TRANSACTIONS OF THE ZOOLOGICAL SOCIETY OF LONDON. VOLUME VIII. LONDON: PRINTED FOR THE SOCIETY : SOLD AT THEIR HOUSE IN HANOVER-SQUARE: AND BY MESSRS. LONGMANS, GREEN, READER, AND DYER, PATERNOSTER-ROW. 1874. " beTS Ve LET igh TES : ‘ ae a) paste apiyety tae ’ ) i Mi CONTENTS. I. On Risso’s Dolphin, Grampus griseus (Cuv.). By Wituiam Henry Fiower, F.2.S., F.R.C.S., V.P.Z.S., Hunterian Professor of Comparative Anatomy and Conservator of the Museum of the Royal College of Surgeons of England . . . .« page 1 Il. A List of the Birds known to inhabit the Island of Celebes. By Artaur, Viscount WALDEN: FRS.; Presidemtiof the SOciety «iin te Aa erste pa dorqeade Fee Ill, Appendix to a List of Birds known to inhabit the Island of Celebes. By ARTHUR, Viscount Wa.peEN, F.R.S., President of the Society . . . . . . . . 109 IV. On Divornis (Part XVII.): containing a Description of the Sternum and Pelvis, with an attempted Restoration, i Aptornis defossor, Ow. By Professor OWEN, PBS: PS Glee) ee 2 RM ei lee 5 ae SPC V. On the Form and Structure of the Manatee (Manatus americanus). By Dr. JAMES Mort, F.L.8., F.G.S., &c., late Prosector to the Zoological Society . . . 127 VI. On the Recent Ziphioid Whales, with a Description of the Skeleton of Berardius ammouxi. By Wituiam Henry Fiower, /.R.S., V.P.Z.S., Hunterian Professor of Comparative Anatomy, and Conservator a the Museum ae the Royal College of Surgeons of England . . . ... : oth i OS VII. On the Organization of the Caaing Whale, Globiocephalus melas. By Dr. JAmzs IMIG REE; -Pe SS. Bs GS 5tQe.. ower OS. ay, eg ee ae eh ees, Zoo VIII. A Deseription of the Madreporaria dredged up during the Expeditions of H.M.S. ‘ Porcupine’ in 1869 and 1870. By Professor P. Martin Duncan, MB. (Lond.), F.R.S., F.G.S., Professor of Geology to King’s College, London, &c. . . . 308 IX. On the Osteology of the Marsupialia. (Part III.) Modifications of the Skeleton in the Species of Phascolomys. By Professor Owsn, F.B.S., F.ZS.,&ce. . . 345 X. On Divoryis (Part XVIII.): containing a Description of the Pelvis and Bones of the Leg of Dinornis gravis. By Professor Owsn, F.RS., P.ZS., de. . . 361 IV CONTENTS. XI. On Dinornts (Part XTX.): containing a Description of a Femur indicative of anew Genus of large Wingless Bird (Dromornis australis, Owen) from a post-tertiary deposit in Queensland, Australia. By Professor Own, F.RS., FLS., de. . 381 XII. On the Axial Skeleton of the Ostrich (Struthio camelus). By St. GEoRGE Mivant, FURS 5 us te SES RS OS es a 80 XIII. On the Osteology and Dentition of Hylomys. By Joun Anverson, M.D., Curator of the Indian Museum, and Professor of Comparative Anatomy in the Medical College Caloutia 0. aos seu ath ol hence ee) a) A eae eed a tOS XIV. Report on the Hydroida collected during the Hes of HM.S. ‘Porcupine. By Professor G. J. AutMan, FLRS. . . 5 = ere «469 XV. On the Osteology of the Marsupialia. (Part IV.) Bones of the Trunk and Limbs, Phascolomys. By Professor Owen, C.B., PRS. FZS. . . . . . . 483 XVI. Researches upon the Anatomy of the Pinnipedia.—Part III. Descriptive Anatomy of the Sea-lion (Otaria jubata). By Dr. James Muniz, PLS. PGS, &c., late Prosector to the SOG a. CoA tat > CME ose KOPF 4oeccucte OSD ae Mae, Te ALN S AOE TONS OF THE ZOOLOGICAL SOCIETY. I. On Risso’s Dolphin, Grampus griseus (Cuv.). By Witttam Henry Fiower, F.R.S., F.R.CS., V.P.ZS., Hunterian Professor of Comparative Anatomy and Conservator of the Museum of the Royal College of Surgeons of England. Read June 6th, 1871. [Puates I. and IT] ON the 28th of February, 1870, an animal belonging to the Order Cetacea was taken in a mackerel-net near the Eddystone Lighthouse, and brought into Plymouth. It was afterwards sent to the Columbia Fish Market, London, was then exhibited in the Kingsland Road, where I first saw it on the 4th of March, and finally lodged at Mr. E. Gerrard’s workshop, where the skin and skeleton were prepared for the British Museum, and where the very carefully executed drawings (PI. I. figs. 1-3) which accom- pany this communication were made, by Mr. Sherwin, under my superintendence’. The animal was a female, and (as afterwards appeared from the condition of the bones) perfectly adult. Moreover there was evidence that she had recently given birth to a calf. The principal dimensions were as follows (all the measurements are taken in a straight lines with calipers, unless otherwise stated) :— ft. in. Length, from anterior edge of upper lip to notch in middle of caudal fin. 10 6 From upper lip to anterior edge of dorsalfn. . . . ..... . 311 Exom.upper'lip toanterior‘angle of eye... soe ee eee Ea ' T have much pleasure in acknowledging the great facilities which Mr. Gerrard afforded me in examining both the specimens described in this communication. /- VOL. VIII.—PART I. March, 1872. B bo PROFESSOR FLOWER ON RISSO’S DOLPHIN. From upper lip to blow-hole (following curve). . . i 7 From upper lip to junction of anterior edge of nectar fin ih ‘thie body . ‘ From upper lip to angie a mouth : . H (ac) (ao) J“) SS) SS) Ro) Rho bole IRIE tole From upper lip to anus Length of eye-aperture Roca From posterior angle of eye to ear-aperture . From angle of mouth to anterior angle of eye Pectoral fin, length from junction of anterior edge with aryae to coe Pectoral fin, from junction of posterior edge with body to tip . Pectoral fin, breadth at base . Pectoral fin, greatest breadth Height of dorsal fin . Breadth of caudal fin 5 Vertical height of body, at the eye . Vertical height of body, immediately behind “pe ssadtorals Vertical height of body, immediately in front of dorsal fin Vertical height of body at pudendal orifice po je tole ble BPD eFrenNorooOorrFCcooOoOOoN CF woonrwwoor Da a In general form the animal more resembled a Glodicephalus than any other Cetacean with which I am acquainted—so much so that I was at first inclined to refer it to that genus. The front part of the head was furnished with a similar rounded adipose pro- tuberance, though developed to a less extent; and the pectoral fins had almost the same low position and narrow falcate form, though they were considerably less elongated; the dorsal fin, however, was higher. A better idea of the general form of the body can be obtained by an inspection of the figure (Pl. I. fig. 1) than by any description. The crescentic aperture of the blow-hole, 13 inch in width, was placed directly over the eyes. The minute external auditory meatus (no larger than a hole made by pricking the skin with a needle) was an inch lower than the eye-aperture, and 34 inches behind its hinder angle. From the blow-hole the upper contour of the head was continued at first nearly hori- zontally forward, then curved pretty rapidly downwards to form the nearly vertical anterior surface of the head. This was somewhat hollowed in the middle line, and expanded below into a thick, rounded, very short snout, which projected 14 inch further forwards than the edge of the lower lip. The lower jaw was narrowed in front. The opening of the mouth was directed slightly upwards towards the angle, but finally took a sweep downwards near the junction of the upper and lower lips. There were no traces of teeth in the upper jaw; but there was a deep narrow groove along the alveolar margin of the palate. In the lower jaw, near the anterior or symphysial region, were three rather small teeth on each side, the apices PROFESSOR FLOWER ON RISSO’S DOLPHIN. 3 of which were worn quite flat. The hinder edge of the most posterior of these was 24 inches from the anterior edge of the chin; and the three occupied a space of 1:7 inch. The middle one was slightly larger than the others, the truncated crown being 0°3 of an inch in diameter, and projecting 0-2 of an inch above the level of the gum. The space between the right and left anterior tooth was 1:2 inch. On the under surface of the body there was a deep median depression containing the vulva and anus, 8 inches in length, and bordered in its anterior portion by prominent Jabia. The nipples were each placed in a slit, 1 inch long, on the side of the median pudendal fissure, and 13 inch distant from it. The mammary glands were largely developed, the internal reservoir being filled with milk. This circumstance, combined with the dilated, vascular condition of the uterus, showed that the animal had recently given birth to a young one. Perhaps the most noticeable external character was the very marked and peculiar coloration. The most prevalent tint was grey, varying in some parts to pure white, and in others to deep black; but the light parts of the head and anterior portion of the body had a yellowish wash, and the dark parts a slight bluish or purple tinge which varied much in different lights. The length of time that the animal had been out of the water may have modified these colours somewhat; but the general disposition of the light and dark shades, as shown in the figure, were evidently natural. The head and the whole of the body anterior to the dorsal fin was generally of a lightish grey, variegated with patches of both darker and whiter hue. The eye was surrounded by a small oval patch of black. The lips were mottled with black. There was a large, nearly black, patch on the top of the head, extending backwards a short distance behind the blow-hole, and on each side towards, but not reaching, the eye. Both surfaces of the pectoral fin were nearly black, very finely mottled or dappled with grey, and becoming darker towards the tip. The neighbourhood of the axilla was of the same dappled colour. Behind the anterior edge of the dorsal fin the general colour of the surface, including the dorsal and caudal fins, was nearly black, though with a large light patch on the upper part of the side directly above the pudendal orifice. The middle of the belly, as far back as the pudendal orifice, was greyish white. But what gave the most remarkable and characteristic appearance to the animal was the presence of conspicuous, but most irregular, light streaks and spots, scattered over the whole of the sides from the front of the head to about two feet from the end of the tail, where they ceased, at least on the lateral surfaces. These markings were naturally most conspicuous in those parts of the surface where the ground-colour on which they were placed was dark. The streaks or lines were of various lengths, and running in all directions in a most fantastic manner, some parallel, some crossing each other, and some forming sharp angles, zigzags, and scribble-like patterns. When most completely developed, and not interfered with by others, each linear marking was of a compound character, consisting of a very narrow, central white line, with an irregular, black, B2 4 PROFESSOR FLOWER ON RISSO’S DOLPHIN. mottled border, which, again, was separated from the general dark colour of the surface by another white line, so that there were three white and two black lines, altogether nearly half an inch in width. In some places the central white line was absent; and then the marking showed only a dark centre, bordered by white. There are also many round and oval patches of white, generally with a dark centre. On close inspection it was seen that these dark lines and patches were really formed by aggregations of minute black dots and fine linear streaks placed transversely to the main line, and that they depended altogether upon the disposition of the pigment in the cuticle, the peculiar coloration passing through its entire thickness. Nothing but the drawing can convey any idea of the extraordinary and irregular manner in which these markings were distributed. Though there was a general corre- spondence between their arrangement on the two sides, there was no symmetry in detail. They were entirely absent from the dorsal, pectoral, and upper surface of the caudal fins, though on the under side of the right lobe of the latter were some broad, rather indistinct, white lines, parallel with the anterior border of the lobe, and following its curve. There were no corresponding markings on the other lobe. The viscera generally, as far as I had an opportunity of examining them, appeared closely to resemble those of Globicephalus. The stomach was nearly empty, containing only a little fluid, and in its last compartment a single crystalline lens, apparently of a small Cephalopod. Skeleton—The condition of the bones showed that the animal was adult but not aged: all the epiphyses of the limb-bones were completely united with the shafts ; and the disk-like terminal epiphyses of the bodies of the vertebra were likewise joined with the rest of the bone, with the exception of a few in the lumbar region, which still remained distinct. In general appearance the skeleton presents the same kind of resemblance to that of Globicephalus that the external figure of the animal does, the proportions of the dif- ferent regions of the trunk being very similar; but as the vertebr are more numerous, especially in the lumbar region, they are individually shorter from before backwards. The spinous and transverse processes are also longer and more slender, approaching in this respect Delphinus and Lagenorhynchus, and deviating greatly from Orca and Pseudorca. The most noticeable peculiarity of the vertebral column, taken as a whole, is the very feeble development of the metapophyses. The total number of vertebre is 68, which may be divided into 7 cervical, 12 thoracic, 19 lumbar, and 30 caudal’. The seven cervical vertebre are all firmly united together by the laminz of their arches and the spines; but the body of the seventh is quite distinct from the sixth, and that of 1 Exactly the same numbers as those found by Fischer in a specimen of Grampus griseus, stranded in 1867, on the west coast of France. (Annales des Sciences Naturelles, 5th ser. viii. p. 363, 1867.) PROFESSOR FLOWER ON RISSO’S DOLPHIN. or the latter only imperfectly joined to the fifth. The bodies of the remainder are com- pletely consolidated together. The spines of all seven are joined into a single conical mass, flattened in front, and compressed from side to side posteriorly. Indistinct traces of their original individuality may be seen on the sides of the lamin. The pedicles of the arches are all distinct, with interspaces for the passage of the nerves; but, with the exception of the first and seventh, they are scarcely thicker than pieces of cardboard. There is no foramen in the arch of the atlas for the passage of the suboccipital nerve, but merely a shallow groove. The transverse process of the atlas is directed nearly straight outwards, is stout, somewhat flattened from above downwards, and rounded at Fig. 1. Ny Ss Anterior surface of cervical vertebre. Side view of cervical yertebre. 2. vie the extremity ; that of the axis forms a small but distinct tubercle, projecting somewhat backwards from the middle of the transverse process of the atlas; that of the seventh is long and slender, inclining forwards and downwards, its apex being but 0-2 of an inch behind the end of the transverse process of the axis, and extending as far laterally. The transverse processes of the intervening vertebre are but slight triangular projections from the roots of the flattened pedicles of the arches. The side of the body of the seventh vertebra has a well-marked articular surface, which receives the head of the first rib. Slight rough elevations on the bodies of the vertebre in front of this, and at a lower level, are all the indications shown of the inferior transverse processes so commonly met with in this region in the Cetacea. The dimensions of these vertebree are’ :— 1 The cervical yertebrz most closely resemble those of Grampus griseus figured by Van Beneden and Gervais (Ostéographie des Cétacés, pl. 54. figs. 8 & 8a). 6 PROFESSOR FLOWER ON RISSO’S DOLPHIN. inches. Length of inferior surface of conjoined bodies. . . - +» + = 2°4 Length of conjoined arches. . - - - + + ee ect Se eat 27Z Height from inferior surface of atlas to apex of es spinous processes . . : pti, siiehereaet eatior Breadth of aftierle sur faces of ace mau pnStenee Be Pete idles Uae Breadth between tips of transverse processes of atlas. . . . . . 77 Breadth between tips of transverse processes of axis. . - . . . 03 Breadth between tips of transverse processes of seventh vertebra. . 5°2 Greatest breadth of spinal canal (inside arch of atlas) . . . . . 20 Height of spinal canal at the same place. - . . . + +s 1-4 The thoracic vertebre are twelve in number. The body of the first is but 0°6 of an inch in length; they gradually increase from this to the fifth; but the remainder are very nearly equal, viz. 1-7 inch in length, and present no marked differences in breadth and height. Articular surfaces for the heads of the ribs are developed on the hinder edge of the base of the pedicle of the arch of the first five; and on the sixth there is a rough tubercle in the corresponding position. The spine of the first is very small and directed forwards; that of the second is equally low, but broader in the antero-posterior direction; that of the third is long and pointed, and sloping much backwards; the remainder increase gradually in length to the last, and become more upright in position ; beyond the sixth they have a slight forward curve. ‘The transverse processes arise in the anterior vertebre high up on the sides of the arch; but, as in other true Dolphins, their position is gradually lowered until, before the termination of the thoracic region, they are transferred to the bodies of the vertebre. In the first seven vertebre they are of nearly equal length, but from the eighth to the twelfth they gradually increase. They all have articular surfaces at their extremities for the heads of the ribs, at first oval from above downwards, but gradually becoming elongated in the other direction. In the twelfth the surface is convex and very slightly marked. Zygapophyses are deve- loped only as far as the articulation between the sixth and seventh. The metapophysial tubercles are slightly indicated on the third, near the outer end of the anterior edge of the transverse process; on the sixth and seventh they are prominent, conical, and close to the base of the process; on the eighth they become less marked, and begin to rise on the side of the arch; and they no longer exist as distinct processes on the eleventh, and thenceforth are only indicated by a slight bulging forwards of the anterior edge of the upper part of the arch, and do not reappear, as is usually the case, in the lumbar or caudal region’. The bodies of all the lumbar vertebre are very nearly equal in length, viz. 1-4 inch. ' The metapophyses of the posterior thoracic region are much better developed in the skeleton of the “ Grampus rissoanus,” figured by Van Beneden and Gervais (op. cit. pl. 54. fig. 1), than in the present specimen. PROFESSOR FLOWER ON RISSO’S DOLPHIN. U In height they gradually increase from 2 inches (first) to 2°4 inches (nineteenth); in breadth they increase from 2°3 inches (first) to 2°4 inches (nineteenth) at the articular ends. The sixth, seventh, and eighth have the highest spines, viz. 6:4 inches from the upper surface of the body of the vertebra to the tip of the spine—the height of the first being 6 inches, that of the last 4-3 inches. The spines are long, slender, upright, and devoid of metapophyses. ‘The transverse processes gradually diminish from the first (where the breadth of the vertebra between the tips of the processes is 11:2 inches) to the last (where the same measurement is but 7-5 inches); they are very nearly equal in antero-posterior breadth throughout, viz. 0°9 of an inch; and they arise from rather nearer the front than the hinder end of the body ; but this is less marked in the pos- terior than in the anterior portion of the series. I have, as usual, reckoned as the first caudal vertebra that which bears at the hinder end of its body the first chevron bone. ‘The bodies of these increase in length from the first (which is 1-4 inch) to the sixteenth and seventeenth (which are 2 inches), after which they again diminish. In height they do not differ greatly, until beyond the eighteenth, when they rapidly decrease. ‘They begin to diminish in breadth after the eleventh. The lateral compression characteristic of this part of the vertebral column of Cetacea continues until the twentieth vertebra; the twenty-second is the first of the series of broad, depressed, terminal vertebre, the twenty-first being of transitional form. The spinal canal ceases at the nineteenth caudal vertebra. The transverse process is reduced to a low ridge on the fourteenth, and disappears altogether on the fifteenth. The vertical vascular canals first appear in the middle of the base of the transverse process of the fifth, though small, and on the right side only; on the sixth they are present on both sides, and they continue as far as the penultimate vertebra. The terminal vertebra is a small, triangular nodule, very inferior in width to that which precedes it. The chevron bones present are twenty in number, all having the two lateral halves united. It is not improbable that some additional ones from the hinder end of the series may have been lost in macerating the skeleton. ‘The first two are small, with no spines developed beyond the union of the lamine. ‘The third shows a sudden increase in length, which augments in each succeeding one until the seventh, after which they diminish in length, but are more expanded in the antero-posterior direction. There are twelve pairs of vertebral ribs, all of which, except the first three or four, are very slender. The anterior six pairs have long necks, reaching in each case to the articular surface on the side of the vertebra in front of that to which the tubercle is attached. The seventh presents, on both sides, a peculiar arrangement. The rib is not developed inwards beyond the tubercle, which articulates (as do all the posterior ribs) with the end of the transverse process of the corresponding vertebra; but, detached from the rib, and fused with the under surface of the transverse process of the vertebra, is a strong spiculum of bone 1:4 inch long, with its free end pointing forwards, downwards, and inwards, and reaching to within half an inch of the before-mentioned tubercle on 8 PROFESSOR FLOWER ON RISSO’S DOLPHIN. the base of the pedicle of the arch of the sixth vertebra. This obviously represents the neck of the seventh rib. There are eight pairs of sternal ribs, the last being very rudimentary. ‘The first pair articulate near the anterior extremity of the sternum, the second at the junction of the first and second segments of that bone, the third at the junction of the second and third segments, and the fourth and fifth to the hinder end of the third segment. ‘The remainder are not directly connected with the sternum. The various elements of the sternum are consolidated into a single bone, though traces of its original formation out of three segments can be seen, and the primordial median fissure is indicated by a slight longitudinal groove on its inner surface and a small foramen near the anterior part of the first segment or manubrium. ‘The entire length is 11:2 inches. The greatest breadth of the first segment is 5:5 inches; the least breadth, at the middle of the second segment, is 1-7 inch. The manubrium is very slightly notched in the middle line in front; behind the attachment of the first pair of sternal ribs its lateral borders expand as usual into rough triangular processes, directed outwards and backwards. The hinder end of the posterior segment is deeply notched’. The pelvic bones are slender and styliform, 4-9 inches in length. In general form the skull resembles the well-known figure of that of G. griseus in Cuvier’s ‘ Ossemens fossiles, pl. 223. In plate 54 of the great work on the osteology of the Cetacea, now in progress, by Pro- fessors Van Beneden and Gervais, are beautifully executed and evidently most faithfully drawn figures of skulls, named respectively Grampus rissoanus and Grampus griseus, from specimens in the collection at Paris, doubtless the type specimens*. There are certain obvious differences between these two figures, especially in the size of the nasal bones and the width and form of the rostrum; but whether these are more than individual differ- ences it would be hard to say, without a comparison of a large series of specimens. It is to be noticed, however, that in all those points in which the figures differ, the present specimen resembles G. griseus rather than G. rissoanus ; indeed the figure of G. griseus (fig. 7) is so close a representation of it, both as to form and size, that, except for a trifling difference in the shape of the anterior edge of the narial aperture, it might very well have been drawn from it. In the flatness and breadth of the cranial part of the skull, and the wide expansion of the maxille above the orbits, it much resembles that of Globicephalus; but it differs in the rounded form and absence of elevation of the region behind the superior narial apertures, in the marked convexity of the premaxille in front of these apertures, and ‘ This sternum appears narrower, in proportion to its length, than that of Grampus griseus figured by Van Beneden and Gervais (op. cit. pl. 54. fig. 9), but otherwise does not differ materially from it. It closely resembles the sternum of Globicephalus. ? Unfortunately the letterpress of this portion of the work has not yet appeared. PROFESSOR FLOWER ON RISSO’S DOLPHIN. 