, where the leashes of nerves are given off (figs. 3 & 4). A separate cartilage, developed at the end of the pterygo-quadrate, or one above the point of the main hyoid, would correspond more truly. The further segmentation and metamorphosis of the skull and its arches will now be described. I have shown above what, from the beginning, was independently chondri- fied, and now will show how the main bars break up. Second Stage: Embryo of Raia maculata, 4 inches long; body 13 in., tail 23 in. ; time from deposit of egg-pouch 3 months. This important specimen was taken for me from the Brighton tank by the same valued friend, Mr. H. Lee; and although the time of its growth was less than twice that of the early specimen, the development and metamorphosis was quite perfect— that is, as to chondrification and segmentation. There is much that is instructive to the morphologist in the external characters of this embryo (Pl. XL. figs. 1 & 2, drawn as far as to the umbilicus, w). In front the beak has become fixed to the anterior angle of the outspread, gigantic, flabelliform pectoral fin (p,f’), which is seen curling round the depressed cheek. Above, the beak is seen to be separated by a deep crescentic sulcus from the rounded cranial sac ; and 218 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT the flattened eye-balls are carefully lodged in sockets, the inner half of which is formed by the semicircular superorbital cartilages, the horns of which are grafted on to the nasal sac in front and to the auditory sac behind. Outside the posterior horn of the superorbital (sphenotic process) is the first postoral cleft, or spiracle (Pl. XL. figs. 2 & 3, cl 1); its narrow inferior part has been filled in, and only the upper expanded end has kept open—widely open, however, and showing on the spiracular or metapterygoid car- tilage a comb-like ‘“‘ pseudo-branchia ” (ps.dr), composed of about eight branchial papille. Behind the spiracle the gill-bearing hyoid and the proper branchial arches are seen elegantly spreading into a U-shaped system of pouches, growing, as it were from a mas- sive stalk, the occiput and spine. Below, behind the basitrabecular beak, the fronto-nasal process (n.f.p) is very large and persistent, forming a free, anterior, emarginate lip, with the valvular nasal openings (na) in their primordial ventral position, with enclosed “ labials” to guard the openings. The angles of the mouth are lipped at right angles to the transverse inferior oral opening; but the mandible has its gums and teeth quite bare. The angulo-labial fold is continued backwards as part of the general opercular skin, which further backwards and outwards is imperfect in five places on each side. These places are the branchial clefts; they are the retained lower ends of the huge primary clefts. The general oper- cular fold which has covered the open grating to so large an extent is seen to develop a special ear-shaped flap to each of the branchial outlets. Embryos as large as this, when carefully examined, have protruding from their branchial clefts what might, at first sight, be mistaken for a parasitic growth of fila- mentous conferve. These, however, are the still retained external branchie. They have begun to shorten in the first opening; but most of them are very long (Pl. XL. fig. 1, e.br). They are spatulate at their free ends; and their single vascular loop is still functional. Between and behind the branchial apparatus is the large, short umbilicus (w), which connects the embryo with a yolk the size of a dove’s egg. The chondrocranium, with its appended basketwork of visceral arches, is now com- plete, both as to chondrification and segmentation. On the whole very similar to that of the Shark at the same stage (Pl. XXXVI. figs. 3, 4, 5, and Pl. XL. figs. 4, 5, 6), it yet differs in several important points. The cranium seen from below (Pl. XL. fig. 6), with the eye-balls removed, is seen to be composed of four pairs of primary elements. Behind, the “ parachordal ” cartilages that invest the notochord (i v, nc) have coalesced with the hinder pair of “ paraneural” capsules, those of the ear. This four-sided shorter hind part of the basis cranii has coalesced in front with another pair of outspread cartilages—the trabecule (tr); and these, also, in front have coalesced with the foremost paraneurals—those of the nose (na). Where the four basal plates meet, the internal carotid arteries (7.c) enter; and in front of these is the appearance of a space and a slit, not so densely chondrified : OF THE SKULL IN SHARKS AND SKATES. 219 these are marks of the primary pituitary space (py), the soft tract between the bowed edges of the trabecule. The trabecule have developed a common basal piece, the “basitrabecular ” rostrum (d.tr); but this is not segmented off as a distinct bar. Where the trabecule have so completely coalesced in front, forming also their basal rostrum, they turn upwards into the frontal wall of the face, and finish the cranial floor. ‘The nasal sacs are hollow inverted cups of cartilage, with their downturned mouth stopped largely by “labials.” The rim of the cup is strong on the outside, and also gives attachment to an ethmo-palatine cartilage (a.0). The roof, also, of the nasal sac is modified in its form by reason of the engrafting upon it of a large bowed cartilage, the “ superorbital ” (fig. 4, s.ob), a cartilage which has but little independence of growth, but the substance of which early appears in the embryo. The lower edge of this cartilage, mesiad of the eye-ball, is continuous with the trabe- cular crest; and the “tegmen cranii” grows directly from it towards the mid line of the roof. The roof and side walls are analogous to the upper portion of a vertebral arch. Here, however, in the Skate, the tegmen is largely undeveloped; the brain-sac is permanently membranous above in front, and rests upon the laminar trabecule. Then, just in front of the nasal region, there is a cartilaginous beam thrown over; but it is narrow, and thence to the ear-sacs the roof is bare of cartilage. Behind, the tegmen reappears, and helps the superoccipital to roof in the hinder brain between the auditory masses (aw). The superorbital are grafts itself also on these sacs; hence the compound region, which ossifies separately in the “ Teleostei,” behind the orbit, the so-called postfrontal or sphenotic. The clear but cheese-like cartilage shows the three canals through its walls (a.s.c, h.s.c, p.s.c); and where the anterior and posterior of these unite, the “ aqueeductus vesti- buli” is seen open. Behind, a mass of notochord is still to be seen, and the parachordal cartilages project backwards outside, to form the occipital condyles (0¢.c). The epiotic elevation over the junction of the anterior and posterior canals is slight; the pterotic ridge outside the horizontal canal is well developed. The oral and pharyngeal visceral arches (the first or trabecular has been described as part of the chondrocranium) are nowhere more instructively developed than in the Rays. Keeping the eye upon the early condition of these parts (Pls. XX XV. and XXXIX.), we shall see what metamorphic results have been brought about. The apex of the first postoral or mandibular arch has been developed as a distinct crescentic cartilage, the “spiracular cartilage” or metapterygoid (Pl. XL. fig. 4, mt.pg) ; it is attached below the sphenotic process, and behind the fifth nerve; it is the bearer of the “ pseudo-branchia ” (ps.b7’), lies in the anterior wall of the first cleft, and answers to the “otic process” of an Amphibian. This detached suspensorium is joined to the quadrate region by a ligament ; it answers to the hinder fork of the visceral rod. The larger anterior fork, the posterior extremity of which is part of the main descending bar 220 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT (the quadrate, g), has grown forwards, and has become a large upper jaw, the “ pterygo- quadrate ;” whilst the remainder of the primary stem hinges with the outturned end of this as the lower jaw—articulo-Meckelian (Pl. XL. figs. 4 & 5, gpg, ar.mn). These two arches (two by metamorphosis) in rest are but slightly bowed forwards, where they meet their fellow bars of the opposite side; they are almost directly trans- verse, and conform, as every thing else does, to the flat outspread form of this peculiar type of Fish. The second postoral (hyoid) was also forked above in the early embryo; the further development of that arch is similar to what takes place in the Teleostean (e.g. Salmon): but it is arrested at a lower morphological level. The anterior fork of the second postoral articulates with the auditory capsule beneath the pterotic ridge (“ tegmen tympani”) by an oblong condyle above (Pl. XL. fig. 4, .m), whilst below it becomes detached entirely from the rest of the bar, turns forwards close behind the spiracle, which it protects, as the metapterygoid does in front, and then fastens itself strongly by ligamentous fibres to the quadrate, becoming its new additional suspensorium. The hinder fork, or proper apex, is wholly freed from the hyo-mandibular ; its upper piece, or “ epihyal,” is in reality only half the epihyal region, having lost the hyo- mandibular wedge. Thus the expanded apex of the archis cleft obliquely, as in Saurop- sida and Mammalia, and not from top to bottom, as in Teleostei, nor simply across without subdivision of the “ epihyal” region, as we have just seen in the Dog-fish. The bar, freed from the anterior fork, is now developed into a rather feeble branchial arch ; it is attached by ligament to the end of the jutting pterotic above ; and the upper or epihyal segment is exactly like the epibranChials (Pl. XL. fig. 4, e.hy, e.br), save that it is smaller than most of them, and, being attached to the corner of the skull, sends no “pharyngo-pleural” segment over the pharynx. There is no secondary cartilage developed in the attaching ligament, as in Teleostei; the “interhyal” (‘stylohyal,” Cuy.) is fibrous. The ceratohyal (c.hy) is feeble, but normally branchial in character; the arch is finished below by a small styliform hypohyal segment; this is attached by ligament to the first hypobranchial (fig. 5, h.hy, h.br 1). Here we see a vast difference between the Shark and Skate; for in the former the stout two-membered bars of the hyoid have their ventral ends strongly articulated to a large basihyal piece; whereas in the Skate the lower ends of the hyoid are nearly as far apart as the breadth of the transverse mouth. The five branchial arches (Jr 1-5) are very uniform, only decreasing gently in size from before backwards. Each is composed of a superpharyngeal, apical piece, the pharyngo- branchial, an epibranchial, a ceratobranchial, and a hypobranchial on each side (p.r, e.br, c.br, h.br). These arches are strongly bowed outwards, and bent on themselves; their lateral parts are thick, and grooved externally. A single series of cartilaginous OF THE SKULL IN SHARKS AND SKATES. 221 branchial rays proceeds from each; these are pedate at their outer ends. The ventral segment (hypobranchial) is more or less adze-shaped (fig. 5, f.dr); but this obtains only in the second to the fourth. The first has this segment very long, at first dilated, and then very slender, and the right and left are early fused together. The last branchial arch has its pharyngo-branchial united to that of the fourth, and its hypo- branchials (/.d7. 5) completely united. These fused elements look like an azygous piece; but in the adult of &. clavata (Pl. XLII. fig. 4, h.br 5) they are very partially united, and, as to form, are seen to be only a modification of the adze-shaped or fan- shaped type. At this stage the branchial arches, including the hyoid, carry two sets of branchize in full function. The eaternal are at their highest development; and the internal plates are perfect, although small. The “extrabranchials” are absent, as far as I can make out, in different kinds of Skates’ (2. clavata, R. maculata, &c.); nor is there a labial on the mandible. But there are four pairs of preoral labials—three acting as nasal valves (/ 2, 3, 4), and the pair (1) attached to the side of the “rostrum,” but not riding upon the nasal sac, as in the Dog-fish. Third Stage: Embryos of the Thornback Skate (Raia clavata) nearly ready for exclusion JSrom the egg-pouch. In the last stage the metamorphosis was complete; this third stage is given for the sake of the vertical and transverse sections, which reveal the architecture of this kind of chondrocranium. A longitudinally vertical section with the brain removed (Pl. XLI. fig. 1) shows a hollow barge-like structure, with a cartilaginous bottom perfect, and projecting as a free prow—the basitrabecular rostrum (0.tr)—whilst the ‘‘ deck” is only cartilaginous fore and aft over the ethmoidal and the auditory region. The various nerye-outlets (1, 2, 5, 7, 8, 9, 10) are easily recognized. ‘This little boat is undergirded by transverse bars that appear in the section—the pterygoid, mandibular, and first and fifth hypobranchials (¢.pg, mn, h.br 1, h.br 5). The first transverse section shows the nasal caps or domes (0/) far apart; these are connected by a large bridge of cartilage, convex below and concave above. ‘Thus there is formed a large precranial space ; for the frontal skin is convex. An inner labial (72) is seen in section ; the palatal skin follows the convexity of the car- tilaginous bridge; that bridge is formed by the trabecule and their commissure (#7.cm). Here are the very elements of the nasal septum and roof of the higher types; but the trabecule have only united below; they are far apart; and their crest applies itself normally to the inner edge of the nasal dome, yet forms a structure widely different from that which obtains when these domes are closely adpressed, and the trabecular 1 In the Torpedo the dilated ends of the branchial rays unite outside the pouches in such a manner as to form a practical “ extrabranchial” band (Gegenbaur, pl. 13. fig. 3, pl. 20. fig. 1). vou. X.—part Iv. No. 5.—WMarch 1st, 1878, 21 222 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT crests close together, ready for fusion at the mid line: the distinct antorbital (a.0) is cut through. In an antorbital section (Pl. XLII. fig. 1), the nasal dome is seen to be thickened by the superaddition of the engrafted superorbital arc; and, in this its hinder part, the nasal sac is some distance from the trabecule and the cranial cavity. The trabecular crest has also received increment from the same “ brow” of cartilage (s.0b) ; and thus the deck is partly covered in—entirely for a short distance, a little further back. Beneath, the pterygoid undergirders are seen, where they join by a fibrous band. A section (Pl. XLII. fig. 2) behind the mouth (m) and through the eye-ball (¢) shows the squared form of this little barge, also that the brain at the time of hatching well fills its cavity. The upper third of the “ orbito-sphenoidal” side wall is due to the thick brow-cartilage (s.0b), the rest to the trabecular crest; the floor, thicker at this part, is a trabecular commissure. The quadrate part of the upper jaw is here cut through, and also the whole extent of the articulo-Meckelian bar, with its distinct symphysis, its projecting angle, and its cupped articular surface for the rounded quadrate condyle. The mouth (m) is shown as in a nearly closed state; it has much mobility, because of its free double suspen- sorium, its well wrought articulo-quadrate hinge, and its anterior and posterior fibrous symphyses at the ventral extremities of the two arches. The last section (fig. 3) to be described is through the auditory sacs, somewhat obliquely, near their posterior wall; for it can be seen that the occipital arch is here complete, crowned with a veritable spinous process, having neurapophysial sides that form a gothic arch, and a broad base, on which lies the diminished and fading cephalic notochord. On the right side the horizontal canal (/.sc) is cut through where it overarches the condyle for the hyomandibular (fm). On the other side, the posterior canal is cut through, the razor cutting further back; and the “interhyal” ligament is seen carrying the epihyal (e.hy) with its three branchial rays, below which part of the ceratohyal (c.hy) is seen, with three more rays. On the right side, under the hyomandibular, most of the ceratohyal is seen (chy) with its flat rays, and below it the little hypohyal (h.hy) attached to the long slender hypobranchial belt of the first proper branchial arch. This figure shows the breadth of the pharynx whilst in a state of rest. Fourth Stage: Skull of adult Thornback Skate (Raia clavata). The skull of the adult Skate may be said to be fiddle-shaped (Pls. XLI. and XLII.); for it has very pinched sides, is very flat, and has a long stem or handle. The narrowing of its body is to make room for the eye-balls; the bulging parts are the auditory sacs behind, and the wider nasal sacs in front; and the projecting shaft is the enormously developed basitrabecular rostrum, which in this and related types acquires its uttermost development. The interauditory part only of the base of this chondrocranium is formed by the “ parachordal ” cartilages’; from the postpituitary wall to the end of the snout all the OF THE SKULL IN SHARKS AND SKATES. 223 rest is trabecular. On the upper surface it is easy to see how much building-material has been used to form the superorbital arcs; whatever their morphological nature may be, they are parts that do not chondrify separately, but are very distinct in early embryos, as I have shown in my paper on the Salmon. Behind, the crested form of the auditory masses is seen to be largely due to the form of the great loops into which the internal sac is developed: the primary opening, or vestibular ‘‘ aqueduct ” (aq.v.) 1s only covered by skin. The interorbital space above is still retained as a large oblong fonta- nelle; then there is a short and narrow bridge of cartilage, and then, in front of that, the lanceolate opening into the great precranial or internasal vacuity. On each side of this gaping space the trabeculz, as they melt into the rostrum, grow over as an inter- nasal eave. These ascending and upper growths of the trabecule represent, by a sort of morphological hypertrophy, the trabecular crests of a bird or mammal that form the lower portion of the mesethmoid. The nasal sacs, very large and wide apart, are modified from their original dome-shape by the condyle on their outside for the rib- shaped antorbital (a.0), and by the addition of a new stratum of cartilage yielded by the great superorbital arc. The same arc still shows its form behind, where it has thickened the periotic mass, and formed the sphenotic process (sp.o). Between the superorbitals the skull is flat above; and below, the basis cranii is only very gently convex. The occipital condyles (0c.c) project but little; and a tract of notochordal jelly is interposed between the median part and the first vertebra. The nerve-outlets are most of them best seen in the side view (Pl. XLI. fig. 4. 2,5, 7). The peculiar mode of suspension of the double dentigerous arch is best seen in the side view (Pl. XLI. fig. 4, q-pq, mn, mt.pg,hm); the hyomandibular running forwards to the base of the metaptery- goid, or spiracular cartilage, to be attached by ligament to both the quadrate and the angle of the mandible. The loose spiracular cartilage bears but little of the weight of the jaw; of which it is the proper apex. In the lateral view the arches are compressed as much as possible for display; in the upper and lower views they are in a state of rest, and therefore greatly depressed. The pterygo-quadrate bar is attached in front by ligament behind the nasal sac; and the ligamentous fibres attached to the round inferior end of the oblique subtrihedral hyomandibular spread fan-like, to insert themselves on the top of the quadrate and the angle of the mandible. The latter is hinged in a somewhat complex manner, a ball on the quadrate slipping into an articular cup, which has three nodules on its rim, whilst the whole mass of the angular region is thin-edged. The articular part of the upper bar—the quadrate—is separated by a cleanly made neck from the proper pterygoid region; but no inspection of the parts in the adult would have led to the discovery of the fact that this upper dentigerous bar, in its toothed part, is a mere process or fork of the mandibular or first postoral bar. The posterior part of the hyoid is seen to be jointed by its epihyal part (¢.hy) to the end of the condyle of the hyomandibular; the ligament is almost the length of the so-called 212 224 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT “ stylohyal-” (“interhyal”) of the Teleostei; but it has stopped in its metamorphosis at that point where, in the osseous fish, there are two subequal hyoid arches (“‘ Salmon’s Skull,” Third Stage, pl. 2. figs. 6 & 7, hm, c.h). In the Teleostean the ligament would be more than halfway down the side of the hyomandibular (three fourths in the Salmon, ibid. pl. 6. figs. 1, 2 (im, st.h). Also in the osseous fish, a nodule of cartilage appears in the ligament, like a rudimentary meniscus, which grows into a terete rod and ossifies (ibid. s¢.h). The small epihyal (e.hy) is not only attached by its apex to the hyomandibular, but also it is closely tied to the front of the first epibranchial at its upper third ; this gill- bearing bar is succeeded below by a ceratohyal of equal size (c.hy); and then there is below it a small styloid hypohyal (/.hy), situated on the notched proximal end of the first hypobranchial (h.br 1). At first sight it would seem as though the hyoid had formed a commisural band across behind the mandible (Pl. XLI. fig. 4, 4.471); but that is the belt formed by the fusion of the first hypobranchials. The normal direction of the pharyngo-branchials (p.br) is backward ; that of the fourth is continuous with the epihyal of the fifth, as there is no separate pharyngo-branchial in the last arch. The epi- and ceratobranchials are tolerably stout, flat within, grooved outside, and strongly tied together by fibrous bands. The arches end below in flat, kidney-shaped, adze-shaped, and even flabelliform hypobranchials (h.dr). That of the last coalesces in some degree with its fellow ; and the notched and split plate thus formed reaches by its fore horns nearly to the hypobranchial girder of the first arch. About twelve or thirteen branchial rays are attached to the hinder edge of each arch (see Pl. XLI. fig. 5, br.7, for those on the hyoid). These are pedate at their distal ends, which lie in the outer wall of the sac ; and in some of the larger Skates their extremities are very large and lobate. It is impossible, in the Skate, to prove that the whole of the trabecular growth is not axial. Attached to the nasal sac of each side is a large, solid, short, rib-like antorbital (a.0), evidently a preoral visceral bar. The “labials” are the only “ extraviscerals” found by me in this type. There are four on each side; and the fourth is but slightly connected with the pterygo-quadrate bar (Pl. XLII. fig. 4, 7. 4). The first pair (/.1) are lanceolate and notched in front; they help the “ cutwater’ or “rostrum ” very little, being carried away far from the nasal sacs, on which they are mounted behind in the Shark (Pl. XX XVII. fig. 2,71). The second labial (/ 2) is in front of the nasal opening ; the third (/ 3) outside; the fourth! (/ 4) is articulated to the second, and lies inside below the rim of the cup and the fore edge of the pterygoid bar. ? 1 This fourth labial of the Skate does not answer to the fourth of the Dog-fish, but to the inner of the two on the pterygo-quadrate of Scymnus, Squatina, and Centrophorus (Gegenbaur, pls. 11 & 12, L; the next is marked L’, OF THE SKULL IN SHARKS AND SKATES. 225 These thin valvular cartilages are of great interest to the morphologist : the second and fourth are exactly repeated in the Snake tribe; the fourth is often large in birds (e.g. Rhea, Turnix, some “ Picide,” and the “Passerine” generally). The fifth, or mandibular labial, which is absent in the Skate, but present in the Shark, I found thirty-five years ago in the Coot, and later in the Gallinule. The third labial (73) has only to be compared with the “ appendix ale nasi” of an embryo mammal for the two to be immediately recognized as representing each other. SUMMARY AND CONCLUSION. The skull of the Selachians may be expected to yield much instruction to the mor- phologist. I shall compare the varieties of it seen in that great group with each other, and then with the skull of various other types. A. The skulls of the Dog-fish and the Skate as compared with each other and with what is seen in the Selachians generally. The Dog-fish and the Skate are representatives of the two main divisions of the Elasmobranchii, and are fairly typical of the two suborders. In their outer form the Sharks differ but little from that of an embryo Teleostean at about the time of hatching; the Skates mask this form by their extreme flatness and the huge expanse of their outspread pectoral fins. Their embryos differ but little in outward appearance, as Mr. Balfour’s figures (Q, pl. 25) and mine clearly show; but as soon as the skeletal elements can be traced their divergent development can be seen. Besides the long external branchials of the hyoid and branchial arches, each type shows four free external branchiz emerging from the spiracle ; and the papilla actually developed are twice that number. The postoral arches are, normally, seven; but there are e7ght in Hexanchus and nine in Heptanchus. One important difference has been shown in this, namely that the postoral “ extra- viscerals,” which are so well developed in the Dog-fish, do not appear in the Skate. Another contrast lies in this, namely that the hyoid rays are more or less split or digitate in the Sharks and undivided in the Skates. As a rule, the cartilage developed in the spiracular operculum in the Shark (it is small in Scyllium canicula) is a vay (or rays); in the Skate it is part of the body of the arch. There is one ray in Sqguatina, Mustelus, and Galeus; two in Scymnus, and three in Centrophorus (Gegenbaur, pls. 11 & 12). In the Torpedo (ibid. pl. 13. fig. 3, Zr, a.b) there are three small subsidiary cartilages besides the main “metapterygoid” segment in front of the spiracle. In Cestracion (Huxley, 4, p. 42, fig. 8, ot.p) the spiracular cartilage is a free “ otic process,” or meta- pterygoid. ‘The transverse position of the mouth, which is so perfect in the Skates, is much more oblique and projecting in the Sharks, and is more like what is seen in 226 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT other fishes in that respect; in Hexanchus and Heptanchus (Gegenbaur, pl. 10) the mouth is very Batrachian. The upper fontanelle is more completely closed in Sharks than in Skates, although some of them have a nearly perfect “ tegmen” (ibid. pls. 7, 8, and 13). The “cutwater,” or facial rostrum,‘is least developed in the Sharks, and attains its highest development in Skates, especially in Pristis. In Sharks the nasal domes approximate in the adult; in Skates they are permanently far apart (ibid. pls. 7, 8, 13, 14). The “ aqueduct” leading to the ear-labyrinth is seen in the roof of the skull in both types ; and they agree in a large number of characters, as the double occipital condyles, &c. But their points of nonconformity are of the highest interest, and this especially in regard to the “ visceral arches.” The structure of this group seemed to me for some time to be most conclusive against the theory of the independence of a palatine arch in front of the mouth, as the pterygo-quadrate arcade is in them manifestly the foreturned upper region of the mandibular arch, or a huge outgrowth or process from that arch. But much comparative study of the Selachian skull and that of the Amphibians has shown me that I had been missing the true “ ethmo-palatine ” element, a very distinct thing from the pedate process of the quadrate or mandibular pier. In some Sharks, and in all the Rays, a rib-like cartilage grows in front of the eye on each side, either attached to the nasal dome itself or to the lateral ethmoidal region. In many of the Sharks it is exogenous, and does not exist in the form of a separate carti- lage; but it is much more clearly seen in the embryo than in the adult (Pl. XX XVII. figs. 1 & 3, a.o). It is most distinct in /Zeptanchus, and is very definite in Hexanchus (Gegenbaur, pl. 1. figs. 1 & 2,1). The process can be seen in Acanthias (ibid. pl. 2. fig. 3, m') ; but all Gegenbaur’s figures show in the Rays what I have found in Raia maculata and clavata—namely, a large antorbital or ethmo-palatine cartilage, whose title to be called a rudimentary visceral arch I shall discuss anon. The trabecule, up to, or even between the nasal sacs, must be considered to be cranial and not facial ; yet in front they send out three facial “ processes,” that in an exogenous manner represent visceral arches. Thus it appears to me that there are visceral rudiments in the face both before and behind the nasal capsules. That these arrested arch-piers derive their nervous supply from the huge crowded nerves that also freely grow down into the postoral region, cannot surely tell against their ventral or visceral character; they are aborted or arrested piers, and have no free inferior arch, like the mandible and the hyoid cornua. The sharpest contrast between the Shark’s and the Skate’s facial basketwork is seen in the manner in which the hyoid arch becomes segmented and specialized. In Scyllium canicula, as we have just seen, both the primary mandibular and hyoid OF THE SKULL IN SHARKS AND SKATES. 227 arches are bent forwards above, and they both simply become segmented at the bowed part into an epi- and ceratopleural element. In the first arch the epipleural element is the pterygo-quadrate, and the cerato- pleural the free mandible; in the hyoid the epipleural is the hyomandibular, and the ceratopleural the free hyoid cornu (Pl. XXXIV. fig. 1, Pl. XX XVIII. fig. 2). This is a very simple piece of morphology; and if the modification of these arches had stopped here their meaning would have been evident. But in the Skate we have at once a hyoid arch as difficult of interpretation as that seen in so many higher types of Ichthyopsida, and of the air-breathing Vertebrata generally ; so that the first, as it were, of a whole series of puzzles is set before us. In the Shark the first and second arches are merely “ branchials,” without distinct pharyngo- or hypopleural elements. In the Skate the hyoid arch abutting above against the skull does not grow over the pharyngeal roof, as in the succeeding arches (Pl. XXXV. fig. 4, hm, hy), and has there- fore no pharyngo-pleural element. But in the forked expansion formed by the primary bar, cartilage commences in two places—a little nucleus in the front fork, and the apex of the main bar in the hind fork: thus the cartilage of the epipleural region is primarily double. A similar puzzle offers itself in the first arch; but the order is inverted: in it the hind fork has its own little nodule of cartilage, and the main bar runs in the front fork; thus the epipleural region of the mandible has two sources of cartilage—a small hinder part, and a large (the main) front part. In the Dog-fish the two hyoid elements get close within the articular region of the mandible, and are strongly strapped to it by a hyo-suspensorial or symplectic, and a -mandibulo-hyal ligament. In the Skate the little front cartilage of the hyoid arch becomes the large hyo- mandibular, being loosely connected with the rest of its own arch, and having the whole mandibular apparatus suspended to its distal end, so that the mandibular arch is “hyostylic,” as in the Shark; yet it is not only suspended on the upper element, but has also a feeble metapterygoid or otic suspension ; it is somewhat “ amphistylic.” B. The Skull of the Dog-fish and Skate, as compared with that of the more generalized Selachians, of the Chimeroids, of the Dipnot, and of the Amphibia. All these types may profitably be compared together; much, however, of this work I find admirably done to my hand, in Professor Huxley’s paper on Ceratodus (P. Z. S. Jan. 4, 1876"). 1 The harmony between the author of that paper and the writer of this is almost perfect. One /ittle and one great difference of opinion exist: namely, his “angulare” in Ceratodus and the Amphibia is my “ arti- ticulare,” p. 34; and the second, or great difference, is the relegation of the trabecul by him to the “pleural” elements (p. 32). I have shown in the present paper my own change of opinion, and now consider them to be pro-parachordal tracts, ending in exogenous ‘‘ pleurals.” 228 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT With regard to the Amphibia, however, there are some points which I have lately discovered that are of great importance. As to the abortion of the upper part of the suspensorium in the Selachians, in such contrast with its trifurcate condition in the Urodeles, and its bifurcate condition in the Batrachia, I find myself, even in details, in the happiest conformity of views with Professor Huxley. As to Cestracion (Huxley, 4, p. 42, fig. 6), I quite agree with him that the spiracular cartilage is the separate ‘ otic process” (of.p); and it is worthy of notice that whilst the spiracular cartilages of the Sharks are mere rays, yet they are chondrified detach- ments of the thin edge of the tissue that fills the primary fold in front of the first cleft. In the Batrachia the annulus tympanicus is a cartilage; and at first, in the Tadpole, it is a process from the angle of the suspensorium ; it becomes then a free ray, and then curves to form the tympanic ring—the analogue but not the homologue of the ‘“‘annulus” of Man and the other Mammalia. In some Urodeles this cartilage reappears, and in them forms an attachment to the stapes above the seventh nerve, or portio dura; this is seen in Menopoma, Spelerpes, Desmognathus. It imitates the “columella,” but is not that organ, only a curiously specialized homologue of the spiracular cartilage of the Shark and the annulus of the Frog. As to the rudiments of the pedicle and otic process seen in some Selachians, I quite agree with Professor Huxley, who shows both in an embryo of Notidamus cinereus (4, p. 44, fig. 9, p.st.p). Gegenbaur (op. cit.), in his exquisite figures, shows these processes, notably in Hevanchus and Heptanchus (pl. 10. figs. 1, 2, 2', p); but they are very evident in Scymnus and Squatina (pl. 11. figs. 1, 2), and are still more clear in Centrophorus (pl. 12, fig. 1). I am also quite satisfied, from the study of a large number of Amphibian skulls (larval and adult), that the “pedicle” is the true primary head or apex of the sus- pensorium. : There are some curious points in the structure of the palatal bars worthy of note. Gegenbaur (op. cit. pl. 11. fig. 1, p) figures in Scymnus a keystone piece to the pterygo-quadrate bars in front. In certain birds, especially the Picide, the two palatine bones are united by a tract of thin cartilage, which ossifies as a medio-palatine bone (Trans. Linn. Soe. ser. 2, vol. i. pl. 3, m.pa). I spoke of the absence of a ‘‘ pharyngo-pleural” element in the upper mandibular segment of the Selachians; I have described such a cartilage attached to the fore end of the pterygo-quadrate in a young specimen of the Aoxolotl (Siredon). A similar cartilage is very constant in the palate of Passerine birds (‘+ Aégithognathe,” part 1. pl. 55. figs. 1, 13, ¢.pa); but this, in both cases, may belong to the ‘‘ antorbital.” OF THE SKULL IN SHARKS AND SKATES. 229 In Birds an “os uncinatum” also often occurs, which is evidently the true homo- logue of the antorbital or ethmo-palatine of the Skate and of the Amphibians. Birds also possess “ labials.” But, to return to the Ichthyopsida, I may mention that the Batrachia, which, like the Selachians, appeared most evidently to show that there was no proper distinct palatine arch, can now be cited as yielding what appear to me to be convincing proofs of the existence of such an arch, whose suspensorial point is the ethmoidal projection of the trabecula. At first the suspensorium of a Batrachian is quite distinct from the trabecular bar; but soon after hatching it becomes attached both before and behind—before to the trabecular elbow, and behind to the ethmoidal region of the bar (see ‘ Frog’s Skull,” pl. 5. figs. 1-4, pg). If, in this condition, segmentation had taken place across this short conjugational bar, the upper half would have belonged to the trabecula, and the lower half to the suspensorium. Instead of this, in the Frog, it grows into a long arch, which projects forward beyond its ethmoidal part, and is continuous with the suspensorium behind. In the Toad, however, (see “‘ Batrachian Skull,” part 2, pl. 54. figs. 3, 4), it is seg- mented off, not only from its ethmo-trabecular attachment, but also from the pterygoid process of the suspensorium, and becomes what its counterpart is at first in the Urodeles and Skates, namely a distinct ethmo-palatine visceral piece. C. Comparison of the Selachian Skull with that of Ganoids and Teleosteans. In the lower kind of Ganoids, such as the Sturgeon (see Month. Micro. Journ., June 1, 1873, pl. 20, pp. 254-257), the skull becomes hyostylic in the highest degree, and the hyomandibular has its lower third segmented off, and separately ossified as a large symplectic segment. The pterygo-quadrate cartilages are not unlike those of the Skate, but more arched, ultimately having three bony tracts in them, namely the pterygoid, palatine, and meso- pterygoid. But there is only a single counterpart of the two spiracular cartilages. It is an arched or convexo-concave plate, lying over the supero-posterior part of the tubular mouth; it is rounded and thick behind, and thin and angular in front, where it fits in beween the right and left pterygo-quadrate bars; it is a double “ pedicle.” In Polypterus (Traquair, Journ. of Anat. and Phys. vol. v. plate 6), the pterygo- quadrate is a large bar of cartilage more or less ossified by a metapterygoid, quadrate, pterygoid, and mesopterygoid centre. But the top of the metapterygoid is very low down ; and the hyomandibular has a small postero-superior osseous centre, and a large bony tract which takes in all the rest of the bar, without an inferior “symplectic” bony centre. VoL. X.—PArRT Iv. No. 6.—Warch 1st, 1878. 2x bo 30 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT The palatine is a small ossicle, placed unconformably to the rest of the bar, at its eatreme fore end, and in its transverse position answers to the ethmo-palatine bone of Amblystoma and the Batrachia. In the Salmon there is much that is in conformity with what is found in the Rays, but greatly modified. The palato-quadrate arcade (“Salmon’s Skull,” plate 6. fig 2) is let down, so that the top, or otic process, is halfway down the hyomandibular bone; and this process is occupied by a square ossification, the metapterygoid, between which and the base of the inverted triangle of bone below (the quadrate) there is a wide tract of cartilage. This tract mounts up in front, and passes into a rounded boss, the arrested pedicle. Then, further forwards, the cartilage is reduced to a very narrow tract; and this wedge of cartilage, between the mesopterygoid above and the pterygoid below, shows how far the pterygoid cartilaginous process grew from the front of the quadrate region of the suspensorium. That tract answers to the fore part of the upper jaw of a Selachian. But the rest of the bar, where the cartilage breaks away from between the two plates of bone (py, m.pg), is attached by a short pedicle to the ethmoid, and then grows forwards as a massive prepalatine rod, with a swollen boss for the head of the maxillary. All this part is formed out of the originally distinct ethmo-palatine rod (ibid. pl. 2. figs. 6, 7, pp.q). Curious and instructive it is to find that at first, as in the Skate and Urodele, this distinct preoral pleural arch is thick above and grows backwards to a point. By the middle of the second week after hatching, the fry of the Salmon shows a pre- palatine spur, which becomes so large in the adult, and which is so well seen in the separated ethmo-palatine of Bufo vulgaris. The hyostylic skull of the Salmon is easily seen to be a further specialization of what is so remarkable in the skull of the Skate—although in the details of the segmenta- tion of the top of the hyoid arch there is some difference, the hyomandibular being an independent nucleus in the Skate, and arising in the Salmon by longitudinal splitting of the primary bar (“ Salmon’s Skull,” plate 2. fig. 3, h.m, ch). In the Skate (Pl. XLI. fig. 4, 7.4.7) the posterior division of the hyoid is attached above to the postero-superior angle of the hyomandibular by an interhyal ligament. This state of things is seen in Salmon embryos just before hatching (ibid. pl. 2. figs. 6, 7, h.m, ch, third stage); but in the adult (ibid. pl. 6. fig. 2) the hinder moiety (ep.h, ¢.h) is suspended from the synchondrosis between the long hyomandibular and the short symplectic, exactly opposite the quadrato-metapterygoid synchondrosis. There this kind of specialization, the hyostylic, is carried to its uttermost degree, and the articulation of the two main hyoidean moieties is but little above that of the mandible and its pier; yet the whole is but an exaggeration of what is seen in the Skate. The structure just described may be found in by far the greater number of Tele- OF THE SKULL IN SHARKS AND SKATES. 231 ostean fishes; but in the Kel (Anguilla acutirostris) and its congeners there is a less degree of specialization of the Raiine type of face. In small, white young of that species, 23 inches in length (the gift of Mr. F. Buck- land; see ‘ Nature,’ June 22, 1871, pp. 146-148), I find the following modifications of the Teleostean face. The hyomandibular has a most extensive otic region, the front and hinder head being far apart, and under the latter the opercular knob. This transverse top grows downwards, as an arcuate rod, with its convexity behind. It is ossified by the hyomandibular centre down to the bent part; then there is a large tract of cartilage, which has its pointed forward end capped by a symplectic bony sheath, which passes inside the quadrate in front of it. On the back of the middle of the synchondrosial tract there is an interhyal nucleus of cartilage, from which a ligament grows; and to this ligament the hyoid cornu is suspended. This cornu does not separate off into a hypohyal below; and the epi- and ceratohyals completely ossify the rod, largely overlapping each other. Below its front condyle the hyomandibular sends forwards and downwards two strong sharp spines of periosteal bone ; and these embrace the ascending but lowered head of of the mandibular suspensorium, which is a shortish straight rod, unossified above and below. Its ossification is a cylindrical, ectosteal quadrate; and the apex receives no meta- pterygoid bony centre. As in the Urodeles, there are a bony and a cartilaginous pterygoid; the former is a delicate f-shaped style, pedate behind, to run up the front face of the suspensorium. The cartilaginous pterygoid (r.pg, rudiment of pterygoid) is a triangular process, growing from the front of the quadrate bar below, exactly as in the larva of every known kind of Caducibranchiate Urodele, as well as in certain Perennibranchiates ; it is also the arrested homologue of the part which, in the Salmon, coalesces with the ethmopalatine, and also of the extended limb of the “upper jaw” of the Selachian. The ethmopalatine is suppressed in the Eel, and its skull is as simple as that of a Snake or of the Proteus. But a skull whose structure has long puzzled me, namely that of the Siluroid Clarias capensis, comes in as the most demonstrative proof of the existence of a preoral post- nasal visceral arch. As in the Eel and its congeners, the hyomandibular of Clarias has an extensive otic region ; and, as in the Eel, the fore part of that bone develops an extraordinary amount of periosteal bone in front, which aborts, or coalesces with the metapterygoid. The whole pterygo-quadrate arcade is a broad, flat, almost transverse plate, not ascend- ing behind into an otic process, but being angular behind and above the quadrate con- dyle, and strongly wedged in between the foregrowths of the hyomandibular and the separate symplectic bone. 2K 2 232 MR. W. K. PARKER ON THE STRUCTURE AND DEVELOPMENT This obliquely placed plate runs almost directly forwards, and articulates by its antero-superior bony centre to the outer angle of the crescentic dentary fore edge of the vomer—a constant relation of these bones in Carinate Birds; it is seen also in Anguis and Hatteria. But the thin lamina of cartilage that formed the “ model” for these three bones— the quadrate (its front part), the pterygoid—and the mesopterygoid, is totally inde- pendent of the palatine bone; and the whole arcade is very loosely connected by liga- ment to that bone. The palatine (or rather “ ethmo-palatine”) is a thick rod of ossified cartilage, lying above the pterygo-quadrate plate, and reaching far in front of it. Behind, it ends above the fore margin of the quadrate ossification, and in front reaches nearly to the angle of the preemaxillary. In front it carries the arrested maxillary—the filament-bearer—which is wedged between it and the premaxillary; whilst the foremost of the two preorbital ossicles, that which becomes the ‘“ septomaxillary ” of the air-breathers, rides upon its fore end and also upon the fore end of the little maxillary. The most projecting part of the massive lateral ethmoid is a little in front of the middle of this ethmo-palatine bar; but the ethmo-palatine process is represented by ligament, and the bone is everywhere loosely attached to the surrounding bones. Everywhere, above the Skate and its congeners, the exoskeletal bone that is related to the “ ethmo-palatine,” as its proper splint, is the maxillary. This may be equally well seen in Teleostean Fishes, Urodeles, Anura, and Carinate Birds. In bringing out this ethmo-palatine element of the face into bold relief, I have done no violence to Nature, but only to my own confused and confusing prejudices. No “labial” element or “ extravisceral” cartilage has ever been allowed to come into my way in considering the true endoskeletal elements. I have used no cutting and con- triving in trying to arrange morphological segments; the parts are allowed to tell their own story; and listening to catch from them the least hint of their real meaning (I hold my opinions almost loosely even), I am ready to cease from the folly of my own wisdom at any moment, and adopt the truth, whenever and wherever I can unearth it}. DESCRIPTION OF THE PLATES. PLATE XXXIV. Fig. 1. First stage. Side view of an embryo of Dog-fish (Scyllium canicula), 8 lines long, seen as a transparency, X 16 diam. 1 Whilst correcting the present paper (August 25th, 1877) I find in the Tadpole of Rana clamata four small internal branchial arches; its pouched arches are “ extrabranchials,” and answer to the branchial “ basket" of the Lamprey. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. 5 Fig. oS Oe OF bo or em © De oy OF De OF THE SKULL IN SHARKS AND SKATES. 233 . Another embryo, 11 lines long, seen as an opaque object, x 14 diam. . The same, seen from below, x 14 diam. . The same, seen from above, X 14 diam. . The same, in longitudinal section, x 14 diam. . Part of fig. 1, x 24 diam. PLATE XXXV. . First stage. Side view of Pristiwrus embryo, 3 of an inch long, xX 20 diam. . The same, from below, x 20 diam. . Dissection, from above, of the same embryo, x 20 diam. . Dissection, from above, of a larger embryo (# of an inch long), x 20 diam. . Dissection, from above, of an embryo of Scylliwm canicula (13 lines long), < 20 diam. . Second stage. Dissection, from above, of a larger embryo of Scyllium canicula 13 inch long, x 14 diam. . The same, dissected from below, x 14 diam. . Part of fig. 6, x 30 diam. PLATE XXXVI. . Second stage. Side view of head of Scyllium canicula, partly dissected, from an embryo, | inch 2 lines long, x 12 diam. . Vertical section of a somewhat larger embryo, x 14 diam. . Third stage. Chondrocranium of a two-thirds-ripe embryo of the same fish, 1 inch 8 lines long, seen from above, x 10 diam. . The same object, seen from below, x 10 diam. . The cranium without the facial arches; side view, x 10 diam. . Section of head of the same, x 14 diam. PLATE XXXVII. . Third stage, continued. Skull of the same, seen from below, x 10 diam. . Fourth stage. Adult Dog-fish, upper view of the skull, x 1 diam. . The same, with facial arches removed, under view, x 14 diam. . Section of head of the same, xX 13 diam. . Part of same object, x 53 diam. PLATE XXXVIII. . Skull of adult Dog-fish, seen from below, x 14 diam. . Side view of same, x 14 diam. . End view of same, X 14 diam. . Vertieal section of cranium, x 14 diam. 234 o> oR oO bo “Io Ot PB OO MR. W. K. PARKER ON THE SKULL IN SHARKS AND SKATES. PLATE XXXIX. . Skate. First stage. Side view of embryo of Raia maculata, 1 inch 4 lines long, 7 weeks old, from the deposit of the ege-pouch, x 14 diam. . Same object, with facial arches shown, x 14 diam. . Part of same, x 20 diam. . Under view of same embryo, x 14 diam. . Upper view of same, x 14 diam. . Part of fig. 5, Plate XXXV., Scyllium, x 40 diam. PLATE XL. . Second stage. Under view of fore part of an embryo of Raia maculata, 3 months from deposit of egg-pouch, x 53 diam. . Upper view of same, x 34 diam. . Side view of same, X 7 diam. . Skull of same, seen from above, x 5 diam. . Same object, seen from below, x 5 diam. . Cranium of same, from below, x 5 diam. . Part of fig. 3, Plate XX XV. x 40 diam. PLATE XLI. . Third stage. Section of cranium of Raia clavata (nearly ripe embryo), < 7 diam. . Transverse section of same through nasal sacs, x 14 diam. . Fourth stage. Skull of adult Raia clavata, from above, X 1} diam. . The same, side view, X 13 diam. . Hyoid arch, with “ branchial rays” of same, inner view, X 1 diam. PLATE XLII. . Third stage, continued. Section through antorbital region of ripe embryo of Raia clavata, x 14 diam. . A similar section of same, through the eyes, x 14 diam. . Another section through auditory sacs, x 14 diam. . Fourth stage continued. Under view of skull of adult Raia clavata, x 1} diam. . Transverse section of skull of Dog-fish, 5rd stage, through the eyeballs, x 14 diam. . Another section, rather oblique, in postorbital region, x 14 diam. . A third section, through the auditory capsule, x 14 diam. W West & C° imp SCYLLIUM CANICULA. Trans. Loot Loc. Vol. 0: FU AKKV. W West & C° imp 1—4 PRISTIURUS, 5—8. SCYLLIUM. rans. Loolb Soo. Vol 70-PY4, XEKV WAICP del. ad nat G.West lith. WWest & C° imp SCYLLIUM CANICULA. Oe nee OR a a ee oe ¢ + s o ca e20G La. W. West & C° inp WACP. del. ad nat. G-West lith SCYLLIUM CANICULA. LAL LEI WELLL OI. evs hubri ae WWeet 8 C° imp SCYLLIUM CANICULA. dur 0) 3 392M VIDOINYO Whitt Ros "9! =~ nyvtno wie iyo a ie od oc, VV ABIL PU PR PCG PO we Or es a or, es Cy ; Bs Bans. Bolt Sse Vl GAN. Gx 5. W.West £C° imp 4 1—é€ RAIA MACULATA. 7. PRISTIURUS. hibr s WEP. del.ad not. G.We'st lith RAIA hbr1 CLAVATA . eSrans ee, Soc. Vol Vb GY aii WWest 2C° imp Trans. a0k Loe. Vol AIO ObI trom eKAE A se as WEP. del. ad not. G West lth. W. West & C®? imp. , 1-4. RATA CLAVATA, 5—7. SCYLLIUM CANICULA. [ 235 ] VI. A Description of the Madreporaria dredged up during the Expedition of H.M.S. ‘ Porcupine’ in 1869 and 1870.—Part Il. By Professor P. Martin Duncan, IB. (Lond.), F.R.S., President of the Geological Society. Received May 10th, 1876. Read May 16th, 1876. [Puates XLIII.—XLV.} ContEnts. I. Introduction. . . . . . . «. page 235 III. Descriptions of New Species, and Notes on II. List of New Species, and Table of their described Species. . . . . . page 237 @lassitication ee) 200) | Ve) General Remarks) sj.) 4p) 2) eee I. INTRODUCTION. THE first part of this description of the corals dredged up by H.MLS. ‘ Porcupine,’ under the direction of Dr. Carpenter, C.B., F.R.S., and Professor Wyville Thomson, F.R.S., was read on May 16th, 1871, and has been already published in the Trans- actions of the Society (vol. viii. pt. v.). A certain number of specimens remained undescribed on account of their admixture with other matters dredged up, and partly on account of their difficulty of determination. It was intended that not much attention should be paid to them until after more specimens of deep-sea corals should have been received, as it was manifestly desirable that one course of study should complete this by no means easy description and analysis of the corals of the deep sea. During the voyage of the ‘ Challenger’ a considerable number of species of deep-sea forms of Madreporaria have been collected; but as it appears, from a com- munication to the Royal Society by Mr. Moseley, one of the staff of the ‘Challenger,’ that the species dredged up will be described by him, I think it best now to offer this concluding essay to this Society. ‘The interesting and most valuable communication on the deep-sea corals dredged up by the ‘ Hassler Expedition, written by Alexander Agassiz and L. F. de Pourtales, has been published since the first part of this essay appeared ; and whilst my former communication was in the press I had the advantage of receiving Count Pourtales’s descriptions of the deep-sea corals collected in the Gulf- stream Expeditions in 1867-1869. Both of these works have been of the greatest use tome. I include also some notes on some of the species already described. VoL. X.—ParT v. No. 1.—Warch 1st, 1878. Pai PROFESSOR P. M. DUNCAN ON THE II. List or New SpEcIEs, AND TABLE OF THEIR CLASSIFICATION. ZOANTHARIA SCLERODERMATA. Section APOROSA. Family Turbinoliide. Subfamily CaRYOPHYLLIIN&. Genus CARYOPHYLLIA. Species Caryophyllia carpenteri.. No. 9 dredging, 539 fms. simplex. NOG « Genus BATHYCYATHUS. Species Bathycyathus minor. ~ Nomi. (acs 1095 fms. Subfamily TRocHOCYATHACEA. Genus ParacyaTHUs. Species Paracyathus insignis.“ Mediterranean. 248 fms. monilis. v of 60 fms. humilis. Ny, Me mornatus. 33 africanus.“ Tunis coast. 40 fms. costatus. J Subfamily TURBINOLIACES. Genus GEMMULATROCHUS, gen. nov. Species Gemmulatrochus simplex, v Mediterranean. Genus FLABELLUM. Species Flabellum minus. “No. 16 dredging. 994 fms. Family Astreida. Subfamily HUPHYLLIACEA, Genus BLASTOSMILIA, gen. nov. Species Blastosmilia pourtalesi. / Mediterranean. Coral zone. MADREPORARIA OF THE DEEP SEA. 237 III. Descriptions or New SpEcIEs, AND NOTEs ON DESCRIBED SPECIES. Genus CaRYOPHYLLIA. CaRYOPHYLLIA CARPENTERI, sp. noy. (Plate XLIII. figs. 28-31.) The corallum is elongate, curved, and conico-cylindrical ; the calice is slightly ellip- tical; and there is a mark of fracture at the base, the form having been adherent. There is an epitheca reaching to the calicular margin, hiding the coste; and there are prominences on the external surface, which indicate irregular growth. ‘The septa are barely exsert, are narrower than the costal ends, are not crowded, and are unequal. There are three cycles in six systems; and the pali, which are distinct, are placed before the secondaries. The septal lamin, especially of the secondaries, are wavy and bent in their course; and the tertiaries are smaller, but well developed. The columella is composed of one twisted process. Height of corallum nearly 45 inch ; diameter of calice jy inch. The specimen was dredged up with Caryophyllia inskipi, nobis, in No. 9 dredging, 2nd expedition of the ‘ Porcupine,’ from a depth of 539 fathoms. The base of the specimen was fractured as the coral was removed from the sea- floor; and the section thus shown is very remarkable. ‘The fracture is round, and shows the inside of the coral, being about 35 inch in diameter. The septal number is nearly complete there; and two of the primaries, Nos. 1 & 4, are united by a small columella; and pali exist before the secondaries of the complete systems. This species, had it no epitheca, would evidently come within Pourtales’s genus Steno- cyathus (op. cit. No. iv. p. 9); and it is in the neighbourhood of my Caryophyllia ver- miformis (‘ Deep-sea Corals,’ p. 316). The generic attributes and alliances of Caryophyllia are noticed in the Deep-sea Corals of the first voyage of the ‘ Porcupine’ (Trans. Zool. Soc. vol. viii. pt. v. p. 309). CARYOPHYLLIA SIMPLEX, sp. noy. (Plate XLIII. figs, 32-34.) The corallum is long, bent, cylindrical, and increases in diameter towards the calice, which is circular in outline, and shallow. The columella is very small; the septa are complete in two cycles in six systems; and there are a few small septa of the third cycle in some systems. The primary and secondary septa are subequal, and extend far inwards; and the pali are like continuations of the primaries before which they are situated. The septa are not exsert. The epitheca is pellicular, and its orna- mentation is of a chevron pattern; and there are faint lines of coste, which are slightly prominent here and there. The base is small, and presents a fracture of former adhesion. Height 3, inch; breadth of calice 7’g inch. Locality. No. 9 dredging, ‘Porcupine’ Expedition; with Caryophylla carpenteri, nobis, from which it differs in its epitheca, coste, smaller columella, and septal number. 2L2 238 PROFESSOR P. M. DUNCAN ON THE CARYOPHYLLIA POURTALESI, Duncan. Trans. Zool. Soc. viii. p. 317. This species was described and figured in the former communication to the Zoolo- gical Society ; and some specimens, which were not examined formerly, are now noticed, in order to explain some points of structure and its distinctness from some other species. The pali in the fully developed corallite are well developed when the columella has only one twist, and are less so when this structure is more complicated; and this peculiarity has been noticed by Pourtales in Caryophyllia formosa, Pourt. The pali are thin and long, but not high; and they are placed before the third cycle of septa, being absent when the fourth and fifth orders are incomplete. The species has, in external form, some resemblance to Caryophyllia cornuformis, Pourtales, and also in the costal arrangement; but this last species has four complete cycles of septa, and the pali are only placed before the secondaries, making it a very marked and peculiar member of the genus. In order to establish the species Caryophyllia pourtalesi, I have again delineated some parts of the specimens, to show the relative size of the coste and septa, and the size of the pali (Plate XLIII. figs. 1-7.) Young specimens of the species :— A young specimen was figured in the former essay (plate xlii. figs. 4-7) ; and 1 append the following remarks on another. The young specimens, like the old, show that they were attached by a round base, which was fractured by the dredge. In the earliest stage there appear to be six primary septa, which are curved, united in two instances, formed of two lamine, and united to a trabecular columella. Then a small secondary is found between each of them. At this stage the epitheca, which comes eventually to be beautifully granular, is readily separable from the cost beneath, which are curved (Plate XLIII. figs. 11-14). As the coral grows, the third cycle appears, and rapidly becomes complete. ‘The septa are wide apart, reach far inwards, and have large papilla-like granules on their sides. The pali are not distinguishable in specimens 3%) inch in height; but there are processes of the single columella which evidently are in relation with the secondary septa, where the tertiaries are complete in a system. CARYOPHYLLIA INSKIPI, nobis. This species was described in the first part of this essay (Trans. Zool. Soe. vol. viii. p- 316), but it was not delineated. A figure is now given; and the remarkably deep- seated columella, tall pali, and externally thick septa will be noticed (Plate XLIII. figs. 8-10). MADREPORARIA OF THE DEEP SEA. 239 CARYOPHYLLIA CALVERI, nobis. (Trans. Zool. Soc. vol. viii. p. 316.) This beautiful coral was not figured in the former description of the deep-sea corals, as the specimen was mislaid; it has fortunately been found, and has been compared * with its nearest ally, Caryophyllia antillarum, Pourtales (Zool. Results of Hassler Exped. i. p. 34). Both have a thick wall, and exsert primary and secondary coste ; but the Barbadian form has them less so than the other, and its cost are more developed. Probably they are races of the same species. One specimen, ;’p inch, exhibits beautiful costz, subequal, granular, and flat. (Plate XLIII. figs. 15-27.) Genus BATHYCYATHUS. BATHYCYATHUS MINOR, sp. nov. (Plate XLV. figs. 1-4.) A small coral, adherent by a broad base to the coste of a dead Bathycyathus atlanticus, nobis, has a circular calice, four cycles of septa, some of which are exsert, but all are largest at the costal ends. The cost are very small inferiorly and granular. They increase in size as the form expands near the widely open calice. The columella is small, and the pali are thin and long. Height 33; inch ; breadth of calice 4 inch. BatTHYCYATHUS MINOR, the young coral. (Plate XLV. figs. 7—9.) This is a coral on the septum of an old Bathycyathus; and it is about 3'g inch broad. It has a broad base, which slopes to the calice, the wall being low, and the six primary septa exsert. The calice is open, and there are six primaries and six secondaries and a small columella. A coral more advanced in growth is on the stem of a larger form; the septa are numerous, and the pali are irregularly distributed. (See the Plate.) Genus PARACYATHUS. PARACYATHUS INSIGNIS, sp. nov. (Plate XLIV. figs. 1-3.) The corallum is attached by a moderate-sized base, above which there is a slight constriction; the general shape is conico-cylindrical, bent and with a widely open calice. The coste are distinct to the base, are subequal, and each is multigranular. The calice is elliptical and open. The columella is very small, being composed of a few minute processes, and it occupies about one fifth of the calice. The septa are slightly exsert, largely granular laterally, and are moderately crowded. ‘There are four perfect cycles in each of the six systems, and sometimes orders of the fifth cycle in some. The orders of the fourth and fifth cycles tend to approach the intermediate septa towards the inner margin and close to the pali. The septa are short, and reach inwards accord- 240 PROFESSOR P. M. DUNCAN ON THE ing to their order. The pali are single, tall, sharply granular, not incised; and those of the tertiary septa are short and most distinct Height 3% inch ; breadth of calice 3% inch. Locality, No. 19 dredging, in the 2nd expedition of the ‘ Porcupine,’ 248 fathoms. The small columella, septal number, large pali, the approach of the higher order of septa towards the next, and the costal structures separate this Paracyathus very readily. It belongs to the group with unlobed pali, and is easily distinguished from Paracyathus pulchellus, Ed. & H., and Paracyathus strictus, Philippi. It differs from Paracyathus agassizi in the smallness of the calice, that of the last- named coral being large and the pali bilobed. ParacyaTuus srriatus, Phil. (Plate XLIV. figs. 4-10.) In the memoir on the deep-sea corals (No. iv. Illustr, Cat. Harvard Coll. 1871), Count Pourtales describes, with some hesitation, Paracyathus confertus, with cost distinct to the base, not prominent and granular, calice oblong, concave; septa crowded, thin, entire, slightly exsert, in 5 cycles, but with considerable irregularity in some of the systems. Pali numerous, and difficult to distinguish from the papille of the columella; and he notices that specimens from the Azores do not differ from this form. In the description of the corals collected in the Hassler expedition, 1874, he notices (p. 38) that he has seen small specimens associated with his type which resemble Paracyathus de filippi, Duch. et Mich., of the West Indies; and he suspects them all to be of one species. With regard to the variation of the Paracyathi, he observes :— The characters are very variable—the type figured in my Deep-sea Corals, pl. vi. figs. 11-13, passing into another with deeply sunk columella, the papille of which are partly twisted like those of a Caryophyllia or Trochocyathus, well-defined pali rising much higher, more exsert septa, and a more regular shape. ‘This latter type is the most common at Barbadoes, the other in Florida. The great variability of these forms inclines me to believe that Paracyathus agassizi, Dunc., can scarcely be separated, especially from the West-Indian form.” There is no doubt that the dif- ficulty of discriminating the species of Paracyathus is very great; but, as a rule, the septal number, the size of the columella, the lobed or not lobed character of the pali, and perhaps the costal development are visible early in the coral-growth. I would rather therefore at present continue to maintain Paracyathus agassizi. The figure given by Count Pourtales of Paracyathus confertus shows distinctly the crowded septa with hardly any interseptal loculi, a large columella, and perfectly well-formed bilobed pali. But in the specimen which I received from him this character of the pali is not present, and the columella is deeply seated, the septa being crowded. Probably, then, there is more than one species of American Paracyathus with close septa. On comparing this last specimen with those obtained in the expedition of H.M.S. ‘ Porcupine,’ some of which were figured in the former Memoir on Deep-sea Corals (Trans. Zool. Soc. vol. viii. pt. v.), MADREPORARIA OF THE DEEP SEA, 241 the closest resemblance was noticed between it and those described and figured as Para- cyathus striatus, Philippi. A specimen of this species is in the collection of corals now under consideration ; and as it is in a good state of preservation, it is delineated with the Paracyathus from 50-100 fathoms dredged by the Gulf-Stream Exploration Survey. The American form has four perfect cycles and none of the orders of the fifth cycle; it has 24 single-lobed tall pali, of which the tertiaries are the largest; the columella is deeply situated, and its papilla are distinct and much smaller than the pali. The columella is oval and moderately large; and the cost are subequal, rather flat, varnished- looking, and minutely cross-grained. (Plate XLIV. figs. 4-7.) The Mediterranean forms are like the American, with the exception of the costal structures, which in the first are more prominent, less glazed, but still granular. There are four cycles of septa. (Plate XLIV. figs. 8-10.) I propose, then, to include the coral so kindly given me by Count Pourtales with the Mediterranean Paracyathus striatus, Philippi, sp. PARACYATHUS MONILIS, sp. nov. (Plate XLIV. figs. 11-13.) The corallum has a broad base and open calice, which is not very shallow. The coste are visible to the base in series of raised lines, with a row of large sharp granules. The columella is formed by a twist of ribbon-shaped sclerenchyma, and is small. The septa are not crowded, are distinct, and the primaries are the largest and the most exsert. They extend far into the fossa; there are four incomplete cycles, and all the lamine are granular. The smaller septa are rather wavy, and correspond to coste larger than themselves, which do not reach far down. ‘The pali are long, narrow, not prominent, and are placed before the tertiaries and some secondaries where the cyclical arrangement is incomplete. In young specimens with 24 septa the pali are placed before the primaries. Height 32; inch; breadth of calice 75 inch. Locality, 60 fathoms, seven miles off Rinaldo’s Chair, Mediterranean. PARACYATHUS INORNATUS, sp.nov. (Plate XLIV. figs. 14-16.) The corallum is short, and the base is almost as wide as the calice. The coste are absent, and are replaced by a plain glistening pellicular epitheca, marked with in- distinct shagreen-looking granulations. The septa are not exsert, are numerous, slender, wavy, not crowded, and very unequal. Those of the last cycle are simple pro- jections from the wall. The tertiaries project more inwards, and have a small palus before them as thin as they are. The secondaries are not to be distinguished from the primaries, are granular, and have a papilliform palus. The columella is very small and trabecular. Height 5 inch; breadth 75 inch. Locality, Mediterranean Sea. . 