9 in the comparative narrowness of the rostrum. The general aspect of the upper surface somewhat recalls that of a Beluga. There are no traces of alveoli in the maxilla; but there is a slight depression in the usual situation of the upper teeth, with numerous openings of vascular canals. At the extreme tip of each premaxilla is a conspicuous rounded foramen of the same nature!. The petrotympanic bones closely resemble those of Globicephalus, and are almost of the same size as those of an animal of that genus of nearly double the length of the present specimen. Fig. 3. Periotic and tympanic bones united, outer surface. Fig. 4. Periotic and tympanic bones united, under surface. Fig. 5. Periotic bone, under surface. Fig. 6. Tympanic bone, upper surface. All the natural size. per, periotie ; ty, tympanic; m, posterior end of the periotic, answering to the mastoid of ordinary mammals; a, principal articular surface between periotic and tympanic; ¢, fissure for Eustachian canal at anterior extremity of tympanic; e¢.a.m, external auditory meatus ; ¢.p./, external posterior lobe of tympanic; i.p./, internal posterior lobe of tympanic ; g, groove between these lobes ; /.7, fenestra rotunda in periotic; a,f, aqueduct of Fallopius; s, stapes. The thyro-hyals are ankylosed to the basi-hyal, but with distinct marks remaining of the original suture. The thyro-hyals are more flattened and less tapering at their free extremities than in Glodicephalus; and the basi-hyal has not the same median prolongation in front for the attachment of the anterior cornua. The principal dimensions of the skull are as follows (the measurements of the 1 This is evidently the foramen thus described by Fischer, “En avant des intermaxillaires on voit deux petits trous ot devaient étre enchassées les incisives rudimentaires.” He appears also to have taken the other vascular foramina in both upper and lower jaws for alveoli of rudimentary teeth (loc. cit. p. 367). vou. viil.—PartT I. March, 1872. c 10 PROFESSOR FLOWER ON RISSO’S DOLPHIN. skull of the animal next to be described are also appended for convenience of com- parison) :— Skull of adult. Skull of young. inches. inches. Mntireviength. oe) 0s. yi) 7s ok ea ee 136 Icenpthiot rostrum! 7s. os ieee ee OS 6-1 Breadth of occipital foramen =) = = = ose = so 1:2 Greatest height of occipital foramen . . . . . . . L9 ils7 Breadth or occipital condyles) manne 4:5 3°2 Greatest breadth of cranium (at pa region, in temporal fossa). . : galt Uh Greatest breadtit a: areal (at Ppotiake: process of squamosals). . . . . Ai He ene lL AY) 9-0 Breadth at anteorbital processes of feontals Sa Poy ee eles 8:2 Breadth of anterior narial apertures. . . 2°8 2°2 Breadth of rostrum at base (bottom of anieorital cotehy (tl 5°6 Breadth of rostrum at the middle . ...... . 44 3° Itength of tympanic bone = = = © - = = = «= = 1:8 1:6 Mandible—Length oframus. . ...... . is, 4 10°6 LGR Oris 6 oo 6 o 9 2 6 J) 1:2 lomanclin eGo . 5 5 6 = 5 6 o o Ill Height at coronoid process . . . .. . 43 2:7 Hyoid.— Breadth between tips of thyro-hyals. . . . . 87 Basi-hyal— Greatest antero-posterior length . . . . . 3 The bones of the pectoral limb generally present a nearer approach to those of Glodi- cephalus than any other Cetacean, or, rather, may be described as intermediate between that genus and Delphinus proper. Of the two scapule figured by Van Beneden and Gervais, one of G. rissoanus, and the other of G. griseus, the present one most nearly resembles the latter in outline, especially in the form of the acromion ; it is rather smaller, however, in all its dimen- sions, in which respect it is more like the former. The humerus is immovably united with the radius and ulna. These bones are not so broad in proportion to their length as in Globicephalus. The carpal bones are five in number, and form a close mosaic, three in the first and two in the second row, and have precisely the same arrangement as in Globicephalus. The pollex consists of a short, nearly square metacarpal, and a single, conical, tapering phalanx, reaching nearly to the end of the second metacarpal. It is certain that no other ossified phalanx was present in this digit on either hand—a circumstance which I note particularly, because in Gervais’s figures’, both of G. griseus and G. rissoanus, there * Measured from a line drawn between the anteorbital notches. ? Op. cit. pl. 54. figs. 10 & 5. PROFESSOR FLOWER ON RISSO’S DOLPHIN. Wi is a small additional phalanx to the pollex, and one is also present in all the specimens of Globicephalus which I have examined. As the bones of the digits have never been separated, there can be no doubt as to their correct number and position. The second digit is the largest, and has ten distinct ossifications including the metacarpal, the last being a rounded nodule rather smaller thana pea. ‘The third has eight ossifications; its metacarpal is considerably longer than that of the second digit; and its proximal phalanges are rather broader, though more flattened and more compressed, especially at their posterior or ulnar edge, than are those of the second digit. The relative condition of the bones of these two digits thus agrees rather with Gervais’s figure of the limb of G. griseus (fig. 11) than of G. rissoanus (fig. 6). The fourth digit is very short, and has but three ossifications ; and the fifth is rudimentary, being mostly cartilaginous, with a nodular metacarpal bone at the base. The minute bone-specks represented in the terminal portion of the cartilages of these last two digits in Gervais’s figure of G. griseus are not present. The principal dimensions of the bones of the pectoral limb are as follows :— inches. Scapulars bleioiin ges aan pee nce Vo au OU) Breadth wre abi bonus checked notes CIRO Ibe CEAOHOMON 6 5 6 & @ a s 6 6 2H) Greatest depth of acromion. . . . . . . 19 Length of coracoid process. . .... . 20 Length from head of humerus to tip of second finger’. . 22:5 mensthyotahumentsieme) at) vee cer) es ee ace EC: Wenge thvotmadiustna’. st Vagecd Sekt) te) ei ee os2: ienethrotgulnageeya ues nate eee) sa sc ee amen ASO Breadth of radius at distalextremity . . . . . . . 24 Breadth of ulna at distalextremity. . . . .... 19 About a month after the capture of this Dolphin (viz. March 31st) Mr. Gerrard gave me an opportunity of examining another specimen which he had bought at Billingsgate Market; but, as it had changed hands several times, he was unfortunately not able to obtain any trustworthy account of the place of its capture, though this was probably somewhere in the Channel. It was quite fresh at the time of my seeing it. This was also a female, but, as the condition of the bones afterwards showed, a very young animal. Although very different from the former in the disposition of the surface colours, the other characters, especially those of the skeleton and dentition, are so closely similar that I have little doubt of its specific identity; and such being the case, it is within the bounds of probability that this might have been the identical young animal which ? Owing to the drying and contraction of the cartilages, this must be somewhat less than in life. 2 a 12 PROFESSOR FLOWER ON RISSO’S DOLPHIN. the former must have had near her at the time of her capture; or, as small herds of these Cetaceans generally travel in company, it might have been a member of the same band. There is no evidence, however, in the present case of more than these two individuals having been seen’. In general form the young animal closely resembles the old one; but the head is more rounded, the dorsal fin is not quite so high, and slightly more posterior in position, and the pectoral fin is decidedly shorter in proportion to the general size of the body. These differences will be best appreciated by the following table of proportions, the entire length of the animal being in each case taken as 100 :— Young. Adult. eng thkofpectoralptii= alae mee nel OL _ 18°8 Breadth) oficaudal tm) 25s 20:0 23:0 Heicht of dorsalein es aan lOe3 12:7 I think that the difference would be even greater, if in the entire length the caudal portion of the body could be excluded; for that this grows more in proportion than the head and trunk, seems to be shown by the relatively more advanced position of the dorsal fin in the adult. It is but natural to suppose that the locomotive appendages should be more highly developed in the full-grown than the new-born animal; and there is certainly a similar alteration with age (at least as regards the pectoral fin) in the allied genus Globicephalus. I am indebted to Mr. Gerrard, jun., for the following dimensions, as well as for the drawing of the animal (PI. I. fig. 4) :— P t. in. Length, in straight line from the upper lip to notch in middle of caudal fin . From the upper lip to anterior rede of dona fin (ellownite) ca) From the upper lip to anterior angle of eye From posterior edge of dorsal fin to middle of tail From posterior edge of dorsal fin to angle of mouth . Length of eye-aperture : Pectoral fin.—Length of anterior Hot der Length of posterior border Height of dorsal fin Antero-posterior length of dorsal fin at Base Width of caudal fin Girth of body immediately in front of acral ffn bole nH e ee eH ont NAN Oo (Sy Ms) | fs nile oo Heo ~) tole Bey (OS) SOS OS SSS) SY eS) Blo (o’9) The upper parts and sides of the body were almost black, the lower parts nearly ‘ It is stated by the late Mr. Jonathan Couch, in ‘ Land and Water,’ March 19th, 1870, that an animal of the same species was observed off the coast of Cornwall in the month of May 1869. PROFESSOR FLOWER ON RISSO’S DOLPHIN. 13 white, the junction between the two colours being very abrupt, passing from the angle of the mouth above the origin of the pectoral fin backwards to the vulva, behind which the whole of the surface was black. On the sides were several large longitudinally disposed nebulous washes of light grey, the most conspicuous of which was between the pectoral and dorsal fins ; and the lateral surfaces of the last-named fin, especially near its base, were yellowish white. The upper lip, the front surface of the head, as high as the top of the vertical anterior wall of the adipose prominence, and the chin and throat were white, with a distinct yellow tinge. The pectoral fin was black on both sides, and showed none of that peculiar delicate mottling or dappling so conspicuous in the larger specimen. Both surfaces of the caudal fin were black. So much for the general ground-colour, upon which were laid certain markings of a somewhat similar character, though very different in number and in arrangement, to those of the adult animal. On each side of the body were six vertical whitish stripes, nearly symmetrically arranged, and almost equidistant, being about six inches apart. They did not extend quite to the middle line of the body above, and were lost below in the light colouring of the abdomen. ‘These stripes were not so complex in structure as those of the adult specimen, being merely broad white lines, shading insensibly at the edges into the general dark colour of the surrounding cuticle, and with an obscure dark central line. In addition to these, however, there were three others, which, though short and straight, had exactly the same characteristic formation as the best-marked streaks of the other specimen, having a white central line, bordered first with black and then again with white. These were situated :—one on each side of the narrow part of the tail, close to the upper median ridge, just before the commencement of the lateral expansions of the caudal fin; and the third obliquely longitudinal, a little to the left of the middle line of the back, just in front of the dorsal fin. This last, especially, was extremely impor- tant, as, being asymmetrical, it showed the tendency to variation in colouring and surface-markings, and so helps to account for the great difference in these respects between the two specimens. On each side of the upper lip were eight extremely short, whitish bristles, their tips only just projecting beyond the level of the cuticle, arranged in two rows, six in the lower and two in the upper row, as shown by the dots in the figure. There were no similar bristles on the under lip or chin. The condition of the teeth was of extreme importance, as helping to establish the normal dentition of the species. There were no teeth visible above the gum in either jaw; and on a careful examination of the soft tissues between the surface and the bone, not a rudiment of any tooth could be found in the upper jaw, while in the lower jaw were the germs of seven teeth—four on the right and three on the left side, close to the symphysis. The crowns of these teeth were conical, pointed, and strongly curved. Calcification had extended rather below the base of the crown or neck of the tooth. The skeleton presents all the signs of immaturity which might be expected in an 14 PROFESSOR FLOWER ON RISSO’S DOLPHIN. animal of so early an age; but in other respects it does not differ materally from that of the other specimen. The total number of vertebre is 69, being divided into 7 cervical, 12 dorsal, 20 lumbar, and 30 caudal; this gives one additional vertebra in the lumbar region, an individual variation by no means uncommon in the Cetacea. The bodies of the first and second cervical vertebre are united; but those of all the others consist of very thin plates still separable from each other, or in some cases, as between the fourth and fifth, united by a partial ankylosis in the centre of the disk. Except in the atlas and axis, the arches have not yet united with the bodies; but the spines are joined into two sets—one consisting of the first, second, and third, and the other consisting of the last four. The metapophyses of the posterior dorsal and anterior lumbar vertebre are more developed than in the adult specimen, resembling those in Van Beneden and Gervais’s figure of the skeleton of G. rissoanus. The chevron bones are twenty-one in number. There are twelve pairs of vertebral ribs, of which the first six have necks; in the seventh the neck is represented by an unossified ligament. The sternal ribs resemble those of the adult in number and connexions. The three segments of the sternum are not united by bone; the last is divided in the middle line into two separate pieces. The skull, as is usual with young animals, differs from that of the adult in the large size, globular form, and smooth outlines of the cerebral portion as compared with the rostrum. It much resembles that of a young Glodicephalus of corresponding age, but can be at once distinguished by the convexity of the upper surface in front of the narial aperture. The under surface of the anterior portion of the maxilla has a well-marked, but narrow, longitudinal groove near the outer border, corresponding to the alveolar line of other Dolphins; but there are no distinct alveoli. The opening of the vascular canal at the apex of the premaxilla is very distinct. In the anterior portion of the upper edge of the mandible is a deep narrow groove, 44 inches long, the anterior portion of which is dilated into a wide alveolar chamber (1°8 inch long on the right side and 1°5 inch on the left), divided by very imperfect septa into chambers for lodging the teeth, four on the right and three on the left side. The principal dimensions of the skull are given at p. 10. The carpal bones and the phalanges, though of course far less developed, are the same in number as in the adult, except that the minute terminal ossifications of the second and third fingers of the latter were not yet apparent’. There can be no doubt that these two animals belong to a peculiar group of the Cetacea constituting the genus Grampus of Gray, of which two species are commonly ? Some notes upon the visceral anatomy and external characters of this individual have been given by Dr. Murie in the ‘Journal of Anatomy and Physiology,’ Nov. 1870, p. 118. The differences in Dr. Murie’s description of the external surface from that given above, are probably due to changes resulting from the greater length of time that had elapsed between the death of the animal and his examination of it. PROFESSOR FLOWER ON RISSO’S DOLPHIN. 15 recognized as inhabiting the northern hemisphere (one having hitherto been met with only in the Atlantic, and the other only in the Mediterranean), and have been named respectively G. griseus (= G. cuviert, Gray) and G. rissoanus'. The earliest account of both of these animals is contained in the ‘ Rapport fait a la classe des sciences mathématiques et physiques sur divers Cétacés pris sur les cétes de France,’ by G. Cuvier (Annales du Muséum, tome xix. 1812). The first is described from a drawing of the external characters, accompanied by the skeleton of the animal, sent to the Paris Museum from Brest. It is stated to have been 34 metres in length, to have had but four teeth in the lower jaw, “toutes trés-usées et prétes 4 tomber,” and to be of a “ grisétre” colour, whence Cuvier bestowed upon it the name of Delphinus griseus. 'The skeleton (in a very imperfect condition) is still in the Museum at the Jardin des Plantes. The drawing, reproduced at pl. 1 of the volume of the ‘ Annales,’ obviously gives but a mere rude approximation to the true external form of the animal; but it must be remarked that there are numerous irregular scratch-like black lines on the face, around the eyes, on the dorsal and pectoral fins, and especially on the tail, which can scarcely have been put in by the artist without some foundation in nature, although they are not alluded to in the exceedingly brief description. To continue the history of the various specimens which have been considered to belong to this species :— In the middle of June 1822, four Dolphins were stranded near |’ Aiguillon (la Vendée), and were described in some detail by D’Orbigny. F. Cuvier (Histoire Naturelle des Cétacés, 1836, p. 185), citing this account, speaks of them under the denomination of “Le Marsouin de d’Orbigny, Phocena griseus,’ and recognizes their specific identity with the Brest specimen. The length of the full-grown individuals of this band is stated to have been ten feet (Fr.). ‘La teinte générale du dessus du corps et de la téte est d’un noir bleuatre; le dessous est d’un blanc sale, qui se fond sur les cétes avec le noir.” A figure of one of them is added to the description; but it is probably not to be depended upon as giving a correct outline of the animal, owing to the decomposed state of the bodies at the time they were examined. ‘There were no traces of teeth in the upper jaw of either. A young specimen is said to have had eight teeth in the lower jaw, and the older ones from six to seven. The statement that “la machoire supérieure est plus longue et s’avance de quatre pouces au-dela de celle d’en bas,” does not accord with the descriptions of other observers; but D’Orbigny’s observations were made under difficulties. In consequence of D’Orbigny’s statement as to the colour, Dr. Gray changed the specific name from griseus to cuviert (Ann. N. H. 1846). 1 A third species (G. richardsonii, Gray) is founded on a lower jaw of unknown locality, and a skull from the Cape of Good Hope, which differs slightly from those above described. (Cat. Seals and Whales, Brit. Mus. p. 299, 1866.) 16 PROFESSOR FLOWER ON RISSO’S DOLPHIN. On April 12th, 1844, a Dolphin attributed to this species was stranded near Cageaux (Gironde)'. A skull is contained in the British-Museum collection, from the Isle of Wight, pre- sented in 1845 by the Rev. C. Bury. On July 22nd, 1867, a Dolphin was cast up by the sea on the shore of the Depart- ment of la Gironde, France, and taken to Arcachon, where, fortunately, it was examined by M. P. Fischer, who has given’ an excellent, and evidently trustworthy, description of its external characters and skeleton, accompanied by a succinct history of the species, and of its relation to the so-called Risso’s Dolphin, to which I shall afterwards have occasion to refer. The specimen was young, measuring but 2°80 metres (9 ft. 24in.). His description of the colour is as follows :—‘“ Le corps est de couleur noire sur le dos et les flancs, blanche en dessous autour des parties génitales et de anus, d'un blanc teint de gris de fer en avant de la verge, blanche enfin au niveau et en avant de la base des nageoires pec- torales. Le dessous de la téte et du cou est d’un gris noiratre, marbré de taches blanchatres, terminé en pointe noire dirigée vers le thorax; le dessus de la téte, le bord des lévres, sont également marbrés de blanc sale. Les nageoires pectorales, caudale, et Yaileron dorsal, ont une coloration noire uniforme.” The dental formula was = There were sixty-eight vertebree—seven cervical, twelve dorsal, forty-nine lumbar and caudal. The second Dolphin alluded to above as having been brought into notice in Cuvier’s ‘Rapport’ was known to him only by a notice and figure communicated to the Academy by M. Risso of Nice, the figure being reproduced in the same plate as that of the Grey Dolphin from Brest, and marked “ Delphinus aries?” It was stated to be 3 metres long, and to have five teeth on each side of the lower jaw only. The figure shows numerous white lines on the surface, mostly in the longitudinal direction. Delphinus (Phocena) rissoanus, of Desmarest’s Mammalogie (part 2, 1822), is founded on this description and figure. Subsequently M. Risso published in his ‘ Histoire Naturelle de ! Europe méridionale,’ 1826, tome iii. p. 23, a fuller description, under the name of ‘‘Delphinus risso, Cuv.,” and a different figure (pl. 1. fig. 2), the accuracy of which may be estimated by that of the wretched caricature of the Globicephalus in the same plate. The description runs thus :— “ D. dorso lato ; capite maximo, obtuso; maxilla superiore longiore. “Des meeurs douces, comme la zone tempérée quil habite, semblent étre le partage de ce cétacé, qui n’approche de nos cétes que dans le temps des amours. Son corps est alongé, arrondi, renflé vers sa partie antérieure, diminuant insensiblement de grosseur ‘ Laporte: Actes de la Société Linnéenne de Bordeaux, 1853, t. xix. p. 215 (fide Fischer). * “Note sur un Cétacé (Grampus griseus) échoué sur les cotes de France,” Annales des Sciences naturelles, Sth ser. vol. vill. p, 363. PROFESSOR FLOWER ON RISSO’S DOLPHIN. 17 vers la queue, qui est déprimée; sa peau est mince, unie, de couleur grise, 4 nuances bleuatres, traversée par des traits irréguliers et des raies inégales, droites ou flexueuses, blanchatres; le ventre est d'un blanc mat; la téte fort grande; le museau arrondi, relevé en arc, obtus, percé vers la nuque par Vouverture des évents; la bouche est ample, arquée; la machoire supérieure pourvue d’alvéole seulement, est plus avancée, et couvre l’inférieure, qui est garnie de chaque cété de cing grosses dents coniques; aigués, un peu courbées, distantes, fortement enchassés dans l’ossement de la machoire ; ces dents sont solides, presque égales, d’un blanc jaunatre, recouvertes d’un émail fort luisant; l’intérieur de la gueule est muni de tubercules émoussés; la langue est libre, unie sur ses deux bords; les yeux sont ovales oblongs, trés-petits, avec Viris doré; la nageoire dorsale, haute, élevée, 4 peu prés en forme de triangle scaléne, est située presqu’au milieu du dos; les nageoires paires sont grandes, épaisses, noiratres; la caudale est forte, divisée en deux grands lobes par une échancrure assez profonde. Long. 5 métres, larg. 1 metre. Séj. Surface des eaux. App. Printemps, automne.” Another and somewhat better figure is given by Laurillard in Fr. Cuvier’s ‘ Histoire Naturelle des Cétacés’ (1836), taken from one stranded, with many others, in the Bay of Saint-Jean, near Nice, in June 1829. The length of these specimens is stated to have been about nine feet (French); and their peculiar colours are thus described by Laurillard :— La couleur de ces dauphins différait suivant les sexes. Celle qui faisait le fond de la peau des femelles était un brun uniforme; les males, au contraire, étaient généralement d'un blanc bleuatre; mais ce qui charactérisait les uns et les autres, c’étaient les singuliéres lignes semées irréguli¢rement sur toutes les parties supérieures du corps, et qui ressemblaient, au premier coup d’ceil, 4 des égratignures produites par des épines. Ces lignes, vues de prés, se composaient de traces plus claires que le fond de la peau, et bordées d’une multitude de petites lignes perpendiculaires d’un brun foncé. De plus, les males, comme le montre la figure que nous donnons, avaient des taches irréguliéres d’un brun foncé sous la moitié postérieure du corps, et les nageoires avaient la méme couleur; mais la dorsale et la pectorale se trouvaient de plus ornées de lignes blanches. Deux lignes brunes garnissent le dessus et le dessous de la bouche, et un cercle de méme couleur entoure l’eil.” The name applied by F. Cuvier to these animals is “ Le Marsouin Risso (Phocena risonnus).” A complete skeleton and a skull of animals from this shoal are preserved in the Paris Museum; the former is that above referred to as figured by Van Beneden and Gervais. The number of teeth is = Some time about 1854, a herd of these Dolphins came into Carry (Bouches-du-Rhone); and the skull of one of them which was killed is preserved in the Museum at Marseilles’. It thus appears that the so-called Delphinus griseus has been met with on five distinct occasions, the localities having been Brest, Aiguillon, Cajeaux, and Arcachon, on the 1 P. Gervais, “Cétacés des cotes Francaises de la Méditerranée,” Comptes Rendus, tome lix. 28th Noy. 1864. VOL. Vill.— Part I. March, 1872. D 18 PROFESSOR FLOWER ON RISSO’S DOLPHIN. west coast of France, and the Isle of Wight; and the so-called Delphinus rissoanus three times, always at Nice or the Mediterranean coast of France, and, as Fischer has par- ticularly pointed out, whenever the date has been recorded, the occurrence has in both cases always taken place in the spring or summer (April to July). From this circum- stance he concludes that this species (for after a comparison of the osteological and other characters he has come to the conclusion that they are one and the same) is migra- tory, visiting the shores of Europe in the summer and passing the winter either to the south towards the coast of Africa or to the west towards the American continent. Although the present examples (occurring in the mouth of the channel so early as the end of February) may be thought somewhat to shake this conclusion, it may on the other hand prove to be merely a case of an unusually early arrival in our seas. Further observations can alone determine the question". Identity of G. griseus and G. rissoanus. It has been mentioned that Fischer came to the conclusion that the two species (D. griseus and D. rissoanus) ought to be reunited, as had also appeared probable to Cuvier’—a conclusion founded on the following considerations :— **1°. Que le Dauphin de Risso apparait dans la Méditerranée 4 la méme époque que le griseus sur les cOtes océaniques de France. “2°. Que sa dentition ne différe pas sensiblement de celle du griseus, puisqu’on peut établir la série suivante :— = ae = (griseus) ; = = (rissoanus). 3°. Que le nombre des vertebres, des cétes, des phalanges, en un mot que tous les caracteres ostéologiques sont identiques dans les deux espéces. “4°, Que les seules différences relevées entre elles portent sur la coloration extérieure éminemment variable, et sur la forme plus ou moins ventrue du corps, qui peut tenir a Yembonpoint des individus ou a la distension de labdomen par des gaz aprés la mort.” Gervais had previously expressed his opinion*® that “Le D. griseus, qu’on appelle quelquefois Marsouin de d Orbigny, est trés-peu différent du D. rissoanus; son systéme dentaire parait néanmoins devoir l’en faire séparer;” and he gives the following diagnostic characters :— ** Dents supérieures caduques, les inférieures au nombre de cing ou six paires, Del- phinus rissoanus. * It is possible that the Grampus, before mentioned, from the Cape of Good Hope, of which there is a skull in the British Museum, named G. richardsonii, may also prove to be of this species; if so, it would indicate that the South Atlantic may be its winter habitat. In size it perfectly agrees; but it is rather narrower in proportion to its length, and the anteorbital processes of the maxilla are more upturned at their edges, and less laterally and anteriorly expanded, and the anteorbital notch is less deep than in the other specimens. The teeth are = * Ossemens fossiles, edit. 1836, tome viii. 2m¢ partie, p. 98. * Zoologie et Paléontologie Francaises, p. 301. PROFESSOR FLOWER ON RISSO’S DOLPHIN. Ty) “Dents caduques; deux paires 4 la partie terminale de la miachoire inférieure seulement, D. griseus.” It now remains to be seen whether Fischer’s opinion is strengthened or the reverse by the new materials afforded by the two specimens described in the first part of this communication. As has just been shown, the differences hitherto noticed between the supposed species have resolved themselves into those of habitat, colour, and number of teeth. 1. According to the previously observed habitats of the two species or varieties, the present specimens should be referred to D. griseus, as all the examples of D. rissoanus hitherto met with have been from the Mediterranean. 2. According to the coloration, they should be D. rissoanus: but additional light is thrown upon this part of the subject by these two specimens; for they show conclusively how extremely variable the species is in this respect. This might be inferred from the absence of bilateral symmetry in the markings of each individual, shown most strongly in the adult example, where the markings were more profuse and complicated in cha- racter. lLaurillard speaks of the ground-colour of the females being of a “ uniform brown,” and therefore quite different from that of the two specimens now described. Then, again, with reference to D. griseus, it must be remembered that the three descrip- tions all differ somewhat as to the colouring, and, especially, that in the figure of the type specimen from Brest numerous distinct irregular linear markings are indicated (as previously mentioned). ‘These with the general “grisdtre” colour would appear to show that it was more nearly allied in external characters to Risso’s Mediterranean Dolphin than to the black-and-white specimens described by D’Orbigny. If all the specimens of alleged D. griseus had been uniform in colour, and all those of D. rissoanus had presented another characteristic coloration, there would have been more grounds for keeping them distinct; but from the facts before us it is safer to conclude that we have here an example, very rare among Mammals, of a species of variable and irregular coloration. As to the teeth, the new specimens completely break down the specific distinction previously drawn; for, with the colouring of D. rissoanus, the adult one has the number of teeth assigned to D. griseus, viz. ; and that this is not the result of loss by age is = satisfactorily proved by the young individual, in which the teeth had not even cut the gum. This last-named specimen was also extremely important, as presenting an undoubted case of original unequal number on the two sides, viz. three and four, showing that the exact number of teeth is a variable character ; and it also set at rest the question as to whether the absence of teeth in the upper jaw is a congenital peculiarity, or arises, as had generally been supposed, from their loss at an early age. It appears then necessary, until any better diagnostic characters are made out, to sink the specific name of rissoanus in that of griseus, though it may be convenient to apply the term “Risso’s Dolphin” to the peculiarly marked variety which was first made 20 PROFESSOR FLOWER ON RISSO’S DOLPHIN. known to science by that naturalist, or even to keep it as a vernacular appellation for the entire species, and thereby continue to associate his name with it. Systematic Position—In Baron Cuvier’s time this species, with a vast number of others now generically separated, was included in the genus Delphinus. Mons. F, Cuvier constituted of a group of short-nosed Dolphins the genus Phocena, embracing with the species now under consideration the Common Porpoise, the Killer (Orca gladiator), the Round-Headed Dolphin (Glodicephalus melas’), and even the Beluga (Histoire des Cétacés, 1836). The genus Globicephalus* was formed by Lesson’ for a still more restricted group, from which Dr. Gray has separated the present animal and its most immediate allies under the name of Grampus*. My first impression, obtained from an inspection of the external appearance of the animal, was that it resembled Globicephalus melas so nearly as hardly to warrant generic separation. There was the same rounded form of forehead, and the same elongated pointed pectoral fins placed low down on the sides of the thorax. It is true that these were developed in a less exaggerated degree than in Globicephalus, and the dorsal fin was more anterior in position and more elevated; but such characters can hardly be considered generic, unless accompanied by other and more important structural differences. The teeth, again, have much the same size and form as those of Globi- cephalus, and are also confined to the anterior part of the jaws; and I was not then aware that the absence of maxillary teeth was congenital, but supposed that it was due to their being deciduous, a circumstance frequently observed in Globicephalus at a late period of life. Consequently in the preliminary notice of the capture of this specimen® the generic name of Globicephalus was assigned to it. After a closer examination of the characters, especially of the skeleton and teeth, of both forms, I am now inclined to think that they may fairly be treated as distinct although closely allied genera, and subjoin the following comparative table of diagnostic characters :— GLOBICEPHALUS.—Teeth in both jaws, 9 to 12 on each side, confined to the anterior half of the rostrum and corresponding portion of the mandible; sometimes deciduous in old age. Vertebre. C. 7, D. 11, L. 12 to 14, C, 27 to 29, total 58 or 59. 1 This species was first correctly described and figured by Trail, under the name of Delphinus melas (Nicholson’s Journ. xxii. 1809, p. 21). Cuvier, unacquainted with Trail’s memoir, described and figured it again in the “ Rapport sur divers Cétacés,” frequently referred to above, as Delphinus globiceps (Annales du Muséum, t, xix. 1812). Lacépéde’s Catodon svineval (Hist. Nat. des Cétacés, 1804, p. 216) may have been founded on some vague idea of this animal; but the description is almost altogether inaccurate. * Often spelt Globiocephalus. 3 Compl. de Buffon, i. 1828 (fide Agassiz, Nom. Zool.). * Zoology of ‘ Erebus’ and ‘ Terror,’ p. 30 (1846). Grampus, Gray, Spice. Zool. 2 (1828), was a heterogeneous group, comprising Grampus griseus, Globicephalus melas, Delphinus acutus, D, heavisidii, and D, obscurus. ® Pp. Z. 8. 1870, p. 128. PROFESSOR FLOWER ON RISSO’S DOLPHIN. 21 Skull. Rostrum and cranium proper of nearly equal length. Upper surface of rostrum very broad and flat’, rounded in front. The premaxille at the middle of the rostrum as wide as, or wider than, at the base, and very nearly or completely concealing the maxille in the anterior half of the rostrum. Upper surface of the premaxille in front of the narial apertures strongly concave. Upper surface of the skull behind the narial apertures raised into a strong prominence, of which the nasal bones form the apex. Manus very long and pointed, the number of phalanges (including the metacarpals) in the different digits being respectively I. 3-4, I. 13-14, ITI. 10, IV. 3, V. 1. Grampus.—TZeeth, none in the upper jaw. In the mandible 3 to 7 on each side, confined to the anterior part of the ramus close to the symphysis. Vertebre. C. 7, D. 12, L. 19, C. 30, total 68. Skull. Rostrum slightly shorter than cranium proper, its upper surface moderately broad and flat, slightly expanding laterally in front of the notch, then tapering gradually to the apex. The premaxille not so wide at the middle of the rostrum as at the base. Upper surface of the premaxille in front of the narial apertures convex. Upper surface of the skull behind the nasal apertures rounded. Manus \ong and pointed, the number of phalanges (including metacarpals) in the different digits being I. 2, II. 10, III. 8, IV. 3, V. 1. DESCRIPTION OF THE PLATES. PLATE I. Fig. 1. Adult female Risso’s Dolphin, caught near the Eddystone Lighthouse, February 28th, 1870. Fig. 2. Upper surface of the head of the same. Fig. 3. Upper view of the tail of the same. From drawings by Mr. R. W. Sherwin. Fig. 4. Very young female Risso’s Dolphin, bought in Billingsgate Market, March 30th, 1870. From a drawing by Mr. E. Gerrard, jun. All 5's the natural size. PLATE II. Fig. 1. Side view of the skeleton of the adult Risso’s Dolphin. Fig. 2. Upper view of the same skeleton. Fig. 5. The sternum of the same, from below. Fig. 4. Skull of the young animal. All } the natural size. 1 Tn the skulls of very young animals these special characteristics of the genus are but slightly developed, the rostrum being much more narrow in proportion to its length than in adults. vou. vill.—ParT I. March, 1872. E ‘+, Vea T ne atieny werd : 9Gaae00- }99999980- »] ® © a * a) rai aie) x g Trraspiza, Kaup. 13. TeRAsPIzA RHODOGASTRA (Schlegel), Valkv. Nederl. Ind. pp. 21 & 60, pl. 12. figs. 5 & 6. Nisus virgatus rhodogaster, Schlegel, Mus. Pays-Bas, Astures, p. 32. Hab. Celebes (Wallace); Menado (mus. nostr.) ; Gounong-Pello, district of Gorontalo (Forsten). EryTHrospiza, Kaup. 14. Eryrurospiza TRINOTATA (Bp.), Consp. Av. i. p. 33, “ Celebes” (1850). Nisus trinotatus, Schl. Mus. Pays-Bas, Astures, p. 45; Valkv. Nederl. Ind. pp. 27 & 65, pl. 19. figs. 1, 2, 3. Hab. Macassar, Menado (Wallace); Menado (mus. nostr.); Gorontalo (Forsten). 1 Tt is a debatable question whether the generic title Cerchnets, Boie (1826), the type of which is 7. rupicolus, Daud., should not be employed rather than that of Zinnwnculus, Vieillot (1807), the type of which is F. columbarius, Linn. VOL. VIII.—PaRT U. May, 1872. H 34 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Eryrurospiza rocastra (S. Miiller), Verhandl. Land- en Volkenk. p. 110, “Amboyna” (1839 ?). Accipiter hyogaster, S. Miller, Jacquin. & Pucheran, Voy. Péle Sud. iii. p. 48, “ Macassar.” Epervier océanien, Hombron & Jacquin. op. cit. Atlas, pl. 2. fig. 1. Jacquinot is our only authority for the occurrence of this species, as identified by Pucheran, in Celebes ; but as his notes relating to localities are not always trustworthy, and as A. dogaster is not given from Celebes by either S. Miiller, Professor Schlegel, or Mr. Wallace, I shall not include it in this list. In this instance Jacquinot’s authority is doubly untrustworthy; for he identified A. rufitorques, Peale, with A. togaster, and noted the two examples, which he figured (/. ¢.) as having been obtained in the Viti Islands and at Macassar. It is quite possible that A. iogaster does occur in Celebes; and it is difficult to determine the A. eruentus, Gould, ap. Schlegel, M. P.-Bas, Astures, p. 42, male adulte, Célébes, Voy. de Reinwardt, “gorge d’un roussditre uniforme,” unless we refer it to A. togaster. And yet, under Misus cruentus (Valky. Nederl. Ind. p. 61), Professor Schlegel does not allude to this specimen, nor does he give Celebes as a locality for NV. cruentus. Tacuyspiza, Kaup. 16. Tacnyspiza soLoinsis (Horsf.), Trans. Linn. Soc. xiii. p. 137, “Java” (1822); Mus. Pays-Bas, Astures, p. 44. Falco cuculoides, Temm. PI. Col. livr. xxii. pls. 110, 129, “ Java” (August 2, 1828). Hab. New Guinea, Batchian, Sumatra, Malacca ( Wallace); Java, Philippines (Schlegel) ; Menado (mus. nostr.) ; North Celebes (Forsten). A QUILINA. Neopus, Hodgson. 16. NEopuS MALAYENSIS (Reinw.), Pl. Col. 117 (26th June, 1824), “ Java, Sumatra.” Hab. Java, Sumatra (Temminck); most of the hilly and jungly districts of India (Jerdon); Simla (mus. nostr.); Nepaul (Hodgs.); Malacca (mus. nostr.); Celebes (Bernstein) ; Ceylon (Layard). Limyaktus, Vigors. 17. Limnaitus Lanceotatus, Bp. Consp. Av. i. p. 29, “ Celebes” (1850). Spizaétus cirratus, Schlegel, Mus. Pays-Bas, Astures, p. 9, ex Celebes. Hab. Celebes (Wall.); Menado (mus. nostr.); Tondano (Forsten); Sula Islands (Wallace). The example in my collection has been identified by Mr. J. H. Gurney as a young individual of the above species. Underneath pure white; thigh-coverts faintly fringed with pale fulvous; entire head and nape pale fulvous-white; remaining upper plumage hair-brown, darkest in shade at the end of each feather, unexposed portion of each VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 35 feather being pure white; minor and major under wing-coverts pure white, a few possessing a terminal light-brown spot or drop; axillaries immaculate white; major wing-coverts brown on outer, white, barred with brown, on the inner web; no trace of an occipital crest; bill exceedingly powerful, height from festoon to culmen being full five eighths of an inch. This bird closely resembles a Cingalese example of a young S. cirrhatus, Gm. (=S. cristatellus, Temm.) in my collection. The only points of dif- ference in the Cingalese individual being, besides its smaller dimensions, a black occipital crest three inches and a half long, the major wing-coverts being mostly white, and the axillaries and thigh-coverts being white, largely dashed, freckled, and barred with a clear tint of pale rufous. In the Celebean bird the tarsal feathers incline to cover the insertion of the toes. Dimensions:—wing 16 inches, tail 11%, tarsus 33, mid toe with claw 23, bill from gape 2. Poxroaiitus, Kaup. 18. Poxioaiitus Humiiis (Miill. et Schlegel), Verh. Ned. overz. Bez. Aves, p. 47, pl. 6, “ Sumatra;” Wallace, Ibis, 1868, p. 14; Schlegel, Mus. Pays-Bas, Aguile, p- 18. Hab. Celebes (Wallace); Sumatra (Miller); Malay peninsula (Blyth); Bengal (Schlegel). Cuncuma, Hodgson. 19. CuncuMa LEUcoGASTER (Gm.), Syst. Nat. i. p. 257 (1788); Valkv. Neder. Ind. pl. 4. figs. 1, 2; Wall. Ibis, 1868, p.15; Mus. Pays-Bas, Aquila, p. 14. Hab. Malacca, Celebes, Gilolo, Batchian, Morty, Aru Islands (Wallace); Macassar (mus. nostr); all over India, chiefly on the coast (Jerdon); Australia, Tasmania (Gould); Timor, Sumatra, Java, Ternate (Schlegel); Ceylon (Layard). Sprtornis, G. R. Gray. 20. SPILORNIS RUFIPECTUS, Gould, P. Z. S. 1857, p. 222, “‘ Macassar ;” Schlegel, Valky. Nederl. Ind. p. 72, pl. 23; Mus. Pays-Bas, Buteones, p. 27. Hab. Menado (mus. nostr.); Celebes (Wallace); Tondano, Gorontalo (Forsten). MILVIN&. Hauiastur, Selby. 21. HALIAsTUR LEUCOSTERNUS, Gould, P. Z. S. 1837, p. 138, “ Australia ;” Birds of Austr. i. pl. 4; Valkv. Nederl. Ind. pl. 4. fig. 3. Hab. Celebes, all the Moluccas, New Guinea (Wallace); northern and eastern shores of Australia (Gould); Macassar (S. Miller); Goenong-Tello, district of Gorontalo, Tondano (orsten). H2 56 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Mitvvs, Cuvier. 22. Mitvus arrinis, Gould, P. Z. 8. 1837, p. 140; Birds of Austr. i. pl. 21; Wallace, Ibis, 1868, p. 16; Valkv. Nederl. Ind. pl. 20. fig. 1. Hab. Macassar, Timor (Wallace); Australia (Gould). I hesitate to include Sumatra within the range of this form, as that haditat rests only on the correct identification of a skeleton in the Leyden Museum. Exanvs, Savigny. 23. ELANUS HypPoLEucus, Gould, P. Z. 8. 1859, p. 127, ‘‘ Macassar;” Wallace, Ibis, 1868, p. 17. ?\Elanus intermedius, Schlegel, Mus. Pas-Bas, Milvi, p. 7 (1862). Hab. Macassar (Wallace); and if the same as EL. intermedius, Schlegel, North Celebes, Borneo, Java (Schlegel). PERNIS, Cuvier. 24. PERNIS PTILORHYNCHA (Temm.), Pl. Col. livr. viii. pl. 44, “Java et Sumatra” (26th July, 1823); Verh. Ned. overz. Bezitt. Aves, p. 49, pl. 7, “Sumatra.” Pernis cristata, Cuv.! R. Anim. ed. 2, i. p. 335, “ Java” (1829). Var. celebensis, Schlegel, Valkv. Neder]. Ind. pl. 26. f. 4; Wall. Ibis, 1868, p. 17. Hab. Celebes only, if distinct from Indian and Malayan species (conf. Wallace, 1. c.). In Mr. J. H. Gurney’s opinion the Celebean Pernis should receive a distinct specific title; and Mr. Wallace writes that it is distinct. I have been unable to examine any examples. Baza, Hodgson. 25. Baza mMaeNiRostRIS, G. R. Gray, List Birds Brit. Mus. 1844, p. 19, “ Philippine Islands ;” Kaup, Isis, 1847, p. 343; Schlegel, Nederl. Tijdschr. Dierk. 1866, p- 328; Schlegel, Vog. Nederl. Ind. Valky. pp. 40, 75, pl. 28. f. 4, 5. Pernis crassirostris, Kaup, Contrib. Orn. 1850, p. 77. Hab. Philippines (type); Celebes, Sulu Islands (Wallace); Borneo (Schlegel). Professor Schlegel (Valkv. p. 77) states that the types of Lophotes reinwardtit, Schlegel & Miiller (Verh. Ned. overz. Bezitt. Aves, p. 37, pl. 5. figs. 1, 2, “ Celebes”), were not obtained by Reinwardt in Celebes, and that the Dutch travellers have never obtained it in that island. The Professor, while identifying the Celebean Baza with the Philippine species, points out differences which may eventually prove sufficient to justify the Celebean bird being specifically separated from the Philippine. In the ‘ Hand-list’ Mr. Gray has introduced B. reinwardtii as a second Celebean 1 Tt seems to have been oyerlooked that, although Cuvier discriminated this species in 1817, he only conferred a Latin title on it in 1829. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 387 species. No authentic account of its occurrence in Celebes has as yet been published, while Professor Schlegel and Mr. Wallace restrict its range to Bourou, Ceram, and Amboyna. BuUTEONINA. Pouiornis, Kaup. 26. PoLIORNIS LIVENTER (Temm.), Pl. Col. livr. Ixxiv. pl. 438, “Java” (September 22, 1837); Schlegel, Mus. Pays-Bas, Buteones, p. 21. Hab. Celebes (Wallace); Java, Timor (Mus. Lugd.); Macassar (S. Miller); Toungoo, Burma (mus. nostr.); Siam (Gurney). 27. Po.iorNis npicus (Gm.), Syst. Nat. ed. 13, i. p. 264. no. 68 (1789), ex Latham. Javan Hawk, ath. Gen. Syn. i. p. 34*. no. 8, 7d, “ Java.” Falco javanicus, Lath. Gen. Hist. i. p. 163. no. 87, ex Lath. (1821). Falco poliogenys (Temm.), Pl. Col. livr. lv. pl. 325, “Ile de Lugon” (February 28, 1825). Buteo pyrrhogenys, Schlegel (lapsu calami), Faun. Jap. Aves, p. 21, pl. 78, “ Japan.” Astur barbatus, Eyton, Ann. Nat. Hist. xvi. p. 227, “ Malacca” (1845). Buteo pygmeus, Blyth, J. A. S. B. 1845, p. 177, “ Tenasserim provinces,” op. cit. 1850, p. 339. Hab. Menado (mus. nostr.); Gilolo (mus. nostr.); Lugon (Dussumier); Japan (Schlegel); Morty Isl., Sanguir Isl. (Mus. Lugd.); Java (Latham); Malacca (Eyton) ; Tenasserim Prov. (Blyth). The designation F. indicus, Gm., is rejected by Professor Schlegel (Mus. Pays-Bas, Buteones), on the ground of its being undeterminable. Gmelin gave that title to the Javan Hawk, described by Latham from an individual which flew on board a vessel off the coast of Java. Mr. J. H. Gurney informs me (i epist.) that, having compared Latham’s description with the three Asiatic species of Poliornis, he agrees with the late Mr. Strickland (and consequently with Mr. G. R. Gray, List B. Mus. p. 68, 1848) in identifying it with /. poliogenys, Temm. “ Latham’s description agrees in all respects,” continues Mr. Gurney, “‘ except that he speaks of five transverse bars on the tail, and I have not seen more than four, and in one specimen only three.” My Celebean and Gilolo examples only possess three bars. But in the ‘Fauna Japonica’ Professor Schlegel states that this species has four or five caudal bands. Mr. Blyth tells me that he considers his B. pyymeus to be the same as F. poliogenys, Temm., and that Mr. Eyton’s description of Astur barbatus sufficiently applies to B. pygmeus. Circus, Lacépéde. 28. Circus assmMILis, Jardine & Selby, Illustr. Ornith. ii. pl. 51, juv., “ New Holland” (1826); Schlegel, Mus. Pays-Bas, Circi, p.9; Valkv. Nederl. Ind. pl. 20. figs. 2, 3. Circus jardinii, Gould, P. Z. S. 1837, p. 141, “ New South Wales ;” Wallace, Ibis, 1868, p. 19, adult. Hab. Celebes (Wallace); Menado (mus. nostr.); Gorontalo (Forsten); Macassar 38 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 8. Miller); New South Wales (Gould); Central Polynesia, if identical with C. approai- mans (Peale); Viti islands (Finsch and Hartlaub). Professor Schlegel (/. ¢c.) mentions that the Macassar example in the Leyden Museum, a female in first plumage, obtained by S. Miller, perfectly agrees with the figures of C. assimilis as given by both Jardine and Selby and by Mr. Gould. But C. assimilis, J.&S., and C. assimilis, J. & S. apud Gould (B. Austr. pl. 26), are two distinct species, both inhabiting Australia, but with different ranges, C. assimilis, J. & S., being the young bird of C. jardinii, Gould, pl. 27, and C. assimilis, J. & S. apud Gould, pl. 26, being a distinct species ranging into New Zealand, but not occurring in South Aus- tralia, and named C. gouldi, Bp. Consp. p. 34, ex Austr., and Rey. de Zool. 1850, p. 491, “‘de la Nouvelle Hollande.” C. gouldi, Bp., was described by its author (fide Schlegel, l.c.) from specimens in the Leyden Museum, “acquis comme provenant de la Pata- gonie;” and Professor Schlegel identifies them with C. macropterus, Vieill. Mr. J. H. Gurney is of opinion that they are not C. macropterus, that the types came from Australia, as twice over stated by Prince Bonaparte, and not from South America, and that they are identical with C. assimilis, J. & S. apud Gould, nec J. & S.C. assimilis, J.&8., =C. jardinii, Gould, Mr. Gurney informs me, has alone been obtained in Celebes. C. wolfi, Gurney, P. Z. 8. 1865, p. 823, pl. 44, ex New Caledonia, and which Messrs. Finsch and Hartlaub (Centr. Polyn. p.7) identify with one of the two Aus- tralian Harriers (for they confound the two), Mr. Gurney assures me is a perfectly distinct species. STRIGIDZ. ATHENE, Boie. 29. ATHENE PUNCTULATA (Quoy et Gaim.), Voy. Astrolabe, Zool. i. p. 165, pl. i. f. 1, “ Celebes” (1830); Mus. Pays-Bas, Striges, p. 29. Hab. Macassar (Wallace) ; Menado (Schlegel). 30. ATHENE OCHRACEA (Schlegel), Nederl. Tijdschr. Dierk. 1866, p. 183, “ Negrilama, Celebes” (1866). Hab. Celebes (Rosenberg). I refer this species to Athene with some doubt, never having seen an example. The description reads like that of a Minox. One specimen only, and that of a female, seems to be known. Professor Schlegel (7. c.) remarks that it resembles generally his Noctua philippensis’, but that it has a longer tail, and that the style of colouring differs. ” Nederl. Tijdschr. Dierk. 1866, p, 183, = Winow philippensis, Bp. Compt. Rend. xli. p. 655 (1855)? VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 39 Epaiattes, Keyserling & Blasius. 31. EpnraLres macicus (Schlegel), Faun. Jap. Aves, p. 25, “ Amboyna, Celebes ” (1842) ; Bp. Consp. p. 46; Schlegel, Mus. Pays-Bas, O¢/, p. 22. Ephialtes leucospila, G. R. Gray apud Wall. Ibis, 1868, p. 25. Hab. Amboyna (S. Miller); Gorontalo (Forsten). The range of the Celebean species referred to the above title has not as yet been ascertained ; and its right to that title even has yet to be proved. Otus magicus is the MS. title given by S. Miiller to a Scops Owl inhabiting Amboyna. Professor Schlegel (Faun. Jap.) seems to have identified the Celebean bird with that of Amboyna; for there is no note in the Mus. Pays-Bas of Celebean examples obtained by S. Miiller. Unfor- tunately, as Miiller never published a description, his title cannot be fixed on the Amboyna bird. If Professor Schlegel is right in considering the Amboyna and Cele- bean species identical, there can be no question that their title must stand EZ. magicus But this view is not adopted by Mr. Wallace, who identifies (/. c.) the Celebean species with the Papuan E. leucospila, G. R. Gray, and leaves E. magicus as the title of the Amboyna and Ceram forms. Mr. J. H. Gurney is doubtful whether Z. lewcospila can be separated from E. magicus, but has had no Amboyna examples for comparison. He has kindly sent me the following note on the subject:—‘‘The Norwich Museum has ten specimens of Ephialtes leucospila, but only one of E. magicus as limited by Mr. Wallace. This specimen does not differ from some of those of E. lewcospila more than they do amongst themselves; and I am therefore disposed to agree with Professor Schlegel in thinking that the two are not really separable, unless it be right to separate the different phases of Z. leucospila, which appear to vary somewhat in measurement, and also a good deal in the relative darkness of their markings. In the - following list of the Norwich specimens I begin with the darkest and finish with the lightest, and I also give the length of the tarsus and of the wing from the carpal joint to the tip:— Wing. Tarsus. Ki. LEUCOSPILA. No. 1. East Gilolo’ .......... 