242 PROFESSOR P. M. DUNCAN ON THE PARACYATHUS HUMILIS, sp. nov. (Plate XLIV. figs. 17-19.) The corallum is small, with a wide base. The coste are distinct, large, multi- granular, and subequal ; inferiorly they are less separate, and the granular structure is more distinct. The septa are large, and the pali also, for the size of the calice; and the primaries are very exsert and arched, the tertiaries being higher than the secondaries. The pali are large and rather square, and are placed before two cycles of septa. There are three cycles of septa, some of the tertiaries being absent. The columella is very deep and very small. Height of corallum 35 inch; breadth of calice 3’; inch. Locality, Mediterranean Sea. PARACYATHUS AFRICANUS. (Plate XLIV. figs. 20-22.) The corallum is conico-cylindrical, with a base more than half the breadth of the calice. The calicular margin is thick, circular in outline, and the fossa is deep. The columella is very small, and the papille are distant and small. The septa are in six systems; and there are four cycles in most; but in a few the fourth and fifth orders are wanting or very slightly developed. The laminz are large, not crowded, largely granular at the sides, and extend inwards not quite one third of the diameter of the calice. The primaries are the largest, the most exsert, and are attached to the largest cost; and all the others project beyond the wall in continuation with the coste. The secondaries greatly resemble the primaries ; and the tertiaries are smaller and less exsert than the rest. Pali are placed before the septa, except those of the higher orders; they are tall, moderately stout, simple, not excised; and those of the tertiary septa are the largest, and are broader than the septal end. All are granular, and slightly bent here and there. The cost are distinct to the base in some places; but elsewhere there is a dense epitheca ; where visible they are slightly spinulose externally, alternately large and small, but rather subequal. Near the calice they are more prominent than elsewhere. Height 33; inch; breadth of calice 7%5 inch. Locality: Coast of Tunis, 2nd exped. ‘ Porcupine,’ in 40 fathoms. PARACYATHUS COSTATUS, sp.nov. (Plate XLIV. figs. 25-26.) The corallum is straight, cylindro-conical, with an attached base and a calice circular in outline and large. The coste are distinct to the base, slightly rounded, sparsely but distinctly granular, unequal; and the highest are the primaries and those next to them. The calicular margin is circular, sharp; and the fossa is deep; the septa are very slightly exsert, and project but little into the calice; they are thin, separate, unequal, granular, and there is little difference between the primaries, secondaries, and tertiaries. The columella is large, slightly concave, and consists of trabecule ending in numerous papille. The pali are tall, not long, are distinct from the septa, stouter than these, MADREPORARIA OF THE DEEP SEA. 243 not bilobed, but ragged on their inner margin; and the largest are before the tertiary septa. There are four cycles of septa, and many members of the order of the fifth cycle. Height of corallum 4 inch; breadth of calice 33; inch. Mediterranean. Coral zone. Genus FLABELLUM. FLABELLUM MINUS, sp. nov. (Plate XLV. figs. 10-13.) The corallum has a distinct base of attachment and is cylindro-conical and compressed superiorly. ‘The epitheca is well developed, and is marked by close curved lines in festoons, which meet along longitudinal linear grooves that correspond with the inter- laminar space of each septum. The calice is elliptical, not deep, and the margin is sharp and thin. There are 16 septa, there being six primaries, six secondaries, and the tertiaries only developed in four half systems at the opposite ends of the long axis. The septa are wide apart, thin, granular, and slightly exsert. The height of the corallum is 35, and breadth of calice ;?; inch. Locality: 2nd exped. ‘ Porcupine,’ 996 fathoms, No. 16 dredging. This small Mlabellum may not be full-grown; but its broad base (for its size) larger than that of the full-grown specimens of Flabellum distinctum, and its low septal number, when of the same size as the young of that species, indicate a satisfactory specific difference. The species has some structural resemblance to Mabellum woodsi of the Crag, espe- cially in the lines on the outside, which corresponds with the middle of the septa; but probably its nearest ally is Flabellum siciliense, Ed. & H., of the Sicilian Tertiaries. Genus GEMMULATROCHUS, gen. nov. The corallum is compound, is fixed by a broadish base, and is conico-cylindrical in shape. The wall is thick; and there is a well-marked epitheca, the cost being rarely visible. The calice is very deep; and there is a rudimentary columella. The septa are stout. Budding takes place from the wall high up; and the buds ascend and fre- quently join by their walls to others of different corallites, so as to constitute a bush- shaped corallum. GEMMULATROCHUS SIMPLEX, spec. nov. (Plate XLV. figs. 18-20.) The parent corallite bears buds on opposite sides; it has a slight constriction above the base of attachment, a well-developed epitheca, and a rather elliptical calice. The septa are distinct, stout, granular, and short, not reaching far inwards; there are six systems; and the fourth cycle is incomplete in all but one. The primaries are the VOL. X—Part v. No. 2.—WMarch 1st, 1878. 2M 244 PROFESSOR P. M. DUNCAN ON THE largest and longest, but they are barely exsert; and the secondaries often have a tertiary united to them very low down in the very deep calice, near the rudimentary columella. The calicular margin is stout. Height ;4, inch; breadth of calice ’-inch. The buds are bent slightly; and the smaller ones have three cycles of septa. The form resembles Blastotrochus ; but the buds do not fall off, but remain to form the tuft-like corallum. It is a genus allied to Smélotrochus, Ed. & H., and Onchotrochus, nobis ; but the gemmation and epitheca separate it. Much resembling in its calice, except in the thick margin, Cenocyathus anthophyllites ; this species, however, has no pal. Locality: Northern shores of Mediterranean, below tide-marks. Genus BLASTOSMILIA, gen. nov. The corallum is compound; and there are repeated gemmations from the wall of the parent corallite, and occasionally from the walls of buds. The corallites are conico- cylindrical, long, bent, except the parent; and the calice is circular in outline and deep. The wall is thin, and is covered with a granular epitheca, the rudimentary coste being only visible close to the calices. The columella is rudimentary, but exists as trabecule from the septal ends. The septa are very thin, slightly exsert, not incised, project but little into the calice; and the primaries, and sometimes the secondaries, unite at the base of the fossa with the small deeply seated columella. There are six systems of septa; and the fourth cycle is usually incomplete in some systems. The dissepiments are wide apart, and are formed at the bottom of the calice by the septal ends becoming oblique and wide and occluding the space below. BLASTOSMILIA POURTALESI, sp. nov. (Pl. XLV. figs. 14-17.) The corallum has a long parent corallite, with long cylindroid curved buds, curving more or less in oblique series. ‘The septa are unequal, the primaries being larger than the secondaries; they are also slightly exsert. The cost near the margin are broader than the septa; and the margin is unequally circular in young specimens, the intercostal spaces bulging out in elegant curves. The septa of the parent are in six systems; and the fourth cycle is in all of them; but there are only three cycles in the next in size. The columella is small. Height nearly 1} inch; breadth of parent calice 3% inch. Locality: Mediterranean, from red-coral zone. Count L. F. Pourtales, in his admirable description of the Deep-sea Corals (Illust. Catalogue, No. iv. p. 21), described and figured a coral, Cwlosmilia fecunda, Pourt., which evidently has the closest alliance with this Blastosmilia. He remarks, after describing his species, that the generic affinities are a little doubtful, and distinguishes MADREPORARIA OF THE DEEP SEA. 246 it from the Cladocoracee and from my Onchotrochus (Monog. Brit. Foss. Corals, 2nd series, part ii. No.1, p. 4, 1869), and places it in the genus Celosmilia. This genus I carefully analyzed in the monograph just referred to (p.5); and out of 15 species I have described six. It is a subgenus of the genus Trochosmilia; and I have never noticed gemmation from any corallite of any species. Believing that the new genus is a good one, and that it is better to form one for the two species, I venture to include Celosmilia fecunda, Pourt., in it, and to term it _ Blastosmilia fecunda, Pourt. The American species ranges from 63 to 315 fathoms, and affords another instance of the affinity of the West-Indian and Mediterranean marine faunas. A further possible alliance is indicated in'the affinities of the species with Cenosmilia arbuscula, Pourtales (Zool. Results of Haslar Expedition, pt. i. p. 39,1874). The genus Cenosmilia is thus defined :— This genus is formed to receive the Parasmiliv propagating by germination, and thus becoming compound. Single corallites are typical Parasmiliw.” Inthe Supplement to the British Fossil Corals I described several Parasmilie, and was always impressed with the great costal development, and that of the endotheca and columella. But I never found one budding. The coste in Ceno- smilia arbuscula are, from the photographic reproduction given, not well developed, nor is the columella. Whilst clearly seeing the distinction between Blastosmilia and Celo- smilia, 1 cannot help thinking that the form described by Pourtales is very closely allied to mine. IV. Genera Remarks. The numbers of the “ dredgings” refer to those of H.M.S. ‘Porcupine ;’ and their exact localities and temperatures are stated in the first part of this essay (Trans. Zool. Soc. vol. viii. p. 338). The Caryophyllie now described are remarkable for their low septal number and slender shape. They have each an epitheca; and in Caryophyllia simples it is beau- tifully marked with a chevron pattern. They come within a section of the genus of which the species C. vermiformis, Duncan, described in the former essay, is the type. They are not without affinities to Pourtales’s Stenocyathus, from the other side of the Atlantic. They are both from deep water in the Atlantic, west of the British Channel. Bathycyathus minor, sp. nov., is without those interesting alliances which rendered the other species with which it was found so interesting. It came from a great depth. off the south-west coast of Spain, in 1095 fathoms. Six species of Paracyathi, all from the Mediterranean, are interesting for the beauty of their construction and their distinctness from the forms already described. The Parathyathus striatus, described in the first essay, I believe to be fouud also in the American part of the Atlantic. One of the new species is remarkable from its entire 2m 2 246 PROFESSOR P. M. DUNCAN ON THE want of cost; and others have them moniliform or multigranular; and a partial epi- theca is found in Paracyathus africanus. The bathymetrical range is great, or from 480 fathoms to a few feet below tide-mark. The Flabellum is allied to the common form in the Sicilian Tertiaries, Flabellum siciliense. Gemmulatrochus simplex is a species of a new genus formed to admit budding Tur- binoliide without pali. The buds do not fall off, as in Blastotrochus, but remain attached to the side of the parent corallum, and grow. The new genus Blastosmilia is a remarkable one, on account of the repeated gem- mation from a parent, the presence of endotheca, and the rudimentary coste and columella. The remarks on Caryophyllia calveri and C. pourtalesi, species which were described in the former essay, are included in the notice of them in the description of the species. These 12 species, added to the 30 described in the former essay, bring the number of species of deep-sea corals dredged up in the voyage of H.M.S. ‘ Porcupine’ to 42. They include a large and predominating number of Turbinoliide; and there is no instance of any form possessing cellular exotheca or exothecal structures binding together the corallites, as in reef-builders. One of the species newly described is allied to a fossil form; and thus the alliances with the old coral faunas stand :—42 species found in the recent fauna; 9 of them lived in the Pliocene, 1 in the Miocene, 1 in the Cretaceous, and 5 of the species have alliances with the corals of former ages. The résumé of the peculiarities of the deep-sea corals given in page 337 of the former essay is shown to be correct by the study of the forms described in this. Fig. Fig. Fig. Fig. Fig. art mo fF oF DH a S) eo) io) | OMOARANKRwWNHS: bo po bw bw one & . 24, bo bo op Or Beale bo oo © O85 oe OO OO Hm CFO DD Re © bo nse ~I MADREPORARIA OF THE DEEP SEA. DESCRIPTION OF THE PLATES. PLATE XLIII. . Caryophyllia pourtalesi, nobis: nat. size. . Caryophyllia pourtalesi, nobis: x 5. . Caryophyllia pourtalesi, nobis: septum and pali, magnified. . Caryophyllia pourtalesi, nobis: calice, x 5. . Caryophyllia pourtalesi, another specimen : nat. size. . Caryophyllia pourtalesi, another specimen: X 5. . Caryophyllia pourtalesi, another specimen: calice, x 5. . Caryophyllia inskipi, nobis: side view. Caryophyllia inskipi, nobis: side view, x 5. . Caryophyllia inskipi, nobis: calice, x 5. . Caryophyllia pourtalesi: young specimens. . Caryophyllia pourtalesi: magnified 4 times. . Caryophyllia pourtalesi : . Caryophyllia pourtalesi : . Caryophyllia calveri, nobis: a corallum. calice, x 4. its base, x 4. . Caryophyllia calveri, nobis : . Caryophyllia calveri,; nobis : . Caryophyllia calveri, nobis : . Caryophyllia calveri, nobis : . Caryophyllia calveri: another form. . Caryophyllia calveri, nobis : . Caryophyllia calveri, nobis: . Caryophyllia calveri, nobis : Caryophyllia calveri, nobis: . Caryophyllia calveri, nobis: . Caryophyllia calveri, nobis : Caryophyllia calveri, nobis: . Caryophyllia carpenteri, sp. . Caryophyllia carpenteri, sp. a corallum, x 3. a calice, x 3. part of a calice, x 8. its base, . ~ hu Hanhart ump. WHWesley adnat.del AXIAL SKELETON OF THE PELECANIDA ; * ‘ { | 4 i a ‘ 4 ‘ es ‘ <. | ; ‘ a pl 4 Hanhart ump WHWesley ad nat del AXIAL SKELETON OF THE PELECANIDA. <0 e : = { 3, \ iss " i . € : i 2 Y ¢ \ “e : he bd + ‘ es ’ j ia %, Wy, -" Oo os i rar Ce Tue * eas 7 . ot ~< ee . t am ~ d nat del o r i} Se eee i. Imr yaeuceypy mt pLey ey , & “ f } t ¢ a ree { cs rf ; had ¥ oer E ‘ 4 : . WCINVOHTHd HHI 40 NOLH TEMS TVIXV dt yx2y0ey 15*MHM pr are * IX. A Monograph of the Ostracoda of the Antwerp Crag. By Grorae StEwaRDSON Brapy, .D., F.L.S., C_M.Z.S., Professor of Natural History in the University of Durham Colne of Physical Science, Newcastle-upon-Tyne. Received April 19th, read May 15th, 1877. [Puates LXII. to LXIX.] AFTER examining and carefully studying a very interesting collection of Ostracoda from the neighbourhood of Antwerp, which has, with great courtesy, been submitted to me by M. Ernest Vanden Broeck, I regret that I am unable to point to any facts of distribution or grouping which can throw the slightest light upon the age or mutual relations of the strata in which they occur. The general character of the collection is about as different as well could be from that of the group described and figured by Professor T. Rupert Jones in his ‘ Monograph of the Tertiary Entomostraca of England.’ Indeed, of the fifty species described in the present memoir, eight only enter into Professor Jones’s list—viz. Cythere woodiana, C. plicata, C. wetherellii, C. macropora, C. scabropapulosa, C. jonesti, Cytheridea pinguis, and C. miilleri; and of these eight, three (C. woodiana, wetherellii, and scabropapulosa) are of extremely rare occurrence in M. Van den Broeck’s collection. Under these circumstances it is obviously impossible, in the present state of our knowledge, to institute any useful comparison between the Tertiary Ostracoda of England and Holland. One noteworthy point in the collection, however, is this—the entire absence of freshwater species. Prof. Jones’s memoir contains eight freshwater species from the Eocene and Pleistocene deposits of England ; M. Bosquet’s Monograph of the French and Belgian species, on the other hand, has no freshwater species ; neither, to all appearance, have the smaller works of MM. Speyer and Egger. In the case of the Antwerp specimens it would appear that the fauna was deposited in water of a moderate depth, probably not less than 15 or 20 fathoms. In the following Table the number of asterisks roughly indicates the comparative abundance of the different species:—one asterisk denoting scarcity; two, moderate quantity ; and three, the greatest abundance. yOL x.—PparT vill. No. 1.—August 1st, 1878. 380 DR. G. S. BRADY ON THE OSTRACODA SABLES INFERIEURS. Sases ovens, _ | SABLES su- PERIEURS. Species. 4 “i Pectunculus- Banepa zone: a 2 e Zone a Iso- Zone a | yo ; Trophon | zone. zoaires, | Cardium cor. antiquum. } Kiel. Edeghem, Paracypris)polita,(Sars..,....h.- 55s ace * * | +Pontocypris faba (Heuss.)...........-.-.. * ot | be PLOPING UA POL Ay) eet eke bad Bairdia oviformis, Speyer .............-. * Cythere cribrosa, B., C.,g¢ R. ............ * | * WOMibheh INOS canenopneonnonnedac rea | * ellipsoidea, Brady ............... BOeE Puce * * It. Siebel, OPT neconscspnbacsenoe HK #*X snide aye % —— plicata, Minster .......... 0050045: * *x * ** t Wale. ERG) scooec.an ho enon osOoue |S oe ee 't+——— »plicatula)(Fewss)) 25. os. eee oe * t———. cieatricosa (Fewss) ......0..5+-2.0- ao * * xe | —— edichilus, Brady.................. ** * Sere * Petrosas Bradley) er ceele a cia lade) wie * + timycolax(Vonrmamn) mare er icertst silent saat Savas eco Bee * t latimarginata, Speyer .............- *RK **K KK ** * * aw ELHOLeliiteye) OES teieleta neater teie weiss #* ** ** tT (GhdaInby JECiHe og gt om odo oda be ** ** ** acuticosta, Hoger .. 6. v anes 20 we ole ** * mE aAbh JEP sobanoancede ceed 2600 St08 one aes * be EO), JOUVE ss dons anndnabaes ado **K **K HH * polytrema, Brady.........+..-...5- ES * * HK scabropapulosa, Jones ..........--+- * ——— GAOL 145 TEP (OL noon caenadnoaos ne Baden * * * 2 subcoronata, Speyer... ..- 6. +..s00-: * eee * i TCO SW) Genancow tassemade antec ee tes % + WORE) Boas gah danonAcnEno 56 aE * * * KKK ———_—-S TM As HCLSSH ata eyes ee iceacicioe nears * +Cytheridea papillosa, Bosquet, var. levis .... (UAT, UMS so agoorbioa nd oowasnec ERX KK HK Se EK * Be cypridioides, Brady ...........-.-.- teers eee Rc: ** t pind eile SATS 6 ond oad beatcoae aos HK EK HK HK * Loxoconcha latissima, Brady ............ ** #* #K lrtnATLACRN a, Ver OT so gonnnsjsosuWens Sener: aeeys *% KK grateloupiana (Bosquct) ............ * “ae * 5 rae * BE TENQENE, Jen00M ppeapatess soo oeoue ** * * HK +Xestoleberis depressa, Sars .............. 008 se0¢ S008 eK +Cytherura broeckiana, Brady ,........... * * * be CONN, BRM aodoanacanad hen aaode Seah * * +Cytheropteron latissimum (Norman) ...... * * * | intermedium, Brady .............- errs * gradatum (Bosquet) ................ * * Pipistrellase Brady le. crleeciel. -stels eo *% +Bythocythere constricta, Sars ............ ane vere * +Cytherideis lithodomoides (Bosquet)........ *#*X ** EK HK Meg, Jens s6a0cs on co ab enouOuo6 * +Paradoxostoma ensiforme, Brady ........ satus are * +Cytherella parallela (Reuss) ............. * atest #* * ellipticasBradaypy.ccteiie:< sient statis git * ao ——=— NOGOSA, Brady) ta ewnielsie wears alee ele wl ** ** * 7 The species marked thus are known as still living. ¢ These are so closely allied to OF THE ANTWERP CRAG. 381 Family CYPRID/. Genus Paracypris, G. O. Sars. Shell smooth, compact, higher in front than behind. Upper antenne shortly seti- ferous, lower strongly clawed. Second maxilla having a branchial appendage, palp elongated, conical, and inarticulate. Last pair of feet like the first, and ending in a long curved claw. Postabdominal rami large, ending in two strong curved claws and a short seta; on the posterior margin two long sete. One eye. PaRAcyPRIS PoLita, Sars. (Plate LXIII. figs. 5 2-5 d.) Paracypris polita, G. O. Sars, Oversigt af Norges Marine Ostracoder, p. 12 (1865). Paracypris polita, Brady, Monograph of Recent British Ostracoda, p. 378, pl. xxvii. figs. 1-4, and pl. xxxvilii. fig. 2 (1868). Paracypris polita, Brady, Crosskey, and Robertson, Monograph of the Post-tertiary Entomostraca of Scotland, &e., p. 131, plate xv. figs. 9, 10 (1874). Carapace, as seen from the side, elongated, siliquose or subtriangular, highest in front of the middle; height equal to rather more than one third of the length; anterior margin evenly rounded, posterior sharply attenuated ; dorsal margin well arched, sloping steeply behind, ventral margin more or less sinuated in the middle. Seen from above, the outline is compressed, oblong, tapering evenly to the extremities; greatest width situated near the middle and equal to more than a quarter of the length. The left valve overlaps the right in the middle of the ventral surface. End view broadly ovate. Surface smooth. Length 3'5 inch (1 millimétre). One or two specimens only were found in the Pectunculus and Panopwa-menardi beds of the “Sables inférieurs.” In the living state the species occurs, though not very commonly, in the North Sea off the coasts of Great Britain and Norway. It has been noticed also sparingly as a fossil in the Post-tertiary beds of Norway and Scotland. Genus Pontocypris, G. O. Sars. Shell thin and fragile, higher in front than behind, elongated, subreniform or sub- certain liying species as to make their distinction somewhat doubtful; the alliances are as follows :— Pontocypris faba closely approaches P. mytiloides (Norman). Pontocypris propinqua closely approaches P. angustata, Brady, and P. trigonella, Sars. Cythere belgica closely approaches C. plicatula, Reuss. Cythere cicatricosa closely approaches C. convexa, Baird. Cythere macropora closely approaches C. lactea, Brady. Cytheridea cypridioides closely approaches C. zetlandica, Brady. Loxoconcha variolata closely approaches L. alata, Brady. Cytherura broeckiana closely approaches C. fulua, Brady and Robertson. Cytherideis lithodomoides closely approaches Cytheridea elongata, Brady. 382 DR. G. S. BRADY ON THE OSTRACODA triangular. Lower antenne bearing, on the last joint but two, a brush of five sete, and at the base of the same joint a pedicellated vesicle. Second maxille having no branchial appendage. Palp large and subpediform, 3-jointed; last joint in the female with two long, curved claws. Second pair of feet flexuous, 4-jointed; last joint short and ending in several short sete, one of which is pectinated. Postabdominal rami large, with two curved claws and a slender seta at the apex and three long sete on the inner margin. PontocyPris FABA (Reuss). (Plate LXIII. figs. 6a, 66 (2), 6c-Ge (¢). Bairdia faba, Reuss, “Ein Beitrag zur genaueren Kenntniss der Kreidegebilde Meklenburgs,”’ Zeitschrift d. deutsch. geol. Ges. 1855, p. 278, pl. x. fig. 2. Carapace of the female, as seen from the side, subtriangular ; greatest height situated in front and equal to one half of the length; anterior extremity broadly rounded, posterior rounded but much attenuated ; dorsal margin very strongly arched and highest in front of the middle; ventral almost straight, slightly sinuated. Seen from above, oblong-ovate, tapering suddenly in front and very gradually toward the posterior extre- mity; greatest width situated at the anterior third, and equal to more than one third of the length. The right valve overlaps the left in the middle of the ventral surface. Surface smooth or slightly granulated. The male differs from the female in the greater size of the shell, in its more attenuated proportions, and in being subacuminate at the hinder extremity. Length of female 75, of male 3!; inch (0°65, 0°75 millim.). Several examples were found in the bed “a Bryozoaires” of the ‘‘ Sables moyens,” and a few in the ‘“ Panopwa menardi” bed (Sables inférieurs). The reference of the two forms figured in Pl. LXIII. to the male and female sexes, respectively, is of course hypothetical; but the general characters of the two forms are so similar, and the points of difference are so exactly those which we observe as sexual distinctions among the recent Ostracoda, that I entertain very little doubt as to the correctness of the diagnosis. Reuss’s figure of a single valve of his “ Bairdia faba” agrees exactly with the present species; I therefore do not hesitate to adopt that name. ‘The recent Pontocypris mytiloides approaches it also very closely in general character. I think it very probable that Egger’s Bairdia dactylus and its variety punctata may also be identical with this species; but the figures given by that author are unfortunately so extremely coarse that it is impossible in many cases to make an accurate diagnosis. PONTOCYPRIS PROPINQUA, nov. sp. (Plate LXIII. figs. 4 a—4 c.) Carapace, seen laterally, subtriangular, highest in the middle; height equal to half the length; anterior extremity evenly, posterior obliquely rounded; superior margin boldly and evenly arched, inferior nearly straight. Seen from above, the outline is ovate, pointed in front and obtusely rounded behind; greatest width situated in the OF THE ANTWERP CRAG. 383 middle, and equal to somewhat less than one half the length. End view subcircular. Surface of the shell perfectly smooth. Length 3 inch (0°75 millim.). A few specimens only in the “Sables moyens,” zone 4 Bryozoaires. This species, when seen laterally, is very like P. trigonella, Sars, but is less attenuated, and when seen from above is found to be much more tumid. Genus Barrp1a, M‘Coy. Valves unequal in size, the left much the larger and overlapping on the dorsal and ventral surfaces. Shell nearly or quite smooth, mostly subrhomboidal or subtriangular. Kyes wanting. Antenne robust; the upper 6-jointed, the first two joints being large and thick, the rest short and bearing long sete. Lower antennz 5-jointed, the second joint having on its base a bisetose tubercle. Mandibles large, having six or seven long, strongly serrulated teeth; palp 4-jointed, bearing a small trisetose branchial plate. One pair of jaws only, 3-branched, and bearing a well-developed branchial appendage. Three pairs of feet, all alike, directed forwards and protruding from the shell, 4-jointed, and clawed at extremity ; first pair bearing at the base a large ovate branchial lamina. Postabdominal rami short, clawed and setose. BAIRDIA OVIFORMIS, Speyer. (Plate LXIII. figs. 7a-7c.) Bairdia oviformis, Speyer, Die Ostracoden der Casseler Tertiarbildungen (1863), p. 44. pl. 1. fig. 6. Carapace, as seen from the side, broadly subtriangular, approaching to elliptical ; greatest height situated in the middle and equal to two thirds of the length; anterior extremity broad, obliquely rounded ; posterior broad and slightly produced, so as to form an almost obsolete beak ; dorsal margin strongly arched ; inferior slightly convex. Seen from above, the outline is regularly ovate, pointed in front and broadly mucronate behind, widest in the middle, the width being about equal to half the length. End view broadly ovate, narrower above. ‘The right valve considerably smaller than the left, somewhat angular on the dorsal margin and distinctly beaked behind. Surface smooth. Length +; inch (1°5 millim.). One specimen only of this fine species was found in the Lsocardium-cor bed of the “Sables moyens.” Fam. CYTHERID &. Genus CytTHERE, Miiller. Valves unequal, mostly oblong-ovate, subreniform or subquadrate; surface smooth, punctate, rugose, spinous or tuberculated, usually bearing a rounded, polished tubercle over the anterior hinge-joint. Hinge formed on the right valve by two terminal teeth, on the left by one anterior tooth and a posterior fossa, between which there is often a 384 DR. G. 8S. BRADY ON THE OSTRACODA bar which is received into a furrow of the opposite valve; the teeth sometimes crenu- late, and sometimes wanting on the left valve. Antenne robust ; superior 5-6-jointed, and bearing on the anterior margin three curved spines ; inferior 4-jointed, Mandibular palp 3-4-jointed, bearing in place of a branchial appendage a tuft of sete. Eyes one or two’. ? CyrHERE criprosa, B.,C., & R. (Plate LXIV. figs. 4a, 45.) Cythere cribrosa, Brady, Crosskey, and Robertson, Monograph of Post-tert. Entom. of Scotland, &c., p. 146, pl. x. figs. 5-7. Carapace compressed, oblong, subreniform ; seen from the side, the anterior extremity is evenly rounded, the posterior oblique and ending above in an obtuse angle; dorsal margin nearly straight, ventral rather deeply sinuated in the middle. Height scarcely equal to half the length. The outline, when seen from above, is evenly compressed, oblong, nearly of equal diameter throughout, the extremities being rather obtuse. The surface of the shell is reticulated, the meshes uniting into obscure longitudinal furrows on the ventral surface. Length #4 inch (0°55 millim.). Of rare occurrence in the Zrophon- and L[socardium-beds. CYTHERE WOODIANA, Jones. (Plate LXV. figs. 4a, 45.) Cythere woodiana, Jones, Monogr. Tert. Entom. England, p. 29, pl. iu. figs. 2a—29. Carapace, seen laterally, oblong subquadrangular ; anterior extremity oblique, slightly rounded ; posterior scarcely rounded, almost truncate ; superior margin almost straight, inferior very slightly convex; height equal to half the length. Seen from above, the outline is oblong-ovate, widest behind the middle. The surface is thickly covered with large rounded or subangular punctations. Length 3/5 inch (1°3 millim.). This is one of the most abundant and characteristic of the Ostracoda of the Pliocene Crag of Suffolk (England). Two detached valves have been found in the “Sables supérieurs” of Antwerp, Zrophon-antiquum bed. The lower or Suffolk Crag, in which only the English specimens of C. woodiana have been found, is that known as the “Coralline” Crag, though, as stated by Professor Rupert Jones, that designation is quite inapplicable, the characteristic fossils of the deposit being not Corals or Corallines, but Sponges and Bryozoa (Polyzoa). CYTHERE ELLIPSOIDEA, noy. sp. (Plate LXV. figs. 1 a—-1d.) Carapace, seen from the side, subelliptical ; height equal to more than half the length and nearly uniform throughout; extremities rounded; dorsal margin very slightly arched, having a slight projection over each hinge-joint fore and aft; ventral margin * In the generic definitions given in this Memoir, I have not thought it desirable to include every anatomical detail, but have been content to give only the more important features. OF THE ANTWERP CRAG. 385 rather convex, with a slight sinuation in the middle. Outline, seen dorsally, oblong, subpentagonal, widest near the posterior extremity; greatest width equalling half the length ; anterior extremity broad, subtruncate, and slightly emarginate in the middle ; posterior tapering abruptly to a bifid submucronate point. End view broadly subtri- angular. Surface of the shell somewhat undulated and marked with impressed puncta, which are irregularly scattered and of various sizes, being larger towards the anterior and ventral margins, where they are arranged more or less regularly in the direction of longitudinal furrows. The left valve is larger than the right; and the junction of the valves on the hinge-line is marked on the dorsal surface by a deep longitudinal depression. Length 3's inch (11 millim.). This species seems to be of rare occurrence, only one or two examples having been noticed in the zones “ & Bryozoaires ” and “ Jsocardium cor.” It approaches closely to Speyer’s Cythere millepunctata, with which I was at one time disposed to identify it ; and I am not sure that Speyer’s figures may not represent the young form of the species. CYTHERE JURINEI, Miinster. (Plate LXV. figs. 2 a-2h.) Cythere jurinei, Von Minster, Jahrbuch fiir Mineralogie, &c., 1830, p. 60, et 1835, p. 445 (fide Bosquet) . Cythere jurinei, Rémer, op. cit. 1838, p. 516, pl. vi. fig. 12. Cythere jurinei, Bosquet, Entom. fossil. des terrains Tertiaires de la France et de la Belgique, p. 56, pl. ii. figs. 9a, 6, c, d (1852). Cythere jurinei, idem, var. 8. tenuipunctata, op. cit. pl. ii. figs. 10, a, , ¢, d. Cythere jurinei, Egger, Die Ostrak. der Miocan-Schichten bei Orenburg in Nieder-Bayern (1858) p. 20, pl. iii. fig. 5 (icones mali), var. ovata, pl. iu. fig. 4. Cythere jurinei, Speyer, Die Ostrac. der Casseler Tertiirbildungen (1863), p. 15, pl. ii. fig. 5. ? Bairdia semipunctata, Bornemann, “ Die mikroskopische Fauna des Septarienthones von Herms- dorf bei Berlin,” Zeitschrift d. deutsch. geol. Ges. 1855, p. 359, pl. xxi. fig. 1. Carapace, seen laterally, oblong, higher in front than behind, the greatest height being equal to half the length; anterior extremity obliquely rounded, posterior rounded, often much narrowed and produced; dorsal margin very gently arched, ventral slightly sinuated in the middle. Seen from above, the outline is oblong-ovate, tapering towards each extremity, the greatest width situated behind the middle. ‘The surface of the shell is either quite smooth, or marked along the middle or on the lower half of each valve with small impressed puncta, usually faint and arranged in curved longitudinal furrows. The right valve is smaller than the left, and is abruptly sinuated at each end of the dorsal margin. Length 3! inch (1-1 millim.). Figs. e-h represent, as I believe, the adult form of the species, whereas an immature stage is shown in the figures a—d, which also seem to be identical with the var. 2. tenui- punctata of M. Bosquet. It is not uncommon in recent Ostracoda to find the young marked with delicate sculpturing, which disappears in advanced age; the shape and 386 DR. G. 8S. BRADY ON THE OSTRACODA proportions of the shell are also liable to considerable variation during the process of growth. Cythere jurinei occurs in moderate abundance in the Pectunculus and Panopea- menardi beds (Sables inférieurs), also in the Trophon-antiquum bed (Sables supérieurs). CYTHERE PLIcaTA, Miinster. (Plate LXV. figs. 5 a-5 d.) Cythere plicata, Minster, Jahrb. fiir Mineralogie, &c., 1830, p. 63, and Neues Jahrb. &c. 1835, p- 446 (fide Jones et Bosquet). Cythere plicata, Romer, Neues Jahrb. fiir Min. &c. 1838, p. 518, pl. vi. fig. 26 (fide Jones et Bosquet). Cypridina laticosta, Reuss, Haidinger’s Abhandl. ui. p. 87, pl. ii. fig. 13. Cythere plicata, Bosquet, Entom. fossil. des terr. Tertiair. de la France, &c., p. 60, tab. ii. fig. 13. Cythere plicata, Jones, Tertiary Entomostraca of England, p. 32, pl. iv. fig. 16, and pl. v. figs. 8a— 8d (? pl. v. fig. 17). Cythere plicata, Egger, Die Ostrak. der Miocain-Schichten bei Orenburg (1858), p. 24, pl. v. fig. 9 (icones male) . Cythere plicata, Speyer, Die Ostrac. der Casseler Tertiarbildungen (1863), p. 29, pl. iv. figs. 2@,4,¢, d. Carapace, as seen from the side, oblong, subquadrangular; length equal to rather more than twice the height. Anterior extremity wide and well rounded, posterior narrowed, and armed with three or four blunt teeth ; dorsal and ventral margins nearly straight in front, but converging towards the posterior extremity. Seen from above, oblong-ovate, tapering gradually toward the front, and abruptly behind; extremities obtusely pointed ; greatest width near the hinder extremity, and equal to half the length. End view irregularly quadrate, the lateral margins having each a large central promi- nence. The valves are marked by three large, curved and rounded longitudinal ribs, the central one being the most prominent: the ribs themselves are smooth; but the intermediate furrows are sculptured with large rounded pittings. The hinge-line is marked by a deep depression. Length 3/5 inch (0°85 millim.). C. plicata occurs in moderate abundance in both beds of the “Sables moyens” and much more rarely in the Panopewa-bed (Sables inférieurs). It is noted by Professor Rupert Jones as occurring in the middle Eocene of the Isle of Wight and Hampshire ; and the same author states that “it has been found in the Miocene deposits of Dax, and in the Eocene of France, Belgium, North-western Germany, Bohemia, Austria, and Moravia.” In some of these deposits it seems to be very abundant, and, indeed, may be looked upon as one of the commonest and most widely distributed of the Tertiary Ostracoda. It is, moreover, very distinct in its characters, and scarcely likely to be confused with any other species, at any rate in its typical form. CYTHERE BELGICA, nov. sp. (Plate LXV. figs. 3 a, 36.) Carapace, seen from the side, subrhomboidal, somewhat higher in front than behind ; height equal to half the length ; extremities obliquely rounded ; dorsal margin straight, sloping gently from before backwards ; ventral slightly convex. Outline, seen dorsally, OF THE ANTWERP CRAG. 387 oblong-ovate, widest behind. Surface of the shell minutely and rather closely punctate, marked by two distinct but slender ribs, one of which crosses the centre of the valve somewhat obliquely in a longitudinal direction; the other lies near the ventral border, and curves upward behind to join the central rib. Length 3'5 inch (0°85 millim.). This is not very unlike a form described by Prof. Rupert Jones under the specific name “ spherulolineata ;” but as in some points it does not agree with the description, especially as regards the central tubercle and the “ beaded” character of the ridges, I have thought it better to assign it here a new name. Only one or two specimens were found in the “ Sables supérieurs.” CYTHERE PLICATULA (Reuss). (Plate LXIV. figs. 6a, 66.) Cypridina plicatula, Reuss, “ Die fossilen Entomostr.” &c., Haidinger’s Abhandl. 1850, p. 44, pl. x. fig. 23, a, b. Cythere plicatula, Bosquet, Entom. fossil. des terr. Tertiair. de la France, &c. p. 92, pl. x. fig. 23, 6. ? Cythere retifastigiata, Jones, Tertiary Entomostraca, p. 36, pl. 3. fig. 7. Cythere plicatula, Egger, op. cit. p. 38, pl. 5. figs. 6, 7, 8; Brady, “On new or imperfectly known Species of Marine Ostracoda,” Trans. Zool. Soc. London, 1865, vol. v. p. 374, pl. lx. fig. la-c. Carapace, as seen from the side, oblong, subquadrangular, higher in front than behind ; height equal to half the length; anterior extremity rounded, posterior scarcely rounded, and armed below the middle with three or four more or less prominent teeth; dorsal and ventral margins nearly straight, the former slightly elevated over the anterior hinge-joint. Outline, as seen from above, oblong-ovate. The valves are marked by three more or less distinct longitudinal ridges, the rest of the surface being covered with rather coarse angular punctations. Length 3’; inch (0°75 millim.). C. plicatula has been found in several localities, but is apparently not very common in any of them. Dr. Reuss records its occurrence in several localities in Austria, Bohemia, and Galicia; Dr. Egger in Germany, and M. Bosquet in the Miocene of the South of France, as well as in the “terrain subapennin supérieur de Perpignan.” If my identification of it with the C. retifastigiata of Rupert Jones be correct, it has also been noticed sparingly in the Suffolk Crag (‘‘Coralline” Crag) of England. The one specimen which I here figure and describe was found in the bed of “ [socardium cor” (Sables moyens d’Anvers). The species occurs at the present day living in the Levant and Eastern Mediterranean. CyTHERE cIcATRICOsA (Reuss). (Plate LXIV. figs. 3 a—3 d.) Cypridina cicatricosa, Reuss, Die fossil. Entom. dsterreich. Tertiar-Beckens, p. 27, pl. ix. fig. 21 a, b (1849). Cythere cicatricosa, Bosquet, Entom. foss. terr. Tertiair. France, p. 76, pl. ii. fig. 18 (1852) ; Brady, Crosskey, and Robertson, Post-tertiary Entom. of Scotland, &e., p. 151, pl. xiv. figs. 7-10. Cythere arborescens, Brady, Ann. & Mag. Nat. Hist. vol. xvi. pl. ix. figs. 5-8 (1865). VoL. X.—Ppart vill. No. 2.— August 1st, 1878. 3G 388 DR. G. 8S. BRADY ON THE OSTRACODA Carapace somewhat peachstone-shaped: viewed from the side, it is highest in the middle, the height being equal to two thirds of the length; anterior extremity well rounded and broad, posterior narrower and produced into a short almost obsolete beak ; dorsal margin strongly arched, inferior slightly sinuated in the middle. Seen from above, the outline is ovate, widest in the middle and tapering evenly to the pointed extremities ; width equal to half the length. End view broadly ovate, narrowed above and broadly rounded below. The valves are very convex, and are marked all over the surface with rather coarse and closely-set circular punctures. Length 3/5 inch (0°66 millim.). C. cicatricosa has been found in some of the Tertiary deposits of France and Germany, associated with the preceding species; also in some English Post-tertiary deposits. I do not know of its existence in the living state at the present day, unless it be consi- dered identical with C. convexa, Baird, to which species I must suppose that the remarks of M. Bosquet probably refer when he speaks of its being found on the coast of Italy. Our figure and description are drawn from a specimen found in the Sables moyens d’Anvers (zone a Bryozoaires). CYTHERE @DICHILUS, noy. sp. (Pl. LXIV. figs. 1 a-1d.) Carapace, as seen from the side, subquadrangular, rather higher in front than behind ; height equal to nearly two thirds of the length; anterior margin well rounded ; poste- rior angularly produced below, rounded off obliquely above ; dorsal margin sinuated in front of the middle; ventral evenly and rather strongly convex. Viewed from above, the shell is broadly ovate, tapering evenly to the extremities, both of which form broadly mucronate projections; the width in the middle is equal to half the length. End view irregularly and broadly ovate, narrowed towards the apex. The margins of the shell are much produced, rounded and swollen; just within the ventral border is a very conspicuous, thick, and rounded ridge, and near the centre of the valve a large, smooth, rounded tubercle; the rest of the surface is marked with rounded pittings, which are arranged in transverse rows. ‘The hinge-line on the dorsal aspect is marked by a longitudinal furrow; the ventral surface shows the strongly developed marginal plates of the shell with an intervening depression. Length 31; inch (0°75 millim.). This species occurs in all the three divisions of the group, inférieurs (Pectunculus- bed), moyens (zone a Bryozoaires), supérieurs (Zrophon-bed). CYTHERE PETROSA, noy. sp. (Plate LXIV. figs. 5 a-5 d.) Carapace tumid, wedge-shaped: seen from the side, subquadrangular ; height nearly uniform throughout, and equal to rather more than half the length; extremities equal and obliquely rounded ; dorsal and ventral margins parallel and almost perfectly straight, the dorsal being much the shorter of the two. Seen from above, the outline is ovate, very tumid behind, the greatest width being fully equal to the height, narrowed and OF THE ANTWERP CRAG. 389 subtruncate in front, broadly rounded behind. End view very broadly subovate, emar- ginate in the middle both at base and apex. Surface of the shell irregularly furrowed or undulated in a transverse direction, and marked with distant pittings, which have a tendency to follow the lines of the furrows. Length 3's inch (0°66 millim.). One specimen only of this very distinct species has occurred in the Sables moyens (zone 4 Bryozoaires). CyTHERE LimicoLa (Norman). (Plate LXIV. figs. 9a, 96.) Cythereis limicola, Norman, Nat.-Hist. Trans. Northumberland & Durham, vol. i. p. 20, pl. vi. figs. 1-4 (1865). Cythere nodosa, G. O. Sars, Oversigt af Norges marine Ostrac. p. 34 (1865). Cythere areolata, Brady, Trans. Zool. Soc. Lond. vol. v. p. 381, pl. lxi. figs. 2 a-d (1865). Cythere complexa, Brady, Brit.-Assoc. Report, p. 210 (1866). Cythere limicola, Brady, Monog. Recent Brit. Ostrac. p. 405, pl. xxxi. figs. 38-41, 48-46 ; Brady, Crosskey, and Robertson, Monog. Post-tert. Entom. of Scotland, &c., p. 154, pl. x. figs. 1-4. Carapace, as seen from the side, subtrapezoidal, rather higher in front than behind ; height equal to about two thirds of the length; anterior extremity obliquely rounded ; posterior produced in the middle and excavated above ; superior margin rather concave, abruptly angular at each extremity, inferior nearly straight. Seen from above, the outline is boat-shaped, widest behind, tapering gently toward the front and abruptly behind. End view subtriangular, base broad and flat, apex obtusely rounded, sides excavated. Shell-surface uneven, rugose, often obscurely reticulated, the interstices being finely punctate; a conspicuous rounded tubercle over the anterior, and two near the posterior hinge; a strongly marked longitudinal ridge runs along the valves just within the ventral margin. Length 35 inch (0°54 millim.). _ One valve only found in the “Sables moyens” (zone a Lsocardium cor). C. limicola is tolerably common in the North Sea in a living state, and has been found in many Post-tertiary deposits in Scotland, and also in Canada. CYTHERE LATIMARGINATA, Speyer. (Plate LXIV. figs. 8 a8 d.) Cythere latimarginata, Speyer, Die Ostrac. der Casseler Tertiirbild. p. 22, pl. iti. figs. 3a-d (1863) ; Brady, Crosskey, and Robertson, Monog. Post.-Tert. Entom. of Scotland, &c., p. 163, pl. xvi. fig. 6. - Cythere abyssicola, G. O. Sars, Oversigt af Norges marine Ostr. p. 43. Carapace, as seen from the side, oblong, higher in front than behind ; height equal to about half the length; anterior extremity broadly rounded, posterior rounded but much narrower ; superior margin sloping rather steeply backwards, and twice sinuated, in front of and behind the middle; inferior margin rather deeply sinuated in the middle. Seen from above, the outline is irregularly oval and compressed, the extremi- ties being wide and truncated; the sides only slightly convex, and marked by one or two rounded protuberances. The substance of the shell is very thick and hard; the 3G2 390 DR. G. 8. BRADY ON THE OSTRACODA surface of the valves usually more or less beset with small circular pits (but sometimes quite smooth), and having a large central rounded tubercle; anterior and posterior margins produced into wide, thickened, and rounded lips, and fringed with numerous fine teeth, usually a large number in front, but only five or six behind. Length 35 inch (0°75 millim.). This is perhaps the most abundant of all the species found in the Antwerp Crag, occurring in great plenty in the Panopea and Pectunculus-beds, not quite so commonly in the sables 4 Bryozoaires, and is quite scarce in the Zrophon- and Isocardium-beds. It is remarkable that the species has not been found in the English Tertiaries; and one specimen only is on record from the Post-tertiary deposit of Hopton Cliff, near Yarmouth. It occurs in a living condition in the northern portions of the North Sea (Norway and Shetland), also in the Gulf of St. Lawrence (Canada). CYTHERE WETHERELLII, Jones. (Plate LXIV. figs. 7a-7d.) Cythere wetherellii, Jones, Quart. Journ. Geol. Soc. x. p. 161, pl. iii. fig. 9; Tertiary Entomost. England, p. 26, pl. iv. fig. 15, and pl. vi. figs. 16a-16d. Carapace, as seen from the side, subquadrate, much higher in front than behind, greatest height equal to nearly two thirds of the length; anterior extremity broad, obliquely rounded; posterior narrowed, subtruncate, scarcely rounded ; dorsal margin sloping steeply, and slightly arched ; inferior somewhat convex and with a slight sinuation towards each extremity. Dorsal aspect broadly ovate, widest behind the middle, width nearly equal to the height ; extremities produced into two broad mucronate processes. End view very tumid, ovate, width and height about equal. Surface of the shell beautifully and sharply reticulated, the reticulations angular (hexagonal or subhexagonal) and coalescing on the ventral surface so as to form longitudinal furrows ; each valve forms a sort of curved aleform ridge along the ventral margin; and there is a large tubercle in the situation of the anterior hinge-joint, forming a distinct angle or gibbosity. Length 34 inch (1:05 millim.). A few specimens only of Cythere wetherellii have occurred in the Panopwa-bed (Sables inférieurs). In England it has been found by Professor Jones in the “ Middle Eocene” of the Isle of Wight. CYTHERE TARENTINA, Baird. (Plate LXIII. figs. 1a-1d.) Cythere tarentina, Baird, Proc. Zool. Soc. 1850, Annulosa, pl. xviii. figs. 31-33. Carapace, as seen from the side, somewhat wedge-shaped, much higher in front than behind, the greatest height being equal to more than half the length, and situated very near the anterior extremity ; anterior margin broad and obliquely rounded ; posterior narrowed almost to a point; superior margin sloping steeply, especially at the hinder end, very slightly arched; inferior gently convex, with a slight sinuation near the middle. The outline as seen from above is rhomboidal, the extremities truncate; the OF THE ANTWERP CRAG. 391 greatest width in the middle and fully equal to half the length. End view subtri- angular with convex sides, nearly equilateral. Surface of the shell slightly undulated, devoid of sculpture, except that a thin curved aleform ridge is developed towards the ventral margin of each valve; the anterior margin bears below the middle a variable number of short, blunt, irregular spines: and there are usually two or three of similar character at the posterior extremity of the shell; the lateral ala often has a single spine a little behind the middle; and the ventral surface of the shell is more or less marked with flexuous longitudinal grooves. Length 3/5 inch (1:3 millim.). Cythere tarentina has not previously been observed in the fossil condition ; but it still lives in many parts of the Mediterranean. It is a very distinct and fine species. Several specimens have been found in the sand from the Panopwa- and Pectunculus-beds. CYTHERE acuTicosTA, Egger. (Plate LXVI. figs. 5 a—5 d.) Cythere acuticosta, Egger, op. cit. p. 40, pl. vi. fig. 7. Carapace, as seen from the side, rhomboidal, equal in height throughout; height equal to nearly two thirds of the length ; anterior margin obliquely rounded off ; posterior obliquely truncate below the middle, rounded off above ; dorsal margin straight ; ventral also straight, except the anterior half, which is a little upturned. Seen from above, the outline is elongated, subhexagonal or subovate, sides nearly parallel but irregularly flexuous, width equal to half the length, tapering rather abruptly towards the extre- mities, which are pointed. End view quadrangular with irregularly waved sides, width equal to three fourths of the height. The surface of the shell is strongly sculptured with sharply cut ridges, running for the most part in a longitudinal direction, but irregularly flexed and anastomotic; one of them, more developed than the rest, is in the median line; the furrows between these ridges are excavated into pits of very irregular size and shape; the dorsal surface is marked by a very distinct longitudinal median groove with raised ridges; and the ventral surface has on each valve a flattened aleform plate which is sculptured with transverse and marginal excavations. Length qs inch (0°54 millim.). Our specimens of this species occurred in the Pectunculus- and Panopea-beds. Though the figures given by Dr. Egger differ rather considerably (especially in the end view) from my own, I am disposed to think that they are meant to refer to the same species, and on that supposition have adopted the specific name acuticosta. CYTHERE TRAPEZIA, nov. sp. (Plate LXVI. figs. 4 a4 d.) ? Cythere corrugata, Egger, op. cit. p. 35, pl. v. fig. 3. Carapace, as seen laterally, trapezoidal, rather higher in front than behind, height equal to more than half the length; anterior extremity obliquely rounded ; posterior subtruncate, rounded a little below, the lower angle not produced ; dorsal and ventral margins nearly straight, the latter very gently situated in the middle. The outline from 392 DR. G. 8S. BRADY ON THE OSTRACODA above is elongated, subhexagonal, sides nearly parallel but deeply sinuated in the middle ; sides gradually tapering to the anterior extremity, which is broad and truncate ; posterior extremity very wide, with a wide central mucronate projection. End view broadly ovate, height and width nearly equal, outline irregularly waved. The shell- surface is sculptured with large pits, which are arranged in obscurely radiating, some- what flexuous lines round a central tubercle; parallel to and just within the inferior margin is a distinct elevated ridge. Length 3); inch (0°75 millim.). This species may perhaps be identical with that called by Egger “ Cythere corrugata, Reuss ;” but if so Dr. Egger’s identification must be wrong, as Reuss’s figures certainly do not apply to the present species. It was found in the 7rophon-antiquum bed (Sables supérieurs). CyTHERE Macropora, Bosquet. (Plate LXVII. figs. la-ld, and Plate LXVI. figs. 6 a-6 d). Cythere macropora, Bosquet, Entom. fossil. terr. Tertiair. France, p. 97, pl. v. fig. 2; Jones, Tert. Entom. England, p. 35, pl. iii. figs. 9a-9e; Brady, Crosskey, and Robertson, Monog. Post-tert. Entom. Scotland, &c., p. 159, pl. xiv. figs. 1-3. ? Cythere logani, Brady & Crosskey, Geological Magazine, vol. viii. (1871), pl. u. figs. 8, 9. Cythere hornesi, Speyer, op. cit. p. 32, pl. i. fig. 7, and pl. iv. fig. 1. Carapace, as seen from the side, oblong, quadrangular, highest near the front, height equal to fully one half the length ; anterior extremity broad and well rounded, posterior rounded but narrower; dorsal and ventral margins nearly straight, the former sloping rather steeply backwards. The outline as seen from above is oblong and very irregular, the margins much jagged and waved; extremities produced, very broad and truncate, width about equal to the height. The shell-surface is beautifully sculptured with large angular pittings, which are arranged somewhat concentrically round a large rounded central tubercle ; the pitted portion of the valve ends at some distance from the hinder extremity, in an irregular, curved, abrupt declivity, and is connected with the anterior margin by a series of about six short radiating ribs. The anterior margin is fringed below the middle with a series of 12-16 short blunt teeth; the posterior margin also bears about six or eight distant irregular teeth of similar character. The ventral aspect of the shell shows a very broad, prominent central ridge formed by the swollen margins of the valves, and on each side a laterally produced expansion ornamented by two longitudinal rows of deep subangular excavations. Length 3; inch (1:05 millim.). C. macropora occurs in both beds of “Sables moyens”’ (abundantly in that of the zone & Bryozoaires), and also very abundantly in the Panopwa-bed (Sables infé- rieurs”). It is certainly one of the most distinctly marked and most beautiful of fossil Entomostraca. A recent Australian species (Cythere lactea) described by the present writer in 1865 (Trans. Zool. Soc. Lond. vol. v.) comes near to it in character, but is quite sufficiently distinct. The form represented in P]. LX VI. of this Memoir seems to belong to the young, and is probably identical with C. hornesi, Speyer. OF THE ANTWERP CRAG. 393 CYTHERE POLYTREMA, noy.sp. (Plate LXVI. figs. 1 a—1 d.) Carapace, as seen from the side, oblong, highest in front; height equal to more than half the length ; anterior extremity broad, well rounded and bordered below the middle with a number of irregularly placed, rather slender, but blunt spines; posterior extre- mity narrower, scarcely rounded, armed with a few (4-6) downward-pointing long and slender spines; superior margin sloping, arched and sinuated in front of the middle; inferior margin nearly straight. The outline viewed dorsally shows a central quadran- gular portion with two broad extremities projecting in the middle line; width equal to about half the length. The lateral aspect of the valves is marked by three longitudinal sharply-cut ribs, one in the middle, the others just within the superior and inferior margins respectively ; the lower is almost straight, the other two are curved; there is also a curved rounded ridge just within and parallel to the anterior margin; the rest of the surface is occupied by large angular excavations arranged in a reticulated manner ; the dorsal and ventral surfaces show also longitudinal ribs, the intervals between which are filled with single rows of angular pits. Length 3's inch (0-98 millim.). Cythere polytrema occurs in moderate numbers in the two beds of the Sables infé- rieurs and in the /socardium-bed of the Sables moyens. Like the preceding, it is an extremely fine and well-marked species; but the valves seem usually to occur separate. CYTHERE SCABROPAPULOSA, Jones. (Plate LXVI. figs. 2 a, 2 6.) Cythere scabropapulosa, Jones, Tert. Entom. England, p. 31, pl. v. fig. 16. Valves, as seen laterally, oblong, subovate, higher in front than behind ; height quite equal to half the length; anterior extremity rounded; posterior narrowed, rounded, and somewhat produced in the middle; superior margin elevated in front, then sinuated, and distinctly arched behind ; inferior nearly straight for the greater part of its course, curved upwards towards the posterior extremity. Dorsal outline oblong ovate, with jagged or crenate margins. The surface of the shell is closely beset with wart-like tubercles of considerable size ; and the anterior portion has, just within and parallel to its margin a row of bead-like tubercles; a prominent tubercle over the anterior hinge- joint. Length 3'5 inch (0:98 millim.). One valve only of this species was detected, in the Pectunculus-bed (Sables inférieurs). Professor Jones’s specimens were found at Bracklesham in the Middle Eocene. CYTHERE DAwson1?, Brady and Crosskey. (Plate LXVI. figs. 3a, 3 6.) Cythere dawsoni, Brady and Crosskey, Geological Magazine, Feb. 1871, vol. viii. pl. ii. figs. 5, 6. Carapace, seen from the side, oblong, subquadrate, highest in front. Greatest height equal to one half the length; anterior extremity well rounded, posterior narrower and not so fully rounded; dorsal and ventral margins nearly but not quite straight; the former with a prominence at each end. Outline as seen from above compressed, sub- 394 DR. G. 8. BRADY ON THE OSTRACODA ovate, with wide projecting extremities. The valves have a rugose, tuberculated rib running diagonally across from before backward, and are otherwise irregularly sculptured in a rugose manner. Length 3/5 inch (0-98 millim.). One specimen in the Zrophon-bed (Sables supérieurs), one in the Panopea-bed (Sables inférieurs), and one or two in the Zsocardiwm-bed (Sables moyens). In general aspect these specimens are rather like Cythere costata, Brady, but differ in their style of sculpture and in the fact of the postero-inferior angle being rounded off instead of being produced as in C. costata, with a dentate projection. CYTHERE SUBCORONATA, Speyer. (Plate LXVII. figs. 4 a—4 d.) Cythere subcoronata, Speyer, loc. cit. p. 38, pl. iv. fig. 9 (1863). ? Cythere latidentata, Bornemann, “ Die mikroscop. Fauna des Septarienthones von Hermsdorf,” Zeitschr. d. deutsch. geol. Ges. 1855, p. 366, pl. xxi. fig. 6. ? Cythere horrescens, Jones, Tertiary Entom. p. 38, pl. v. figs. 9, 17a, 176 (not Cythere horrescens, Bosquet ; nor Cythereis subcoronata, Brady, Trans. Zool. Soc. Lond. vol. v. p. 384, 1865). Carapace, as seen from the side, oblong, subovate, somewhat higher in front than behind; height equal to rather more than half the length; extremities rounded, the anterior armed below the middle with a series of strong, blunt, projecting spines, the posterior with a few broader and less developed somewhat triangular teeth ; dorsal margin sloping in an almost straight line, but tuberculated and terminating behind in a very large and much elevated blunt spine or tubercle ; ventral margin convex. Outline as seen from above subovate or almost hastate, irregularly jagged or laciniate, widest towards the hinder extremity, the width being equal to the height. End view sub- triangular. The surface of the valves is beset with irregularly scattered, large, rounded tubercles, and along the ventral and dorsal margins with a row of blunt tooth-like processes. Length 3; inch (0°98 millim.). Though Professor Jones’s figures differ, more especially in the sharply spinous character of the armature, from the Antwerp specimens, I think it extremely likely that they really apply to mere varieties (perhaps sexual) or to stages of growth of the present species. And I also strongly suspect that the species itself, as illustrated in Pl. LXVII. figs. 2a-d of this Memoir, may only be the immature form of Cythere mucronata, a strikingly developed specimen of which is shown in figs. 3 a-d of the same Plate. If the two series of figures be carefully compared, it will be seen that they differ scarcely at all, except in the degree of development of the various parts; and though I hesitate, in the absence of a series of specimens exhibiting the inter- mediate stages of growth, to unite them under one specific name, I really entertain very little doubt as to the propriety of doing so. C. subcoronata has been found sparingly in the Pectunculus-bed (Sables inférieurs) and in the zone a Jsocardium (Sables moyens). OF THE ANTWERP CRAG. 395 CYTHERE MucRoNATA, Sars. (Plate LXVII. figs. 3 a-3 d.) Cythere mucronata, G. O. Sars, Oversigt af Norges Marine Ostrace. p. 48. Cythere spinosissima, Brady, Trans. Zool. Soc. (1865), vol. v. p. 386, pl. lx. figs. 10 a-e. Carapace, as seen from the side, subquadrangular, of nearly equal height throughout, height equal to fully half the length; extremities rounded; superior and inferior margins nearly parallel ; the entire circumference (except the upper half of the anterior margin) beset with blunt squamous spines, which are often dilated at the free extre- mity, those of the postero-inferior angle being very long. Seen from above, the outline is rhomboidal, much broken up with spinous projections. The sides of the valves are beset with flattened squamous spines, often so much dilated at the end as to become quite fan-shaped. End view hatchet-shaped. Length 35 inch (1°35 millim.). The spinous armature of this fine species is subject to great variation in the extent of its development, the figures given in Pl. LXVII. exhibiting the most extreme form with which I am acquainted; it is found in the living state off the coasts of Norway. The fossil specimens occurred in the “Sables moyens,” zone a Bryozoaires. CyTHERE JONES (Baird). (Plate LXVII. figs. 2 a-2 d.) Cythereis jonesii, Baird, Brit. Entom. p. 175, pl. xx. fig. 1 (1850) ; Norman, Nat.-Hist. Trans. Northumberland and Durham, vol. i. p. 21, pl. vii. figs. 5-8 (1865). Cythere ceratoptera, Bosquet, Entom. foss. terr. Tert. France, &c. (1852), p. 114, pl. vi. fig. 2. Cythereis ceratoptera, Jones, Monog. Tert. Entom. Eng. p. 39, pl. iv. fig. 1 (1856). ? Cythereis cornuta, Jones, Entom. Tert. Form. Eng. p. 39, pl. iv. fig. 19. Cythereis finbriata, Norman, Ann. & Mag. Nat. Hist. ser. 3, vol. ix. pl. ui. fig. 9 (1862). Cythereis spectabilis, Sars, Oversigt af Norges marine Ostrac. p. 46. Cythere jonesii, Brady, Monog. Recent Brit. Ostrac. p. 418, pl. xxx. figs. 13-16 (1865); Brady, Crosskey, and Robertson, Monog. Post-tert. Entom. Scotland, &c., p. 171, pl. xii. figs. 4-7. Carapace, as seen from the side, oblong subovate, higher in front than behind ; height equal to more than half the length; anterior extremity rounded, fringed with blunt spines ; posterior also rounded, bearing a smaller number (about 5) of longer spines ; dorsal and ventral margins nearly parallel and much broken up into irregular spines ; over the anterior hinge is one very large and strong spine. Outline, as seen from above, rhomboidal, greatest width behind the middle and equal to about two thirds of the length. End view triangular. The surface of the valves, except along the margins, is smooth and gently undulated ; there is a distinct elevated and rounded ridge just within the anterior and posterior extremities ; and the general surface of the valves suddenly sinks to a lower plane at a little distance behind the middle. Length 35 inch (1-1 millim.). I was at one time disposed to think that the form described by M. Bosquet as C. cera- toptera might well be kept apart as a distinct variety of C. jonesii; but, after exami- nation of a large number of recent and fossil specimens, I now believe that there is no character sufficiently persistent to warrant even this separation, the chief variations VoL. X.—PartT vill. No. 3.—August 1st, 1878. 3H 396 DR. G. 8. BRADY ON THE OSTRACODA being in the shape, regularity, and degree of development of the spines. ‘The spines vary in shape from that of evenly rounded blunt teeth of regular size to that of long and slender, or flat, squamous processes. In the best-developed recent specimens (those especially from the North Sea) the first-named condition occurs, and the rows of spines are arranged with much regularity ; in others (as for instance in many specimens from the Mediterranean and in most of the fossil examples) the spines are less regular in arrangement, and tend either to become few, long and slender, or flattened and squa- miform: but there are all shades of gradation between these extreme types. The figures in Pl. LXVII. are taken with great accuracy from a fine fossil specimen illustrating an intermediate condition, but with rather a marked tendency to a squamous form of the spines, except on the posterior margin. C. jonestt occurs in all the Antwerp beds except in that of Zrophon antiquum, but nowhere in much abundance. CyTWERE LIMA, Reuss. Cythere lima, Reuss, Zeitschrift d. deutsch. geol. Ges. 1855, p. 280, t. x. fig. 7. One valve, referable apparently to this species, was found in the Panopewa-menardi bed; but, owing to. its having been lost or mislaid, I am unfortunately not able to describe it. Genus CYTHERIDEA, Bosquet. Valves unequal, ovate or subtriangular, highest near the front; smooth or marked with scattered circular papille, with impressed puncta or concentric furrows; hinge composed of two crenulated crests on one valve, which articulate with depressions of the opposite valve. Upper antenne very robust, mostly 5-jointed, spinous; lower 4-jointed. Mandibular palp 3-jointed, and having a distinct branchial appendage. Right foot of the first and second pairs of feet, in the male, different from the rest, that of the first very strong and prehensile, of the second very feeble; the apex rudimentary and destitute of a terminal claw. Eyes distinct. CYTHERIDEA PAPILLOSA, Bosquet. Cytheridea papillosa, Bosquet, Entom. fossil. terr. Tertiair. France, p. 42, pl. u. fig. 5 (1852) ; Brady, Trans. Zool. Soc. Lond. (1865), vol. v. p. 370, pl. lvii. figs. 8a-g; Brady, Monog. Recent Brit. Ostrac., Trans. Linn. Soc. vol. xxvi. (1868) p. 423, pl. xxvii. figs. 1-6, pl. xl. fig. 1; Brady, Crosskey, and Robertson, Monog. Post-tert. Entom. Scotland, p. 176, pl. vi. figs. 12-15 (1874). Cythere bradiit, Norman, Nat.-Hist. Trans. Northumberland and Durham, vol. i. p. 15, pl. v. figs. 5-8 (1865). Cyprideis bairdii, Sars, Oversigt af Norges Marine Ostracoder, p. 52 (1865). Var. levis. (Plate LXII. figs. 1 a1 d.) Carapace, as seen from the side, subovate, highest a little in front of the middle, height equal to half the length; anterior extremity broadly and evenly, posterior only OF THE ANTWERP CRAG. 397 slightly rounded ; dorsal margin forming a continuous curve from its highest point to the infero-posterior angle; ventral margin almost straight. Seen from above the out- line is ovate, widest in the middle, and tapering evenly to the extremities, which are rather obtuse; width nearly equal to the height. End view nearly circular. Surface of the shell perfectly smooth. Length 3/5 inch (0°85 millim.). This differs from the typical form of the species only in being entirely destitute of papillose sculpture. The papillose form (which, however, varies very much in its surface ornament) has been found in the fossil state in most of the Post-tertiary beds of Scotland, as well as in Norway and Canada, and by M. Bosquet in many of the Tertiary beds of France. In the living state it occurs plentifully in the seas off Norway and Great Britain and Spitzbergen, as well as in Baffin’s Bay and the Gulf of St. Lawrence. My memoranda of the particular bed or beds in which this form occurred have unfortunately been lost. CYTHERIDEA PINGUIS, Jones. (Plate LXII. figs. 3 a—3 d.) Cytheridea pinguis, Jones, Tertiary Entomostraca of England, p. 43, pl. ii. figs. 4a—4h. ? Cytheridea fabeformis, Speyer, op. cit. p. 52, pl. ii. fig. 1. Carapace, as seen from the side, ovato-triangular, highest a little in front of the middle; height equal to more than half the length; extremities rounded, dorsal margin very boldly arched, almost gibbous; ventral slightly convex; seen from above the outline is ovate, scarcely at all tapering to the extremities, which are rather broadly rounded; the width is nearly the same throughout, and is equal to half the length. End view nearly circular. The shell-surface is nearly smooth, but is usually covered with closely-set small punctures. Length 3'; inch (0°9 millim.). Professor Jones’s specimens were found in the Pliocene of Suffolk. Those described in this memoir are from all the Antwerp beds except only that of the “Sables 4 Bryo- zoaires.” It is one of the more abundant species. CYTHERIDEA CYPRIDIOIDES, nov. sp. (Plate LXIX. figs. 6 a—6 e.) Carapace tumid, ovate: seen from the side the outline forms about two thirds of a circle, the dorsal margin being excessively arched, and the ventral slightly convex; the extremities are rounded, the posterior much the narrower and rather flattened; greatest height equal to two thirds of the length. Outline seen from above regular ovate, extre- mities obtuse ; width equal to more than half the length. Surface of the shell perfectly smooth. Length <5 inch (0°85 millim.). A few specimens of this species were found in the “Sables moyens” (zone a Bryo- zoaires). Iam by no means sure that the form called in this memoir C. papillosa, var. levis, may not be merely the male of this species. CYTHERIDEA MULLERI (Minster). (Plate LXII. figs. 4 a—4 e.) Cythere miilleri, Minster, Jahrb. fir Mineralogie, 1830, p. 62, and Neues Jahrb. 1835, p. 446 (fide Jones and Bosquet). 3H 2 398 DR. G. 8. BRADY ON THE OSTRACODA Cytherina miilleri, Romer, Neues Jahrb. fiir Mineralogie, 1838, p. 516. t. vi. fig. 6 (fide Jones and Brady) ; Reuss, Haidinger’s Abhandl. 