7” 6” | 1” 3!" | Very dark. Mr. Gray’s type specimen. 2. Morty Island .......... “© 1 2 | Dark and rufous. 3. Morty, Island’). : = c-s-..: GG ae) 4. Morty Island,9........ (an0 1 2 By Gai sc odoeicde candee 6 9 1 2 | Intermediate in colour between the darkest and (; Ulsremr Gy Gon tance doac ek 1 2 lightest extremes, and all nearly alike. Tio BEEN CO Gasarioes ad 7 0 Abo he) SL Cele bes; 'D) e ysreisayeysheleys-3 7 6 13 GL TiiiibGy ooseon nooete Ws iB e . RAV Beaet ay. ok. mp. \wedes } Mipet oplor pad nearly alka. 1.6 ea 7 9 13 Rather more mottled on the back than the specimen AOD Mas ige Orr) aay of E. leucospila, but comes very near to no. 8. 1 Erroneously given in P. Z. 8. 1860, p. 345, as 6” 6". 40 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. “Tf the races are separable, I should think that probably the birds from Ceram, and Amboyna also (according to Wallace), and perhaps those from Celebes, should stand as E. magicus, and those from Morty, Gilolo, Ternate, and Batchian as E. leucospila, from which the pale-coloured birds from Bouru may be also separable. But the differences are too slender to form a basis for specific distinction, and very probably are not constant.” Mr. G. R. Gray (Hand-list, i. p. 46) treats these forms as distinct species, but makes them both to be inhabitants of Celebes. 32, EPHIALTES MENADENSIS (Quoy et Gaimard), Voy. Astrolabe, Zool. i. p. 170, pl. 2. fig. 2, “ Menado” (1830); Schlegel, Mus. Pays-Bas, Of’, p. 20; Wallace, Ibis, 1868, p. 25. Hab. Gorontalo (Forsten); Macassar, Menado, Island of Flores (Wallace). Dr. Hartlaub (Faun. Madagasc.) identified the Madagascar brown form, Scops mada- gascariensis, Grandid., with the Celebean EF. menadensis, but retained S. rutilus, Pucher. (Archives du Mus. iv. pl. 22), as a distinct species. Professor Schlegel (Rech. s. 1. Faun. Mad.) concurs with Dr. Hartlaub, but besides points out that S. rutilus is nothing but the rufous phase. Mr. J. H. Gurney (Ibis, 1869, p. 452) admits the identity of the two Madagascar forms, but considers the Madagascar to be a larger local race of the Celebean E. menadensis, and (in epist.) “ would be disposed to rank it as one for which a specific name is convenient.” One of Forsten’s Celebean examples (Mus. Pays-Bas, /. ¢.), “‘ teintes tirant fortement au roux,” leads us to expect that E. menadensis will yet be found in Celebes exhibiting the rufous livery of S. rutilus, Pucher. The Flores habitat rests solely on the authority of Mr. Wallace. Celebean examples only are contained in the Leyden Museum. Ninox, Hodgson. 33. NINox JAPONICUS (Bp.), Consp. i. p. 41 (1850), ex Schlegel, Faun. Jap. pl. 9. Noctua hirsuta japonica, Schlegel, Nederl. Tijdschr. 1866, p. 182. Hab. Celebes (Von Rosenberg); Japan, China (Schlegel). The occurrence of a species of Minow in Celebes was first made known by Professor Schlegel (/. ¢.). One example, collected by Von Rosenberg, is stated by the Professor to be absolutely identical with Japanese and Chinese individuals. A second Celebean example, obtained by the same collector, Professor Schlegel considers to be more nearly related to the Minox of continental India. A third example, sent from the island of Sanghir, the same author regards as most nearly resembling the Bornean form Athene borneensis, Bp., but with larger dimensions. The range of the subgenus Ninow is extensive. Its members are found in Ceylon, which furnished the type of Strix hirsuta, Temm.; in Southern and Central India, §. lwgubris, Tickell; in the Himalayas, J. nipalensis, Hodgs., whence they extend eastward and north-eastward to Japan, where VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 4] they become A. japonica, Bp. To the southward they are found in Bengal, Burma, and Cochin China. In the Malaccan peninsula they bear the title of A. malaccensis, Eyton, in the Andamans, NV. affinis, Tytler; while of the Indo-Malayan Islands, Sumatra contains the type of S. scutulata, Raffles; Borneo, A. borneensis, Bp.; and the Philippines, WV. philippensis, Bp. (Compt. Rend. xli. p. 655, 1855). A skeleton in the Leyden Museum is our only evidence of Java possessing a species of this group, to which A. florensis, Wallace, ex Flores, appears also to belong. The Madagascar JV. madagascariensis, Bp., so closely resembles the Indian Minox, that Dr. Hartlaub (Faun. Madagasc.) considers that it can hardly be separated as even a local race (conf. J. H. Gurney, Ibis, 1869, p. 453). Enough has been said to show that all the local varieties have yet to be rigidly compared with one another before the exact title of the Celebean Minox can be absolutely determined. Strix, Linneus. 34. STRIX ROSENBERGI, Schlegel, Nederl. Tijdschr. iii. p. 181, ‘ Celebes” (1866); Wallace, Ibis, 1868, p. 26. Hab. Molido, Boni, Gorontalo (Rosenberg); Menado (mus. nostr.); Macassar (Wallace). A very distinct and fine species. PICARLA. PICIDZ. MULieripPicus, Bonaparte. 35. MULLERIPICUS FULYUS (Quoy et Gaimard), Voy. Astrol. i. p. 228, pl. 17. f. 2, 6, “ Celebes” (1830); Malh. Monogr. i. p. 53, pl. 14. f.1,¢, £.2,2. Hab. Macassar, Menado (mus. nostr.). The affinities of this interesting species are nearer to W/. pulverulentus (Tem.) than to the group of large black-and-white species represented by P. javensis. Malherbe (J. ¢.) erroneously referred P. fulvigaster, Drap. (Dict. Sc. Nat. viii. p. 621, ex Java), to this species, instead of to P. javensis, 3, Horsf. (1822), =P. horsfieldii, 3, Wagler (1827), =P. leucogaster, Reinw. (1830). Yunerpicus, Bonaparte. 36. Yuneiricus TemMMincKi (Malherbe), Rev. Mag. Zool. 1849, p. 529, “Celebes;” Malh. Monogr. i. p. 155, pl. 36. f. 3, 2. Hab. Celebes (Mus. Lugd.) ; Macassar (Wallace). VOL. VIII.—PaRT Il. May, 1872. I 42 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Founded on a single example of a female in the Leyden Museum. Allied to Y. kisuki, but considered a good species by Temminck and Bonaparte. Prince Bonaparte (Consp. i. p. 129) described a specimen of a Woodpecker, Picus sanguineus, Lichtenst. (Cat. Hamb. p. 17), which was wrongly labelled in the Leyden Museum as coming from Celebes, under the title of Venilia albertuli. MEROPID. Meroprs, Linnzus. 37. MeRops PHILIPPINuS, Linn. Syst. Nat. ed. 13' (Vindob.), i. p. 183. no. 5 (1767), ex Brisson. Apiaster philippensis major, Brisson, Orn. iv. p. 560, “ Philippine Islands.” Hab. Menado (mus. nostr.); Indian region. Examples of the Bee-eater, usually referred to Brisson’s Philippine species, from North-east India, Candeish, Malabar, Coorg, Ceylon, Sumatra, and Java, are undistin- guishable ; and my Celebean specimens do not appear to differ. In the Hand-list, no. 1208, Mr. G. R. Gray keeps the species which inhabits India, Ceylon, Java, Flores, Lombok, and Timor separate from the Philippine bird, and refers it to Merops daudini, Cuvier. Cuvier bestowed this title (Régne Anim. i. p. 442) on Levaillant’s Guépier daudin (pl. 14). Levaillant distinctly states that he described his species from examples brought from the Philippines by Sonnerat and Poivre. The title of Merops daudini therefore applies to a Philippine species, and cannot be used for the Indian species even if the Indian bird really does differ. 38. MERops ornatus, Latham, Ind. Orn. Suppl. p. xxxv, “New Holland” (1801); Wallace, Ibis, 1860, p. 147; P.Z.S. 1862, p. 338. Hab. Celebes (Wallace); Java (mus. nostr.); Flores, Lombock, Timor, Sula Islands, Sumbawa, Ternate, Mysol, New Guinea (Wallace); Gilolo (Bernstein); New South Wales, South Australia (Gould); Clarence River, Port Albany (mus. nostr.). Sula-Island examples perfectly agree with Australian. The Philippine Bee-eater referred to this species by Von Martens (J. fiir Orn. 1866, p. 17), seems to belong to another species allied to MZ. viridis, Linn. Meropogon, Bonaparte. 39. MrRopocon Forsteni (Temm.), Bp. Consp. i. p. 164, « Celebes” (1850); Schlegel, Mus. Pays-Bas, Merops, p. 8; Meyer, J. fiir O. 1871, p. 231. Hab. Tondano (Forsten) ; Rurukan (Meyer). Mr. Wallace failed in obtaining this species (Ibis, 1860, p. 142). * In the twelfth edition (“‘ Holmiz ”) the title of this species was omitted by the printer’s mistake. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 43 CORACIIDA. Coracias, Linnzus. 40. Coractas TEMMINCKH (Vieillot), N. Dict. Sc. Nat. xxix. p. 435, “Inde” (1819), ex Levaillant, Hist. Nat. Promér. & Guép. iii. Suppl. p. 46, pl. G; Schlegel, Mus. Pays-Bas, Coraces, p. 138; Wallace, Malay Archipel. i. p. 337. Coracias papuensis, Quoy et Gaimard, Voy. de l’Astrol. Zool. p. 220, pl. 16, “ Dorey,” errore (1830). Hab. Kema (Forsten); Gorontalo, Modelido (Von Rosenberg) ; Menado (mus. nostr.) ; Macassar (Wallace). In the ‘ Hand-list, no. 899, Mr. G. R. Gray extends the range of this species to the Sula Islands. Mr. Wallace is unable to confirm this statement, but writes to me that it is probable. I have failed in finding any confirmation among the Dutch writers. Evrystomus, Vieillot. 4]. Evrystomus oRieNTALIS (Linn.), S. N. ed. 12, i. p. 159, ex Briss.; Schlegel, Mus. Pays-Bas, Coraces, p. 139. Eurystomus pacificus (Lath.), ap. Wallace, P. Z. S. 1862, p. 339. Hab. Limbotto, Gorontalo, Bongka, Ayer-pannas, Boné (Von Rosenberg); Menado (mus. nostr.); Indian region. The Eurystomus of Celebes belongs to the Asiatic and not to the Australian type, E. pacificus (Lath.). It is not to be distinguished from Ceylon and Indian examples. ALCEDINID. DAcELONIN2. Monacuatcyon, Reichenbach. 42. MoNACHALCYON PRINCEPS (Forsten), Mus. Lugd.; Bp. Consp. i. p. 154, “ Celebes,’ adult; Schlegel, Vog. Ned. Ind. Jjsvogels, pl. 7. f. 1, 2, 3. Dacelo cyanocephalus, Forsten, Mus. Lugd. monachus, Temm. Mus. Lugd. Monachalcyon monachus, Sharpe, Mon. Alcedinide, pt. xi. no. 87. Hab. Kema, Menado (Forsten). Mr. Sharpe, in his excellent Monograph, has adopted the specific title of monachus, given by Temminck to the very young bird. I have preferred, in the absence of any recognized rule in such a case, to use the title bestowed by Forsten on the fully adult bird. Ceram and Ternate are cited by Mr. G. R. Gray (Hand-list, no. 1068) as additional habitats of this remarkable species; it appears, however, to be a purely Celebean bird. ee 44 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. SAvRoPATIS, Cabanis. 43. SAUROPATIS CHLORIS (Bodd.), Tabl. Pl. Enl. p. 49 (1783), ex Buff.; Sharpe, Monogr. pt. xii. no. 102. Alcedo collaris, Scopoli, Fl. Faun. Insub. ii. p. 90 (1786), ex Sonn. Hab. Macassar (Wallace). The geographical distribution of this species will be found fully given by Mr. Sharpe (6S @). 44. Sauropatis sancra (Vig. & Horsf.), Trans. Linn. Soc. xv. p. 206, ‘“‘ New Holland” (1825); Sharpe, Monogr. p. xii. no. 104. Mr. Sharpe (/.c.) has not included Celebes within the range of this species; but Mr. Wallace has informed me that he obtained it, as well as S. chloris, at Macassar. 45. SAUROPATIS FoRSTENI, Temm. Bp. Consp. i. p. 157, “Celebes” (1850); Schlegel, Mus. Pays-Bas, Alcedines, p. 37; Schlegel, Vog. Ned. Ind. Jjsvogels, p. 29, pl. 11. f.1; Sharpe, Monogr. pt. xii. no. 103. Hab. Gorontalo (Forsten). The type specimen, an adult female, preserved at Leyden, is the only individual known. In Prince Bonaparte’s diagnosis Professor Schlegel (/. c.) substitutes the words “subtus nigrescens” for “subtus alba.’ My. Sharpe informs me that “it is close to H. chloris, of which perhaps it is only an accidental variety.” Todiramphus funebris (Forsten), Bp. (J. c.), is from Gilolo, and not from Celebes, nor has Alcedo diops, Temm., been found there since Temminck described the species. CaLLiaLcyon, Bonaparte. 46. CALLIALcyon RuFA (Wallace), P. Z. S. 1862, p. 338, “Sula Islands and Celebes.” Halcyon coromanda (Lath., pt.), Sharpe, Monogr. pt. ix. no. 69. Hab. Celebes, Sula Islands (Wallace); Macassar (Wallace). The Celebean Callialcyon is the largest and most brilliantly coloured of the group. In both these respects it differs; and I therefore do not hesitate to retain Mr. Wallace’s title. Cirtura, Reichenbach. 47. Cirrura cyanotis (Temm.), Nouy. Ree. livr. xliv. pl. 262, “Sumatra” (!), errore (March 27, 1824); Sharpe, Monogr. pt. ii. no. 10, “ Celebes;” Wallace, Ibis, 1860, p. 142. Hab. Kema (Forsten); Celebes (Wallace); Menado (mus. nostr.). The true habitat of this species was made known by Professor Schlegel some seven years ago (Mus. Pays-Bas); and to Mr. Wallace we owe not only a confirmation of the fact, but interesting notes on the habits of the bird. It is not improbable that the differences whereon Mr. Sharpe founded his C. sanghirensis will prove to be common to the Celebean bird in certain phases of plumage. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 45 Crycopsis, Salvadori. 48. Crycopsis FALLAX (Schlegel), Nederl. Tijdschr. Dierk. 1866, p. 187, “Celebes;” Sharpe, Monogr. part v. no. 37. Hab. Edges of creeks in the mountainous parts of Celebes (Schlegel) ; Menado (mus. nostr.). Tanysiptera riedelii, J. Verreaux, N. Arch. du Mus. ii. Bulletin, p. 23, pl. 3. f. 1 (1866), is not a Celebean bird. See P. Z. 8. 1872, p. 1. ALCEDININA. PELARGoPSIS, Gloger. 49, PELARGOPSIS MELANORHYNCHA (Temm.), Pl. Col. 391, ‘Celebes” (10th June, 1826); Sharpe, Monogr. pt. ix. no. 66. Hab Celebes (Reinwardt); Menado (mus. nostr.); Sula Islands (Wallace). Aucrepo, Linneus. 50. ALcEDO MoLUccENsIS, Blyth, J. A.S. B. xv. p. 11, “ Moluccas” (1846); Cat. Calc. Mus. no. 215, p. 49, ‘‘ Celebes, Moluccas ;” Sharpe, Mongr. pt. x. no. 74. Alcedo minor moluccensis, Schlegel, Mus. Pays-Bas, Alcedines, p. 8; Schlegel, Vog. Ned. Ind. Tjsvogels, p. 5, pl. 1. f. 4, 5. Hab. Gorontalo (Forsten); Celebes, Bouru, Gilolo, Flores (Wallace); Salawati, Ceram, Batchian, Mysol, Amboina (Von Rosenberg). It was probably Celebean examples of this species which Temminck mistook for the common European Kingfisher (Pl. Col. 272, note). 51. ALCEDO ASIATICA, Swainson, Zool. Illustr. Ist ser. i. pl. 50, “some part of India” (1820-21); Sharpe, Monogr. pt. x. no. 75. Alcedo meninting, Horsf. Trans. Linn. Soc. xiii. p. 172, “ Java” (1822); Temm. Nouv. Ree. pl. 239. f. 2, “ Java and Sumatra ;” Schlegel, Vog. Ned. Ind. [/svogels, p. 6, pl. 3. f. 2, 3. Hab, Indo-Malayan region, Macassar (Wallace); Gorontalo (Von Rosenberg); Lom- bock (Wallace). CYPSELIDZ. MacroprTeryx, Swainson. 52. MAcROPTERYX WALLACII (Gould), P. Z. S. 1859, p. 100, ‘‘ Macassar.” Hab. Celebes, Sula Islands (Wallace); Macassar, Menado (mus. nostr.). This species is closely allied to M. klecho, but differs in being considerably larger and in having the crown of the head, the shoulder-coverts, the edgings of the quills, and the upper surface of the rectrices of a deeper shade of blue-green. Dimensions of wing in M. wallacii, seven inches and a quarter; in JZ. klecho, six inches and a quarter. 46 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. CoLLocaLiA, G. R. Gray. 53. CoLLocaLIA ESCULENTA (Linn.), Syst. Nat. ed. 12, i. p. 343. no. 2 (1766), ex Rumph. Herb. Amb.; Wallace, P. Z. S. 1863, p. 384. Collocalia hypoleuca, G. R. Gray, P. Z. S. 1858, p. 170, “ Aru Islands ;””? Hand-list, no. 749; Ann. Nat. Hist. 3rd ser. vol. xvii. p. 120. Hab. Celebes, Timor, Moluccas, Aru Islands (Wallace). Notwithstanding the reasons advanced by Mr. G. R. Gray (/. c.), Mr. Wallace’s argu- ments in favour of this species being the true Hirwndo esculenta, Linn., appear to me to be decisive. Rumphius does not speak of “ the concealed white spots on the tail-feathers as if there were one on each” (Gray, op. cit. p. 126). On the contrary, by the expres- sion ‘‘ only when the feathers are separated the white spots become visible,” Rumphius leaves it to be inferred that all the white spots are concealed, and therefore that the middle pair of tail-feathers are immaculate. The statement of Linnzus that “ail the tail-feathers are spotted with white,” is an inaccurate rendering of the description given by Rumphius. 54. ConLocaLia FucIPHAGA (Thunberg), Act. Holm. xxxiii. p. 151, pl. 4, “Java” (1772); Wallace, P. Z.S. p. 384. Collocalia nidifica, var., G. R. Gray, Ann. Nat. Hist. 3rd ser. vol. xvii. p. 119; F. Moore, Cat. Mus. E. Ind. Co. i. p. 98. Hab. Macassar (Wallace); Java (H. esculenta, ap. Horsf.); Sumatra (H. esculenta, ap. Raffles); Bourbon, Mauritius (var. francica, Gm.); Neilgherries (H. wnicolor, Jerdon, =C. concolor, Blyth); Malabar coast and Western Ghauts (Jerdon); Ceylon (Layard); Darjeeling (Zickell); Assam (H. brevirostris, McClelland); Bootan (Pem- berton); Sikim (Blyth); the whole of the Malay islands (Wallace). The further limits of this species depend on the true value of H. vanicorensis, Quoy & Gaim., =. leucophea, Peale. Hirunpinapus, Hodgson. 55. HiRUNDINAPUS GIGANTEUS (V. Hasselt): Pl. Col. livr. xli. pl. 364, “Java” (27th August 1825). Chetura gigantea, var. celebensis, Sclater, P. Z. S. 1865, p. 609, “ Menado.” Hab. Sava (type), Sumatra, Celebes (Mus. Lugd.); India (Jerdon) ; Ceylon (Layard). Dr. Sclater (/. ¢.) points out differences which distinguish the Celebean Hirundinapus from typical Javan and Sumatran examples. As one of these distinctions he mentions “a well-marked narrow white patch on the front on each side of the nostrils.” Dr. Sclater also alludes to the Celebean bird as “a very distinct form.” The white frontal marks are also found in the Indian bird, while in a Penang specimen, along with other slight differential characters, Dr. Jerdon (B. of Ind. i. p. 173) found the white frontal patches wanting. This Penang individual thus agreed with the type as described by Temminck. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. AT But it seems possible that the absence and presence of the white frontal spots only denote phases of plumage. If not, the Indian bird will belong to a different species, while the Celebean may be either the same as the Indian (in itself highly improbable), or repre- sent a third form. CAPRIMULGID. Lyncornis, Gould. 56. LyNcORNIS MACROPTERUS, Bp. Consp. i. p. 62, “Celebes” (1850); Wallace, Ibis, 1860, p. 141. Hab. Menado (Wallace). BUCEROTIDZ. Buceros, Linneus. 57. Buceros Exaratus, “ Reinw.,” Temm. Nouy. Recueil, livr. xxxvi. pl. 211, 2, “Celebes” (2nd August, 1823); Schlegel, Mus. Pays-Bas, Buceros, p. 10. (Pl. V. fig. 1, 5; fig. 2, 2.) Hab. Tondano (Forsten); Menado (mus. nostr.); appears to be restricted to the north-eastern parts of Celebes. The male is distinguished from the female by having the throat, cheeks, ear-coverts, sides of neck, and superciliary stripes springing from base of mandible white. In my examples the white supercilium has light ferruginous-brown feathers intermixed. In dimensions the female appears to be somewhat smaller. The example I note from is marked by the collector “female,” while the entirely black individuals are marked “males.” According to Professor Schlegel (J. ¢.) the subject of Temminck’s plate was a female; and, together with Salomon Miller, he describes the male as having the throat and sides of the head white. As this curious form does not belong to any of the established subdivisions of the family, I leave it for the present in the old Linnean genus. It is certainly not a Hydrocissa, as classed by Prince Bonaparte. It belongs to the group of Hornbills, in which the casque and the true maxilla are completely blended together, the pro- longation of the casque forming, in old birds, the apex of the maxilla. CRANORRHINUS, Cabanis. 58. CRANORRHINUS CASSIDIX (Temm.), Pl. Col. 210, 3, “ Celebes” (2nd August, 1823). Buceros cassidiz, Temm.; Schlegel & Miiller, Verhandel. Zool. Aves, p. 24, pl. 4 bis, 2; Schlegel, Mus. Pays-Bas, Buceros, p. 9; Wallace, Malay Archip. i. p. 364. Hab. Tondano (Reinwardt); Menado (mus. nostr.); district of Maros, Macassar (Wallace). The types of the two plates above cited came from Tondano. In the old males the 48 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. colouring of the neck is pale tawny, with scarcely any of the bright ferruginous tint exhibited by the younger birds. Thus the dark chestnut-brown feathers on the crown, occiput, and nape appear more isolated, the much paler hue of the neck-plumage forming a greater contrast. In other respects there are no characters whereby the younger may be distinguished from the older birds, save the somewhat smaller general dimensions, and the form, proportion, and adjuncts of the bill. After the full plumage has been acquired, the bill still passes through three very distinct stages of structure. In the younger (fig. 1) the casque looks more like an inflation of the Fig. 1. Cranorvrhinus cassidix, § jun. culmen than a separate part of the maxilla, so little is it detached. It is swollen posteriorly, and already reaches to above the eye. Anteriorly it falls rapidly towards VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 49 the culmen without exhibiting an erect edge. The cutting-edges of the mandibles are not broken or serrated. In the mature bird the bill measures two inches more than in the younger; yet in the younger bird the mandibles are as high, or are higher, throughout their length than in the fully adult ; or, in other words, in the latter the bill Fig. 3. Cranorrhinus cassidiz, g ad. Cranorrhinus cassidix, 2 . is prolonged at the expense of its height. In this stage there are no traces of the basal lateral plates. The walls of the mandibles are quite smooth, without any indications VOL. VIII.—PART Il. May, 1872. K 50 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. of lateral folds. But the position which is occupied by the lateral plates in older individuals is indicated by a dingy reddish brown colour. In the next stage (fig. 2, p. 48) the bill measures about one inch longer, and has acquired the form which exists in the old bird; but the cutting-edges are unbroken. The casque is more inflated, appears more detached from the culmen, and reaches further back on the crown of the head. In front it stands up at a right angle to the culmen, and is much compressed. A thin, smooth plate has grown on the basal half of the two mandibles; but there are no traces of folds or grooves. The substance of these plates seems to be secreted from the walls of the mandibles. In the fully adult bird (fig. 3, p. 49) the commissure is serrated, notched, and broken. The casque extends back past the line of the eye. Anteriorly it is less compressed than in the previous stages, although not so much swollen as the posterior portion. The anterior edge stands at an acute angle tothe culmen. ‘The casque displays five distinct folds or, rather, undulations. At the base of the mandibles the lateral plates are much thickened. On each side of the maxilla they are divided by a single, deep, diagonal groove into two equally broad flat folds. On the sides of the mandible there are two grooves thus forming three similar folds. In the adult female (fig. 4, p. 49) the commissures are much broken and serrated. The casque is smaller, the anterior edge rising at a right angle to the culmen. It is also divided into five almost equal undulations or folds. The basal plates are divided both on the mandibula and maxilla into three broad flat bands. In a second example of a female (mus. nostr.) a third band has been partially arrested in its development, the groove being being partly obliterated. While the female has certainly three flat bands at the base of both the mandibles, it will be interesting to know whether the male has never more than two at the base of the maxilla. In Temminck’s plate (/. ¢.) the male is figured with only two; while in that given by Schlegel and S. Miiller (J. ¢.) the female is figured with three both above and below. ‘The following dimensions are taken from Menado examples in my collection. The bill is measured in a straight line from the gape to the apex. Dimensions. | Wing. Tail. Bill. Circumference of casque. Casque. Outer. Inner, Height. | inches. inches. inches, inches. inches. inches. Malley younger reels crn. 163 113 62 53 4 13 Male, intermediate...... 173 14 73 53 34 23 iMaletadultiva-ieniierirere Wee 14 82 3 3f 33 Female; adult,.......... 15 12 3 5 37 2 It will be seen that the bill increases in length after the wings and tail have reached their maximum. The inner circumference of the casque is greater in the youngest than in the adult. The anterior part of the casque, at its union with the culmen, VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 51 appears to become absorbed, and to retreat as the bird increases in age; or, as the anterior edge becomes more and more perpendicular to the culmen, it perhaps wears off, or is broken off. This can be traced in one example—the indent or hollow from which the fore part of the casque sprung, and in which it was attached to the culmen, a groove shaped like a V, three quarters of an inch long, not being filled up. Buceros sulcatus, Temm., from the Philippines, and B. corrugatus, Temm., from Borneo, belong to the same genus. CUCULID. ScyTHROPINA. Scyturops, Latham. 