1850, p. 55. pl. viii. fig. 21. Bairdia hagenowi, Reuss, Zeitsch. d. deutsch. geol. Ges. 1855, p. 60, pl. ix. fig. 93. Cytheridea heterostigma, Reuss, ibid. p. 60, pl. 1x. fig. 94. Cytheridea miilleri, Bosquet, Entom. foss. terr. Tert. France, p. 39, pl. ii. fig. 4; Jones, Tert. Entom. England, p. 41, pl. v. figs. 4a-e & 5, pl. vi. figs. 10a, 106, & 11-13; Egger, Die Ostrak. der Miocin-Schicht. Orenburg, p. 18, pl. ii. fig. 7; Speyer, Die Ostrac. der Casseler Tertiirbild. p- 48, pl. i. fig. 8; Brady, Trans. Zool. Soc. Lond. vy. p. 371, pl. lviii. figs. 1la-d. Cytherina intermedia, Reuss, Haidinger’s Abhandl. ii. p. 86, ‘pl. xi. fig. 12. Cytherina seminulum, Reuss, ibid. p. 59, pl. ix. figs. 5-8. Carapace of the female subovate, tumid: seen from the side, ovate-triangular, highest a little in front of the middle; height equal to rather more than half the length; ante- rior extremity well rounded, posterior narrowed, obliquely rounded, and often somewhat exserted at the inferior part, forming a rounded obtuse angle; dorsal margin arched, usually more or less angulated in front of the middle; ventral margin nearly straight. Seen from above, ovate, sides subparallel, tapering abruptly to the extremities, width less than half the length. End view very broadly ovate. The surface is marked with numerous impressed rounded punctures, which often tend to arrange themselves in curved transverse furrows, and on the ventral surface to coalesce into longitudinal furrows; the margins are often entirely smooth ; but frequently the anterior border is armed below the middle with a row of six or eight sharp spines on each valve; the posterior extremity also sometimes bears a single spine at its lower angle; this is situated on the right valve. The shell of the male is, as usual, more compressed and elongated. Length 3; inch (0-9 millim.). This is one of the most common species in the Antwerp beds, and has been found in all of them except the zone of Lsocardium cor.; it has also been found in most of the Tertiary formations of Europe, in Austria, Bohemia, Hesse, Westphalia, France, and the Netherlands (Kocene), in Touraine (Miocene) and in the Netherlands (Pliocene) ; it has also been noticed by Professor Rupert Jones in many of the Tertiary beds of England, and in a Tertiary Clay from Australia. I have myself seen recent specimens from Smyrna, the Levant, and Australia. Genus Loxoconcna, G. O. Sars. Valves nearly equal, subrhomboidal, and mostly flexuous in outline ; surface smooth, or marked with concentrically arranged impressed puncta; or with polygonal fosse, often also with minute circular papille; ventral margin usually forming a prominent compressed keel behind the middle; postero-superior angle obliquely truncate ; hinge- joint formed by two small teeth at the extremities of the hinge-line of each valve. Limbs of the animal slender and colourless. Upper antenne very slender, 6-jointed, the last joint very long, linear, and bearing long simple sete ; lower antennz 4-jointed, the third joint long and narrow. Flagellum long and biarticulate. Mandibular palp OF THE ANTWERP CRAG. 399 3-jointed, bearing a distinct branchial appendage. Lowest seta of the branchial plate of the first pair of jaws deflexed. Feet long and slender, alike in male and female. Abdomen terminated by a hairy conical process; postabdominal lobes bearing two moderately long, subequal sete. LOXOCONCHA LATISSIMA, nov. sp. (Plate LXVIII. figs. 1 ald 2, 1le-Lh3¢.) Carapace of the female very tumid: as seen from the side, subrhomboidal, of nearly equal height throughout; height equal to two thirds of the length; the extremities are obliquely rounded, dorsal margin almost straight, ventral sinuated in front of the middle. Outline as seen from above broadly and regularly ovate, greatest width situated in the middle and equal to nearly two thirds of the length, extremities mucronate. Surface smooth, marked with numerous small impressed punctures, and on the anterior and ventral surfaces by longitudinal strie. Length 3; inch (0°60 millim.). The shell of the male is longer (;'3 inch) and more compressed. LL. latissima occurs in both beds of the Sables inférieurs and also in the “zone a Bryozoaires” of the Sables moyens. It was moderately abundant in all these formations. LoxXocoNcHA BITRUNCATA, nov. sp. (Plate LXVIII. figs. 2 a-2 d). Carapace, as seen from the side, oblong, subrhomboidal; extremities rounded, the posterior produced in the middle; dorsal margin nearly straight, ventral strongly convex; greatest height equal to fully half the length. Seen from above the outline is oblong, quadrangular, with wide truncate extremities and subparallel sides, which, however, bulge out in the middle; width equal to half the length. Surface of the shell covered with large and closely-set polygonal excavations. Length ~5 inch (0:65 millim.). This species was found in the Zrophon-bed of the Sables supérieurs and in the Isocardium-bed (Sables moyens). In the former it is very abundant, and in the latter moderately so. It is a very well marked and distinct species, very similar in style of surface-marking to L. guttata (Norman), but wholly different from it or from any described species in the shape of the shell. LOXOCONCHA GRATELOUPIANA (Bosquet). (Plate LX VIII. figs. 3 a—3 g.) Cythere grateloupiana, Bosquet, Entom. foss. terr. Tertiair. France, &c., p. 81, pl. iv. fig. 3. Carapace of the male (?), as seen from the side, subrhomboidal ; height equal to more than half the length; extremities obliquely rounded, the posterior bevelled off above the middle; dorsal margin straight, ventral slightly sinuated in front of the middle. Outline as seen from above regularly ovate, with projecting, sharply mucronate extre- mities, width about equal to the height. End view nearly circular. Surface marked with moderately large subrotund pittings, which have a tendency to arrange themselves in flexuous sublongitudinal rows, especially towards the margins of the valves. Length gz inch (0°75 millim.). 400 DR. G. 8. BRADY ON THE OSTRACODA Found rather sparingly in both beds of the Sables inférieurs, and also in the zone 4 Isocardium cor (Sables moyens). LOXOCONCHA VARIOLATA, nov. sp. (Plate LX VIII. figs. 4 a—4 d.) ? Cythere subtriangularis, Speyer, Ostracoden der Casseler Tertiarbild. p. 26, pl. il. fig. 6. Carapace, as seen from the side, oblong, subrhomboidal; anterior and posterior extremities almost exactly alike, rounded below and bevelled off above the middle; dorsal margin straight, ventral convex and slightly sinuated in front; height equal to more than half the length. Seen from above, ovate, tapering rather suddenly to the extremities, which are sharply mucronate; width about equal to the height. End view subquadrate, somewhat narrowed and rounded off at the apex. Shell-surface marked by rather large polygonal excavations, which tend to run in longitudinal rows, especially towards the anterior margin and on the ventral surface. Behind the middle of the ventral margm the valves are produced into a distinct but not very prominent angulated aleform ridge. This, however, is absent or obsolete in some specimens, which seem to belong to this species, as they agree with it in other respects. Length », inch (0°66 millim.). This species may very possibly be identical with Cythere subtriangularis, Speyer, the figures of which agree almost entirely with it, except in the absence of any angulated ridge; for the present, however, I prefer to consider it as distinct. Cythere hastata, as figured by Egger and (more doubtfully) by Reuss, are also either the same or nearly allied forms; without examination of authentic specimens it is impossible to pronounce with certainty. Lastly, the species described by myself (Loxoconcha angustata and L. alata) are both nearly allied but distinct, as also is L. multifora (Norman). L. variolata has been found pretty plentifully in the Pectunculus-bed and in the zone & Bryozoaires, and less abundantly in the Panopwa-menardi bed (Edeghem and Kiel). Genus XESTOLEBERIS, G. O. Sars. Shell smooth and polished, ornamented with small, round, distant papille, much lower in front than behind, in the female very tumid behind; hinge-joint formed by a dentated projecting crest of the left, which is received into an excavation of the right valve ; ventral margin of both valves incurved in front of the middle. Upper antenne 6-jointed, the last four joints successively decreasing in length and bearing very short simple sete ; lower antenne short, 4-jointed. Flagellum of moderate length. Mandible- palp 4-jointed. Branchial appendage small and bearing only two sete. Maxille as in Loxoconcha. Feet short. Postabdominal lobes bearing two sete. yes distinct. Ova and immature young borne within the shell of the female. XESTOLEBERIS DEPRESSA, Sars. (Plate LXVI. figs. 8 a-8 d.) Xestoleberis depressa, G. O. Sars, Oversigt af Norges marine Ostracoder, p. 68 (1865); Brady, OF THE ANTWERP CRAG. 401 Monogr. Recent Brit. Ostrac. p. 438, pl. xxvii. figs. 27-33 (1868) ; Brady, Crosskey, and Robert- son, Monog. Post-tert. Entom. Scotland, &c. p. 190, pl. vii. figs. 13-19 (1874). ? Cytherina tumida, Reuss, Foss. Entom. ésterr. Tertiir-beck. p. 57, pl. viii. fig. 29 (1850); Egger, Ostrak. Miocén-Schicht. Orenburg, p. 17, pl. ii. fig. 11 (1858). ? Cytherina impressa, Reuss, Foram. u. Entom. Kreidemergeis v. Lemberg. (Haidinger’s Abhandl. band iv. 1850) p. 48, pl. vi. fig. 5. Carapace of the female, as seen from the side, subsemicircular or subreniform, highest in the middle, height equal to more than half the length, depressed and rounded off in front, broadly rounded behind; superior margin boldly arched, inferior very slightly sinuated in front of the middle. Seen from above, broadly ovate, widest behind the middle, width equal to two thirds of the length, pointed in front and broadly rounded behind. End view subtriangular, with rounded angles and convex sides, width greater than height. Surface of the shell perfectly smooth. Length 5 inch (0°65 millim.). This is a widely distributed and common recent species, living usually in depths of from 2 to 30 fathoms, occurring in all parts of the North Sea and as far as Spitzbergen and Canada. It occurs commonly as a Post-tertiary fossil in Scotland and in Norway and Canada. In the Antwerp sands a few specimens only occurred in the Sables moyens (zone a Bryozoaires). Genus CyTHERURA, G. O. Sars. Valves unequal and dissimilar in form, right more or less ovarlapping left on dorsal margin ; surface reticulated, punctated, deeply excavated or bearing irregularly-disposed ribs or protuberances, mostly marked with a central areola of darker colour than the rest of the shell; posterior extremity produced into a more or less prominent beak. Superior antenne shortly setose, 6-jointed, gradually tapering; second joint bearing a rather long seta on the middle of the posterior margin; inferior antenne 5-jointed, terminal claws short. Flagellum long, inarticulate. Mandibles robust, with very blunt teeth. Palp 3-jointed. Branchial appendage small and bearing only two recurved sete. Terminal lobes of the first pair of maxillz long and narrow. Branchial plate bearing on its external margin two non-ciliated sete, which are directed downwards and arise from a separate lobe. Feet small, the terminal claws short and curved. Eyes distinct. Copulative organs of male very complex, provided with several irregular processes, and a very long spirally convoluted tube; usually very minute. CYTHERURA BROECKIANA, nov. sp. (Plate LXIX. figs. 5 a—6 d.) Carapace, seen laterally, suboval, highest in the middle; height equalling more than half the length; anterior extremity rounded ; posterior produced about the middle, but scarcely beaked ; dorsal margin boldly arched, somewhat flattened in the middle; ventral slightly convex. Seen from above, ovate ; sides nearly parallel, tapering suddenly to the extremities, which are obtusely pointed; width equal to the height. End view sub- 402 DR. G. 8. BRADY ON THE OSTRACODA triangular, with rounded angles and broad base, tapering upwards; surface of the shell smooth, marked by a few very small, distant punctations, and by exceedingly faint longitudinal strie. Length 3'5 inch (0°5 millim.). The shell of the male (fig. 5 ¢) is much more attenuated and tapering at the hinder extremity. This species occurred very sparingly in the two beds of the Sables inférieurs. It is very closely allied to (though somewhat different in shape from) C. fulva, a species described by myself and Mr. Robertson from dredgings made at the Scilly Islands. I have much pleasure in naming this species in honour of M. Ernest Vanden Broeck, not only as a personal acknowledgment of his courtesy to myself, but as a tribute to his services to science in the investigation of the geology of Belgium &c. CYTHERURA cornuTa, Brady. (Plate LXVI. figs. 7 a—7 d.) Cytherura cornuta, Brady, Monogr. Recent Brit. Ostrac. p. 445, pl. xxii. figs. 12-15 (1868) ; Brady, Crosskey, and Robertson, Monogr. Post-tert. Entom. Scotland, &c., p. 199, pl. xiii. figs. 23-25 (1874). Carapace of the male, as seen from the side, oblong, rhomboidal, nearly equal in height throughout; height equal to half the length; anterior extremity rounded, posterior obliquely truncate and produced above the middle into a large obtuse beak; dorsal and ventral margins nearly straight. Seen from above, the outline is ovate, with two rect- angular aleform projections behind the middle; anterior extremity acutely, posterior obtusely pointed ; width equal to about two thirds of the length. End view subtri- angular, with convex and irregularly crenate sides. Shell-surface sculptured with conspicuous waved longitudinal ridges, and having a rectangular transverse ridge near the posterior extremity, which ridge ends near the ventral margin in a strong cornute projection. Length ;}5 inch (0°44 millim.). One or two examples only of this species have been found in the “zone 4 Bryozoaires ” and “‘zone & Panopwa menardi.” The specimen figured I judge, from its elongated form, to have been probably a male. Genus CytHERopreron, G. O. Sars. Valves mostly subrhomboidal, tumid, unequal, and different in shape, the right more or less overlapping the left on the dorsal margin; surface of the shell smooth or variously sculptured, punctate, papillose, reticulated, or transversely rugose; ventral margin produced laterally into a prominent rounded or spinous ala, posterior extremity into a more or less distinct beak ; hinge formed by two small terminal teeth on the right, and by a minutely crenated median bar on the left valve. Upper antenne shortly setose and composed of five joints; penultimate joint elongated, and bearing on the middie of the anterior margin two hairs; lower antenne distinctly 5-jointed. Flagellum Jong. Mandibles of moderate size. Palp 3-jointed. Branchial appendage bearing two OF THE ANTWERP CRAG. 403 very small sete. Jaws as in Cytherwra. Feet long and slender, terminal claws slender. Abdomen ending in a long, narrow process; postabdominal lobes bearing three short hairs. Copulative organs of the male armed behind with three spiniform processes, one of which is trifurcate. Eyes wanting. CYTHEROPTERON LATISsiMuM, Norman. (Plate LXIX. figs. 1 a1 d.) Cythere latissima, Norman, Nat.-Hist. Trans. Northumberland and Durham, vol. i. p. 19, pl. vi. figs. 5-8 (1865) ; Brady, Trans. Zool. Soc. vol. v. p. 381, pl. Ixii. figs. 4a-e (1866). Cytheropteron convexum, Sars, Oversigt af Norges marine Ostrac. p. 80 (1865). Cytheropteron latissimum, Brady, Monog. Recent Brit. Ostr. p. 448, pl. xxxiv. figs. 26-30; Brady, Crosskey, and Robertson, Monog. Post-tert. Entom. Scotland, &e. p. 202, pl. viii. figs. 19-23. Carapace, as seen from the side, subovate, highest in the middle, height equal to two thirds of the length; anterior extremity rounded, posterior rounded but angulated above, scarcely beaked; dorsal margin boldly and evenly arched; ventral convex, sinuated in front. Seen from above, nearly rhomboidal, greatest width behind the middle and equal to two thirds of the length, extremities acuminate. End view equilaterally triangular, sides convex. Surface marked with oblong pits, which are arranged in flexuous transverse grooves; ventral surface longitudinally furrowed ; lateral aleform process rounded and not very prominent. Length 35 inch (0°60 millim.). This species is widely distributed in the present day over the North Sea, and as far as Baffin’s Bay and Spitzbergen. It occurs commonly as a fossil in the Post-tertiary beds of Scotland, at Bridlington in England, and also in Canada. The specimens here described are from the Panopwa-bed (Sables inférieurs), and the zone 4 Bryozoaires (Sables moyens). CYTHEROPTERON INTERMEDIUM, nov. sp. (Plate LXIX. figs. 3 a-3c.) Carapace, as seen from the side, oblong, rounded in front, produced behind into a broad obtuse median beak; dorsal margin moderately arched, ventral nearly straight. Seen from above, rhomboidal, greatest width situated behind the middle, and equal to two thirds of the length; the lateral margins end abruptly in a rectangular projection behind the middle, the hinder third of the shell being much narrower, and forming an acutely tapering process ; the anterior extremity is also sharply pointed. End view triangular, with nearly straight margins and acute angles. Surface of the shell slightly furrowed transversely, otherwise smooth. Length 5 inch (0°50 millim.). A few examples only were found, in the Panopwa-bed (Edeghem). CYTHEROPTERON GRADATUM (Bosquet). (Plate LXIX. figs. 4a-4 d.) Cythere gradata, Bosquet, Entom. fossil. terr. Tertiair. France, &c. p. 127, pl. vi. figs. 11 a-d (1852). Cythere papilio, Egger, Ostrak. der Miocin-Sch. Orenburg, p. 42, pl. vi. fig. 9. Cythere bilacunosa, Speyer, Die Ostrac. Casseler Tertiirbild. p. 34, pl. iv. fig. 6 (1863). Carapace, as seen from the side, oblong, quadrangular, slightly higher in front than VOL. X.—PART vill. No. 4.—August 1st, 1878. 31 404 DR. G. 8. BRADY ON THE OSTRACODA behind, height equal to nearly half the length ; anterior extremity rounded, subtruncate ; posterior produced in the middle into a very acute beak, the upper half being obliquely truncate and jagged, the lower half excavated into two large rectangular indentations ; dorsal margin straight, ventral slightly sinuated. Outline as seen from above lancet- shaped, with a large triangular projection in the middle, behind; widest behind the middle, width equal to nearly two thirds of the length; the anterior extremity is pointed and broadly rounded, as a lancet; lateral margins nearly straight to near the hinder extremity, then suddenly sinking so as to form one and then a second rectan- gular excavation ; posterior extremity acuminate. End view irregularly arcuate, width considerably greater than the height; surface of the shell irregularly waved and nodulated, the posterior portion marked by two prominent and sharply defined trans- verse ridges, which end below in sharp rectangular processes, and above are continued in longitudinal curves toward the middle of the valves. Length 7; inch (0°60 millim.). A few specimens only found, in the Pectunculus-bed (Sables inférieurs) and zone a Bryozoaires (Sables moyens). It has been found by M. Bosquet in several Eocene deposits in France and Belgium, by Egger in the Miocene of Orenburg, and by Speyer in the Eocene of Cassel. CYTHEROPTERON PIPISTRELLA, noy. sp. (Plate LXIX. figs. 2 a—2 d.) Carapace, as seen from the side, subrhomboidal, with a very large and acutely produced triangular alzeform process, which projects below the middle of the ventral margin ; anterior extremity rounded, posterior produced in the middle into a wide truncated beak; dorsal margin excessively arched, gibbous, highest in the middle; ventral margin convex, hidden in the middle by the projection of the ventral ala. Seen from above, broadly sagittate, the lateral ale spreading widely, and forming behind the middle of the shell two acute backward-pointing processes, between which and the posterior extremity the margins form a deeply excavated arch; the extreme width between the apices of the ale is just equal to the length of the shell. End view acutely triangular, the lateral margins nearly straight, ventral margin deeply arched and obscured in the middle by a triangular projection formed by the anterior margin. Surface smooth; lateral ale channelled and marked by a deep hollow at the base. Length 35 inch (1 millim.). This very fine species occurred not uncommonly in the Sables moyens (zone a Bryo- zoaires. Genus BytnocyTHERE, G. O. Sars. Valves subequal, smooth or very sparingly sculptured, almost destitute of hairs; thin and fragile ; hinge-joint quite simple, or composed of a slight bar and furrow; no teeth. Upper antennz elongated, 7-jointed; second joint large and thick, with a seta on each margin, the other joints much narrower. Lower antennx moderately robust, 4-jointed ; second joint large. Mandibles strongly toothed ; palp 4-jointed, with a well- OF THE ANTWERP CRAG. 405 developed branchial plate. Terminal lobes of first maxille short and thick. Branchial plate large and ovate, setose, and with four long deflexed sete at base. Feet elongated, with long, slender claws. Abdomen ending in a large acuminate process; post-abdo- minal lobes narrow, and bearing three hairs. Eyes mostly absent. BYTHOCYTHERE CONSTRICTA, Sars. Bythocythere constricta, G. O. Sars, Oversigt af Norges marine Ostrac. p. 85 ; Brady, Monog. Recent Brit. Ostr. p. 451, pl. xxv. figs. 47-52 ; Brady, Crosskey, and Robertson, Monog. Post-tertiary Entom. Scotland, &c., p. 208, pl. xvi. figs. 9, 10. Carapace, as seen from the side, rhomboidal; height nearly equal throughout, and exceeding half the length; anterior extremity obliquely rounded; posterior obliquely truncate, rounded off at its upper angle; dorsal margin straight, ventral slightly sinuated near the middle, convex and inclined upwards behind. Seen from above, lozenge-shaped, distinctly constricted in the middle. Surface marked with delicate grooves running mostly in a subconcentric manner, but frequently anastomosing so as to form an irre- gular reticulation ; lateral protuberance rounded and not very prominent; centre of the valves marked by a deep and wide transverse furrow. Length 35 inch (0-80 millim.). One imperfect specimen only was found, in the Sables moyens (zone 4 Bryozoaires) ; I have therefore been unable to give figures, and have drawn up the description from Scottish Post-tertiary specimens. B. constricta occurs in the living state on the coasts of Norway and the British Islands, usually in depths exceeding 10 fathoms. Genus CYTHERIDEIS, Jones. Carapace elongated, subovate, depressed in front; hinge-margins nearly simple; shell smooth, punctate, or sometimes grooved ; right valve overlapping the left in the centre of the ventral surface. Superior antenne slender, sparingly setose; last joint short and bearing six short terminal sete. Mandible slender and curved, with about four very small indistinct teeth: palp 4-jointed, its first joint bearing a conical tooth-like process; third joint bearing a comb-like series of straight, equal sete, in other respects like that of Cythere. First segment of the maxille much stouter than the rest. CYTHERIDEIS (?) LITHODOMOIDES (Bosquet). (Plate LXIII. figs. 2 a2 d.) Bairdia lithodomoides, Bosquet, Entom. foss. terr. Tertiair. France, &c., p. 36, pl. ii. figs. 3a-d. Carapace, as seen from the side, oblong ovate, compressed, greatest height situated behind the middle and equal to scarcely half the length, depressed and rounded off in front, wider behind; dorsal margin gently arched, ventral straight or slightly sinuated. Seen from above, regularly ovate, compressed, widest behind the middle, pointed in front, rounded behind, width equal to the height. End view subcircular. Surface smooth, marked on the anterior portion of the shell with a number of shallow concentric grooves. Length 35 inch (0°98 millim.). 312 406 DR. G. 8. BRADY ON THE OSTRACODA This is one of the commoner species in the Antwerp beds, occurring in most of the deposits which we have examined, and usually in considerable quantity. M. Bosquet obtained it from several of the Eocene and Miocene deposits of France, and states also that he had seen a living specimen from the coast of Holland. In all probability, however, this must have been an example of Cytheridea elongata, Brady, which bears a close resemblance to the present species, but is, so far as I can judge, distinct. I have, however, considerable doubt as to the genus to which C. lithodomoides should be referred. CYTHERIDEIS RECTA, noy. sp. (Plate LXIII. figs. 3 a—3 d.) Carapace compressed, oblong, somewhat like a grain of rice in shape; seen from the side, linear-ovate, depressed in front, of nearly equal height throughout, height equal to rather more than one third of the length, obtusely pointed in front, rounded behind, dorsal and ventral margins both straight or very slightly convex. Seen from above, narrow ovate, extremities nearly equally pointed, width not more than one third of the length. End view nearly circular. Shell-surface perfectly smooth. Length 5 inch (0:90 millim.). Occurs very rarely in the zone 4 Bryozoaires. Genus Parapoxstoma, Fischer. Shell thin and fragile, smooth, shining; valves subequal, mostly much higher behind than in front, usually elongato-ovate; hinge-joint simple; ventral margins notched in front, so that when the valves are closed there is still an elongated orifice through which the suctorial mouth can be protruded. Upper antennz very slender, 6-jointed ; lower more robust, 5-jointed. Flagellum large and almost as thick as the antenna itself. Mouth suctorial; labrum and labium forming together a large process, projecting downwards and ending in a disk in the middle of which is the orifice of the mouth ; mandibles very slender, styliform; palp slender, indistinctly jointed and without branchial appendage. ‘Terminal lobes of first maxilla very narrow. Branchial plate elongated, and having two deflexed sete at the base. Feet short and robust, claws very short and curved. One eye. PARADOXOSTOMA ENSIFORME, Brady. (Plate LXIV. fig. 2.) Paradoxostoma ensiforme, Brady, Monogr. Recent Brit. Ostrac. p. 460, pl. xxxv. figs. 8-11; Brady, Crosskey, and Robertson, Monog. Post-tert. Entom. Scotland, &c., p. 215, pl. x. figs. 27, 28. Valves, as seen from the side, somewhat scimitar-shaped, much higher behind than in front, greatest height scarcely equal to half the length and situated behind the middle ; anterior extremity obtusely pointed; posterior obliquely rounded, slightly produced above the middle; superior margin boldly arched, inferior gently sinuated in the middle. Seen from above, compressed, ovate, with pointed extremities. Shell smooth. Length 3'; inch (0°75 millim. ). OF THE ANTWERP CRAG. 407 P. ensiforme occurs living in the British Seas and in the Levant, and has been found in many of the Post-tertiary deposits of England, Scotland, Wales, and Ireland. One or two valves only occurred, in the Sables moyens (zone a Bryozoaires). Fam. CYTHERELLID. Genus CYTHERELLA, Bosquet. Valves elongated, flattened, thick and hard, very unequal; the right much larger than the left, and overlapping throughout the whole circumference, presenting round the entire inner margin a distinct groove, into which the valve of the opposite side is received. CYTHERELLA PARALLELA (Reuss). (Plate LXII. figs. 2 @-2c.) Cytherina parallela, Reuss, ‘‘ Foram. u. Entom. Kreidemergels v. Lemberg,’ Haidinger’s Abhandl. vol. iv. p. 47, pl. vi. fig. 1 (1850). ? Cytherella pulchra, Brady, Trans. Zool. Soc. Lond. vol. v. p. 361, pl. lvii. figs. 1 a-d (1865). Carapace, as seen from the side, oblong elliptical, nearly equal in height throughout, height equal to about half the length; extremities rounded; dorsal margin nearly straight, ventral sinuated in the middle. Seen from above, compressed, ovate, width equal to one third of the length. End view ovate. Surface perfectly smooth. Length gig inch (0°85 millim.). This species occurred sparingly in both beds of the Sables inférieurs and in the zone a Bryozoaires of the Sables moyens. : CYTHERELLA ELLIPTICA, nov. sp. (Plate LXII. figs. 6a, 6.) Valves, as seen from the side, elliptical, width equal to two thirds of the length; extremities rounded ; dorsal margin nearly straight, ventral convex. Seen from above, ovate, widest behind. Surface smooth, with distant round impressed puncta. Length 3g inch (0°85 millim.). One valve only of this species was found, in the Panopwa-bed of Kiel. It seems to be distinct, but very nearly approaches C. beyrichi as figured by Speyer and Bornemann. It is not, however, the C. beyrichi of other authors, a form which is in all probability identical with C. abyssorum of G. O. Sars, and is so considered by that author. CYTHERELLA NoDoSsA, nov. sp. (Plate LXII. figs. 5 a—-5 d.) ? Cythere varians, Bornemann, Zeitschr. d. deutsch. geol. Ges. 1855, p. 365, pl. xxi. figs. 4, 5. Carapace, as seen from the side, elliptical, higher in front than behind, height equal to two thirds of the length; extremities rounded; dorsal margin well arched, ventral nearly straight; the anterior margin raised into a distinct rounded lip; and near the 408 DR. G. S. BRADY ON THE OSTRACODA posterior extremity of each valve are situated two large, polished bead-like tubercles— one near the dorsal, the other near the ventral surface. Seen from above, the outline is somewhat boat-shaped, subtruncate behind, and with wide produced keel in front; greatest width situated in the middle, and equal to less than half the length. End view ovate. Surface of the shell closely pitted with circular puncta. Length 345 inch (0°65 millim.). C. nodosa occurred in moderate numbers in the Pectunculus-bed, and also in the Panopea-beds of Kiel and Edeghem. In general character it is very nearly allied to C. leioptycha, Reuss, but differs in minor details, being more tumid, less angular in its contours, and wanting in longitudinal ribs. DESCRIPTION OF THE PLATES. The letters a, , c, d in these illustrations refer to the various aspects of the shell, as follows :— a, seen laterally ; 4, seen from above; c, seen from below; d, seen from the front. PLATE LXII. Figs. 1 a-ld. Cytheridea papillosa, var. Figs. 4a, 46. Cytheridea miilleri?. levis. Figs. 4c-4e. Cytheridea miilleri ¢. Figs. 2 a-2 ¢. Cytherella parallela. Figs. 5 a-5 d. Cytherella nodosa. Figs. 3a-3 d. Cytheridea pinguis. Figs. 6 a, 66. Cytherella elliptica. PLATE LXIII. Figs. 1 a-l1d. Cythere tarentina. Figs. 5a-5 d. Paracypris polita. Figs. 2 a—2 d. Cytherideis (?) lithodomoides. Figs. 6a, 66. Pontocypris faba °. Figs. 3a-3 d. Cytherideis recta. Figs. 6 c-6 e. Pontocypris faba é. Figs. 4 a—-4.c. Pontocypris propinqua. Figs. 7 a-7 c. Bairdia oviformis. PLATE LXIV. : Figs. la-ld. Cythere edichilus. Figs. 6 a, 66. Cythere plicatula. Fig. 2. Paradoxostoma ensiforme. Figs. 7 a-7 d. Cythere wetherellii. Figs. 3a—3.d. Cythere cicatricosa. Figs. 8 a—8 d. Cythere latimarginata. Figs. 4a,4b. Cythere cribrosa? Figs. 9a, 9b. Cythere limicola. Figs. 5a—d d. Cythere petrosa. PLATE LXV. Figs. 1 a-l1d. Cythere ellipsoidea. ‘Figs. 4a, 46. Cythere woodiana. Figs. 2 a-2h. Cythere jurinei. Figs. 5 a-5 d. Cythere plicata. Figs. 3a, 36. Cythere belgica. Figs. Figs. Figs. Figs. Figs. Figs. Figs. Figs. Figs. Figs. Figs. Figs. Figs. Figs. 1 a—1d. 2a, 20. 3a, 36. 4a-4d. 5a-d d. la-ld. 2 a-2 d. la-ld. le-lLh. Loxoconcha latissima ¢. 2a-2d. la—ld. 2 a2 d. d=) 6. 4a-4d. OF THE ANTWERP CRAG. 409 PLATE LXVI. Cythere polytrema. Figs. 6 a-6 d. Cythere macropora Cythere scabropapulosa. (young). Cythere dawsoni ? Figs. 7 a—7 d. Cytherura cornuta. Cythere trapezia. Figs. 8 a—8d. Xestoleberis depressa. Cythere acuticosta. PLATE LXVII. Cythere macropora. Figs. 3 a-3 d. Cythere mucronata. Cythere jonesit. Figs. 4a—4d. Cythere subcoronata. PLATE LXVIII. Loxoconcha latissima ° . . Figs. 3a—39. Lowoconcha grateloupiana : 3a—dd, young male? de, adult Loxoconcha bitruncata. female; 37,3, adult male. Figs. 4a-4d. Loxoconcha variolata. PLATE LXIX. Cytheropteron latissimum. Figs. 5 a-5 d. Cytherura broeckiana &. Cytheropteron pipistrella. Fig. 5¢. Cytherura broeckiana ¢. Cytheropteron intermedium. Figs. 6 a—6 e. Cytheridea cypridioides. Cytheropteron gradatum. ue. SG OF W, DL. OY fy Grp ror Tra 2d. L000 UE Ope 70 PYIXNT GSB.del. WPurkiss lith. M&N Hanhart. imp FOSSIL OSTRACODA = = t t- ‘ s ~ ae > GSB del. WPorkiss ith. M&N Hanhart map POS Sly Cis wmReACOiwA S GSB.del. WPurlass hth y Of fF l, CILLA. oot re C4 FOSSIL OSTRACODA. i] “4 O C M&H Hanhart imp He V ve L. Bi Vi F Ge G OFS ICN ' ' 1 ' ' ! i Ww) M&N Hanhart imp GSB. del WParkiss hth. FOSSIL OSTRACODA M&N Hanhart imp GSB. del. WPurkiss hth. FOSSIL OSTRACODA. ~ a ; = ’ t = i z % wel (re 3 t ie ‘ *, } 4 + ' ~ E's ~ r . el E Ae = ‘ é by i ry, oe ‘ da ~ ce ‘ = ' i . ‘ cap. ‘a - ~ 4 ‘ * ni . lrand Loot, Soe Vet 10 FL LEVT. od yA x \\ y/ \ Dd et SONS L JS NN KN AS d Vv g = }) ma | {~ 1114 { Ty) IW OS i» | \ Yoel) © EEE in) oe lp Fy iin © ; CN CaM a MM yi "x@senaad SIs M& N Hanhart ump GSB del, WPurlass hth. FOSSIL OSTRACODA Va SYD 2 af Weer Yy SVHTMEA. L004 SH od aa 2b Se ad 3° 1 5h 38 42 4e 4° 4° \ \ } \ GSB del. WFurkiss hth M& N Hanhart imp FOSSIL OSTRACODA. eh, ' } t fl = F — \ x uy \ a —_ a) SS . ee ‘ -* _ 74 > wf if ‘ C “> a a . ¢ k ‘ \ ‘ P ‘ ; f “~~ ane > 7 al - ~ of * ie 9 (¥ WA AS FCUSTEE mL GSB.del WPurkiss lth FOSSIL OSTRACODA. vi -OC 5e M&NHanhart imp oe . é ms A= x = re re be p 4 : 2 ~~ : tira ‘ ' Pa > » x - eyo J naan gi X. On the Brain of the Sumatran Rhinoceros (Ceratorhinus sumatrensis). By A. H. Garrop, W.A., F.R.S., Prosector to the Society. (Received June 19th, 1877.) [Puate LXX.] IN a communication to this Society, published in its ‘Proceedings’ in 1873 (p. 92), I had the opportunity of describing the visceral anatomy of the Sumatran Rhinoceros (Ceratorhinus swmatrensis) from the first specimen received by the Society. A second individual of the species, a female (as was the first), was deposited in the Gardens by Mr. C. Jamrach in July 1875, and was subsequently purchased. It unfortunately died on May 30th of this year, with symptoms of lung disease, a post-mortem examination demonstrating that both lungs were uniformly and throughout implicated. My friend Dr. James F. Goodhart, of Guy’s Hospital, late Pathological Registrar at the College of Surgeons, has kindly examined these organs, and reports to me that they “show a very extensive catarrhal pneumonia, degenerating in the centres of most of the patches. There is, in addition, some peribronchial inflammation, evidenced by a large growth of nuclei in the submucous and deeper tissues of the bronchi. The disease therefore precisely corresponds with the caseous pneumonia to which man is subject.” The specimen is the one referred to by Mr. Sclater in his valuable and superbly illustrated memoir in the Society’s ‘ Transactions,’ vol. ix. p. 651 (foot-note 3), Feeling how important it is to obtain all possible information with reference to the species, and not having removed the brain in the earlier specimen, I took the oppor- tunity of doing so in the second, and on the present occasion place before the Society the drawings of the brain from different aspects (Plate LXX.), for verification of which I would refer the reader to the Museum of the College of Surgeons, where the original will be found preserved and mounted. The brain of the Indian Rhinoceros (Rhinoceros unicornis) is represented in its different aspects, and in its internal detail, by Professor Owen, in the ‘ Transactions’ of this Society, volume iv. pls. 19-22, and is described shortly on page 58 et seg. of the same volume. To this it is my desire that the figures here given should form a companion. By comparison it will be seen at a glance that the brain of Rhinoceros unicornis is slightly more simple than that of Ceratorhinus sumatrensis, although the greater size of the former species would have favoured an opposite conclusion. So complicated and numerous are the convolutions that the general type-plan of their disposition is to a considerable extent disguised. They very closely resemble the same VOL. X.—PaRT 1X. No. 1.— August 1st, 1878. 3K 412 MR. A. H. GARROD ON THE BRAIN OF THE SUMATRAN RHINOCEROS, in the Equide, as might have been surmised. The whole brain, however, is broader, especially near the posterior portion of the cerebral hemispheres, where the breadth is considerably greater than further forward. The accompanying diagram will facilitate the description. It represents the disposition of the main convolutions upon the superior aspect of one hemisphere, and exhibits the direction of the sulci which divide them. ‘Two diagonal sulci cut up the posterior part of each lobe into three oblique gyri, which may be called the (1) external, (2) middle, and (8) internal gyrus. The middle and internal of these fuse together near the transverse line which joins the two rudimentary Sylvian fissures, anteriorly to which there is, in the Equide, no indication of further primary longitudinal division. The external oblique gyrus continues, from this line, directly forwards, and independent. In Ceratorhinus sumatrensis the internal oblique gyrus Upper view of left cerebral hemisphere of Ceratorhinus sumatrensis, showing general direction of sulci. is triangular in shape, its inner boundary being the great longitudinal fissure of the hemispheres, into which it de- scends a short distance. In the Equide the inner boundary of this gyrus is more superficial, and can be seen as a straight longitudinal line, just external to the fissure itself, in the superior view of the brain. The whole gyrus is much broken up by minor foldings of its elements, especially in its median portion, its outer moiety consisting of a minor gyrus, whose general direction is a continuous oblique line, fairly regularly bent upon itself, first one way and then the reverse. The median oblique gyrus is divided into two nearly equal moieties by a fissure running parallel to its direction, each half being much doubled upon itself. Anteriorly bridging minor convolutions blend it with the internal oblique gyrus, about one third distant from the anterior extremity of the hemisphere, in front of which the broad oblong cerebral surface is divided by a longitudinal sulcus into two equal moieties, both convoluted. In the great breadth and division of this anterior portion the Rhinoceros under consideration differs from the Equide, and agrees with Rhinoceros unicornis. The external oblique gyrus is much doubled on itself, and separated from the Sylvian fissire, which it surrounds, by minor conyolutions, more strongly differentiated anteriorly. On the inner surface of the hemisphere the hippocampal gyrus is seen to be traversed by minor sulci and slight folds which run parallel to its length, as in the Equide, the calloso-marginal sulcus following the anterior bending of the corpus callosum, and not, as in so many Artiodactyla (but not in the Kquide), becoming superficial anteriorly. MR. A. H. GARROD ON THE BRAIN OF THE SUMATRAN RHINOCEROS. 415 The fissure of Sylvius forms an open angle, at the bottom of which are situated a number of small convolutions radiating from a point, which I take to be the island of Reil. The under surface of the brain exhibits the smooth surfaces of the middle lobes of the hemispheres and the smooth broad roots of the equally broad olfactory nerves, which are not lobate at their anterior extremities. The optic chiasma is short, the two optic nerves springing from its anterior surface quite close together. The pons Varolii is not large, the reverse being the case with the crura cerebri and the corpora albicantia. The lateral lobes of the cerebellum are small compared with the median portion, as is the case in the Ungulata generally. DESCRIPTION OF PLATE LXX. Brain of Ceratorhinus sumatrensis. Fig. 1. Lateral view of right cerebral hemisphere. 2. Inferior view of left half of brain. 3. Superior view of left half of brain. 4. Internal view of right cerebral hemisphere. J Smit hth Hanhart imp BRAIN OF CERATORHINUS SUMATRENSIS, 9 ‘ “ R J f ‘ ' d s ‘ . 1 = “i ‘ a ' eed [ 415 J XI. A further Contribution to the Knowledge of the existing Ziphioid Whales. Genus Mesoplodon. By Witiiam Henry Fiower, F.R.S., V.P.Z.S. Received July 31st, 1877. Read November 6th, 1877. | Puates LX XIJ.-LXXIII.] DURING the six years that have elapsed since I communicated to the Society a memoir on the recent Ziphioid Whales (Trans. Zool. Soc. vol. viii. p. 203) very consi- derable additions have been made to our knowledge of the group. Instead of being so rare as was then supposed, since the attention of naturalists resident in our colonies has been directed to the importance of losing no opportunity of securing such specimens as accidents of wind and waves may cast upon their shores, it has been proved that in the seas of the southern hemisphere these Whales exist in considerable numbers both as species and as individuals, and that one species at least is gregarious, having been met with in two instances in “ schools’’ of considerable numbers. On the other hand, no remarkable deviations from the forms already known have been met with; and all additional information as to their osteology has fully confirmed the value of the division of the Ziphioid Whales into four distinct types or genera, at that time indicated. With perhaps one exception (to be noticed further on), which presents some signs of transition, all the known individuals lately discovered can be arranged without any hesitation either as Hyperoodon, Berardius, Ziphius, or Meso- plodon, as defined in the previous memoir. It is in our knowledge of the animals of the last-named genus that the greatest advances have been made of late ; and it is to these that the present communication will chiefly relate. After examining all the available specimens and published descriptions, I have arrived at the conclusion that evidence exists at present of six distinct specific modifica- tions of this form, of which four inhabit the southern temperate seas. To these I shall have to add two more, although in neither case on evidence so satisfactory as might be desired. The synonymy and habitats of the hitherto known species are as follows :— 1. M. sipens (Sowerby). Physeter bidens, Sowerby, Brit. Miscellany, p. 1 (1804). Delphinus (Heterodon) sowerbiensis, Blainville, Nouv. Dict. d’Hist. Nat. 2nd ed. tome ix. p. 177 (1817). D. sowerbyi, Desmarest, Mammalogie, p. 521 (1822). Delphinorhynchus micropterus, Cuvier, Régne Animal, t. i. p. 288 (1829) ; Dumortier, Mém. Acad. Roy. Bruxelles, t. xii. (1839). VoL. X.—PART 1X. No. 2.—August 1st, 1878. Bit 416 PROF. W. H. FLOWER ON THE GENUS MESOPLODON. Mesoplodon sowerbiensis, Van Beneden, Mém. Acad. Belgique, coll. in 8vo, t. xvi. (1863). M. sowerbensis, Gervais and Van Beneden, Ostéographie des Cétacés, p. 392 (1877). Micropteron bidens, Malm, “ Hvaldjur i Sveriges Museer ar 1869,” K. Svenska Vetenskaps-Akad. Handl. 1871. Hab. Coast of Elginshire (Sowerby) ; Havre (Blainville); Ostend (Dumortier); Sale- nelles, Calvados (E. Deslongchamps); Brandon Bay, Ireland (Andrews); Norway (Malm). 2. M. Evropaus (Gervais). Dioplodon europeus, Gervais, Zool. et Pal. frangaises, 1"¢ édit. t. ii. Explic. No. 40 (1850). D. gervaisii, E. Deslongchamps, Bull. de la Société Linnéenne de Normandie, tome x. (1866). A single specimen found floating in the sea at the entrance of the British Channel about 1840. The skull is now in the Museum at Caen. It should be mentioned that some doubts have been thrown upon the supposed distinctive characters between this and the last species. 3. M. DENsrRostTRIS (Blainville). Delphinus densirostris, Bainville, Nouv. Dict. d’Histoire Nat. 2nd edit. tome ix. p. 178(1817). Mesodiodon densirostris, Duvernoy, Annales des Sciences Naturelles, sér. 3, t.xv. p. 59 (1851). Ziphius sechellensis, Gray, Zool. H. & T. p. 28 (1846) ; Krefft, P. Z. S. 1870, p. 426. Dioplodon densirostris, Gervais, Zool. et Pal. francaises, 1" édit. t. ii. Explic. No. 40 (1850) ; Ann. des Sciences Nat. sér. 3, t. xiv. p. 16 (1850). Hab. Seychelles Islands (Duvernoy); South Africa (Mus. R. Coll. Surgeons); Lord Howe’s Island (Arefft). 4, M. LayarDI (Gray). Ziphius layardi, Gray, P. Z. S. 1865, p. 358; Owen, Crag Cetacea, p. 12, pl. i. (1870). Ziphius (Dolichodon) layardi, Gray, Cat. Seals and Whales Brit. Mus. p. 353 (1866). Dolichodon layardii, Hector, Trans. New-Zealand Inst. vol. v. p. 166, pl. iii. (1872). Mesoplodon longirostris, Krefit, MS., M. giintheri, Krefit, MS., and Callidon giintheri, Gray, Ann. Mag. Nat. Hist. ser. 4, vii. p. 368 (1871). Dolichodon traversii, Gray, Trans. New-Zealand Inst. vol. vi. p. 96 (1874). Mesoplodon floweri, Haast, P. Z.S. 1876, p. 478. Hab. Cape of Good Hope (Layard) ; Near Sydney (Krefft) ; New Zealand (Haast) ; Chatham Islands (Hector). Two skulls, an adult and a young specimen, of this well-established species were brought home by the naturalists of the ‘ Challenger,’ and will be described in detail by Professor Turner in his Report on the Marine Mammalia collected during the expedition. 5. M. HECTORI (Gray). Berardius arnuxii, Hector, Trans. New-Zeal. Inst. vol. ii. p. 27 (1870). Smaller Ziphioid Whale, Knox & Hector, T. New-Zeal. Inst. vol. iii. p. 125, pls. xiii., xiv. & xv. (1871). PROF. W. H. FLOWER ON THE GENUS MESOPLODON. AIT Berardius hectori, Gray, Annals & Mag. N. H. ser. 4, vol. viii. p. 117 (1871). Mesoplodon knoxi, Hector, Trans. New-Zeal. Inst. vol. v. p. 167 (1873). Hab. Titai Bay, New Zealand. Very young; length 9! 3", 1866 (Anox and Hector). 6. M. erat, Haast, Proc. Zool. Soc. 1876, p. 7. Oulodon grayi, Haast, P. Z. 8. 1876, p. 457". Hab. New Zealand. These six species do not differ greatly in size, and present a close family or, rather, generic resemblance ; but still there are certain cranial and dental characters by which the skulls can readily be distinguished. The most easily recdgnized of these are :— a. In some species there is a deep lateral longitudinal groove at the base of the rostrum, commencing posteriorly in a blind pit below the tubercle of the maxilla situated in front of the anteorbital notch, and bounded above and below by sharply defined prominent ridges, both formed by the maxilla. This groove characterizes M. grayi and M. densirostris. No trace of this groove, but a prominent ridge instead, which gives a wide base to the rostrum, as seen from above or below, exists in M. bidens, M. europeus, and M. hectori. The groove is slightly developed in MW. layardi, which in this respect presents an intermediate condition. b. The relative position of the foramina for the exit of the facial branches of the second division of the fifth nerve appears to afford constant distinctive characters between certain of the species. The principal foramina are in two pairs:—l, the larger or more external, situated in the maxilla, and often double, corresponding to the infraorbital foramen in man (Plate LXXI. mf’); and, 2, the smaller and nearer the middle line, situated in the premaxilla (pf). The latter are rather in front of, or on the same level with, the maxillary foramina in J. bidens, M. europeus, M. hectori, and M. layardi, and placed decidedly on a posterior level in M/. densirostris and M. grayi. c. The constant presence of a row of small teeth in the upper jaw is said to dis- tinguish M7. grayi from the other species, though, as will be discussed presently, this can scarcely be considered of generic importance. d. The position of the large tooth in the lower jaw, whether close to the apex of the ' Tn the notice of this paper in the Proceedings of the Zoological Society, 1877, p. 684, for “7, M. haasti,n. sp. ...... present memoir,” read :— 7. M. haasti, n. sp. Hab. New Zealand. Known only by a portion of a cranium in the Museum of the Royal College of Surgeons. 8. M. australis, n. sp. = M. hectori (Gray). Hector, Trans. New-Zealand Inst. vol. vi. p. 86, and vol. vii. p. 362. Hab. New Zealand. A complete skeleton, now in the British Museum, described in the present memoir. 3L 2 418 PROF. W. H. FLOWER ON THE GENUS MESOPLODON, mandible, as in JV. hectori, or near the hinder part of the symphysis, as in all the others, forms another good distinctive character. e. Finally, the character and form of this tooth presents some important variations. In UM. layardi, its apex is everted, and seated upon a flat strap-like base, which grows upwards, backwards, and finally inwards, closing over the upper jaw and meeting its fellow of the opposite side. In MV. densirostris the apex is directed vertically, placed on a very massive base, which is implanted in a greatly expanded alveolar margin of the jaw, not found in any of the other species. In WM. grayi the tooth is triangular and compressed, and the apex is vertical at all ages. In MV. hectori the apex of the tooth in the young animal is directed forwards. In MW. bidens the apex of the tooth is at first directed backwards, but during growth becomes rotated forwards. In the only specimen of M. ewropeus known (an adult) the apex of the tooth is directed somewhat forwards. These characters may be thus placed in a synoptical form :— Lateral Premaxillary ae basirostral foramen to Position of Other dental characters. groove. maxillary. mandibular tooth. Absent. Level. (OLE Ite), Gree ||" Ganoos snaaudoonbon babe | M. hectori. mandible. Absent. Level. Api ONG || Goontoomepobuccoconeso M. europeus. dibular sym- physis. Absent. Level. Near hinder edge | Apex of tooth at first di- | MW. bidens. of mandibular rected backwards, and symphysis. during growth rotated forwards. Slight. Leyel. Ditto. Apex of tooth everted and | M. layardi. mounted on a flattened base, which increases in a length with age, and finally curves round the maxilla. Deep. Behind. Ditto. Apex of tooth vertical. Al- | MW. densirostris. | veolar portion of jaw expanded. Deep. Behind. Ditto. Apex of tooth vertical. A | M. grayi. row of small teeth in upper jaw. Besides the above, there are many minor differences, more appreciable by a com- parison of the specimens than by description, some of which will be pointed out in the sequel. The special materials upon which the present contmunication is based are :— 1. The skull of a young specimen of WV. hectori, the type and at present only known example of that species. It was taken in Titai Bay, Cook’s Strait, January 1866 ; and the skull is figured by Dr. Hector in the Trans. New-Zeal. Inst. vol. iii. 1870, pls. xiv. PROF. W. H. FLOWER ON THE GENUS MESOPLODON. 419 & xv. The length of the animal is stated by Dr. Knox to have been 9 feet 3 inches. The skull is now in the British Museum. 2. The skeleton of a perfectly adult animal, attributed by Dr. Hector to the same species, prepared from a specimen cast ashore in Lyall Bay, and described by Dr. Hector in Trans. New-Zeal. Inst. vol. vii. p. 262, 1875. This is also in the British Museum !. 3. The skeleton of an adolescent male MW. grayi, stranded, with others, on the coast near Saltwater Creek, about thirty miles north of Banks’s Peninsula, New Zealand. 4. The rostrum, mandible, and mandibular teeth of an old male animal, stranded in December 1875 on the east coast of the North Island, attributed by Dr. v. Haast to the same species. The last two were presented to the Museum of the Royal College of Surgeons by Dr. Julius v. Haast, F.R.S., of Christchurch, New Zealand. Of the specific distinction or identity of the 2nd, 3rd, and 4th I am not perfectly satisfied, as the materials at hand are unfortunately not sufficiently complete for the purpose of arriving at a definite conclusion. No. 3, though a perfect skeleton, is stillimmature. No. 2, though adult, is incom- plete as regards the very important evidence furnished by the dentition. No. 4 is only a fragment, and, although identified by Dr. v. Haast with his species WM. gray, presents strongly marked differential characters which seem beyond the range of individual variation. Under the circumstances, it is somewhat difficult to know what course to pursue with reference to the names by which these specimens are to be respectively distinguished; but on the whole it will lead to less confusion if I designate them, provisionally at least, by specific appellations, bearing in mind that it is quite possible that further information and more abundant materials may cause a modi- fication of this view. I shall therefore, in the present memoir, speak of the British- Museum skeleton from Lyall Bay as WZ. australis, and the fragmentary skull from the east coast of the North Island as WZ. haasti. Skull.—The skull of animals of the genus Mesoplodon is very easily distinguished from those of the other Ziphioids by the characters previously given (T. Z. S. vol. viii. p- 208). The ossification of the mesethmoid cartilage, and its coalescence with the surrounding bones to form a solid rostrum, appear to be greatly dependent upon the influence of age. It has not hitherto been found wanting in any thoroughly adult example of any species of Mesoplodon or of Ziphius; on the other hand, it appears never to occur either in Hyperoodon or Berardius. ‘ T am indebted to Dr. Giinther’s kindness for facilities in examining and comparing these specimens, which were sent to the British Museum by Dr. Hector, F.R.S., of Wellington, New Zealand. * Almost simultaneously with my memoir appeared an excellent description of the skull of WM. sowerbyi, com- pared with that of Ziphius cavirostris, by Professor Turner (Transactions of the Royal Society of Edinburgh, vol. xxyvi.). 420 PROF. W. H. FLOWER ON THE GENUS MESOPLODON. The tympanic bone of Mesoplodon is quite different from that of Ziphius', the groove between the lobes at the posterior end being very well marked as in the true Dolphins, whereas in Ziphius it is obsolete. In Hyperoodon also it is scarcely apparent, while in Berardius it is nearly as well marked as in Mesoplodon. Thus, by the form of this very characteristic bone, Ziphius allies itself to Hyperoodon, and Mesoplodon to Berardius, and the two former approach nearest to Physeter, and the latter to the less-modified Dolphins. The relationship between Berardius and Mesoplodon is undoubtedly close. One of the specimens (J/. hector’) now under consideration has been placed sometimes in one genus and sometimes in the other ; and it does certainly present some transitional cha- racters ; but as it is only known by the skull of a very young individual, it is scarcely safe to decide its position, except provisionally, especially as it is not yet known in which direction the alterations which must take place during the progress towards maturity tend. The animal stranded at Lyall Bay, which I have called M. australis, is perfectly adult, as shown not only by the union of the sutures at the base of the cranium, but also by the condition of the other bones of the skeleton, the terminal epiphyses of the bodies, even of the thoracic vertebre, being completely consolidated. The mesorostral or mesethmoid cartilage is densely ossified throughout its whole length. There is still much to be learned with regard to the mode of ossification of this cartilage. All the specimens which I have had an opportunity of examining are either so young that ossification has not commenced, and the trough of the vomer in the rostrum proper is completely empty in the dried skull, or so old that the consolidation of the cartilage and its union with the surrounding bone has been completed. But it must be observed that, although the cartilage appears to be nothing more than a con- tinuation forwards of the ordinary mesethmoid lamina or septum of the nose, the ossification is not a simple extension forwards of that which occurs in all Cetacea (in all Mammalia, in fact) in the hinder or internarial portion of the septum, but appears to be an independent production, peculiar to the genera Mesoplodon, Ziphius, and certain allied extinct forms. It is separated by an interval (which appears to diminish with age, but of which traces can always be seen on the upper surface of the rostrum near its base) from the true mesethmoid ossification. It differs from the latter in being intensely hard and compact, whereas the mesethmoid is, especially at its anterior part, somewhat spongy in texture. It differs also in showing strong indications of being formed by a pair of lateral ossifications, united in the middle line, as the upper surface in many parts and the anterior apex show a marked median groove. I think it will be well therefore to adopt Professor Turner’s name of “ mesorostral” bone for this solid bar forming the centre of the rostrum, restricting “ mesethmoid” to the part lying between the nares and a short distance in front of them, which is ossified in the * Figured by Prof. Turner, loc. cit. plate xxx. PROF. W. H. FLOWER ON THE GENUS MESOPLODON. 421 young animal and in all other species of Cetavea. The median ossicle in the snout of the Pig offers a similar case of an independent ossification in the anterior extremity of the mesethmoid cartilage. The skull, as seen from above (Plate LX XI. fig. 1), presents several characters by which it can be readily distinguished from MV. didens and the closely allied WM. ewropeus. The principal of these is the sudden narrowing of its outline at the commencement of the rostrum, due to the presence of the deep lateral groove mentioned above. In the British species the rostrum is wider at the base, and passes more insensibly into the cranium. The next point is the relative position of the preemaxillary ( pf’) and the maxil- lary (mf) infraorbital foramina—the former being behind the latter in the southern, and on the same level in the northern species. The third is the form of the premaxille, which are more expanded and flattened out (especially the right) immediately in front of the narial apertures in MV. australis. In the conformation of the upper ends of the premaxill and of the nasal bones the differences are but slight. M. layardi, as mentioned above, agrees with the northern species in the disposition of the infraorbital foramina, and the absence of any marked lateral basirostral groove ; and by these as well as by its very characteristic dental characters it is distinguished from the one now under consideration. MV. densirostris is much more closely allied to our new species, having (judging by the figure in Van Beneden and Gervyais’s Ostéo- graphie, pl. xxv.) a very similar outline as seen from above. It differs, however, independently of the mandibular and dental characters, in the far more massive rostrum, which, although no longer, has exactly double the vertical height at the middle, as seen in the profile view of the skull. A marked difference is also seen on the lower surface of the skull. In MM. densi- rostris the palate-bones completely surround the anterior pointed ends of the pterygoids, widely separating the latter from the maxillaries ; while in MW. australis the palatines on the inferior surface of the skull lie altogether on the outer side of the pterygoids, and do not even extend so far forwards as the pointed extremities of those bones, which thus by their free end and inner side come into direct contact with the maxillaries. It remains only to compare this skull with the other New-Zealand forms—the typical IZ. grayi of Haast, the rostrum which I propose to call I. haasti, and MW. hectori. With the former it has very near affinities—so much so, that from an examination of the cranium alone I should scarcely think of placing them in different species. It is needless to dwell upon differential characters which may be merely individual ; but I may indicate the following as perhaps of more importance. a. Size: MJ. australis would appear to be smaller than YZ. grayi. ‘The two complete skulls are practically of the same size; but that of the former is adult and that of the latter young, and the tympanic bone of the latter is slightly larger than that of the former. Dr. Haast gives 17! 6" as the length of an adult female of WZ. grayi, whereas the adult MV. australis (sex not 422 PROF. W. H. FLOWER ON THE GENUS MESOPLODON. stated) is given by Dr. Hector as 15! 6"; this difference, however, can scarcely be con- sidered beyond the limits of individual variation. 6. The form of the tympanic bone differs slightly, the inner side of the posterior edge (as seen from below) being more bevelled and the inner posterior lobe less prominent in W/. australis. c. The lateral groove at the base of the rostrum is very much deeper in MV. australis, running up- wards into the maxillary bone, and forming a blind pit at its upper termination, of the depth of nearly one inch, whereas in 1. grayi this pit is quite shallow. It is possible, however, that this may be a question of age. Fig. 1. Fig. 2. Mesoplodon australis (adult). Mesoplodon grayt (young), Fig. 3. Fig. 4. Mesoplodon haasti (adult). Mesoplodon heciori (young). Outlines of section of middle of rostrum. Natural size. While the section of the rostrum of both the above (figs. 1 & 2) agree very closely in general outline and size, that of the supposed W/. haasti (fig. 3) differs very greatly, being altogether more compressed and wanting the lateral wing-like ridges near the lower border ; its length is also considerably greater, being 224", while that of the adult M. australis is only 18". Making every allowance for individual variation, it scarcely seems possible that a rostrum such as that shown in fig. 2 could change in the course PROF. W. H. FLOWER ON THE GENUS MESOPLODON. 423 of growth to that in fig. 3. If so, most of the determinations of the fossil species based solely upon the form of the rostrum are quite valueless. It should be mentioned that in this fragmentary rostrum the lateral groove and pit _at its base are deeper even than in M. australis, and that it departs, therefore, in this character, still further from (the young) MW. grayi. The cranium of I. hectori, from Titai Bay, New Zealand, now in the British Museum, presents a conformation quite distinct from that of either of the specimens last under consideration. As is shown in the synoptical table at p. 418, it more nearly resembles the northern species in the relation of the maxillary foramina, the absence of the basirostral groove, and, it may be added, the correlated greater breadth of the base of the rostrum. The skull is evidently that of a very young animal: the teeth are still concealed beneath the gum; the symphysis of the lower jaw is not united; the meso- rostral cartilage is entirely unossified; the basisphenoidal suture is open (as seen from the interior of the cranium; for on the outer surface it is covered by the vomer) ; and the bones generally are very spongy and greasy. On the other hand, the elements of the occipital bone have all coalesced with each other and with the basisphenoid The principal difference between this cranium and that of I. didens and, in fact, of all others of the genus, is that the upper extremities of the premaxille are less developed laterally and less everted, and therefore approximate more to the form of Berardius ; but unfortunately a large piece of one of these bones is broken off from the right side, or that which is usually most strongly marked, and it is impossible to say how much of the absence of characteristic eversion may be due to the immaturity of the specimen. There is also a greater elevation of the longitudinal ridge on the upper surface of the maxilla, immediately to the inner side of the anteorbital notch, than in any other Mesoplodon. This ridge, it should be remarked, forms a strongly pronounced elevation in Berardius, and is developed in Hyperoodon into the immense bony mass which forms such a characteristic feature of the skull of that genus. The nasals, however, are com- pletely sunk between the ends of the premaxille, whereas in Berardius they form prominent masses rising to the vertex. The palate-bones only appear on the palatal surface as narrow strips on the outside of the pterygoids, not reaching as far forwards as they do. The vomer is visible in the middle third of the inferior surface of the rostrum. The mastoids and squamosals are quite free; and the former are largely developed, forming the principal part of the great post-tympanic processes of the skull. The tympanic is very slightly larger than that of WZ. australis, and therefore equal to that of WZ. grayi, and presents a very close resemblance to both, but is nearer in the form of the lobes at the posterior part to the former. I have not had an opportunity of comparing it with the tympanic of If. bidens. As the tympanic bone of Berardius is constructed on exactly the same type as that of Mesoplodon, this bone gives no assist- ance in determining the supposed relationship of IZ. hectori to Berardius. VoL. X.—ParT 1x. No. 3.—August 1st, 1878. 3M 424 PROF, W. H. FLOWER ON THE GENUS MESOPLODON. The principal dimensions of these crania are as follows :— M. australis, M, grayi, M. hectori, | adult. young. young. | inches. millim. inches, millim. inches. millim. Extreme length of cranium .............. 30°3 770 30°31 770* || 22:3 567 | Length of rostrum (from the apex of the | preemaxilla to the middle of a line drawn | | | between the anteorbital notches?) ......| 19-9 505 20:2 510 12°6 320 | From middle of hinder edge of palate (formed by the pterygoids) to apex of rostrum....| 23-8 605 24-5 | 622 || 17-4 442 Greatest height of cranium from yertex to | | PUCEY SOG es ce scsi eberseue eis ciecis ce osrereratete 11:0 278 104 | 264 || 95 241 Breadth of cranium across middle of superior WEAPON OA, Goheatoagegouyoob AGE 11-2 983 || 105 | 268 9-2 235 | Breadth of cranium between zygomatic pro- | ceases of squamosals .................. iile7/ 298 11-1 282 10°2 259 Breadth between anteorbital notches ...... 74 187 7:2 183 | 5:3 135 | Breadth of middle of rostrum ............ 17 44 Gay 20 15 37 | Breadth of occipital condyles ............ 3-9 99 37 | 95 3°6 92 Pramaxille, greatest width behind anterior [SPiriaresier meer tthe tek carne ht - ws | 88) Weeeo. Ge > Kes ese fe) WBOG00e APVPAUSTRALIS. [ 439 ] XII. Notes on the Fins of Elasmobranchs, with Considerations on the Nature and Homologues of Vertebrate Limbs. By Sv. Grorce Mivart, V.P.Z.S. Received December 22nd, 1877. Read February 5th, 1878. [Puates LXXIV.-LXXIX.]} IN the following paper I describe certain fin-structures which I have not found to be described elsewhere. Before proceeding to do so, however, I must express my grateful sense of the kindness of my friend Dr. Giinther in placing at my disposal for examination and illustration duplicate store-specimens of Elasmobranchs which I should otherwise have had no means of investigating. The species referred to are Zygena malleus, Mustelus antarc- ticus, Notidanus cinereus, Scyllium canicula, Ginglymostoma cirratum, Chiloscyllium ocellatum, Acanthias blainvillii, Spinax niger, Pristiophorus japonicus, Pristis cuspidata, Rhynchobatus djeddensis, Trygonorhina fasciata, and Callorhynchus antarcticus. Besides these, I have made use, for comparison, of certain skeletons preserved in the Museum of the Royal College of Surgeons, drawings of some of which have been made through the kind permission of my friend Professor Flower and the other authorities of that Institution. These are Lamna cornubica, Cestracion philippi, Squatina angelus, Polyodon folium, and Polypterus bichir. It may be well first to describe the specimens examined, and then to make such remarks as have suggested themselves in regard to Vertebrate limbs generally, and the relations borne by the fins of fishes to the extremities of higher animals. ZYGHNA MALLEUS. Dorsat Fin (Plate LX XIV. fig. 1). The dorsal fin in this species is sustained by a great number of very elongate and closely set cartilaginous rays, or, as they may be for distinction termed, “ radials.” which form three superimposed longitudinal series—one basal, one median, one distal. The number of these radials bears, as far as I can perceive, no exact relation to the number of subjacent vertebre ; and the whole fin-skeleton is separated from the sub- jacent axial skeleton by interposed fibrous membrane, the interval, however, being less wide than half the depth of the shortest of the three superimposed series of cartilages. The basal cartilages ave twenty-six in number ; but the second and the eighth from the preaxial end of the series bifurcate distally, and the last but one seems to be made VOL. X.—PaRT x. No. 1.—February 1st, 1879. 30 440 PROF. ST. GEORGE MIVART ON THE up of three cartilages fused together. After the first, they are of nearly equal length till about the seventeenth, whence they rapidly diminish in length postaxiad. The median cartilages are slightly larger than the basal ones, and are twenty-nine in number ; but one or two bifurcate distally. They are more equal in length than either of the other series, and are longest at the preaxial end, 7. ¢. at about the sixth radial. The distal cartilages are far the longest and most unequal (being shorter both pre- and postaxially) and far longer medianly than any of the median or basal cartilages. They are twenty-nine in number, the twelfth, thirteenth, and fourteenth being the longest. Each reposes proximally on the distal end of a median cartilage, except the last but six, which seems intercalated between the distal ends of the adjacent median cartilages. TuHE PEcToRAL FIN. The skeleton of this fin I found to closely resemble that of Carcharias glaucus, as represented by Professor Gegenbaur1, except that the metapterygium was more seg- mented and the propterygium somewhat larger and more prolonged. Tue VENTRAL Fin. The order much resembled that of C. glaucus, as shown by the last-named author ?, save that there was no thick preaxial radial ; but I found that four slender ones joined the _pelvic cartilage, while, instead of one stout radial ; three slender ones were appended to the distal end of the basal supporting cartilage of the ventral fin. LAMNA CORNUBICA. Dorsat Fin (Plate LX XIV. fig. 2). The skeleton of this fin exhibits much general resemblance in outline to the corresponding one of Zygwna malleus ; but the basal cartilages are very much elongated at the expense of the median series, which are found only in the postaxial half of the fin. The basal cartilages are twenty-four in number, and, after the first two, attain at once their greatest length, decreasing postaxiad after about the fifteenth. The five most postaxial radials are segmented ; but they may have been fractured. The median cartilages ave only twelve; and all are very short, and interposed between the more postaxial radials of the other two series. The distal cartilages are extremely elongated, and number twenty-three, the eleven more preaxial of which repose directly on the distal ends of the basal cartilages, the others resting on the short median ones. 1 Untersuchungen, Heft ii. plate ix. fig. 5. 4 « Die Gliedmassen der Wirbelthiere im Allgemeinen,” in Jenaische Zeitschrift, vol. xy. plate xv. fig. 9. FINS OF ELASMOBRANCHS. 441 They seem more segmented than in Zygena, and irregularly so; but this may be due to accidental fractures produced in preparing the dry specimen examined. Tue Cavupat Fin (Plate LX XV. fig. 1). The skeleton of this fin is remarkable for the extreme difference between its ventral and dorsal portions. The dorsal supports are very small and inconspicuous, while the ventral ones are large, and vary extremely in size, the ninth to the thirteenth being exceedingly broad as wellas long. Notwithstanding this change of form and proportion, the ventral-fin cartilages bear an almost exact numerical relation to the vertebra—one to each—save that in passing postaxiad from the tenth vertebra, one seems to disappear ; but it is possible this may be due to the drying and preparation of the specimen. MUSTELUS ANTARCTICUS. Dorsau Fin (Plate LXXTV. fig. 3). The cartilaginous radials in the dorsal fin of this species form a triple longitudinal series of unequal number, without bifurcation, and with hardly any coalescence. The whole cartilaginous fin-skeleton is separated from the subjacent axial cartilages by an interposed fibrous membrane, which at its narrowest part is a little wider (dorso-ventrally ) than the length of the longest of the basal series of cartilages. The basal cartilages are eighteen in number. They increase gradually in length from the first to the eighth and ninth, and then more rapidly diminish, the four last being shorter than the first. The median cartilages are twenty-four, and for the most part are articulated to the dorsal ends of the basal cartilages, one to each. But the first three and the last three median cartilages have no such support, and are only connected ventrally with the supravertebral fibrous membrane. ‘The two most postaxial of these cartilages coalesce together at their ventral ends. There are twenty distal cartilages attached to the distal ends of the twenty posterior median cartilages, one to each. They increase somewhat in length and thickness to the fifth, and then decrease gradually in length, but not in thickness, to the fifteenth, and then become considerably longer till the last, which is a little shorter than the first. Tue Caupa Fin (Plate LXXIV. fig. 6). This fin is supported by a number of cartilaginous rays, which on the ventral side are equal in number with the vertebre, there being one hemal cartilage to each vertebra. The seven or eight most preaxial hemal cartilages are distinctly segmented off (with a line of fibrous tissue indicating the separation) from the superincumbent parapophysial vertebral elements (p). The more postaxially situated hemal cartilages, however, are without any such segmentation. 302 449 PROF. ST. GEORGE MIVART ON THE The more preaxial dorsal cartilages are two in number to each vertebra. The most postaxial ones are, like the hemal cartilages, one to each vertebra; and in the inter- mediate space they pass gradually from one condition to the other. Both the dorsal and ventral cartilages are single and unsegmented, save as above mentioned, and there is neither bifurcation nor coalescence. The dorsal cartilages are not only much more numerous but also much inferior in size to the ventral cartilages. Tue Ventral Fin (Plate LXXIV. fig. 5). This fin is supported by a series of elongated closely set cartilages, to the greater number of which a second (distal) cartilage is appended, there being thus a distal series of smaller cartilages. Proximally, all the rays but the first unite with a continuous thick bar of cartilage, which runs more or less antero-posteriorly and at right angles to the elongated cartilages which join it. The first ray, which is stouter than the others, does not join this bar, but is united with the pelvic girdle (p). The number of elongated cartilages is twenty; and there are seventeen distal car- tilages attached respectively one to each of the elongated cartilages, except the three most postaxiad. The first seven or eight distal cartilages are of subequal length; they then rapidly shorten postaxiad. If we compare this fin-skeleton with that of the dorsal fin, the resemblance between its elongated and distal cartilages and the median and distal cartilages of the dorsal is striking. May, then, this antero-posteriorly elongated bar answer to the basal cartilages of the dorsal fin coalesced together? This elongated cartilage has not a slight resemblance to the metapterygium of the pectoral fin; and the large preaxial radial may be com- pared with the propterygium. Tue Pectoran Fin (Plate LX XIV. fig. 4). The pectoral limb is supported by a cartilaginous skeleton, consisting, as usual in Elasmobranchs, of a pro-, meso-, and metapterygium, with appended segmented rays. The metapterygium (c) is the largest of the three, club-shaped, and narrowing proximad. The mesopterygium (6) is less than half the size of the metapterygium, and acutely subtriangular in shape, with the apex proximad. The propterygium/(q) is a little smaller and also acutely subtriangular, but with the apex distad. To it is appended the most preaxial ray (or radial), which is rather wider than the succeeding ones, and is twice segmented, the terminal segment ending distally in a sharp point. The broad distal end of the mesopterygium supports the three next radials, each of which also consists of three segments. The terminal segment of the first of these three rays is also pointed distally ; but that of each of the two succeeding rays is obliquely truncated. All the other rays, or radials, are attached to the metapterygium, radi- FINS OF ELASMOBRANCHS. 443, ating in part from its distal end. The most postaxiad ray is the shortest, and has a broad and short root-segment. The next two, if not three, rays are connected with the metapterygium through the intervention of a small cartilage, broader than long, which seems to represent the basal ends of these cartilages coalesced into one small piece. The series of segments which continue postaxially the terminal segments of the most preaxiad rays, do not extend throughout the whole of the limb, but get smaller and smaller postaxiad till they are quite minute. They are all truncated distally, except, as before said, the two most preaxial. The more preaxial segments of the distal series are each attached not only to the truncated distal end of the penultimate seg- ment of their own ray, but also to the preaxial side of the penultimate segment of the ray next postaxiad. The more postaxial segments of the same series are also so counected; but as here the penultimate rays are not truncated but end distally in a point with a short facet on each side, it comes about that the most postaxiad terminal segments appear rather wedged in, and equally divided between, the distal ends of the two contiguous penultimate segments, instead of appearing specially attached to one of them. The arrangement and proportions of these cartilages cause the skeleton of the limb to extend distally to a somewhat greater extent towards its preaxial than it does towards its postaxial side, though this predominance is much less than the pre- dominance of the more preaxial fin-rays over the more postaxial ones. NOTIDANUS CINEREUS. Dorsat Fin (Plate LXXV. fig. 2). The cartilaginous skeleton of this fin consists of a series of nineteen radials, consider- ably and somewhat irregularly segmented, the whole being attached proximally to one deep and antero-posteriorly elongated cartilage (4). The most preaxiad of the radials, or rays, is very short, and unsegmented. The second ray consists of two, and almost all the rest consist of three, segments, whereof the deepest is in each case the largest. The four most postaxiad rays consist of but two segments each ; but then they are all attached to one considerable supporting cartilage, which has all the appearance of being formed of each of the deepest segments of the four rays combined together into a single cartilaginous plate. The long basal cartilage is convex ventrally and nearly straight dorsally, with small concavities answering to the bases of the several radials. ‘This basal cartilage certainly corresponds in position, and more or less in shape, to such a structure as would be formed by the coalescence of the whole series of basal cartilages in a fin consisting of three superimposed series of parts like that of Mustelus antarcticus. The whole cartilaginous dorsal-fin skeleton is in this fish placed at a great distance above the axial skeleton, the fibrous membrane between the two being almost equal in depth to the whole cartilaginous mass. 444 PROF. ST. GEORGE MIVART ON THE Tue VenTRAL Fin (Plate LXXV. fig. 4). The ventral-fin skeleton consists of a closely set series of rays, nineteen being attached to an elongate basal cartilage (4), and the remaining four (mostly much shorter) more preaxial rays being attached to the pelvic cartilage (p). Very short, more distally situated cartilages are attached to and between the distal ends of the more proximal rays ; and there is a trace of yet another, third, or most distal series, in the form of five minute cartilages, situated, again, between the more preaxial (excepting the first) of the small cartilages just described. The general resemblance of this structure to that of the dorsal fin is very noteworthy. THE Ana Fin (Plate LXXV. fig. 5). Here we have again a closely set longitudinal series of eighteen cartilaginous rays, or radials, attached to a more solid basal structure. The seven more preaxial rays are much larger than the succeeding ones; always segmented, they are somewhat irregularly so. ‘The succeeding more slender rays are not all segmented. The nine most postaxial rays are attached to one large and long continuous basal cartilage. The next six have a solid cartilaginous support for each pair. The ray next preaxiad has a cartilaginous support to itself, preaxial to which the basal cartilage seems irregularly segmented, longitudinally as well as vertically. The base of this fin has every appearance of having been formed by the coalescence of the basal parts of its constituent radials. Tue Pectorau Fin (Plate LXXV. fig. 3). The propertion borne by the basal cartilages, as a whole, to the rays is very large in the pectoral of this species. The metapterygium (c) is very elongated, very narrow, and once segmented towards its proximal end. It expands greatly distad. It supports fifteen radials. The mesopterygium (0) is also very large, triangular, with its shortest side proximad. It supports ten radials, including the most preaxial ones, as it extends to the preaxial margin of the limb; on which account I am disposed to regard this cartilage as consist- ing of mesopterygium, with the distal part of a segmented propterygium confluent with it. Indeed I know of no definition applicable to these cartilages, save what can be drawn from their position with respect to the limb-axis. The propterygium (a)—or, if the view above stated is correct, the proximal part of the propterygium—is exceedingly small, broader than long, and is entirely excluded from the radials by the extension preaxiad of the mesopterygium to the preaxial margin of the limb. I find the first (most preaxiad) radial to be, in the specimen depicted by me, slightly the broadest at its base, and to consist of four segments; but the segments of the radials generally are somewhat irregular. Those attached to the mesopterygium are FINS OF ELASMOBRANCHS. 445 slightly broader than are the metapterygial radials; but the bases of the two most post- axiad mesopterygial radials have coalesced into a single cartilage. The same is the case with the bases of the penultimate and antepenultimate postaxiad radials, also with the three next, and again with the two following—. e. with the seventh and eighth, count- ing from the postaxial margin. The most postaxial radial is unsegmented. Although the preaxial lobe of the fin, as formed by the fin-rays, extends distally much beyond the postaxial lobe, nevertheless the preaxial part of its cartilaginous skeleton does not extend distad nearly so much as does its postaxial parts, thus exhibiting a marked contrast to the condition presented by Wustelus antarcticus. SCYLLIUM CANICULA. Dorsa Fin (Plate LXXV. fig. 6). This fin has a cartilaginous skeleton, which is separated from the subjacent axial skeleton by a wide interval of fibrous membrane. The skeleton itself consists of twelve juxtaposed cartilaginous rays, or radials, which slope obliquely backwards and upwards, and are all, except the most preaxiad, doubly segmented. The segmentation takes place at nearly the same level in all, so that there comes to be a triple series of cartilages—one basal, the second median, the third distal. The basal cartilages are but ten in number—as the most preaxiad radial has no basal cartilage, and as those of the fifth and sixth radials have coalesced to form one large cartilage. This large cartilage is also the longest of the series of basals, whence their length regularly decreases both pre- and postaxiad. There are twelve median cartilages; and all, even the most preaxiad, are longer than the longest of the basal cartilages. They increase in length from the first to the sixth, and thence slowly decrease. The distal cartilages are eleven, since one is appended to the apex of each median cartilage except the most proximal one. - They increase slightly in length to the fifth, and thence slowly decrease, and are more equal in development than are the basal cartilages. Their apices are mostly more or less pointed. There is no coalescence amongst them, any more than in the median cartilages. Tue CaupaL Fin. This fin has inferior cartilaginous supports, which are hemal vertebral processes, equal in number to the vertebree whence they spring. Dorsally the fin is supported by much smaller but much more numerous cartilages, which have not an exact numerical relation to the vertebra, being rather over two to each vertebra. There is here, there- fore, as great a contrast between the dorsal and ventral skeletal elements of the caudal fin as in Mustelus antarcticus. 446 PROF. ST. GEORGE MIVART ON THE GINGLYMOSTOMA CIRRATUM. Dorsau Fin (Plate LX XVI. fig. 2). This small fin consists altogether of only fourteen radials, with fourteen elongated, juxtaposed, median cartilages. All but the first two of these median cartilages have distal ones attached to their summits; and these, again, may be more or less segmented. There are but nine basal cartilages, because the first two median cartilages have no basal ones, while what should evidently be the four most postaxial basal cartilages have coalesced into a small antero-posteriorly directed cartilage (b), the rudimentary repre- sentative of the large basal cartilage of the dorsal fin of Notidanus cinereus. This small cartilage has a certain resemblance to a pectoral metapterygium. The whole fin-skeleton is separated from the axial skeleton by a considerable mem- branous interval. THE VENTRAL Fin (LXXYV. fig. 7). This fin is supported by a series of sixteen continuous radial cartilages, attached to an elongated basal cartilage ()), except the first (most preaxial) one, which joins the pelvic cartilage (p). The distal ends of each of these cartilages, except the five most postaxial, bears a distal cartilage, which may be once or even (as in the second) twice segmented. Tue PectoraL Fin (LX XVI. fig. 1). The somewhat complex skeleton of this fin is in one respect open to two interpreta- tions; the mesopterygium may be regarded as present or wholly or partly absent, there being but two elongated basal cartilages, with an interposed membranous interval. As to the metapterygium there can be no doubt; it is elongated, articulates proxi- mally with the shoulder-girdle, is unsegmented, and supports ten radials. Between it and the next basal cartilage is a distinct membranous interval, which cannot be regarded as other than the representative of the membranous interval, which we shall find to exist in the pectoral of Chiloscyllium ocellatum (Plate LXXVI. fig. 4). The large elongated preaxial basal cartilage (p¢) seems evidently the homologue of that of Chiloscyllium, and may therefore possibly be the propterygium. But, just as in Notidanus the great preaxiad cartilage seems to consist of part of the propterygium in union with the mesopterygium, so here we may perhaps regard this cartilage as answering to portions of both meso- and propterygia united. This complex cartilage has at its proximal end a small transversely extended cartilage (p’) like the small similarly shaped proximal part of the propterygium of Notidanus. In Chiloscyllium ocellatum (Plate LX XVI. fig. 4) the most preaxiad radial creeps up beside the preaxial margin of the propterygium to its middle; but here two or three preaxial radials, having more or less coalesced (¢7), seem to creep up its preaxial edge, and abut against the separate proximal piece of the propterygium, while a border of FINS OF ELASMOBRANCHS, 447 small pieces of cartilage again fringes the preaxial edge of the series formed by the coalesced preaxial radials. It may be, however, that the coalesced piece (c7) may be more properly termed the propterygium. The two next radials also coalesce proxi- mally to form a subquadrate cartilage which joins the preaxial part of the distal end of the elongated more preaxial cartilage. The accession of cartilages to the normal preaxial edge of this fin-skeleton has been so great that I should be tempted to consider these coalesced cartilages as the propte- rygium, and the propterygium as the mesopterygium, but for the examples of Chilo- scyllium, Chimera, and Callorhynchus, and but for the close resemblance of this small proximal piece (p') to the small propterygial cartilage of Notidanus. The radials generally are much segmented, and extend distad much further on the preaxial than they do on the postaxial side of the limb. Not one (as in Chiloscyllium), but five radials are crowded together, and interposed between the meta- and propterygium; and of these the median one extends slightly further proximad than do the others, which are so crowded and superimposed that they cannot be distinguished till the parts are more or less separated. Of the radials attached to the propterygium, the first two coalesce proximally, the five others, more postaxial, are single. All the radials are much segmented, and mostly expand somewhat distally, while none appear to bifurcate. The skeleton of the limb projects distad most towards its preaxial margin. CHILOSCYLLIUM OCELLATUM. Dorsat Fin (Plate LX XVI. fig. 3). This small fin has a skeleton which consists of three superimposed series of conjoined cartilages, as in Wustelus antarcticus—a series of elongated median cartilages, a series of short distal, and another series of short basal cartilages. The whole structure is separated from the subjacent axial skeleton by a side interval occupied by strong fibrous membrane. The basal cartilages are thirteen in number, and are all nearly equally short, though those towards the postaxial end are somewhat less so. ‘There is no sign of coalescence between even any two of them. The median cartilages are also thirteen in number. The most preaxial pair are very short indeed; but all the others are much elongated, the sixth being most so, and the rest slightly decreasing in length postaxiad. The distal cartilages are but eleven, there being none to the first two median carti- lages. The second of these eleven is the longest of all, and then the first and third. They then decrease in length to the eighth and ninth, while the tenth and eleventh are slightly longer. The four shortest are shorter than any of the basal cartilages, while the three most preaxiad are longer than any of the basal cartilages. The VOL. X.—PartT x. No. 2.—February 1st, 1879. 3P 448 PROF. ST. GEORGE MIVART ON THE apices of the two most preaxial distal cartilages are pointed ; but the rest are truncated apically. Tue VENTRAL Fin (Plate LX XVI. fig. 5). Here we have a very simple skeleton, consisting of sixteen radials, whereof two are attached to the pelvic cartilage (p), and the rest to an antero-posterior basal cartilage (4) of nearly equal width throughout. The fourth radial is the longest ; and thence they decrease gradually in length in each direction. Distal cartilages are attached, one to the apex of each elongated cartilage, from the second to the eighth inclusive. That attached to the fourth long cartilage is the longest ; and thence they rapidly diminish, both preaxiad and postaxiad. THE Pecrorau Fin (Plate LX XVI. fig. 4). In this fin there are but two basal cartilages (m and p), separated by a wide mem- branous interval. They are both long cartilages, and of subequal length. The metapterygium is undoubtedly represented by the more postaxial of these two cartilages ; it supports distally six radials. The mesopterygium may possibly be represented, as well as the propterygium, by the more preaxial basal cartilage; but I am inclined to regard it as absent, since (as before said) it is hard to find any definition of a “ mesopterygium ” save that of ‘a basal carti- lage occupying the middle region of a limb-axis.” This absence is, moreover, the less surprising when we consider the conditions which exist in Chimera and Callorhynchus. The propterygium is represented by the preaxial basal cartilage, which is as elon- gated and large as the mesopterygium. It has a strongly concave mediad margin, and expands distally to give attachment to nine radials. The most preaxiad radial creeps up, as it were, along the preaxial margin of the propterygium to about its middle, thus presenting a much less development of what we found in Ginglymostoma. It consists of three segments. In the radials attached to the propterygium the middle segment is the broadest, and the distal segments are so applied to their supporting cartilages that the adjacent sides of the distal ends of these (from the fourth to the ninth) support a distal cartilage between them. One radial has its proximal apex placed, not against, but between the pro- and metapterygiun. Its distal cartilage is medianly divided, so that there are two little terminal cartilages placed side by side. The same is the case in the terminations of the three next more postaxial rays. The last three (most postaxial) rays do not show any distinct signs of segmentation in my specimen. It is the proximal part of the limb-skeleton which projects furthest distad. FINS OF ELASMOBRANCHS. 449 CESTRACION PHILIPPI. Pecrorau Fin (Plate LX XVI. fig. 6). In this form ! there are but two basal cartilages in the pectoral fin, one of which, the undoubted metapterygium, is in the form of an elongated triangle, with the apex proximad. It supports about ten radials, pairs of which may or may not coalesce together proximally. The other basal cartilage, apparently answering to both pro- and mesopterygium, is a large rhomboidal plate, with its narrowest border proximad, and supports the rest of the radials, which appear to be about seven in number. The bases of these radials, however, have either coalesced into, or have connately arisen as, thin, large, irregularly formed, but rather oval plates, two of which adjoin the distal margin of the pro-mesopterygium, which margin they entirely occupy except towards its postaxial end, where it gives attachment to five uncoalesced radials. The third large oval plate is attached to the distal margin of the more postaxiad of the other two oval plates, and to the postaxiad margin of the more preaxiad one, extending distad beyond the latter. The two preaxial oval plates, thus succeeding each other distad end to end, form the preaxial margin of the limb-skeleton, save for one or two very small pieces of cartilage, which are continued on distad by the most pre- axial of the radials. This exceptional form serves to demonstrate how radials may coalesce (or arise con- nately) at their proximal ends, so as to be replaced by large cartilaginous plates. We may, if we please, call these plates segmented pro- and mesopterygia; but there seems no good reason for doing so, since they appear to represent merely coalesced radials. Rather, I venture to think, they suggest the probability that the pro-, meso-, and me- tapterygium may themselves be nothing but plates formed by the coalescence of the bases of radials which were primitively distinct. The whole limb-skeleton extends distad to a much greater extent preaxially than postaxially, in harmony with the external form of the whole fin. ACANTHIAS BLAINVILLII. Dorsau Fin (Plate LX XVII. fig. 1). In this fin the skeleton may be considered as formed of the normal three super- imposed longitudinal series of adjacent cartilages, specially modified with respect to the sharp spine which it bears at its preaxial end. The basal cartilages are represented by the cartilaginous base of the spine, by one 1 Represented by Prof. Huxley in the P.Z. 8. 1876, p. 51, fig. 11; and by Prof. Gegenbaur, ‘ Untersuchungen,’ Heft 2, plate ix. fig. 3. 3P 2 450 PROF. ST. GEORGE MIVART ON THE large sheet of cartilage, and by a small triangular piece attached postaxially to the latter. The preaxial end of the base of the large cartilage is applied to, but is not continuous with, the subjacent axial skeleton. The median cartilage appears as a small piece applied to the postaxial side of the spine, and as three large subquadrate cartilages placed more postaxially and resting upon the basal cartilages—the first two upon the large preaxial basal cartilage, and the third upon the small triangular more postaxial basal cartilage. The distal cartilages are five small cartilages of nearly equal depth, placed side by side, and resting on the distal edges of the median cartilages. THE Caupau Fin (Plate LX XVII. fig. 3). The skeleton of this fin, again, shows a great contrast between its dorsal and its ventral portions. Ventrally, one large heemal cartilage is attached to each vertebra. At first it looks as if they were slightly less in number than the vertebre; a careful exa- mination, however, proves that this apparent disaccordance is due only to the varying degree of prolongation preaxiad of the root-portions of the hemal cartilages. Really the correspondence in number is accurate and complete. Dorsally the cartilages are more numerous. There are about two to each centrum preaxially, while, as we proceed postaxiad, they come by degrees to be one to each vertebra. They are more inclined postaxiad distally than are the hemal cartilages. Tue Pectorau Fin (Plate LX XVII. fig. 2). In the pectoral fin of this species all the three normal cartilages exist, but the middle one is considerably the largest. The metapterygium (c) is in the form of a very acute isosceles triangle, with the apex proximad. It gives attachment to fourteen radials. The mesopterygium (6) is also triangular, with its acute angle proximad; but it is much broader, and the distal margin is oblique. It gives attachment to ten radials. The propterygium (a) is very small, but, unlike the other basals, it is broadest proxi- mad. It gives attachment only to one small radial. The radials, except the two most preaxial ones and the three most postaxial, all broaden more or less distad, and they are all once or twice segmented. With the same exceptions, their apices are truncated. No radials bifurcate, even at their ex- tremity ; but the bases of the five most postaxiad appear to have coalesced into a single elongated cartilage which is appended to the pestaxial end of the distal margin of the metapterygium. The whole skeleton. extends distad very much more on its postaxial than on its pre- axial side, in spite of the greater extension distad of the more preaxial fin-rays. FINS OF ELASMOBRANCHS. 451 SPINAx NIGER. Dorsau Fin (Plate LX XVII. fig. 4). The dorsal fin of this Shark presents a great similarity to that of Acanthias blainvillii ; only coalescence has here been carried still further. One basal cartilage, besides the spine, represents by itself the two of Acanthias, and forms a longer and narrower triangular plate than in the last-mentioned genus. The median cartilages are but two in number, besides the elongated piece applied to the postaxial side of the spine. This piece is longer than in Acanthias. The median cartilage postaxial to it (or second median cartilage) is much as in Acanthias, save that it rises to form the distal margin of the most preaxial part of the fin-skeleton behind the elongated piece just mentioned. The third, or most postaxial median cartilage (m3), is greatly enlarged antero- posteriorly, compared with that of Acanthias blainvillii, and seems to answer to the two postaxiad median cartilages of that species fused together. The distal cartilages of Acanthias blainvillii are here represented by a narrow elon- gated cartilage, which shows an obscure and doubtful segmentation into three parts. As in Acanthias blainvillii, so here, the cartilaginous skeleton of the dorsal fin is applied to the axial skeleton by the proximal end of the spine and the most preaxial part of the base of the large cartilage which is placed just behind the spine. More postaxially this skeleton is connected with the axial skeleton by fibrous membrane only. SCYMNUS LICHIA. PeEcToRAL FIn. This species is very interesting!, because it serves to show how one single cartilage may represent the pro-, meso-, and metapterygium—all three. For there is but one basal cartilage, which supports all the radials (some seventeen in number), and is in the form of a large sub- triangular plate, with its truncated apex proximad. The postaxial border is the longest; and its distal margin slopes obliquely postaxiad and ventrad. 1 Represented in Gegenbayr’s ‘ Untersuchungen,’ plate ix. fig. 9. Right pectoral fin of Seymnus lichia. 452 PROF. ST. GEORGE MIVART ON THE SQUATINA ANGELUS. Dosau Fin (Plate LX XVII. fig. 5). This fin is supported by two large basal plates, to the distal margins of which four much smaller subquadrate cartilages are attached. The anterior basal plate is subquadrate; the postaxial one is triangular. They are evidently in series with the long cartilaginous spines, which surmount more preaxially placed vertebre. Both these spines and the basal cartilages seem discon- tinuous with the subjacent axial skeleton. ‘This is the case not only in the large dry specimen figured, but also in a very young specimen which I have had the oppor- tunity of examining. ParreD FINs. If the pectoral and ventral fins of Squatina be compared, it seems evident that the elongated basal cartilage of the ventral answers to the metapterygium of the pectoral, and the preaxial plate, apparently formed of coalesced radials of the ventral, to the propterygium of the pectoral. PRISTIOPHORUS JAPONICUS. First Dorsau Fin (Plate LX XVII. fig. 6). This fin is supported by two very large cartilaginous plates, which are directly con- tinuous proximally with the interneural plates of the axial skeleton. Between them is interposed another small plate continuous with a single interneural cartilage. Distally they support a row of small cartilages, to a few of which yet smaller cartilaginous nodules are distally appended. The plates are in series with other superaxial plates more anteriorly placed, and which are also continuous with the subjacent interneural cartilages. Tue Second Dorsau Fin (Plate LX XVII. fig. 7). This is very like the preceding, save that the small cartilage interposed between the two large basal plates is somewhat bigger. Tue VENTRAL Fin (Plate LX XVIII. fig. 2). Here the radials are mostly attached to the elongated basal cartilage; but the pelvic cartilage supports a single short radial, to the side of which four progressively elongated radials are successively attached. The more anterior radialseare segmented; the more postaxial pair have coalesced proximally, and join the postaxial end of the basal cartilage. FINS OF ELASMOBRANCHS. 453 Tue Pectoraw Fry (Plate LX XVIII. fig. 1). In this fin the mesopterygium is absent or defective, as there is a certain interspace left between the metapterygium (c) and the larger basal cartilage. This larger cartilage (which should probably be reckoned as both meso- and propterygial) is greatly ex- panded distally, and is the main solid constituent of the limb. On its preaxial side is a single elongated cartilage, somewhat broadest proximally, which may be the proptery- gium, or answer to the preaxiad upcreeping radials which are found in Chiloscyllium and Ginglymostoma. The proximal ends of the three most preaxial of the undoubted radials have coalesced into a single plate. The radials of the postaxial part of the limb are very irregularly arranged and segmented, and have, in part, coalesced into two plates, placed close to the distal postaxial angle of the metapterygium, projecting postaxiad much beyond it. The whole skeleton projects a little more distally on the postaxial than it does on the preaxial side of the limb ; but the fin-rays project more on the preaxial side. Tue Caupa Fin (Plate LX XVIII. fig. 3). The skeleton of this fin shows the same discrepancy between its dorsal and ventral supports as has been before noted in other species. Ventrally there are cartilages numerically corresponding with the vertebra, whence they continuously proceed. Dor- sally the cartilages are separate, and have no definite numerical correspondence with the subjacent vertebre, which they much exceed in number. PRISTIS CUSPIDATA. First Dorsau Fin (Plate LX XVIII. fig. 4). The dorsal fin of this form is supported by large cartilaginous plates, which are in series with other narrower superaxial cartilages situated more preaxially. The hindmost and longest cartilages are separated proximally by a membranous interval. Distally they support a series of twenty elongated, slender, unsegmented cartilages. The more postaxial of the two large basal cartilages appears to be segmented off from the vertebra subjacent; but the more anterior is intimately united with the axial skeleton beneath. The large basal cartilages seem to correspond with the median and basal series of cartilages of those dorsal fins which have three superimposed series of cartilaginous elements, or, if not to both, certainly to the basal series, and therefore to the conti- nuous basal cartilage, of the dorsal fin of Notidanus. RHYNCHOBATUS DJEDDENSIS. Dorsat Fry (Plate LX XVIII. fig. 5). The first dorsal of this species presents a series of eighteen moderately elongated juxtaposed radial cartilages, which descend to unite with the skeletal axis. What is 454 PROF. ST. GEORGE MIVART ON THE the nature of these eighteen cartilages? Are they homologous with the basal, the median, or the distal cartilages of the triserial dorsal fin-skeletons ? It is sufficient to compare their position, with reference to the fin they sustain and the body, with the relation borne to the same parts by the basal cartilages of one or all the other forms described, to see at once that they cannot be basal cartilages. They must therefore answer to distal, or to both median and distal cartilages in an undifferen- tiated condition. They are simple in form, neither coalescing, bifurcating, nor showing any plain evidences (in the specimen examined) of segmentation. They increase gradually in length postaxiad from the first to the eighth and ninth, and thence still more gradually decrease. The basal cartilages of other dorsal fins must then be represented by the two large cartilaginous plates which sustain these eighteen radials, The more preaxial of these two cartilages is much the larger, though it supports but seven of the radials. It is subquadrangular in shape, and seems continuous proximally with the cartilaginous axial skeleton. It is adjoined preaxially by the most postaxiad of a series of vertically elongated more or less oblique cartilages, extending upwards from the neural arches towards, and nearly to, the dorsal mid line of the body, which vertically elongated cartilages have the appearance of so many enlarged neural spines. The more postaxial cartilage is much smaller than its predecessor, and is subtri- angular in shape, with one angle ventrad to join the axial skeleton. Its dorsal margin is nearly as elongated as the dorsal margin of the more preaxial cartilage, and it supports eleven radials. The dorsal half of its oblique preaxial margin (which slopes ventrad and postaxiad) joins the postaxial margin of the more preaxial cartilage ; but its ventral half is separated from the preaxially sloping postaxial margin of the cartilage in front by a considerable membranous interval, triangular in shape, with its most acute angle dorsad. We have therefore in all three forms, Pristiophorus, Pristis, and Rhynchobatus, a triangular interval between the bases of the two large basal cartilages of the dorsal fin, which interval is either occupied by membrane or (as in Pristiophorus) by a small separate cartilage. TRYGONORHINA FASCIATA. Dorsat Fin (Plate LX XVIII. fig. 6). The cartilaginous skeleton of this fin extends but a very little way into the relatively very large expanse formed by the fin-rays. It consists of the longitudinal series of contiguous cartilages, supported upon two very large cartilaginous plates, which are continuous ventrally with the subjacent axial skeleton. The distal cartilages ave twelve in number, and increase slightly in length from the most preaxial to the fifth, and thence slowly decrease. ‘Their apices are rounded. FINS OF ELASMOBRANCHS. 455 The median cartilages are exceedingly small, and are also twelve in number. They are all of subequal length, and shorter than any of the distal cartilages except the first. The first median cartilage is more extended antero-posteriorly than are the others. They gradually decrease in size postaxiad from the first. The more preaxial of the two basal cartilages, the more preaxial of the two large cartilaginous plates, is very elongated and subquadrangular in shape. It supports the first five median cartilages and part of the sixth. The other basal cartilage is subtri- angular, with one angle ventrad. It supports part of the sixth median cartilage and all the more postaxial ones. These two cartilages are evidently serial homotypes of the dorsal portions of the cartilaginous plates which are situated preaxially and postaxially to them. The latter are continuous with the interneural axial cartilages ; but the basal cartilages of the dorsal fin show more or less evidence of segmentation existing between them and the inter- neural cartilages which ventrally support them. RAIA MACULATA. Dorsa Fin (Plate LX XVIII. fig. 7). Here there are two elongated basal cartilages which support radials, many of which bifurcate. The more preaxial basal cartilage is much the larger, and lies upon the interneural axial plates. It may be said to support nine radials, of which the seventh and ninth bifurcate distally. The more postaxial basal cartilage supports postaxially one single radial and a large cartilage which ends distally in six filamentary prolongations. Another, much-divided radial is supported by the dorsal margin of the more postaxial basal cartilage. The resemblance presented by these radials, with their supporting basal cartilages, to the cartilages of the ventral fins of Sharks is striking, with the exception of the distal bifurcations and subdivisions which here exist. CALLORHYNCHUS ANTARCTICUS. First Dorsau Fin (Plate LX XIX. fig. 1). The first dorsal-fin skeleton closely resembles that of Chimera, and exhibits a degree of concentration exceeding that of the spine-bearing Elasmobranchs Acanthias and Spinax. It consists of one single triangular plate only, which is probably articulated ventrad by its apex to a cartilaginous upgrowth of the vertebral column, and bears in front of it a very elongated spine. This plate, when compared with the dorsal fin- skeleton of the genera just mentioned, seems to answer to the basal, median, and distal cartilages all fused together into one mass. VoL. X.—Part x. No. 3.—February 1st, 1879. 3Q 456 PROF. ST. GEORGE MIVART ON THE Srconp Dorsa Fin (Plate LX XIX. fig. 2). The skeleton of the second dorsal jin is of the simplest possible character, consisting, as it does, of a longitudinal series of subvertically placed, not quite contiguous, carti- laginous radial pieces, forty-one in number. They increase very slowly in length post- axiad from the first to about the eighth, remain much the same to the tenth, then rapidly decrease in length to the fifteenth, then very slowly tothe end. This skeleton is nowhere in contact with the axial skeleton, but is separated from it by a wide tract of fibrous tissue, which shows very plainly the lines of attachment of the intermuscular septa sloping dorsad and preaxiad. The number of these does not at all correspond with that of the fin-radials. Tue VenTRAL Fin (Plate LX XIX. fig. 4). The cartilaginous skeleton of this fin is very unlike that of other Elasmobranchs’ ven- trals, and recalls to mind the skeleton of their pectoral fins. Thus, ¢. g., ifin the skeleton of the pectoral fin of Acanthias blainvillii the protruding preaxiad radials were shortened and the basals fused and contracted, we should have very nearly the condition we find in the ventral of Callorhynchus. We have, indeed, already in Scymnus lichia an example of a single cartilage taking the place of the pro-, meso-, and metapterygium. In the ventral of Callorhynchus we have one large basal cartilage, the proximal margin of which exhibits an articular concavity to fit on to the articular surface (Plate LX XIX. fig. 5 a) of the pelvic cartilage. This basal cartilage must answer to the pro-, meso-, and metapterygium of an Elasmo- branch pectoral, and is deepest on the postaxial or metapterygial side. Its distal curved margin is even and continuous, except a preaxial continuity with the radials, like that found in Chimera!. Here, however, the process is less distinctly marked. These radials, at first very short, increase in length postaxiad to about the eleventh, and decrease from about the fourteenth. At the apex of each is a small cartilaginous segment, longest in the more preaxial radials, 7. e. postaxiad of the second radial. The first radial bears two small cartilages, side by side, at its distal end. The bases of the twelfth, thir- teenth, fourteenth, and fifteenth radials have combined to form a small plate interposed between their distinct proximal ends and the distal margin of the basal cartilage. The skeleton of the fin projects distad most a little protaxial to its antero-posterior middle, about opposite the bottom of the concavity of the distal margin of the fin as formed by its fin-rays. Tue Petvis (Plate LX XIX. fig. 5). The pelvic cartilage is very remarkable, and far more like the pelvis of a Batrachian than is any other fish-pelvis known to me. ‘The right and left halves are separate. Each consists of a solid mass of cartilage (¢p), which may be called tschio-pubic, from Gegenbaur’s ‘ Das Skelet der Gliedmassen der Wirbelthiere im Allgemeinen,’ plate xvi. fig. 22, R. FINS OF ELASMOBRANCHS. 457 the dorsal postaxial angle of which a tall iliac process (il) extends dorsad. Behind the root of this process is another (@), which is the articular surface for the basal cartilage of the ventral fin. From the outer dorsal and outer ventral parts of the ischio-pubic cartilage two cartilaginous bars proceed preaxiad for some distance, and then meet, thus enclosing a space (0), which may be called an “ obturator foramen.” THE Prcrorau Fin (Plate LX XIX. fig. 3). The skeleton of this fin presents only two basal cartilages, whereof the preaxiad one (p), short and quadrate, must, I think, be deemed the propterygium}, while the other much elongated one (m) must include the mesopterygium as well as the metapterygium, unless we consider the former to be absent. Annexed distally to the propterygium is a triangular cartilage*, which appears to consist of, or to represent, parts of the four most preaxiad radials in one mass. At the distal end of the metapterygium is an hexagonal cartilage, whence no less than fourteen radials diverge, coalescing in various ways so as to produce a curious kind of mosaic pattern. Eleven radials are attached to the distal margin of the meta- pterygium; and two pairs of these coalesce together proximally for a short space. Almost all the radials have a small cartilaginous segment appended to their distal end ; and these small cartilages increase very gradually in size, from the one attached to the triangular preaxiad cartilage, to that of the last metapterygial radial: thence they decrease in size postaxiad with much greater rapidity. Minute extra cartilages are intercalated between the apices of the last radial attached to the metapterygium and those of the three or four next postaxiad—one between each. The proportion borne by the cartilaginous skeleton to the whole fin is small ; and it projects distally more towards the postaxial than towards the preaxial limb-margin. Compared with the pectoral of Chimera *, the coalescence at the bases of the radials is less, save near the distal end of the metapterygium, where it is more. ‘The series of small distal cartilages is less complete ; but this may be due to the defective condition of the specimen at my disposal. POLYODON FOLIUM. Ventral Fin (Plate LXXVIII. fig. 8). The ventral fin of this fish exhibits a most interesting structure, although unfor- tunately I have been able to examine only a dried specimen. ‘The interest consists in the fact that the cartilaginous skeleton is made up exclusively of elongated radials ' Tt is also thus designated by Gegenbaur in Chimera. 2 This is referred to the mesopterygium by Gegenbaur in Chimera ; but with such an interpretation I cannot agree. It seems to me to have no warrant, and to tend to mislead. 3 As represented by Gegenbaur, ‘ Untersuchungen,’ pl. ix. fig. 15. 3Q 2 458 PROF. ST. GEORGE MIVART ON THE which do not coalesce together at all, are not attached to any basal cartilage, and have no pelvic cartilage. There are eighteen such radials, each once segmented; and the distal segment of the most preaxial radial is longitudinally divided into two closely applied rods. I suspect, from the form of the distal ends of the distal segments, that other small cartilages were appended to their apices. There also appears to be some- thing wanting between the two segments of each radial '. POLYPTERUS BICHIR. Dorsa Fin (Plate LX XIX. fig. 6). This fin is supported by radials which give off on one side small secondary rays, proceeding dorsad and postaxiad. VENTRAL Fin anp ANAL Fin (Plate LX XIX. figs. 7 & 8). These two fins, when compared together, exhibit much similarity. In the ventral fin the fin-rays are attached to the apices of four cartilaginous radials, while those of the anal fin are borne by six essentially similar radials. Let but the two postaxial radials of the anal abort, and let the apices of its two or three preaxial radials elongate, coalesce, and become segmented off, and the form of the anal fin would be changed into that of the ventral one. ON THE NATURE AND HOMOLOGIES OF VERTEBRATE LIMBS. The examinations I have been so kindly permitted to make have partly been under- taken with a view to two special questions, which are :— (1.) What is the nature of vertebrate limbs generally ? (2.). What is the relation of piscine to other limbs? To answer these questions satisfactorily, replies must be found to four questions of a subordinate nature. These are :— A. Are the paired limbs structures of a nature distinct from that of azygos fins? B. Are paired limbs essentially axial structures, which have become more or less detached from the skeletal axis, or peripheral structures which have become secondarily more or less connected with it? C. What is the probable nature of the parts to which they are attached, 7. e. of the limb-girdles ? D. Cau we yet accurately determine the line of genesis of the cheiropterygium 2 ‘ Since this paper was delivered, I have received from Mr. James K. Thacher a paper on the ventral fins of Ganoids (from the Transactions of the Connecticut Academy, vol. iv.), in which he describes and figures the ventral fin of Polyodon. I find it has small apical cartilages, while at the point of segmentation a process is given off dorsad from a greater or less number of the radials—a process which may be called a sort of separate iliac process for each such radial. In Mr. Thacher’s specimen also a few of the radials coalesce proximally, but there is still no continuous basal cartilage as in the ventral-fin skeleton of Sharks. FINS OF ELASMOBRANCHS. 459 Although it may seem easier and more prudent to consider the various conflicting views with respect to the special homologies of piscine and non-piscine vertebrate limbs, before attacking the wider question—-that as to the essential nature of all vertebrate limbs—I nevertheless believe that the contrary course is the one which is less likely to mislead. Recent experience has abundantly shown to what different interpretations, more or less hastily assumed, special homologies may lend themselves. Thus even so distin- guished a naturalist as Professor Gegenbaur has found it needful to change his teaching upon this subject no less than three times since 1865. These repeated modi- fications, however, are far from being any discredit to that gifted anatomist; on the contrary, they testify to his flexibility of mind, and to his eagerness to adopt new truths rather than to adhere with pertinacity to any original view with which his name may have become associated. Nevertheless such changes show that the subject is one apt to mislead; and I am the more persuaded of this, since I believe he will find it desirable to make yet a fourth change, as I venture to think that even in his last view he, in common with Professor Huxley, is mistaken. Obviously, if we can arrive at any tolerable certainty as to the nature of vertebrate limbs in general, the recognition of that general nature, or fundamental condition, may be expected to throw light on the structures derived from that fundamental form, and, to a greater or less extent, on their special homologies. Vertebrate limbs may be, and have been taken to be, either parts derived (in one way or another) from the axial skeleton, or else special parts which have been ap- pended to and have become more or less intimately connected with that axial skeleton. In 1843, Oken! taught that arms and legs were so many liberated ribs; and Carus, following him to a certain extent, regarded them as elements radiating from the exterior of a rib-like arch. In 1848, Professor Owen? propounded the view that the vertebrate limbs are diverging appendages attached to ribs, and serially homologous with such parts as the uncinate processes of the ribs of birds and the branchiostegal rays of fishes. He also taught that the shoulder-girdle is an axial structure made up of pleurapophyses and hemapophyses. In 1832, Maclise* represented the limbs as modified ribs, the parts beyond the elbow and the knee, however, corresponding with the interspinous bones and fin-rays of fishes’ azygos fins. In 1857, Professor Goodsir # described the limbs as being parts of so many radiating actinapophyses, actinapophyses being parts radiating outwards from the sides of the 1 Lehrbuch der Natur-Philosophie, p. 330. 2 Archetype and Homologies of the Vertebrate Skeleton. 3 Todd's Cyclopedia, vol. iv. page 70, fig. 490. 4 Edinburgh New Philosophical Journal, vol. vy. (new series) 1857, page 178. 460 PROF. ST. GEORGE MIVART ON THE axial system, as do the epipleural spines of fishes, e.g. those so well developed in the Herring. He appears to have considered the limb-girdles as hypertrophied radial parts left after others primitively (or ideally) existing with them had disappeared. He says they are formed “ under the tegumentary covering, and therefore external to the proper visceral wall of the body.” The limbs of higher animals than fishes he regards as formed in the following way :—‘*‘ By atrophy or otherwise, one or more of the segments in the successive transverse rows of actinapophysial elements disappear, so as to leave in Man one in the arm, two in the next row for the coracoid and clavicle, and one in the proximal row for the scapula.” He is “inclined to believe that the coracoid is an acti- napophysial segment between the humerus and scapula, prolonged downwards towards the hemal margin of the body; that the scapula is a proximal element, elongated towards the neural margin of the body; that the clavicle is the only other retained element in the same transverse row as the coracoid ”!, In 18717, Professor Humphrey, recalling to recollection the fact that the dorsal and ventral mid lines of the trunk are formed by the junction of the bifold amine dorsales and ventrales respectively, thence deduced the conclusion that the azygos fins of fishes must be of bifold origin. He also puts forward the view that the pectoral and ventral fins are but certain portions of the inferior azygos fins, which are separated (or, rather, hindered from uniting) by the interposed body-cavity. Thus he regards the pelvic bones of the ventral fins and the so-called “ carpals” of the pectoral fins as modified interspinous bones. In this way the limbs of higher animals would be (as Maclise had before represented them to be) modified portions of the ventral azygos fin. In 1872, Professor Gegenbaur ° threw out the suggestion that the shoulder-girdle was a modified arch of similar nature to the branchial arches (7. e. a part of the visceral skeleton), and that the limbs may have been formed from the diverging rays of such an arch. This view he has confirmed and supported in more recent publications in 1874* and 1876°. The azygos fins of fishes are considered by him to be parts seg- mented off from the neural or hzmal spines of the vertebral column °. In April, 1877, Professor Macalister, in treating’ of the development of the muscular system, and in a later publication *, has spoken of vertebrate limbs in such a way as Discs pelle > Cambridge Journal of Anatomy and Physiology, vol. v. (second series, vol. iv.) page 58, plate ii. * Untersuchungen, Heft 3, p. 181, note. ‘ Grundriss d. vergl. Anat, page 494. His words are, ‘ Das ganze einem Fiederblatte ihnliche Skelet- gebilde stimmt auffallend mit manchen Stiitzapparaten der Selachier-Kiemen, und lisst dadurch ein Streif Licht auf die Frage von der Phylegenese der Gliedmassenbildungen fallen.” ° Morphologisches Jahrbuch, ii. Band, drittes Heft, p. 417, fig. 4. ° Grundriss der vergl. Anatomie, 1874, p. 488. 7 Dublin Medical Journal for April 1877. * Address to British Association of 1877, reported in ‘ Nature.’ FINS OF ELASMOBRANCHS. 461 seems to indicate that he is disposed to consider them to be derivatives from the axial skeleton—as formed, perhaps, from modifications of diverging appendages of the ribs, i. é. from rib-elements proceeding out laterally. He thus more or less returns, if I have not misunderstood him, towards Goodsir’s speculations. According to these various naturalists, the paired fins of fishes are derivations from the axial skeleton or from the branchial arches, being of the nature of productions ventrad or externad from such internal parts. Some of the authors cited regard the skeleton of the limbs as essentially similar to that of the azygos fins. Such are Maclise, Humphrey, Macalister, and (not certainly) Goodsir. Professors Owen and Gegenbaur, on the contrary, consider the paired and azygos limbs as fundamentally different skeletal structures; and Cuvier and Professor Huxley also appear to have shared this view. I do not recollect to have heard or read any distinct statement by Professor Huxley on this subject; but in his ‘ Anatomy of Vertebrates’! he speaks of ‘‘ the interspinous cartilages or bones,” of “ the median fins,” as “ as cartilaginous or osseous elements of the exoskeleton.” The further fundamental distinctness between the paired limbs and the axial (including the branchial) system appears to have been a view entertained by Cuvier and also by Professor Huxley. I have myself distinctly enunciated the same view in 1870°; and later® I have declared my conviction “‘ that the appendicular skeleton is no mere portion of the axial skeleton, but a distinct system of parts appended to and more or less closely and variously connected with the axial system.” ‘To this conviction I now adhere more firmly than ever. As with the limbs, so also the azygos fins may be, and have been taken to be, either parts of the axial skeleton, or parts fundamentally distinct therefrom. The third pos- sibility that they may be partly axial and partly peripheral, is an opinion which I believe has not been entertained hitherto by any one. The first view (that they are essentially axial) has been held by Geoffroy St.-Hilaire * and Maclise, probably by Professor Goodsir, and certainly by Professor Gegenbaur, who tells us° that their skeletal supports, “erscheinen im einfachsten Zustande als Gliedstiicke bedeutend ausgedehnter oberer Domfortsatzbildungen, die unter Ablésung vom Wirbel zu grosserer Selbstandigkeit gelangen.”’ The doctrine of the distinctness of these parts from the axial system appears to have had the sanction of Cuvier; and it has been taught by both Professors Owen and Huxley that the solid supports of the azygos fins were exoskeletal structures ; and this view I have myself held and put forward °. Altogether, then, the following views have been held :— 1 1871, page 43. 3 Lessons in Elementary Anatomy, 1873, p. ° Grundriss d. yergl. Anat. p. 488. ? Linnean Trans. yol. xxvii. p. 388. 30. 4 Mémoires du Muséum, vol. ix. 1822, p. 89. ° Lessons in Elementary Anatomy, p. 275 bo 462 PROF. ST. GEORGE MIVART ON THE (1) That both the azygos and paired fins are modified axial structures— Maclise, and probably Geoffroy St.-Hilaire, and not certainly Goodsir and Humphrey. (2) That neither the azygos nor the paired fins are modified axial structures— Cuvier, Huxley. (3) That the azygos fins are axial structures, but that the paired ones are not so— Gegenbaur at first, and even now if the branchial arches are not considered to be parts of the axial skeleton. (4) That the azygos fins are not axial structures, but that the paired fins are so— Owen. All these writers, and all naturalists as far as I know, have held all the azygos fins to belong to the same skeletal category. It is, however, obviously possible that such uniformity of nature and origin may not obtain, but that the apparent similarity may be due to a kind of homoplasy. All these writers have also hitherto, as far as I know, held that, whatever be the nature of the supports of the azygos fins, the skeletal parts of the limbs and paired fins are funda- mentally endoskeletal structures. There is, however, obviously another possibility. Not only may the paired and azygos fins be of fundamentally distinct origin and nature from the axial skeleton, but the hard parts of all these structures (the limbs as well as the azygos fins) may have arisen through centripetal chondrifications or calcifications, and so be genetically exoskeletal. This possibility first assumed to my mind some importance when examining the structure of certain Elasmobranch fins ; and the possibility became changed for me into a probability upon reading Mr. Balfour’s account of the mode of development of Elasmobranch limbs. In 1876, Mr. F. M. Balfour described’ the development of the limbs of Elasmo- branchs as “ special developments of a continuous ridge on each side, precisely like the ridges of epiblast which form the rudiments of the unpaired fins,” and extending “ on a level slightly ventral to that of the dorsal aorta,” “ from just behind the head to the level of the anus.” He adds, “ If the account just given of this development of the limbs is an accurate record of what really takes place, it is not possible to deny that some light is thrown by it upon the first origin of the vertebrate limbs. The fact can only bear one inter- pretation, viz. that the limbs are the remnants of continuous lateral fins.” As to the limb-girdles, supposed by Gegenbaur and Dohm to be modified gill-arches, he tells us, “‘ None of my observations on Elasmobranchs lends any support to these views; but perhaps, while regarding the limbs as the remains of a continuous fin, it might be permissible to suppose that the pelvic and thoracic girdles are altered ' Cambridge Journal of Anatomy and Physiology, vol. xi. part i. p. 132. FINS OF ELASMOBRANCHS. 463 remnants of the skeletal parts of some of the gill-arches which have vanished in existing vertebrates.” This discovery and happy suggestion as to limb-origin made by Mr. Balfour led me to examine the dorsal, ventral, and caudal fins of such Elasmobranchs as I could obtain for the purpose, to see if I could find :— (1) Whether any such coalescence of skeletal parts exists in azygos fins as might justify the belief that the cartilages of Elasmobranch paired fins had arisen by similar coalescence. (2) Whether there is any such solid fixation of the azygos fins to the axial system as might harmonize with the fixation of the paired limbs by means of the limb-girdles. (3) Whether I could get any light as to the nature of these limb-girdles. To take the third point first: I have been long convinced that the shoulder-girdle could not be a branchial arch or form of coalesced branchial arches, as also that the branchial arches could not be, as some have supposed, serial homologues of costal arches; for the branchial arches are situated within instead of external to the aortic arches, which vessels I took! to ‘‘ indicate the line of the pleuro-peritoneal division of the ventral lamine.” This conviction has been remarkably justified by Mr. Balfour’s recent discovery of the continuation of the pleuro-peritoneal cavity into the head, and externally to these aortic canals’. The branchial arches being then eliminated, there remained the costal arches, the outer cartilages of the gills of Sharks, and other parts external to the body-cavity as possible sources of the shoulder-girdle. It is just these parts which Professor Parker seems disposed to consider as the source of the shoulder-girdle. He tells us (‘ Mor- phology of the Skull, p. 343), “The extra branchials of the Dogfish are, at any rate, superficial cartilages related to the branchial arches, and they appear to be homo- logous with the scapulo-coracoid cartilages.” Nevertheless a careful consideration of all the leading modifications of piscine structure has ended by convincing me that such could not be the origin of this enigmatical girdle, as to the origin of which I could myself only arrive at negative results. With respect to the two former problems, it appeared, and it appears, to me evident that, if the limbs and their hard parts arise in a similar way to that in which the azygos fins and their hard parts arise, they must all probably belong to the same category of peripheral, non-axial structures. Moreover, since this is the mode of development of the individual, there arises of course, as Mr. Balfour indicates, an &@ priori probability that the primeval vertebrate limbs were a pair of continuous lateral folds, serving to balance the body in swimming. 1 Linn. Trans. J. c. and Lessons in Elem. Anat. pp. 222 and 225, fig. 196. 2? Cambridge Journal, vol. xi. part iii. April 1877, page 474. His words are: ‘“ It occupies” a “‘ position on the outer side of the aortic trunk of its arch,” VOL. X.—PART x. No. 4.—February 1st, 1879. 3B 464 PROF. ST. GEORGE MIVART ON THE Again, since in the Sharks the limbs are formed by the differentiation and predo- minant growth of two portions of such an, at first, continuous fold, so, as different species successively arose with ‘different wants and requirements, different portions of the fold may have been developed in different early forms! with a resulting difference of innervation in such differently arising paired limbs, the resemblance between such diversely arising limbs being due to homoplasy. The results of my examinations appear to me to confirm the belief that the nature of azygos and paired limbs is fundamentally the same. I suppose no one will dispute the truth of the doctrines :—(1) that when hard sup- porting structures first appeared in the dorsal fin, such structures formed a longitudinal series of similar, separate, more or less numerous parts—a condition they still present in most existing fishes; (2) that more solid and complex structures in the dorsal fin are secondary and derivative. These points being concceded, have we evidence of actual coalescence in the skeleton of the pectoral and dorsal fins ? Whatever view may be taken of the primitive pectoral fin (whether the archyptery- gium be conceived like the limb of Ceratodus, of Lepidosiren, of Raia, or as having arisen from diverging branchial rays), coalescence and segmentation must have taken place to produce the existing Elasmobranch pectoral. As to the dorsal fin, we have incipient coalescence between radials in Scyllium cani- cula (Plate LXXV. fig. 6) and Ginglymostoma cirratum (Plate LX XVI. fig. 2); and we have found in Notidanus cinereus (Plate LX XV. fig. 2) a very remarkable instance of such coalescence, most of the radials having come to repose upon one continuous basal cartilage, an instance of coalescence equal to any thing found in the pectoral fin. Again, in Acanthias (Plate LXXVII. fig. 1), Spinax (Plate LXXVIL. fig. 4), Callo- rhynchus (Plate LXXIX. fig. 1), Rhyncobatus (Plate LXXVIII. fig. 5), Pristis (Plate LXXVIII. fig. 4), and Pristiophorus (Plate LX XVII. fig. 6), as well as in Squatina (Plate LXXVII. fig. 5), we have more or less striking examples of considerable, though less complete or extensive, coalescence. But the pectoral fin has, as every one admits, the same essential nature as the ventral fin; and some ventral fins present a striking resemblance to dorsal fins. To see this it is only necessary to compare the ventral fin of Chiloscyllium (Plate LXXVI. fig. 5) with the dorsal fin of Raia (Plate LX XVIII. fig. 7), or the ventral of Notidanus (Plate LX XV. fig. 4) with its dorsal (fig. 2). The anal fin may also be sometimes made use of to show this community of nature between the azygos and paired fins. If the anal and ventral fins of Motidanus be compared (Plate LX XV. figs. 5 and 6), it will be difficult to believe them to be of radically different nature ; and a comparison between the anal and ventral fins of Poly- ' This consideration has also been brought forward by Mr. Balfour, /. c. p. 134. FINS OF ELASMOBRANCHS. 465 pterus will teach the same lesson (Plate LXXIX. figs. 7 & 8). But perhaps the most instructive ventral fin is that of Polyodon (Plate LXXVIII. fig. 8), where the skeleton of this fin remains in its, as I suppose, primitive condition—a double longitudinal series of simple parallel radials, without enlargement at either end or coalescence. That is to say, it is in the condition of the simplest kind of dorsal fin as regards its cartilaginous supports. The ventral fin of Callorhynchus antarcticus is, as we have seen, very remarkable for another reason—namely, from its close resemblance to the pectoral. A close ¢ea- ternal resemblance between these two fins appears in many Elasmobranchs; but here the resemblance is in the cartilaginous fin-supports, which is such as exists in no Shark or Ray known to me. This resemblance proves either (1) that both the pectoral and pelvic fins can be simultaneously and similarly modified, or (2) that a normal Elasmobranch ventral can assume the general appearance of a pectoral, or else (3) (if, as I am far from supposing, the ventral shows the original form) that the normal Elasmobranch pectoral has been changed in the overwhelming majority of cases from a condition more or less like that now found in the normal Elasmobranch ventral. If the ventral fin in the same great group can have the form either of a ‘ dorsal” or of a “pectoral,” it is a strong argument that the “ pectoral” may also have at one time borne the aspect of a “ dorsal,” or of such a “ ventral” as that of Polyodon. One formidable objection, however, remains against the similarity in nature of dorsals and pectorals. It is that which is afforded by the fact that in most fishes the pectoral fins acquire a firm fixation to the axial system, through a shoulder-girdle, while the dorsal fins have no similar support. But we have seen that in Pristiophorus and Pristis (Plates LX XVII. and LXXVIIL.) the dorsal fin becomes directly continuous with the axial skeleton by a mass of car- tilage large enough to warrant comparison with the shoulder-girdle itself, while it is more or less firmly united to the axial skeleton or movably connected with it in a number of forms, ¢. g. Rhynchobatus, Squatina, Acanthias, Spinax, Chimera, and Callo- rhynchus. It must be conceded, however, that the direction of this attachment is continuous, direct, and longitudinal (7. ¢. antero-posterior), and therefore at right angles to the line of extension exhibited by the shoulder-girdle; and it may be asked, “Tf the pectoral fin is similar in nature to the dorsal, why, when it comes to contract adhesion to the axial system, does it not contract that adhesion by means of a con- tinuous, direct, and much extended antero-posterior connexion, as the dorsal fin does?” To this question it may be replied :— (1) That, whatever be the nature of the shoulder-girdle, whether composed of one arch or of several arches united, or of whatsoever other parts, it could not cohere continuously and antero-posteriorly with the axial skeleton without 3R2 466 PROF. ST. GEORGE MIVART ON THE impeding the lateral flexure of the body in swimming, because the plane of its antero-posterior longitudinal adherence must be at right angles to that of the longitudinal adherence of the dorsal fin. (2) That generally the pectoral fins join the body at too low a level to enable them to abut directly upon the vertebral centra (or their representatives) or upon the neural arches, while the solid paraxial skeletal elements are very short in Elasmobranchs. (3) That they could not be directly connected in a straight line, even obliquely, with the skeletal axis without interfering with the body-cavity of that region. On all these accounts the pectoral fins, and the ventrals also, must (if they are to rest on a solid support for their more or less obliquely up-and-down mode of flapping) have a narrow connexion with a sustaining structure not directly continuous, in a straight line, with the skeletal axis; and these exigencies wonld account for the difference existing between the mode of attachment of the pectoral fins generally and the mode of attachment of the dorsal fin of Notidanus. But it may be said that the radial cartilages of the dorsal fin are really the prolon- gations outwards of the axial skeleton. This is the teaching of Gegenbaur, who, as we have seen, considers the dorsal radials to have been originally but productions of the neural] spines. The almost universal absence, however, of concordance or any definite numerical cor- respondence between these elements and the subjacent vertebra seems conclusive against this view. We have (in Notidanus, Spinax, Acanthias, Pristiophorus, Pristis, and Rhynchobates) found a series of forms which agree well with a process of coalescence and centripetal extension, but which quite disagree with the opposite view; for, according to the latter view, we should have to suppose that the neural spines became segmented, that they then enlarged and serially cohered down to their very bases, and that subsequently such solid base became absorbed close to the vertebral column, while remaining more or less coherent at a greater or less distance from it—a supposition which seems to me a very unlikely one. It may be objected that in Pristiophorus, Pristis, and Squatina there are elongated cartilages preaxial to the dorsal fin, rising up from the axial skeleton, and seeming at the same time to be neural spines and serial homologues of the large dorsal fin-plates of those genera. But are these parts really neural spines? In Sguatina not only do all or part of the basal plates of the dorsal fin seem discontinuous with the subjacent skeletal axis, but the neural spine-like cartilages, situated yet more preaxially, also seem separate from the axial skeleton on which they rest. Are these structures, then, neural spines segmented off, or are they dorsal radials which have as yet imperfectly cohered? The study of development can alone solve this problem; but if it should turn out that that even these cartilages are centripetal chondrifications, it would fully FINS OF ELASMOBRANCHS. 