59. ScyTHROPS NOVH-HOLLANDL&, Lath. Ind. Orn. i. p. 141, “ Nova Hollandia” (1790) ; Schlegel, Mus. Pays-Bas, Cuculi, p. 36; Temm. PI. Col. 290. Cuculus presagus, Reinw. MS., ex Celebes. Hab. Menado, Macassar (mus. nostr.); Kema (lorsten); Ceram, north coast (Jus. Ingd.); Ceram, south coast, adult males, April (Hoedt); Obi-major, adult male, 29th of June, Batchian, adult male and female, end of June, a male, 8th of September (Bernstein); Flores (Wallace); New South Wales, between October and January (Gould); Cape York (mus. nostr.). Two individuals from the vicinity of Menado are, in their colouring and markings, almost identical with an example from Cape York. The dimensions of the wing and tail also agree. But the bill of the Menado male, measured from the nostril, is full two inches and three quarters in length, and that of the female two and five eighths, whereas that of the Cape-York bird is only two inches and a quarter. In form the pill of the Celebean bird differs from that of the Cape-York example. In the latter the culmen is rounded, smooth, and broad, and there is only one lateral channel or groove present. This starts from above the nostril, and runs in a line more or less parallel with the culmen. In the Menado male the culmen, on leaving the forehead, forms a distinct narrow ridge; on each side of it is a depression or shallow valley, formed and bounded by a second ridge, below which again is the channel observable in the Cape-York bird. In the bill of the Menado female the culmen is sharper and still more clearly defined; and the lateral channels, while being deeper, are prolonged nearly to the apex of the maxilla’. The type of structure is essentially that of the bill in some species of the Bucerotide. We know nothing of this form out of Australia. In that country it is migratory. Its geograpical distribution in the archipelago, as at present known, is anomalous; for it occurs in Flores, and is not recorded from Lombock or Timor. It has been found in Batchian, but not in Gilolo; in Ceram, but not in Bourou. ‘ A Macassar example, since obtained, presents a similar structure. . K2 52 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. PH@NICOPHAINA. Puanicopnaks, Vieillot. 60. Pua@nicopHais caALoRHyNcHUS (Temm.), Nouy. Ree. livr. lix. pl. 349, “Celebes” (25th of June 1825); Wallace, Malay Archip. ii. p. 340. Le Malcoha & bec peint, Less. Complém. de Buffon, ii. p. 618, pl. —. fig. 1. Hab. Gorontalo (Forsten); Menado, Macassar (mus. nostr.). MM. Verreaux proposed (Rev. & Mag. Zool. 1855, p. 356) to restrict Vieillot’s generic title Phenicophaés to a small group consisting of this species, of P. curvirostris, Shaw, P. erythrognathus, Temm., and a fourth species, P. eneicaudus, Verr., not since obtained. And they suggested a new generic title, Alectorops, for the reception of Cuculus pyrrho- cephalus, Forster. But, as Forster’s Ceylon Malkoha is the type of Phenicophaés, this arrangement cannot be recognized. Dr. Cabanis (Mus. Hein. iv. p. 85), concurring in the propriety of separating the Ceylon species from the others, retained it, Vieillot’s type species, in Phenicophaés, and proposed Rhamphococcyx for the small Indo-Malayan group. The grounds for this separation are the great extent of naked space surrounding the eye, the abnormal colouring of the plumage, the form of the bill, and the position and shape of the nostrils in P. pyrrhocephalus. The naked space is certainly more extended than in P. curvi- rostris or P. erythrognathus ; but then P. calorhynchus has the ophthalmic region almost entirely clothed. The colouring of the plumage differs principally in that white replaces the rufous of P. curvirostris and P. erythrognathus, thus evincing an affinity to Rhopo- dytes, Cab. (Zanclostomus of Indian authors, but not of Swainson). ‘The tail is tipped with white instead of rufous; but the upper plumage in all three is green. In P. calorhynchus green is entirely absent, and the tail is uniform in colour. In colouring P. calorhynchus is as much an isolated species as P. pyrrhocephalus. The form of the bill in all four species is very similar; but the position and shape of the nostrils is different in each of the four. The nostril of P. pyrrhocephalus (fig. 8) is placed in a narrow, depressed, lengthened, oval slit, which runs almost parallel with the commis- sure, yet slightly descending. Its situation is almost on the edge of the commissure, and at an unusual distance from the base of the maxilla. In P. curvirostris (Shaw) (fig. 6) the nostrils are set at the commencement of a deep narrow groove or channel. In P. erythrognathus, Bp.' (fig. 7), the nostril is a simple round hole. The nostril of P. calorhynchus (fig. 5) isan elongated slit, like that of P. pyrrhocephalus, but running quite parallel with the commissure, and not so near its edge; nor is it as advanced from the base of the maxilla. The position and shape of the nostrils in these four species is so peculiar and distinctive, that the species could be determined from a fragment of the maxilla alone. ‘The striking difference in the shape of the nasal opening of the Javan P. curvirostris and Sumatran, Moluccan, and Bornean P. erythrognathus (forms which * Consp. i. p. 98. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 53 are otherwise difficult to recognize as distinct species) is very remarkable. ‘The four species form a natural group which cannot be consistently subdivided, unless P. calo- rhynchus be also made the type of a separate genus. Within the limits of Phanicophaés I am also inclined to include Melias diardi, Less., and also Cuculus sumatranus, Raffles Fig. 5. Pheenicophaés calorhynchus. Fig. 7. Pheenicophaés erythrognathus. CUCULIN A. Evupynamis, Vigors & Horsfield. 61. EvpyNaMIs MELANoRHYNCHA, S. Miiller, Verhandel. p. 176; Schlegel, Mus. Pays-Bas, Cuculi, p. 20; Walden, Ibis, 1869, p. 344. Hab. Kema, Tondano, Gorontalo (Forsten); Menado (mus. nostr.). Cacomantis, S. Miiller. 62. CACOMANTIS LANCEOLATUS, S. Miiller, Verhandel. p. 178. Hab. Macassar (Wallace, mus. nostr.); Java (type). The synonymy of the species usually comprised in Cacomantis, 8. Miiller, is still so entangled, that a few general remarks on the Plaintive Cuckoos of the Indian and Australian regions are necessary to enable us to establish the identity of the Celebean member of the genus. 04 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. In India there are two species: 1, C. passerinus (Vahl), without any rufous in the adult plumage; 2, C. tenuirostris (Gray, apud Jerdon), with a rufous belly. Both pass through an hepatic phase. C. passerinus (Vahl) chiefly inhabits western and south- western India and Ceylon; C. tenuirostris, Gray, ap. Jerdon, frequents Bengal and the countries to the eastward, including Burma. In Bengal the two species are said to meet and interbreed. C. passerinus (Vahl) has no representative ; but C. tenuirostris, Gray, ap. Jerdon, is represented in the Malay peninsula by C. threnodes, Cab.; in Borneo by C. borneensis, Bp.; in the Philippines by C. merulinus (Scopoli) verus ; and in Java by Cuculus flavus, Gm., apud Horsf., S. Miiller, &c., =Cacomantis merulinus, Scop., ap. Cab., and Polyphasia merulina, Scop., ap. Horsf. & Moore. The Javan bird, in the hepatic stage, is probably the C. danceolatus, S. Miiller. When fully adult it has the head, nape, throat, and breast pale ashy; the remaining lower parts fulvous, more or less inclining to rufous; the caudal bands are white; and the quills uni- colorous. In the young and in the transition stage these bands, which are broad, equidistant, and unbroken, are rufous, and the quills are either all or partly rufous- banded. This description will apply more or less to all the races above alluded to. C. sepulchralis, 8. Miller, is the title of a third very distinct species, which inhabits Java. When adult it may be at once recognized from C. merulinus of Java by its longer bill, and from all the races of that species by its much longer wings and tail, by the chin, cheeks, and ear-coverts only being pale ashy, the head dark grey, the upper surface bronze-green, and by the whole under surface, the chin excepted, being ruddy fulvous. The white markings on the rectrices are fewer, smaller, and chiefly consist of triangular edge-spots, and not of bands running right through. In transition plumage this is in all probability the C. pyrogaster, Drapiez. C.. sepulchralis, S. Miiller, belongs to the group which includes C. flabelliformis, Lath., C. dumetorum, Gould, and C. insperatus, Gould, from Australia, also several races of small Cuckoos of the Austro- Malayan archipelago, as C. assimilis, G. R. Gray, Aru Islands, C. infaustus, Cabanis, Mysol, and some undetermined species in Goram, Batchian, Morty, and Salawati, likewise C. simus, Peale, Feejee Islands, C. castaneiventris, Gould, Cape York, and C. bronzinus, G. R. Gray, in New Caledonia. No member of this group has been identified as inhabiting Continental Asia; yet the Bengal specimen, stated by Dr. Jerdon (B. of Ind. i. p. 355) to have the rufous extending to the chin, may belong to it. A fourth group of Plaintive Cuckoos is represented by C. tymbonomus, S. Miiller, from Timor; to it belongs the C. pallidus (Lath.) of Australia, and an undetermined species from Waigiou. In C. tymbonomus the upper surface is pale olive-brown, inclining to ashy on the head and rump; the under surface is paler and more cinereous ; under tail-coverts tawny, or pallid rufous; middle pair of rectrices imma- culate, but broadly tipped with brown; the remainder tipped with white, and partially toothed on the inner webs with white. This species and its allies also pass through a rufous phase. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 55 C. sonnerati, Lath., founded on Sonnerat’s Petit coucou des Indes (Voy. Indes, il. p. 211), from its being more or less rufous at all ages, and a small species, has been often confounded with either one or other of the foregoing. Its Javan representative, but slightly differing, is the C. pravatus, Horsf., =C. fasciolatus, 8. Miller, =C. rufo- vittatus, Drapiez. The group is also represented in Sumatra, Malacca, Borneo, and Ceylon. This form, raised to generic rank by Dr. Cabanis (Penthoceryx), has the bill long, broad at the base, and uncompressed throughout its entire length, the maxilla overlapping the mandibula. In old birds the rufous and dark brown bands of the upper plumage are washed with bronze-green. From the chin to the under tail- coverts each feather is white, traversed by usually three narrow, dusky, irregular lines ; the white interspaces being three or four times as broad as the dusky lines. A uniform transverse striated appearance is thus imparted to the under plumage, never found in any other group of the small Asiatic Cuckoos. The middle pair of rectrices are, according to age, either almost entirely dark brown with a bronzy gloss, or else have both sides of the shaft dark brown, indented with bright rufous. The lateral rectrices are never evenly barred through, are always bright rufous with dark cross marks, have a white or else a pale fulvous terminal spot and a penultimate broad dark brown band. Many of the frontal plumes are white at their base and in the centre—a character alone sufficient to distinguish this group from any of the Plaintive Cuckoos in hepatic plumage. C. infuscatus, Hartl., is either another type of the Plaintive Cuckoos, or else it belongs to the same subsection of C. passerinus; or it may prove to be only a phase of C. simus. A Macassar specimen, collected by Mr. Wallace, appears to belong to the group of which 0. merulinus is typical. It has six of the secondary quills with rufous bars, part of the unmoulted hepatic dress; otherwise it is undistinguishable from Javan examples of C. lanceolatus. The lateral rectrices are, as in that species, broadly barred with pure white. It is, however, a larger bird, with wings and tail somewhat longer. Wing 43, tail 48. CENTROPODINA. PyrRHOocENtTOR, Cabanis. 63. PYRRHOCENTOR CELEBENSIS (Quoy & Gaimard), Voy. Astrol. Zool. i. p. 230, pl. 20, “‘ Menado” (1830). Centropus bicolor, Cuvier, Mus. Paris, fide Pucheran, Rey. et Mag. Zool. 1852, p. 472; Lesson, Tr. p. 137 (1831). = Hab. Menado (mus. nostr.); Gorontalo (Forsten). I cannot find that Cuvier ever published his title of C. bicolor. A second species of this subsection inhabits the Philippines (P. unirufus, Cab.). But it is not unlikely that 56 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Cabanis’s species is the same as C. melanops, Less. ex Cuv., said to have been obtained by the Paris Museum from Java (conf. Pucheran, op. cit. p. 473). C. melanops is cer- tainly not a Javan bird; and though Professor Schlegel has identified it with C. rufi- pennis, Illiger, it belongs to a different group of Coucals. Notwithstanding the opinion of the learned Professor, of Prince Bonaparte, who made it equal to C. medius, Miiller, and Dr. Cabanis, Mr. Cassin appears to have correctly identified it with C. nigrifrons, Peale. C. ateralbus, Less., ex New Ireland, is a closely allied form. In P. celebensis, the fully adult bird loses the bright yellow-rufous chin-, throat-, neck-, and breast-plumage of the younger bird. These parts become very pale fulvous, and contrast with the dark chestnut of the remaining lower region. In this state Cuvier’s title of bicolor is applicable. The young bird is bright rufous throughout; and, judging by analogy, the Philippine P. wnirufus, Cab., is the young bird of C. melanops, Less., =C. nigrifrons, Peale. CEntrococcyx, Cabanis. 64. Cenrrococcyx AFFINis (Horsf.), Trans. Linn. Soc. xiii. p. 180, “Java” (1821). Centropus medius, Bernst. Nat. Tjdschr. Ned. Ind. xxi. p. 27; J. fiir Orn. 1860, p. 269. rectunguis, Strickl., ap. Schlegel, Mus. Pays-Bas, Cuculi, p. 69. Hab. Macassar (mus. nostr.); Java (mus. nostr.). The red-and-black Coucals of the Indian region form a natural and well-defined group; and I concur with Dr. Cabanis in the propriety of separating them from the African genus Centropus. Notwithstanding the labours of Dr. Cabanis and Professor Schlegel, the species are far from being clearly established. Examples of two species from Celebes are in my collection, and would, were I to follow Professor Schlegel, be referable to C. rectunguis, Strickl., a title made by the learned Professor to include most of the smaller Asiatic Coucals and even an African species. An examination of a con- siderable series of this group has led me to conclusions widely differing from those contained in the Catalogue of the Leyden Museum. The difficulties which meet a student of the genus Centrococcyz arise from the general resemblance in the plumage of its members, the blue, the green, or the purple hue of the black portion, and the deeper or less intense shades of the rufous not being sufficiently striking and well marked, except in perfect plumage, to be relied on as distinguishing characters. We also find in the Coucals, as in other natural groups the members of which are numerous, the colouring of the adult in one species representing, more or less, the transition colouring of the young of another species. Thus the dingy greenish brown hue of the rectrices in an immature C. rectunguis changes to glossy dark green in the next stage, and is again converted into deep blue in the adult bird. But in the common Indian Crow-Pheasant the colour of the rectrices is arrested at the green stage, and green remains the hue of that part of the plumage in the fully adult bird. A complete series of fully adult examples from all parts is consequently essential VISCOUNT WALDEN ON THE BIRDS OF CELEBES. oT before characters founded on the colouring can be relied on. Another source of diffi- culty is the extreme variability of the plumage in the first and second years, the young wearing a livery greatly resembling in general characters the adult garb of Aus- tralian C. phasianus and its allies. It is likely that this variability is more apparent than real, and that each species, as in Eudynamis for instance, has special phases of immature plumage peculiar to itself. Happily certain characters are always present whereby typical specific groups can be separated from one another. The most important are the dimensions and the form of the bill. Three distinct groups of Asiatic Coucals can thus be separated ; and as Java is the only known locality where all three are repre- sented, we may use the Javan species as standards :— First, C. bubutus, Horsf., as the type of the large species, such as :—C. sinensis, Stephens, ex China; C. rufipennis, Illiger, India and Ceylon; C. ewrycercus, A. Hay, Malacca; C. borneensis, Bp., Borneo; and the very distinct C. chlororhynchus, Blyth, ex Ceylon. This group appears to be unrepresented in Celebes; nor has it been discovered in the Philippines or in Formosa. The second group comprises species smaller in size, with short, but proportionally very high bills, the diminished length of the bill making it appear disproportionately broad at the base. The Javan representative is C. affinis, Horsf. An identical form inhabits Flores; and a somewhat larger species is found in Ceram, which, if the same as the Amboyna Centrococcyx, must stand as C. medius, Bp., ex Miiller. A Macassar individual, collected by Mr. Wallace, although in young plumage, bill pale flesh-colour, and plumage streaked and mottled, so closely resembles Javan C. affinis in dimensions and form of the bill, that I cannot separate it. The third group consists of the smallest species, represented in Java by C. javanensis, Dumont, =C. lepidus, Horsf. In these the bill is a miniature resemblance of that of the second group. The upper tail-coverts are highly developed, or, in other words, they are the tail-coverts of the C. affinis group. C. bengalensis, Gm., of India; C. viridis, Scop., of the Philippines; C. moluccensis, Cab., ex Bernst., of Ternate, very near to C. javanensis, but with a proportionately longer tail; and C. dimidiatus, Blyth, =C. lig- nator, Swinh., of Formosa and China, come within these limits. Malaccan and Banjar- massing individuals scarcely differ from the Javan species; and judging from the mea- surements given by Professor Schlegel, representatives occur in Ceram, Sambawa, Bangka, Sumatra, and Amboyna. A Celebean example of a young bird, with a pale bill and buff-streaked plumage, offers no distinctive characters, either in its proportional dimensions or in the form of the bill, whereby it can be separated from C. javanensis. Yet it must be recollected that the adult bird may present characters more than enough to stamp it as distinct. C. rectunguis, Strickland, a perfectly distinct species, as yet only known from Malacca, is a miniature C. rufipennis. By its external structure it belongs to the first group. The bill is long and much compressed. The wings are short and much rounded, the VOL. VIMI.—PaRT U. May, 1872. L 58 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. seventh quill, if any thing, the longest. The claw of the hallux is short, and therefore appears straight; in character, it is the claw of the large species, shortened in pro- portion to the size of the bird. In the second and third groups the hind claws are not proportionately diminished, and therefore appear disproportionate when compared with those of the large species. The upper surface of the tail is a deep, rich, almost purple blue. The nape is glossy violet-blue, contrasting with the duller green-black of the head. The adolescent plumage possesses characters peculiar to the species. From this it will be seen that it has no characters in common with the members of the second and third groups. It does not appear to be contained in the Leyden Museum. The following is a synopsis of the smaller Asiatic Coucals, together with the dimen- sions of the individuals I have had access to. Cuculus tolu, Gm., ex Madagascar, seems to belong to the Asiatic and not to the African section of Centropodine. 2X, 1. Centrococcyx affinis (Horsf.), Trans. Linn. Soc. xiii. p. 180, “Java” (1821). Cuculus totu, Gm., ap. Raffles, Trans. Linn. Soc. xiii. p. 285, is either the Sumatran form of this species or else of C. javanensis. Hab. Java, Sumatra (?), Celebes, Flores. 2. Centrococcyx medius (Bp.), ex Miiller, Consp. i. p. 108, “‘ Amboyna” (1850). Professor Schlegel applies this title to the Amboyna species. Prince Bonaparte includes the Javan form. Hab. Amboyna, Ceram. B. 1. Centrococcyx javanensis (Dumont de Ste. Croix): Dict. Sc. Nat. xi. p. 144, “Java” (1818). Centropus lepidus, Horsf. Trans. Linn. Soe. xiii. p. 180, “ Java” (1821). Centropus pumilus, Less. Traité, p. 136, “ Java” (1831). Lesson described partly from the example on which Dumont founded the title of javanensis. My Bornean example is smaller than Javan individuals. Hab. Java, Malacca, Banjermassing, Celebes. 2. Centrococcyx viridis (Scop.), Del. Fl. Faun. Insub. ii. p. 89 (1786), “ Philippines,” ex Sonn. Centropus philippensis, Cuv. R. An. i. p. 426 (1817), ex Pl. Enl. 824, “ Philippines.” Corydonix pyrrhopterus, Vieill. Enc. Method. iii. p. 1353, ex Pl. Enl. 824, “ Philippines” (1828). Centropus molkenboeri, Bp. Consp. i. p. 108, descr. orig. (1850), ‘ Philippines.” In the first edition of the Régn. An., Cuvier erroneously quotes Pl. Enl. 884. In the edition of 1829 this error is corrected. Vieillot founded his species on Pl. Enl. 225—also VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 59 an obvious error. Jam not acquainted with this species. It is admitted as perfectly distinct by Professor Schlegel. Hab. Philippines. 3. Centrococcyx moluccensis, Cab., ex Bernst. Mus. Hein. iv. p. 113 (1862), “'Tinor” (Timor 2). The types of Bernstein’s MS. title moluccensis, in the Berlin Museum, were from Ternate. Is Tinor a misprint for Timor or Tidore? Hab. Ternate. 4, Centrococcyx bengalensis (Gm.), Syst. Nat. ed. 13, i. p, 112 (1788), “ Bengal,” ex Brown, pl. 13. Corydonix maculatus, Vieill. Enc. Meth. iii. p. 1353 (1823), ex Brown, pl. 13. Centropus pygmeus, Hodgs. Gray’s Zool. Mise. p. 85 (1844), “ Nipaul.” Centropus viridis, Scop., ap. Blyth, Jerdon, Horsf. & Moore, and Swinhoe. The following titles are usually associated with this species:—Polophilus lathami, Leach, Zool. Misc. pl. 56, described from a British Museum specimen ; locality unknown. The species is undeterminable, Leach’s plate and description being insufficient. Centropus rufinus, Cuv. R. An. (1817), p. 426, and Polophilus rufus, Stephens, Gen. Zool, Aves, ix. p. 44 (1815), titles founded on Levaillant’s 221st plate (Ois. d’Afr.), would take precedence of C. affinis, Horsf., if, as suggested by Professor Sundevall, Levaillant figures the Javan bird (Krit. Framst. p. 48). Dr. Cabanis deems it more probable that the “Lesser Indian Coucal” formed the subject of Levaillant’s plate. From the figure it is impossible to decide which of these two opinions is correct; while Lyeaillant’s mendacious account only tends to mislead us. Hab. Bengal, Mysore, Central India, Burma. 5. Centrococcyx dimidiatus (Blyth), J. A. 8. B, 1842, p. 945, “ Chusan.” Centropus lignator, Swinhoe, Ibis, 1861, p. 48, ex Formosa, Amoy, Hong Kong, Centropus viridis (Scop.), ap. Swinhoe, P. Z. S. 1863, p. 266, “ South China, Formosa ;” Ibis, 1870, p. 235, “ Hainan.” It is not as yet satisfactorily determined whether the Chinese Lesser Coucal is a distinct species—and if not, whether it is the same as C. viridis (Scop.) or C. benga- lensis (Gm.). Hab. South China, Hainan, Formosa (Swinhoe). 60 VISCOUNT WALDEN ON THE BIRDS OF CELEBES., Dimensions. Wing. | Tail. | Tarsus. | Bill. Claw. inches. | inches. | inches. | inch. inch. C. affinis, Horsf. ...... 7-000} 9-750 | 1-8750) 0-6825} 1-125 Java. Not quite adult. > pa cseaoes 6°875 | 9°375 | 1:7500) 0-6250| 1-250 |Macassar. @ young. 5 Fay 4 | Se Oe 6-750 | 9-625 | 1-7500) 0-6825| 1-125 |Flores. Moulting into adult plumage. C. medius, Bp., ex Miill...) 7-875 |10-000 | 2-1250) 0-7500} 1-375 |North Ceram. adult. C. javanensis, Dumont ..| 6-250) 7-875 | 1-5000 0°6250} .... \Java. Not quite adult; claw broken. 3 ..| 5°875 | 7-250 | 1-5000) 0-6250) .... |Banjarmassing. Adult; hind claw broken. .| 5°375 | 6:875 | 1:5625) 0-5625) 0-875 |Penang. Very young. 6:125 | 8°375 | 1-7500, 0-6250) 1:125 Malacca. Adult. 6-000 | 8-625 | 1-5625) 0-6250) 1:125 |Macassar. Young. ” 2C. moluccensis, Cab. ....| 6125} 9-000 | 1-6250) 0-6250| 1-000 |Timor. Moulting into adult plumage. C. bengalensis, Gm. ....| 6-500} 8-000} 1-2500) .... | 1-000 \Jerd. Birds of Ind. i. p. 351. C. rectunguis, Strick. ..| 6°375| 9-000 | 1-7500) 1-0000) 0-625 |Malacca. Adult. Ns 5 ..| 5°875| 8-375 | 1°7500, 0-9375] 0-625 |Malacca. Young. 65. CENTROCOCCYX JAVANENSIS (Dumont de Ste. Croix), Dict. Sc. Nat. xi. p. 144 (1818), “ Java.” Centropus lepidus, Horsf. Trans. Linn. Soe. xiii. p. 180 (1821), “Java.” Centropus pumilus, Less. Traité, p. 1386 (1831), “Java;” Pucheran, Rev. Mag. Zool. 1853, p. 56. Centropus affinis, Horsf., ap. Bernst. Nat. Tijdschr. Ned. Ind. xxi. p. 27; J. fiir Orn. 1859. p. 185, 1860, p. 269. Hab. Macassar, Java, Malacca, Banjarmassing (mus. nostr.). An interesting account of the habits and nesting of this species in Java, and of the peculiar structure of its spinal column, has been given by Bernstein (J. fiir O. /. c.); also detailed observations on parts of its internal anatomy, and of that of C. affinis (Horsf.), by the same author, in the Tijdschrift (/. c.). The skeleton of the Celebean bird will have to be compared with that of the Javan before the absolute identity of the two species can be established. PASSERES. ORIOLID. BroveEripus, Bonaparte. 7 66. BRropDERIPUS coRoNATUS (Swains.): An. in Menager. p. 342 (1837), “ Java.’ Oriolus horsfieldii, Bp. Consp. p. 348 (1850), “ Jaya.” Oriolus galbula, ap. Horsf. Trans. Linn. Soe. viii. p. 152, “ Java.” Oriolus hippocrepis, Wagler, part., Syst. Av. Oriolus (1827). Oriolus indicus, Schlegel, Mus. Pays-Bas., Coraces, p. 102. Hab. Java (mus. nostr.); Macassar, Menado (Wallace); Bougka, Gorontalo (Von Rosenberg). VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 61 I have compared two Macassar male examples collected by Mr. Wallace with a large series of Javan individuals, and have failed in detecting any valid specific differences. The black-naped Orioles, before attaining their full plumage, pass through a stage wherein the two centre rectrices retain the olive-green hue found in younger birds, while they have already put on the black feathers which surround the head, and the full bright adult yellow plumage of the entire under surface, the crown, the neck, and the rump, the plumage of the back alone showing immaturity by traces, more or less, of dingy greenish-yellow. It would seem that the central pair of olive-coloured rectrices are not moulted and replaced by a pair of new black feathers, but rather that the olive- green hue changes gradually into black, commencing from near the tips, which are pure yellow at the earliest stage, and thence passing upwards. In adult Javan examples the lesser wing-coverts are tipped with yellow, thus forming a conspicuous yellow speculum. But in Javan examples in the stage of plumage above described, these yellow tips are frequently absent, or only commencing to be developed. ‘The two Macassar examples are in the intermediate stage of plumage described above: one has no yellow tips to the lesser wing-coverts; in the other they are just appearing. Whether in perfect plumage the yellow alar bar is wanting, as in the Sula B. frontalis, has yet to be ascer- tained. In the mean time I shall retain the Macassar Oriole under the title of the Javan bird. The Macassar species is somewhat larger. Wing 5, tail 4$, bill $. The only Menado example I have been able to examine is in the intermediate stage of plumage, with green middle rectrices and no alar bar. It differs in that the black _ coronal ring does not unite at the nape, the yellow of the crown being thus confluent with that of the nape. As indications of the complete black circle in Broderipus appear in the earliest stages of plumage, this break in the coronal ring cannot be a sign of nonage. The dimensions differ from those of the southern form. Wing 53, tail 43, bill 3. It possibly represents a distinct species. TURDIDZ. GeocicHLa, Kuhl. 67. GrocicHta ERYTHRONOTA, Sclater, Ibis, i. p. 113, “ Macassar” (1859). (Pl. VL. fig. 2.) Hab. Macassar (Wallace). This species and G. interpres (Kuhl) form a section of the genus which perhaps deserves a subgeneric title. Turdus avensis, J. E. Gray, Griffith, Anim. Kingd. Birds, i. p. 530, pl. —, named from an Indian drawing, is either G. interpres or else an unknown Burmese representative form. 62 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. TIMALIID. TricHostoma, Blyth. 68. 'TRICHOSTOMA CELEBENSE, Strickland, Contr. Ornith. 1849, p. 128, pl. —, ‘“‘ Celebes.” Hab. Macassar (Wallace). A species of the above genus, collected by Mr. Wallace, is referred, with some doubt, to the bird figured and described by the late Mr. Strickland. The chin and throat are white; the rest of the under surface is washed with pale ferruginous faintly tinged with brown. The upper plumage and wings are dark olive-brown, the loose plumes of the lower back being tinged with rusty, and the upper tail-coverts being distinctly rust- coloured. The outer edges of the rectrices are rusty brown. Lores and cheeks dingy white. Wing 24, tail 23, tarsus 1. While evidently belonging to the genus Trichostoma, this species differs structurally from T. bicolor (Lesson) of Sumatra and Malacca, by having the rictal bristles but slightly developed and the tail proportionally short. PITTID. Me anopirra, Bonaparte. 69. MELANOPITTA FORSTENI (Bp.), Consp. i. p. 256, “ Celebes” (1850). Pitta melanocephala, Forsten (nec Wagler), Schlegel & S. Miiller, Verhandl. Zool. Aves, p. 19, “Tondano ;” Westerman, Bijdr. i. pt. 6. p. 46, pl. 2; Schlegel, Vog. Ned. Ind. Pitta, p. 5, pl. 2. f. 1. Brachyurus forsteni, Bp. Elliot, Monogr. p. 88, pl. 24; Wallace, Ibis, 1864, p. 106. Hab. Kema, Tondano (Forsten). EryTuropitta, Bonaparte. 70. ERYTHROPITTA CELEBENSIS (Forsten), Schlegel & S. Miiller, Verhandel. Zool. Aves, p- 18, “ Tondano ;” Westerman, Bijdr. i. p. 6. p. 46, pi. 3; Schlegel, Vog. Ned. Ind. Pitta, p. 17, pl. 4. fig. 4; Wallace, Ibis, 1864, p. 106. Brachyurus celebensis (Forst.), Elliot, Monogr. p. 67, pl. 17. Hab. North Celebes (Forsten). This species was found to be scarce by Mr. Wallace (Ibis, 1860, p. 142). When remarking that three species of Pitta inhabited Celebes, Mr. Wallace (/. c.) was probably misled by Bonaparte’s Conspectus, wherein P. miilleri, Bp., is stated to be from Celebes instead of Borneo. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 63 SAXICOLID. Monticoxa, Boie. 71. Monticona sonitarta (P. L. 8. Miiller), Syst. Nat. Suppl. p. 142. no. 46 (1776), ex Buffon, Pl. Enl. 564. f. 2. Le Merle solitaire de Manille, Montb. Hist. Nat. Ois. iii. p. 363. no. 1, descr. orig. ex Sonnerat; Pl. Enl. 636, 3; 564, f. 2, ¢ vel 9 adolesce. Turdus manilla, Boddaert, Tab. Pl. Enl. 636 (1783). Merula solitaria philippensis, Briss. Orn. ii. p. 272, no. 32, “Ins. Philipp.” descr. orig. ex Poivre (avis juv ?). Le Merle solitaire des Philippines, Montb. op. cit. p. 364, no. 2; Pl. Enl. 339, ex Brisson, no. 321. Turdus philippensis, P. Li. S. Miller, op. cit. p. 145. no. 59, ex Buffon, Pl. Enl. 339 (1776). Turdus philippensis, Bodd. op. cit., ex Buffon, Pl. Enl. 339 (1783). Turdus eremita, Gm. Syst. Nat. 13th ed. i. p. 833 (1788), ex Brisson, no. 32. Merula solitaria manillensis, Brisson, op. cit. p. 270. no. 31, “ Manilla,” descr. orig. Turdus manillensis, Gm. op. cit. p. 833 (1788), ex Brisson, no. 31. Turdus manillensis, auct.; Schlegel, Faun. Jap. Aves, p. 67. Hab. North Celebes (Forsten) ; Philippines (type); China, Formosa, Japan (Swinhoe). There seems little doubt that the Merle solitaire de Manille and the Merle solitaire des Philippines of Montbeillard are the same species in different phases of plumage. This was Montbeillard’s own opinion (op. cit. p. 365). The most recent authors, how- ever, have continued to treat them as distinct. Pratincoia, Koch. 72. PRATINCOLA CAPRATA (Linn.), Syst. Nat. ed. 12, i. p. 335. no. 33, “Luzon” (1766), ex Brisson, Orn. iii. p. 440. Hab. Macassar (Wallace) ; Philippines (mus. nostr.) ; common all over India (Jerdon) ; Tongoo (mus. nostr.); Aracan (Blyth); Java (Horsfield); Nipaul (Hodgson); Moul- mein, Lombock, Timor, Flores (mus. nostr.); Simla (Beavan); Coorg, Candeish (mus. nostr.). An example of a young male individual of this species was collected by Mr. Wallace at Macassar. It in no way differs from Philippine specimens in my collection. Examples from the localities above cited agree well in their dimensions. Those from Candeish are larger, but not so large as the Ceylon P. atrata, Blyth. * Montbeillard’s account contains internal evidence sufficient to prove that it was compiled from Brisson’s description. The plate (339) appears also to have been drawn from Brisson’s description only. This will explain the unnatural colouring of the head. 64 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. SYLVIIDZ. AcROcEPHALUS, Naumann. 73. ACROCEPHALUS ORIENTALIS (Bp.), Consp. i. p. 285 (1850), ex Schlegel. Salicaria turdina orientalis, Schlegel, Faun. Jap. Aves, p. 50, pl. 21, “ Japan.” Acrocephalus magnirostris, Swinh. Ibis, 1860, p. 51, “ Amoy, Shanghai.” Hab. Celebes (Schlegel) ; Menado (mus. nostr.); Japan (Schlegel) ; China (Swinhoe). Two examples of a large Reed-Warbler from Menado agree best with Amoy indivi- viduals. I therefore provisionally refer them to the Chinese species. ‘They, however, differ from my examples of A. orientalis (Amoy), A. brunnescens (Coorg), and A. arun- dinaceus (Linn.) (Holland), in having the rectrices conspicuously tipped with dirty white. The proportion of the quills in these examples does not exactly coincide with the proportions existing in the other species alluded to; nor do the dimensions com- pletely agree. Bill. Wing. Tail. Tarsus. inch. inches. inches. inch. A. arundinaceus (Linn.) .......-.. 4375 3°5625 3°250 1-0000 A. brunnescens (Jerd.) ..........+- 5625 3:4375 3-250 1-1250 A. orientalis (Bp.) ..........<..--- 5625 3°2500 3:000 1:1250 Acrocephalus, sp., ex Cashmere...... *6250 3:4375 3°500 1:1875 Bs a Menado ...... “5625 3:4375 3-000 1-1250 A. arundinaceus (Linn.). First long primary nearly as long as second, which is longest ; third shorter than first. A. brunnescens (Jerd.). First much shorter than third and fourth, which are longest. In one example the third is longest; in another the fourth is longest. A. orientalis (Bp.). Second longest, third nearly equal to second, first equal to fourth. Ex Menado. Second longest, first nearly equal to third, first longer than fourth. Ex Cashmere. Second equal to fourth, third longest; first somewhat shorter than second and fourth, which are nearly equal to third. The Cashmere example seems to belong to a distinct species, and differs from A. brunnescens of Southern India in its longer and stouter bill, longer tail, and in the upper plumage being darker brown. CisticoLa, Kaup. 74. CISTICOLA CURSITANS (Franklin), P. Z. S. 1831, p. 118. Sylvia cisticola, Temm. Man. d’Orn. i. p. 228 (1820). Cisticola schenicola, Bp. Birds of Europe, p. 12 (1838). Hab. Macassar (Wallace). For complete range cf. Von Heuglin, Orn. N.-O. Afr. pp- 269, 270. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 65 A Macassar example of a male Cisticola, kindly lent to me by Mr. Wallace, I am unable to distinguish from Assamese and Daccan individuals of C. cursitans. It is labelled C. lineocapilla, Gould, with the note, “tail rather more distinctly marked.” Wing 1%, tail 143. The range of this tiny species is very extensive. MOTACILLIDA. Bupytes, Cuvier. 75. BupytEs viripis (Gm.), Syst. Nat. ed. 13, i. p. 962 (1788), “Ceylon,” ex Brown, pl. 33. Hab. Menado (mus. nostr.). One example, in winter plumage. Olive-green above. Upper part of breast sulphur- yellow; rest of under surface pure white, some of the ventral and under tail-coverts dashed with sulphur-yellow. Supercilium conspicuous, broad, and pure white. Agrees perfectly with examples from continental India. Motacilla flavescens, Stephens, Gen. Zool. Aves, x. p. 559, is enumerated in the ‘Hand-list’ by Mr. G. R. Gray as a distinct species, with the habitats of the Moluccas, Celebes, Timor, and Java assigned. Stephens gave this title to Buffon’s “‘ Bergeronnette de Vile de Timor,” Hist. Nat. v. p.275. Buffon’s bird belongs to that phase of plumage of B. viridis (Gm.) in which the superciliary stripe is yellow, the upper plumage ash- coloured, and the under yellow. HIRUNDINID#. Hirvnpo, Linneus. 76. HiruNDO GUTTURALIS, Scopoli, Del. Fl. Faun. Insubr. ii. p. 96. no. 115, ex Sonn. (1786). LT? Hirondelle d’ Antigue, Sonn. Voy. Nouy. Guin. p. 118, pl. 78. Hirundo panayana, Gm. Syst. Nat. i. p. 1018, ex Sonn. (1788). Hab. Menado (mus. nostr.) ; Indian region. Celebean examples agree with specimens from India, Japan, China, Java, Malacca, and Morty Island. In one the crown is ashy brown, the forehead albescent. The black pectoral band is present, and the chin and throat are dirty rufous; on the outer tail-feathers the white mark is in the form of a diagonal oval drop. An example of an adult bird has the head steel-blue; forehead, chin, and throat deep rufous, as in the European H. rustica, the rufous breast being bounded by the usual black pectoral band. Wing 43 inches. Whether this and the other races of Chimney-Swallows which inhabit the Malay archipelago and Eastern Asia are or are not of the same species as the European bird, they undoubtedly belong to Sonnerat’s Hirondelle d Antique. VOL. VIM.—PART Il. May, 1872. M 66 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 77. Hirunpo savanica, Sparrman, Mus. Carls. pl. 100, “ Java” (1789). Hab. Indo-Malayan region. Mr. Wallace informs me that he found this species common at Macassar, “ building its mud nests in verandas in the town.” MUSCICAPID. Cyornis, Blyth. 78. CyorNIS RUFIGULA, Wallace, P. Z. S. 1865, p. 476, “ Menado.” (PI. VII. fig. 3.) Hab. Menado (Wallace). MyrALestes, Reichenbach. 79. MYIALESTES HELIANTHEA [(Wallace), P. Z. S. 1865, p. 476, “ Menado. (Pl. VII. fig. 1.) Hab. Menado (Wallace). This is a representative form of MW. cinereocapitlla (Vieill.), differing from that species by wanting the ashy head, nape, throat, and breast of the Indian bird. The head is subcrested. Hyportuymis, Boie. 80. Hypornymis PUELLA (Wallace), P. Z. S. 1862, p. 340, “Sula Islands and Celebes.” (Bo Va sic 2) Hab. Sula Islands and Celebes (Wallace). The azure Flycatchers form a natural section consisting of several very closely allied species, which have yet to be worked out. The group is characteristic of the Indian as distinguished from the Australian region; and Boie’s generic title is here adopted in preference to classing MZ. azwrea, Bodd., and its allies with the Australian Myiagra rubeculoides, Vig. & Horsf., and its allied species. 81. HyporHyMis MANADENSIS (Quoy et Gaimard), Voy. Astrol. Zool. i. p. 176, “* Menado” (1830), pl. 3. fig. 3. Hab. Menado (Quoy et Gaim.). Prince Bonaparte (Coll. Delattre, p. 81) refers this form to Hypothymis, where I place it with doubt, being unacquainted with the species. Butalis hypogrammica, Wallace, Ibis, 1862, p. 350, is recorded from Celebes by Mr. G. R. Gray (Hand-list, no. 4814), Mr. Wallace cannot assure me positively that it occurs in that island. But as it is a summer visitant in China, and was obtained in Ceram and Morty Island by Mr. Wallace, it is not unlikely to be a winter resident in Celebes. Hemichelidon griseosticta, Swinhoe, is undoubtedly the same species; and that title takes precedence (Ibis, 1861, p. 330). VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 67 ARTAMIDE. ARtTAmvs, Vieillot. 82. ARTAMUS MONACHUS, Temm.; Bp. Conspectus, i. p. 343, “ Celebes” (1850); Wallace, P. Z. S. 1862, p. 340; Ibis, 1860, p. 141. (PI. VI. fig. 1.) Hab. Mountain districts of North Celebes, as well as the Sula Islands (Wallace). The diagnosis by Mr. Wallace was taken from Sula examples. Neither does it, nor do Sula individuals (mus. nostr.) altogether agree with the description given by Prince Bonaparte (J. ¢.). 83. ARTAMUS LEUCORHYNCHUS (Linn.): Mantissa Plant. p. 524, ex Brisson, “ Manilla” Gen): Lanius manillensis, Briss. Ornith. ii. p. 180, “ Manilla,” descr. orig. (1760). : leucorhynchus, Gm. S. N. ed. 18, i. p. 305, ex Brisson (1788). dominicanus, Gm. op. cit. p. 307, ex Sonnerat, Voy. Nouv. Guin. p. 55, pl. 25. ? Lanius leucogaster, Valenc. Mém. du Mus. vi. p. 27, “Timor” (1820). Artamus leucopygialis, Gould, P. Z. S. 1842, p. 17, “Australia.” papuensis, Temm.; Bp. Consp. i. p. 344, “ Nov. Guinea, Timor.” Leptopteryx leucorhynchus (Linn.) ; Horsf. Linn, Trans. xiii. p. 244, “ Java.” Lanius leucorhynchus, Linn.; Raffles, Linn. Trans. xiii. p. 806, “ Sumatra.” Artamus leucogaster (Valenc.) ; Wallace, P. Z. S. 1868, p. 28; Walden, P. Z. S. 1866, p. 555; Beavan, Ibis, 1867, p. 324. leucorhynchus, Wallace, Ibis, 1860, p. 141. Hab. Timor, Flores, Lombock, Bouru, and the whole archipelago from Sumatra to New Guinea, Celebes (Wallace); Sumatra (Raffles); Java (Horsfield); Andamans (Beavan) ; Cape York, Moreton Bay, Queensland, Mysol, Menado, Manilla, Andamans, Java (mus. nostr.). I am unable to distinguish individuals of the white-bellied Swallow-shrike inhabiting the Philippines, Andamans, Java, Lombock, Mysol, Australia, and Celebes. In colora- tion they appear to be absolutely identical. In dimension, with the exception of the large Celebean form, they vary but little. I have therefore included all under the oldest title given by Linneus to the Philippine bird. The Celebean is much the largest, and ought, perhaps, to receive a separate specific name. Mr. Wallace (P. Z. 8. 1863, p. 485) entitles the Timor bird A. leucogaster, var.; but it was from Timor specimens that Valenciennes described A. leucogaster. A. mentalis, Jard. (Fidjee Islands), and A. melaleucus (Forsten), a good species (New Caledonia), belong to this group. A. monachus (Temm.) ought, perhaps, to be also included, M2 68 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Dimensions of Artamus leucorhynchus. x : Bill ill Wing. Tail. Tarsus. from forehead. peat inches. inches. inch. inch. inch. Manillapeencirecicicee cee 5:250 2-625 *625 *7500 5000 ombocksereeem ee nce 5-250 2-750 “625 +8125 5625 Miysolics tes xn cee ee ....| 5°250 2-625 *625 +7500 "5625 Queensland............. 5:375 2-625 “625 “6875 -5625 Cape Yorker atcha erie tea newer ean 5-250 2:625 *625 *6875 5000 IMoretoneBaygerer cir crnceuanicrere 5125 2-750 *625 “6875 -5000 Va Venie Ravccts loreixvecereuvts ctekes assis cetizcassraga cies coe 5:250 2-500 *625 +7500 *5625 AM GAM ANE! frevanve tara atess avsheilecsuciacatsvoranere: 6 5125 2-500 5625 “6875 -5000 Men ado? tna) actonte crate aus boloittessesteteaar 5-625 3°000 625 +8750 +5625 CAMPEPHAGID/. GRAUCALUS, Cuvier. 84. Gravcawus aTriceps, 8. Miill. Verhand. Land- en Volk. p. 190, “Celebes ” (1839-44); Hartlaub, J. fiir Orn. 1864, p. 437. Hab. Celebes (S. Miiller) ; Ceram, Sumbawa, Flores (Hartlaub). In his admirable monograph, Dr. Hartlaub (/. c.) describes from a Ceram male and a Sumbawa female. It is not stated whether they were compared with Celebean individuals, I therefore include these localities with some doubt. Mr. Wallace (P. Z.S. 1863, p. 485) notes only one Graucalus as inhabiting Flores, G. personatus, 8. Miller. 85. GRAUCALUS LEUCOPYGIUS, Bp. Consp. i. p. 354, “ Celebes” (1850); Hartlaub, J. fiir Orn. 1864, p. 443. Hab. Macassar (Hartlaub); Macassar, Menado (mus. nostr.). 86. GRAUCALUS TEMMINCKII, S. Miiller, Verhandel. Land- en Volkenk. p. 191, “ North- eastern Celebes”’ (1839-44); Hartlaub, J. fiir Orn. 1864, p. 446. Hab. Gorontalo (Forsten, fide Hartlaub). This is a most remarkable form, and seems to be rare. In 1864 only one example was contained in the Leyden Museum. Another, a male, is preserved in the British Museum. The types (for S. Miiller also described the female, /. c.) were obtained by Forsten in North-eastern Celebes. Mr. Wallace (P. Z. S. 1862, p. 342) has added the Sula Islands to its range, on S. Miiller’s authority. I have failed in finding any state- ment of 8. Miiller to that effect. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 69 Votyocivora, Hodgson. 87. Vo.vocrvora Mortio (S. Miiller), Verhandel. Land- en Volkenk. p. 189, “ Celebes” (1839-44); Hartlaub, J. fiir Orn. 1865, p. 155. (Pl. VIII. fig. 1.) Edboliosoma melanolema, G. R. Gray, Hand-list, no. 5099, “ Celebes” (1869). Hab. Tondano, Gorontalo (Hartlaub); Macassar (mus. nostr.). This and several Indo-Malayan and Papuan species are classed by Dr. Hartlaub (/. c.) under Campephaga, Vieillot, the type of that genus being the African Campephaga nigra, Vieill. This species, in its turn, Dr. Hartlaub transfers to Lesson’s genus Lanicterus. I venture, however, to refer the Celebean bird to Volvocivora, Hodgs., as it is nearly allied to the type of that genus, Lanius silens, Tickell (1833), = Ceblepyris lugubris, Sundey. (1837), = Volvocivora melaschistos, Hodgs. (1837). On examination I find that Edoliosoma melanolema, a title published without descrip- tion, refers to S. Miiller’s species; while the EZ. morio, of the Hand-list, no. 5097, appears to be C. fimbriatus, Temm. The British-Museum examples of the last are noted from Celebes ; but that locality requires further confirmation. La.aGE, Boie. 88. LALAGE LEUCOPYGIALIS, n.s. (Pl. VIII. fig. 2.) Hab. Menado (mus. nostr.). S. Miiller, Hartlaub, O. Finsch, and others have hitherto included Celebes within the range of the Lalage of Java, Turdus dominicus, P. L. 8. Miiller, =T. terat, Bodd., =T. orientalis, Gm. Two examples of a Lalage, one of an adult male, and the other of an adult female, received by me from Menado, are to be readily distinguished from the Javan bird by having the lower back and rump pure white, the long upper tail- coverts only being grey. In this respect the Celebean Lalage agrees with L. mela- noleuca (Blyth) from the Philippines; but that species is without a white supercilium (fide Hartl. J. fiir Orn. 1865, p. 163). This is probably the Z. leucopygialis of Mr. Gray’s Hand-list ; but as no description is given, his title cannot be noticed. The Lalage which inhabits South-eastern Borneo differs from the Javan form in its longer wing and broader though not longer bill. Bill from nostril. inches. | inches. inch. inch, Java. ¢ adult ................400. 3:375| 3-000| +7500 500 L. dominica. Banjarmassing. g¢ adult............ 3-750] 3-250] -*7500 500 ” ” Menado. ¢ adult...............005 3:500] 3:125| -6875 500 L. leucopygialis. Menado. 9 adult...............5.. 3:500| 3:125| +6875 500 ” ” 70 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Lalage aurea (Temm.), Pl. Col. livr. Ixiv. pl. 582. fig. 2, “Timor” (21st December 1825); Bp. Consp. i. p. 355; Coll. Delattre, p. 78; Hartlaub, J. fiir Orn. 1865, p. 168. This species is stated to occur in Celebes by Bonaparte and Dr. Hartlaub. Mr. Wallace, however, is of opinion that it cannot be considered a Celebean bird. ARTAMIDES, Hartlaub. 89. ArTAMIDES BIcoLoR (Temm.), Pl. Col. livr. xlvii. pl. 278, “ Sumatra” (!) (June 26, 1824); S. Miller, Verh. Land- en Volk. p. 191, “Celebes;” G. R. Gray, Hand- list, no. 5091, “ Banda, Celebes, Sumatra.” Hab. Menado (mus. nostr.). Salomon Miiller (7. c.) expressly states that this species inhabits Celebes, and neither Sumatra nor Banda. DICRURIDZ. Dicrurtvs, Vieillot. 90. Dicrurus Levcops, Wallace, P. Z. S. 1865, p. 478, “‘ Celebes.” Hab. Macassar, Menado (Wallace). Closely resembles D. pectoralis, Wallace, of the Sula Islands, but is somewhat larger in all its dimensions. The irides are stated by Mr. Wallace to be invariably milk- white, while in the Sula species and in all others known they are red. The type of Vieillot’s genus Dicrurus, Corvus balicassius, L., appears to stand alone ; and it will be perhaps necessary to form a separate genus for the reception of all the Austro-Malayan Dicruride, whose affinities seem to be with Chibia, Hodgs. NECTARINIIDA. NECTARININA. ARACHNOTHERA, Temminck. 91. ARACHNOTHERA q Arachnothera longirostra (Lath.), S. Muller, Verhandel. Zool. Aves, p. 69, ex Celebes. I have not had an opportunity of examining an example of the Celebean Arachno- thera, and am therefore unable to determine its correct title. ANTHREPTES, Swainson. 92. ANTHREPTES MALACCENSIS (Scopoli), Del. Fl. et Faun. Insub. ii. p. 90. no. 62 (1786); Walden, Ibis, 1870, p. 47. no. 38. Hab. Celebes (Wallace); Menado (mus. nostr.); Java, Sumatra, Borneo, Malacca VISCOUNT WALDEN ON THE BIRDS OF CELEBES, 71 (S. Miiller); Sula Islands, Flores (Wallace); Aracan, Tenasserim (Blyth); Labuan (Motley & Dillwyn); Banjermassing (Sclater) ; Siam (Gould); Cambodia (Walden). This must be a common species in the neighbourhood of Menado, judging from the number of examples sent from that locality. CHALCOSTETHA, Cabanis. 93. CHALCOSTETHA PORPHYROL&MA (Wallace), P. Z. S. 1865, p. 479, “ Macassar ;” Walden, Ibis, 1870, p. 46. no. 35. Nectarinia aspasia (Less.), Schlegel & S. Miiller, Verhand. Zool. Aves, p. 58, “ Macassar” (?). Hab. Macassar (Wallace). ARACHNECHTHRA, Cabanis. 94, ARACHNECHTHRA FRENATA (S. Miiller), Verhandl. p. 173, “New Guinea, Menado” (1843); op. cit. Zool. Aves, p. 61, pl. 8. f. 1; Walden, Ibis, 1870, p. 26. no. 6. Hab. Celebes, Sula Islands, Mysol, Moluccas, Kaisa Island (Wallace); Batchian, Ternate, Aru Islands, New Guinea, Islands of Torres Straits (@. &. Gray); North-east coast of Australia (J. Macgillivray). A. flavigastra (Gould), ex New Ireland, is closely related to this species. The male, as described by Lesson and Garnot (Voy. Coq. Zool. i. p. 344, note), is undistinguishable. NEcTAROPHILA, Reichenbach. 95. NECTAROPHILA GRAYI (Wallace), P. Z. S. 1865, p. 479, “ Menado;” Walden, Ibis, 1870, p. 42. no. 30, pl. 1. f. 2. Hab. Menado (Wallace). /MTHOPYGA, Cabanis. 96. AETHOPYGA FLAVOSTRIATA (Wallace), P. Z. S. 1865, p. 478, pl. 29. £2; Walden, Ibis, 1870, p. 35. no. 18; Wallace, Ibis, 1860, p. 141. Hab. Menado (Wallace). In the Proceedings of the Zoological Society (J. c.) Mr. Wallace states Menado to be the habitat of this species; but elsewhere (Ibis, /. c.) that gentleman states that he obtained this Sun-bird in a forest district beyond the Lake of Tondano, at an elevation of about 1500 feet. A sixth species of Nectarinia appears to inhabit Celebes (conf. Walden, Ibis, 1870, p- 42. no. 30). 72 VISCOUNT WALDEN ON THE BIRDS OF CELEBES, DicaINA. Diczum, Cuvier. 97. Dicaum cetesicum, S. Miiller, Verhandel. p. 162, ‘Celebes;” Wallace, P. Z.S. 1862, p. 324. Diceum leclancherii, Lafr. Rev. Zool. 1845, p. 94, “ Menado ;” op. cit. 1846, p. 42; Hartl. op. cit. 1846, pp. 4, 47, 111. Hab. Celebes, Sula Islands (Wallace). PRIONOCHILUS, Strickland. 98. PRIONOCHILUS AUREOLIMBATUS, Wallace, P. Z.S. 1865, p. 477, pl. 29. f. 1, “North Celebes;” Salvadori, Atti Accad. Scien. Torino, 1868, p. 420. Parus ——?, Wallace, Ibis, 1860, p. 141. Hab. Mountains of Minahasa (Wallace). The female scarcely differs from the male. MELIPHAGID. ZOSTEROPS, Vigors. 99. ZosTEROPS INTERMEDIA, Wallace, P. Z. S. 1863, p. 493, ““Macassar and Lombock ;” Hartl. J. fiir Orn. 1865, p. 16. (Pl. IX. fig. 2.) Hab. Macassar, Lombock (Wallace). The above specific title was attached to a Macassar example in the British Museum by Mr. G. R. Gray, and was adopted by Mr. Wallace, who first discovered and first described the species. 100. ZostErors atrirrons, Wallace, P. Z. S. 1863, p. 493, ““Menado.” (Pl. IX. fig. 3.) Zosterops nigrifrons, Temm. Mus. Lugd.; Hartl. J. fiir Orn. 1865, p. 22. Hab. Menado (Wallace); Gorontalo (Mus. Lugd.). PLOCEID. Pappa, Reichenbach. 101. Pappa oryzivora (Linn.), Ameen. Acad. iv. p. 243. no. 16 (1759), ex Edwards, pl. 41. Hab. Macassar (Wallace); Java (Horsf.); Sumatra (Raffles); Malacca (Cantor) ; Lombock (Wallace); Banjarmassing (Sclater); South China (Swinhoe); Manilla (Von Martens). Mr. Wallace informs me that this species is abundant near the town of Macassar. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 73 Monta, Hodgson. 