467 explain the condition of these parts in Ceratodus and Lepidosiren, and would lead us to regard the spinous processes of the human skeleton as essentially or originally exoskeletal structures (dorsal radials) which have adhered to and grown to be connate with the subjacent axial skeleton. Moreover it seems almost impossible not to regard at least the distal parts of the basal plates of the dorsal fin in Pristis and Pristiophorus as answering to the basal cartilage of the dorsal of Notidanus. But if the dorsal and anal fins have been developed centripetally, how about the paired fins? Did the hard parts developed within them shoot forth from the skeletal axis, or grow inwards towards that axis? Now, in the first place, the reasons which have just been enumerated why the paired fins cannot have an antero-posteriorly longitudinal attachment to the skeletal axis are equally reasons against the growing forth from that axis of the hard parts formed within them. Again, if we look at the ventral fins, and consider the multiplicity of their radial parts, together with the simplicity of their support, it seems incredible that the former should have been formed by a centrifugal chondrification. But if the centripetal process be conceded as that by which the paired fins were formed, it reflects additional probability on the centripetal formation of the azygos fins also. My examinations and reflections had proceeded thus far when my friend Prof. J. Reay Greene called my attention to a paper on “median and paired fins,” noticed in Silliman’s Journal as having been published in the third volume of the ‘ Transactions’ of the Connecticut Academy. Only the first part of the third volume was to be obtained at any scientific library known to me in London; and the ‘ Transactions’ of this Academy do not seem to find their way to our National Library at the British Museum. Under these circumstances I addressed myself directly to the author, Mr. James K. Thacher, who, with extreme kindness, sent me from America a separate copy of his valuable paper. The paper was to me of the highest interest. I found by it that its author and I had been simultaneously following out to generally similar results a similar line of thought, though in one important point he had gone beyond me. The memoir contains good figures and descriptions of the dorsal-fin cartilages of Petromyzon marinus, Mustelus canis, Galeocerdo tigrinus, Eulamia milberti, Sphyrena zygena, Odontaspis litoralis, Acanthias americanus, Raia levis, Myliobatis fremenvillei, and Acipenser brevirostris, as well as of some other structures, with a well-noted his- torical summary of recent publications regarding the genesis of the cheiropterygium and a reasoned statement of his own views as to the nature and homologies of verte- brate limbs. Singularly enough, the author states that “the origin of his paper lay in an observa- tion of a fin of Raia” —apparently the very resemblance between the dorsal and ventral fin-structure, which so greatly and so early impressed me. 468 PROF. ST. GEORGE MIVART ON THE He arrives at the conclusion that the paired and azygos fins are of similar nature, and are “ specializations of two lateral folds, as the dorsal and anal fins were specializa- tions of the median folds.” He has, moreover, the merit of having arrived at this con- clusion before becoming acquainted with the mode of development of the Elasmobranch paired fins described by Mr. Balfour. The important point on which Mr. Thacher has gone beyond me is in the suggestion that not only have the paired fins themselves been developed centripetally, but that the pelvic girdle itself is but due to the further extension inwards of that centripetal growth :— Fig. 2. — —— —_—" —— ae — ——— 7 — — — a — es = Ss gm — — = _— —— a — — — — ~ —a —~ SS. SS A ieee > Diagram illustrating Mr. Thacher’s conception of the genesis of the pelvic girdle. A. Left-ventral-fin skeleton seen from beneath in its primitive condition. B. The same, with the proximal ends of the radials coalesced into a longitudinal basal cartilage. C. The same with the basal cartilage segmented. D. Both ventral fins, showing the development of a median process from the preaxial basal cartilage of each. E. The same, with the preaxial cartilages medianly adherent, and the proximal part of the preaxial radials coalescing. I regard this as a most happy suggestion, and I give it my entire adhesion. Mr. Thacher says little! about the nature or formation of the shoulder-girdle, and nothing as to the causes of its mode of attachment to the skeletal axis and of the manner of its growth inwards from the pectoral fins. But the very different conditions of the two limb-girdles in Fishes seems to me to readily explain these phenomena. The more powerful action of the pectoral fins, as compared with the ventrals, requires for them a firmer point d@’appui than is needed for the latter. The ingrowth from them, therefore, seeks a direct solid fixation, just as in Pristis and Pristiophorus the resistance to lateral pressure, in using their saw, is facilitated by the direct solid fixation of the dorsal fin to the skeletal axis. But the reasons before enumerated ? as hindering the direct ingrowth of the pectoral-fin base, or the formation of an antero-posterior longi- tudinal connexion between it and the spinal column, determine both the shape and 1 His words are :—“ The cartilages, spreading, met in the middle line; and a later extension of the cartilages dorsad completed the limb-girdle.”—Z. c. p. 298. ? See ante, pp. 465, 466. FINS OF ELASMOBRANCHS. 469 attachment of the shoulder-girdle. Now, inasmuch as we have seen that the pectoral fins must have an antero-posteriorly narrow point of support, so it is also necessary for that point of support, if it seeks a firm basis, to do so by extending, not antero- posteriorly, but vertically ; and thus it has come to abut against its fellow of the oppo- site side ventrally, and against the ventral column (as in Raia), or against the skull (as in the Teleostei) dorsally. It must also grow in two directions, and so develop into a limb-girdle, as, had it but one oblique attachment, it would not be firmly fixed. In this way the mystery of the limb-girdles seems to be satisfactorily explicable ; they are neither modified hypaxial nor paraxial parts. They have, as I have before contended, neither the nature of branchial arches, nor of modified ribs, but are parts respectively swi generis, due to the ingrowth of originally superficial structures—exo- skeletal hardenings which have grown inwards, and become endoskeletal. It might possibly be objected against these views, that in fishes which have no pectoral fins, such as Amphipnous, Monopterus, and Symbranchus, there is, none the less, a pectoral arch, and one, moreover, which, coming, as it does, just behind the branchial arches, and being of very small size, has much the appearance of, and structure serially homologous with, these arches. But the absence of paired fins in these fishes is manifestly due to degradation, since their shoulder-girdle, instead of being a nascent rudimentary structure, has its normal Teleostean form, each lateral half consisting of three bones, the uppermost of which, in Monopterus and Symbranchus, forks proximally in the usual fashion. Therefore these fishes are in a similar case with the legless lizards, which are provided with a small, but normally developed, shoulder-girdle. As to the pelvic girdle, it might perhaps be expected that it would in some fishes take on a development and adhesion to the spinal column approximating to what we find in tailed Batrachians. But, in fact, no fish known to me uses its ventral fins at all in the way in which these Batrachians use theirs; and it is therefore not wonderful that the structure is never such as theirs. In some fishes, however (e.g. Cotylus cephalus, Chorisochismus dentex, and Sicyases sanguineus), the pelvic arch may acquire considerable complication. But this compli- cation is peculiar, and does not present any approximation to higher, non-piscine, pelvic structure. In Chimera and in Callorhynchus, however, the pelvis, as we have seen 4, does assume much the appearance of the pelvis of air-breathing vertebrates (Plate LX XIX. fig. 5), though it does not acquire a direct fixation to the spinal column. Altogether, I conclude then that the limb-girdles are ingrowths from the paired fin- skeleton, and that such skeleton is the modified remnant of a longitudinal series of similar radial parts, like those of the ventrals of Polyodon, formed primitively in a continuous lateral fold, such fold being similar, save in situation, to the folds which form the azygos fins, which therefore, together with the limbs, form one distinct genus, of parts of which the paired limbs are species. 1 See ante, p. 456. 470 PROF. ST. GEORGE MIVART ON THE Mr. Thacher does not share my views as to the nature of the solid supports of the azygos fins in Ceratodus and Lepidosiren. In order to consider this point, however, I must recall to recollection conditions existing in the dorsal fin of Pristis, Pristio- phorus, and Notidanus. As to the first two genera, it seems to me, as I have before said, impossible not to regard their continuous ray-bearing cartilages as homologous with the continuous ray-bearing cartilage of Notidanus. The study of development will show how they arise in the individual, but not necessarily how they arose in the race; for having once acquired union with the axial skeleton centripetally, they may subsequently have come to arise in continuity with that skeleton. The fact that in Ceratodus and Lepidosiren the azygos fins are supported by rays which are, though segmented, continuous with the neural and hemal spines, does not, to my mind, necessarily prove them to be of axial origin. Mr. Thacher, as I have said, thinks otherwise; he says!:—‘The cartilaginous supports of the median fold of the Dip- noans are very long and segmented; they are simply elongated neural spines, and not primordial fin-rays in any homological sense. If they were formed by the reduction in number of the primordial fin-rays and their coalescence with the neural spines, it is impossible that we should not have here and there an extra one, and some evidence in the case of others of such a junction.” He is even disposed? to associate the Dipnoi with the Amphibia and Amniota as one great generic group, on the strength of their having entirely lost the “ primordial median fin-rays.” If they were really so lost, it would be indeed an important character: but I do not see the force of his argument in favour of their having been so; for since the median fin-rays have become, in some Elasmobranchs, equal in number to the vertebre for a varying extent of the vertebral column, what is there improbable in that correspondence having become, in a few exceptional forms, complete? Again, the Dipnoi must at least be allowed to be allied to the Ganoids; and in Lepidosteus and Amia we find a near, though not an accurate, correspondence in number between the dorsal-fin radials and the subjacent vertebre. But, as I have before suggested, even if these spines of Ceratodus and Lepidosiren are neural spines, does that necessarily forbid their having been derived centripetally from the dorsal fin? It is possible that all neurapophyses may have been so derived. As to this question, however, the study of development can alone decide. At any rate undoubted neural spines may retrogressively assume the condition of dorsal radials, as in those lizards (e. g. Basiliscus) in which these processes send off delicate prolongations into the long tegumentary processes of the dorsal crest. Mr. Thacher, in controverting the view here advocated, and affirming ? that ‘‘ neither are median fin-rays derived from neural spines, nor neural spines from fin-rays,” brings forward as an argument the condition of these parts in Acipenser as showing the distinctness between the neurapophysial and fin-ray elements. But this very instance 1 DL. c. p. 292. 2 DL. c. p. 293. 2 Lc. p. 292, FINS OF ELASMOBRANCHS. 471 appears to me to show how slight a modification might have produced that separation and aggregation of the dorsal radials which now exists. It is evident how little change Fig. 3. MU Figure of the dorsal-fin cartilages of Acipenser, after Thacher. would be needed to make the three preaxial radials continuous with three neural spines subjacent to them. Again, let the neural spines of Acipenser be somewhat more developed, and let the preaxial parts of the radials coalesce together (as in Notidanus) and then with the neural spines, and we at once get the condition presented by Pris- tiophorus and Pristis. But what are we to consider the nature of the skeleton of the caudal fin to be? We have again and again found a discordance between its dorsal and ventral portions. In Elasmobranchs the dorsal cartilaginous radials are ever more numerous than the vertebree, are unconformable to them, and small, while in the ventral portion of the fin they conform in number to the vertebre (however at first apparently discordant from differences of size) and are large. In the former they are not continuous with the vertebre ; in the latter (ventral portion) they are so. This seems to point to a differ- ence in nature between the dorsal and ventral portions of the caudal fin in, at least most, Elasmobranchs. This distinctness of nature seems pointed to also in Polyodon, where the dorsal part of the caudal fin is supported by bones homotypal with the dorsal-fin supports, while its ventral part is quite devoid of such structures (thus differing from the adjacent anal fin), and has for supports extensions of the axial skeleton only, to which fin-rays are distally attached. As to the condition of this part in Teleosteans generally, I will not as yet venture to say any thing, except that it is plain that in such forms as Murena the dorsal and ventral parts of the caudal are similar in nature and homotypal with ordinary dorsal and anal fins. If the neural spines of vertebre are ingrowths from the dorsal azygos fin-fold, one would naturally expect that the hemal spines would be similar ingrowths from VoL. x.—part x. No. 5.—February 1st, 1879. 358 472 PROF. ST. GEORGE MIVART ON THE the ventral azygos fin-fold. The peculiar and apparently essentially axial nature of the solid supports of the ventral part of the caudal fin seems to tell against this hypo- thesis; but it need not necessarily do so, since, if we may suppose these subcaudal ingrowths in certain forms to have taken place very early, they would have acquired an integration with the skeletal axis much surpassing that of their dorsal homotypes, with a resulting difference of development and appearance such as we find now in the caudal structures here noticed. But the caudal fins of different kinds of fishes may have arisen in different ways in different cases, and apparent superficial resemblances may be due to the action of homo- plasy. The ventral portion of the caudal fin of Murena, that of most Teleosteans, and that of Elasmobranchs may have been derived independently from the primeval caudal fin}, It remains to consider the relations existing between the paired fins of fishes and the limbs of higher Vertebrates, and also the development of the cheiropterygium. Before entering upon detailed comparisons, it must first be determined which part of the pectoral fin is preaxial and homologous with the radial side of the arm and hand? As to the first question, I think there can be no doubt that the view entertained by Cuvier and Huxley is the right one, and that the preaxial margin of the pectoral fin is that margin which is turned obliquely dorsad. To this view Professor Gegenbaur, though at first opposed to it, has now adhered 2 as the more probable; and I marvel that Mr. Thacher still prefers the opposite one ®. I marvel all the more because his own paper seems to me to be by itself capable of demonstrating the truth of the view which he is disposed to reject; for it shows ¢ (as was long ago shown by Swan”®) that the more preaxial spinal nerves go to the dorsad or propterygial side of the fin, while the more postaxial spinal nerves go to the ventrad or metapterygial side of the fin. If, then, the view here advocated is correct, 7. e. if the true pectoral archipterygium had a skeleton like that of the first dorsal fin of Chiloscyllium, the second dorsal of 1 Since this paper was written, Mr. Alexander Agassiz has very kindly sent me a paper of his (from the 13th vol. of the ‘ Proceedings’ of the American Academy of Arts and Sciences) on the development of the tail in some species of fishes. In this very interesting memoir the frequent absence of vertical homology between the parts respectively on the ventral and dorsal sides of the upturned extremity is made very evident, the ventral portion being a secondary growth in the Flounder, and in Atherina, Batrachus, Lumpus, Ctenolabrus, Poronotus, Lophius, and Gasterosteus. * Morpholog. Jahrbuch, 2nd. vol. 3rd Heft, “‘ Zur Morphologie der Gliedmassen der Wirbelthiere,” p. 396. 3 He says, ‘“‘ The weight of evidence seems to me in favour of the view... . . that the metapterygial edge of the fish-fin corresponds with the radial or thumb-side of the hand.” ‘ In a section on ‘* The Innervation of the varied Fins in Mustelus canis” (1. c. p. 304). 5 See Swan’s “ Illustrations of the Comparative Anatomy of the Nervous System,” 1864, p. 32, plate xi. FINS OF ELASMOBRANCHS. 473 Callorhynchus (Plate LX XIX. fig. 2), or the ventral of Polyodon, the preaxial primitive radials laid the foundation of the propterygial part of the fin, and the postaxial radials of its metapterygial part. The same analogies apply to the ventral fin, the continuous basal part of which is evidently the homotype of the pectoral metapterygium, while the preaxial radial or radials serially repeat the propterygium, as is most plainly shown by the limbs of Squatina. The radials, by increase, approximation, and coalescence, may be conceived as pro- ducing the form of pectoral found in the Rays, and by further coalescence (through Squatina-like forms) that found in the Sharks and in most other fishes. According to this view, the fin of Ceratodus cannot (as supposed by Giinther and Huxley) represent the archipterygium, and, far from being a primitive form! whence the piscine limb may have been derived by a progressive shortening (as suggested by Professor Huxley”), is, on the contrary, a very special and peculiar structure, which is carried to a still more abnormal development in Lepidosiren, by progressive elongation and by atrophy of the postaxial radials. And this relationship is surely what might have been expected @ priori. Surely, on the evolution theory, air-breathing vertebrates were later developments, and such a structure as the limb of Ceratodus must have long postdated that of the limbs of primeval Elasmobranchs, if not that of the earliest Teleosteans also. A fortiori, then, Professor Gegenbaur’s view, that the cheiroptery- gium is due to a further continuation of that process by which the Elasmobranch fin has, in his opinion, been formed from a Ceratodus-like limb, is quite fallacious °. Much more probable is the hypothesis of Professor Huxley, that the cheiroptery- gium, “as an organ of support and locomotion,” requiring “length, strength, and 1 Both Dr. Giinther and Professor Gegenbaur have suggested that the Ceratodus-limb may have resulted in different ways from the coalescence of a longitudinal series of parts which, according to Professor Gegenbaur, may have been like branchial rays. See Phil. Trans. vol. clxi. p. 534, and ‘ Untersuchungen,’ Heft 3, p. 181, note. ? Professor Huxley says:—‘ The most highly specialized forms of ichthyopterygium result from the shorten- ing of the skeleton of the fin, the approximation of its distal elements to the shoulder-girdle, and the multi- plication of its rays.”—P. Z. S. 1876, p. 56. 3 Dr. Giinther says (/.c. pp. 532, 533), as to the second cartilage of Ceratodus :—* Although externally it appears as a single, flat, broad, short piece, unevennesses of its surface indicate that several primary pieces are coalesced in it. Iam confirmed in this view by a horizontal section in which the lines of the former divisions are preserved in the shape of tracts of a white connective tissue.” Professor Huxley, in his specimen, found “no trace of such divisions,” though examined in a microscopic section. Mr. Thacher observes, as to this divergence (J. c. p. 300):—* The fact that Huxley could find no sign of division in his specimens seems of little weight in view of the complete fusion which we know takes place here and there in median fins.” I may add that I have carefully examined the specimen described by Dr. Giinther, and I am perfectly con- vinced of the correctness of his observation. 382 474 PROF. ST. GEORGE MIVART ON THE mobility,” has resulted from “ the lengthening of the axial skeleton, accompanied by a removal of its distal elements further away from the shoulder-girdle, and a diminution in the number of its rays.” At the same time the possibility should be borne in mind— a possibility which, I am inclined to think, should rather be deemed a probability—that such forms as Ceratodus and Lepidosiren may have arisen as a special lateral offshoot from the main stem, an offshoot not leading to the parents of Batrachians and higher Vertebrates. Mr. Thacher suggests! the possibility that “the fringing rays are new developments,” and is “strongly inclined to suspect that the three portions of the second piece of the limb of Ceratodus, which Giinther describes, indicate three fin-rays, and that the feathering of one of them is a later development.” Dr. Peters long ago” pointed out (a fact adverted to by Mr. Thacher) that there is a resemblance between the fringing rays of the Dipnoi and the secondary filaments borne by the dorsal-fin rays of Polypterus, which are unquestionably new developments (Plate LX XIX. fig. 6). Professor Gegenbaur at first * derived the Elasmobranch limb from that of the Lepi- dosiren type of limb as below :— I o 7. Y; LEA A. Lepidosiren type; B. Ventral type; C, D, E. Ray pectoral type; F. Shark pectoral type. Subsequently he adopted* the Ceratodus-limb as the fundamental form, explaining the formation of the Elasmobranch pectoral as due to the great increase in length and coalescence of the diverging radials of one side, these growing into the pro-, meso-, » L.c. p. 304. * See Miiller’s Archiv, 1845, p. 3, ‘‘ Die einzige Analogie dazu liefern die abgesonderten Riickenflossen des Polypterus bichir, welche aus einer Flossenstange und einer dayon ausgehenden Flossenfahne bestehen.” 3 Jenaische Zeitschrift, vol. v. Heft 4, 1870. 4 Jenaische Zeitschrift, vol. vii. Heft 2. FINS OF ELASMOBRANCHS. 475 and metapterygium, with their annexed cartilages, while all, or nearly all, the radials of the opposite side became aborted. He has thus come to consider the metapterygium and the parts serially continuous with it as the representatives of the main axis of the Ceratodus-limb ; and he has been followed in this by A. Bunge}, who has brought forward interesting examples of what he considers to be persistent traces of the originally biserial arrangement, persisting towards the distal end of the line of the metapterygium. Professor Huxley has given? another interpretation. He considers the mesoptery- gium and the parts continuous with it as the representatives of the limb-axis of Cera- todus. This is an interpretation which logically follows from his conception of the Elasmobranch pectoral as a drawing-in of the Ceratodus-fin. He says*:—‘‘ In my judgment the mesopterygium of Gegenbaur is tle proximal piece of the axial skeleton, which constantly retains its primary articulation with the pectoral arch.’ But, as we have seen (Plate LX XIX. fig. 3), this element is certainly wanting, as a distinct part in Chimera and Callorhynchus, and most probably altogether from the skeleton of Chiloscyllium (Plate LX XVI. fig. 4). In Polypterus, again, its proximal part disappears, so that articulation with the shoulder-girdle devolves entirely on the pro- and metapterygium. But whether absent or not, such reduction in size and partial atrophy tells against its represent- ing the true limb-axis of higher forms. Such axis must be rather represented by the pro- or metapterygium; but as to these a few observations will be more in place later on. As has been said, Professor Huxley is inclined to derive the higher Vertebrate limb from the Ceratodus type, and this by the atrophy of its proximal fore-and-aft radials and the hypertrophy of its distal radials. ‘Thus the main axis of the Ceratodus-limb becomes the middle digit of the cheiropterygium, and the four other digits are the terminations of the ultimate and penultimate radials of the two sides. His‘ special interpretation of the genesis of the hand from the Ceratodus-limb is as follows:—‘“ The parts which are traversed by a line drawn through the humerus, the intermedium, the centrale, the third distal carpal, and the third digit in the cheiropterygium, may be regarded as so many mesomeres, representing the axis of the archipterygium. Two pairs of parameres are retained on each side. The preaxial are:—(1) the radius, the radiale, the first distal, carpal, and the pollex; (2) the second distal carpal and the index. The postaxial parameres are:—(1) the ulna, the ulnare, the fifth distal carpal, and the digitus minimus; (2) the fourth carpal and the annularis.” * Jenaisch Zeitschrift, vol. viii. (1874), p. 293, plates 8 & 9. 2 P. Z. 8. 1876, p. 24. * Lc. p. 55. 4 L.c. p. 56. 476 PROF. ST. GEORGE MIVART ON THE This view may be represented (with respect to the arm and hand) as in the annexed diagram :— Fig. 5. 1, humerus; 2, intermedium (lunare); 3, centrale; 4, magnum; 5, 6, 7, & 8, metacarpal, and then pha- langes of middle digit; 9, radius; 10, radiale (scaphoid); 11, trapezium; 12, 13, 14, metacarpal and pha- langes of pollex; 15, trapezoides; 16, 17, 18, 19, metacarpal and phalanges of index digit ; 20, median half of unciforme; 21, 22, 23, and 24, metacarpal and phalanges of fourth digit ; 25, ulna; 26, ulnare (cunei- forme); 27, external half of unciforme ; 28, 29, 30, and 31, metacarpal and phalanges of the fifth digit. As Professor Huxley truly says!:—‘‘The confirmation or refutation of this hypo- thesis is to be sought in development, and in the condition of the limbs in the Paleozoic Amphibia.” And he tells us that his suggestion is made “ mainly in the hope of stimulating investigation in both these directions.” Professor Huxley’s view is, I consider, the best yet suggested; but I cannot feel much confidence in its accuracy for all that. I cannot do so on account of the poly- morphic nature of the Elasmobranch fins, the multitudinous differences of which seem to point to protean transformations, and therefore to an extreme plasticity, which may have generated the cheiropterygium from any one of numerous possible sources. More- over Professor Huxley’s view seems to demand the unity of the centrale, while this carpal ossicle, deemed double by Gegenbaur in Jchthyosaurus and Plesiosaurus, has now been shown to be actually double, not cnly in Cryptobranchus, but in both limbs of three species of Siberian Urodeles*, while the representations given in the memoir referred to certainly do exhibit the tarsal and carpal structures rather as portions of oblique rays, proceeding radiad and distad from the ulnar side of the limb, than as disposed in harmony with Professor Huxley’s hypothesis. . P. Z. 8. 1876, page 57. * Namely, in Ranodon sibiricus, Salamandrella keyserlingii, and Salamandrella wosnefsenskyi. See Morpho- log. Jahrbuch, vol. ii. 3rd Heft, page 421, and plate xxix.: “Die iiltesten Formen des Carpus und Tarsus der heutigen Amphibien,” by Dr. R. Wiedersheim. FINS OF ELASMOBRANCHS. 477 Altgether, valuable as are such suggestions as hints for the direction of inquiry, I do not think we are yet in a position to do more than guess at the exact form and precise mode of origin of the archipterygium. There is also the objection that something may be said in favour of the line of the propterygium having furnished the air-breathing limb-axis. Before entering upon this point, however, it may be well to review Professor Huxley’s! special interpretations of fin-homologies. In the first place, he regards the Notidanus-pectoral as the least modified form of Elasmobranch pectoral, that in which the Ceratodus type is least departed from. The mesopterygium he regards as the remains of the shrunken axis; the propterygium he appears to deem a modified preaxial ray ; and the metapterygium he considers’ as “ formed by the coalescence of the axial ends of the pectoral rays.” He does not, however, explicitly refer to the small cartilage called by Gegenbaur propterygium, and figured by him *, and found by me (Plate LXXV. fig. 3, a) to be divided off from the bases of the preaxial rays by a preaxiad extension of the meso- pterygium. Moreover I did not find the most preaxial ray much “ broader than the others,” or “ two-jointed,” but distinctly divided into four segments, while Gegenbaur represents it in W. grisews* as much narrower than the others, and with only one joint. In WV. cinereus? he repre- sents it with five segments and a very small, still more pre- axiad cartilage annexed to it very near its base. Evidently, then, there is much individual variation in these parts, as it might well be expected @ priori that there would be. I cannot, therefore, regard the above explanation of Notidanus by Ceratodus as more than an ingenious speculation. In Cestracion® we have, as we have seen, a great cartilage which evidently represents both the pro- and mesopterygium undifferentiated, or coalesced; and we have evidence of the great plasticity of those structures in the appearance of three large cartilages connected (two directly and the third by these two) with the large compound cartilage. That this great cartilage is so complex in nature is made evident, I think, by the pectoral of Scymnus lichia’, in which one great cartilaginous mass must answer to the pro-, meso-, Pectoral fin of Scymnus - lichia. and metapterygium, all three. Profesor Huxley next speaks of Scylliwm*°, which has a greatly reduced mesoptery- gium (still considered, of course, by Professor Huxley as the reduced limb-axis) and a 1p. Z.8. 186. ? L. c. p. 50. 3 Untersuch. ii. plate ix. figs. 1 & 2. 4c. fig. 1. 5 L. c, fig. 2. ® Huxley, J. c. fig. 11, page 51, and Gegenbaur, pl. ix. fig. 3. 7 Gegenbaur, plate ix. fig. 9. ® L. c. fig. 10, p. 48. 478 PROF. ST. GEORGE MIVART ON THE metapterygium “ extended upwards along the postaxial face of the mesopterygium, until it has not only reached the articular surface of the pectoral arch, but furnishes a large part of the articular cavity. In like manner the proximal preaxial ray (proptery- gium) has ascended along the preaxial face of the axial cartilage,until it also is able to furnish a facet which completes the anterior part of the cup for the condyle of the pectoral arch.” But even in Notidanus both the pro- and the metapterygium contribute to form the articular cavity; it is no wonder, therefore, that they do so in Scyllium. The Professor then goes on to speak of Sguatina and Raia, treating them as steps further and further removed from the archipterygium. But these structures are at least as explicable on the other (as I believe, correct) view, namely that they are more and more near to the archipterygium. Evidently, on the centripetal theory, the rays incipiently coalescing proximad would first form various cartilages like those of the anal of NMotidanus (Plate LXXYV. fig. 5), or the pectoral of Myliobates aquila (Gegenbaur, plate ix. fig. 14), before coalescing into three cartilages, viz. into the pro-, meso-, and metapterygia. He then proceeds to consider the pectoral of Chimera, which he interprets as having a small proximal articular axial cartilage (answering to the mesopterygium of Votidanus), a large triangular cartilage distal to it and formed of preaxial rays coalesced, and a metapterygium formed of coalesced postaxial rays. A comparison of Chimera (or of Callorhynchus) with Scyllium shows, indeed, that the metapterygia of the two are evidently homologous ; but a comparison of Chimera or Callorhynchus with Notidanus can, I think, render it as little doubtful that Professor Huxley’s axial cartilage, instead of being mesopterygial, is the homologue of the pro- pterygium of Notidanus—as Gegenbaur has determined it to be. The triangular carti- lage distally united to the propterygium is also, I think, manifestly propterygial. In this case the mesopterygium has disappeared altogether as a distinct part; but in Scyllium it is much reduced—so much so that a slightly greater reduction would give the condition which we find in Chimera. This comparison strengthens the view that it is the mesopterygium which is absent in Chiloscyllium ocellatum. I regard the Rays as showing pectorals more approaching the true archipterygium than those of other Elasmobranchs. It is, I think, evident that there is a tendency to an inverse development between the lateral and azygous folds and their derivatives. In fishes in which the paired fins are minute or absent (Murena, Symbranchus, &c.) the azygos fins are extensively developed ; and when, as in the Rays, the paired fins are in excess, the azygos fins tend to disappear. On this account, if on no other, we should, I think, regard the Ray form of dorsal fin-skeleton as less primitive than the simpler form of Sharks. By analogy the simple, multiradiate, paired limb-skeleton of the Rays seems to point FINS OF ELASMOBRANCHS. 479 to a possibly multiplied, but at the same time little differentiated, form of skeleton. That they are more primitive in this matter seems evidenced by the persistence in many forms of a continuous lateral fold, reaching far back along each side of the tail, and seeming to be a persistent remnant of the primitive lateral continous fold from which the paired limbs were developed. As to Teleosteans, not only do I think, with Giinther, that “ the arrangement of the limb-skeleton of Ceratodus is foreshadowed in the pectoral fin of Acipenser,” but I think it probable that the Teleostean form of limb is in some respects more primitive, not only than that of Ceratodus, but than that of even any Elasmobranchs. I con- sider, of course, the Teleostean limb a specially modified one, and do not believe the Elasmobranch limb to have been formed from it, any more than that it was formed from that of the Elasmobranchs; but I think that in such forms as Anguilla (Parker, Shoulder-girdle, pl. ii. fig. 9) and Blennius (page 46, figs. b, c) we have indications or reminiscences of a more primitive type, whence Teleosteans may have been derived on the one hand, and Elasmobranchs on the other. To return to the question as to which part of the Elasmobrach limb may most pro- bably be conceived as having developed into the limb-axis of air-breathing vertebrates, I think, as before said, there are some considerations which seem rather to point to the line of the propterygium as that genetic part. ‘These considerations are :— (1) That it is more probable that the cheiropterygium was derived from a shark- like form of limb than from such a fin as the pectoral fin of the Rays, which (on the theory here advocated) is more primitive and more thoroughly un- fitted for supporting the body on land. (2) That on turning the pectoral limb of a Shark ventrad in such a way that its dorsal surface looks outwards, as in higher vertebrates, it is the propterygium which seems to be in the line of support needed for the body, the fore limb of a quadruped having necessarily to extend more or less preaxiad distally. (3) That this circumstance renders it very unlikely that the metapterygium should have remained and developed into the cheiropterygial axis; and that conse- quently, if the mesopterygium is to be deemed absent in Chiloscyllium, it can only be the propterygial element which has so persisted. (4) The large size which the propterygium has obtained in Chimera, Callorhynchus, Cestracion, Scyllium, and Pristiurus seems to demonstrate that this element is at least susceptible of great increase, and of predominance over the meta- pterygial element, and so far renders it less improbable that it was the genetic element of the cheiropterygial axis. As to the structure of Elasmobranch yentrals, it may be unhesitatingly affirmed that they lend no yestige of support to the theory that the Ceratodus-limb was the archi- 1 See Gegenbaur’s ‘ Untersuchungen,’ vol. ix. figs. 3, 7, 8, and 15. you. x.— Part x. No. 6.—February 1st, 1879. 37 480 PROF. ST. GEORGE MIVART ON THE pterygium. No structure exists in them which can be certainly considered the relic of a primitively biserial limb, while their genesis from primarily distinct radials, significantly indicated by a comparison of the dorsal fin of Notidanus with its ventral fin on the one hand, and with the dorsal fins of such forms as Chiloscyllium on the other, seems all but demonstrated by the simple ventral fin of Polyodon. Thus, while I agree with Professor Humphrey as to the ventral and pectoral fins being skeletal developments in lateral body-folds, I believe him to be quite mistaken in regarding them as divaricated portions of the fold which forms the anal fin (the ventral fold). The paired fins are of different origin; and the ventrals have only assumed their approximately mid-ventral position through adaptive developmental changes. On the whole, I feel strongly persuaded that Vertebrate limbs have been formed as follows :— (1) Two continuous lateral longitudinal folds were developed, similar to dorsal and ventral median longitudinal folds. (2) Separate narrow solid supports (radials), in longitudinal series, and with their long axes directed more or less outwards at right angles with the long axis of the body, were developed in varying extents in all these four longitudinal folds. (3) The longitudinal folds became interrupted variously, but so as to form two prominences on each side, 7. ¢. the primitive paired limbs. (4) Each anterior paired limb increased in size more rapidly than the posterior limb. (5) The bases of the cartilaginous supports coalesced as was needed, according to the respective practical needs of the different separate portions of the longi- tudinal folds, 2. ¢. the respective needs of the several fins. (6) Occasionally the dorsal radials coalesced (as in Notidanus &c.) and sought centripetally (Pristis &c.) adherence to the skeletal axis. (7) The radials of the hinder paired limb did so more constantly, and ultimately prolonged themselves inwards by mesiad growth from their coalesced base, till the piscine pelvic structure arose, as, ¢.g.,in Squatina. (8) The pectoral radials with increasing development also coalesced proximally, and thence prolonging themselves inwards to seek a point d’appui, shot dorsad and ventrad to obtain a firm support, and at the same time to avoid the visceral cavity. ‘Thus they came to abut dorsally against the axial skeleton, and to meet ventrally together in the middle line below. (9) The lateral fins, as they were applied to support the body on the ground, became elongated, segmented, and narrowed, so that probably tbe line of the propterygium, or possibly that of the mesopterygium, became the cheiroptery- gial axis. (10) The distal end of the incipient cheiropterygium either preserved and enlarged preexisting cartilages or developed fresh ones to serve fresh needs, FINS OF ELASMOBRANCHS. 481 and so grew into the developed cheiropterygium; but there is not yet enough evidence to determine what was the precise course of this trans- formation. (11) The pelvic limb acquired a solid connexion with the axial skeleton (a pelvic girdle) throngh its need of a point dappwi as a locomotive organ on land. (12) The pelvic limb became also elongated; and when its function was quite similar to that of the pectoral limb, its structure became also quite similar (e. g. Ichthyosaurus, Plesiosaurus, Chelydra, &c.); but for the ordinary qua- drupedal mode of progression it became segmented and inflected in a way generally parallel with, but (from its mode of use) in part inversely to, the inflections of the pectoral limb. The amount of apparently spontaneous change needed for the processes enumerated may appear excessive ; but I believe that in general we far too little appreciate the excessive plasticity of the animal organism—a plasticity which results in, and is evidenced by so many instances of the independent origin of similar structures. The plasticity of animals might be expected to be great; for plasticity is bodily reaction in response to external stimuli. ‘The response which is most rapid and complete, is “sensation ;” and an animal is a creature the essence of which is “ sen- ’ sitiveness” or “ impressionability.” Indeed an animal may be described as a more or less complex arrangement for carrying about, nourishing, and perfecting a plexus of sensations. In conclusion, then, the replies to our initial inquiries will be as follows :— A. The paired limbs and azygos fins are of similar origin and nature. B. Paired limbs are essentially peripheral structures which have become more or less closely connected with the skeletal axis. C. The limb-girdles are ingrowths from the bases of the limbs. D. The line of genesis of the cheiropterygium cannot yet be accurately determined. (1) Vertebrate limbs are differentiations of continuous lateral folds; they are therefore not limited to four, save for locomotive convenience!. There might be several successive paired limbs on each side, as there are sometimes several successive dorsal fins, which are differentiations of a continuous dorsal fold. Thus limbs and azygos fins are structures differing mainly by position, the limbs being lateral, the azygos fins median. ‘They may all be viewed as different species (ichthyopterygia, cheiropterygia, dorsal fins, &c.) of one fundamental 1 The Rey. Dr. Haughton has shown good reason why the number should be thus limited in most land yertebrates. I do not at all understand why the lateral fold should not sometimes haye resulted in the forma- tion of more than four limbs in Fishes. 37 2 Fig. PROF. ST. GEORGE MIVART ON THE set of parts (pterygia), for the sum total of which the term sympterygiwm may be employed. (2) Piscine fins ave related to the limbs of higher vertebrates as structures which have diverged to a less degree from their primitive condition—and this not only because the piscine body, as a whole, is a more primitive structure, but also because their fins are still used for locomotion in that medium in which their primeval form, the continuous lateral fold, was first developed. DESCRIPTION OF THE PLATES. PLATE LXXIV. . First dorsal fin of Zygena malleus. . First dorsal fin of Lamna cornubica. . First dorsal fin of Mustelus antarcticus. . Left pectoral fin of ditto. a, propterygium ; 6, mesopterygium ; c, metapterygium. » . Left ventral fin of ditto. Pp, pelvic cartilage. . Caudal fin of ditto. p, parapophysial vertebral elements. PLATE LXXV. . Caudal fin of Lamna cornubica. . First dorsal fin of Notidanus cinereus. 6, basal cartilage. . Left pectoral fin of ditto. a, propterygium ; b, mesopterygium ; c, metapterygium. . Left ventral fin of ditto. b, basal cartilage; p, pelvic cartilage. . Anal fin of ditto. . First dorsal fin of Scyllium canicula. . Left ventral fin of Ginglymostoma cirratum. b, basal cartilage; p, pelvic cartilage. PLATE LXXVI. . Left pectoral fin of Ginglymostoma cirratum. cr, coalesced radials; p', small proximal piece; pt, propterygium ? Fig. S- Fig. Fig. Fig. to} Fig. Fig. FINS OF ELASMOBRANCHS. . First dorsal fin of ditto. b, coalesced basal cartilage. . First dorsal fin of Chiloscyllium ocellatum. . Left pectoral fin of ditto. m, metapterygium ; p, propterygium. . Left ventral fin of ditto. b, basal cartilage ; p, pelvic cartilage. . Left pectoral fin of Cestracion philippi. PLATE LXXVII. . First dorsal fin of Acanthias blainvillii. . Left pectoral fin of ditto. a, propterygium ; 5, mesopterygium; c, metapterygium. . Caudal fin of ditto. . First dorsal fin of Spinax niger. m 3, most postaxial median cartilage. . First dorsal fin of Squatina angelus. . First dorsal fin of Pristiophorus japonicus. . Second dorsal fin of ditto. PLATE LXXVIII. . Left pectoral fin of Pristiophorus japonicus. c, metapterygium. . Left ventral fin of ditto. . Part of caudal fin of ditto. . First dorsal fin of Pristis cuspidata. . First dorsal fin of Rhynchobatus djeddensis. . First dorsal fin of Trygonorhina fasciata. . First dorsal fin of Raia maculata. . Left ventral fin of Polyodon foliwm. PLATE LXXIX. . First dorsal fin of Calloriynchus antarcticus. . Second dorsal fin of ditto. . Left pectoral fin of ditto. m, meso- and metapterygium ; p, propterygium. . Left ventral fin of ditto. 483 484 ON THE FINS OF ELASMOBRANCHS. .5 Right side of pelvic cartilage of ditto. o, obturator foramen ; a, articular surface ; iJ, iliac process ; ip, ischio-pubic mass. y. 6. Two spines of dorsal fin of Polypterus bichir, showing the secondary postaxiad rays. . 7. Left ventral fin of ditto. ig. 8. Anal fin of ditto. WHWesley Lith 4 IW $ GQ GEL 4) O4 TV SOME. L001 TOC OtU USA LAX Hanhart ump FINS OF ELASMOBRANCHS. Ts Hanhart amp WHWesley Lith FINS OF ELASMOBRANCHS \ \ ~ . mt t = x * r 2 — ‘ - ~~ ’ cs wauuouuuuia Ca ee a ee ee pes ee pec pe 4 al ti! WHWesley hth. Hanha>: imp FINS OF ELASMOBRANCHS. WWesley lith Ip IYO, Cs 2 Lh LIONS ~L000/ VO FINS OF ELASMOBRANCHS i Y) UOC a a. bet ar Re Ly ~~ ane , ® RS WHWesley ith Hanhart imp ° FINS OF ELASMOBRANCHS. ; é \ } y % { . ) i | ‘ 1 Sy i ‘ap ec i ‘ “ i fei ‘ : - es ! « Ui @ £ bs . \ ee , A ; Poti ' m tn, * 7 . lea ie r é Sty ‘ a] a ‘ of * re y ‘ * 1 ~ ' ' i / } , é he ‘ ’ -s an Fart Deol Ibe Vel 10 FULM WHWesley hth. Hanhart imp FINS OF ELASMOBRANCHS. Qo mede5 id XIII. On the Mechanism of the Odontophore in certain Mollusca. By Parrick GEDDES. (From the Biological Laboratory of the Royal School of Mines.) Received November 23rd, 1877. Read February 19th, 1878. (Prarns LXXX.-LXXXIL] AFTER consideration I have resolved to omit describing in full detail the variations of the odontophore in the three types of Mollusca here selected (Loligo, Patella, and Buccinum), and even to omit naming all the muscles figured, since for description the drawings, with their explanation, may suffice, while no nomenclature can be completely rational and permanent till the organ has been exhaustively studied throughout the Odontophora. It is only proposed briefly to attempt an explanation of the principle of the apparatus, incidentally reviewing the leading existing theories of its structure and function. Most of the misconceptions which have prevailed on this subject appear to have arisen by commencing with the highly specialised organ of the Whelk, and taking that asa type. A priori it seems better to begin with the Cephalopod, of which the organi- zation in certain other respects appears to have departed much less from the ancestral Molluscan type’. From this point of view the buccal mass of Nautilus should be of extreme interest. The reader is doubtless aware that the odontophore in all Cephalophora consists of a toothed radula, placed on a partly muscular and partly cartilaginous support. The single pair of buccal “ cartilages” in the Cuttlefish (Loligo) are very slightly developed, mere thin sheets, to which muscles are attached. Their lateral position, sloping upwards and forwards, and their connexion even at the apex by transverse muscular fibres only, are to be noted (figs. 1-9, dif). Over the sharply crescentiform apical edge formed by this ventral adductor muscle of the cartilages, there fits a stout and equally sharp-edged infraradular sheet (figs. 4 & 5, irs) of muscular fibres with abundant connective tissue, into which is inserted a pair of muscles on the ventral side (vis, fig. 8), while above it is continuous with the longitudinal muscles (fig. 4, ed/m, dlm) both internal and external to the transverse fibres which connect the cartilages on the dorsal side. Very deep within the inner mass of longitudinal muscles, and as close as possible to the framework, lies the sheath of the radula (figs. 4, 5, 9, rs). The radula itself has a quite sharp and permanent bend (figs. 4 & 5, r), where it passes round the edge of its flexible muscular support, to the sides of which it is also 1 Gegenbaur, ‘ Grundziige der Vergleichenden Anatomie.’ VoL. X.—PART x1. No. 1.—February 1st, 1879. 3U 486 MR. P. GEDDES ON THE MECHANISM closely fitted by its non-dentigerous flaps. In action the powerful internal retractor muscle (d/m), aided by the external retractor (ed/m), must drag slightly upon the apex of the radula, but mainly upon the apex of the infraradular sheet ; and both, prevented from sliding over the apical edge of the framework by their sharp permanent curvature in both longitudinal and transverse sections, as well as by the parallel curvature and absence of firmness in the framework itself, must be pushed upwards and forwards, the teeth of the radula thus scraping harshly against an object held in the jaws immediately above. During this upward and forward curvature, if the dorsal transverse fibres of the framework (df) contract, they will assist the stroke, at the same time deepening and narrowing its cavity. The return is effected by the elasticity of the whole apparatus, assisted by the contraction of the ventral longitudinal muscles of the infra- radular sheet (v/s, figs. 1, 4, 5), as also by the ventral transverse fibres of the support (rtf, figs. 1, 4, 8). A pretty good rough idea of the action may be gained by laying the open hand on the table with the palm upwards and bringing up the four finger- tips till they nearly meet, the knuckles being meanwhile pushed forward along the table, and then drawn backwards as the hand reopens—or even by moving the tip of the tongue, as in licking, curling up its sides at the same time, to represent the sulcus. Attention is particularly called to the fact that the construction and adjustment of each and all of the three parts of the whole mechanism—the radula, the infraradular sheet with its muscles, and the framework, in all the details above described—is such as to render a strap-over-pulley motion practically impossible. In Patella (figs. 10-23) the framework is very highly developed, containing three pairs of cartilages. The anterior pair is continuous for a short distance in front, and the pointed apex, though slightly rolled downwards, is very thin and flexible. The ventral pair of protractor muscles, united into a continuous sheet (pr, fig. 14), originate in the lower lip ; the postero-lateral (/pr, fig. 20) at the sides of the head. Both pairs are inserted into the posterior buccal cartilages; but while the lateral muscles are attached at about the middle of these, so as to pull the whole buccal mass forward without moving its parts upon each other (as also do the anterolateral pair), the lower sheet is attached to the ventral edge of the cartilages (figs. 14, 20); and these, therefore, rock upon the anterior pair at the joint, thrusting them upwards and forwards. The infraradular sheet fits over the anterior pair of cartilages, but is too thin to have a permanent angle at the apex as in Loligo. Its ventral muscles originate in the posterior buccal cartilages’; and their contraction must push the anterior pair, and thus the apex, downwards. The powerful muscles which antagonize them on the dorsal side must bend the cartilages upwards at the apex through a not inconsiderable arc, the sulcus at the same time deepening. It will be seen, however, that in this case the \ These cartilages seem to be represented in Loligo by a membranous sheet similarly placed, and from which the dorsal and ventral muscles of the infraradular sheet originate. OF THE ODONTOPHORE IN CERTAIN MOLLUSCA. 487 infraradular sheet is not prevented from sliding a little way over the apex of the carti- lages, and bearing the radula with it. And thus the movement, mainly due to the protraction and retraction, the apical bending, and the approximation and divarication of the cartilages by means of their own muscles and those of the infraradular sheet, is complicated with a less important factor—a slight sliding of the radula over the apex of the cartilages. (In this respect, though not in general structure, the odontophore of Patella may be less specialized than that of oligo.) This hypothetical interpretation I have been able fully to verify by simply watching the action in live Limpets turned oyer on their backs. The complex apparatus of Buccinum (figs. 24-34) will now present little difficulty. The single pair of cartilages are united and curved upwards at the apex as usual, and connected dorsally and ventrally by transverse fibres (dtf, vif). As usual, too, this apex is slender and sharp, bending upwards when a slight pull is given to the radular sheath (7s) to imitate contraction of its muscles. The infraradular sheet is much reduced, and at first sight appears to be absent; but, as may be seen by scraping the lower surface of the radula, or by taking transverse sections, it is closely attached to this, and is hence not drawn separately in figs. 29 & 34. The sheet is attached to the walls of the mouth on the ventral side about twice as far down as on the dorsal (figs. 26, 28, 29, & irs, 33), where it is even with the opening of the gullet. Thus a slight contraction of the highly differentiated longitudinal dorsal radular muscles must bend the weak cartilages upwards at the latter comparatively fixed point, their sulcus being deepened partly by the dorsal transverse fibres (dtf), aided by the slips which pass obliquely from the radular sheath to the cartilages (rs, figs. 26 & 33). The return is effected by the ventral longitudinal muscles (v/s), probably assisted by the contraction of the transverse fibres (vf) which connect the cartilages on the ventral side. Little of that sliding movement over the apex of the cartilages which we saw in Patella can here take place, owing partly to the weakness and curvature in two planes of the cartilages, partly to the sharpness of their apex, eminently unfitting it fora pulley-block, partly to the slight fixed flexure of the radula and its want of pliability, and largely also to the attachment of the infraradular membrane to the sides of the mouth all round (figs. 26, 27, 28), which thus fixes the radula very steadily over the cartilages. Some little yielding may take place; but it must be evident, from the above considerations, that the movements of the radula are similar to, and dependent upon, that licking action impressed upon the buccal cartilages in the way we have seen. Thus the explanation here put forward has something more in common with that of Cuvier’, ‘‘ that the tongue-plate is essentially passive, and that its movements depend upon the protraction, retraction, divergence, and approximation of the cartilages” (though, of course, these have nothing to do with rudimentary jaws), than with the later theory proposed by Professor Huxley”, who, chiefly from the unquestionable fact 1 «Mémoires sur les Mollusques.’ See also Lacaze-Duthiers on Dentalium, Ann, Se. Nat. 1856. ? Huxley, Phil. Trans. 1853, “‘ On the Morphology of the Cephalous Mollusca.” 3U 2 488 MR. P. GEDDES ON THE MECHANISM that the radula in Buccinwn has muscular attachments, both dorsal and ventral (figs. 24 & 28), to the cartilages over which it is bent, has argued that the radula “travels over the ends of the cartilages just like a band over its pulley, the cartilages being entirely passive in the matter,” and, further, that “the tongue has the same chain-saw-like mode of operation throughout Cephalopoda and Gasteropoda.” The anatomical facts, however, above described appear to me to render these propositions, on the whole, no longer tenable, though an observation remains. In the transparent bodies of some Heteropoda, Prof. Huxley! describes a chain-saw movement; so, if the framework remains quite stationary, I can only suggest that the sliding of the radula over its support, which we saw as a secondary factor in the Limpet, though impossible in the Cuttlefish and highly improbable in the Whelk, may in these animals have acquired greater importance. A remaining argument, however, of high importance in confirming the view here taken is the embryological. In developing embryos of Limnea, days before they leave the egg, the action of the incipient odontophore may be observed with a low power. Neither radula nor muscles are yet to be seen; but the future cartilages rock steadily, licking upwards and forwards with deepening sulcus, quite as described above in the adult mechanisms of Patella. The fully developed structure in Pulmogasteropods has the usual type; and a transverse vertical section, say of the apparatus in the common Snail’, is sufficient to prove its use in licking; while in some cases, e.g. the adult Zimnea, this may easily be proved by direct observation. In Aplysia, too, the same structure may be observed. Though not coming strictly, perhaps, within the limits of the present paper, a by- product of the inquiry may here be noted. Dr. Herman Fol? recently published some interesting observations upon Ascidians, from which it appears that the endostyle secretes an inverted hollow cone of mucus lining the pharynx, which gradually descends into the gullet, bearing attached to it the numerous nutritive particles which enter the pharynx in currents of water. An odd analogy to this mode of feeding is furnished by the small Limnea stagnalis, which, when creeping about, foot uppermost, at the surface of the water in a slightly stagnant aquarium, covered here and there with a film of resting Bacteria, is wont cautiously to curve the middle of its foot downwards a little, thus producing a shallow boat-shaped depression below the surface, but free from water. A very active secretion of mucus now commences over the whole surface of the foot; and part of the adjacent film of Bacteria is permitted to slide gently down- wards into the hollow, over the anterior edge of the foot, which is not depressed enough to let any appreciable quantity of water enter as well. When the cup is nearly full the mollusc raises its head and laps up the mixture of mucus with Bacteria, &c. in large mouthfuls. For such a purpose a tongue is necessary, a chain-saw would be unayailable. 1 Loe. ett. + Bronn’s ‘ Thier-Reich,’ Malacozoa, plate xey. 3 Gegenbaur’s Morphologisches Jahrbuch, Bd. i. OF THE ODONTOPHORE IN CERTAIN MOLLUSCA. 489 In conclusion I have only to express my warmest thanks to Professor Huxley for suggesting, encouraging, and supplying the materials for the above inquiry. Fig. Fig. bo aS oO EXPLANATION OF THE PLATES. PLATE LXXX, Figs. 1 to 9. Odontophore of Loligo, sp., natural size. . Left side: ¢, the soft spongy “ taste-organ ” with its muscles; 7, the radula; 7’, old half dissolved portion; 7s, infraradular sheet ; ed/m, its single median ex- ternal dorsal longitudinal muscle ; v/s, its left ventral longitudinal muscle ; ¢, the anterior so-called “cartilage; ” vtf, ventral transverse fibres connecting the cartilages ; e/, external longitudinal muscle of left cartilage ; g, the gullet. ‘Taste-organ and gullet removed. External longitudinal muscle of left cartilage removed, showing d¢f, dorsal transverse fibres uniting cartilages, c. Radula half pulled out, left ventral muscle of infraradular sheath cut away ; dtf and vtf, dorsal and ventral transverse fibres of cartilages cut across, showing 7s, sheath of radula, and d/m, dorsal longitudinal muscles. Median longitudinal vertical section completed*. Dorsal aspect, ed/m partly cut away. edim entirely cut away ; dtf cut open, to expose dorsal longitudinal muscles of infraradular sheet; radula half pulled out. Ventral aspect, showing @7's with its muscles, v/s; also the cartilages c, and their ventral transverse fibres vtf. . Median transverse vertical section. Figs. 10 to 23. Buccal mass of Patella vulgata, x 3. 10. Dorsal surface: pr, antero-lateral and postero-lateral protractors; 1, radula ms, rouscles of infraradular sheet. Fig. 11. Radula removed ; iis, infraradular sheet ; ms, its muscles. Fig. Fig. 12. Infraradular sheet and muscles removed, showing cartilages with their muscles ; ac, anterior cartilage ; pc, posterior cartilage ; /c, lateral cartilage ; alm, muscle uniting anterior and lateral cartilages. 13. alm removed, showing cartilaginous framework ; vér, ventral transverse fibres of anterior cartilages. See also Bronn’s ‘ Thier-Reich,’ Malacozoa, plate exyi. . 24. MR. P. GEDDES ON THE MECHANISM . Ventral aspect of buccal mass: /, lips; y, horny jaw; vpr, sheet of ventral protractor fibres. 5. Ventral protractors removed: v/s, ventral longitudinal muscles of infraradular sheet ; vt7, ventral transverse fibres; v't7’, inner ventral transverse fibres. . Radula turned forward: infraradular sheet and its muscles shown. . Infraradular sheet cut away ; 77s, its cut edge; ac, the slightly inrolled apex of anterior cartilages. . Transverse muscles all cut away, leaving framework ; vf7 and v't7", their cut edges. . Cross section of framework towards apex, showing deep sulcus. PLATE LXXXI. Left side of buccal mass in situ: f, foot; J, lips; 7, jaw; s, skin of head; vp, ventral protractor ; /pr, lateral protractor; g, gullet. Superficial muscles removed, showing left of framework. . Deep dissection, lateral cartilage removed; 77s, cut edge of infraradular sheet. . Left anterior and posterior cartilages removed, thus completing median longitudinal vertical section. Figs. 24 to 34. Buccal mass of Buccinum. (24-31, x3; 32-34, x 6.) Skin laid back, dorsal aspect ; 7, radula, seen through window cut in gullet, g ; mg, small muscles which attach gullet to skin; d#f, dorsal transverse fibres sheathing upper part of apparatus ; ms, muscle of radular sheath; d/m, other dorsal longitudinal muscles. . Gullet removed, showing cpr, circular protractor of lips, and lpr, lateral protractors. 5. Dorsal transverse fibres laid back, showing mode of attachment of radular and underlying sheet; c, the cartilages, their edges thickened by muscles; rs, radular sheath, whence pass slips of muscle to cartilages; d/m, the complex dorsal longitudinal muscles. . Radular sheath and muscles, all cut save the deepest set, dim’; v/s, ventral longitudinal muscles. PLATE LXXXIT. Ventral aspect, showing lower attachment of radula ventrally: v/s, ventral longitudinal muscles ; vf/, ventral transverse fibres ; g, depression leading into gullet. Fig. Fig. Fig. to} Fig. Fig. Fig. Fig. 30. 31. 32. 33. 34. 35. OF THE ODONTOPHORE IN CERTAIN MOLLUSCA. 491 . Longitudinal ventral muscles, v/s, cut, radula and adherent infraradular sheet laid back, showing cartilages, c; cpr, downward prolongation from circular protractor of lips. Ventral transverse fibres of cartilages (vtf) cut open, showing dim’, the deepest set of longitudinal dorsal muscles; ¢, backward prolongation of the cartilages. Cartilages completely cut apart, showing dorsal and ventral transverse fibres. Left side of apparatus in situ; letters as in figs. 26-28. Superficial muscles of left side cut away, showing left of framework ; 7s, line of attachment of radula and infraradular sheet to side of mouth, deeper ventrally ; apex therefore rocks forward and upward at 7s. Left cartilage completely removed, and median section completed. Lateral vertical section of anterior portion of Molluscan odontophore, diagrammatic. EXPLANATION OF THE COLOURS. Crrilicesmeyrmn as = cn Sens = oe Blue: Infra-radular sheet, with its dorsal and ventral muscles. . Red. Redulaeeee era i 2 ee seeds oe eee ae ellows 14 FULL 4/7 W. bl. G, Z d ee Z 0. ACC BA ost G Frand, 13 nw i 10 Tike ac 19 Hanhart Sup PGeddes del. J Smit bth. MECHANISM OF THE ODONTOPHORE IN MOLLUSCA bl. 10 GY, LAGE. tL. Soc. D oa (ze Be FL7S (2 rw i S Hanhart imp PGeddes del JSmrt hth MECHANISM OF THE ODONTOPHORE IN MOLLUSCA rant. Loot, $e. Vol 0 GY LAGE dim -- ss dla Y} rs PGeddes del. [Smit bth. Hanhart amp MECHANISM OF THE ODONTOPHORE IN MOLLUSCA [ 493 ] XIV. On the Hearts of Ceratodus, Protopterus, and Chimera, with an Account of unde- scribed Pocket Valves in the Conus arteriosus of Ceratodus and of Protopterus. By %. Ray Lanxesrer, WA., F.R.S., Professor of Zoology and Comparative Anatomy in University College, London. Received May 28th, read June 4th, 1878. [Puates LXXXITIT., LXXXIV.] INTRODUCTION. AMONG the numerous interesting structural features made known to zoologists by the admirable monograph of Dr. Giinther on the recently discovered Dipnoous form Ceratodus, none was more striking or more clearly indicative of the special zoological relationships of that remarkable animal than that exhibited by the arterial cone or basal portion of the arterial system. In describing this part of the heart of Ceratodus, Dr. Giinther says (Philosophical Transactions, 1871, p. 537):—‘“ The conus arteriosus differs from that of Lepidosiren less externally than internally. Its anterior wall is thin, though provided with a muscular stratum, which is thickest along the posterior rim of its spiral course. Its beginning is indicated by the absence of trabecule carne, the inner surface being smooth. The valvular arrangement is entirely different from that which was consi- dered to be the characteristic of the subclass Dipnoi. No valve exists on the boundary line between conus and ventricle. Before the conus turns to the left, its interior is rather spacious; but this compartment is separated from the transverse portion of the conus by a cartilaginous valve, which, from a certain view, appears as a merely papil- lary prominence, but when viewed from various sides proves to be a spiral performing a half turn, to the lower end of which a muscle is attached, which reaches down into the ventricle. This valve closes the lumen of the conus most effectually during the diastole of the heart. In the systole, the muscle attached to it contracts, and draws the valve from its position downwards, thus opening free egress to the blood out of the heart. When the muscle relaxes during the systole, the valve resumes its position through its own elasticity, shutting up the communication between the heart and arterial system. “ Beyond this valve the conus turns towards the left, and then for a very short dis- tance forwards. Quite at the end of it, and immediately before it bifurcates, there are two pairs of (ganoid) valves, narrow and rather long, with stiff non-collapsing walls, thicker along the middle than at the sides, and without tendinous chorde, in a single VoL. X.—PART x1. No. 2.—February 1st, 1879. 3X 494 PROF. RAY LANKESTER ON THE HEARTS OF transverse series. Their tunics are continued in four narrow raised stripes behind their bases. So far I have found the arrangement of this part of the heart nearly identical in two examples. On examining the first example a pair of small papillary promi- nences were found in a line between the series of stripes and the spiral valve, imme- diately in front of the latter. These stripes and papille appeared to me to represent rudiments of a second and third series of valves, analogous to the plurality of series in other ganoid fishes. Remembering, at the same time, the fact that individual variations in the development and number of valves are not of uncommon occurrence in these fishes, I examined the heart of a second (smaller, female) specimen, and had the satis- faction of finding my supposition confirmed. In this specimen there are, in the series corresponding to the pair of small papille, four valves corresponding in position to, but much smaller than, the permanent large valves which I have described above.” Further on the same distinguished zoologist observes, in speaking of the relationship of Lepidosiren and Ceratodus to one another and to the Amphibians (p. 552) :—“ The singular arrangement of the valves in the conus arteriosus is a point of much deeper interest [than the differences in the lungs and ovaries respectively]. There cannot be a question that Ceratodus should be referred to Miiller’s subclass of Ganoids, and ex- cluded from that of Dipnoi, according to the chief characteristics by which he has defined these divisions.” Finally, in giving these genera their systematic position, Dr. Giinther places them in a suborder ‘“‘ Dipnoi” of the order Ganoidei, and each in a distinct subfamily of the family “Sirenide.” The subfamily ‘“ Ceratodontina,” with the single genus Ceratodus, is defined by the following characters :—‘ Conus arteriosus with transverse series of valves. Ovaries transversely lamellated. One continuous vertical fin.” The subfamily “‘Protopterina,” with the two genera Lepidosiren and Protopterus, is thus defined :— “Conus arteriosus with two longitudinal valves. Ovaries closed sacs. One continuous vertical fin.” The quotations above given from Dr. Giinther’s admirable monograph are sufficient to show the importance attached to the valvular structures in the arterial cone of these fish-like organisms. The fact that Dr. Giinther had noticed considerable variation in the number and form of the valves in the specimens examined by him led me to make a careful examination of these parts, and of the heart generally, in one of two fine specimens of Ceratodus recently purchased for the Museum of Comparative Anatomy in University College, London. I am glad to be able to gratefully acknowledge, on the present occasion, the liberal manner in which the Council of that Institution have seconded my efforts to form a typical collection of zoological specimens for the puy- poses of class-instruction. The heart had been removed from one specimen of Ceratodus, at my request, by Mr. Ewart, the Curator of the anatomical collections; and, accordingly, I shall not now have any remarks to offer on the course and mode of entrance of the large veins CERATODUS, PROTOPTERUS, AND CHIMERA. 495 which bring blood into the heart, but shall confine myself to the structure of the heart itself, more especially of the arterial cone. Asa chief result I show that there exist in the lower part of the arterial cone of Ceratodus a numerous series of small and deli- cate “ ganoid ” or “ pocket” valves provided with “ chorde tendinez,” which valves have not been described by Dr. Giinther as occurring in the two hearts studied by him. And further, led by the observation of this “lower-end” series of valves to examine the corresponding region in the heart of Protopterus annectens (of which species the Museum of University College contains several fine examples bequeathed by the late Professor Grant), I have arrived at the interesting result that the Protopterina, as well as the Ceratodontina, possess “ ganoid” valves in the conus arteriosus, though in a rudimentary condition. Accordingly the absolute distinction between the subfamilies Ceratodontina and Protopterina, based on the presence and absence respectively of these valves, breaks down, although the spirit of the distinction, which consists in the more complete substitution of “longitudinal” valves for “ pocket” valves in Proto- pterina, as compared with Ceratodontina, remains. Lastly, I have examined the struc- ture of the heart in the fish most closely related to the Dipnoi, namely Chimera, and have given a brief account, with figures, of the general architecture of the heart in the three genera Chimera, Ceratodus, and Protopterus. I. Heart or CERATODUS. A view of the heart of Ceratodus, as seen when removed from the pericardium, with the right side facing the observer, is given in Plate LX XXIII, figure 1. The heart is chiefly remarkable for the great bulk and solidity of the conus arteriosus, from which the branchial vessels spring like a stack of chimneys. ‘The fleshy ventricle comes to a point in the middle line (here turned to the right); and from this apex a fibrous band extends to the pericardium. The auricle has much thinner walls than either ventricle or conus. It is not possible to distinguish on the surface (nor indeed interiorly) a demarcation between auricle and sinus venosus. This fact is not alluded to by Dr. Giinther, but is important, because in Protopterus and, according to Hyrtl, in Lepidosiren an internal separation of auricle and sinus can be observed and a corresponding external constriction. On the surface of fig. 1 the dotted line represents the direction of the cut which was made so as to obtain the view given in fig. 2. By this cut the cavity of the ventricle is thrown open; a flap, constituting its right wall, is thrown back. At the same time the upper part of the conus arteriosus is opened, and the cut edges of its muscular walls are divaricated so as to exhibit the valves which lie within it. The lower part of the arterial cone, which is twisted so as to lie in a different plane from the upper part, and which connects the upper part with the cavity of the ventricle, has not yet been cut into. The figures of dissections will be best understood if the draw- ing fig. 7A, representing diagrammatically the arterial cone and its longitudinal (and 3x2 496 PROF. RAY LANKESTER ON THE HEARTS OF therefore spiral) valve, be at once examined. It will be seen from that diagram that the arterial cone may be represented as a cylindrical tube which is spirally twisted on itself leiotropically—that is, in such way that in ascending the spire one would turn to the left. There is a lower vertical limb, and an upper vertical limb, and a transverse oblique segment connecting the two vertical limbs. A longitudinal muscular fold projects from the interior of the tube, and makes with it a spiral leiotropic turn. This fold is the single “longitudinal” or “ spiral valve” of the arterial cone of Ceratodus. It is not continued in that animal into the upper limb of the cone, but stops short in the transverse segment (fig. 7c). In fig. 2 the transverse portion of the arterial cone is partly exposed (though much of this cavity lies deeply behind what is there exposed), and the upper termination of the spiral valve (Sp.v) is seen. If we could follow the cavity, we should pass in the direction indicated by the dark shading behind this bit of the spiral valve, and then plunging downwards we should emerge in the cavity of the ventricle at the point indicated by the deep shadow where the dotted line termi- nates inferiorly. This dark shadow, then, indicates the lower end of the arterial cone. Its deep-lying wall is seen to form a broad arch-like flap, which is tough, though membranous (AVV). Immediately below this flap a second dark shadow indicates another recess, which is, in fact, the passage from ventricle to auricle. This arch-like flap, which is the inferior internal wall of the arterial cone, may therefore be regarded as the auriculo-ventricular valve. It has been fully recognized by Dr. Giinther; and an exactly similar arch-like valve is described by Owen in Protopterus and by Hyrtl in Lepidosiren (see figs. 10 & 11). The cut flap of the ventricular wall, which is reflected to the right, shows chiefly cut muscular fibre on its surface; but the more central part is the natural right wall of the ventricular cavity, and is marked by a meshwork of trabecule carnee. The floor of the ventricular cavity, as exposed in the dissection (fig. 2), is seen to present a smooth convex prominence (FC), which, taking its origin in the muscular substance at the apex of the ventricular cavity, passes as a solid ridge beneath the auriculo-ventricular archway. Its actual form and extension are best seen in fig. 4. This remarkable structure has been recognized by Hyrtl in the heart of Lepidosiren, and by Giinther in Ceratodus and Protopterus. It is no doubt efficient in causing a closure of the auriculo-ventricular passage, being brought up against the arch-like flap or curtain during the ventricular systole. Of its extension into the auricle I shall speak in describing the further dissection exhibited in fig. 4. The dotted line running in figure 2 from the front edge of the spiral valve across the floor of the transverse part of the arterial cone, and then across the cut muscular sur- face of the cone to its lower opening, indicates the next step in the dissection. A cut being made as indicated by that line, we obtain the appearance drawn in fig. 3. In fig. 3 the lower limb of the arterial cone is opened, and the transverse portion (at the left-hand side of the figure) is more widely exposed than in fig. 2. The lower CERATODUS, PROTOPTERUS, AND CHIMARA. 497 segment of the longitudinal (spiral) valve is seen projecting from the left-hand side of the divided wall of the lower part of the conus (compare fig. 7,Aandc). ‘The wall of the lower part of the conus is continued downwards to form the arch-like auriculo- ventricular valve. The free depending portion has a perfectly smooth surface ; but this is not the case higher up. In fact the wall of this lower segment of the arterial cone is richly provided with peculiar pocket valves, which have not hitherto been described. These valves are so delicate that they might easily escape detection, or, in a specimen which was less excellently preserved than the one now examined, might be destroyed. An enlarged view of this portion of the wall of the arterial cone, pinned out flat and advantageously illuminated for the purpose of showing the valves in question, is given in fig. 5. It will be observed that there are three rows of these valves, of which the lower is the most complete, exhibiting as many as eight separate flaps, whilst these, and more especially the flaps of the middle series, exhibit a tendency to reduplication by transverse division. The three rows of valves correspond to three folds on the concave surface of the spiral longitudinal valve. The valves thus disposed consist of delicate membranous flaps (some as large again as, others smaller than, the area of a common pin’s head), which are fixed to the wall of the cone by their lower borders ; and most of them hang freely by the other three sides, except for the presence of more or less numerous chord tendinew, which attach them loosely at various points (see fiz. 6). Some of the larger valves are attached by the two sides right and left, as well as by the base, so as to form regular watch-pocket valves like those in the upper limb of the cone. The delicate and diminutive nature of some of these valves suggests very forcibly that we have here the remnant of an ancestral condition in which more powerful pocket valves were present throughout the extent of the conus, and that these have dwindled in proportion as the longitudinal spiral valve, characteristic of Dipnoi, has developed. That this is the case is further pointed to by the fact that in Ceratodus pocket valves are still strongly developed in the upper part of the conus, into which the longitudinal valve does not reach, whilst, on the other hand, in Lepidosiren and Protopterus, where the longitudinal valve does extend into the upper vertical limb of the conus (fig. 7B), no pocket valves are present in that region. Seeing that the delicate valves of the lower part of the conus of Ceratodus were such as to escape attention, it occurred to me as possible that similar delicate pocket valves might exist in the corresponding part of the conus of Protopterus; and accord- ingly I searched for them in that position and found them (fig. 12). Before leaving the heart of Ceratodus and passing on to that of Protopterus, there are three additional points which a comparison of my own notes with Dr. Giinther’s description leads me to consider as worthy of mention. First, as to the number and form of the pocket valves in the upper vertical segment of the cone in my specimen. Dr. Giinther examined the hearts of two specimens of Ceratodus; and, knowing the 498 PROF. RAY LANKESTER ON THE HEARTS OF tendency to individual variation in the number and character of the pocket valves of the cone of Selachians and Ganoids, he was prepared to find that his two specimens differed in the disposition of these parts. In the one there was found an upper row of four large watch-pocket valves (close to the origins of the branchial vessels), each thickened at its base by a vertical pad, such as Gegenbaur describes in Acanthias ; and below this series were observed two papille placed side by side. ‘These Dr. Giinther took to be rudimentary indications of a second row of pocket valves. His second spe- cimen proved this supposition to be correct; for in it he found the upper row of four pocket valves with their pads, and a second complete row of four small pocket valves in place of the two papille. The specimen examined by me has the valves in this part of the cone more strongly developed than was the case even with Dr. Giinther’s second specimen. The upper series of four large pocket valves are present, and possess thickened pads which appear as the bases of the pockets, not as separated tubercles such as are drawn in Dr. Giinther’s plate. The second series of four pocket valves is complete, and the valves are relatively larger than in the second specimen figured by Dr. Giinther. It would seem possible, from the amount of difference between my specimen and Dr. Giinther’s first specimen, that a considerable range of valve-development obtains in the Ceratodus cone ; and it is even possible that the delicate valves of the lower vertical segment of the cone do not exist, excepting in those specimens which are, so to speak, highly valvular. The second point is one on which I venture to differ from Dr. Giinther. It appears to me that the longitudinal fold in the conus, both of Ceratodus and Protopterus, is not to be regarded as consisting of separate parts, viz. valve and muscle, but that the whole fold is a muscular contractile ridge which acts as a valve in consequence of the spiral disposition of the tube in which it exists. Dr. Giinther, however, describes that part of the longitudinal valve which lies in the lower vertical limb of the conus as a muscle, whilst only the twisted portion running into the transverse portion of the conus is regarded by him as “valve.” Functionally this may be true; morphologically the division is, I think, inadmissible. The third structural feature to which I will draw attention is the remarkable “‘fibro-cartilaginous mass” first described by Hyrtl in Lepidosiren, which, originating in the floor of the ventricle, extends through the auriculo-ventricular passage, expands into a large knob in the floor of the auricle, and can then be traced further as a dense fibrous ridge right across the floor of the sinu-auricular cavity. This structure, and the fact of the complete continuity of “ sinus” and “ auricle,” are exhibited in the dissection fig. 4 (FC). A dotted line on the ventricular wall in fig. 3 indicates the direction of the deep cut which has now been made, the point of the scissors being passed completely through the auriculo-ventricular passage. Dr. Giinther speaks of the sinus and auricle as distinct structures, the former open- CERATODUS, PROTOPTERUS, AND CHIMARA. 499 ing into the latter without a valve. The fact is that they form one cavity, not even divided by a constriction ; it is all the more necessary to insist upon this confluence of the two cavities, because in Lepidosiren and Protopterus a sinus is distinctly marked off from the auricle (fig. 9), though there are no proper valves between sinus and auricle. The form and extension of the basal fibro-cartilage, as we may call the body marked FC in fig. 4 in Ceratodus on the one hand, and Protopterus on the other, are correlated with the suppression and expression of the auricular antechamber. The general features of this remarkable body in Ceratodus are accurately described by Dr. Giinther, who also notes the fact that a corresponding structure exists in Protopterus. Owen, in his classical account of that fish, does not mention it. Hyrtl carefully described the basal fibro-cartilage of Lepidosiren, comparing its action, in relation to that of the hanging curtain valve or archway, with that of a piston and fan valve, whilst he sug- gested that it should be considered an imperfect septum ventriculi. There is not, I think, at the present time, any ground for regarding this structure as the forerunner of the septum ventriculi; but Hyrtl was apparently led to regard it as having the nature of a vertical cardiac “septum” from the fact that in Lepidosiren the auricular end of this mass gives attachment to muscular trabecule, which, according to that observer, incompletely divide the auricle into two chambers, a right and a left, of which the left receives only the pulmonary vein. Hyrtl gives no adequate figure of these parts, whilst both Bischoff and Owen are silent concerning it in Lepidosiren and Protopterus respectively. Dr. Giinther, in his description of the basal fibro-cartilage of Ceratodus, mentions the existence of one in Protopterus. In fig. 10 and fig. 11 the heart of Protopterus is represented in such a way that in the former the observer has the auricular termination of the cartilage (FC) facing him, whilst its ventricular origin is exhibited in the latter. The difference between the auricular termination of this structure in Protopterus on the one hand, and Ceratodus on the other (as exhibited in fig. 4), is of some interest. I have cited above Hyrtl’s observation, that in Lepidosiren the incomplete trabecular septum of the auricle is inserted into this fibro-cartilage, which he regards as an incomplete septum of the ventricle. Compare with this the condition in Protopterus. In Protopterus there is no indication of an auricular septum ; the broad knob-like end of the basal fibro-carti- lage projects into the auricle, and there ends in a point, giving attachment inferiorly on each side to a membranous fold (fig. 10, Sav), the two folds thus forming a feebly developed valvular separation between sinus and auricle. In Lepidosiren, according to Hyrtl, the sinus opens separately into the right and the left divisions of the auricle without valves in either case. On the other hand, in Ceratodus, where, as above re- marked, there is not only no division of the auricle, but no separation between sinus and auricle, we find the knob-like swelling of the basal fibro-cartilage (fig. 4, #C) occu- pying the same position as in Protopterus (that is, projecting from beneath the auriculo- 500 PROF. RAY LANKESTER ON THE HEARTS OF ventricular curtain into the auricle); but instead of ending abruptly in the auricle, it is continued across the sinu-auricular floor in the form of a strong fibrous ridge, and can actually be traced into the tough fibrous walls of the large vena cava, which, sur- rounded by liver, opens mesially into the hinder part of the sinu-auricular cavity. It would thus seem that the basal fibro-cartilage, ‘‘ eine Vorrichtung, die noch bei keinem Wirbelthiere beobachtet wiirde,” as Hyrtl remarked when first describing it in Lepidosiren, has its most extensive development in Ceratodus (it is shown throughout its course in fig. 4), and its smallest in Protopterus; whilst in Lepidosiren it acquires a special relation to muscular trabecule of the auricle, which gives it a share in the formation of the imperfect auricular septum present in that alone of the three genera of Dipnoi. I]. Heart or PRroroprerus (LEPIDOSIREN) ANNECTENS. Professor Owen, in the description of the anatomy of this form, has not given special illustrations of the general appearance and interior structure of the heart ; and, I believe, it has not been figured by any later writer. In a monograph which contained so much that was new and of the greatest importance, it is not surprising that the smaller details of one organ should be passed over. ‘The main features of the structure of the heart (excepting the basilar fibro-cartilage) were briefly described by Owen ; his atten- tion was especially given to the structure of the conus arteriosus, concerning which he notes the absence of watch-pocket valves, the spirally-twisted shape, and the ane of longitudinal folds or valves. : In figs. 8, 9, 10, 11, 12 I have given drawings of the entire heart of Protopterus! and of dissections. In figure 8 a front view of the heart removed from the pericardium is given. It is more symmetrical than that of Ceratodus, and differs most markedly from the latter in the presence of large auricular appendages, and in the constriction of the neck of the arterial cone. A fibrous band connects the apex of the ventricle with the pericardium, as in Ceratodus. On turning the same heart we obtain a more complete view of auricle and sinus (fig. 9). The sinus is elongated posteriorly, and has tough fibrous walls where it becomes con- tinuous with the large median vena cava. ‘Towards the auricle its walls become much more delicate, and are loosely folded in the collapsed condition. Externally there is no marked constriction between the sinus and the median portion of the auricle. The lateral expansions of the auricle are very voluminous, and are vesicular or even digi- tate. ‘Chey extend around the ventricle, and fit with their upper processes into the ‘I must point out here that the observations recorded in the present paper on the hearts of Ceratodus, Protopterus, and Chimera have, owing to special circumstances, been made on eacised hearts. I have not, therefore, in any ease entered into the question of the number and position of the openings of vessels into the auricular sinus, nor of those from the conus. CERATODUS, PROTOPTERUS, AND CHIMARA. 501 constriction between conus and ventricle. The right auricular appendage presents a strongly marked additional lobe, which reaches far forwards (fig. 9, 7) and renders the auricle asymmetrical. The dotted line on the surface of fig. 9 indicates the direction of the cut by which the appearances in fig. 10 were obtained. The sinus and median portion of the auricle have been opened by a cut along the median dorsal line, and the cut edges pinned back (fig. 10). The floor of the sinus is seen to be fibrous and tough. In front the walls become more delicate, and are turned in so as to present two folds (Sav), which are inserted into the fibro-cartilage (FC). These folds indicate the separation between sinus and auricle, and serve as valves. The head of the great basilar fibro-cartilage (FC) is seen projecting from beneath the auriculo-ventricular archway (AVV) into the auricle. It terminates in a point, from which the septal folds take origin. The dotted line on fig. 8 indicates the direction of a cut by which the whole length of the ventricle and conus arteriosus was divided so as fully to expose their cavities, the cut edges being pinned back. ‘The resulting appearances are given in fig. 11, which presents a view of the heart of Protopterus very closely similar to that of the heart of Ceratodus given in fig. 3. ‘The chief difference depends upon the fact that the upper limb of the conus of the Protopterus heart has been completely bisected. The very small ventricular cavity and the very thick ventricular walls are remarkable; arising from the floor of the ventricular cavity we have the ventricular extremity of the basal fibro-cartilage (FC). We here see it disappearing beneath the auriculo-ventricular curtain or archway, just as (in fig. 10) we see it emerging from beneath the same screen on the auricular side. The auriculo-ventricular curtain valve is seen to be the extended inferior margin of the conus arteriosus, as it is in Ceratodus, and can be traced on the right-hand side of the figure into trabeculae carnee—with which compare the similar disposition in Ceratodus (fig. 3). On the left-hand side of the lower segment of the conus arteriosus, as thus exposed, we see a broad fleshy fold which has a free slightly incurved lower margin. This is the longitudinal valve. The arrow indicates the hidden course of the transverse segment of the arterial cone. ‘This region is indicated in the surface-view of the heart by the dilated region above the constriction marking off cone from ventricle. The cone turns here with a sharp curve to the left, and then resumes its vertical course (see fig. 74). The longitudinal valve does not, however, terminate in this region as in Ceratodus, but is continued along the upper vertical segment of the cone (fig. 7B, and fig. 11, Sp.v). In this upper region only it is accompanied by a second longitudinal production of the wall of the conus (fig. 10, w/v), which is of a denser substance, and is more probably to be regarded as a septum than as a muscular valve, such as the longer spirally-turning fold certainly is. No trace of “ganoid” valves, such as those of Ceratodus, are to be found in the upper segment of the heart of Pro- topterus. The continuation of the spiral valve and the second septal fold entirely replace them. VOL. X.—PART XI. No. 3.—February 1st, 1879. 3Y 502 PROF. RAY LANKESTER ON THE HEARTS OF But in the lower segment of the cone of Protopterus, as I have mentioned above, I have found minute pocket valves in the same position as the delicate “ganoid” or pocket valves which I discovered in Ceratodus. 'The position of these valves is indi- cated at vv in fig. 11. A magnified view of the valves is given in figure 12. They are so exceedingly small as to be functionally of very little, if any, significance, and may be regarded as evanescent organs. Just asin Ceratodus, so here do we find the under surface of the free margin of the longitudinal spiral valve (fig. 12, Sp v) thrown into transverse folds corresponding with these valves. In fact these folds must be regarded, both in Cera- todus and Protopterus, as indications of pocket valves arranged in transverse series on the concave face of the outstanding longitudinal spiral valve. Though so minute, the pocket valves of the conus of the Protopterus heart can be lifted with the point of a fine needle; and the fact that they are free in front, and attached basally, can be demonstrated. I did not observe chorde tendinee in connexion with these valves, I have as yet only examined two hearts of Protopterus annectens ; and in the better-pre- served specimen I found the valves as figured in fig. 12. In the second specimen I could not detect them, and am inclined to ascribe their absence to the inferior state of conservation of the heart, which was obvious enough. III. Heart or CHIM&RA MONSTROSA. The Holocephali are those fish which assuredly come nearest to the Dipnoi of all living forms. The resemblances in some small points are quite striking. Beyond the important points of identity in the structure of the skull and spiral column, we have:—the remarkable scissor-like teeth, closely agreeing in the two forms; the position of the fold of the nares in Chimera, and of the anterior and posterior nares in the Dipnoi; the short oval mass of intestine in both groups, really rendered of considerable length by the close-fitting windings of the spiral chamber within; the detailed agreement of the urino-genital organs of both sexes respectively ; and, perhaps most strikingly, the identity in Chimwra and the Protopterina of the lines of mucous glands disposed on the head—an agreement the closeness of which was first pointed out by Hyrtl. I am not aware that there is any detailed account of the heart of Chimera or of Callorhynchus extant, and I therefore turned to the examination of a specimen in my possession with considerable interest. The result has been disappointing, so far as the prospect of finding new points of agreement between Chimera and the Dipnoi is con- cerned. The heart of Chimera is widely separated from that of Dipnoi, is, in fact, eminently ichthyic and Selachian. The heart is not one of those organs with “a charmed life,” the characteristic structure of which remains unaffected by, indifferent to, the immensely important physiological changes passed through on the way from Chimera to Protopterus. Protopterus, with its mud-lurking habits and air-breathing capacity, has lost, or never developed, the Selachian limb, and has acquired a large functional CERATODUS, PROTOPTERUS, AND CHIMZ#RA. 508 lung, the representative of which is absorbed in Chimera. With these changed conditions of other organs the heart and the brain have varied concomitantly, without our being able to say exactly of what value to the organism the more amphibian brain and the more amphibian heart of Protopterus, as compared with those of Chimera, may be. In figure 13 is given a surface-view of the heart of Chimera monstrosa, as seen from the front, the heart being freed from pericardium. The heart has the usual form and proportions of that organ in Selachians—an obtuse broad ventricle free at the apex, voluminous auricle, and a straight cylindrical cone. The ventricle and cone were opened by the removal of a slice not quite sufficiently large to bisect these structures. The appearance with the slice removed is given in fig. 14. The conus was found to be perfectly straight, leading into a relatively much larger ventricular cavity than we find in the Dipnoi. In fact there seems little room for doubt that the cavity of the ventricle between the auriculo-ventricular valve and the base of the conus, as delineated in Chimera, by the change in the character of the walls represents genetically the lower vertical and the transverse segments of the arterial cone of Dipnoi. I am inclined to think that it is only the upper vertical segment of the cone of Dipnoi which should be compared with the outstanding “ cone” of the Chimeera’s heart. And when these two portions are compared in Ceratodus and Chimera the resemblance is found to be exact. Chimera (the specimen at least examined by me) has two transverse rows of watch- pocket valves in the conus, four in each row, precisely as in the upper limb of my specimen of Ceratodus. The agreement extends to the longitudinal pads which exist below the uppermost or larger series of valves, but not below the lower and smaller row. In fig. 14 only one complete and two cut valves in each row are seen, the rest of the wall of the cone and ventricle having been taken away altogther, and not merely pinned back (as in the dissections of Ceratodus). If, then, we may regard that part of the ventricular cavity to the left in my drawing as what in the Dipnoi becomes drawn up out of the heart and developed as an addition to the existing cone, we may look in that part of the cavity to the extreme right (the animal’s left) for indications of parts corresponding to the auriculo-ventricular curtain valve and the basilar fibro-cartilage. I think that it is possible to recognize these in the com- ponents of the tricuspid valve by which the auricle and ventricle of the Chimera’s heart communicate. In order to understand the form of this valve it must be looked at first of all from the auricular side ; and it is so seen in fig. 15, A, AVV. A hole like a leech-bite is there seen leading from auricle to ventricle. Three pads are placed between the three arms of the opening. One of these pads(AVV) is more prominent and membranous than the other two. It may be taken to represent the auriculo-ventricular curtain valve of Dipnoi. Of the remaining two pieces one is larger and firmer than the other, and 3x 2 504 PROF. RAY LANKESTER ON THE HEARTS OF when traced into the ventricular cavity is found to form a considerable lobe there (A in figs. 14 and 15). This possibly is the representative of the great basilar fibro-cartilage of Dipnoi, whilst the third pad is relatively small, and might be expected to disappear were the larger one (A) to encroach upon it. I do not attach any importance to these hypothetical identifications as showing affi- nities between Chimera and the Dipnoi, but rather I must insist on the fact that there is so little community of structure. The auricle of Chimera is not sacculated like that of Protopterus, but has a very distinct pinched-off sinus (fig. 15) which opens into the auricle by a limited slit-like orifice protected by valvular folds (Sav). It is of some interest in connexion with the development of an auricular septum in this and in that animal to notice that in Chimera the auricle is incompletely divided by a septal ridge into two compartments, into one of which the sinus opens, whilst from the other leads the small tricuspid auriculo-ventricular passage. DESCRIPTION OF THE PLATES. PLATE LXXXIII. Fig. 1. Heart of Ceratodus Forsteri, right lateral view, natural size: BA, branchial arteries; Per, remnant of pericardium; AC, arterial cone; Aw, auricle; Vn, ventricle. Fig. 2. Heart of Ceratodus: the ventricle and upper limb of the arterial cone have been opened and the cut edges pinned back: Sp v, spiral valve; FC, basal fibro-cartilage ; other letters as in fig. 1. Fig. 3. The same heart further dissected, the lower limb of the conus being now opened by a cut indicated by the dotted line in fig. 2. AVV, auriculo-ven- tricular curtain valve ; vv, situation of the small pocket valves; other letters as in figs. 1 & 2. Fig. 4. The same heart further dissected by a cut passing along the dotted line marked in fig. 3. Iw, liver; VC, vena cava; other letters as in preceding figures. Fig. 5. Inner wall of lower limb of arterial cone spread out flat, and enlarged so as to show the undescribed pocket valves. Sp. V, spiral valve. Fig. 6. A single valve from fig. 5, enlarged so as to show the chorde tendinez. Fig. 7. Diagrams of the arterial cone and its longitudinal spiral valve. A, the cone itself; the upper and transverse limbs have their anterior walls removed. The dotted line, ss’, indicates the line of origin of the longitudinal ridge or “spiral” valve. B, spiral valve of Protopterus; the shaded border (ss’) is that which is attached to the wall of the conus. C, the shorter spiral valve of Ceratodus; ss', attached border. or CERATODUS, PROTOPTERUS, AND CHIMERA. 50 PLATE LXXXIV. Fig. 8. Heart of Protopterus annectens, x 4 linear, seen from the front. CA, arterial cone; Au, auriclar appendages ; x, additional lobe of right side of auricle , Vn, ventricle ; Per, attachment to pericardium. The right appendage of the auricle is thrown back so as to expose the constricted neck of the conus. Fig. 9. The same heart, back view. S/, sinus venosus ; ve, ve’, apertures of veins enter- Fig. 10. Fig. 11. Fig. 12. Fig. 13. Fig. 14. Fig. 15. vou. X.—ParT x1. No. 4.—February 1st, 1879. ing the sinus; other letters as in fig. 8. The same heart, back view, with sinus and auricle opened by a median slit. AVY, auriculo-ventricular curtain valve; FC, basal fibro-cartilage ; Sav, incom- plete sinu-auricular valves ; other letters as in fig. 9. Heart of Protopterus opened from the front by a median vertical cut as indi- cated by the dotted line in fig. 8. CA, arterial cone; Aw, auricle ; Vn, ven- tricle ; FC, basal fibro-cartilage ; vv, rudimentary pocket valves; Sp v, spiral valve ; ulv, second long valve of the upper limb of the conus. View of the lower border of the arterial cone, showing the pocket valves (greatly enlarged). Heart of Chimera monstrosa, front view. CA, arterial cone; Vn, ventricle; Au, auricle. The same heart with the anterior half of the conus and ventricle entirely removed. (A, conus arteriosus; B, limit of arterial cone corresponding to the lower limit of the upper vertical limb of the Dipnoous arterial cone ; AVY, auriculo-ventricular valve; A, fleshy pad; Aw, auricle; Vn, ventricle. The same heart seen from behind, with the auricle opened. Sav, sinu-auricular valve; Si, sinus venosus ; other letters as in fig. 14. 3 y a 7 Qe Q (Pl 7 VVVIUT Dns Geol be OL 0 GLI STH} Li 400, COL 2 LAAALM HRay Larkester ad nat del Smt kth Hanrart 119; HEART OF CERATODUS ERa tyLankester, adnat del HEART OF PROTOPTERUS ¢€ RF USL ans. Loot $e CHIMARA \ Smit hth Hanbart imp [ 507 ] XV. Observations on the Uraniide, a Family of Lepidopterous Insects, with a Synopsis of the Family and « Monograph of Coronidia, one of the Genera of which it is composed. By J.O. Wxstwoop, M.A., F.LS., &e. (Piates LXXXV.-LXXXVIII.) Received March 18th, read April 16th, 1878. IT is now more than forty years ago since there appeared in the Transactions of this Society a memoir by Mr. W. S. MacLeay, and a notice in the ‘ Annales’ of the French Entomological Society by M. Boisduyal, in which were first made known the trans- formations of two of the most splendid of Lepidopterous insects. These insects had, up to that time, been regarded by most writers as butterflies, but were proved, by the details then made known, to belong to the Heterocerous division of the order, although their day-flying habits, and the extraordinary brilliancy of their colours, had naturally led to their having been considered as belonging to the Rhopalocera or true butterflies. M. Boisduval has well described one of these insects as ‘‘ ce magnifique Lépidoptere, le plus beau de la création.” Hence, as well as in consequence of the singular manner in which systematic writers on the order haye treated the position of the different members of the group to which these brilliant insects belong, and their interesting metamorphoses, it will not be considered irrelevant to the special subject of this me- moir to enter into some details upon the subject, more especially as some very difficult questions as to the rules of nomenclature are involved in the inquiry. Amongst the species of his great genus Papilio, containing the whole of the day- flying Lepidoptera, Linnzus introduced Papilio leilus, P. orontes, P. patroclus, and P. lunus, to which were added in the last century Papilio rhipheus by Drury, and P. sloanus and P. empedocles by Cramer. Another species belonging to this group was added by Cramer, but regarded by him as a moth, under the name of Phalena orithea. In 1807 there appeared in the sixth volume of Illiger’s Magazine a posthumous sketch of the proposed division of the Lepidoptera into genera by Fabricius, who had previous to his death published his separate works on the Coleoptera (Hleutherata, F.), Hymenoptera (Piezata, F.), Diptera (Antliata, ¥.), and Hemiptera (Rhyngota, F.), in which each of these orders of insects had been cut up into very numerous genera. In this sketch of the Lepidoptera’ Fabricius placed at the head of the order (followed by the other numerous genera of butterflies) his new genus Urania, shortly charac- 1 Mr. J. G. Children published an English abstract of the proposed system of Fabricius in the ‘ Philosophical Magazine and Annals’ for February 1830. Vou. X.—ParT xl. No. 1.—Jumne 1st, 1879. 4a 508 PROF. J. 0. WESTWOOD ON THE URANIIDA. terized, with P. leilus and P. patroclus mentioned as two of the seven species of which it was stated to be composed, the names of the other five species not being given. The name Urania proposed by Fabricius for this new genus had unfortunately been, then recently, applied by Schreber to a genus of plants of the natural order Musacez, from Madagascar, to which Adanson, Sonnerat, and Jussieu had previously given the name of Ravenala, which the tree also bears in Madagascar. It is true that the name Urania is still “ unjustifiably”’' retained for the tree in botanical works; but surely it ought to give way to Ravenala, in which case Urania would be free for use in ento- mology”. It moreover does not appear to me objectionable to employ a generic name for a group of objects in one kingdom of nature which has already been, and is even still in use in a different kingdom. I should therefore not hesitate on both these grounds to retain the generic name of Urania in entomology, and to consider the first species named by Fabricius (P. /eilus) as its type, that particular species being a good repre- sentative of the group of new species which have subsequently been described as most closely allied to P. leilus. In like manner, as P. lei/us and the other new species allied to it are good representatives of the whole group, I did not hesitate in my new edition of Drury’s work on Exotic Insects, nor do I now, to consider Urania as the type of the family to which to which I applied and still retain the name of Uraniide, it having in more recent times been deemed advisable to split up the genus Urania into smaller genera. In 1816, Hiibner, evidently profiting by the publication of the Fabrician system in Illiger’s Magazine, retained the previously mentioned species in a single subdivision of his great phalanx Geometre, forming them into the first family Heroice of his 4th geometrideous stirps, Lares, and subdividing them as follows in his ‘ Verzeichniss,’ pp. 289, 290:— Coitus 1. Larunpz: LARUNDA, Orithearia (Orithea, Cram. 262. C, D). Coitus 2. Lyssm: LYSSA, Achillaria (Patroclus, Cram. 198. A), and Patroclaria (Patroclus, Linn., Cram. 109. A, B). Coitus 8. Aucipes: ALCIDIS, Orontiaria (Orontes, Linn., Cram. 83. A, B). Coitus 4. Curystrip1z: CHRYSIRIDIA, Riphearia (Ripheus, Cram. 385. A, B). Coitus 5. Uranra: URANIA, Sloanaria (Sloanus, Cram. 85. E,F); U. leilaria (Leilus, Linn., Cram. 85. C,D). Coitus 6. Maniax: MANIA, Empedoclaria (Empedocles, Cram. 199. A, B); M. candilunaria (Lunus, Linn., Cram. 200. A) ; M. lunigeraria (Lunus, Cram. 200. B,C). These six subdivisions form so many genera which it is desirable still to retain. ' To use the phrase of Loudon, Encycl. of Plants, p. 245. 2 M. Boisduval (Nouy. Ann. du Mus. ii. p. 260) perceived the inconvenience of the same name being em- ployed in entomology and botany, but considered that the long usage of the name Urania in entomology out- weighed such inconvenience. ‘Si le nom de ce genre n’étoit adopté depuis long-temps par la plupart des entomologistes, il seroit conyenable de le changer parce qu'il existe déja un genre des plantes appelé Urania.” PROF. J. O. WESTWOOD ON THE URANIIDZ. 509 It has happened, however, unfortunately for these names of Hiibner that nearly all of them have given rise to disputes upon points of nomenclature. ‘The name of Hiibner’s first group, Larunda had been used by Dr. Leach in 1814-15 for a genus of Crustacea, to which the name of Cyamus had been given by Lamarck in 1801, and adopted by Latreille (as quoted by Leach himself in Linn. Trans. xi. p. 364, where the name Larunda was proposed). It is still undecided whether a name which has been proved to be a synonym can subsequently be used for a different genus in a different group. ‘The difficulty here might have been solved by the employment of Latreille’s generic name Coronis (proposed in the ‘ Familles Naturelles’ in 1825, and in the second edition of the ‘ Régne Animal’ in 1829) for the Larunde of Hiibner, only that it still more unfortunately happened—Ist, that the name Coronis itself had been proposed by Hiibner in his ‘ Verzeichniss,’ 1816, p. 265, for a totally different genus belonging to the Noctuide (Phal. stolliz, Cram. pl. 310. figs. A, B); 2ndly, that Coronis had also been actually proposed by Latreille himself for a genus of Squillideous Crustacea, to which it is still applied, in the work in which he also gave it to the Lepidopterous genus in question; and, Srdly, that in 1827 Coronis had been given to a genus of passerine birds by Gloger, to which, however, the name of Gymnoderus had been given in 1809 by Geoffroy. This is the more annoying because for nearly fifty years the generic name of Coronits has been universally applied to Phalena orithea, Cramer, and its congeners by all entomologists, except Dr. Felder, who in 1875 used the objection- able Hiibnerian name Larunda, but with reservation. Although, as I have said above, I do not consider it unadvisable to employ the same generic name in two different kingdoms of nature, I can scarcely go so far as to think it proper to use the same name for a genus of birds, a genus of Crustacea, and a genus of Noctuideous moths (for if used in Lepidoptera, Hiibner’s appropriation of the name has the priority). It has, I know, been proposed recently to treat the ‘ Verzeichniss’ of Hiibner as a nullity; but I cannot agree to the proposal. In many respects, indeed, this work is most unsatisfactory; but where the author has made (as he has often done) good arrangements of the multitudinous species of Lepi- doptera, which had up to his days been in a chaotic state of confusion (either as regards their family distribution or the juxtaposition of the different species), I think it is an act of justice to give him credit for his work. To prevent further confusion I therefore propose in this memoir to employ the name Coronis in a slightly modified form, Coronidia. Hiibner’s second name, Lyssa, is also liable to the objection that in 1815 Dr. Leach had used Lissa for a genus of Crustacea, for which it is still retained. The same name was also used generically in Diptera by Meigen in 1826. ‘To avoid this confusion, I propose to modify Hiibner’s name into Lyssidia. The name Alcides, proposed by Dalman in 1826 for a genus of Coleopterous weevils, is, of course, posterior to the employment of Alcidis by Hiibner in 1816; I propose, VOL. X.—PART x11. No, 2.—June 1st, 1879. 4B 510 PROF. J. O. WESTWOOD ON THE URANIIDA. therefore, to retain Hiibner’s name, with a slightly altered termination, for uniformity, into Alcidia. The name Chrysiridia, given by Hiibner to P. ripheus, is a happy one, and may be retained for that brilliant insect; but it has unfortunately happened for Hiibner’s last name, Mania, that it has been employed by Treitschke for a genus of Noctuide, for which it is still retained by Guenée and H. Doubleday ; I therefore have the less hesi- tation in altering Hiibner’s name into Manidia, and in rejecting the subsequent name given to it of Sematura. In 1825 the Swedish naturalist Dalman, in his monograph of Castnia, neglected the arrangements of Fabricius and Hiibner, blaming the former for inserting P. orontes, Linn., in the genus Castnia, with the observation “ que vero species minime hujus generis sed ad NycrantpEAs nostras pertinet, etsi transitum memorabilem Papilionides inter et Nyctalideas bene offert. Genera huc pertinentia sequenti modo distribuere tentavimus : “Cypimon; Ripheus, Sloaneus, Leilus et forte Lavinia Fabr.=Gen. Urania Fabr. Latr.! NyctaLtemon: Orontes, Patroclus. Semarura: Lunus, Afgistus Fabr., Empedocles Herbst.” To which he adds Thysania with agrippina, odora, &c., as types, evidently identical with the Noctuideous genus Hrebus, and also doubtfully the genus Agarista of Leach (Zool. Misc. i. tab. xv.). It will be seen that the genus Cydimon of Dalman is equivalent to the Chrysiridia and Urania of Hiibner, that Nyctalemon, Dalm.,= the Lysse and Alcides of Hiibner, and that Sematura, Dalm., = Mania of Hiibner. In 1840 (Introd. Mod. Class. Ins. ii. p. 369) I separated these insects into a single family placed between the Sphingide and Anthroceride. In 1854 Mr. Walker, in his Catalogue of the Collection of the British Museum (Lepidoptera Heterocera, Pt. 1), adopted the genera Urania for Leilus and Rhipheus, Nyctalemon for Orontes and Patroclus, to which he added Lunus and its allies (gen. Sematura, Dalman), and Coronis; which last, notwithstanding its intimate relation- ship with Sematura, Lunus, &c., he placed in the family Castnii, as Latreille had done. Lastly, in 1857, M. A. Guenée (Hist. Nat. Ins. Lep. tom. ix.) has arranged all the preceding genera into one group, Uranides, at the head of his Phalénites (Geometra, Linn.), divided into four families : . Cyprmonip&: Gen. Cydimon (Leilus &e.). Uranipm: Gen. Urania (Rhipheus). Nycraramonip&: Gen. Alecidis (Orontes) and Nyctalemon (Patroclus &c.). Sematunipe: Gen. Sematura (Lunus &c.) and Coronis (Orithea &c.). mo we * “Obs. Uranive nomen pro insectorum genere nullo modo admittendum, etenim jamdiu plantarum genus, notissimum quidem, sic vocavum.” PROF. J. O. WESTWOOD ON THE URANIIDA. 511 Adopting the opinion of M. Boisduval, that the group before us will not enter into any of our generally received families, and that it is “ une de ces créations a part, qui envoie 4 la fois un rameau vers plusieurs groupes, mais que l’on ne peut faire entrer convénablement dans aucun” (Mon. Agarist. p. 7, extr. Rev. et Mag. de Zoologie), we must now investigate the natural relationships of this most interesting group of insects, which have been alternately regarded as butterflies and moths. Thus Linneus regarded the more typical species as butterflies, and Fabricius even placed them at the head of the day-flying genera. Dalman, as we have already seen, considered that P. orontes formed the transition between the Papilionide and the other Uraniide. Latreille (Gen. Crust. et Ins. iv. p. 207) gave Urania and Hesperia as the two terminal genera of the Diurna, and Castnia (into which he introduced P. orontes) as the first genus of the Sphingides. The same arrangement was employed in the 9th volume of the ‘ Encyclopédie méthodique,’ where Godart divided the genus Urania into four groups:—A, Ripheus; B, Sloaneus and Leilus; C, Orontes and Patroclus ; D, Lunus and Empedocles. In 1825 (Fam. Nat. du Regne An. p. 470), and in 1829 (Régne An. 2nd edit. iv. p. 387), Hesperia and Urania are still given by Latreille as two genera of Hesperiide, and Castnia, Coronis, and Agarista of Leach as forming the first tribe (Hesperi-Sphinges) at the head of the Sphingide. This arrangement continued unchanged by Latreille to the last, and was adopted by his more immediate French followers. It had, however, in the meantime met with opposition in Germany and Sweden, Hiibner in 1816, as we have already seen, having placed the entire group amongst the Geometride, while Dalman had removed them from the Diurnal Lepidoptera to form his uncharacterized group Nyctalidee with P. orontes as the connecting link between them and the butterflies. M.Guenée, however, did not hesitate in 1857 entirely to reject their relationship with the Diurna, showing that with respect to the characters derived from the spring and socket at the base of the wings, the form of the antenne and palpi, the structure and armature of the legs, and the venation of the wings, together with the form of the larve, so far as known at that time, these insects had no real relation with the Diurnal Lepidoptera (Hist. Nat. Ins. Lép. ix. p. 3), that they formed one entire group, and that they ought to be placed among the Nocturna; in fact, although by being placed by some writers at the head of the Heterocera their supposed relationship with the Hesperiide has been in a manner kept up, the pointed tips of the antennz of some of the species, and the spines on the hind legs favouring such a view, they exhibit no real relationship with the Hesperiide. In like manner a comparison of the structural details which I have given in the accompanying plates, with those of the Castniide and Hepialide published in my re- cent memoir on the former family in the ‘Transactions of the Linnean Society,’ will clearly show that the relationship of Coronis with Castnia as indicated by Latreille, and that of Urania with Castnia as suggested by Macleay in this Society’s ‘ ‘Transactions,’ i. p- 188, must be completely ignored. 4B 2 512 PROF. J. O. WESTWOOD ON THE URANIID£. Rejecting, then, the Rhopalocera (including the Hesperiide), the typical Sphingide, Castniide, and Hepialide, together with the whole of the Microlepidoptera, we have to inquire which of the remaining Macrolepidopterous groups show the greatest amount of affinity to the Uraniide. If we rezard the Noctuide, we find a robust body with comparatively small wings formed for powerful flight, and generally marked with a peculiar reniform and a cir- cular spot or patch in the middle of the fore wings; the antenne are also almost invariably slender and setaceous, becoming gradually attenuated to the tip. In this family, however, is found a group (Erebus!) with the palpi elongated, terminated by a slender joint, which probably induced Dalman to place them, under the name of Thysania, with the Urantide. Plate LXXXVI. fig. 4 represents the head of Erebus (Patula) macrops, Linn., Guén. (Bubo, Fabr., Donovan, Ins. China, pl. 44. f. 1). The venation of the wings, however (Plate LXXXVI. fig. 1 fore wing, and fig. 2 hind wing of the male, and fig. 3 hind wing of the female), of the same Indian species of Hrebus denuded of scales, is entirely different from that of any of the Uraniide, the fore wings having the small subcostal cell (se. c) and the lower discoidal vein (¢ 3*) arising close to the base of the third branch (¢ 3) of the median vein from a very short transverse discocellular vein. Mr. MacLeay, in his memoir on Urania, noticed the resemblance between the more or less spherical eggs of Urania and Catocala. ‘The last-named genus, however, is an aberrant one in the family Noctuide; and the oology of the Lepidoptera has not been sufficiently studied to allow much weight to be given to the character of the eggs of these insects. At all events, as Mr. MacLeay remarked, the form of the eggs of Urania is a very common one in Lepidopterous insects. Hence we may reject the Noctuide from amongst the near relations of the Uraniide. Of the remaining families, typified by the Linnean genera Bombya and Geometra, M. Guenée is decidedly in favour of the latter :—‘ Il me semble,” says he (Hist. Nat. Ins. Lép. ix. p. 4), “qu’aucune ne peut lutter a cet égard avec les Géométres. Nous retrouvons d’abord dans la premiére famille de ces derniéres que personne ne sera tenté de disputer aux Phalénes une nervulation [venation] exactement semblable. Les antennes quoique légérement renflées prés du sommet chez plusieurs Uranides, sont fili- formes ou plutdt sétacées, et tout le monde sait que ce n’est que chez les Geometra que cette forme est vraiment normale. L’absence des stemmates et des taches réniformes et orbiculaires suffit pour les éloigner des Noctuelles et les rapprocher des Geometra ou ces caractéres manquent également. Les queues des ailes inférieures, avec les taches ocellées qui les accompagnent ne se retrouvent que chez les Géométres de la premiere famille ou chez les Saturnides qui les précéderont dans la distribution que j’ai adoptée. L’aspect général des deux derniéres familles, leurs ailes minces, étendues, leur vol diurne ‘ M. Eoisduval (in Nouy. Ann, du Mus. ii. p. 260) introduced the genus Urania between Erebus and the Geometride. ; PROF. J. O. WESTWOOD ON THE URANIID&. 515 les poussent encore vers cette division. Quant aux premiers états, il y a sans doute du pour et du contre, cependant ceux de l'Urania rhipheus sont tellement décisifs, que personne n’a hésité a la rejeter dans les Phalénes.” Mr. Packard, in his fine monograph upon the American Geometrideous Moths, goes even further than M. Guenée. On account of the structure of the head! in the Ura- nides he cannot agree with Guenée in regarding them as a family distinct from the Geometride. He adds, “The venation is also much as in the Phalenide” (a group of the Geometride), there being only three median nervules, on which account he regards them as forming the highest subfamily of the Phalenide. “Iam aware that the larve have sixteen feet, no other Phalenid having more than fourteen” (Mon. p. 22). Mr. Packard, who has devoted six plates and nearly 150 figures to represent the venation of the wings of the genera of Geometrideous moths, gives the following as the general character of the venation of the family :—‘ Usually six subcostal venules, always but three median venules; no submedian vein, sometimes a fold representing it. A subcostal cell often present, sometimes two, the cell being formed almost invariably by the anastomosis of the first subcostal venule with its vein. Independent vein well marked, usually in the middle of the discal space” (p. 16). I have copied from Mr. Packard’s plates one of the most characteristic figures of the fore wings of the Geometride (Plate LXXXVI. fig. 7), representing that of Hydria undulata, Packard (pl. i. fig. 14), in which we see two small prediscoidal cells distinct from the costal vein, the upper discoidal vein (B5*) arising at a distance beyond the discoidal cell from what I consider as normally constituting the basal portion of the branch B2 or more probably of B3. In the fore wing of Eumacaria brunnearia, Packard (pl. iii. fig. 7), we see only a single small prediscoidal cell, the first branch of the subcostal vein uniting with the costal vein near its extremity, whilst the upper discoidal vein (B5*) extends backwards so as to form the upper extremity of the discoidal cell, the basal portion of the upper discoidal vein being the upper discocellular veinlet of E. Doubleday, and the lower discoidal vein (c3«) arising halfway between the branch B5* and c3 from an oblique veinlet, the upper part of which forms the middle discocellular veinlet of E. Doubleday, and the lower part being his lower dis- cocellular veinlet. On looking over Mr. Packard’s numerous figures of the veining of the wings, we see 1 The following are the characters of the head of Cydimon leilus (which, following Guenée, Mr. Packard has introduced at the end of the Geometride) given by the last-named writer :— “ Cydimon leilus (pl. vi. f. 25'). The occiput and epicranium are small and narrow, the antenn being in- serted on the summit of the head; the epicranium is very small, the basal joints of the antennz being large and near together. The clypeus occupies the entire front, being much longer than broad, narrow, the sides parallel, not narrowing in front, and the surface flush with the eyes; the front edge is slightly arcuate, being slightly produced in the middle of the edges, with lateral foramina distinct. Mandibles rather long, incurved, and with the usual dense golden sete lying over the base of the maxille. Labrum small, narrow. Maxille well developed.” 514 PROF. J. O. WESTWOOD ON THE URANIID®. that the tendency to form one, two, or even three small cells by the branchlets of the subcostal vein more or less anastomosing together, near the middle of the fore margin of the fore wing, is one of the most constant characters of the Geometride. No such small cell occurs in the Uraniide'; nor, as Mr. Packard remarks, is any Geometrideous larva known having the normal number of sixteen feet, in consequence of which the peculiar mode of locomotion exhibited by the ‘‘looper” caterpillars of the latter family is rendered necessary. In the Uraniide, on the contrary, the larve have sixteen feet; for, even in that of Urania rhipheus, as described by M. Sganzin, ‘il n’y avait aucune interruption de pattes,” although when walking they are said to have * quel- ques rapports avec les chenilles dites Arpenteuses et dans le repos elles formaient entierement la boucle.” On casting our eyes over the extensive family of the Geometride there are a few species which, in their larger size and in the possession of a short tail to each of the hind wings, approach more nearly to the Uraniide than the rest. Urapterya, with which M. Guenée commences the series of the Geometride, forming “un assez bon passage aux Uranides,” is remarkable for the arrangement of the veins of the wings, recalling to mind that of the Saturnides, and differing from the general types of the family. Plate LXXXVI. fig. 5 represents the veins of the fore wing, and fig. 6 those of the hind wings of U. sambucaria. It is true that we here see three branches to the median vein and the lower discoidal (c3«), or the independent vein of Mr. Packard, arising from the middle of the extremity of the discoidal cell; but both on the fore and hind wings a branch (representing the upper discoidal vein, B5*, in the fore wings) is wanting, as is also the small subcostal discoidal cell or cells. There are, however, certain moths, natives of the Malayan archipelago, which ex- hibit a much closer resemblance to the Uraniide than Urapteryx in the arrangement of the veins of their wings, the hind pair of which are likewise furnished with a short broad tail, marked (like that of Urapteryx) with a somewhat eye-like black spot. These form the genus Strophidia of Hiibner and Felder (Micronia, group 1, H. N. Lép. x. p. 24), the first species of which (Micronia astheniata, from Borneo) is named by M. Guenée after my genus Asthenia (upon which observations will be found in the later part of this memoir). Other species are:—WM. caudata, Fab. ( fasciata, Cram. pi. 104. f. D); JL. obtusata, Guen. pl. 5. f. 6 (errore caudata); M. aculeata, Guen. pl. 13. f. 8; I. striataria, Linn., Clerck, pl. 55. Two very typical species have also been figured by Messrs. Felder and Rogenhofer (Strophidia pannata, Novara Exp. pl. cxxviii. fig. 39, from Halmaheira and Salwatti, and S. phantasmah, ib. fig. 40, from Gebeh, Java (Bern- stein)). Plate LX XVI. fig. 8 represents the venation of the anterior, and fig. 9 that of the posterior wing of a typical species of this genus closely allied to S. phantasmah, which * In Chrysiridia rhipheus (P1.LXXXV. fig. 15) there is a very narrow elongated subdiscoidal cell, resulting from the abortion of the extremity of the second branch of the postcostal vein (6 2) and its coalescing with the base of the third branch (6 3), quite unlike that of any of the Geometride. PROF. J. 0. WESTWOOD ON THE URANIIDZ. 515 I received from the Leyden Museum, and which appears to be undescribed’. On com- paring the venation of this insect with that of the Uraniide (e.g. sloanus &c.), it will be seen that they are so nearly identical that no doubt could be entertained of their affinity if we were assured that the larve of the Strophidie were not loopers, and possessed the full complement of sixteen legs. It will, however, be remarked that M. Guenée (H. N. Lép. x. p. 24) is by no means absolutely satisfied that the Strophidie are really - geometrideous, since he enters into a description of the differences which separate them from Asthenidia, Westw., which he assumes to belong to the Bombycides. In instituting this genus Asthenia’ in the volume of Exotic Moths in Jardine’s ‘ Natu- ralist’s Library,’ Entom., vii. 1841, p. 209, pl. xxix. f. 1, I regarded the type, A. poda- liriaria, as belonging to the Geometride, pointing out its chief distinctive characters, especially the short strongly bipectinated antennz, the very short and weak legs and body, and the venation of the wings. The latter character is here represented (Plate LX XVI. fig. 10, fore and hind wings of A. podaliriaria), whence it will be seen that, whilst differing considerably from Urapterya, it approximates more closely to Strophidia, from which it, however, differs at once in wanting several of the branches of the postcostal vein of the fore wings. M. Guenée directs attention to the woolly forehead, the bipectinated antenne, the short indistinctly jointed palpi, the rudimental spiral tongue, the short woolly body, the short legs, woolly anterior tibiz and basal joint of the tarsi, the vena- tion of the wings, and especially to the fact that “tous les rameaux costaux [of the subcostal vein of the fore wings] sont retranchés comme chez toutes les Saturnides,” as so many characters separating Asthenidia from Strophidia, and consigning it to the Bombycides—an opinion in which I am fully prepared to acquiesce from a careful ex- amination of various Saturnideous types. Regarding, then, the relationship between Asthenia and Urapteryx as one of analogy, and not of affinity, and considering Asthenia as belonging to the great group of Bom- bycide, and that Strophidia is more nearly allied to Asthenia than it is to Urapteryx, I conceive that we shall be warranted in placing the Uraniide at a distance from the Geometride and amongst the Bombycide, on account, Ist, of the structure of the larva of Coronidia, Uranidia, and Chrysiridia, and, 2ndly, the venation of the wings destitute of a small cell. The long slender terminal joint of the palpi, and the prolongation of the hind wings into caudate appendages, appear to me to ‘be of secondary importance in determining the relations of the group. 1 The following are its short characters :— SrropHip1a voutennovu, Westw. Alis albis, anticarum margine tenui antico et apicali fasciisque duabus transyersis mediis pallide fuscis, posticarum fascia e medio coste ad angulum analem alteraque apicali pallide fuscis, maculis duabus nigris caudalibus. Exp. alar antic. une. 23. Hab. In ins. Malayanis (Mus. Hopeiano Oxoniz). 2 As the name Asthenia has been elsewhere employed in Diptera, it may be well to slightly modify it to prevent all confusion. I propose, therefore, to change the Lepidopterous name to Asthenidia, 516 PROF. J. 0. WESTWOOD ON THE URANIIDA. The fact that Godart, in the ‘ Encyclopédie méthodique,’ described a species of Coro- nidia as an Agarista, that Latreille placed the two genera in juxtaposition, and that M. Boisduval had united the three genera named by him Urania (Chrysiridia), Cydimon (Uranidia), and Nyctalemon (Alcidia and Lyssidia) as three of the tribes of his family Agaristidées in his ‘Monographie des Agaristidées,’ render it necessary to inquire how far this relationship is real. It is true that the shape of the antenne in Uranidia agrees with that of Agarista, and that the armature of the four posterior tibiz is similar ; but the arrangement of the wing-veins in the two groups is wholly unlike, as may be seen by comparing my figures accompanying the present memoir with that of Agarista lindigii given by me in the ‘ Transactions of the Linnean Society,’ Ser. 2, Zool. vol. i. pl. xxix. fig. 24, which exhibits the small lozenge-shaped cell (occurring also in Othria augias, tab. cit. fig. 18, Hespagarista interjecta, ibid. fig. 22), of which there is no trace in Uranidia, where the lower discoidal vein is far removed at its base from the terminal portion of the median system of veins. A more careful examination than has hitherto been published of the arrangement of the veins of the wings in the chief types of the Bombycide discloses the fact, that there is more variation amongst them in this respect than is to be met with in other great groups, such as the Sphingide, Noctuide, and Geometride, and hence that a discrepancy between the veining of the wings of the Uraniide and the Bombycide is not a sufficient argument for their rejection from the latter. In the accompanying Plate LXXXVI. I have given figures of a few of the leading types of Bombycide, commencing with those in which some of the branches of the veins are obsolete, and terminating with some which have the full complement of veins and branches. It is remarkable that the gigantic types of the Bombycide (Attacus atlas and its allies), not- withstanding the comparatively large size of the wings, should have several branches wanting ; and it is not easy to trace the precise analogies of some of those which remain. In this species the strong costal vein of the fore wings (a) extends about three fourths of the length of the entire costa; the subcostal vein has apparently only three instead of five branches ; the first branch (6 1) arises at a short distance before the extremity of the discoidal cell, and reaches the costa halfway between the extremity of the costal vein and the tip of the wing; the second branch (4 2) of the subcostal arises at a moderate distance beyond the cell and reaches the tip of the wing, whilst the main branch (4 3) extends to the middle of the rounded hook or apex of the wing. From the underside of the subcostal vein, just beyond the branching of the first branch, extends a vein obliquely, forming a portion of the anterior margin of the glassy disk, and branching into two branches at the middle of this vitreous spot. Are these two branches the fourth and fifth branches of the subcostal vein ? or are they the two discoidal branches of Mr. E. Doubleday, one of which I have considered as supplemental to the subcostal series, and the other to the median series? In their position they exactly correspond to the arrangement of the veins in Morpho, as represented by Mr. Doubleday (Gen. D. PROF. J. O. WESTWOOD ON THE URANIIDA. 517 Lep. pl. A. fig. 3), the only difference being that in that figure we have all the five branches and the postcostal vein. In that work these two branches are rightly regarded as the two discoidal branches, whilst in Attacus, as in Morpho, we only find three clearly defined branches of the median vein, ¢ 1, ¢ 2, ¢ 3. In Attacus pavonia minor (Plate LXX XVI. fig. 11) we have the same small number of branches in the fore wing as in A. atlas, with the exception that the first branch of the postcostal vein (B') is so very short and slender as to be scarcely visible, arising nearly at the tip of the wing and almost immediately joining the costa; the difficulty as to the two discoidal branches (B4=? B5x* and B5 =? c3x) is the same as in A. atlas. In Aglaia tau (fig. 14) the same arrangement of the veins of the fore wings exists as in A. atlas. In the large long-tailed Bombycide (Actias luna &c., Plate LKX XVI. fig. 15, fore wing, fig. 16, hind wing) we have the same arrangement of the median and two supposed discoidal veins ; but the postcostal vein has an extra fourth branch ; the first (6 1) arises at about three fourths of the length of the costal margin, and extends to its five-sixth portion, whilst the second branch (6 2) is exceedingly short and close to the tip of the wing, just as in Pavonia minor; the third branch extends (6 3) to the tip of the wing, and is followed by a long branch (4 4), which agrees equally with the branch (6 3) in Pavonia minor; the fifth postcostal branch is wanting, whilst the two discoidal ones (b5« and ¢3«) arise from a very short basal vein near the extremity of the discoidal cell. In the tailed hind wings of this group (by which they are rendered to a certain degree analogous to some of the Uraniide) the three branches of the me- dian vein (¢ 1, ¢ 2, ¢ 3) run to the extremity of the long tail. Such is also the case with the remarkable Phalena brachyura of Cramer, from tropical Africa, whilst in Eudemonia semiramis (Plate LX XXVI. fig. 13) the long tail is strengthened by the third subcostal branch as well as the second and short third median branches. (In the Uraniideous genera Manidia and Coronidia the tail of the hind wings is strengthened by two of the branches of the median vein, whilst in Uranidia leilus it bears only the third median branch. ) In Saturnia certhia, Fabr., we find a deficiency in one of the branches of the post- costal vein, the first branch arising at about three fourths of the length of the wing, the second branch extending to the tip, the third and fourth arising at about the middle of the wing at some distance beyond the discoidal cell, whilst the two discoidal branches, as well as the third branch of the median veins, arise from the transverse extremity of the discoidal cell. In Endromis versicolor (Plate LXXXVI. fig. 12) we find the full complement of veins and branches in the fore wings; and here the two discoidal branches are so placed as completely to prove (as it seems to me) that the upper one (B5x) is a portion of the postcostal series, and that the lower one (c3*) is a portion of the median series, The discoidal cell is here closed by a much curved discocellular vein, emitting two veinlets running towards the base of the wing within the cell, which might be assumed to be VOL. X.—PART x11. No. 3.—June Ist, 1879. 4c 518 PROF, J. O. WESTWOOD ON THE URANIID&. the representatives of the lost basal portions of the discoidal branches, if these should be considered distinct from the two series, postcostal and median. In Gastropacha quercifolia there is also the full complement of veins and branches Plate LXXXVI. fig. 17); but here the binary system of the branches into postcostal and median is still more marked, the branch marked c3« being as completely a branch of the median vein as either c 2 or c 3 (in which respect it agrees with Papilio), whilst B5* is as completely a portion of the postcostal series as B 5 or B 4, the latter of which will be seen to arise at the upper extremity of the discoidal cell. In Gastr. expansa, B 4 arises beyond the cell, whilst the basal portion of B 2 and B 3 arises at the extremity of the cell (where B 4 arises in G. guercifolia) ; so that the first postcostal branch (B 1) is the only branch arising from the upper margin of the discoidal cell. The only other Bombycideous insects which I shall here notice belong to the singular genus Epicopeia, Westw. (Arc. Ent. i. pl. 5), founded upon certain Indian moths which have entirely the appearance of some species of Papilio. The late Mr. Edward Double- day, to whom I applied for his opinion on the relations of this singular genus, ob- served “that it seems to partake of the characters of Papilio, Urania, and that group of Bombyces to which B. luna belongs. The last-named species has no bridle to the wings, no maxillz ; and there is some resemblance in the neuration of the wings. But I must confess that I see no real connexion between the two insects. My impression is that it must be nearer the Uraniw, some of which, in form, nearly resemble it ; but all these have maxill and the discoidal cell of the posterior wings open, and two pairs of spines, I think, to the posterior tibie. The one middle spine to the anterior tibie is found in some Uranie.” In Epicopeia polydora 9 (Plate LXXXVI. fig. 18) we have the full complement of the veins of the fore wings arranged almost as in Saturnia certhia, except that the second and third postcostal branches arise from the front margin of the fourth branch, which, as well as the fifth branch, arises from the upper extremity of the discoidal cell, which emits the two discoidal branches and the third median branch from its truncated extremity, as in Saturnia certhia. The hind wings are very remarkable, terminating in a very broad tail, which is traversed by all the three branches of the postcostal vein. Although the relationship of Epicopeia with Urania, pointed out by Mr. Doubleday, does not appear to me so strong as it did to him, the nearer relationship of the former being, as it seems to me, towards certain of the Chalcosideous Bombycide, it is impos- sible not to see that we have in all these insects a proof of the stronger relationship of Urania with the Bombyces than with any other of the Heterocera. It remains to notice the structure of these insects in their preparatory states, which equally supports the opinion advanced above, that the group is to be referred to the great division of the Bombyces. The transformations of Urania boisduvalii (U. fernan- dine, MacL.) are fully described by Mr. MacLeay in his memoir in the first volume of the ‘Transactions’ of this Society. The larva cannot, from his account and figure, PROF. J. O. WESTWOOD ON THE URANIID®. 