102. Monta nisoria (Temm.), Pl. Col. 500. f. 2, “Java” (8 May, 1830); conf. Blyth, Ibis, 1870, p. 172; Walden, Ibis, 1869, p. 211, note. Hab. Macassar (Wallace); Java (mus. nostr.). A single Celebean example in Mr. Wallace’s collection, the only individual I have been able to examine, agrees well with Javan specimens. The upper tail-coverts and edges of the rectrices, however, are olive-green, and not grey as is the case in all my Javan examples. Mr. Blyth (J. c.) observes that the Celebean race has no pale shafts to the feathers of the upper parts; but in this Macassar individual the pale shafts are very conspicuous. The two principal characters which distinguish the Javan WV. nisoria (T.) from the Indian VW. punctularia (L.), are the rufous colouring of the breast-markings and the grey colour of the upper tail-coverts and edges of rectrices. In the Indian bird these are golden yellow, and the breast-markings are almost black. Moulmein indivi- duals, again, differ from those of India in having the breast-markings rufous, and from both Javan and Indian in having the upper tail-coverts and edges of the rectrices yellowish green; nor are the breast-markings in the Moulmein race as well defined. In the race which inhabits Flores the upper tail-coverts are pale olive-green, as in the Celebean bird. M. punctularia and M. nisoria, in young plumage, before the breast-markings appear and the upper coverts assume the waxy lustre found in the adult, are extremely difficult to distinguish. The Indian bird, however, is considerably larger, and has the bill much stouter. From I. rubro-nigra and its allies, when in first plumage, they are likewise difficult to separate. The only sure characters are the sinuated commissure and massive form of the bill in I. rubro-nigra. 103. Munta motucca (Linn.), Syst. Nat. ed. 12, i. p. 302 (1766), ex Brisson, Orn. iii. p- 241. no. 10; Wallace, Ibis, 1860, p. 147. Hab. Macassar (Wallace); Flores (mus. nostr.). A Celebean example of an adult male collected by Mr. Wallace perfectly agrees with Brisson’s description of Count Bentick’s specimen obtained in the Moluccas, on which Linnzus bestowed the above specific title. 104. MunIA BRUNNEICEPS, n.sp. (Pl. IX. fig. 1.) Conf. Blyth, Ibis, 1870, p. 171. Hab. Macassar (Wallace). Head, chin, throat, and breast brown; abdominal stripe, vent, and under tail-coverts black; remainder of plumage dark chestnut. From a Macassar example of a male collected by Mr. Wallace. In another example from the same locality, marked a female, the head and nape are of a lighter and less decided shade of brown. Wing 2 inches. VOL. VilI.— Part u. May, 1872. N 74 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Were it not that Mr. Blyth had already remarked the imbrowned colouring of the head and neck in examples from Celebes, contained in the Leyden Museum, I should have felt less confidence in considering these Macassar individuals distinct from VW. rubro- nigra, Hodgs. CORVID. Corvus, Linnzeus. 105. Corvus eNnca (Horsf.), Trans. Linn. Soc. xiii. p. 164, “Java” (1822); Schlegel, Bijdrag. part vill. p. 13, pl. 1. fig. 23; Mus. Pays-Bas, Coraces, p. 29. Corvus validus, var., Wallace, partim, P. Z. 8. 1862, p. 343. Hab. Macassar (Bernstein) ; Limbotto, Gorontalo, Kema, Toulabello (Von Rosenberg) ; Java (type). The species inhabiting Celebes has not been satisfactorily identified. By Professor Schlegel it is considered the same as that found in Java, while true C. validus, Temm., Bp. (Consp. i. p. 385), is from Sumatra, and does not occur in Java. GazzoLa, Bonaparte. 106. GazzoLa typica, Bp. Compt. Rend. xxxvii. p. 828, “ Nouvelle Calédonie,” errore (5th December 1853); Notes Orn. Coll. Delattre, p. 6 (1854); Sclater, Ibis, 1859, p. 113. Corvus advena, Schlegel, Bijdragen tot de Dierk. pt. viii. p. 3, pl. 2, “Sumatra,” errore (1859) ; Mus. Pays-Bas, Coraces, p. 6, “ Macassar ;” Wallace, Mal. Archip. i. p. 375. Hab. Macassar (Wallace). This species has hitherto been found only in the Macassar district. Mr. Wallace (J. c.) alludes to it as rare. It is an anomalous form, hardly exceeding a Lycos in size, but with a bill equal to that of Corvus corone, and of much the same character. The arrangement of the quills is peculiar. The fourth much exceeds the others; and the first is very short. Prince Bonaparte separated it generically, but placed it next to Corvus (Physocorax) moneduloides (Less.), another unique and aberrant Corvine form, with which it has nothing in common beyond its general family relations. This species has partly been the subject of some of the most curious mistakes in ornithological literature; and the position of the generic title Gazzola, Bp., whether among the Campephagide or the Corvide, depends on a correct history and explanation of how the confusion arose. In the thirteenth edition of the ‘Systema,’ Gmelin gave the title of Corvus caledonicus to Latham’s “ New-Caledonian Crow,” a species described by Latham (General Syn. i. p. 377) from a drawing belonging to Sir Joseph Banks. This is a true New-Caledonian Graucalus. In the second supplement to the ‘ Synopsis, Latham inserted a distinct bird (Labillardiére’s “Pie de la Nouvelle Calédonie”) under the title of “Caledonian Crow,” and called it in the supplement to the ‘ Index VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 7) Omithologicus’ Corvus caledonicus. Thus there became a Corvus caledonicus, Gm., and a Corvus caledonicus, Lath., the first being a Graucalus, the last a Streptocitta, the first being a really New-Caledonian species, the last being only found in Celebes. In 1850 Bonaparte founded his genus Gazzola, making C. caledonicus, Gm., the type, and associating with it the correct synonyms of true C. caledonicus, Gm. Still it is evident that Bonaparte was confounding the then unique specimen in the Paris Museum of the Celebean black-and-white Crow (which was labelled.“ Corvus dauricus de la Nou- velle Calédonie”) with Corvus caledonicus, Lath., the black-and-white Streptocitta ; for the Prince would never have identified a true Graucalus with either a Pica or a Corvus, and he made Gazzola the connecting link between the Garrulide and the Corvide. Thus the elements of confusion were these :—one Corvus caledonicus, Gm.; two species under that title in Latham, one of them being described as black and white; a black- and-white Corvus in the Paris Museum labelled “C. dauricus de la Nouvelle Calédonie,’— only one of the three species being a New-Caledonian bird. Three years later Bona- parte partly cleared up the confusion. He (Notes Ornith. /.c.) changed the title from Gazzola caledonica (Gm.) to that of Gazzola typica, Bp., on the ground that the type of his genus Gazzola was neither of the “deux C. caledonicus, de Latham,”! nor that of Labillardiére, nor that of Gmelin. The question now arises whether Corvus caledonicus, Gm., ought to be considered the type of the genus Gazzola. It has been so treated by Mr. G. R. Gray (Hand-list, no. 1246). But as the Prince has described the species he founded the genus on, I have thought it best to retain Gazzola for that species, which is the same as Corvus advena, Schlegel. MM. Verreaux and O. des Murs (Rev. & Mag. Zool. 1860, p. 432) included Gazzola typica, Bp., in their list of New-Caledonian birds, trusting, in all probability, to the erroneous locality on the label of the Paris-Museum specimen. Srreprocitra, Bonaparte. 107. SrrEprocirra cALEDONICA (Lath.), Ind. Orn. Suppl. p. xxv. no. 3 (1801), ex Voy. Entrecasteaux, ii. p. 226, pl. 35 (39 2), “‘ New Caledonia,” errore. Pie of New Caledonia, Labillarditre, Voy. Entrecasteaux, Eng. Tr. (Stockdale), ii. p. 227, pl. 39; G. R. Gray, Cat. Birds Trop. Isl. p. 25. Pica albicollis, Vieill. N. D. Sc. Nat. xxvi. p. 128, ex Voy. Entrecast. pl. 39. Streptocitta caledonica, Bp. Consp. i. p. 382. albicollis, Sclater, Ibis, 1859, p. 113; Wallace, Malay Archip. i. p. 430. Hab. Macassar (Wallace, fide Sclater ; Mus. Brit.). Although Labillardiére (/. c.) tells us, very circumstantially, the date and the occasion when and where he obtained his Pie de la Nouvelle Calédonie, Mr. Sclater’s explanation 1 This is a good illustration of the confusion that may be created by not quoting the names of the original authors, or by replacing them with the names of subsequent authors, who may have quoted or misquoted. n2 76 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. (l. c.) of the probable cause of the error is most likely correct. Yet it must be borne in mind that Labillardiére never set foot on the island of Celebes proper; nor does Entrecasteaux’s expedition appear to have had any direct communication with that island on either of the occasions of its presence in the Moluccas. On its way from Bouru to Sourabaya, in October 1793, the expedition, after failing in its attempt to pass the Straits of Tioro, occupied several days in passing those of Boeton, and remained a day at the town of Boeton itself. During this period the French naturalist made several excursions on shore, and, as he particularly mentions, in the island of Pangasane, and one, of two hours’ duration, in the neighbourhood of the town of Boeton. It is most probable therefore that this form of Streptocittg was obtained either on the island of Pangasane or of Boeton; for the expedition did not touch the mainland of Celebes, nor at the island of Saleyer when passing the straits of that name. I identify the species which inhabits the district of Macassar with Labillardieére’s bird, because it best agrees with his short description. By him the bill is stated to be “ofa light black from the root to within one third of the point, the remainder is yellowish.” This and the green hue of the black portion of the plumage easily distinguish the South from the North Celebean species. The bill is also more slender than that of the following species. 108. Srreprocirra Torquata (Temm.), Nouv. Rec. 75°™° livr. Pl. Col. 444, “ Celebes” (Jan. 5, 1828); G. R. Gray, Birds Trop. Isl. p. 25. Hab. Menado (Mus. nostr.). I quite agree with Mr. G. R. Gray in regarding this form as specifically distinct from the true S. caledonica, from which it differs by its strong, jet-black bill, and by having the black portions of its plumage glossed with dark blue. Mr. G. R. Gray (J. c.) states that the actual individual from which Temminck’s figure was drawn is in the British Museum. In this species the first quill is barely one inch long; the fourth and fifth are nearly equal, the fifth being slightly the longest; the third is somewhat shorter than the fourth ; the second still a little shorter than the third. The wing measures 53 inches. The second pair of rectrices exceeds the first by § of an inch; the third the second by 1;; the fourth the third by 13; the fifth the fourth by 14; and the sixth, or middle pair, the fifth by 2$; the total length of the middle pair is 114; bill from nostril § of an inch; tarsus 12. Temminck’s surmise that this species occurs in Borneo has not been, as yet, realized. Professor Schlegel has generically separated his Charitornis albertine from Strepto- citta; but it is difficult to seize the characters wherein it generically differs. The structure of the wings, tail, and feet is identical. The colouring of the plumage is congeneric. The nostrils are similar in form and position. The bill differs in being VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 77 more arched and stouter, but it does not differ in form from that of S. torquata so much as the bill of §. torguata does from that of S. caledonica. In C. albertine, how- ever, the naked spaces, which are confined to the ophthalmic region in the Celebean birds, extend to under the throat. In it also the frontal plumes are not developed and curved back as in the two species of Streptocittw. Indeed the normal condition of the frontal or nasal plumes is the only external character in which Charitornis differs from Streptocitta. It seems more in accordance with the facts to regard the three species as belonging to the same natural genus, with S. caledonica as the connecting link. In the colouring of the plumage ©. albertine only differs from S. caledonica by having the "head white. By the black-and-yellow colouring of the bill, the South-Celebean species occupies an intermediate position between the completely black bill of S. torquata and the completely yellow bill of C. albertine. Mr. Wallace has led us to infer (Malay Archip. i. p. 430) that Charitornis is confined to Celebes; but this is doubtless an error. Professor Schlegel’s types were obtained in the island of Soula Mangouli; and the species has not been recorded from any other locality. I cannot concur with the Leyden Professor in placing Streptocitta among the Gracu- lide ; though a most anomalous form, its nearest affinities seem to be with the Corvide. BasILeornNis, Temminck. 109. BASILEORNIS CELEBENSIS, Temm. (Mus. Lugd.); G. R. Gray, P. Z. S. 1861, p. 184. no. 2, fig. 2; Wallace, Malay Archip. i. p. 430; Ibis, 1861, pl. 9. fig. 2. Basileornis corythaix (Wagler), Bp. Consp. p. 420 (?), nec Wagler ; Sclater, Ibis, 1859, p. 113. Hab. Menado, Macassar (Wallace). Prince Bonaparte’s description is so vague that it is impossible to decide whether he described from the Celebean or the Ceramese bird. AcRIDOTHERES, Vieillot. 110. AcrmporHERES cINEREUS, Miiller (Mus. Lugd.); Bp. Consp. i. p. 420, “ Celebes” (1850). (Plate X. fig. 1.) Hab. Celebes (Mus. Lugd.); Macassar (mus. nostr.). This is a well-marked species, most nearly allied to A. javanicus, Cab. (= Pastor griseus, Horsf., nec Wagl.), but readily distinguishable by the upper and lower plumage being light grey, and not dark iron-grey. All its dimensions are less; and it has the base of the mandible with traces of black, but not as marked as in its other congener, A. fuscus (Wagler), ap. Jerd., of continental Asia. 78 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. SturN1A, Lesson. 111. Srurnia? pyrruogenys (Schlegel), Faun. Jap. Aves, p. 86, “ Japan, Borneo” (1842). Lamprotornis pyrrhopogon, Schlegel, op. cit. pl. 46. Hab. Japan, Borneo (Schlegel); Philippines (Swinhoe, P. Z. S. 1863, p. 302. no. 217) ; Celebes (Salvin). Three examples, agreeing in every respect with Japanese individuals, are contained in Mr. O. Salvin’s collection; and that gentleman tells me that they were all procured in Celebes. Although I have adopted Professor Schlegel’s name, I have little doubt that eventually, after comparison has been made with Philippine examples, it will have to give way to dominicanus, Bodd., =T. dominicanus, Gm. These similar titles were founded on Le Merle dominiquain des Philippines of Montbeillard (Hist. Nat. Ois. iii. p. 396), who described it from a Philippine individual obtained by Sonnerat which was figured by Daubenton (Pl. Enl. 627. f. 2). Gmelin’s title has hitherto most unaccountably been applied to the Sturnus dauricus, Pall. Pastor rujicollis, Wagler, Syst. Nat. Av. p. 92, ex Manilla, is also clearly a synonym of 7. dominicanus, Bodd. & Gm., and not a distinct species as enumerated by Prince Bonaparte. One of Mr. Salvin’s specimens has the chin, tips and outer edges of the quills, the under and upper tail-coverts, and the rectrices deeply tinged with bright rusty red. Traces of this hue appear in other parts of the plumage. This peculiarity in members of this group has been remarked upon by Mr. Swinhoe (P. Z. 8. 1863, p. 302), and is said by him to prevail during the breeding-season. Is the species, therefore, a per- manent resident in Celebes ? In S. pyrrhogenys, Schlegel, and S. dawricus, Pall., the first quill is longest, and the second nearly as long, the third and following quills being much shorter. This indi- cates an affinity of these two species to true Sturnus; and the structure of the tail and the metallic hues of the plumage strengthen the evidences of the relationship. The form of the bill is peculiar, being short and stout, albeit Stwrnine. On the other hand, Oriolus sinensis, Gm., the type of Stwrnia, Lesson, is a true Temenuchus, Cab. ; consequently all the species falling under the latter generic title must be referred to Sturnia, and Temenuchus will have to be suppressed. A distinct subgenus will probably have to be made for S. pyrrhogenys and S. dawricus, and another for the reception of the isolated Cingalese form Pastor senex, Temm., =Sturnia albofrontata, Layard. Enopes, Temminck. 112. Envopes ERyTHRoPHRYS (Temm.), Nouv. Rec. xlv° livr.; Pl. Col. 267, “ Celebes, environs de Menado, et dans Vile Taguatto (1st of May 1824);” Wallace, Ibis, 1860, p. 141. Hab. N.-E. Celebes, confined to the interior mountain districts, never abundant (Wallace). VISCOUNT WALDEN ON THE BIRDS OF CELEBES, 79 Catornis, G. R. Gray. 113. CALORNIS NEGLECTA, 0. Ss. Calornis obscura, var., Wallace, P. Z. S. 1862, p. 343. Hab. Celebes (mus. nostr.); Sula Islands (Wallace). Having carefully compared examples of nearly all the described species of this genus, I have no hesitation in considering the Calornis of Celebes and the Sula Islands distinct. In colouring it most nearly approaches C. chalybea (Horsf.), ex Java, with the allied races from Sumatra, Malacca, Borneo, and Cambodja; but its dimensions are much greater. From C. obscura (Forst.), ex Gilolo and Batchian, it can be readily discriminated by its bright green colouring. From all the members of the C. metallica group it may be known by the total absence of any iridescent colours. It perfectly agrees with examples from the Sula Islands. The individuals on which this species is founded were sent from Menado in a box which contained nothing but Celebean birds. Notwithstanding, therefore, Mr. Wallace’s statement (Mal. Archip. i. p. 431) that the genus does not occur in Celebes, we may, I venture to think, conclude that that island is not an exception to the general rule which prevails in the geographic distribution of Calornis. The following attempt at an analysis of the species belonging to this difficult genus may perhaps assist in clearing up the confusion in which the synonymy of its members is involved. All the species are divisible into two distinct groups:—first, those in which the plumage is uniform green, varying from light to very dark green; secondly, those which have, added to the prevailing green colour of the plumage, metallic reflec- tions of purple and violet. The uniform green species may be further subdivided into light green and dark green; while the metallic-green species are usually also distin- guished by having the middle pair of rectrices much prolonged. In general terms it may be said that the first subdivision embraces all the Indo-Malayan, the second and third all the Australian forms. A. Uniform green plumage. a. Light green. 1. Muscicapa panayensis, Scop., ex Sonn. pl. 73, =TZurdus cantor, Gm., ex Sonn. pl. 73, =Turdus columbinus, Gm., ex Montbeillard, ex Sonn., “‘ Philippines.” 2. Turdus chalybeus, Horsf., =Turdus strigatus, Horsf. (av. juv.), =Lamprotornis cantor, Gm., ap. Temm. PI. Col. 149, “ Java.” 3. Turdus insidiator, Raffles, “‘ Sumatra.” 4. Calornis affinis, A. Hay, ‘‘Tipperah, Arracan, Tenasserim, Nicobars.” It is very questionable whether these last three species are separable. To them 80 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. belong the Malaccan, Sarawak, and Cambodjan races, which are as yet without titles, but exhibit certain differences. 5. Calornis neglecta, nob., “ Celebes, Sula Islands.” Turdus palmarum, Bodd., =Turdus mauritianus, Gm., both titles being founded on Le Merle vert de Vile de France, of Montbeillard; and Pl. Enl. 648. f. 2 belongs to one of these species; but to which, it is now impossible to say. b. Dark green. 6. Lamprotornis obscura, Forst.; Bp. Consp. p. 417, “ Gilolo.” Given also from Batchian by Mr. G. R. Gray. A good species. Dark purplish green. 7. Calornis erassirostris, nob., “ Lombock.” Collected by Mr. Wallace. Very dark green. Bill high and stout as in Aplonis. The locality may not be correct. 8. Calornis mysolensis, G. R. Gray, “Mysol.” Closely allied to, if not the same as, C. obscura. 9. Calornis cantoroides, G. R. Gray. Like C. mysolensis, only that the tail is shorter and nearly square. Considered by Mr. Wallace to be a good species (P. Z. S. 1862, p. 343). 10. Lanius pacificus, Gm., ex Lath., =Calornis kittlitzi, O. Finsch, = Lamprotornis columbinus, ap. Kittlitz (“ Mariannes, Carolines, and Puynipet”’), from the descrip- tion, seems to belong to this subgroup. B. Green with purple and violet reflections. 11. Lamprotornis metallica, Temm. Pl. Col. 226. Described from Timor and Celebes. The type was probably from Amboyna, perhaps from Australia, possibly from Timor. 12. Calornis purpurascens, G. R. Gray, =C. metallica, Temm., ap. Gould, “ North and East Australia.” 13. Calornis amboinensis, G. R. Gray, “Amboyna.” Closely resembles the Australian species, but is smaller. 14. Lamprotornis minor, Mill.; Bp. Consp. 417, “Timor.” Wallace gives it also from Flores and Lombock. A very distinct species. 15. Calornis viridescens, G. R. Gray, “Aru Islands.” Near to C. amboinensis. Also given from Dorey by Mr. G. R. Gray under the inaccurate title of C. virescens (P. Z. S. 1859, p. 158). 16. Calornis gularis, G. R. Gray, ‘‘ Mysol.” Apparently nothing but C. viridescens. 17. Calornis nitida, G. R. Gray, “ New Ireland,” ex Less. & Garnot, Voy. Coq. Zool. i. p- 343. Seems to belong to the green and purple group. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 81 Dimensions. Wing. Tail. Bill. Tarsus. Locality. inches. inches. inch. inch. C. purpurascens, G. RB. Gray.... 6... cece ce eee 4-250 4-750 5625 “8750 | Cape York. C. amboinensis, @. HR. Gray . 0... .2.05 5s sees 3°875 3500 5000 *8125 | Amboyna. Cy sina (WiGiD) 34 cakeuootene anoAasienoTee 3875 2-500 5000 “7500 | Timor. Cachaly bea (Ho7sf-)rrstry « fgets alee 02 4 sete is sere 3°750 2500 5000 -7500 | Java. CFinsidiator Chapa wa. cveraceverterclare ie iar 4-000 2875 5000 *8125 | Malacca. on Eigse eat ate uoariite eet tary create kaG 3°625 2-625 5000 *8125 | Sarawak. 4 PG ete oeernie, Cceae incr en oar 3°750 2-750 +5000 *8125 | Cambodja. Oimeplectac mops raty vel svsiiste Sevan ys) sera eiaivys a. 4125 3-250 5625 *8750 | Celebes. CXerdssirostrisAmob.st ane os esl e amentoe coord 4-000 4:125 -5000 *8750 | Lombock. C. cantoroides, G. R. Gray 0....... 10+ ese 3°750 2-500 5000 *8125 | Mysol. ScisstrostRuM, Lafresnaye. 114. Scisstrosrrum DuBium (Latham), Ind. Orn. Suppl. p. xviii. no. 5 (1801), ex Lath. Syn. Suppl. i. p. 73. no. 11, descr. orig. Scissirostrum pagei, Lafr. Rev. Zool. 1845, p. 98, “ Manado;”? Mag. Zool. 1845, pl. 59; Wallace, Ibis, 1860, p. 141; Malay Archip. i. p. 430. Hab. Scarce at Macassar, plentiful near Menado (Wallace). We owe the identification of this most anomalous form with the Lanius dubius, Lath., to Dr. Hartlaub (Arch. Nat. xiii. 2. p. 57). Notwithstanding Prince Bonaparte’s incredulity (Consp. i. p. 423), a reference to Latham’s original description, taken from a specimen “at Mr. Thompson’s, Little St. Martin’s Lane, London, but without any history of its manners or country annexed,” leaves no doubt of its identity. The sequence and relative proportions of the quills in this species are the same as in Calornis. The structure of the tail is similar to that of Calornis metallica (Temm.). The bill resembles most nearly, in its massiveness and general outline, that of Eulabes Jwanus, Cuy.; but the peculiar position of the nostrils, situated in narrow and deep ascending grooves, is quite unique. The sole existing representative of a subfamily (?) long since extinct, its systematic place seems to between Calornis and Eulabes. COLUMBZ. TRERONIDZ. OsmorreRoN, Bonaparte. 115. OsmorrEron veRNANS (Linn.), Mantissa, p. 526 (1771), ex Briss. Om. i. p. 143, “Philipp. Ins. ;” Gm. Syst. Nat. i. p. 789, ex Linn. ; Wallace, Ibis, 1863, p. 320. Columba viridis, Scop. Del. Fl. Insub. ii. p. 94 (1786), ex Sonnerat, Voy. Nouv. Guin. p. 110, pl. 64, 3, pl. 65,2, “iles de Lucon et D’ Antigue.” VOL. VIII.—PART I. May, 1872. i) 82 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Treron viridis (Scop.), Wallace, Ibis, 1865, p. 374. vernans, Steph., Schlegel, Nederl. Tijdschr. Dierk. i. p. 70. Hab. Philippine Islands (type); Penang, Sumatra, Borneo, Macassar (Wallace) ; Java, Gorontalo, Sumatra, Bangka (Schlege/). The Celebean form is here retained under the title of the Philippine bird, as I have not been able to compare examples from the two localities. But both from Mr. Wallace's and Professor Schlegel’s remarks on the differences existing in examples from the dif- ferent Indo-Malayan islands, it seems probable that the species inhabiting the localities given above will be all found to differ from one another specifically. On the Sumatran, Javan, Bankan, and Celebean birds, Professor Schlegel has bestowed the title of griset- capilla. And yet he distinguishes the Javan and Celebean forms from the Sumatran and Bankan species by remarking that the former has the head and throat dark greyish- green, while the latter has those parts “jolie gris bleuatre.” 116. OsmorrERon’ GrisErcaupa (G. R. Gray), Mus. Brit. Columbe, p. 10, “ patr. incert.” (1856); Wallace, ex Gray, P. Z.S. 1862, p. 344, “Sula Islands, Celebes;” Schlegel, Nederl. Tijdschr. Dierk. 1866, pp. 210, 346 ; Wallace, Ibis, 1863, p. 319. Hab. Celebes, Sula Islands (Wallace). Professor Schlegel (J. c.) is unable to discover any sufficient and constant distinctions between the Javan 7. pulverulenta, Wallace, and this Celebean species. The Sanghir bird, on account of its stouter bill, the learned Professor considers to possess greater claims, but to be very closely allied. The Sula and Javan examples I have had an opportunity of comparing exhibit the differential characters Mr. Wallace has insisted on, and they seem to me sufficient. It would perhaps be convenient to separate the maroon-backed members of Osmotreron under a distinct subgeneric title. LAMPROTRERON, Bonaparte. 117. Lamprotreron rormosa (G. R. Gray), P. Z.S. 1860, p. 360; Wallace, Ibis, 1865, p- 379, “ Celebes.” Hab. Macassar, Menado (Wallace). Closely allied to P. superbus (Temm.), and hardly admitted as distinct by Professor Schlegel. 