519 be referred either to the Hesperideous, Sphingideous, or Geometrideous divisions of the order. “Its head,” says Mr. MacLeay, “is not set on the body by means of a narrow neck, as in the larve of the true Hesperide. It can run about as quickly as the larve of any Bombycide, and shows little affinity to the caterpillars of the diurnal Lepidoptera, which usually have a slow motion. In form it agrees very closely with the caterpillar of Agarista, as figured by Lewin, but is more simple, having no hinder pro- tuberances on the penultimate segment.” It, indeed, appears to me to bear a close resemblance to the larva of the buff-tip moth, B. bucephala, Linn. Mr. Macleay de- scribes the cocoon of this insect as made of loose dirty yellow silk, the meshes of which were so few as to allow the inmate to be easily seen ; but his figure represents a much more solid structure. The chrysalis, which is not at all angular, moreover, is said to repose in a horizontal position—circumstances which bear upon the question whether the chrysalis is supported by a thread girt round the middle of the body, as in the butterflies with perfect fore legs, and which does not appear likely to be the case with the Urania-chrysalis. The transformations of U. rhipheus were observed by M. Sganzin in Madagascar, and communicated to M. Boisduval, by whom they were published in his ‘‘ Monographie des Agaristidées,” in the ‘Revue et Magasin de Zoologie, 1874. On the 9th of August “la femelle pondit un grand nombre d’eufs blancs agglomérés ensemble et placés sur les feuilles de manguier. Au bout d’une douzaine de jours les ceufs vinrent d’éclore.... . La chenille en grossissant changeait de couleur et devenait de plus en plus noirdtre. Quand elle eut quelques lignes de grandeur, on apercevait trés-distinctement de petites épines jaunatres et deux petits cornes roses -retractiles ; elles les faisaient mouvoir 4 volonté et les rentraient quand on les inqui- était le moindrement et surtout lorsqu’on les exposait 4 l’ombre. En grossissant mes chenilles prenaient une singuliére forme, elles s’élargissaient vers le milieu et devenaient minces aux extrémités. On apercevait sur les cétés un feston 4 dents de loup, com- posé de plusieurs bandes réguliéres de points blancs, verts et jaunes; les cornes deve- naient d’un rose foncé, tirant sur le carmin: elles avaient dans leur marche quelques rapports avec les chenilles dites Arpenteuses, et dans le repos, elles formaient entiére- ment la boucle; cela m’a paru dautant plus extraordinaire, qu'il n’y avait aucune interruption de pattes comme cela a lieu chez les Arpenteuses. La plus grande des chenilles pouvait avoir trois pouces de longueur ; toutes les autres étaient plus petites. Tl est possible que dans l’état de liberté elles prennent plus d’accroissement. Lorsque les chenilles ont fait leurs chrysalides, elles se sont attachées au moyen d'un fil passé au milieu du corps; je crois qu’elles etaient aussi attenantes par la queue. La chry- salide était verte, conique, et un peu anguleuse, avec des bandes dorées placées hori- zontalement depuis la téte jusqu’a la queue. L’extrémité etait d'un vert bien plus foncé et parsemée d’un grand nombre de points dorés. Les papillons sont restés 21 jours en chrysalides.” Three circumstances in M. Sganzin’s account merit notice, as showing differences not 4c2 520 PROF. J. O. WESTWOOD ON THE URANIID#. mentioned by Mr. MacLeay :—1st, the retractile rosy-coloured tentacles of the larve ; 2ndly, its somewhat Geometriform movement; and, 3rdly, the want of a cocoon (at least no mention is made thereof by M. Sganzin), and the girt condition of the chry- salis. I am able to add descriptions and figures of the larva and pupa of a species of the genus Coronidia, allied to C. orithea, which had been observed by Herr Moritz at Bogota, and to which he had applied the name of C. regina. The specimens of the Jarva and pupa were carefully prepared, and were in the collection of Dr. Kaden, of Dresden, where I examined and made the drawings of them here published (Pl. LXXXV. figs. 1, 2). The caterpillar has a small head, and the prothorax is smaller and thinner than the following segment. The general colour is luteous, with black spots and a black head ; on the prothorax are ten small shining points, forming an oval patch placed transversely ; the segments of the body are furnished with two small conical luteous tubercles, and on the second, sixth, and seventh segments are two erect black, slightly curved spines. The chrysalis is oval, entire, without conical protuberances; the tongue- case is considerably elongated beyond the wing-cases; and the extremity of the body forms an acute deflexed spine. It is enclosed in a loose open-work cocoon spun at the base of a folded leaf. The following Table combines the principal characters of the sections and genera comprising the family Uraniide, details of which are illustrated in Plate LXXXV. Family Urantipa, Westw., Introduction, ii. p. 369 (1840), Uranides, Guenée, Sp. Gén. ix. p. 1. Cydimonii, Blanchard, H. N. Ins. ii. p. 848; Walker, List Brit. Mus. Lep. Het. p. 4. A. Fore wings with the fifth branch of the subcostal vein (65) emitting the upper discoidal vein (65 *) at a considerable distance beyond the discoidal cell. a. (Cydimon Dalm.) Palpi with the terminal joint very short. * (Cydimonide Blanchard, Guenée.) Hind wings with the second and third branches of the median vein produced into the long tail ; second branch of the subcostal vein free.— Gen. Unanipia, Westw.; type U. leilus (figs. 18, 14). *%* (Urania Blanchard, Guenée.) Each of the veins of the hind wings prolonged into a tail ; second branch of the subcostal vein of the fore wing coalescing with the third branch halfway between the cell and the tip of the wing.—Gen. Curysiripia, Hb.; type U. rhipheus (figs. 15, 16). b. (Nyctalemon Dalm., Nyctalemonide Guenée.) Palpi with the terminal joint long, slender, and pointed. * Each of the veins of the hind wings produced into a short scallop, postcostal vein of fore wings wanting the 2nd branch (6 2).—Gen. Aucrpta, Westw.; type U. orontes (figs. 17, 18). ** Hind wings with the second and third branches of the median vein produced into a long tail.—Gen. Lyssip1a, Westw.; type U. patroclus (figs. 19, 20). PROF, J. O. WESTWOOD ON THE URANIID&. 521 B. (Sematuride Guenée.) Fore wings with the upper discoidal vein (65 *) arising at a short dis- tance beyond the discoidal cell, between the cell and the origin of the second branch of the subcostal vein (4 2). : a. Discoidal cell of the hind wings terminating at a distance before the emission of the first branch of the median vein.—Gen, Manip1a, Westw.; type U. lunus (figs. 10-12). b. Discoidal cell of the hind wings extending nearly to the emission of the third branch of the median vein.—Gen. Coronip1a, Westw.; type U. orithea (figs. 1-9). Genus 1. Uranipia, Westw. (Details, Plate LXXXV. figs. 13, 14.) Urania (pars) Fabricius, in Illig. Mag. vi.; Hiibner, Verz. ; Latreille. Cydimon (pars ult.) Dalman (1824), Kongl. Vet. Acad. p. 392. Cydimon Guenée, Felder, Blanchard, Boisduval, Rev. Zool. 1874, p. 10. Urania (pars prior), Walker, List B.M. Lep. Het. p. 4. Leilus (pars), Swainson, Zool. Ill. n. ser. pl. 129. The species of this genus are remarkable for the habit of migrating in large numbers, as described by Mr. MacLeay in his paper in the Society’s ‘ Transactions.’ Such flights of U. leilus are also described in ‘ Nature,’ iv. pp. 12, 13, 494; and the annual migrations of U. fulgens from east to west in August and September, across the Isthmus of Panama, are described in the same journal (‘ Nature,’ viii. p. 536). Sp. 1. URANIDIA LEILUS. Papilio leilus, Linn. 8. N. ii. 750; Clerck, Icon. pl. 27. f. 1; Cramer, Pap. Exot. i. pl. 85. f. C, D; Fabricius, Syst. Gloss. (Urania, L.) ; Guenée, Sp. Gén. ix. p. 7; Boisduval, Rev. Zool. 1874, p- 10 (Cydimon, L.). Lars heroica leilaria, Hiibner, Samml. ex. Schm. (L. B. heroice), pl. 4. f. 1, 2. Leilus surinamensis, Swainson, Zool. Ill. 2nd ser. pl. 125. Hab. Cayenne, Surinam. Sp. 2. URANIDIA AMPHIELUS. Cydemon amphielus, Boisduval, Rev. Zool. 1874, p. 11. Cydimon leilus, var., Guenée, Sp. Gén. ix. p. 7. Hab. Trinidad. Sp. 3. URANIDIA BRASILIENSIS. Leilus brasiliensis, Swainson, Zool. Ill. 2nd ser. pl. 126; Guenée, Sp. Gén. ix. p. 7, pl. 1. f. 1. Hab. Brazil. Sp. 4. URANIDIA CACICA. Cydimon cacica, Guenée, Sp. Gén. ix. p. 8; Boisduval, Rev. Zool. 1874, p. 11. Hab. Mexico, Acapulco. 522 PROF. J. O. WESTWOOD ON THE URANIID&. Sp. 6. URANIDIA POEYI. Urania poeyi, Herrich-Schiffer, Corr. zool. Ver. Regensburg, xx. 1868 (p. 43, Lep.); Felder, Novara, Lep. v. pl. cxxi. figs. 6, 7. Hab. Cuba. Sp. 6. URANIDIA BOISDUVALII. Urania boisduvalii, Guérin, Icon. R. An. p. 490, Ins. pl. 82. f. 1; Griffith, An. Kingd. Ins. pl. 99; Guenée, Sp. Gén. ix. p. 8; Boisduval, Rev. Zool. 1874, p. 12. Urania fernandine, Macleay, Trans. Zool. Soc. i. 1834, p. 180, pl. 26. Hab. Cuba. Sp. 7. URANIDIA FULGENS. Urania fulgens, Boisduval, MSS.; Walker, List B.M. Lep. Het. p. 5; Guenée, Sp. Gén. ix. p. 9; Boisduval, Rey. Zool. 1874, p. 16, Lepid. Guatemala, p. 77. Hab. Columbia, Guatemala, Mexico. Sp. 8. URANIDIA SLOANUS. Papilio sloanus, Cramer, Pap. Exot. i. pl. 85. figs. E, F; Sloane, Jamaica, 1. pl. 239. f. 11, 12; Godart, Enc. M. ix. p. 709 (Urania, 8.) ; Guenée, Sp. Gén. ix. p. 9; Boisduyal, Rey. Zool. 1874, p. 15. Papilio leilus, var. 3, Gmelin, i. pp. 5, 2, 237; Fabricius, Ent. Syst. iii. p. 122. Urania sloanaria, Hubner, Verz. p. 289. Leilus occidentalis, Swainson, Zool. Ill. 2nd ser. pl. 129. Hab. Jamaica. Genus 2. Curysiripia, Dalm. (Details, Plate LXXXYV. figs. 15, 16.) Chrysiridia, Hiibner (Verz.). Urania, Boisduval, Rey. Zool. 1874, p. 7; Guenée, Sp. Gén. ix. p. 11". Urania (pars ult.), Walker, List B. M. Lep. Het. p. 4. Cydimon (pars prima), Dalm. 1825. Thaliura, Duncan in Jardine, Nat. Lib. Lep. p. 195 (1837). Leilus (pars), Swainson, Zool. Ill. n. ser. pl. 130. Rhipheus, Swainson, Zool. Ill. n. ser. pl. 131. Sp. 1 (9). CurysrRIDIA RHIPHEUS. Papilio rhipheus, Drury, ii. pl. 23. f. 1, 2, Index, col. 2; Cramer, Pap. Exot. iv. pl. 385. f. A, B; Godart, Enc. M. ix. p. 709 (Urania rh.) ; Boisduval, Faune Ent. Madagascar, Lép., pl. 14. f.1, 2, and Rev. Zool. 1874, p. 8; Crochard’s edit. R. An. Atlas Ins. pl. 144. f. 3; Guenée, Sp. Gén. ix. p. 12. Urania prometheus, Drapiez, Dict. Class. d’Hist. Nat. iui. pl. 8. f. 1, 2. Leilus orientalis, Swainson, Zool. Ill. 2nd ser. pl. 130. 1 Fabricius does not mention U. rhipheus amongst his types of Urania. PROF. J. 0. WESTWOOD ON THE URANIIDZ. 528 Rhipheus dasycephalus, Swainson, op, cit. pl. 131, Urania druryi, Boisduyal, Rey. Zool. 1874, p. 8. Urania rhipheus, var., Guenée, Sp. Gén. ix. p. 12. Chrysiridia riphearia, Hiibner, Verz. p. 289. Urania madagascariensis, Lesson, Illustr. d. Zool. pl. xxxiii. Hab. Madagascar; China? (Drury'); Bengal? (Cramer); Coromandel?; St. Helena ? (Bory de St. Vincent). I am quite satisfied that the figure given by Drury was taken from a specimen which had the hind wings mutilated, and the head replaced by that of a Papilio. Drury gave China as the locality of his specimen, which is also clearly erroneous; but whether the specimen had been brought from Madagascar or from some part of the eastern coast of Africa is uncertain. In the latter case his insect may be geographically distinct from the Madagascar individuals; and this would enable us to account for the variations in the markings of the wings pointed out by Swainson, Guenée, and Boisduval. Until, however, we receive individuals agreeing with Drury’s figure, I should be inclined to think that the variations pointed out resulted from the evidently imperfect and partially rubbed condition of Drury’s insect. According to M. Sganzin, who captured numbers of C. rhipheus in Madagascar, the small neighbouring island of Sainte-Marie possesses a smaller species, thus noticed by him in his communication to M. Boisduval :—“ II] existe aussi une petite espéce que je crois trés-différente de la premiére (U. rhipheus). On ne l’a pas rencontré 4 la méme époque, et je ne crois pas qu’on la trouve a la grande terre; je ne l’ai jamais vue qu’a Sainte-Marie; elle parait en mars et avril,” the true U. rhipheus appearing in the winged state in September. Sp. 2 (10). CurysiriDIA MACLEAYII. Urania macleayii, Montrouzier, Ann. Sci. Phys. et Nat. Lyon, sér. 2, viii. p. 410. Hab. Woodlark Island. Sp. 3? (11%). Curysrripia crasvs. Thaliura cresus, Gerstaecker, in Archiv f. Naturg. xxxvil. p. 361, and V. d, Decken, Reisen in Ost-Afrika, Ins. p. 383, tab. xvi. f. 4. “ A Thal. rhipheo, Cyr., differt alarum anticarum fascia media smaragdina latiore, mar- ginem externum versus furcatim divisa, strigis dimidii apicalis densioribus, posticarum area smaragdina et igneo-cuprea uberius nigro conspersa et fasciata, appendiculis 3 posticis brevioribus, alis anticis infra ubique subequaliter viridi-undulatis, posticarum area basali obscurius ceruleo-viridi.’ Exp. alar. mill. 78 g, 929. Hab. Ins. Zanzibar. I examined specimens of this supposed species in the hands of M. Deyrolle in Paris, 1 On the supposed Asiatic locality of this species see Trimen in Journ. Linn. Soc. Zool. xi. p. 284. 524 PROF. J. O. WESTWOOD ON THE URANIIDZ. which did not, however, appear to me to be specifically distinct from C. rhipheus. Such is also the opinion of M. Lucas, who states that all the transitional forms occur be- tween U. rhipheus and U. cresus. Bull. Soc. Ent. Fr. [5] vi. pp. exxvii, exxviii. Genus 8. ALcipiA, Westw. (Details, Plate LXXXYV. figs. 17, 18.) Alcidis, Hiibner (Verz. 1816) ; Guenée, Sp. Gén. ix. p. 13 (nec Alcides, Dalm. 1826, Gen. Curcul.). Nyctalemon (pars prior), Dalman, Walker, List B.M. Lep. Het. i. p.7; Boisduval, Rev. Zool. 1874, p- 17. Sp. 1 (12). Axcmra orontEs. Papilio orontes, Linn. Syst. Nat. ii. p. 750; Clerck, Icon. pl. 26. f. 1; Cramer, Pap. Exot. i. pl. 83. figs. A, B; Godart, Enc. M. ix. p. 710 (Urania o.); Hopffer, Neue Schm. Heft 2, tab. 11, figs. 1, 2. Alcidis orontiaria, Hiibner, Samml. exot. Schm., Lar., B. her. 3, f. 3, 4. Hab. Amboina, Ceram, Gilolo, New Guinea, Cape Grafton, north-east coast of New Holland. Sp. 2 (13). Axcrpra LIRIS. Alcidis liris, Felder, Wiener ent. Monatschr. Bd. iv. p. 250; Reise d. Novara, Lep. pl. exxi. fig. 2. Hab. Batchian (Wallace). Sp. 3 (14). Aucrp1a ARNUS. Alcidis arnus, Felder, Reise d. Novara, Lep. pl. exxi. fig. 1. Hab. Insula Aru? (Lorquin). In Mus. Brit. &c. Sp. 4 (15). ALcrpia cypNus. Nyctalemon cydnus, Felder, in Wiener ent. Monatsch. iii. pl. 3. f. 1, p. 179. Var. With band of hind wings very broad, extending nearly to the anal angle (Aru). An JN. metaurus, Hopff.? Hab. Dorey, New Guinea (Mus. Brit.) ; Polynesia? (Felder). Sp. 5 (16). AucrpIa zoDAIca. Nyctalemon zodaica, Butler, in Entom. M. Mag. vol. v. p. 273. “39. Ale supra nigre, ad basin virescentes, fascia media communi lata aureo-viridi ; antice fascia altera lineolari subapicali, pallidiore striolisque costalibus ad basin aureo- albidis ; posticee cauda ceruleo-alba, ciliis albis; macula squamisque submarginalibus analibus, corpus virescens abdomine pallidiori. Ale subtus pallide virescentes, fasciis fere velut in V. orontiaria, Hiibner (orontes, Linn.), maculis autem posticis subapica- libus in margine subrotundatis viridibus; corpus thorace albido, abdomine aurantiaco, cirris maris perlongis ochreis. Expans. alar. une. 4, lin. 7.” PROF. J. O. WESTWOOD ON THE URANIIDZ. 525 Except in being more bronzed in the colour of the wings, this species scarcely differs from A. agathyrsus. Hab. North China (Fortune). In Brit. Mus. Sp. 6 (17). ALcIDIA AGATHYRSUS. Nyctalemon agathyrsus, Kirsch, Mittheil. zool. Mus. Dresden, Heft u1. p. 129, pl. vii. figs. 8, 8a. Hab. New Guinea, In Mus. Dresden. Sp. 7 (18). Atcrpra Boops, sp. nov. (Plate LXXXVII. fig. 1.) Alis supra cyaneo-nigris, anticarum dimidio basali cost striolis, fascia submedia postice intus incurva, fasciolaque subapicali czruleo-viridibus; posticis fascia media dilatata ejusdem coloris, cauda czruleo-alba, ciliis albis. Capite maximo, ano maris fulvo bar- bato. Expans. alar. antic. unc. 43. Hab. Aru (Wallace). In Mus. Hopeiano Oxonie. This insect agrees so entirely with A. agathyrsus that I should have had no hesitation in so naming it; but the extraordinary size of the head and eyes, far exceeding those of any other males of the genus which I have examined, seems a sufficient character to allow it a distinctive specific rank, although I am aware that in certain insects which are distinguished by the large size of their heads there is much difference in the extent of this character. I know no such instance, however, amongst Lepidopterous insects. Sp. 8 (19). Aucrra MeTAvRUS. (Plate LXXXVII. fig. 2.) Nyctalemon metaurus, Hopfter, Neue od. wen. bek. Schmett. 11. Heft, pl. 2. figs. 3, 4; Boisduval, Rev. Zool. 1874, p. 18. Alcides orontes, Guenée, Sp. Gén. ix. p. 14. Note—tThe figure given by Hopffer is far too highly coloured. I have therefore given a new representation of it. It may be distinguished by the very broad conical central fascia of the fore wings. Hab. New Guinea and north of Australasia. In Mus. Hop. Oxoni. Sp. 9 (20). ALcrDIA AURORA. Alcides aurora, Salvin & Godman, Proc. Zool. Soc. 1877, p. 150, pl. xxiii. figs. 5, 6 Hab. New Ireland, Duke-of-York Island. In Brit. Mus. Genus 4. Lyssip1a, Westw. (Details, Plate LXXXV. figs. 19, 20.) Lyssa, Hiibner (Verz. 1816), nec Lissa, Leach, Gen. Crust. (1815). Nyctalemon, pars ult., Dalman, Boisduval, Rev. Zool. 1874, p. 16. VOL. X.—Part. xu. No. 4.—June 1st, 1879. 4p 526 PROF. J. O. WESTWOOD ON THE URANIID#. Nyctalemon, pars med., Walker, List B.M. Lep. Het. p. 7. Nyctalemon, Guenée, Sp. Gén. ix. p. 14. Sp. 1 (21). Lyssipra PATROcLUS. Papilio patroclus, Linn. Mus. Reg. 204; Syst. Nat. ii. 749; Clerck, Icon. pl. 37. fig. 1; God. Enc. M. ix. 710; (Urania p.) Guenée, Sp. Gén. ix. p. 15. Lyssa patroclaria, Hiibuer, Verz. p. 289. Note.—The fascia of the fore wing is broader, white, and extends nearly to the anal angle of the fore wing. Hab. Amboina, Sylhet, Java, Duke-of-York Island. Sp. 2 (22). Lyssipia MEN@TIUS. Papilio patroclus, Cramer, Pap. pl. 109. figs. A, B; Drury, Ins. i. pl. 7, 8; Guenée, Sp. Gén. ix. p. 15 (Nyctal. patr.), nec Linn., Clerck. Nyctalemon menetius, Hopffer, Neue od. wen. bek. Schmett. 11. Heft, pl. 3. fig. 1 (1856) ; Boisduval, Rev. Zool. 1874, p. 20. Nyctalemon zampa, Butler, Ent. M. Mag. vol. vy. p. 273 (1869). Nyctalemon crameri, Boisduval, Rey. Zool. 1874, p. 19. Lyssa achillaria, Hiibuer, Verz. p. 289. Nyctalemon achillaria. Guenée, Sp. Gén. ix. p. 15. Papilio patroclus, var., Cramer, Pap. pl. 198. figs, A, B. Note.—The fascia is narrow, white, and does not reach near to the anal angle of the fore wings. Hab. China, India, Singapore. Sp. 3 (23). Lyssmpra HECTOR. Nyctalemon hector, White, MS.; Walker, List Brit.-Mus. Lep. Het. vii. p. 1771. Nyctalemon longicaudus, Schauff. Nunq. Otios. i. p. 13 (Manilla). ZI. fusculo-cinerea, alis anticis posticisque fascia recta obliqua alba; alis anticis margine basali nigro, albo subundulato ; alis posticisad apicem externum nigro plagiatis vel subfasciatis. Hab. Borneo and Philippine Islands. In Mus. Brit. Smaller than NV. patroclus, inner tail of hind wings proportionally longer. Sp. 4 (24). LyssipIa NAJABULA. Nyctalemon najabula, F. Moore, Proc. Zool. Soc. 1877, p. 620. Hab. Andaman Islands. In Mus. Brit. Smaller and of a much darker fuliginous-brown colour both above and below than the Indian LZ. menetius and the Malayan Z. docilis, and has a narrower median trans- PROF. J. 0. WESTWOOD ON THE URANIID#. 527 verse pale band on both wings above than the former species; and on the underside these bands are well defined and narrow. Expanse 5 inches. S. Andamans (Port Blair). In Coll. F. Moore. Sp. 5 (25). Lyssrp1a Dociuis. Nyctalemon docile, Butler, Journ. Linn. Soe. (Zool.) xiii. p. 197; Druce, P, Z. S. 1878, p. 642. Hab. Malacca. In Mus. Brit. Genus 5. Manipia, Westw. (Details, Plate LXXXV. figs. 10-12.) Mania, Hiibner (Verz.). Sematura, Dalman, Act. Holm. 1824, p. 407; Guenée, Sp. Gén. ix. p. 17. Nyctalemon, pars ult., Walker, List B. M. Lep. Het. p. 8. ——, Guenée, H. N. Lép. ix. Sp. 1 (26). Manrpia Lunus. Phalena lunus, Linn. Syst. Nat. ii. 810 ; Clerck, Icon. pl. 52. figs. 3, 4; Godart, Enc. Méth. ix. 711. Sematura acteon, Felder, Novara, pl. exxi. fig. 5 (=“* Lunus Clerck,” Felder, in text). Foem.? Papilio empedocles (Fabricius, Mant. Ins. ii. 10?), Cramer, Pap. Exot. iii. pl. 199. figs. A, B ; Godart. Enc. Méth. ix. 711; (Urania emp.) Guenée, Sp. Gén. ix. p. 19. Mania empedoclaria, Hiibuer, Verz. p. 290; Samml. exot. Schm. Lar., B. her. B, f. 314. Hab. Brasilia, Honduras. Note.—Cramer’s figures B, C, plate 220, named cunigeraria by Hiibner, Verz. p. 290, have been referred to this species; but they represent a different insect, having undu- lating, not straight, fasciz on the wings. Sp. 2 (27). Manip1A SELENE. Sematura selene, Guenée, Sp. Gén. ix. p. 18; an Nyct. egistus, W1k.? Hab. Para. Sp. 3 (28). Manipra DIANA. Sematura diana, Guenée, Sp. Gén. ix. p. 18. Hab. U Sp. 4 (29). Manrpra pHa. Sematura phebe, Guenée, Sp. Gén. ix. p. 19, Ann. Soc. Ent. France (5), v. pp. 117, 118 ( 2). Hab. t Sp. 5 (30). MaAnmpIa CAUDILUNARIA. Papilio lunus 9, Cramer, Pap. pl. 200. fig. A. Sematura caudilunaria, Hibner, Verz. p. 290. 4p 2 628 PROF. J. O. WESTWOOD ON THE URANIIDZ. Sematura phebe 3, Guenée, Ann. Soc. Ent. France (5), v. p. 117 (1874). Hab. Brasilia, Cayenna, Guiana. In Mus. Hopeiano Oxonie. Sp. 6 (31). Manipia Excavata. Nyctalemon excavatus, Walker, List Brit.-Mus. Lep. Het. p. 9; Butler, Ill. Lep. Het. B. M. pl. 1. fig. 1. Hab. St. Domingo, Haiti, and Jamaica. In Mus. Brit. et Hopeiano Oxonie. Sp. 7% (32). Manipia &a@IsTus. Papilio egistus? Fabr. Mant. Ins. ii. p. 10. Nyctalemon egistus, Walker, List B.M. Lep. Het.i. p.9; an S. selene, Guenée ? Hab, Jamaica (Gosse). In Mus. Brit. Sp. 8? (33). Manipia? Lavinia. Papilio lavinia, Fabricius, Ent. Syst. iii. p. 22. An hujus generis? an Papilio ? Hab. America. Genus 6. CoronipiA, Westw. (Details, Plate LX XXYV. figs. 1-9.) Coronis, Latreille et alior.; nec Coronis, Hiibner (Verz. Gen. Noctuidarum) ; nec Coronis, Latr. (Gen. Crust.) ; nee Coronis, Glog. (Gen. Avium). Larunda, Hiibner (Verz.), Felder; nec Larunda, Leach (Crust.). Section 1. Alis posticis ceeruleo fasciatis. Sp. 1 (34). Coronm1a onitHEA. (Plate LXX XVII. fig. 3.) Alis brunneo-nigris, anticis fasciis duabus albido-carneis rectis parallelis distantibus, secunda e costa ad angulum internum extensa intus dentata ; alis posticis macula magna discoidali semilunari ceruleo-azurea in lineam curvatam ad marginem analem prope angulum analem extensa; lunulis tribus nigris supra violaceis inter angulum et caudam, hujus apice subalbido. Phalena orithea, Cramer, Pap. Exot. iii. pl. 262. fig. C; (Larunda or.) Hiibner, Verz. no. 2807 ; (Coronis or.) Walker, List Lep. Het. Brit. Mus. i. 87; Guenée, Sp. Gén. ix. p. 21. Coronis @urvillii, Boisduval, in Régne An. 2nd edit. i. p. 440, & v. p. 389, pl. xx. fig. +; Godart, Enc. Méth. ix. p. 803; Guenée, Sp. Gén. ix. p. 21. Hab. Surinam, Cayenne, Brazil (Espirito Santo), Guiana. The original figure of this species given by Cramer exhibits the large semioval violet- coluored patch extending more than halfway from the costa of the hind wings, PROF. J. O. WESTWOOD ON THE URANITID#. 529 and reaching in a fine curved line to the anal margin just above the anal angle, with several dark lunules between it and the tail. M. Boisduval, evidently overlooking Cramer’s figure, described and figured the same species in Latreille’s Appendix to the second edition of the ‘ Regne Animal’ under the name of C. d@urvillii, with the following description, which is stated to be from his pen :— “‘ CORONIS D’URVILLII.—Dessus des premicéres ailes d’un brun olivatre, ayant prés de la base et vers le milieu une bande oblique dentée en scie blanchatre, celle de la base plus ou moins violatre, celle du milieu un peu lavée d’olivatre sur son cété interne qui seul est denté, l’extrémité offre prés de la frange une double ligne grisatre, dont la plie externe denticulée: ailes postérieures se terminant par une queue médiocre, un peu spatulée et offrant sur le milieu une bande d’un bleu violet vif, trés-large pres de la coté et finissant en pointe prés de Vangle anal. Dessous des quatres d’un brun olivatre pale, avec une bande blanche sur le milieu de chacune, et l’extrémité dun gris jaunatre. Cayenne; de la collection de M. Boisduval.” Notwithstanding the statement of M. Boisduval that the type was in his collection, M. Guenée, who was allowed by him the unlimited use of his collection, overlooking Boisduval’s description, says that he only knew C. d’urvillit by Latreille’s figure, and that he could not, therefore, describe its colours; whilst he describes C. orithea from “Cayenne, un ¢ Coll. Bdv. Cette magnifique espéce est toujours trés rare.” It is evident, therefore, that M. Boisduval had altered the name of the specimen in his collection from d’urvillii to orithea, probably in consequence of the observation of M. Blanchard in the Crochard edition of the ‘ Regne Animal,’ that the two insects were synonymous. The insect which I have represented in P]. LXXXVII. fig. 3, appears to me to be sufficiently similar to Cramer's figure of the species to allow it to be described as the male thereof, the figures both of Cramer and Boisduval above referred to evidently representing female insects, whilst the two specimens in the Hopeian Collection, one from Columbia (collected by Chesterton) and one from Papagaya (collected by Rogers), are males, having a triangular patch of luteous hairs on the underside of the fore wings near the middle of the posterior margin nearly concealed by the costal portion of the hind wings (and hence overlooked by all writers on these insects), whilst the hind wings have a patch of rough black scales on the upperside between the base and the blue spot. The fore wings above are of a very rich maroon-brown colour, the basal fourth part of the wing varied with very slender purplish white lines, forming several more or less oval dark patches. The outermost of these lines is slightly waved at its junction with the principal veins. At a slight distance beyond the middle the wing is traversed by a distinct, nearly straight, narrow, whitish fascia, extending from the costa, where it is a little dilated, to the anal angle, where its inner margin is a little incurved. This fascia is marked on the costa with two short brown marks, the inner one of which extends in a much more slender condition along the inner edge of the fascia. The 530 PROF. J. O. WESTWOOD ON THE URANIID#. apical margin of the wing is marked with two very slender pale lines, which unite together about half the distance from the apical angle. The first of these lines is angulated near the angle; and the outer one is scalloped. The rich purple oval spot occupying the whole of the outer angle, and a great portion of the disk of the wing, terminates in a narrow flesh-coloured line extending to the anal angle, from which upwards the anal margin of the wing is marked with several short, obscure, transverse marks, whilst between the anal angle and the tail there are three rather large semioval black spots. On the underside of the Columbian specimen the fore wings are uniform brown, the basal third portion as far as the vein closing the discoidal cell being luteous; the costa is narrowly luteous with three small black dots between the end of the cell and the pale fascia, and six similar dots between the fascia and apical mark. A fascia of clear, pale buffish white occupies the place of the band of the upperside, with both its margins slightly scalloped or waved at the veins; the apical margin is pale luteous buff, dilated a little at the apex, and marked with very minute black transverse lines ; this pale border vanishes between the first and second median branches. The hind wings have the basal half luteous and the apical half brown, the former colour extending to within two lines of the anal angle. In the Columbian specimen the pale basal portion terminates in a slight curve, the middle of which is towards the base of the wing ; it is not defined by a distinct paler streak; and the longitudinal veins in the middle of the wing are marked with three or four small luteo-fulvous spots on each. The apical margin of the hind wings is narrowly luteous buff between the tail and outer angle, and bears a somewhat triangular spot between the tail and anal angle. There is considerable difference between the two specimens in the Hopeian Collection. The Columbian one is more elegant in its form, the fore wings of the Papagayan individual are more ovate, and the hind wings shorter. The dark central portion of the fore wings is relatively very much broader in the Columbian specimen, in which the costal margin of the fore wing is unspotted beyond the middle, whereas it bears four pale dots in the other ; the blue patch on the hind wings is of a whiter blue in the Papagayan specimen, and the pale basal portion of the hind wings terminates in a narrow wayed paler fascia preceded by a darker and more distinct waved line, this pale fascia slightly curving outwards in the middle, whilst the pale fascia of the fore wings on the underside is nearly straight and rather broader than in the other. Sp. 2 (35). CoronipIA ERECTHEA, sp. nov. (Plate LXXXVII. fig. 4.) Precedenti similis; differt margine lato apicali fusco-luteo alarum anticarum, macu- laque multo minore elongato-trigona ceruleo-purpurea posticarum, cauda spatulata alba nigro bipunctata. Expans. alar antic. unc. 2. Hab. Brasilia. In Mus. Hopeiano Oxonie et Mus. Brit. PROF. J. 0. WESTWOOD ON THE URANIID&. 531 In its general character this species closely resembles C. orithea. The specimen in the Hopeian Collection is a male, having the fore wings triangular and pointed at the tip, and the oval patch of rough black scales on the upperside of the base of the hind wing, and the tuft of luteous hairs on the hind margin of the underside of the fore wing, whilst two individuals in the British Museum, from Espirito Santo, want these characteristic sexual markings. In these two specimens, however, the antenne are slightly bipectinated, proving the anomalous fact that simple antenne may be characteristic of the males of a species of which the female has those organs more or less pectinated. In the Hopeian specimen the coloured patch of the hind wings is elongate-conic, broadest at the costa, with its sides nearly even and straight, and terminating in a somewhat flesh-coloured streak extending nearly to the anal angle; but in the British- Museum specimens the outer margin of this blue patch is somewhat more scalloped. The terminal joint of the palpi is very short and clavate. On the underside the wings are brown, the basal portion being only slightly tinted with ochreous buff; the fore wings without the pale marks at the base, the central one entire and white with a pearly tinge, the apical margin nearly as on the upper side. The under wings beneath have a slightly defined, central, oblique, waved fascia of palish lilac, preceded towards the base by two slender, brown, dentated, waved lines. The patch of luteous hairs on the underside of the fore wings is covered by the costal margin of the hind wing, and forms a triangular patch lying flat on the wing, and reaching to the first branch of the median vein. Beyond the fascia the hind wing is uniform brown with a small pale triangular patch between the anal angle and the tail, which is brown and tipped with white. _ Sp. 3 (36). Coronipia BOREADA, sp. noy. (Plate LXX XVII. fig. 5.) Preecedentibus similis, differt antennis feemine circiter 80-articulatis, articulis bipec- linatis ; fascia media alarum anticarum intus denticulata, extus recta, e costa ad angulum posticum extensa, et ibi intus curvata, costa parum lutescenti-guttata, margine apicali bilineato, linea interna prope angulum apicis angulata, linea externa bistrigulata et cum precedente pone medium marginis apicalis confluente; alis posticis macula anguste semi-ovata cerulea, in lineam augustiorem carneam ad angulum analem extensa, margine anali luteo-subfasciato, margine apicali inter angulum analem et caudam maculis duabus semiovalibus, et inter hanc et angulum externum margine tenui albido striolam nigram includente. Expans. alar. antic. 23 unc. Hab. Brasilia. In Mus. Hopeiano Oxonie et Britann. The remarkable elongated and bipectinated antenne of the female in this species distinguish it at once from all its congeners. As the wings do not exhibit the secondary sexual characters noticed in the males of the preceding species, 1 am compelled to 532 PROF. J. O. WESTWOOD ON THE URANIIDA. consider the specimens with these antennz females. ‘They are now for the first time noticed ; and I know no other instance amongst Lepidopterous insects in which pectina- tions are developed in the female antenne whilst those of the male are simple. On the underside the wings are brown, buff at the base, the fore wings without basal markings and with a central fascia shaped as on the upperside, white, very slightly tinged with luteous; the apical margin with a sharply marked pale buff edge, slightly scalloped within and finely irrorated with brown scales; the costa indistinctly marked with four dark spots beyond the fascia: the hind wings have a central, strongly scalloped, narrow, pale buff fascia, very oblique, preceded and followed by a slender black waved line, the veins beyond the fascia being dark brown, dotted with minute buff marks; between the anal angle and the tail, and between the tail and the outer angle, the margin of the hind wings is buff, irrorated with brown scales. Sp. 4 (87). CorontDIA HYPHASIS. C. alis fusco-nigris, fascia media continua anticarum recta alba, posticarum curvata albo-cerulea micante; subtus pallidioribus fascia communi angustiore alba lilacino micante (fem.). Expans. alar. antic. unc. 2, lin. 2. Antenne in figura Hopfferi parum serrate videntur. Coronis hyphasis, Hopfter, Lep. nov. pl. Ixxxvii. fig. 2, 3. This species is distinguished by the middle band of the fore wings being more trans- verse, reaching considerably within the posterior angle of the wings, to which it is united by a short marginal extension, so that it is much more continuous with the band of the hind wings than in any of the other blue-banded species. fab. Mexico. Sp. 5 (38). CoronrDIA HYSUDRUS. C. alis fusco-nigris, fascia media non continua anticarum obliqua alba, posticarum subcurvata, cyaneo micante ; subtus fuscis, anticis fascia lilacino-alba, posticis testaceo irroratis (mas. et fem.). Expans. alar. antic. lin. 24, 25. Coronis hysudrus, Hopfter, Lep. nov. pl. Ixxxvii. figs. 4, 5. This species is closely allied to C. boreada, from which it differs in the form of the pale band of the hind wings, and in the structure of the antenne. Hopffer figures a female, but states that specimens of both sexes are in the Berlin Museum. It is therefore evident that, had the antenne in either or both sexes been pectinated, he would have represented or described them so. Hab. Brasilia et Mexico, or oo oo PROF. J. O. WESTWOOD ON THE URANITIDZ. Section 2. Alis posticis rufo fasciatis. Sp. 6 (39). CoRronrpia EGINA. “ Ailes brunes ; les antérieures présentent alternativement des lignes ondulées plus obscures et plus pales et en outre quatre bandes transversales d’un gris blanchatre interrompues par des lignes brunes. La premiére de ces bandes est presque basilaire, la 4™° presque terminale, la 2™° en deca et la 3" audeladumilieu. Les ailes postérieures, plus rembrunies dans leur moitié inférieure que vers la base, ont une large bande rouge atténuée avant darriver au cété interne de l’aile. En outre il existe pres du bord terminal des stries transversales et 3 ou 4 taches noiratres.” Coronis egina, Boisduval, in R. An., Crochard edit., Atlas, Ins. pl. 165. fig. 4; Walker, List Lep. Het. B. M.i. 38; Chenu, Enc. Hist. n. Pap. p. 234, fig. 399; Guenée, Sp. Gén. Lép. ix. p- 21, pl. 1. fig. 3. Hab. Brazil. In Mus. Brit. CoRONIDIA EGINA, var.? (Plate LXX XVII. figs. 6, 7.) I have represented in figs. 6 & 7 an insect from Sao Paulo, Brazil, in the Hopeian Collection, which agrees in nearly every respect with the figure given of this species in the Crochard edition of the ‘Régne Animal;’ but the precise form of the markings of the fore wings is either incorrectly given in that figure, or we must regard the insect before us as distinct. The red band of the hind wings also is more ovate than in the Crochard figure, with its outer edge scalloped ; and it only extends to the middle of the wing instead of arriving in an attenuated form at the anal margin; the tail also is very broadly spatulate in the Hopeian specimen ; the longitudinal lower discoidal vein of the fore wings is also white beyond the fascia, uniting with the white inner edge of the pale apical border, and enclosing therewith a broad dark patch, widest at the costa, which space on the underside of the wing is tinged with rosy. The underside of the wings of this insect is so beautiful that I have given a separate figure of it. If it ’ should ultimately prove to be distinct from C. egina, I propose the name of paulina for it. The specimen measures 23 inches in the expanse of the fore wings, which are rather sharply pointed at the tip, with the hind margin dilated and rounded. The characteristic sexual characters of the wings are wanting. Sp. 7 (40). Coronrp1a cANACE. C. alis fuscis, subtus pallidioribus: anticis supra strigis obliquis luteo-albidis tribus, exteriore rectissima, subtus fascia media obliqua albida rectissima; posticis supra fascia angulata sanguinea, subtus rubescenti-albida (mas et fem.). VOL. x.—PART xII. No. 5.—Jumne 1st, 1879. 4k 584 PROF. J. O. WESTWOOD ON THE URANIID&. C. egine proxima. Coronis canace, Hopfter, Lep. noy. pl. lxxxvii. fig. 6. Hab. Brazil. Sp. 8 (41). Coronrpia COLUMBIANA, sp. nov. (Plate LXX XVIII. fig. 4.) (. alis anticis magnis ovatis fuscis, strigis numerosissimis obscurioribus obliquis plus minusve undulatis fasciisque 4 subalbidis, 1ma angusta obliqua prope basin, 2nda ante medium, linea tenui fusca interna notata, 3tia pone medium marginibus ejus parum undulatis, ad angulum posticum extensa et ibi incurva, 4ta ante apicem et cum margine apicali, supra angulum posticum, confluente, lunulis subapicalibus nigris ; alis posticis fuscis, fascia curvata submedia extus lobulata, obscure punicea seu chermesina, e medio ale ad angulum analem augustata, obscuriore et recurva, margine apicali inter caudam et angulum analem maculis magnis semiovalibus nigris, et inter caudam et angulum externum pallide fusco-lineato. Alis anticis subtus basi pallidis, macula nigra ad apicem cellule discoidalis, fasciis duabus pagine superioris obliteratis, spatio inter fascias 3tiam et 4tam obscuriores fusco, haud undulato ; alis posticis, basi pallidis, fascia media latiori et ad angulum analem late extensa, margineque postico etiam maculis continuis pallide chermesinis notato. Expans. alar. antic. unc. 25. Hab. Columbia. In Mus. Hopeiano Oxonie. The specimen in the Hopeian Collection, obtained from M. Depuiset, appears to be a female, being destitute of the patch of scales on the hind wings and the tuft of hairs on the fore wings. It is closely allied to C. canace, which inhabits Brazil. Sp. 9 (42). CoRonrpIA NICARAGUANA, sp. nov. (Plate LXXXVIII. fig. 3.) C. alis supra fuscis, anticis acute trigonis, strigis numerosis plus minusve undulatis nigris, lineola tenui subbasali, fascia ante medium alteraque undulata pone medium obscure luteo-fuscis et lineolis undulate notatis, margine apicali paullo pallidiore, intus subundato; alis posticis fuscis, dimidio externo obscurioribus, longe pone medium fascia intus recta et extus triloba, sanguinea e margine costali ad medium ale extensa, spatio inter medium et angulum analem obscurius undulato, margine postico inter caudam et angulum analem maculis semiovalibus nigris notato et inter caudam et angulum externum lineis fulvis nigrisque undulato, cauda spatulata macula ovali nigra notata. Expans. alar. unc. 24. Hab. Nicaragua (Belt). In Mus. Hopeiano Oxonie. The fore wings are of a much more triangular and acute form than in C. canace. The tuft of hair and patch of scales are wanting ; but the shape of the wings induces me to regard the Hopeian specimen as a male. The underside closely resembles that of PROF. J. O. WESTWOOD ON THE URANIIDA. 535 C. paulina, except that the central fascia is more elbowed and undulated towards the costa, the pale subapical margin is more irregular, and the middle fascia of the hind wings is light rosy ferruginous, with the veins crossing the fascia whitish, and a minute dot of white on each of the black spots of the tail. Sp. 10 (43). Coronipia £01, sp. nov. (Plate LX XXVIII. figs. 1, 2.) C. testaceo-fusca, alis anticis striolis obscuris numerosis transverse undulatis, fascia transversa albida tenuissima prope basin alarum, 2nda ante medium paullo curvata angusta, 3tia pone medium fere recta e costa extus integra, intus denticulata, ad angulum posticum extensa et intus striola media fusca notata, inter fascias 2dam et 3tiam serie transversa ocellorum nigrorum ; lineola abrupte angulata prope apicem alarum e fascia tertia ducta, serieque apicali lunularum nigrarum; alis posticis basi pallide fuscis, dimidio externo obscurioribus, fascia sanguinea e costa (ultra medium) in medio ale recurva et ibi obscuriore et angustiore marginem analem attingente, spatio pone fasciam obscurius undulato, margine postico inter angulum analem et caudam nigro maculato, inter caudam et angulum externum albido anguste lunulato, cauda late spatulata nigro ocellata: alis infra basi pallidis, singula macula parva obscura in cellula discoidali, anticis fascia recta pone medium alba, dimidioque antico marginis apicalis pallido ; alis posticis fascia submedia pallidiore sanguinea intus nigro marginata et antice et postice striolis obscuris undulatis numerosis notatis. Expans. alar. antic. unc, 22-3. Hab. West Indies, Guatemala, Brazil. In Mus. Hopeiano Oxonie. The four specimens of this species in the Hopeian Collection described above and illustrated in the accompanying figure, appear to be females, and are destitute of the patches of hair and scales distinctive of the males of some, at least, of the species of this genus. The straight pale middle fascia of the fore wings separates it from C. paulina, rosina, and nicaraguana, Sp. 11 (44). Coronipia Rosina. (Plate LXXXVII. figs. 8, 9.) C. alis fuscis, anticis strigis numerosis augustis transversis parum undulatis nigris, fascia tenui prope basin transversa, 2da ante medium recta integra et bilineata, 3tia pone medium subundulata atque etiam intus fusco-bilineata lineolaque subapicali albida, strigisque marginalibus subpallidis angustis; alis posticis fascia submedia dimidiata puniceo-rufa antice fere recta, postice undata, e costa ad medium ale extensa et ibi anguste incurva multo obscuriore et retro directa, spatio pone fasciam ad angulum externum nigricante, ad angulum analem pallidiore et tenuissime undulato, maculis submarginalibus inter angulum analem et caudam nigris anguloque externo albido, cauda spatulata lutea macula ocellari nigra. Alis infra basi pallide luteo fuscis, anticis macula parva subocellata in cellula discoidali, fascia curvata paullo pone medium 4n 2 536 PROF. J. O. WESTWOOD ON THE URANIID#, alteraque cum margine apicali parallela albidis ; alis posticis fascia rufa media magis continua alteraque minus distincta inter hanc et marginem posticum, cauda nigra apice albo. Expans. alar. antic. unc. 25. Larunda rosina, Felder, Novara, tab. cxxxi. figs. 3, 4 (sine descriptione). Hab. Bogota. No description of this species has, I believe, hitherto been published by Dr. Felder ; but his characteristic figure leads me to infer that the insect represented in my figure, from a specimen in the Hopeian Collection, obtained from M. Depuiset, is identical. This specimen does not exhibit the tuft of hairs or patch of scales, although from its general appearance I suppose it to be a male. Section 3. Alis posticis fulvo fasciatis. Sp. 12 (45). Coronipia ECHENAIS. C. alis utrinque fuscis, anticis supra lineis transversis undulatis obscurioribus et pallidioribus alternis, medio nigro-fusco subfasciatis, subtus testaceo-marginatis, medio albo-fasciatis; posticis supra fascia media antice flava postice brunnea, subtus tota rubenti-flava, nigro marginata (mas). Expans. alar. antic. unc. 2, lin. 4. Coronis echenais, Hopffer, Lep. nov. pl. Ixxxviil. fig. 1. Hab. Mexico. C. ECHENAIS, var.? (C. granadina, Westw., provis.) (Plate LXX XVIII. fig. 5.) C. alis pallide fuscis, anticis supra fusco valde undulatis, fascia valde undulata media nigricante, pone medium pallidiore at sensim versus apicem obscuriore, margine apicali pallidiore, intus albido-denticulato ; alis posticis in medio obscuris, fascia undulata pone medium late fulva, inter medium et marginem analem fere obliterata, maculis margi- nalibus nigris inter caudam et angulum analem, cauda etiam macula angulata nigra notata: alis anticis subtus basi pallide fuscis, puncto obscuro in cellula discoidali, fascia media undata albida, intus fusco late marginata, pone fasciam fuscis margine lutescenti intus undato ; alis posticis subtus magis fulvescentibus, puncto in cellula, fascia fusca irregulari media, pone fasciam fulvis, extus magis brunneis in fuscum vergentibus, margine fulvo. Expans. alar. antic. fere unc. 23. Hab. Nova Granada. In Mus. Hopeiano Oxonie. The gradually dark subapical mark of the fore wings and the different shape of the fulvous fascia of the hind wings seem to separate this insect from the smaller Mexican species figured by Hopffer. The Hopeian specimen is, I presume, a female, having neither patches of hair nor of scales on the wings. PROF. J. O. WESTWOOD ON THE URANIIDZ. 537 C. ECHENAIS, var. (C. subpicta, Wk.) C. nigro-fusca: alis anticis lineis transversis undulatis obscurioribus et pallidioribus alternis, puncto subcostali ante medium fasciaque nigro-fuscis, subtus basi testacea, fasciisque duabus albido-testaceis; alis posticis fascia brevi obliqua undulata lutea, postice ferrugineis lineis obliquis angulatis fuscis ocellisque caudalibus. Long. corp. lin. 11; expans. alar. antic. lin. 34. Coronis subpicta, Walk. List Lep. Het. B. M. i. 39; Butler, Ill. Lep. Het. B. M. pl. 3. fig. 2. Hab. Venezuela (Dyson). In Mus. Brit. This variety of C. echenais, as I regard it, is much darker, and suffused with brown on the fore wings; the hind wings also are darker; and the fulvous band is obliterated, except in the middle of the wing; but the characteristic markings are identical in both. Sp. 13 (46). CoronrpIA BIBLINA, sp. nov. (Plate LK XXVIII. fig. 7.) C. alis anticis supra nigro-fuscis, obscurius valde undulatis, fascia alba pone medium alarum, e costa ad angulum posticum extensa, fere recta, ante medium ale tamen extus paullo curvata, intus presertim ad costam obscurius guttata, margine apicali lunulis fulvis obscuris notato ; alis posticis fuscis fascia angusta lete aurantiaca extus castanea in medio valde curvata, pone medium notatis, puncto pallido ad angulum externum, cauda spatulata alba ocello nigro ad basin albo biguttato notata, lunulisque nigris inter caudam et angulum analem: alis anticis infra fuscis basi pallidioribus, puncto parvo obscuro discoidali, fascia alba ut in pagina superiore, margine pallido in dimidio antico alarum; alis posticis ut supra coloratis, fasciis presertim castaneis magis distinctis. Expans. alar. antic. unc. 23. Hab. Nicaragua, Venezuela, In Mus. Brit. et Hopeiano Oxonie. The large rounded wings of this species, destitute of the sexual patches of hairs and scales, indicate the specimens I have hitherto seen to be females. Sp. 14 (47). Coronipia Jaret. (Plate LX XXVIII. fig. 6.) “Les quatre ailes d'un brun grisdtre; les antérieures présentent plusieurs lignes ondulées plus obscures, et deux bandes transversales blanchatres, l'un prés de la base, l'autre au dela du milieu. Les ailes postérieures, plus obscures, vers leur extrémité, ont une bande transversale ondulée d’un jaune orangé, et une rangée de taches noires circlées de grisdtre prés du bord terminal.’’ Hab. Brazil. Coronis japet, Boisduval in ‘Régne Animal,’ ed. Crochard, Ins. pl. 145. fig. 3; Chenu, Enc. Hist. n. Pap. 234, p. 400; Walker, List Lep. Het. B. M.i. 38. Coronis leachii 2, sec. Guenée, Sp. Gén. ix. p. 22. 538 PROF. J. O. WESTWOOD ON THE URANIIDZ. The Hopeian Collection possesses five specimens which I regard as belonging to this species, apparently of both sexes, from Brazil and Parana (Papagaya), and Nicaragua (the latter collected by Mr. Belt). One of these, apparently a male (although destitute of the sexual patches of scales and hairs), is represented in fig. 6. The central fascia is nearly straight, being only curved a little outwardly rather in front of the middle of the wing; the apical margin is pale and very irregular, marked, however, below the middle with two large conical pale spots. On the underside the basal half of the fore wings is pale greyish brown with a dark dot in the middle of the discoidal cell, the basal fascia is wanting, the middle very distinct and nearly white, the anterior half of the apical margin with a pale brown patch ; the hind wings are rich brown, pale at the base, with the fascia of a pale rosy buff colour, preceded and followed by pale undulated lines, especially marked on the veins; the outer and anal angles are marked with white. Although there is considerable general resemblance between this species and C. leachit, I cannot agree with M. Guenée in regarding these fulvous banded individuals as the females of the black C. leachii. The shape of the band of the fore wings is of itself quite sufficient (independent of the want of the band of the hind wings) to separate the two species. Sp. 15 (48). CoronipiA BRISEIS, sp. nov. (Plate LXX XVIII. fig. 9.) C. alis pallide fuscis, anticis prope basin lineolis 4 angulatis fuscis per paria dispositis, tunc fascia obliqua albida extus nigro marginata, spatio medio obscure fusco prope costam valde dilatato extus nigricante, tunc fascia lata albida, fuscescenti parum nebulosa, striolisque fuscis parum undulatis sequentibus, margine irregulari pallidiore pone medium in maculas tres conicas intus dilatato ; alis posticis fuscis fascia submedia lata parum curvata flavida, angulo externo apiceque caude albo notatis, hac nigro bimaculata ; alis infra coloribus magis uniformibus, fascia media anticarum magis distincta, venis posticarum punctis fuscis et albidis obscurius notatis. Expans. alar. antic. unc. 2. Hab. t In Mus. Hopeiano Oxonie. Sp. 16 (49). CoRonipIa INTERLINEATA. (Plate LXX XVIII. fig. 8.) Fusca, subtus testacea; antennis fulvis; alis anticis lineis alternis obscurioribus et pallidioribus transversis undulatis, quatuorque distinctioribus albidis, duabus basalibus angulatis, 3tia undulata, 4ta subapicali angulata; alis posticis fascia obliqua undulata lutea, subtus fusco trifasciatis. Long. corp. lin. 10; expans. alar. antic. lin. 29. Coronis interlineata, Wik. List Lep. Het. B. M. i. 38; Butler, Ill. Lep. Het. B. M. pl. 3. fig. 1. Hab. ——?* In Mus. Brit. Bogota vel Venezuela. In Mus. Hopeiano Oxonie. The Hopeian Museum possesses a specimen of this species presented to me by the late Dr Kaden, from Bogota or Venezuela. The general colour of the upperside of PROF. J. O. WESTWOOD ON THE URANIIDA. 539 the fore wings is pale brown, very much varied with lighter and darker very undulating and angulated lines, which form several macular spots rather before the middle of the wings, and with a very much curved whitish line terminating the central darker part of the wing; the apical margin is pale and marked with darker lines, and with three prominent conical spots between the middle of the margin and the posterior angle, preceded by dark reversed-conical dots. The hind wings are pale brown at the base. On the underside the fore wings have the base quite pale brown and uniform, with two square dark brown spots in the middle towards the costa, separated by a light fulvous one, and followed by the very much curved fulvous fascia; the apical margin also fulvous, with the conical spots strongly marked; the hind wings are fulvous, with a dark fuscous fascia in the middle. All the veins are marked with blackish dots. Expansion of the fore wings 1 inch. Sp. 17 (50). CoRonIDIA GENEVANA, sp. noy. (Plate LXXXVIII. fig. 10.) C. alis supra fuscis, anticis puncto parvo nigro versus basin strigaque transversa nigra ad apicem cellule discoidalis, pone medium lineis duabus approximatis gracilibus valde undatis, spatio sequente fulvescente, venis nigro punctatis, apice alarum obscuriore ; alis posticis fuscis paullo obscurius variegatis, fascia luteo-fulva inter medium et angulum externum ad medium al tantum extensa, cauda brevi conica, ad basin ocello parvo nigro notata: alis anticis infra pallide fuscis, guttis quatuor nigris in medio coste punctoque parvo stigmaticali nigra extus, striga pallidiore in medio ale, fascia pallida luteo-fulva pone medium, spatio apicali pallide fusco obscurius undulato versus costam et undulis fasciam angustiorem et obscurlorem versus angulum posticum formantibus ; alis posticis pallide luteo-albidis, costa lineis obscurioribus et curvatis, et fascia valde undulata medium ale occupante e lineis fuscis formata, margine parum pallidiore guttis hastatis in venis notato. Expans. alar. unc. 2. Hab. Mexico. In Mus. Geneve. Sp. 18 (51). Coronipia ocytus, Boisduval, Consid. Lépid. Guatemala, p. 76. ‘ Port et taille de notre Coronis japet figurée par M. Emile Blanchard. Dessus des ailes d’un brun clair, traversées au dela du milieu par une bande commune d’un blanc jaunatre sur les supérieures et d’un jaune d’ocre sur les inférieures ; de chaque coté de cette bande, la teinte générale est presque noiratre et l’extrémité marginale d’un gris cendré. Les secondes ailes ont en outre un point noir 4 la base de leur appendice caudal ; dessous grisdtre, celui des supérieures avec un point discoidal noir et deux bandes jaunatres. Une seule femelle de Guatemala. Nous possédons un exemplaire un peu plus petit venant du Mexique. II est possible que le male, que nous ne connaissons pas, différe beaucoup de la femelle.” 540 PROF. J. O. WESTWOOD ON THE URANIIDZ. Sp. 19 (52). Coronipia DUCATRIX. Coronis ducatrix, Schaufuss, Nunq. Otios. i. p. 12. “*Oben heller braun als Evenus, Oberfliigel mit zwei fahlen, etwas gezackten, linirten Querbinden, welche nach aussen zu dunkel abschattirt sind ; Rand mit heller Querwolken nach aufwarts, jede vorn mit dunklem irregulirem Strich begrenzt; Unterfliigel mit querer eckig ausgebogener, gebuchteter orange-gelber Binde ; Aussenrand mit gelber und schwarzer Binde umgeben, an die beiden ausseren Enden weiss gefranzt, nach unten mit vier runden oder halbmondférmigen schwarz brauen Flecken. Unten der C. leachi, God., ganz ahnlich, nur die Unterfliigel mit weisslicher, rosa angeflogener Binde.” War mit dem Namen Inge Moritz bezeichnet. Lat. 43-56 mm. Hab. Venezuela. Section 4. Alis posticis haud fasciatis. Sp. 20 (53). Coronipia LEacn. (Plate LX XXVIII. fig. 11.) C. nigro-fusca: alis anticis lineis transversis undulatis pallidioribus, fasciaque recta alba in anticis discoidali, in posticis marginali, anticis strigis anticis subapicalibus rufescentibus, subtus basim versus pallidioribus, gutta costali nigro-fusca ; alis posticis ocellis caudalibus, subtus linea transversa valde undulata testacea apud venas albida. Coronis leachii, Godart, Enc. Méth. ix. 803; Guérin, Icon. R. An. Ins. pl. 83. fig. 3; Griffith, Anim. Kingd. Ins. pl. 53. fig. 8; Boisduval, Spec. Gén. Lép.i. pl. 14 (10 B). fig. 2; Walk. List B.M. Lep. Het. i. 39; Guenée, Sp. Gén. ix. p. 22, pl. 1. fig. 2. Hab. Mexico (Hartweg). Spirito Santo, Brasilia. In Mus. Brit. et Hopeiano, Oxonie. M. Guenée considers this species to consist of males, of which ©. japet are the females. Of four specimens in the Hopeian Collection, from Brazil and Guatemala, all are destitute of the sexual patches of hair and scales; the size of the abdomen nevertheless indicates them to be males. The dark-coloured central portion of the wings is much broader towards the hinder margin than in C. japet. Its outline also on both sides is much more irregular. The white patch at the outer angle of the hind wings is also much larger. Sp. 21 (64). CoronipIA EVENUS. ‘* Ailes entiérement d’un brun assez foncé, les supérieures presentant dans toute leur étendue des lignes transversales ondulées, et dans leur milieu il en existe deux qui circonscrivent un espace plus foncé. Les ailes postérieures offrent au bord terminal des taches semilunaires et sur l’appendice cordiforme une antre tache arrondie. Expans. alar, antic. une. 2, lin. 2°3. PROF. J. 0. WESTWOOD ON THE URANIIDZ. 541 C. evenus, Boisduval in Cuv. R. An. ed. Crochard, Ins. pl. 145. fig. 2; Chenu, Enc. Hist. n. Pap. 234, fig.401; Walker, List Lep. Het. B. M.i. 40; Guenée, Sp. Gén. ix. p. 23. “ C, leachii, var. ??’? Walker, J. c. Hab. Venezuela (Dyson). In Mus. Brit. et Hopeiano, Oxoniz. This species differs from C. Jeachii in the uniform nearly black colour of the hind wings, without any white patch at the outer angle, and by the more distinct pale apical margin of the fore wings, dilated into two or three conical pale patches towards the posterior angle. The space preceding the pale apical margin is also much darker than in C. leachii. A specimen in the Hopeian Museum is either from Venezuela or New Granada. Norr.—Of C. ducalis, Schaufuss (Nunquam Otiosus, i. p. 12), from Venezuela, and C. dutreuati, E. Deyrolle, in Rev. Zool., from Costa Rica, no descriptions have been published. EXPLANATION OF THE PLATES. PLATE LXXXV. Fig. 1. Larva of Coronidia (sp. allied to C. orithea), pp. 521, 528. Fig. 2. Pupa of ditto. Fig. 3. Palpus of Coronidia eola. Fig. 4. Fore leg of ditto. Fig. 5. Middle leg of ditto. Fig. 6. Hind leg of ditto. Fig. 7. Fore wing of ditto. 8 Fig. 8. Hind wing of ditto. Fig. 9. Head and antenne of Coronidia boreada. Fig. 10. Head and antenn of Manidia lunus, p. 521. Fig. 11. Fore wing of ditto. Fig. 12. Hind wing of ditto. Fig. 13. Fore wing of Uranidia sloanus, pp. 520, 521. Fig. 14. Hind wing of ditto. Fig. 16. Fore wing of Chrysiridia rhipheus, pp. 520, 522. Fig. 16. Hind wing of ditto. Fig. 17. Fore wing of Alcidia orontes, pp. 520, 524. Fig. 18. Hind wing of ditto. Fig. 19. Fore wing of Lyssidia patroclus, pp. 520, 525. Fig. 20. Hind wing of ditto. VOL. X.—ParT xu. No. 6.—June 1st, 1879. 4r 542 Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. fo) Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. PROF. J. 0. WESTWOOD ON THE URANIID#. PLATE LXXXVI. 1. Fore wing of Erebus (Patula) macrops, p. 512. 2. Hind wing of ditto (male). 3. Hind wing of ditto (female). 4, Head of ditto. 5. Fore wing of Urapterya sambucaria, p. 514. 6. Hind wing of ditto. 7. Fore wing of Hydria undulata (from Packard), p. 513. 8. Fore wing of Strophidia vollenhovit, p. 514. 9. Hind wing of ditto. 10. Wings of Asthenidia podaliriaria, p. 515. 11. Fore wing of Attacus pavonia minor, p. 517. 12. Fore wing of Endromis versicolor, p. 517. 13. Hind wing of Eudemonia semiramis, p. 517. 14. Fore wing of Aglaia tau, p. 517. 15. Fore wing of Actias luna, p. 517. 16. Hind wing of ditto. 17. Fore wing of Gastropacha quercifolia, p. 518. 18. Wings of Epicopeia polydora, p. 518. PLATE LXXXVII. 1. Alcidia boops, male, p. 525. Fig. 7. Coronidia egina, underside, p. 533. 2. Alcidia metaurus, p. 528. Fig. 8. Coronidia rosina, p. 535. 3. Coronidia orithea, var., p. 628. Fig. 9. Coronidia rosina, underside. 4, Coronidia erecthea, p. 530. 5. Coronidia boreada, p. 531. 6. Coronidia paulina, p. 533. PLATE LXXXVIII. 1. Coronidia eola, p. 535. Fig. 7. Coronidia biblina, p. 537. 2. Coronidia cola, underside, p. 535. Fig. 8. Coronidia interlineata, p. 538. 3. Coronidia nicaraguana, p. 534. Fig. 9. Coronidia briseis, p. 538. 4. Coronidia columbiana, p. 534. Fig. 10. Coronidia genevana, p. 539. 5. Coronidia granadina, p. 536. Fig. 11. Coronidia leachii, p. 540. 6. Coronidia japet, p. 537. JOW. del RHF Rippon hth. Hanhart imp SYNOPSIS OF THE URANIIDA. JW. del RHFRippon hth Hanhart mp SYNOPSIS OF THE URANIIDA ar = Rn & } A Pe iN) NGS NS JOM del Hanhart imp Koppon hth OY Oy (eZ OK PW W ITIVVVIV7 UT VOU U J.OW. del a RHE Rippon, lth Hanhart imp SYNOPSIS OF THE URANIIDA [ 543 J XVI. Supplementary Notes on the Curassows now or lately living in the Society's Gardens. By P. L. Scuarer, W.A., Ph.D., F.R.S., Secretary to the Society. Read June 18th, 1878. [Pirates LXXXIX.-XCV.] SINCE I read my paper on the Curassows living in the Society’s Gardens, five years ago!, many additional examples of these birds have been received by the Society. From this source, and from the examination of specimens living in other Gardens, as also from the study of examples in several Collections and Museums, I am enabled to offer a few additional remarks, which, I trust, may tend to elucidate further the history of this difficult group. 1. Crax GLopicera. (Plate LXXXIX.) Crax globicera, Linn. Scl. Trans. Zool. Soe. ix. p. 274. T have already mentioned the fact that at Panama the male of this species shows a slight white terminal margin to the tail, thereby approaching C. daubentoni. The female of the Panama form is likewise different from the same sex of the more northern bird, in having the tail blackish brown strongly barred with six or seven light cross bands, and terminated bya similar margin. ‘This is shown in Mr. Smit’s drawing (Pl. LXXXIX.), taken from a skin in the collection of Messrs. Salvin and Godman. In a skin from Costa Rica in the same collection the bands are likewise well defined, but much narrower. There are likewise examples of both sexes of the Panama form in the gallery of the British Museum, and in the Society’s living Collection. The latter are said to have been obtained at Cartagena, which is the most southern locality yet recorded for this species. 2. CRAX ERYTHROGNATHA. (Plate XC.) Craz erythrognatha, Scl. et Saly. P.Z.S. 1877, p. 22. Purpurascenti-nigra, ventre imo crissoque albis; cristee plumis nigris, recurvis ; loris nudis; cera et rostro basali rubris; hujus apice flavicante; pedibus in pelle sicca pallide corylinis: long. tota 28:0, ala 16:0, caude 11:0, tarsi 4-5, Fem. mari similis, sed criste plumis fasciis minutis albis variegatis. Hab. Columbiam interiorem. Mus. 8.-G. Obs. Similis C. alectori et omnino ejusdem forme, sed rostro rubro diversa. 1 See Trans. Zool. Soc. vol. ix. p. 273. VoL. X.—PART xu. No. 1.—October 1st, 1879. 46 544 MR. P. L. SCLATER ON THE CURASSOWS This near ally of C. alector was recently described from a pair of “ Bogota” skins in the Collection of Messrs. Salvin and Godman. As in that species, the female (Pl. XC. fig. 2) seems to be only distinguishable from the male by the slightly banded crest. I regret to say I have no further information to offer respecting this little-known bird. 3. CRAX GLOBULOSA. (Plate XCI.) Crazx globulosa, Spix. Scl. Trans. Zool. Soe. ix. p. 279. On the 15th of October last we purchased, along with other birds, a Curassow, of which I exhibit a drawing (Pl. XCI.), said to have been obtained from the Amazons’. I have referred this bird provisionally to the female of Craax globulosa of Spix ; but it differs from Spix’s figure (Av. Bras. ii. t. 66), as likewise from the female of the same species which I formerly examined in Mr. Lawrence’s collection (cf. Trans. Zool. Soc. ix. p. 279), in having the whole bill black, and in possessing white markings to the crest, as in the female of Crax alector. This bird is still alive in the Gardens, though in bad plumage and with the toes defective, as is too often the case with these birds in captivity. The legs are red, the bare space round the eye dark bluish black ; the irides are light brown. 4. CRrax vinipiRostRis. (Plate XCII.) Crazx viridirostris, Sclaier, Trans. Zool. Soc. ix. p. 282. Nitenti-nigra, ventre imo cum crisso et caud apice albis; criste plumis brevibus, exstantibus, retroflexis ; rostri cera tumida et carunculis mandibule inferioris pallide viridescenti-flavis : long. tota 32-0, ale 15-0, caudee 13-0, tarsi 4:5. I now exhibit the skin of the Curassow which I described under the above- mentioned name in my former paper. Unfortunately I have not access to a male example of Orax alberti to compare it with; but that is unquestionably its nearest ally, and it is possible that the change of the cere from blue into greenish-yellow may be an individual variation. 5. Crax 1ncommMopa. (Plate XCIII.) Craz incommoda, Scl. Trans. Zool. Soe. ix. p. 281. On October 30th, 1873, we purchased, along with other South-American Mammals and birds, what appears to be a second example of the singular Curassow on which I bestowed the name above given. It lived until December last. Its skin, as likewise that of the former example, is now in the collection of Messrs. Salvin and Godman ? See Report of Z. 8. for 1877, p. 38. LIVING IN THE SOCIETY'S GARDENS. 545 This specimen, as will be seen by the figure of it, is not very different from the former, but has more white on the under surface, the white cross markings being continued all over the breast and chest up to the chin. The lower belly and vent are also slightly tinged with fawn-colour. The sex was ascertained by Prof. Garrod, on dissection, to be female, as in the former specimen ; so that we are still ignorant of the male of this species. 6. Noruocrax urRuMuTUM. (Plate XCIV.) Nothocrax urumutum (Spix). Scl. Trans. Zool. Soc. ix. p. 282 et P. Z.S. 1877, p. 681. In July last (1877), we obtained our first living example of this rare bird, which was said to have been brought down from the Upper Amazons. As will be seen by Mr. Smit’s figure (Pl. XCIV.), taken from the living bird, the colour of the naked skin round the eye was incorrectly given in the former plate (vol. ix. Pl. L.). The upper part of it, above and in front of the eye, instead of being yellow, like the lower portion, is of a bright yellow. The colour of the legs is pale flesh-colour in the living bird. [There was a specimen of this Curassow living in the Jardin des Plantes at Paris, June 1879, which was similarly coloured. | Read February 4th, 1879. 7. Mirva satvini. (Plate XCV.) Mitua salvini, Reinhardt, Vid. Medd. For. Kjébenh. 1879-80, p. ; Sclater, P. Z. 8. 1879, p. 108. Nigra purpureo nitens, ventre imo et caudz apice albis; pilei plumis elongatis sicut in WW. tuberosa jacentibus; loris et capitis lateribus dense plumosis; rostro sicut in M. tomentosa formato, sed paulo longiore et minus alto, toto rubro; pedibus rubris: long. tota circ. 2°10, ale 15, caudee 12°5, tarsi 4°7. Hab, Brazil, prov. Bahia ? Obs. Sp. ventre albo satis distincta, quoad rostrum ad J/. tomentosam, quoad cristam magis ad MW. tuberosam appropinquans. Through the kindness of Prof. Reinhardt I am able to add to my paper on the Curassows, read in June last, a figure of the fine new species of Mitwa which he has lately described from a specimen formerly living in the Zoological Gardens of Copenhagen. In 1876 Prof. Reinhardt first called my attention to this Curassow, which had been received in April of that year from the Danish Consul at Bahia. The bird having died at the end of 1878, was presented to the Museum of the University. It was ascertained by dissection to be a female, and was necessarily quite adult, having been nearly three years in captivity. The male would probably not be different in colour, the sexes being alike in the two other species of Mztua. Unfortunately the exact locality of this Curassow could not be ascertained. It is not likely that it was from the immediate vicinity of Bahia, as that is one of the 546 MR. P. L. SCLATER ON THE CURASSOWS. best-known portions of South America. Jfitua salvini is probably a native of one of the less-known districts of the interior of Brazil. DESCRIPTION OF THE PLATES. PLATE LXXXIX. Crax globicera, female of the Panama form, from a skin in Messrs. Salvin and Godman’s collection. PLATE XC. Crax erythrognatha, male and head of female, from specimens in the same collection. PLATE XCI. Crax globulosa, female, from an example living in the Society’s Gardens (purchased October 15, 1877). PLATE XCII. Craz viridirostris, from the typical specimen, now in Messrs. Salvin and Godman’s collection. PLATE XCIII. Crax incommoda, female, from a living example, received October 30, 1873. PLATE XCIV. Nothocrax urumutum, from an example living in the Gardens, received July 1877. PLATE XCV. Mitua salvini, female, from the typical example in the Museum of the University of Copenhagen. JSmit delet lith thart u Oo nn) & N.Hanhart Ar N Lith J.Smat pF ” Poon f H Hanhart imp J Smit hth IDA i¢ CRAX INCOMM } lanhard anor ee : anhar OF JSrmit hth | te = a i < in > ra —_ Soa . Fr —" a 7 a 54T LIST OF THE PAPERS CONTAINED IN VOL. X. Brapy, GxrorcE Srewarpson, M.D., F.L.S., C.M.Z.S., &e. A Monograph of the Ostracoda of the Ant- Werpi Crags isrste,stisisie cudiecners anvspscuese Busk, Groner, F.R.S., F.Z.8., &e. On the Ancient or Quaternary Fauna of Gibraltar, as exemplified in the Mamma- lian Remains of the Ossiferous Breccia . Duncan, Professor P. Martin, M.B. (Lond.), F.R.S., &e. A Description of the Madreporaria dredged up during the Expedition of H.M.S. ‘Porcupine’ in 1869 and 1870. Part IT. Frower, Wri11am Henry, F.R.S., V.P.Z.S., &e. A further Contribution to the Knowledge of the existing Ziphioid Whales. Genus Mesoplodon Garrop, A. H., M.A., F.R.S., &e. Notes on the Manatee (Manatus americanus) recently living in the Society’s Gardens On the Brain of the Sumatran Rhinoceros (Ceratorhinus sumatrensis) Geppes, Parricr. On the Mechanism of the Odontophore in certain Mollusca : Lannester, FE. Ray, M.A., F.R.S., &e. On the Hearts of Ceratodus, Protopterus, and Chimera, with an Account of unde- scribed Pocket Valves in the Conus arte- riosus of Ceratodus and of Protopterus. . VoL. X.—PaRT xt. No. 2.—October 1st, 1879. Page 235 415 137 411 485 493 | Mivaxt, Sr. Groner, F.R.S., V.P.Z.S., &e. On the Axial Skeleton of the Struthionide On the Axial Skeleton of the Pelecanide . Notes on the Fins of Elasmobranchs, with Considerations on the Nature and Homo- logues of Vertebrate Limbs Owen, Professor, C.B., F.R.S., F.Z.S., &e. On Dinornis (Part XXI.): containing a Restoration of the Skeleton of Dinornis maximus, Owen. With an Appendix, on Additional Evidence of the Genus Dro- mornis in Australia Parker, Wioiram Kircuen, F.R.S., F.Z.8., &e. On the Structure and Development of the Skull in Sharks and Skates .......... On the Skull of the Agithognathous Birds. Scrarer, Puiip Luriey, M.A., Ph.D., F.RS., &e. Supplementary Notes on the Curassows now or lately living in the Society’s Gardens Westwoop, J. O., M.A., F.LS., &e. Observations on the Uraniidx, a Family of Lepidopterous Insects, with a Synopsis of the Family and a Monograph of Coro- nidia, one of the Genera of which it is composed Page 1 315 439 147 189 251 543 iia — ¥ WOW Bl GARTATHOD eiSdA® ABT B0- Tell :) a Ma wt ' ar = : ey ERAT EET eine ae: Papas iE eat Ae srtinaaivo sweat ail i= ay alder: A malt geal fe TH oth = ri wh ASSES Sel LS. abel ld Fey RES, Lota dt et) © 2b sity seca att-to, dart i Psd Bityete baie” | by-tS ir. i o gett ine wf gin LS aa Anger’ 1! IE oo P fee A eae ananad 9 = a “ar ee tine pete Ae eae’ ea =n He Bie : waa Dv ae aU eae, ryotae. nati Hh ld of tail So29 "sn sotbenlif): ee : pce ot VF Segal TAX Panes ahihratl it? ee iN -L gery wad eel wild Nee ex liegelt aE © a goa petal he Or eas rr pe aa vl retanh hs ae i a rhage: a3 ct eT, Mirisars ee Bie ee | - ie es. en) an ne ata = Wo 4 é . 7 > ie rae 2 5, : +r ay a - ‘ ; _ " - fe MF in walt eh dc 4 mt ei i : a i Shes zl 3 rf ; ‘ At iV iy z a ih ey Mii; we i i ox - ¥ 7.) ARP ean ag oF ee ae i : . 5 = ‘ A rm ss . Ae fy ad . _ wep me | mm ; wee S a 2 er a ’ a” « ; f “eds a An * : nM 7 es _ a — ' ~~ & 7 a a . Py ‘ | 1 # ng i ofa *) > 5 TRANSACTIONS OF THE ZOOLOGICAL SOCIETY ie OF LONDON. Voi. X.—Parr 1. LONDON: PRINTED FOR THE SOCIETY : SOLD AT THEIR HOUSE IN HANOVER-SQUARE; AND BY MESSRS. LONGMANS, GREEN, READER AND DYER, PATERNOSTER-ROW. May 1st, 1877. Price 12s. ‘Taylor and Francis, Printers. ] [Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. : To Fellows. To the Public. £s. d. ee A VOLUME I. (1833-1835, containing 59 Plates) . . . Price 313 6 . . . 418 O* Pain 12? (VSB) % tains | 3. Sea ee eat feet ren ct ey a DAO POR tea eae » 2. (1834)* x ; BE Osh (ISSO) (As o% fy ico Saka tienuhe Ue atin os ae ean ee OLE SL O avian hall are ean) Be Br (h:siti) peu ene SS ASS Resins we Pad oS arr 7 114 0 VOLUME II. (1836-1841, containing 71 Plates). . . Price 4 0 0 . .. 5 6 6* PART --a(UGSG) oye wame meni along, fe sl EB Stig he Aces On st een 2 ea eae » 2 (1838)* : Beoege” (TREO Gap tpi Lhe Menge’ ole Hee Oopalis, Secu es) ed mtr hay Leeman Ove , Woe, nae awk 6 NOTRE OF de RBA oer) Onli: REM Rae das uke eet ent A Le eet ole AON De » d» (1841)* VOLUME ITI. (1842-1849, contaming 63 Plates) . . Price 3 8 6 . . . 411 OF Parr 1. (1842)* GSSiatee 21 lst: 3) Meype apm aine PPB cl es eC HY iy Pera eR vem pe pape ysis O oe] Elton | uteenrntnrnm yt CS) LOY SY Fy Se ro RO AAD ac ten: tas cre vt ale am oes ae Aretha ce SO EAN ee On gage) sie el Oe » 4. (1846)* HIST TBAB) Sor aie ee sea 20 iL el ine ee ates LO nd IO cua ee OT PERG hy Giee(LBAO) ieee Seve is 2 NN Bea Ne! oni Meee nae oh OS Ze cee chess OE a ey VOLUME IV. (1851-1862, containing 78 Plates). . . Price 6 2 o ’ (oe) tw) D> Paws (S51) Sue ee faa eee ee Ones eel eates! Oye Oa) 012 0 Pd Ye (03359) Ware AM Nh OP De CRN EGS AP SEL Neer sa 15 0 » 8. (1853)* nee (USB) RAGE bak: ch ase eget aaneay Soa) Op TOEO 1 4 0 SBA TBS) sy tea eRe ne a he bee SR eG 1 4°0 BR: Sp (str eee eae Si edn Oreu haga anad Gites Gate TVA ar 140 » 7, sec. 1. (1861) ee SORT 29 VED aD » 7, sec. 2. (1862) SOO 1.24520 VOLUME VY. (1862-1866, containing 67 Plates). . . Price 5 3 6 619 0 Pash s 01862) a as (4 oc eae, Ca eR RR PaaS 12:4040 Seb as(EBGB) SK: APSHA MCR 2) cleo le Nene aaa ats od 2 Boe) ag | Bes (UBGAN cect ek ace kny cee pe Dare eek = ee eT ngs 110 0 oo. Ae, (BBB) Cn Po Oe Se os amen ane Se a Ca) avi Be (USGEYn: sey ne: tes te) Sot reiitee ka. oy Ue MMR ale ey Grn Ware Seta Ea) [Continued on page 3 of Wrapper.) * The volumes and parts thus marked are out of print. y, ‘ To Fellows. To the Public. et Shad: ey Ma. i : % VOLUME VI. (1866-1869, containing 91 Plates) . . Price 11 5 0 .. .15 0 O Para Tee CVOEB aia cece eine se eae RR. tt eV EO 116 0 » 2. (1867). steed gna 116 0 » 3. (1867). PAHs AN 2 Pe bee 116 0 kee LTE ig Won ay AO OS a a ae WA el aa A A oe ea PR Is Pasi E78: PA Nah Oe Rea eC Oa Mk rR UR Qe 1 Ya 116 0 » 6. (1868). pea oS 7 eee 116 0 » 7. (1868) . ait’ 1G 0 28 0 » 8. (1869). PL 72-0 116 0 VOLUME VII. (1869-1872, containing 73 Plates) . . Price 817 0 1116 0 Beas dea ASGO) ho eo eee en aa 110 0 » 2. (1870). pete dy 2 "6 110 0 5 8. (1870). e NGS? Go eee GND 5, 4.*(1870) . mage hie ame 116 0 Soe etinial (iteye Biss So Le a ZO A 2 4 0 eG a(TSz1). oe 2 So) gd vO. Bey ies vote i owionl Cite ya eae tees Sang Ob 116 0 Fu0- 0 (LO72) ~ spin kiieaN Or - 110 0 VOLUME VIII. (1872-1874, containing 82 Plates). . Price 9 8 3 . - 1211 0 OT epee ies : _ TNO Reden Ol 1a BEA OR Sy aye cick Sy Ses) Coc oN MB ae So ADT eg i. a CESS At To) US ENLaRE OME a ML, Wal eA y MEM I Ait | 4 sue saaggana eal sy te Patent (MEHTA) “elie (svedsil oe ies ct Lane ane aeek apc me ease Ol ROYR Oa tyme LesitcrO pelt ese (LBs ech 1) 6 aaah Gon Ra 3 Weal Saget Se OR IB. 30h ts 140 BEN (TOY ay a Me tankeg eo ON en MaMa ca 2 OO. 6 ee TON Tate Be TST AN SS ME ALU Ai Me Ae TN ae ss OF 16: Gane 11.25,0 ; PS ML OTA Pesce tel t-sak ct oh 7 eek te pt aa tok oe Op Oe 1 4 0 Pehe O.te (LOT AN Arties i sraniatet ns anptath eke ony taal ole ko, shite Peay eal bl Wir x 8 2 2 0 - VOLUME IX. (1875-1877, containing 99 Plates) . . Price 12 0 O LG VOLO Parr 1. (1875, contaming 22 Plates) . . . --. 4, (1 6 ee O » 2- (1875, containing 12 Plates) . . . . - then Gal O16) 2 2 0 » 98. (1875, containing 5 Plates). . . .. - raat Ong i kOl ie 4g O. » 4. (1875, containing 14 Plates) . . . . +. » IJ11 6 Zine) » 5. (1875, containing 9 Plates). . . . .- » 1 26 110 0 >, 6. (1876, containing 4 Plates). . . . - . er nOwlOnsG 014 0 » 7. (1876, containing 7 Plates). . . . - . ai Damo rai sera PT 8 », 8. (1876, containmg 10 Plates) . . . . . Fe Op ow 140 5, 9. (1876, containing 6 Plates). . . . . - ae ee bee eA) 112 0 » 10. (1877, containing 5 Plates). . . oa niles 0 LAO ,, 11. (1877, containing 5 Plates and Title sind Taden} Aye) POND Seo eL OnG The «Transactions? and other publications of the Society may be obtained at the Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. CONTENTS. I. On the Axial Skeleton of the Struthionide. By St. GEORGE Mivarr, FR.S., Sec.L.8., Professor of Biology at University College, Kensington. . . . . . pagel THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological Society are of two kinds—* Proceedings,” published in an octavo form, and “ Transactions,” in quarto. According to the present arrangements, the ‘“‘ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published by the Committee of Publication. From fifty to seventy coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. ~* The “ Proceedings ”’ for each year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the preceding year. They may be obtained with black or coloured illustrations. The “ Transactions ” contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are - better adapted for publication in the quarto form. They are published at irregular intervals, but not less than three parts are usually issued in each year. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five pounds. The publications of the Society may be purchased at the Society’s Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. ; May Ist, 1877. : P. L. SCLATER Secretary. «TRANSACTIONS | OF (OOLOGICAL SOCIETY f teas 4 [or LONDON 7 i 4 1 Pei Rute ee | PRINTED FOR THE SOCIETY: = SOLD AT THEIR HOUSE IN HANOVER-SQUARE; s, LONGM: EEN, RI ND DYER, PATERNOSTER-ROW. a ee ws. ‘ - lugust Ist, 1877. ty | ‘e TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. VOLUME I. (1833-1835, containing 59 Plates) Part 1. (1833) . » 2. (1834)* 508.1838) . MR es Paci (URS Cay aire, aaa VOLUME II. (1836-1841, containing 71 Plates) . Part 1. (1836). » 2. (1838)* » 3. (1839). » 4. (1840) . » 5. (1841)* VOLUME III. (1842-1849, containing 63 Plates) Parr 1. (1842)* . (1843) . . (1844) . (1846) * . (1848) . . (1849) . c™) » 2 » ” 2 VOLUME IV. (1851-1862, containing 78 Plates). Part 1. (1851) . SRL Beek td Sue) (LS02 5 ie Many ten ene) cue » 9. (1853)* » 4 (1857). » 5. (1858) . » §. (1859) . » 7,s8ec. 1. (1861) » 7, sec. 2. (1862) VOLUME V. (1862-1866, containing 67 Plates) . Part 1. (1862). »» 2. (1868) . » 93. (1864). ened, R(US65) soa 3 De (LSGG)ic-s er ve * The volumes and parts thus marked are out of print. To Fellows. To the Public. ean aie by sds Price 313 6 418 O* » O14 8 019 0 3 0,152.9 1 0 Bee oad Ea eet) 14 0 Price 4 0 O D\Gyeom Pe wk Ae 8 112 0 so AONE besO 100 » O14 3 019 0 Price 3 8 6 411 O# » O14 8 019 6 » O14 8 019 6 2S ONO 010 6 EO vedi a 010 6 Price 6 2 O 8 2 6 » O09 O 012 0 sO LBS 9 aa aso UML ONO 1 40 Be OBIS: 0 140 Sau Ons se0 1 4 0 tr eOTkS 9 a Wore Cla 0 ive OREO Oo 140 Price 5 3 6 619 0 » O18 O 140 yee O15) 9 I baO Ps Wes OT 110 0 Aiea! Geer thal 0 110 O Dee at OS ire ed a a [Continued on page 3 of Wrapper.j To Fellows. To the Public. VOLUME VI. (1866-1869, containing 91 Plates) . . Price i 3 0 i 15 0 0 Papen dA (1866)) se. yes P on 197200 116 0 » 2. (1867). eee Med feet 116 0 » 3. (1867) . Bel sr 40 116 0 » 4. (1867) . eee 220"9 116 0 » 9. (1868) . Bie a A hie) 116 0 » 6. (1868) . al 78 116 0 » 7. (1868) . eae G': 0 2 3S x0 ‘ » 8. (1869). oe hi a aa 116 0 VOLUME VII. (1869-1872, containing 73 Plates) . . Price 817 0 1116 0 Et amunoey aC emtalte pk te Gl he eg og 110 0 » 2. (1870) . ig RTa G 110 0 » 3. (1870). pe AS S8 eG 110 0 iy A. (1870) « Pras Io er) 116 0 SMa 1871) . hk 1anO 24 0 pth Ss): Sage We ean!) 116 0 M7 (1871). & ba Aa! Surety 116 0 » 8 (1872). Taso g 110 0 VOLUME VIII. (1872-1874, containing 82 Plates). . Price9 8 3 1211 0 Part 1. (1872). 5 ee D LOM 014 0 » 2. (1872). pea eS 2 2 0 PUT ELS YA) URN Ser tae Senthil Verte a8) 116 0 Seidl (URRY cm LG hime ee ATR » O11 9 1G Peay IRIB abn anh RK a. Tae eeae Be 18) 0 140 » 6. (1878). 3 OV LOE'G 1 6 0 » 7. (1874). EOI. 16 Seg » 8. (1874). i 1BEO Lao » 9. (1874). pha a hte s 220 VOLUME IX. (1875-1877, containing 99. Plates) . . Price 12 O O .16 0 0 \ Part 1. (1875, contaming 22 Plates) eh Stee) LG 2 2 0 eee anon contamine Ve iPlates)i oan tee os gL AT) 6 wee? 2220 » 9. (1875, containing 5 Plates). . . . . . » O18 O . 14 0 », 4. (1875, contaming 14 Plates) . . .. . Pog nee KS) 2 2 0 yy D- (1875, contaming 9 Plates)... 5... Haier ari Wale sek) 110 0 », 6. (1876, containing 4 Plates). . . . . . pee ORLOeAG - 014 0 if (lor Go contamine7) Plates) 20.6). jay 1 skewed yyy 9.50 1D) 9 L210 é » 8. (1876, containing 10 Plates) . . . . . 4, 018 0 140 », 9. (1876, containing 6 Plates). . . . . . sy, 4:10. 112 0 », 10. (1877, containing 5 Plates). . . Bao 26. 110 0 mae ',, 11. (4877, containing 5 Plates and Title ues ries em 15.4.9 Tiare 2a VOLUME X. a *~ Parr 1. (1877, with numerous woodeuts). . . . Price O 9 0... 012 0 The ‘Transactions’ and other publications of the Society may be obtained at the Office (11 Hanover Square, W.), at Messrs, Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. CONTENTS. Il. On the Ancient or Quaternary Fauna of Gibraltar, as exemplified in the Mammalian Remains of the Ossiferous Breccia. By Guorce Busk, F.R.S., FEZ IS MOG | sok ei sh hea le sein Ps hahaa LN argos eee ee a THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological Society are of two kinds—“ Proceedings,” published in an octavo form, and “ Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published by the Committee of Publication. From fifty to seventy coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings ”’ for each year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the preceding year. They may be obtained with black or coloured illustrations. The “ Transactions ” contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are published at irregular intervals; but not less than three parts are usually issued in each year. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five pounds. The publications of the Society may be purchased at the Society’s Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. : August Ist, 1877. P. L. SCLATER ‘ Secretary. ° TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Vou. X.—Parr 3. LONDON: PRINTED FOR THE SOCIETY: . SOLD AT THEIR HOUSE IN HANOVER-SQUARE; AND BY MESSRS. LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. October 1st, 1877. Price 24s. Taylor and Francis, Printers,] [Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. VOLUME I. (1833-1835, containing 59 Plates) Parr 1. (1833). . » 2. (1834) » 8. (1835). » 4. (1835) . VOLUME II. (1836-1841, containing 71 Plates) . Part 1. (1836). » 2 (1888) » 98. (1839). » 4 (1840) . » 5. (1841) VOLUME III. (1842-1849, containing 63 Plates) Part 1. (1842) » 2. (1848) . . (1844) . . (1846) SAL BAB) eons caged aca Sapo MMC SE aa RS . (1849) . ee) Ph » Do f OO eh VOLUME IV. (1851-1862, containing 78 Plates). Panwl (USDD) ie nae pen he hones » 2. (1852). (1853) . (1857) . - (1858). . . (1859). . . » 7, sec. 1. (1861) » 7, sec. 2. (1862) be) S One VOLUME V. (1862-1866, containing 67 Plates) . Pape Ui (1862) shes ren visemes scl te nie » 2. (1868) . > 8. (1864) . » 4 (1865)... st Be (IBBG) ah To Fellows. £ s. Price 3 13 » O14 Price 4 0 1 4 » 018 ao nL he » 1 2 2? 1 2 10 10 AARwoonD 6 To the Public. £ 0 o j=) — me me OD 1 8. 19 a aS 19 4 10 10. 10 d. 418 0 0. 0 0 6 i=) for) o o oooco [Continued on page 3 of Wrapp eococo°o er. | To Fellows. To the Public. VOLUME VI. (1866-1869, containing 91 Plates) . . Price il 3 a : : is 0 6 Pannell (USGG) heard hal wee. eo) re ‘5 L7-> Oe 116 0 Scan LSC ene Rea Re a bres Oe Vi hah LS Qalag 116 0 Pee neem nna a yo Re pe Quik 116 0 » 4. (1867). yeh West 2a oe 116 0 » 5. (1868) . Persie ea Whee Oh oe 116 0 » 6. (1868). Fi jean We rhea Tete 116 0 » 7. (1868) . aoe Gta Oia 28 0 » 8. (1869). ye ee an Ray a ee 116 0 VOLUME VII. (1869-1872, contaming 73 Plates) . . Price 817 0 . - 1116 0 Parr]. (1869)... ... SEU SRB ibe 110 0 »» 2. (1870) . Bek NG 110 0 » 3. (1870). oye ips) Uy yan 110 0 » 4. (1870). eget dO 116 0 2S b> (871)% Peas Gri 240 » 6. (1871). rinse! ky esi t ge 116 0 geiong. "(LSZE) BRSoin ety cn 116 0 » 8. (1872). Bie 8 ia 110 0 VOLUME VIII. (1872-1874, containing 82 Plates). . Price 9 8 3 1211 0 Part 1. (1872). SPCR LM fit ahs 8s » O10 6 014 0 » >2- (1872). yep! bee Li bal ede oa 22 0 » 8. (1872). PE Ree aN ee: 116 0 b) SALUT BTS). o issue ste 0.45, 9 eae ie Bie rice (Ey 8) Vi See pene ot) OFS 40) 1 4 0 ‘i. » 6. (1873). » 019 6 160 st oe (1874); a8 1646 120 SEPT el TRO NT SS a ma SF 140 ULSI AN teen At in MS Pia itor i ee ALT OG 22 0 VOLUME IX. (1875-1877, containing 99 Plates) . . Price 12 0 O 16 0 0 Parr 1. (1875, containing 22 Plates) eT ayy py et UL ie 2 2 0 » 2. (1875, containing 12 Plates) . . . . . ers esi 8 Wares 2 2 0 » 3% (1875, containing 5 Plates) . Av toe emt rrrene bee He er ( tae 14 0 Peeuisvo,,containin= 14 Plates) 9520 seo. ogo MN TL Ge. 2ee2. 0 » 9- (1875, containing 9 Plates). . . . . . Soret, ki Gee ns 110 0 5, 6. (1876, containing 4 Plates). . . . . . 4» O10 6. 014 0 » @. (1876, containing 7 Plates). . . . . . Bay OL LD ig es LOE-O », 8. (1876, containmg 10 Plates) . . . . . 4 O18 O. 140 5, 9. (1876, containing 6 Plates). . . . .. ye Liaw Ae ee: 112 0 ,, 10. (1877, containing 5 Plates). . . Sa tred BY -Tem! Ceor 110 0 » 11. (1877, containing 5 Plates and Title ea ihdbe) sie Ons Sry! Bae dig pet RP VOLUME X. Part 1. (1877, with numerous woodcuts). . . . Price O 9 O .. . 012 0 » 2. (1877, containing 27 Plates) . . . . . PROS Soap aaa Oe Lysates FO LET The ‘Transactions’ and other publications of the Society may be obtained at the Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. CONTENTS. III. Notes on the Manatee (Manatus americanus) recently living in the Society's Gardens. By A. H. Garrop, W.A., F.Z.S., Prosector to the Society . page 137 IV. On Dinornis (Part XXI.): containing a Restoration of the Skeleton of Dinornis maximus, Owen. With an Appendix, on Additional Evidence of the Genus Dromornis in Australia. By Professor Owen, C.B., P.RS., F.ZS., dc. . 147 THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological Society are of two kinds—“ Proceedings,” published in an octavo form, and “Transactions,” in quarto. According to the present arrangements, the “ Proceedings”’ contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published bythe Committee of Publication. From fifty to seventy coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings ”’ for each year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the preceding year. They may be obtained with black or coloured illustrations. The “ Transactions” contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are published at irregular intervals; but not less than three parts are usually issued in each year. : Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five pounds. The publications of the Society may be purchased at the Society’s Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. P. L. SCLATER, October Ist, 1877. Secretary. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Vou. X.—Parr 4. LONDON: PRINTED FOR THE SOCIETY : SOLD AT THEIR HOUSE IN HANOVER-SQUARE; AND BY MESSRS. LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW March 1st, 1878. Price 30s. Taylor and Francis, Printers, } [Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. * To Fellows. To the Public. { EER) 8p! SE ih VOLUME I. (1833-1835, containing 59 Plates) . . . Price 313 6... 418 O Pawn 1). (UB83) oes cee Beh, ete ty eae AOL Ae en tr OSG amt » 2. (1834) 5 BE (LBBB) mee eh UI Re ee CTI CGT at Se NIC) a CR Ost hg ME sty Ane) (LBD) Ahern. Sih droit ey Wait tifa SO aint Whe pene aT pnt Gin sara ene at ONE VOLUME II. (1836-1841, containing 71 Plates). . . Price 4 0 0 .:*. . 5 6 6 Pinel? (USEPA <9) Meta anes deen cent kee NUE aT HO paneer » 2. (1838) SB CIBER). ASS UR cass too SS CA Dy ae ear ne ae con et a Cae Orde MET RAED), 5 Uae aE eae eet iyo a See el eu Nee Cin Tat oeser UR. rae ean ane » D. (1841) VOLUME III. (1842-1849, containing 63 Plates) . . Price 3 8 6 ... 41211 0 Part 1. (1842) SP NRALEBABY: 38 saaehatiss WMS Oe 8 nA aL Eon aN Rapin ath Se iS aS eae NA 7) eae ee ene a tees HET AS panda hr Yhs Fla Weare ay 7s » 4. (1846) Abie” (LBABY 8 (ky Se es RR Tk ARO oye nc aimed SG. UBAG) rcs rea pan Ii A an BC Op cy WP Bis nea OS Cy VOLUME IV. (1851-1862, containing 78 Plates). . . Price 6 2 6... 8 2 6 D Are 1s {(USB1) a ROE. eC eee a Rs Die Gare Gta a am Beer LO) iciar eatin Scrat RUM ayer tana agra ane ere AURA WO) Lib OF » 3. (1853) Fava LB a? ett Meee yi Be arg veal MN, Ce aga 140 Bear UE) -) eerie mctoh aces Hla nireeey Samim Mra (memery. (ete 14 0 SiGe, (B59) or oN OLB RO 140 » 7, sec. 1. (1861) SSRN A de on CORE in Ee ai 1 he Dale RECS ALOU ier laa cree) Ce Lae Fe alisha) 14 0 VOLUME V. (1862-1866, containing 67 Plates). . . Price 5 3 6 619 0 Bann © (L862). Fee aera oi 5 es, Boe ae a eee eT nO Ade Se teh (E BGS) ee ear peo Laie aaa gt ys a a a Pn EL SSE 1S ee O jos (BGS) er PE ie See ea oe I eh a AR 110 0 yo 1; 4ee (LBGB I dies GRE 02" hots 1 SRT ge ee Amie esse crea eee RR ET) Piste Com is! 314) Nea pein es tueae nr ihe ae Minintabr an agn een ne Ua Thon Digi bene ues Tlalriay Up [Continued on page 3 of Wrapper.| To Fellows. To the Public. VOLUME VI. (1866-1869, containing 91 Plates) . . Price i ae eae Panne) (1866) 0 Jah as ee Ge 116 0 » 2 (1867). Bere yc Mi he 116 0 st, 8 (1867). BP ts Wie (0 is 116 0 » 4 (1867) . S eAmlic meRUa eee 116 0 » 5. (1868) . SA Meee cs ae 116 0 » 6. (1868) . nan NH ae sien 116 0 >» 7%. (1868) . i 116 0 2 8 0 » 8. (1869). es Leegen0 116 0 VOLUME VII. (1869-1872, containing 73 Plates) . . Price 817 0 . PUG EO Martel BOO) i Ns) Ne J SAG 110 0 » 2. (1870) . 1 926 110 0 2018. (1870) des ae 110 0 » 4. (1870). Males 0) D16 970 » d. (1871). Fionn 1340 240 Om OV )es epee] Mar at) 3 ile Loge} Sg A871) - ay ie aah 116 0 » 8. (1872). DRESHONG. 6 110 0 VOLUME VIII. (1872-1874, containing 82 Plates). . Price 9 8 3 120A E50 Part 1. (1872). sity OaLOLNG 014 0 2. (1872) . ak nS 2 2 0 5» 8: (1872) . Pam a ere oa) 116 0 AN (BYE). ian Peeaivn tne PL Pls NBEO ey MOS 1 0 5 Ge (A873): = Qiks 160 5 7. (1874). f » O16 6 ING) » 8. (1874). Pair 30) LAreG » 9. (1874). HORT Tarsng 220 VOLUME IX. (1875-1877, containing 99 Plates) Price 12 0 O 16 0 0 Part l. (1875, contaming 22 Plates) speed core OL Chae, 2 2 0 » 2 (1875, containing 12 Plates) . . . .. a G 2 2.0 » 9 (1875, contaiming 5 Plates). . . . .. Ontos O 1 4 0 », 4. (1875, containing 14 Plates) . . ... . Saeed amd 2 20 » 9. (1875, containing 9 Plates). . . . . . CCL DON LO 110 0 Dior (evo, containing 4: Plates).9 22). -. ay O-1LOe-6 O14 0 Evite (bOZo;;contaimme 7 Plates): 20 89 24 av Oi aye DENA tC: », 8. (1876, contaiming 10 Plates) . . . . . ee OOM 1 4 0 » 9. (1876, containing 6 Plates). . . . . . Bot cde Ae 1 el ORS) », 10. (1877, containing 5 Plates). . . ele OF 110 0 », 11. (1877, containing 5 Plates and Title aa Tiles) 015 9 Evite yout Oe: Part 1. (1877, with numerous woodcuts). . . . PriceO 9 O 012.0 2. (1877, containing 27 Plates) . . .. . eta aid! ahs 2) F100 3. (1877, containing 6 Plates) . . . . . » 018 0 1 4 0 », 4. (1878, containing 9 Plates) . . .. . Aa ee O. 110 0 y Dan(lLerey contains isd) Plates). 5 iyi le. Hr ONO LO 0.12 0 The ‘Transactions’ and other publications of the Society may be obtained at the Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, H.C.), or through any bookseller. CONTENTS. V. On the Structure and Development of the Skull in Sharks and Skates. By Wi IP Areva PF 1k EAS aa Sea Sa eA ace ed THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological Society are of two kinds— Proceedings,” published in an octavo form, and “ Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- ‘mended to be published by the Committee of Publication. From fifty to seventy coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings ”’ for each year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the preceding year. They may be obtained with black or coloured illustrations. The “ Transactions ”’ contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are published at irregular intervals; but not less than three parts are usually issued in each year. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five pounds. The publications of the Society may be purchased at the Society’s Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. ‘ P, L. SCLATER, March Ist, 1878. Secretary TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Vou. X.—Parr 5. LONDON: PRINTED FOR THE SOCIETY: SOLD AT THEIR HOUSE IN HANOVER-SQUARE; AND BY MESSRS. LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. March 1st, 1878. Price 12s. Taylor and Francis, Printers, | [Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. To the Public. ees eds £ 3s. d. VOLUME I. (1833-1835, containing 59 Plates) . . . Price 313 6... 418 0 Pann der (L888) 9k scrreclssayir OS (IT Aan Mar een) AT Quan » 4. (1846) Soil ete OAB) SARs cc dee Mie We ona HP iat cat oh iuuae in Oke SZ aa TO UMMER Et TLY ne CORT Oks SOL BAO) 58 ARs ORC NG NS Te rao at reer aan Save ae oman des nom) ORG VOLUME IV. (1851-1862, containing 78 Plates). . . Price 6 2 6 3.2 6 ARs) (LED) is erie tealige Dayne) celeritete ORO eI) 012 0 Brg ee: (LOD 2) oo ao eas ay. bent QUAM Ee ee ee NAT DLO TB s3O) » 93. (1853) » 4. (1857) . » 018 0 140 > D. (1858) . » O18 0 140 » 6. (1859). » 018 0 140 » 7, sec. 1. (1861) » 9015 9 110 »» 7, sec. 2. (1862) » 018 0 140 VOLUME V. (1862-1866, containing 67 Plates). . . Price 5 3 6 619 0 Pawns) (1862) nny iiregmee tach) 1s < ncaa fh See OMSL 140 bie (LOGS) 5 Sete tii SANA dic iha aa p eae Une RST) 15 0 » 98. (1864) Pel Grur ste a 110 0 3518 AL ACB GB) it Vecre esi 1 Pea a EE De agg MNT a Go orn eT ST Bag 99 Heo (LBGE re hare apenas rv ie ist i kc eT EOI : 110 0 [Continued on page 3 of Wrapper.) To Fellows. To the Public. VOLUME VI. (1866-1869, containing 91 Plates) . . Price it 3 6 A is 0 0 Paro Pa (LSGG).. yee ss Las7nO 116 0 » 2 (1867) . Ar aie AD orien 116 0 » 3. (1867) . Rees The a i hae 116 0 a LO a rane NO Rar Cs Hl (RE 116 0 LO PONTE IU Sa Ra RR es A A 116 0 » 6. (1868) . ate eae it. 116 0 » 7. (1868) . Jeo ae sie 2 8 0 » 8. (1869). Saad sg 116 0 VOLUME VII. (1869-1872, contaming 73 Plates) . . Price 817 0 . 1116 0 Parte weogy sree bir de rane 110 0 » 2. (1870) . Re onnG 110 0 | » 3. (1870). Be Lr eh Guine 110 0 " EC TUS Ns ae aa Bek whee 116 0 ‘ Mee IRAE sy eae RT vara 1a Ome 240 » 6. (1871). Ce am ee 116 0 » 7% (871). van Ba Ae 116 0 » 98. (1872). eed Qa 110 0 VOLUME VIII. (1872-1874, containing 82 Plates). . Price 9 8 3 1211 0 Parr 1. (1872). BPP NE Gin: PROT OK Gi 014 0 Sy eek LST2)): sity DLE Gye 2 2 0 PSE amare) es nee Pe ian Poe 116 0 SEO MTGE Se alt eh caen Ondonto9 jie aa A GAS jp oy 2a hat oe Lave E800 1 4 0 Fits ARMY GS) anf or eh caer » 019 6 160 on NIB ZANE e i Sar sO nl Gee 120 ¥ LORS OCIC EE Ue Sie. | ER ce NO 08) 08? a lat4a70 RGA REE sy), 5 lve odd, oe oo OIG 22.0 VOLUME IX. (1875-1877, containing 99 Plates) . . Price 12 0 0 .16 0 0 ‘t Part 1. (1875, containing 22 Plates) . . . . . ross lee hin ta De taenO ni » 2. (1875, containing 12 Plates) . . . . . Ser lL yaiG 2,2 0 » 93 (1875, containing 5 Plates). . . . =. . sn OL ESO 14 0 5, 4. (1875, containmg 14 Plates) . . . . . » (Lill 6 2 2 0 en lo73, containing OWPlates) < ajfrs in a) 3% asthe pre en 110 0 5, 6. (1876, containing 4 Plates). . . . . . 35 BO LONG: 014 0 » 7%. (1876, containing 7 Plates). . . . . . Bs ROMO a Tia Ian 4) ( » 8. (1876, containing 10 Plates) . . . . . Se OL LS 2c0 140 » 9. (1876, containing 6 Plates). . . . . . 5b 4 O 1096 Puen) 5, 10. (1877, containing 5 Plates). . . Tp ReT Meet ped 6 110 0 5, 11. (1877, containing 5 Plates and Title ae Tides) i Olbee9) 1k 0 VOLUME X. Parr 1. (1877, with numerous woodeuts). . . . PriceO 9 O 0.12 0 » 2. (1877, containing 27 Plates) . . . . . See ERY pe Oe s0 » 93. (1877, containing 6 Plates). .). . - Od 85:0 14 0 5, 4. (1877, containiug 9 Plates) - |... >: ees oa a RE » 5. (1878, containing 3 Plates) . . . . - Sil Onan nO: 012 0 The ‘Transactions’ and other publications of the Society may be obtained at the Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. = CONTENTS. VI. A Description of the Madreporaria dredged up during the Expedition of H.M.S. ‘ Porcupine’ in 1869 and 1870.—Part II. By Professor P. Martin Duncan, /.B. (Lond.), F.RS., President of the Geological Society . . . . . = page 235 o. THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tur scientific publications of the Zoological Society are of two kinds— Proceedings,” published im an octavo form, and “ Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published by the Committee of Publication. From fifty to seventy coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings ” for each year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the preceding year. They may be obtained with black or coloured illustrations. The “ Transactions ” contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are published at irregular intervals; but not less than three parts are usually issued in each year. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five pounds. : The publications of the Society may be purchased at the Society’s Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller, March Ist, 1878. Be eh — TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Vou. X.—Parrt 6. LONDON: PRINTED FOR THE SOCIETY: SOLD AT THEIR HOUSE IN HANOVER-SQUARE; AND BY MESSRS. LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. June ist, 1878. Price 30s. Taylor and Francis, Printers, ] [Red Lion Court, Fleet Street. TRAN SACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. To the Public. cE Sark ae 25) SOS ICB VOLUME I. (1833-1835, containing 59 Plates) . . . Price 3138 6 . . . 418 O Pabr Uy (ESBS) ooo ak ep UR ae oy Se al Beceem » 2. (1834) SPASOACUBB), 20g, 08 (abc? -8" 2 MNS adh cere aa ae yd I Lae Bytes (UBD) es dase a iaintec 2a! sls) BaNReien lee eae ue ky as yal Male te 3) 114 0 VOLUME II. (1836-1841, containing 71 Plates). . . Price 4 0 O 5 6 6 PAu (PAsp) ee Oeees is fH SOARS ON deta hay 112 0 » 2. (1838) i eS URRY? ae 5 nC UR tenia ae ay 1 SO576 ya Ate (LSA) tees tite fie ce Phe adic ee eelpe nee ee te pair DOTA 13 019 0 » 5. (1841) VOLUME III. (1842-1849, containing 63 Plates) . . Price 3 8 6 ... 411 0 Part 1. (1842) bs HAS) teens 35 Sones telly (Ub En eee a Ah » O14 8 019 6 aN esa SAM Vat 27 Ro hht CORON LL YN EM nae aire, OTA, 8 019 6 » 4 (1846) ‘ Petlstg (it) 3) aera MiGs GERARD Oe ate ele ee eK aCe NANO 010 6 OA BGT Gl ot! AC) MORN Area a ete ARAMA “a ah and Ds » O 710 010 6 VOLUME IV. (1851-1862, containing 78 Plates). . . Price 6 2 6 8 2 6 Bane IO UBol ee ee Sg oe Sec Le OD at) 012 0 Ge SNEED DS: Se PELs eR LS AAA te AAS gg Fg a 1 500 » 9. (1853) are ORS V A Reysteen talc epee ge De Ream Ue eae » 018 O 140 pena (LSDS) sans se sera alia: Mech A oateeA crt ie » 018 0 1 4 0 » 6. (1859). Pind sain Weyer tae Gs: 0 140 » 7, sec. 1. (1861) os LR rea GI Re ar NON » O18 9 110 NG eee (AMER IN. ail MMM A MELD eo CL RRL) 1400 VOLUME V. (1862-1866, containing 67 Plates)... . Price 5 3 6 619 0 Parr 1. (1862). AWS Va ban DaN Sa Pde RMR OA cols 1ovaa SV GISGAN CEL Seas PO SN ie Ge RT gy aa tea 15? x0 Ss ia@eyh (LOOM) ich can nue uae pny a. Nah Rene roan nga Ay 6 110 0 Sstanteticns (LGD) cena came Petia. Waly nL Tae ea cage Tp aaa Wimp jrea 3 110 0 Sei) Nbsi (ROUO) enact ears aetna aha nics tate aaee ke ake O 110 0 VOLUME VI. (1866-1869, containing 91 Plates) . . Price 1l 5 O 15 0 0 PAR TCLS (URGE) cry aa aeceettY te eben cone Nene iae 3 Piers 0 116 0 » 2. (1867). i ow An) 116 0 5 3 (1867). De 17 0 116 0 » 4 (1867) . us D827. 30: 116 0 » 5. (1868). ai Avie gent) 116 0 » 6. (1868). Bs I ny Aaee 8) 116 0 » 7. (1868) . UY iV) 28 0 » 8. (1869). a 17 0 116 0 [Continued on page 3 nee Wrapper. | To Fellows. To the Public. VOLUME VII. (1869-1872, containing 73 Plates) . . Price 817. 0 whe oO ‘ PRE ME SD) stick. oust. ea atee ONIN OMAR LO UO gs 110 0 ; » %. (1870) . Pi wes 110 0 i fo GSC Ea ae ens oy: 110 0 ¢, MASLOV A, wer hes Path (6° | Ay Mi CU aN ams SoD Jr cag) 116 0 a sty “UPON CN SEU an ctr Re parte Ea i aan ss a 2 4 0 ir » §6. (1871). Jey rem in arent) 116 0 3 » 7. (1873). Bek era NO 116 0 | es » 8. (1872). dt DEE 110 0 cy VOLUME VIII. (1872-1874, containing 82 Plates). . Price 9 8 3 - 1211 0 j LAE ce) ae na apa Bre0 1026 014 0 BY » 2. (1872) Raa A156 220 ; » 8. (1872) Ba TY das) 116 0 A » 4 (1878) Hipeie «a Pog 141 o » 5. (1878) eeOeLs~ 0 140 » 6. (1873) » 019 6 160 » 7. (1874) P » 016 6 CTE » 8. (1874) inh LEO 140 » 9. (1874) irae ai tos PAD EA : VOLUME IX. (1875-1877, containing 99 Plates) . . Price 12 0 O .16 0 0 : Part 1. (1875, containing 22 Plates) SCL CAL cid wd mao) 2 2 0 Fs: » 2. (1875, containmg 12 Plates) . . . .. 4 +=%d1dil 6 2..2-0 re » 8. (1875, containing 5 Plates). . . . .. «4 O18 0 1 40 m » 4. (1875, containing 14 Plates) . . . . . 4 +%Idill 6 2 2 0 pe » B= (1875; containing-9 Plates)... 4... +, 1 2 6 110 0 ‘se », 6. (1876, contaming 4 Plates). . . . .. 4 O10 6 014 0 ‘3 Ba 2 (erG;contaimme 7 Plates) 7 i)o 20 8) 58 20) 15.9 1 AO », 8. (1876, containing 10 Plates) . . . . . 5 OMS 320 1 4 0 ‘ y 5, 9. (1876, containing 6 Plates). . . . . . of abead Lait 2 a 9) 112 6 ,, 10. (1877, containing 5 Plates). . . i dese ene 110 0 » 11. (1877, containing 5 Plates and Title an mye 015 9 Lat tae VOLUME X. Part 1. (1877, with numerous woodcuts). . . . PriceO 9 O 012 0 » 2%. (1877, containing 27 Plates) . . . . . Be eG 110 0 » 9. (1877, containing 6 Plates) . . .. . » 0918 0 1 4 0 » 4. (1878, contaming 9 Plates) . . .. . pon ee nO 110 4 » 9. (1878, contaming 3 Plates) . . .. . SOLO Eko 012 0 » 6. (1878, contaming 9 Plates) . . . . . eon eisai a 110 0 The ‘Transactions’ and other publications of the Society may be obtained at the Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. CONTENTS. VII. On the Skull of the Aigithognathous Birds.—Part Ul. By W.K. Parker, /.2.S., PDS: PEAS RGA Ge eke, CaM elses cat deteaae Nr iain, 4 ena Ma aon OEE ‘ THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. THE scientific publications of the Zoological Society are of two kinds—“ Proceedings,” published in an octavo form, and ‘“ Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published by the Committee of Publication. From fifty to seventy coloured plates~ and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of anima]s described in them. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings” for each year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the precedmg year. They may be obtained with black or coloured illustrations. The “ Transactions” contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are published at irregular intervals; but not less than three parts are usually issued in each year. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five pounds. The publications of the Society may be purchased at the Society’s Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. P. L. SCLATER, © June ist, 1878. Secretary. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Vou. X.—Parr 7. LONDON: PRINTED FOR THE SOCIETY: SOLD AT THEIR HOUSE IN HANOVER-SQUARE; AND BY MESSRS. LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. August 1st, 1878. Price 24s. Taylor and Francis, Printers, | [Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. To the Public. Ea £ s. d. VOLUME I. (1833-1835, containing 59 Plates) . . . Price 313 6 ... 418 O Pann il. (L888) ea heen a veut. Nbenca em ati aie 39 le OMAN Sree ae uO! 19000 » 2. (1884) si Ss) CLBEB) liege as Satna Rh UE Ne eM Sing WATO 5p esate SEE) erin eae eerie SW ant ace eile apie oT nite ey doa es 114 0 VOLUME II. (1836-1841, containing 71 Plates). . . Price 4 0 0 5 6 6 Pann eh SsG) isin) ee) eames: heymetie a nee . 597 AO 112 0 » 2. (1838) Bets Feed (5312) orem Kae io Mbarara aay ato alte bal be Se OMT, Oto en teil edie Fi FEEL SAO). ee RMR SS Sun aan eats as ven aie mameeaieays (pn arr eRINan Comes Miike (NT mn » 5. (1841) VOLUME III. (1842-1849, containing 63 Plates) . . Price 3 8 6 . . . 411 0 Part 1. (1842) pyFe eee (OAS) PeReen «., tn Sew alan Mie elias Saaremaa yites » O14 8... 019 6 et CLBAAD tant 2's ee bree re eect mda ate nr area 019 6 » 4. (1846) Pe Pag RUT aa eR race aay OEE OS NER anette ANGRY et Cy 010 6 seamen (ESAQ a Ua uN VR Ua MBS Wa al ca Neen gy 010 6 VOLUME IV. (1851-1862, containing 78 Plates). . . Price 6 2 6 8 2 6 V2.9 ul Lik ks 5\t PRA ay tp ARG erg ee a aL Fae De rac ON 29.0 012 0 Si2e (LSBRN A ate as, ek Li Mek TB eg att » 3. (1853) » 4 (1857). » 018 0 ae » 5. (1858). Ys OTRO iY a:0 oo Ga LEDO) angie ee 3) OAS 20 14 0 » 7, sec. 1. (1861) SAR iets Ne eaten CC Dar OND ey 2 spit, Mp BECLP AAC UBER) stanly Oe MEH jy tans Mh greta hh eel) TRIG i Hee hea VOLUME V. (1862-1866, containing 67'Plates). . . Price 5 38 6 619 0O amr tT BG2) Mirae si 1 nabs OT YS ang ox Deg » O18 O 1 4 0 » 2. (1863) . PVA SARA PE SRN he Apo ANAL Dl ra) 15 0 peg, Seb PEGA) Anat ARN. Gun Pineau Osis ona bea pile uaa GI 110 0O x 4. (1865). Soe Leh 110 0 » 5. (1866) . Sede so 110 © VOLUME VI. (1866-1869, containing 9i Plates) . . Price 1] 5 O 1 0 0 PARE L866) every oe ha nemesis p se Eh) 116 0 » 2. (1867). es O 116 0 33) 08s CBG) ale EAN Di Ge an ae 116 0 39 gee ASOT) BOSON NE Tae a ae a) Me ON Saat 17 0 116 0 » 5. (1868). a 17 0 116 0 » 6. (1868) . Pah ey anes 8 116 0 53) as (UBER) coh 98, Vaastiat aes spe LRP Ry rhe TES eT 2 8 0 ap Se AT SGD) cc on eae vO ann a five ERR 2 AE LOZ 116 0 [Continued on page 3 of Wrapper. | To Fellows. To the Public. VOLUME VIL. (1869-1872, containing 73 Plates) . . Price 817 0 it 460 JEST TIN TE (G that) aioe an ni oe 5 Oe 110 0 » 2% (1870) . Pi aot PAC 110 0 » 8. (1870). ja are ts 110 0 Ss eat YOST Q)) SE TS lt So cl Bod alae) 116 0 Sei Bie ile LO: Ae. a pete a a » 1138 0 2 4 0 » 6. (1871). Rey dy CO 116 0 » 7. (1871). ie AL ee O 116 0 » 8. (1872). Bik .2ie6 110 0 VOLUME VIII. (1872-1874, containing 82 Plates). . Price 9 8 3 SABA 0 LAOS S) 23 NE), el ie eg » 9010 6 014 0 Lo SA OIC LE ae are Ss Bie Be 220 ee(iSr iene SS eg 116 0 eat (Ieper epee mee ck Tk a OTE. og 1 ¥10 een is a neni me nie I A OBL O Li4e ; WG ise a eS Se oe 0.19. 6 1 60 Beethoven eee Nass otis Ae tie Nh So O66 12 E10 S SATE, Od” oS A a ec 4 FY) 1 40 <, » 9. (1874) . 3 AWE 2 2 0 VOLUME IX. (1875-1877, containing 99 Plates) . . Price 12 0 O .16 0 0 Part 1. (1875, containing 22 Plates) . . . . . «4 (dill 6 2°20 » 2. (1875, containmg 12 Plates) . . .. . 9a LG 2° 250 » 98 (1875, containing 5 Plates). . . . . . «4 #+O18 O 1 4 0 A » 4. (1875, containing 14 Plates) . . . . . Sis VLE 2 2 0 » 5. (1875, containing 9 Plates). . . .. . Ss eek ein Le a0) », 6. (1876, containing 4 Plates). . . . . . » 9010 6 014 0 5 » 7. (1876, containing 7 Plates). . . . . . 4 OD 9 1 Ween en) », 8. (1876, containing 10 Plates) . . . . . 5 018' 0 .1 4 0 » 9. (1876, containing 6 Plates). . . . .. ay ls Ae 112 0 », 10. (1877, containing 5 Plates). . . Peis ot gave rue 110 0 », 11. (1877, containing 5 Plates and Title sd les) per onse 1 ess a VOLUME X. Parr 1. (1877, with numerous woodcuts). . . . Price 0 9 O 012 0 » 2. (1877, containing 27 Plates) . . .. . Coe MO Aes 3 110 0 » 9. (1877, containing 6 Plates) . . .. . ee OLS) ah 1 4 0 » 4. (1878, containing 9 Plates) . ... . BAR yy ieah BLOG » 5. (1878, containmg 3 Plates) . ... . eigen aw 0) 012 0 », 6. (1878, containing 9 Plates) . . . - . 2 ee BiG 110 0 » 7. (1878, containmg 7 Plates) . . .. - » 918 O 14 0 The ‘Transactions’ and other publications of the Society may be obtained at the Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. CONTENTS. VIII. On the Axial Skeleton of the Pelecanide. By Sv. Grorce Mivart, F.R.S., Pro- fessor of Biology at University College, Kensington. . . . . . page 315 THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. THE scientific publications of the Zoological Society are of two kinds—“ Proceedings,” published in an octavo form, and “ Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published by the Committee of Publication. From fifty to seventy coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings” for each year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the preceding year. They may be obtained with black or coloured illustrations. The “ Transactions” contain such of the more important communications made to the scientific meetings of the Society as; on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are published at irregular intervals; but not less than three parts are usually issued in each year. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five pounds. The publications of the Society may be purchased at the Society’s Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, H.C.), or through any bookseller. P. L. SCLATER, August 1st, 1878, Secretary TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Vou. X.—Parr 8. PRINTED FOR THE SOCIETY: SOLD AT THEIR HOUSE IN HANOVER-SQUARE: AND BY MESSRS. LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. August 1st, 1878. Price 20s. Taylor and Francis, Printers, | [Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. To the Public. - onsen ae Lige, meds VOLUME I. (1833-1835, contaming 59 Plates) . . . Price 3138 6 . . . 418 0 Pane’ 13(1838):-5 45 MEN Ge he eek rere ea aun COLA ROM ane tO) Ou aG » 2 (1884) PR @ Roh 1) WAN ee Petra net Orta ee ae apear aul retention MaRS a Wash Ce 110 5p dts (LBOD) eran Acee dete ie oll sale hee te alter ae imc TAS eal ee nn 114 0 VOLUME II. (1836-1841, containing 71 Plates). . . Price 4 0 O 5 6 6 Part 1. (1886). 2 2... aie a oe Ceieansse a Wie: eh 0) 112 0 » 2. (1838) Pea iis (to?) Verges CEN aOR nr Mare iy ea iat eH tA USCA O)/E a eh Noah ysis Sid tes Ujleat@ 359 ASN (UBAO) AER OSC PRE Ter Tinea hin dae Prenat ened (Dud aor es amerriareraco: dal OO » 5. (1841) VOLUME III. (1842-1849, containing 63 Plates) . . Price 3 8 6 . . . 411 O Part 1. (1842) pa ee (LOA) seis! 6, UCR NEA REN Sts c IIe Wd Cne 2 AIR Sliee Lea Lats eae ser) cone OME My ype (LO 4A) ears Ste easy heb n on oe etme eae OUEAG iG 019 6 » 4. (1846) Bye Dek (BAB) 252 ot SY ue saa ala Ory earrea el ceri Ac Rr Si ie een UO 010 6 SEO (NGA) oof seo en iin Mik: hetrea Nama chc ber (eae one Mal 010 6 VOLUME IV. (1851-1862, containing 78 Plates). . . Price 6 2 6 8 2 6 Part 1. (1851) . Sei avoeent ar (aregecsa ear kenny Myc ee ee Oa () 012 0 By Mec LOD S) corso cuear Merce ey ney ath OM ea oe Ceae a ost) taba ok A NG) SLR Hp La » 98. (1853) ee Aes LSBT ce ties en le MRP tO 50, RSLS SO 1 40 Saylor (LBDB) je arse oral sy ay eae oun ad Near py aoe geh oe ORSLG yO 1 4 0 mah (LEO) Te AE AIRS OO GEC cc cement as 0 a8" (0 140 pies BeGs Le (TRG DS he (bight. sheep ek ome he MG eae et RO MAD =O 1 10 Boi id y SECIAk| (LOGS) Mi eae Le NM Rr nega NUN yes eee ORES) 10) 14 0 VOLUME V. (1862-1866, containing 67 Plates). . . Price 5 3 6 619 0 PART a(LSG2) os seps he aoe Neale ean wal Os Fie TIRES) 1 4 0 Sar Otel 0Ts hae Mey an at my Ame Ml NN ncanbiiat MeN a” BapeeU mT Od Ue Stals) Li eon O Piling tae (2) Pesach cet ete Re rae eT Pag APES SAR Ged Peal OO 110 0 gait iL BGD) eset tee Se tea Hp ah pn cao Rian a ue See aren Renna 110 0 asnwdet( LBOG) apie amen ten) ab, shone: te <0 Mea ta eRe eT RE A 110 © VOLUME VI. (1866-1869, containing 91 Plates) . . Price 1] 5 O 15 0 0 Part l. (1866). . . . .. Sr Ppllctictiy.* O 116 0 » 2. (1867). Pare Wee gent 0) 116 0 » 98. (1867). Pei ray Aiea} 116 0° » 4. (1867) . Ae creme tog oe 116 0 » 5. (1868). . : Bhd ge O 116 0 » 6. (1868). Ss eat. 116 0 » 7. (1868). » 116 0 2 8 0 5 8. (1869). Peers Wit aad) 116 0 Lonel on page 3 Ba Wrapper.] . To Fellows. To the Public. VOLUME VIL. (1869-1872, ee 73 a gaan cebae feng te Part l. (1869). . . a apes Hn aa , PrN Peete, 110 0 » 2 (1870). Dy aL eAeenG 110 0 » 3. (1870). Pi We a1: 110 0 Berea ioe se ee a ay ps ELKO 116 0 PROS ee i a ha. 8 2 40 » 6, (1871). Pee at om) 116 0 » 7, (1871). Pra Seri 116 0 » 8. (1872). en a WO: 110 0 VOLUME VIII. (1872-1874, containing 82 Plates). . Pricee9 8 3 . . 1211 0 PARE ES (Lue) ai tren caster LEAL a) ce | yi OAC Gis 014 0 » 2 (1872). PME Es aes ae: WAG 220 Waeiemeyriite wen. a SUE gg 116 0 » 4 (1873) - ORO EGC ag er Sar EAU enorme Wel a eg B79 140 Pa emer ee eet eo O19 6 160 Polite e ee nd cre ey eo ies. oe ee -6 120 Bee aay Maree kT her Se ag jae pe Pn eh ents encima sd: Fe foes ho} eS, 220 VOLUME IX. (1875-1877, containing 99 Plates) . . Price 12 0 O -16 0 0 Parr 1. (1875, containing 22 Plates) . . . .. «4 +%1ll 6 2 2.0. 3, 2. (1875, containing 12 Plates) . . . . . 3 eh S6 2 2 0 » 98 (1875, containing 5 Plates). . . . . . pie Om oenO 1 4 0 » 4. (1875, containing 14 Plates) . . . .. 4 lll 6 2°20 » 5. (1875, containing 9 Plates). . . . .. 4 +126 110 0 », 6. (1876, containing 4 Plates). . . . . . 3 OvIOK'6 014 0 » %. (1876, containing 7 Plates). . . . . . 4 O11 Q 1 1 0 », 8. (1876, containmg 10 Plates) . . . |. » O18 0 140 » 9. (1876, containing 6 Plates). . . . .. 4 +%140 112 0 », 10. (1877, containing 5 Plates). . . Paps anti ead ¢) 110 0 », 11. (1877, containing 5 Plates and Title aid Tiles} 5 ODD) Fete O VOLUME X. Parr 1. (1877, with numerous woodcuts). . . . PriceO 9 O 012 0 » 2. (1877, containing 27 Plates) . . . . . SL 206 110 0 » 9%. (1877, containing 6 Plates) . . . . . S57 G18.20 14 0 » 4. (1878, containing 9 Plates) ... . . . Ba Ok ae Gs 110 4 » 5. (1878, containmg 8 Plates) . . . .. Pre! eee? Jha 0) 012 0 », 6. (1878, containing 9 Plates) . . . .. Siok neo 110 0 » 7. (1878, containing 7 Plates) . . .. . » 018 O 1 4 0 », 8. (1878, containing 8 Plates) . . .. . 55 Or Eap Oe. 1 0 0 The ‘Transactions’ and other publications of the Society may be obtained at the Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. CONTENTS. IX. A Monograph of the Ostracoda of the Antwerp Crag. By Gzorcr SrewaRrpson Brapy, U.D., .LS., CM.ZS., Professor of Natural History in the University of Durham College of Physical Science, Newcastle-upon-Tyne . . . page 879 THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON, Tue scientific publications of the Zoological Society are of two kinds—* Proceedings,” published in an octavo form, and “ Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published by the Committee of Publication. From fifty to seventy coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Amongst such illustrations, figures of the new or rare species acquired in a, living state for the Society’s Gardens are often given. The “ Proceedings ” for each year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the preceding year. They may be obtained with black or coloured illustrations. : The “Transactions ” contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are published at irregular intervals; but not less than three parts are usually issued in each year. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five: pounds. The publications of the Society may be purchased at the Society’s Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. P. L. SCLATER, August Ist, 1878. Secretary TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Vout. X.—Parr 9. PRINTED FOR THE SOCIETY : SOLD AT THEIR HOUSE IN HANOVER-SQUARE; AND BY MESSRS. LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. August 1st, 1878. Price 12s. Taylor and Francis, Printers, | {Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. To the Public. Lee d. EE Let p VOLUME I. (1833-1835, containing 59 Plates) . . . Price 313 6... 418 0 Pawn 1. (WBBB) 5 gb ee ks sce eC AEE NS ne eee as nd a »» 2. (1834) ELS Bed (ksi 17) omen aMn as eae cae aT Om Toner M Aid gO WEY ET 8) ape a) phe Ase (LBB a Reick Maweery «kc ert altatels ae ceOl Tolan ive) top annie ed OED Naas 114 0 VOLUME II. (1836-1841, containing 71 Plates). . . Price 4 0 O 5 6 6 PR TUL y (LRSO)reWase ee We teste ig a admel mae aNen ite GN aM Tales I 8Cyl Uy De Pea he » 2. (1838) ESS FURBO). 8 (Sue 1 MRE Nes ORILEY WR es CoA a Bae OS Gt Oger eee MOT tne CEMA ante Tas a ance UN eb ka he den te sae TOES Ke) 200) » 5B. (1841) VOLUME III. (1842-1849, containing 63 Plates) . . Price 3 8 6 . . . 411 O Part 1. (1842) Pa TO apa ee RE MOR PURE At wer Scy tO 019 6 Ne eos s (MO AA Yar i? eer Ret ae MRS Sia amar os Na Cnee smd t MM SARA Ee 019 6 5, 4. (1846) peti: (ESAS). ico cette hier Acer saa.c meet an aN ty are tw Ree ny Olea) 010 6 Ryn Ose (LAG) ica tee Medicerens tt or tre eae WORE km ano mer Ote Othe gC LO) 010 6 VOLUME IV. (1851-1862, containing 78 Plates). . . Price 6 2 6 8 2 6 Panel M(TS5T) cP ecaeehebeak ie Nel keno nem e 53 sO 29.0 012 0 EDM BE OD os, ak aC MMS in -Re Sre O() LER 1 5ead » 8. (1853) een (LBD CEA tem lara both where oe eam cn uae ana, MORE OH O 14 0 SPR (ISHS) icc te) Petes opel ieee hh pom Aire Nhe panera TORN Sag) Lodo » 6. (1859) . Alpe chasse eR ee bn OP LS O 14 0 by tis SEC2 Le (1861) Be a gee eRe amedie Me aR et GOD LONED dacs be) Behl Lp HOCH LAOS) tik. Cees ne Ameo ai eue tant ann soto al sity O 14 0 VOLUME V. (1862-1866, containing 67 Plates). . . Price 5 3 6 619 0 aR Dal (USG2) ee hoy Vetoes MnrwMNancs aul Cait i Role ats IO MALS SD 14 0 pe CL SES) sua mane Macao lols tncmmey co on cn ONL ae Ae SO: Fee (LUBE Mose LeMay Pha, Nae annie ness gear SOAS 110 0 SAE ROG) eek Meee tetas 2 t's Shey ANI Sapa Ww a ne 110 0 Spt a ( DOGG) 5 ina mecca emer inert mod ice ee ed Us RUMORS aie tor GS 110 C VOLUME VI. Gerd containing 91 Plates) . . Price ll 5 O 15 0 0 Pan 1 (18GG ET. Se Me cea ee an a a 116 0 99 ae ABB Ze ae tana Uhh aea itl) Men Ov anh AUR aM ee aL aoa) 116 0 Pigs bie eo oy A RAE Ramen nt ce NCR NMS CURE neg abd, 1D, AOAC) 116 0 SPICE LUBE ZY) RS BP Wn BN NY dt A ae 116 0 » 5. (1868) . oe vile es O 116 0 » 6. (1868) . Site Be ATR) 116 0 » 7. (1868) . SoG O 28 0 » 8. (1869). pan Dera O 116 0 [Continued on page 3 ahs Wrapper.] To Fellows. To the Public. VOLUME VII. (1869-1872, containing 73 Plates) . . Drie atte eO: io TE ay ES W((IU SEES) 1S A a he. T10e0 » 2. (1870) . Se Ma ONS 110 0 » 93. (1870). bp eel ea L1O.%0 MAAR EVO pra aia a Sl ES gO 116 0 Perm wir. eRe ag 2 4 0 EWN e rey a Ne Fetey dee A: 116 0 yt 871): ies ety eet 116 0 » 8 (1872). Heaton Bye = 110 0 VOLUME VIII. (1872-1874, fag 82 Plates) . Price 9 8 3. . 12 11-0 Perks (ESR) iste coat ety crea 53. 0-10.56 014 0 » 2. (1872). iy el A Dare iO » 3. (1872). Cera 116 0 » 4 (1873). » 015 9 Ets 32/5. (1873): a0 ES ea 140 » 6. (1873) . pyaO? 19 226 i GO 3» 7 (1874) . 0 1628 120 » 98. (1874). OPI SEO 1 4 0 » 9. (1874) . 3 AAG 220 VOLUME IX. (1875-1877, containing 99 Plates) Price 12 0 O 1630.20 Parr 1.. (1875, containing 22 Plates) . . ... 4, (I11 6 2.2 0 » 2. (1875, containing 12 Plates) . . . .. «4 Ill 6 220 » 8. (1875, containing 5 Plates). . . . . . » O18 0 14 0 » 4. (1875, containmg 14 Plates) . . . . .- pod eG 2 2 0 » 5. (1875, containing 9 Plates). . . . . . 5 %I1 2 6 110 0 aber », 6. (1876, containing 4 Plates). . . . . . oO Oy 26 014 0 » %. (1876, containing 7 Plates). . . . . . Ae “On TSE SD a [pant EL », 8. (1876, containing 10 Plates) . . . . -. 5 O18 0 14 0 >, 9. (1876, containing 6 Plates). . . . . . Bie Aes Ae 30 112 0 », 10. (1877, containing 5 Plates). . . Ay te <6 110 0 Lie Sth containing 5 Plates and Title a Taded) Ree jroo bagel) A NRee! AAA 0 VOLUME X. Parr 1. (1877, with numerous woodcuts). . - . Price O 9 O 012 0 » 2. (1877, containing 27 Plates) . . ..- - ge A a G 110 0 sie. 0S7 7. containme 6 Plates)? 2 2. 0-18, 0 1 4 0 » 4. (1878, contaiming 9 Plates) . . - ~ - oper Dy HG) tag Oe » 5. (1878, containmg 3 Plates) . ... - Se Ug at! Tiss) 012 0 », 6. (1878, containing 9 Plates) . - -- - Drolae eae 110 0 » 7. (1878, containing 7 Plates) . - . . - eae OU LGe a fae: gy 6 » 8. (1878, contaming 8 Plates) . - - - - POLLO 1 -G0 » 9. (1878, containing 4 Plates) . - - - - op One OO 012 0 The ‘Transactions’ and other publications of the Society may be obtained at the Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. CONTENTS. X. On the Brain of the Sumatran Rhinoceros (Ceratorhinus sumatrensis). By A. H. Garrop, /.A., F.R S., Prosector to the Society . . . . . . . page 411 XI. A further Contribution to the Knowledge of the existing Ziphioid Whales. Genus Mesoplodon. By WituiaM Henry Firower, MRS, V.P.ZS. . . . . «415 THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological Society are of two kinds—“ Proceedings,” published in an octayo form, and “ Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published by the Committee of Publication. From fifty to seventy coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings ”’ for each year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the preceding year. They may be obtained with black or coloured illustrations. The “ Transactions” contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are published at irregular intervals; but not less than three parts are usually issued in each year. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five pounds. The publications of the Society may be purchased at the Society’s Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. a P. L. SCLATER August Ist, 1878, Secretary. ; —_—-..-- TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Vout. X.—Parr 10. LONDON: PRINTED FOR THE SOCIETY : SOLD AT THEIR HOUSE IN HANOVER-SQUARE; AND BY MESSRS. LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. February 1st, 1879. Price 16s. Taylor and Francis, Printers, | [Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. To the Public. £s. d. £s. d. VOLUME I. (1833-1835, contaming 59 Plates) . . . Price 313 6... 418 0 PART LT: (1SBB) or osh. Venue. Sethe ike: Leora ch ons h Onda itea ee ae eee Lang) » 2 (1834) pial; Aa (131215) a ae a Bee nO tn estes even hati eek Olid ire pe a Seis (ESBS) ei oem 7 aha at Ula Cate tad ee ie mene cata Reva 114 0 VOLUME II. (1836-1841, containing 71 Plates). . . Price 4 0 O 5 6 6 Pawnee! S86) cir eee, oy yh 2 ne aes camel age ares abi Were AE 112 0 » 2 (1838) SO Be (BBO) eh apne can) Le aaa Me aon reieeyale lee Mullssst ath OMe iy Openness nN tual nO 6 Ae (IS4O)2 Me oO A ee ee iar i Ears Ree ee OTT » 5. (1841) VOLUME III. (1842-1849, containing 63 Plates) . . Price 3 8 6 . . . 411 0 Part 1. (1842) iP Raa. ea ee Ree JRO) IAS 8 019 6 Sp Ose (A OAAIES cen ayia be Ws Sara C its » O14 8 019 6 » 4. (1846) tye a ks 1S) eR eae ae eae Loe ly tM es ery 0) ycrel (D) 010 6 SIGS LSAQ) , 6. (1876, containing 4 Plates). . . . .. » O10 6 014 0 ae » 7. (1876, containing 7 Plates) . Bi a oe » 015 9 LAO, ; » 8. (1876, containing 10 Plates) . . . . . » 018 0 1 4 0 ay », 9. (1876, containing 6 Plates). . . . .. Paved ore: taal) 112 0 rhe », 10. (1877, containing 5 Plates). . . AMA Ch 2 NG 110 0 aig » 11. (1877, containing 5 Plates and Title Saal Index) Bo sOslore 9 LEO VOLUME X. Part 1. (1877, with numerous woodcuts). . . . PriceO 9 O 012 0 aa » 2. (1877, containing 27 Plates) . . .. . Pe Dae ie 110 0 » 9. (1877, containing 6 Plates) . . .. . » 918 0 1 4 0 » 4. (1878, contamimg 9 Plates) . ... . Py Wee St) 110 0 » 5. (1878, contaimmg 3 Plates) . . .. . » O 9 O 012 0 » 6. (1878, containmg 9 Plates) . . .. . SAL ae 110 0 » 7. (1878, containing 7 Plates) . . . . . » 918 O 140 > 98. (1878, containing. 8 Plates) . .- ..- .- 5 Oma O 100 » 9. (1878, containmg 4 Plates) . ... . > Oren O 012 0 », 10. (1879, containing 6 Plates) . . .. . Oza O 016 0 The ‘Transactions’ and other publications of the Society may be obtained at the Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. CONTENTS. XII. Notes on the Fins of Elasmobranchs, with Considerations on the Nature and Homo- logues of Vertebrate Limbs. By St. GzorcE Mivart, V.P.Z.S.. . page 439 THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological: Society are of two kinds—“ Proceedings,” published in au octavo form, and “‘ Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published by the Committee of Publication. From fifty to seventy coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings” for each year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the preceding year. They may be obtained with black or coloured illustrations. The “ Transactions”? contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are published at irregular intervals; but not less than three parts are usually issued in each year. Fellows and Corresponding Members, upon payment of a Subscription of £1 Is. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five pounds. The publications may be purchased at the Society’s Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. P. L. SCLATER, February Ist, 1879. Secretary TRANSACTIONS OF THE ZOOLOGICAL SOCIETY. OF LONDON. Vou. X.—Parr 11. LONDON: PRINTED FOR THE SOCIETY: SOLD AT THEIR HOUSE IN HANOVER-SQUARE; AND BY MESSRS. LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. February 1st, 1879. Price 12s. Taylor and Francis, Printers, ] [Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. To the Public. Ana eh CIS CE VOLUME I. (1833-1835, containmg 59 Plates) . . . Price 3138 6 . .. 418 O Pam] 5) (1838) i acais kerosene eee tie aes oie Shik MO Lat Bini hieen Ya Ook OA AO »» 2 (1884) : 1 BRUT SBD) oe io ah eR te ge kako ds ROL LO OR Titec EAI (ISEB) ec tty eo) Nek mth tate depart 50 HOGA ROT tings 114 0 VOLUME II. (1836-1841, containmg 71 Plates). . . Price 4 0 O 5 6 6 Pammule) (1836) sok ok Beh wae) hese eee als Pann rnc cA ek: ed 0) 1 as P5250) » 2. (1838) CeiB (UBBO) or cs aU «egy Rees Sie) A ON SUM OM nee OS en non ee aay NAS (TREO) Oe SPC are Sanh Sec te St OR Ca anaes cate ch UTS NU » 5. (1841) VOLUME III. (1842-1849, containing 63 Plates) . . Price 3 8 G6... 411 0 Parr 1. (1842) ‘ Sa LOLS) acter ols omate Va Mp up paneer at Mane oe aun A on eomerns hove Au) aT Oats ADE CRM) lek set. ©, 20 Re Ne ae Tt NAY Samy QUAI AB Ria, Aad ANG I) Cengs » 4 (1846) A bse GL BAB yagi he PRINS She Ly ge eal ee) 010 6 pO (UGAD) eM ae oii hs Stead aes COLIN ant GN LOL ey 010 6 VOLUME IV. (1851-1862, containing 78 Plates). . . Price 6 2 6 8 2 6 Parke (ASR Lea eet SSR ine kee rare A teak ae arg naaE) 012 0 sa ietss CRB) rey: SenBeoNh al AU cs, Vlafireter sy eae eigen BS TS ay SieaiG. » 8. (1853) yy 4. (1857) . OT ean0 140 yD. (1858) . » 018 0 1 4 0 PWe@sis (Lewy)! scars any Old toda 0) Tey AsO » 7, sec. 1. (1861) 0 BL <9 IIe 0 » 7, sec. 2. (1862) oye On US. 0 1 4 0 VOLUME V. (1862-1866, containing 67 Plates). . . Price 5 3 G 619 0 Part 1. (1862). » 9018 0 14 0 » 2. (18638) . Ben! OREN a Bee Searcher ak gead LO) Keteat®) 5 is cn 0) PS SRG ale een ae ps. Un eae a Rem Reig pce ee 110 0 » 4. (1865) . ta sae Poo SO. SOG) Mie Pe lipald bata ay 110 ¢ VOLUME VI. (1866-1869, containing 91 Plates) . . Price 11 5 O 15 0 0 PART DL, ((USGG)isers. eee ee oe 33 RL SO. 116 0 » 2 (1867). PN tra (0 116 0 » 8. (1867) . Raving ee 116 0 ys) 4. (1867) . Peer ts Gna 2 116 0 > 9. (1868) . is Sheek epi 0) 116 0 » 6. (1868). BARNS ie//iNy 0 116 0 » 7. (1868) . ee) a GO 2 8 0 » 8. (1869). Waele a kO) 116 0 Continued on page 3 of Wrapper] ‘ To Fellows. To the Public. VOLUME VII. (1869-1872, containing 73 Plates) . Pals BPO jure o DAO OTC GIL ee ita gia ae SR EEOC ee eR 2 I 126 110 0 PPA ee eR eG SAR ia ge 110 0 So ROMULUS can a ee ante Maer a a ay 110 0 » 4. (1870) . pie Barna 30 116 0 » 5. (1871). Pea me ONAL, 2 4 0 PoOROreHLGm yey Si aet aoe dl, em Rue Lt 116 0 52 Ceo N OSV es ete PaNGe) ii AR YG) 116 0 reese eo) iy. oe hig 110 0 VOLUME VIII. (1872-1874, containing 82 Plates). . Price 9 8 3 1211 0 Puree laeReeyeeh oo. Ss: £20.10) 6 014 0 » 2. (1872) . el Lee G Penn Oy » 93. (1872). parimeadlt » Veiscy Oh 116 0 » 4s. (1873) . Bey aes, ey 6 » 5. (1878). » 018 0 140 » 6. (1873) . ip OLIO 276 160 Lee, (LSTA) i aun” G see) 8. (1874) . a GELS... 0 14:0 » 9. (1874). 1s eG a 2 0 VOLUME IX. (1875-1877, containing 99 Plates) Price 12 0 O 16 0 0 Parr 1. (1875, containing 22 Plates) Peed Bema Was 2 2,0 » 2. (1875, containing 12 Plates) jal al SL Gi 2.2.0 » 3. (1875, containing 5 Plates) . Sea key) SOR aek Oras 1 4 0 » 4. (1875, containing 14 Plates) . . . . . yo SG 2° 2.0 3, 9. (1875, containing 9 Plates) . Bee RONG 110 0 f », 6. (1876, containing 4 Plates) . oe, OW LOR KG O14 0 d » 7. (1876, containing 7 Plates) . A On Lon ug, LEV O. » 8. (1876, containing 10 Plates) $720 LO 1 4 0 », 9. (1876, containing 6 Plates) . asiey Mab aon O) 112 0 », 10. (1877, containing 5 Plates) . nay SERS SL aan 110 0 ‘ », 11. (1877, containing 5 Plates and Title and Index),, 015 9 UAE TEX) VOLUME X. Parr 1. (1877, with numerous woodcuts). . . . Price 0 9 O 0 12 >» 2-| (1877, containing 27 Plates) . . . .. SR BENG 110 0 » 9. (1877, containmg 6 Plates) . . . . . » O18 O 14 0 z weal (azo. containing 9:-Plates)* Mh) ) LL: (1879, containing ‘5 Plates) 9. 2.) SAO OO O12 The ‘Transactions’ and other publications of the Society may be obtained at the Offive (11 Hanover Square, W.), at Messrs, Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. CONTENTS. XIII. On the Mechanism of the Odontophore in Certain Mollusca. By Parrick GEDDES. page 485 XIV. On the Hearts of Ceratodus, Protopterus, and Chimera, with an Account of undescribed Pocket Valves in the Conus arteriosus of Ceratodus and of Protopterus. By HE. Ray Lanxester, 1.4., P.R.S., Professor of Zoology and Comparative Anatomy in University College, London . : : : . page 493 THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tux scientific publications of the Zoological Society are of two kinds—“ Proceedings,” published in an octavo form, and “ Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published by the Committee of Publication. From fifty to seventy coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “‘ Proceedings ” for each year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the preceding year. They may be obtained with black or coloured illustrations. The “ Transactions” contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are published at irregular intervals; but not less than three parts are usually issued in each year. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five pounds. Such of these publications as are in stock may be obtained at the Society’s Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. P. L. SCLATER, February Ist, 1879. Secretary TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON Vou. X.—Parr 12. LONDON: PRINTED FOR THE SOCIETY: SOLD AT THEIR HOUSE IN HANOVER-SQUARE; AND BY MESSRS. LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-hOW. June 1st, 1879. Price 21s. Taylor and Fraucis Printers, ] [Red Lion Court, Fleet Street. TRANSACTIONS OF THE ZOOLOGICAL’ SOCIETY OF LONDON. To Fellows. To the Public. £s. d. El Bb ee VOLUME I. (1833-1835, containing 59 Plates) . . . Price 318 6 . . . 418 O PART Voi(LSOs) ac latans yer ek uelauee vat cy a be seth cote 55 tO TAWA, Bon Ce eae oO Ou ae » 2. (1834) Ae RU CEBBBY oo ads # Mapamenpe Plies Snir ean eee? elle es Tg I ies a NGL REB nae sty aaNet c eaten vie A) Jentem aa ee Neg 114 0 VOLUME II. (1836-1841, containing 71 Plates). . . Price 4 0 O 5 6 6 Parr (VGEO) ecient pills velia | ki anpie eerste aC Ns ae St ARID 112 0 » 2. (1838) SAM HOG kst31°) Palas Deen Die | UNE ae Ai iS UME brat iLAn eeyniniieser opt OEY, A ONY SAL Aon EA ILS) AiG Re ECCS) panei Ame Me va OL emule ae T eMac: PMS NSE) » Dd. (1841) VOLUME III. (1842-1849, containing 63 Plates) . . Price 3 8 G6 . . . 411 0 Parr l. (1842) WU 2 oT SMB CIE: (C1 ORR iment Ma taka Ra 019 6 Bats eo (oe) arti ok Par eo Au en Us moe alas OT fo, 019 6 » 4. (1846) BERN pH (RS SS) OTS mien ara Tit or AIAN Ree He edn Sul O07) ll 0) 010 6 ESTB 3s (LRAQ\ ORO stn eee Saelcclianeine ORNS a ee re oan OMe FG) 010 6 VOLUME IV. (1851-1862, containing 78 Plates). . . Price 6 2 6 S26 PART LOO LES (0) 045. Mehmhe rei ee ae eee nnronn Os aO.AGO 012 0 SOs (LSE 79671. tae ass gute caaeae Ot Sian ep reas » 018 9 ASSO » 3. (1853) 4. (1857) . » 018 0 140 » 5. (1858) . RTO 140 Dv Gei(PBa9) ose: Ewes (0 Lane » 7, sec. 1. (1861) JOU OS 8 REO » 7, Sec. 2. (1862) » 018 0 14 0 VOLUME V. (1862-1866, containing 67 Plates). . . Price 5 3 6 619 0 arunale, (MSGR) Yee ater saan woes ance ae ean Se On SHO 140 SM es (LOGO) Si eae We SH Bieeids Wea in eae Yai wastes ar sels OEE Que Lag 0: Eta SGA) eo caput kuch te siknons A vp ateasernnttes Skat Gola ae iO 110 0 site Ai. 0# (SGD) recive ges ese pyetaawerile re ay a’ vc maya Sy cede 2 TaN Sake AGRA SOHN 110 0 Bh Diy (LOO) ra igen eits cukaak Pe Gm EL oul a eye eeke ls Hike beara 110 0 VOLUME VI. (1866-1869, containing 91 Plates) . . Price 11 5 O 15 0 0 Pann 19 (D8GG6) soe rice 5 ha cael. + BSD Saas Mier gts) 116 0 » 2. (1867). itn Dar 116 0 i Bo(seayn Rial er Ale! 116 0 » 4 (1867). at ARO 116 0 » 5. (1868) . #5 Tero) 116 0 » 6. (1868). Pasay Malar ie to) 116 0 » 7. (1868) . . 116 0 2 8 0 » 8. (1869). A dB ea) 116 0 Continued on page 3 of Wrapper.) To Fellows. To the Public. VOLUME VII. (1869-1872, containing 73 Plates) . . Pape Bitold re Part 1.(1869))) 3.) nes IED” bak kok GE P26 110 0 » 2%. (1870) . Bl iG 110 0 » 3. (1870). Pei a 2 le 110 0 » 4. (1870) . Pe LO 116 0 » 5. (1871). #18) 30 2 4 0 aa von Lome) is ae eee O 116 0 Ait PePnCLeva)ies ie laeg vu) 116 0 » 8. (1872). Prodi Us ee 8} 110 0 VOLUME VIII. (1872-1874, containing 82 Plates). . Price 9 8 3 -1211 0 PART USSR e Rares eens fo) hati oe OO nG: 014 0 3 2. (1872). spetival oad Ee e 220 4 at (872). ee 70 116 0 50.) Aa (1873). » 01 9 AVEO Pon Uys) heen » 9018 0 1 4 0 5 Go (1878) 4 Ew, 19. 6 160 » 7. (1874). » O16 6 1 2 0 » 8. (1874). oN OnIS uO 140 De Gat sz4y. SOLIS G 284.0 VOLUME IX, (1875-1877, containing 99 Plates) . . Price 12 0 O -16 0 0 Parr 1. (1875, containing 22 Plates) sb bie'G, 2 2 0 » 2 (1875, containing 12 Plates) Sih Se eee G 22-0 » 8 (1875, containing 5 Plates). . . . . . » 9018 0 1 4 0 » 4. (1875, containmg 14 Plates) . . . .. foi asd taal ln (3) 2 2 0 5 D- (1875, contaming 9 Plates). . . . . . Pe ag Gadde 5: PAO » 6. (1876, containing 4 Plates) . SAO 6 014 0 » 7. (1876, containing 7 Plates) . Aues » O15 9 nS Len », 8. (1876, contaming 10 Plates) . . . . . » 918 0 14 0 Pie (leas contam« Gi Plates) ii! eo ek yd) Lae 0 112 0 5, LO. (1877, containing 5 Plates). . . byte )id haba) 110 0 » 11. (1877, containing 5 Plates and Title send Tides) al OL aca Leb ‘ VOLUME X. : re Part 1. (1877, with numerous woodcuts). . . . PriceO 9 O 012 0 Ne Se ac(lor7eontaming 27 Plates): ji... 20. «ce. ap tdi) 6 110 0 ‘ » 98. (1877, containing 6 Plates) . . ... «4 018 0 ERA O » 4. (1878, containmg 9 Plates) . ... . De ee 110 0 » 5. (1878, containg 3 Plates) . . . =. Oe LO 012 0 » 6. (1878, containmg 9 Plates) . . .. . ADS ean 110 0 » 7@. (1878, containing 7 Plates) . . . . . » 018 O P 14 0 », 8. (1878, contaming 8 Plates) . . .. . Sy Che tal Oh stake 027-0 » 9. (1878, containmg 4 Plates) . . ... an OS OLTO, 012 0 », 10. (1879, contaming 6 Plates) . . .. . eA OwLe 50. 016 0 ,, 11. (1879, containing 5 Plates) . . . . . fg: O29) 20: 012 0 The ‘Transactions’ and other publications of the Society may be obtained at the Offive (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. CONTENTS. XV. Observations on the Uraniide, a Family of Lepidopterous Insects, with a Synopsis of the Family and a Monograph of Coronidia, one of the Genera of which it is composed. By J.O. Westwoon, M.A., F.L.S., &e. é : . page 507 THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological Society are of two kinds—‘“ cea published in an octavo form, and “ Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended to be published by the Committee of Publication. From fifty to seventy coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings ”’ for each year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the preceding year. They may be obtained with black or coloured illustrations. The “ Transactions’ contain such of the more important communications made to the scientific meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are published at irregular intervals; but not less than three parts are usually issued in each year. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five pounds. Such of these publications as are in stock may be obtained at the Society’s Office (11 Hanover Square, W.), at Messrs. Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookseller. P. L. SCLATER June Ist, 1879. ; Secretary. : aia = ie ee ee ee TRANSACTIONS OF ZOOLOGICAL SOCIETY - PATERNOSTER-ROW. aah i. ae ees way te ie . ; Price 248, TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. To Fellows. To the Public. £8... Bee ME VOLUME I. (1833-1835, containing 59 Plates) . . . Price 313 6 . . . 418 O Pann VATSSS) iy ew senso erieer oot ae dne lela Ka ate EOL uatso te oe tee ae katy an » 2. (1834) st eae LBOO) cnet lace catens” oetotpeeat mira nt sete) Wallvee ends Me ODEON ENO. i Gera a4) ree ise F 2 att he a ee coset NUS Cc Mag oh 114 0 VOLUME II. (1836-1841, containing 71 Plates). . . Price 4 0 O 5 6 6 Par U(TSEG) celeste chk eased ae he nete eater pine Be. ak 0) 112 0 » 2 (1838) Bec 3s (L889) LN aS et bead werk a ater meee eee eb O 5 Gace teide g: One Ere Ae TRAD i Lee Sal onto eee Reenter toda WeDo 2a dhe aL AN te. «ete aN OTL Oita » Dd. (1841) VOLUME III. (1842-1849, containing 63 Plates) . . Price 3 8 6 . .. 411 O Parr 1. (1842) ; S OPERTRABN EL a. OMEN ie Be) wea se Oh kes Ale PN ee Se OER 1 eae LOAAN IS Pore eo A rele Ne a eee tw RV Vistar hula bios SEO AGa ass 019 6 » 4 (1846) ' Sa 1 2) ey aS rami eee oes Sah ey Sipe es 010 6 er fen (bc! |S) Mien, ste Ags Pantie may haa as sain Doc Ai KO) 010 6 VOLUME IV. (1851-1862, containing 78 Plates). . . Price 6 2 6 8 2 6 ; PART VIs God). ON imap ence aatach vas coetetee ion Tau ies nl RO RO 012 0 Zh A LGOe) nes eres oe nee tel eis ee Meet at OLS no Brie) » 93 (1853) Since SCIBBZ) ia os See cls se Bad nl De ED is 40 Sie CISHSY Oe AON ie ae Os a een agate tt iE) 140 PPO LBDO) nee awe nents Srkts ene DRA t bac Tense) dese 140 A UP AIREC Les GLOGL)) ir ite thee loi tale cae we Sina cere OVER 1G Vesa Oy » 7, sec. 2. (1862) at OrlscuO 1 4 0 VOLUME V. (1862-1866, containing 67 Plates). . . Price 5 3 6 619 0 PAmew i CESC 2x aye en crests: aeons ae ire oka eae ROMA GD 140 ade Oe BOS sac Meee nae rei Mapu eal sonal | SUE ae EL OL AO) Leb fy eben TOGA) spe teae caer ieee, os koje a ie on eaha tee a We RRND hay ee Late orn 110 0 a Ao (SGD) ict | canara ein sHi ns a Top eeley eens lee ta 1.10;-0 pork Ben SOO) aca cer ue tit als 2 20a re Pa en Ae ena 110 0 VOLUME VI. (1866-1869, oe 91 Plates) Price 11 5 0 . 15 0 0 Parti. (1866)... .. : eal? pitt ea O 116 0 » 2. (1867). Pipl el geist ane) TG 70 5» 8. (1867). Meese: NO 116 0 », 4. (1867) . aU ong uO), 116 0 > 9. (1868). eget al Sy Am 0) 116 0 » 6. (1868). eal bree is 116 0 59: al. i (EBGB) shies Ar VAM vail PON alee Veemestnis ghee eae Mace aaa ean) 28 0 99 (oe LSBO) Ver a eens ote Waa ee ST ts Ne ae 116 0 Continued on page 3 an Wrapper. } To Fellows. To the Public. VOLUME VIL. (1869-1872, containing 73 Plates) . . wees 8 17-0 “aT ge 8 PARTI ASOD) a Maeemte py stinepaeen scales ess ti iG Oe a) 1 10)).0 » 2 (1870) . SOLE G 110 0 » 8. (1870) . Seat Te G 110 0 » 4 (1870) . praises Maly 116 0 De (1871). Peat) wu) ks Jane 0) 2 4 0 » 6. (1871). ai eaa (ag 116 0 » 7. (1871). Sian hue 7/0) 116 0 3 8 (1872). Spain Mie bas eee 110 0 VOLUME VIII. (1872-1874, containing 82 Plates). . Price 9 8 38 12,11 0 eS Rawabi (USAR a CONTENTS. XVI. Supplementary Notes on the Curassows now or lately living in the Society's Gardens. By P. L. Scuater, M.A., PhD., F.R.S., Secretary to the Society. . page 543 Last/of the ‘Papers:contained in Volo M8: ote tek ee ORT ‘ritex. of Species, ete. in Vol. Ko ea ee meade Liee tree ele oh Ne ee Titlepage and Contents to Vol. X. THE PUBLICATIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. Tue scientific publications of the Zoological Society are of two kinds—* a isa A published in an octavo form, and “ Transactions,” in quarto. According to the present arrangements, the “ Proceedings” contain not only notices of all busi- ness transacted at the scientific meetings, but also all the papers read at such meetings and recom- mended tc be published by the Committee of Publication. From fifty to seventy coloured plates and engravings are attached to each annual volume of the “ Proceedings,” to illustrate the new or otherwise remarkable species of animals described in them. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings” for cach year are issued in four parts, on the first of the months of June, August, October, and April, the part published in April completing the volume for the preceding year. They may be obtained with black or coloured illustrations. The “Transactions ” contain such of the more important communications made to the scientific — meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are published at irregular intervals; but not less than three parts are usually issued in each year. Fellows and Corresponding Members, upon payment of a Subscription of £1 1s. before the day of the Anniversary Meeting in each year, are entitled to receive all the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged for them to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1861, if they exceed the value of five pounds. Such of these publications as are in stock may be obtained at the Society’s Office (11 Hanover Square, W.), at Messrs, Longmans’, the Society’s publishers (Paternoster Row, E.C.), or through any bookselier, P. L. SCLATER, October Ist, 1879. Secretary. ——— oe bate “<< “at