1 The type of Treron, Vieillot, is C. curvirostra, Gm., ex Lath., a species as yet not satisfactorily identified, and not C. aromatica, Gm., as stated by Mr. G. R. Gray (Gen. and Subgen. no. 1654). To whatever species Latham’s Hooked-billed Pigeon belongs, it is evident from the plate (Syn. ii. pl. 59) that in it the corneous culmen extends to the forehead. Prince Bonaparte (Consp. ii. p. 10) reduced Torta, Hodgs., to a synonym of Treron, but associated C. psittacea and C. aromatica with Toria nipalensis, species not possessing the characters on which Mr. Hodgson founded his genus. ’reron = Toria contains only two species, 7’. nipalensis and 7. nasica ; C. curvirostris belongs to either the one or the other, probably (as already suggested by Mr. Wallace) to 7. nasica, Schlegel. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 83 IorreRoN, Bonaparte. 118. IorRERON MELANOCEPHALA (Forster), Zool. Indica, p. 16, pl. 7, “Java” (1781'). Ptilopus melanocephalus, Schlegel, Neder]. Tijdschr. Dierk. iii. p. 207. Hab. Sava (type); Flores, Sumbawa, Celebes, Sula Islands, Ceram, Sanghir (Schlegel) ; Lombock (Wallace). Professor Schlegel (/. ¢.) has detailed the characters which distinguish the several races of this Pigeon inhabiting the islands of Java, Flores, Celebes, Sula, Ceram, and Sanghir. They undoubtedly should receive distinguishing titles; for until they and analogous forms are separately named, the physical geographer will only find half the truth when studying zoological catalogues. The Celebean bird has the yellow gular patch tinged with orange (conf. Schlegel, /. c.). LrvcoTrERoN, Bonaparte. 119. Levcorreron euLARIS (Quoy et Gaimard), Voy. Astr. Zool. i. p. 247, pl. 29, “*Menado ” (1830). Hab. Menado (Wallace). C. diademata, Temm., C. monacha, Reinwardt, and C. hypogastra, Reinwardt, belonging to the Pétilopodine, were erroneously described by Temminck as inhabiting Celebes (conf. Wallace, Ibis, 1865). Carpopuaca, Selby. 120. CarporpHaca PavLINA, Temm. Mus. Lugd. (Columba enea,?, Temm., Knipp, Pig. i. pl. 4); Bp. Consp. ii. p. 35; Wallace, Ibis, 1865, p. 385; Schlegel, Nederl. Tijdschr. Dierk. iii. p. 200. Hab. Macassar, Menado, Sula Islands (Wallace). A Philippine example in the Leyden Museum is stated by Professor Schlegel (J. c.) to resemble the Celebean bird. But the differential characters it possesses render it likely that the Philippine bird is specifically distinct. ‘The examples in the same collection, said to have been brought from the Mariannes (?), differ but slightly from the Celebean species, according to Professor Schlegel. Both Prince Bonaparte and Mr. Wallace rank this fine Fruit-Pigeon under Ducula, Hodgs. It appears to me to be a typical Carpo- phaga, Selby. 1 T have not been able to refer to the first edition of Pennant’s ‘ Indian Zoology ;’ but if this species is there named, it will have to take Pennant’s title (1769). 02 84 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Ducua, Hodgson. 121. Ducuta rosacea (Temm.), Pl. Col. 578, “Timor” (1835); Wallace, Ibis, 1865, p. 386; Schlegel, Nederl. Tijdschr. Dierk. iii. pp. 201, 345. Hab. Timor (type); Macassar, Flores (Wallace); Tolofoko (northern peninsula of Halmaheira), Little Key Island (Schlegel). The Celebean habitat of this Pigeon rests on the oibee a of Mr. Wallace (. c.). The Gilolo bird discovered by the late Dr. Bernstein is stated by Professor Schlegel (J. c.) not to differ from the type species. Myristicivora, Reichenbach. 122. Mynisticivora LucTuosa (Reinw.), Temm. Pl. Col. livr. xlii. pl. 247 (26th Feb. 1825); Wallace, Ibis, 1865, p. 386. Hab. Menado, Macassar, Sula Islands (Wallace); Menado (mus. nostr.). Professor Schlegel (Nederl. Tijdschr. Dierk. iii. p. 343) mentions the fact that, in this species only, the breast and abdomen are sometimes washed or even spotted with black. Mr. Cassin (United States Exped. p. 266) pointed out that while C. dicolor (Scop.) possesses fourteen rectrices, the North-Australian C. luctuosa (=. spilorrhoa, G. R. Gray) has only twelve. An examination of examples in my collection fully bears out this observation ; for I find that examples of 1. MW. bicolor (Scop.), ex New Guinea, has fourteen rectrices. 2. PA 3 ex Batchian, has fourteen rectrices. 3. M. luctuosa (Reinw.), ex Sula Islands, has fourteen rectrices. 4, zt a ex Menado, has fourteen rectrices. 5. es m3 ex Menado, has twelve rectrices. 6. WM. spilorrhoa, G. R. Gray, ex Port Albany, has twelve rectrices. 7. 553 33 ex Somerset, has twelve rectrices. The Menado example, with only twelve rectrices (no. 5), appears to have originally possessed two more, which have been lost. ZoNa@NAS, Reichenbach. 123. Zona@Nas RADIATA (Quoy et Gaim.), Voy. Astrol. Zool. i. p. 244, pl. 26, “‘ Menado” (1830). Hab. Macassar, Menado (Wallace). HemipHaca, Bonaparte. 124. HEMIPHAGA FORSTENI (Temm.), Knipp. Pig. ii. pl. 47; Bp. Consp. ii. p. 39 ; Wallace, Ibis, 1865, p. 387. Hab. Menado; appears to be confined to the mountainous district of Minahassa (Wallace). VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 85 COLUMBID. - Macropyeta, Swainson. 125. Macropyaia ALBICAPILLA, Temm. Mus. Lugd.; Bp. Consp. ii. p. 57, “Celebes” (1857). Hab. Macassar, Tondano, Sula Islands (Wadlace). 126. MAcRoPYGIA MACASSARIENSIS, Wallace, Ibis, 1865, p. 389. Hab. Macassar (Wallace). Mr. Blyth (Ibis, 1870, p. 173) observes that M. leptogrammica (Temm.) is not from Java, but from Celebes. Its author (Pl. Col. 560) states that it inhabits Java and Sumatra. Mr. Wallace (op. cit. p. 390) restricts its range to west Java, where it is found up to an elevation of 7500 feet. TurAceéNA, Bonaparte. 127. TuRAcm@NA MENADENSIS (Quoy et Gaim.), Voy. Astrol. Zool. i. p. 248, pl. 30, “Menado ” (1830). Hab. Macassar, Menado, Sula Island (Wallace). Prince Bonaparte (Consp. ii. p. 59), apparently on Temminck’s authority (Nouv. Rec. Pl. Col. 248), cites Celebes as the habitat of Reinwardtena reinwardti (Temm.). Mr. Wallace (Ibis, 1865, p. 391) does not include Celebes within its range. Turtur, Selby. 128. Turtur ticrina (Temm.), Knipp, Pig. pl. 43 (1811); Wallace, Ibis, 1865, p. 391. Turtur chinensis, ap. Wallace, Ibis, 1860, p. 147. Hab. Java, Malay peninsula, Lombock, Flores, Timor, Ternate, Celebes (Wallace) ; Menado (mus. nostr.). GOURIDZ. Pxioca@nas, Reichenbach. 129. PaLoca@nas rristiemata (Temm.), Mus. Lugd.; Bp. Consp. ii. p. 87, “'Tondano” (1857); Wallace, Ibis, 1865, p. 393, pl. 10; Malay Archip. i. p. 413. Hab. Macassar, Menado (Wallace). CHALCoPHAPS, Gould. 130. CHALCOPHAPS STEPHANI, Jacq. et Puch. Voy. Péle Sud, Zool. p. 119, “ Nouv. Guinée, cote occidentale” (1853); Peristére d Etienne, Homb. et Jacq. Atlas, pl. 28. f. 2 (January 1846); Wallace, Ibis, 1865, p. 394; Schlegel, Nederl. Tijdschr. Dierk. 1866, pp. 265, 345. Hab. North Celebes (Wallace). 86 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Mr. Wallace (/. c.) has separated the New Guinea, Waigiou, and Mysol race from that inhabiting Celebes, and conferred on it the title of Ch. hombroni. But as the type of Ch. stephani was obtained in New Guinea, if the two races are distinct, the Celebean, and not the New-Guinea bird requires a new title. 131. CuatcopHaps rnpica (Linn.), Syst. Nat. 12, i. p. 284 (1766), ex Edwards, pl. 14; Schlegel, Nederl. Tijdschr. Dierk. 1866, p. 267. Only two species of this subgenus are recognized by Professor Schlegel :—first, Ch. stephani, as restricted above; secondly, all the remaining races of Asia, its islands, Australia, New Caledonia, and the islands of the Gilolo and Ceram groups. Members of this second species are stated by the Professor (/. c.) to also inhabit Celebes and New Guinea, but to be exceedingly rare in those two localities. Mr. Wallace does not appear to have met with it in either country. GEOPELIA, Swainson. 132. GEOPELIA striATA (Linn.), Syst. Nat. ed. 12, i. p. 282 (1766), “India orientali” (1766), ex Brisson. Hab. Macassar (Wallace); Sava (Sparrman) ; Queda (Sonnerat) ; Lombock ( Wallace) ; Philippines (Von Martens). I include this species on the authority of Mr. Wallace. Catawas, G. R. Gray. 133. CaL@NAS NICOBARICA (Linn.), Syst. Nat. ed. 12, i. p. 283, “insula Nicombar” (1766), ex Albin, pl. 47; Wallace, Ibis, 1865, p. 400; Von Pelzeln, Reise der Novara, Vogel, p. 110. Hab. Malacca and Singapore, Celebes, Batchian, New Guinea (Wallace); Treis Island, Nicobars (Von Pelzeln). This species is given from Celebes by Mr. Wallace in his table of distribution (J. ¢.); but it is to be inferred, from the interesting account given by the same author of its range and habits (Malay Archip. ii. p. 65), that the Nicobar Pigeon is not found on the main island. GALLIN . PHASIANID. Gaus, Linneus. 134. GaLLus BANKIVA, Temm. Pig. et Gallin. ii. p. 87, “Java” (1813). Hab. Java (type); Macassar (Wallace). Mr. Wallace has informed me that this species occurs in Celebes. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 87 Gmelin’s diagnosis of G. ferrugineus was undoubtedly taken from Latham’s sixty-sixth plate, which represents the hen of the red Indian Jungle-fowl. But Gmelin first quoted Sonnerat’s Grande caille de la chine (It. ii. p. 171), a bird that cannot, by its description, be referred to the genus Gallus, and which seems to haye been described from an example of 7. perlatus, Gm. Latham having erroneously identified Sonnerat’s species with his own Hackled Partridge, was copied by Gmelin; hence two distinct birds are included under Tetrao ferrugineus, Gm. It will be necessary to compare Celebean examples with those from other parts of Asia before we can decide to which species they belong. TETRAONID. ExcaLractoriA, Bonaparte. 135. ExcaLractoriA Minima, Gould, P. Z. S. 1859, p. 128, “ Macassar,” Birds of Asia, pt. xiii. Hab. Macassar (Wallace). A representative form of E. chinensis (Linn.), if admissible as distinct. TURNICID. Tournix, Bonnaterre. 136. Turnix RuFILATUS, Wallace, P. Z. S. 1865, p. 480, “ Macassar.” Hab. Macassar (Wallace). MEGAPODIID. MEGaPopivs, Quoy et Gaimard. 137. Mecapopius GiLBerTI, G. R. Gray, P. Z..S. 1861, p. 289, “Celebes;” Schlegel, Nederl. Tijdschr. 1866, p. 263. Megapodius of small size, Wallace, Ibis, 1860, p. 142. Hab. Celebes (Wallace, Schlegel); Island of Siao (Sanghir group ?) (Schlegel). M. rubripes, Temm. Pl. Col. 411, “Celebes,” is neither from Celebes nor Amboyna (conf. Schlegel, op. cit. p. 260). MEGACEPHALON, Temminck. 138. MEGACEPHALON MALEO, Temminck. Megapodius rubripes, Temm., apud Quoy et Gaim. Voy. Astrol. Zool. i. p. 239, pl. 25, av. juv., nec Temm. Megacephalon rufipes (Quoy et Gaim.), Gray & Mitch. Genera, ii. pl. 123. maleo, Wallace, Ibis, 1860, p. 142. —— rubripes, Wallace, Malay Archip. i. p. 413. 88 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Megacephalon maleo, Temm., Bp. Compt. Rend. xlii. p. 876 (1856). rubripes, G. R. Gray, P. Z.S. 1861, p. 288; op. cit. 1864, p. 42, nec Temm. Hab. North-east Celebes (Wallace). Although we owe to Messrs. Gray and Mitchell (/.c.) an excellent figure, and to Mr. Wallace (/. c.) a most interesting account of this species, no description, with a distinctive title, appears ever to have been published of the adult bird. The specific title adopted above is the name by which this Megapode is known to the natives of North Celebes. Temminck’s only published notice of the species is in these words :— “Le grand Mégapode, connu aux Célébes sous le nom de Maleo ne nous est point encore parvenu” (Pl. Col. 411); and he then states that it must not be confounded with the other Celebean Megapode, I. rubripes, Temm. It was, however, so confounded for many years after, until Prince Bonaparte (/. c.) enumerated it as a distinct species in his ‘ Tableaux Paralléliques.’ Temminck does not appear either to have published the characters of his genus Megacephalon. A fine male from North-east Celebes (mus. nostr.) has the head, chin, throat, and entire upper half of the neck naked, with a few straggling, short, brown feathers interspersed. The quills, rectrices, upper and under tail-coverts are deep brown, nearly black, with a dark green gloss. Upper breast and entire upper surface dark brown. Under surface and flanks salmon-colour. Fifth and sixth quills equal, and longest ; fourth and seventh a trifle shorter, and equal; third somewhat shorter than fourth; the second an inch shorter than the third, and the first an inch shorter than the second. GRALLA. CHARADRIIDZ. Cuaraprivs, Linnzus. 139. CHARADRIUS FULVUS, Gmelin, Syst. Nat. ed. 13, i. p. 687, ex Lath. Syn. iii. p. 211, “ Otaheite ;” Schelgel, Mus. Pays-Bas, Cursores, p. 30. Hab. Gorontalo, April, males passing into perfect plumage, female passing into perfect plumage, April 20 (Rosenberg); Gorontalo, passing out of perfect plumage, September 24 (Forsten). The complete range of this species cannot be given until we have agreed upon the races which ought to be included under the above title. For an exhaustive essay on the subject, conf. Finsch & Hartl. Faun. Centralpolyn. p. 188. Evpromias, Boie. 140. EvpRoMIAS VEREDUS (Gould), P. Z.S. 1848, p. 38, “ Northern Australia ;” Harting, Ibis, 1870, p. 209. Hab. Macassar (Wallace); Northern and Eastern Asia, Malay archipelago, New Guinea, Australia. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 89 ZEGIALITES, Boie. 141. Harauires pusius (Scop.), Del. Faun. et Fl. Insub. ii. p. 93. no. 81 (1786), ex Sonn. Voy. Nouv. Guin. p. 84, pl. 46, “* Luzon.” Charadrius philippinus, Lath. Ind. Orn. ii. p. 745. no. 11 (1790), ex Sonn. J. c. alexandrius, Hasselq. var. 5, Gm. S. N. ed. 12, i. p. 684, ex Sonn. J. c. —— philippinus, Scop. (!), Schlegel, Mus. Pays-Bas, Cursores, p. 28. ? Aigialites minutus (Pall.), ap. Jerdon, Birds Ind. iii. p. 641. Hab. Ayer-pannas, 6th of August (Von Rosenberg). A Celebean example of a Ring-Plover, collected by Von Rosenberg, has been identified by Professor Schlegel (/.¢.) with Le petit Pluvier @ collier de Lugon of Sonnerat; and he has further united it with the Lesser Ring-Plover of Europe. A Philippine Ring-Plover has also been identified by Dr. von Martens (J. fiir O. 1866, p- 26) with the European bird, i. e. C. curonicus, Gm. (ex Beseke, Schr. Berl. Gesellsch. nat. Freunde, vii. p. 463. no. 48, who gave no title)—the C. minor, Meyer, of recent authors. In India, besides C. curonicus (=C. minor, or else C. intermedius, Ménétr., if really distinct), another small Ring-Plover occurs, the 4. minutus (Pall.) ap. Jerdon, a species distinct from C. curonicus, Gm.; and the question arises whether this is not the species Sonnerat figured. As is the case in India, it is not impossible that both species inhabit the Philippines and also Celebes. Without inquiring into the validity of C. minutus, Pallas, and whether or not it indicates only C. curonicus in young plumage, as maintained by O. Finsch and Hartlaub (Orn. Ost-Afr. p. 661), these gentlemen seem to have been somewhat hasty in identifying Sonnerat’s bird with 4. curonicus (Gm.). Sonnerat states that the bill and feet are “noiratres.” Both Sonnerat and Buffon (Hist. Nat. viii. p. 95), who refers to Sonnerat’s Philippine specimen, say that the Philippine species differs but little from the European Little Ring-Plover. But both those authors included it also among North- and South-American species, and Buffon hardly recognized the specific distinctness of 4. hiaticula. Af. minutus (Pall.) ap. Jerd., is a smaller and more delicately formed species. In plumage it closely resembles 4. cwronicus, but has the head-markings better defined than those of any example of that species I have as yet seen. Its chief distinction is to be found in the smallness of the feet and shortness of the legs. A Katmandoo specimen has the legs dark reddish brown, instead of yellow. It behoves naturalists in India to investigate these differences. I am inclined to believe in there being two species, but have not been able to examine a sufficiency of individuals to form a decided opinion. Should the Philippine smallest Ring-Plover prove identical with the European C. curo- nicus, Gm., both will have to take the title of dubiws, Scop. 48, minutus (Pall.) ap. Jerd. may be identical with Charadrius pusillus, Horsf. Trans. Linn. Soc, xii. p. 187, ex Java. VOL. VIII.—PART I. May, 1872. P 90 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Table of Dimensions. | | | Wing. | Tarsus. | Tail. Bill. | inches. | inch. | inches. | inch. ZB. curonicus (Gm.) ..| 4:500| -8750) 2°875}| -5000| Piedmont. 9. May; not quite full plumage. . BS ..| 4750) -8750) 2:875| -5000| Europe. Not quite full plumage. 5 » -+| 4625) -8750| 2-625) -5625) Bengal. Not quite full plumage. 5 » «-| 4500] -8750) 2-625) -5000| Malta, April 4. Not quite full plumage. - 4 ..| 4250| -8750| 2*875| -5000| Malta, April 4. Not quite full plumage. > 5) ow s| ee et | °8750} 2°750) 5625 | Calcutta, February. Not full plumage. “3 os .| 4875 | -8750| 2-750) -5000) Turkey, May 7. Almost full plumage. = 3 .| 4625 | -8750| 2-750) -5000 Abyssinia, R. Amba, August 21. ¢. Young plumage. Ae 5 .| 4875 | -8750| 2°750| -5000)|Coorg. Winter? or first plumage ? BS » «+| 4625} -8750| 2-750} -5000|Coorg. Winter? or first plumage? AR. minutus (Pall.), ap. TTA SO ABA Hers OID O10 4-250) -7500} 2:°375| -4375)|Candeish. Full plumage. e .| 4125 | -7500| 2-875} -5000|Candeish. Full plumage. a 53 .| 4375 | -7500| 2875) -4375 | North-western India. Full plumage. 5 5 .| 4375 +7500) 2°875| -5000 Maunbhoom, December 16. Full plumage. hs a .| 4:125 | -6875 | 2°875| -3750|Katmandoo. Full plumage. 142. Mearaires peronw (Temm.), Schlegel, Mus. Pays-Bas, Cursores, p. 33, “VArchipel Indien” (March 1865); Swinhoe, P. Z.S. 1870, p. 139. (Pl. X. fig. 2.) Hab. Borneo, Java, Semao (Schlegel) ; Macassar (Wallace). An example of this species in summer plumage was obtained by Mr. Wallace in Celebes. Mr. Swinhoe, who has also examined this individual, agrees with me in identifying it as above. It belongs to the subsection of which 4. cantianus may be regarded as the type. As it is a rare species, I append the following description :— Forehead, from the base of the bill, pure white; a broad white superciliary stripe, confluent with the white forehead, terminates above the black ear-coverts; narrow frontal band, lores, ear-coverts, and a broad band crossing the back and reaching to the sides black ; a black pectoral stripe, continued from the black sides, is narrowed into a thin line on the breast, where it does not quite meet; this excepted, the entire under surface, cheeks, under wing-coverts, and a broad nuchal collar pure white. Upper plumage pale earthy brown, most of the feathers with albescent edgings, conspicuous on the wing-coverts, some of which are edged with a purer white; primaries reddish brown; secondaries paler brown, broadly margined on the inner webs, and tipped with white, more or less cinereous; all the shafts white; three outer pairs of rectrices pure white; the next pair pale brown, much mixed with white (the rest of the rectrices are absent in this example); bill jet-black, no trace of any other colour; legs, in dried skin, pale yellow brown. Wing 33, bill from forehead §, tarsus 1, tail 1$. The frontal white patch is broad, more so than in European examples of 4. hiaticula. in proportion the black frontal band is narrow, and is not posteriorly edged with white. ’ First primaries wanting. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 91 STREPSILAS, Illiger. 143. STREPSILAS INTERPRES (Linn.), Syst. Nat. ed. 12,i. p. 248; Schlegel, Mus. Pays-Bas, Cursores, p. 43. Hab. Celebes (Mus. Lugd.): almost universal. Esacus, Lesson. 144. Esacus maa@nirostris (Geoffroy St.-Hilaire): Vieill. N. Dict. xxiii. p. 231 (1818), nec Latham. Cidicnemus magnirostris, Geoftroy, Temm. Pl. Col. 387, “ Celebes ;” Wagler, Syst. Av. Charadrius, no. 3, “ New Holland” (1827) ; Schlegel, Mus. Pays-Bas, Cursores, p. 22. Charadrius giganteus, Lichtenst.; Wagler, Isis, 1829, p. 647, “ New Holland.” Esacus magnirostris, Geoffroy, Gould, Hand-b. B. Austr. 1, p. 213. (Latham), Wall. P. Z. S. 1862, p. 346, nec Latham. Hab. Celebes (Reinwardt); Island of Raou, near Morty, Island of Moor, east coast of Gilolo, Waigiou, Bangka (Mus. Lugd.); northern and north-western parts of Australia (Gould); Sula Islands, New Guinea (Wallace). The Australian ‘“ Great-billed Plover” of Latham (Syn. Supp. il. p. 319, C. magni- rostris, Lath. Ind. Orn. Supp. p. 66) has been shown by Mr. Strickland (Ann. Nat. Hist. xi. p. 337) to be nothing but @dicnemus grallarius (Lath.). Consequently IMliger’s genus Burhinus (Prodrom. p. 250, 1811), founded on C. magnirostris, Lath., is syno- nymous, not with Hsacus, but with Cdicnenws, over which generic title it takes priority. The name magnirostris, Geoffroy, seems to have been an unpublished museum title, I can find no earlier description of the species than that of Vieillot’s (/. ¢.), who adopted the name from the Paris Museum. Temminck figured (/. c.) a Celebean example collected by Reinwardt; but he united with it as belonging to one species individuals from India, Java, and Jes iles Papous. The Celebean bird in size, he observes, holds a middle place between the Indian and the Papuan, the last being the largest and having the plumage very dark-coloured. The Indian E. recurvirostris (Cuv.) is a recognized species; but may not the Celebean bird prove to be a species distinct from the Australian? Professor Schlegel unites the archipelagic with the Australian; but have they been compared? Himantorvs, Brisson. 145. Hiantorus LEucocepnanus, Gould, P. Z. S. 1837, p. 26, “Australia, Java, Sumatra;” Birds Austr. vi. pl. 24; Schlegel, Mus. Pays-Bas, Scolopaces, p- 106; conf. Blyth, Ibis, 1865, p. 35. Hab. Gorontalo, October 9 (Forsten); Ayer-pannas, August 14; Limbotto, August 29; Wawou, a few days old, August 27 (Rosenberg); Bengal (H. intermedius, Blyth, PQ 92 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. J.A.S. B.?; Cat. Mus. Cale. no. 1573); rare in India, J. A.S. B. 1845, p. 459 (Blyth) ; Java, Borneo, Amboyna, Ternate, Sumbawa, Timor, Lobo (New Guinea) (Mus. Lugd.) ; Australia (Gould). RALLID. Porpuyrio, Brisson. 146. Porpuyrio inpicus, Horsf. Trans. Linn. Soc. xiii. p. 194, “ Java” (1822, read 18th of April 1820); Schlegel, Mus. Pays-Bas, Ralli, p. 55; Finsch & Hartl. Faun. Centralpolyn. Aves, p. 170, pl. xii. f. 2. Porphyrio smaragdinus, Temm. Pl. Col. 421, “Java” (February 1827). Hab. Macassar, Menado (mus. nostr.); Tondano, 21st of April (Forsten); Gorontalo, 18th of April, 24th of May, 26th of June; Ayer-pannas, 11th of August (Von Rosenberg); Java (type); Ceram, Bouru (Mus. Lugd.); Banda (G. Rk. Gray); Sumatra (Cassin) ; Samoa Islands (Peale). The absolute identity of the race of purple Coots inhabiting the islands above cited has yet to be established. To the Ceram race Temminck applied the title of mela- nopterus ; that of Samoa has received the designation of samoénsis, Peale. It is true that the late Mr. Cassin could detect no difference between the Samoan and the Javan Porphyrio; and Messrs. Finsch & Hartlaub (/.¢.) agree in uniting them. On the other hand, Professor Schlegel has observed slight distinctions between the individuals inhabiting Java, Celebes, and Ceram. I have not been able to compare a sufficient series in full plumage to form a decided opinion. But Celebean birds appear to have the throat, upper breast, and shoulder-coverts of a much richer and deeper blue than what I have found in Javan examples. I am unable to discover sufficient evidence to justify Latham’s title of poliocephalus (Suppl. Ind. Orn. p. 58) being applied to the Philippine Porphyrio, rather than to the one of Continental India (P. neglectus, Schlegel). Latham’s Grey-headed Gallinule (Syn. Suppl. ii. p. 375) was described by him from a drawing by General Davies, of an individual in Exeter Change. The description agrees well enough with the Indian bird, and better than with the Philippine. It is certainly not sufficiently minute to enable us to refer it without doubt to the latter species, P. pulverulentus, Temm. (Pl. Col. 405, erroneously given from Africa); while the probabilities are in favour of the type having come to London from India, and not from the Philippines. HypraLector, Wagler. 147. HyDRALECTOR GALLINACEUS (Temm.): Pl. Col. 464, “ Moluques” (5th of July 1828) ; Gould, Birds Austr. vi. pl. 75. Parra cristata, Vieill., Schlegel, Mus. Pays-Bas, Railli, p. 68. Hab. Menado, Macassar (mus. nostr.); Ayer-pannas, adult male, 21st of August, VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 93 adult female, 18th of August, male partly moulting, 22nd of August, young female, 21st of August; Limbotto, adult male of small dimensions, 3lst of August, female moulted, 29th of August; Gorontalo, adult female, 30th of April; Wawou, very young male, 27th of August (Von Rosenberg) ; Gorontalo, young female, 29th of June (Forsten); Port Essington, Eastern Australia (Gould) ; Queensland (mus. nostr.). Temminck (7. c.) states that this is a bird of passage at Amboyna. Mr. Gould (Hand- book, Austr. ii. p. 331, where an interesting account of its habits is given) mentions that it is a native of New Guinea. No authority is quoted; and I can find no con- firmation of the statement. Professor Schlegel confines its range to Celebes and Australia. It breeds in Eastern Australia (Gould, /.c.); but unfortunately the month is not stated. I cannot follow Professor Schlegel (J. c.) in referring this species to P. cristata, Vieill. (N. D. xvi. p. 430, ex Ceylon). Vieillot’s title was given to Le Grand Jacana verd & créte of Temminck (Cat. Syst. Cabinet d’Ornith. p. 265. no. 403, 1807), whose description Vieillot reproduces almost word for word. Temminck’s Jacana, as has already been shown by Dr. Hartlaub (Syst. Index, in Jard. Contrib. Ornith. 1849), is clearly Parra indica, Lath. (Ind. Om. ii. p. 765, 1790). Wagler (Isis, 1832, p. 280) gives both P. cristata and P. gallinacea as the types of his genus Hydralector. But the generic character, “Ein aufrecht stehender Fleischkamm am Kopfe,” evidently indi- cates P. gallinacea as the generic type. My Macassar example, an adult, only differs from a Menado individual by being much smaller. Wing 42 against 5§. All the other dimensions proportionally less; it is therefore probably a male. A Queensland example, a young bird, crown and nape rich rufous intermixed with black, only differs in having a much stouter bill. GALLINULA, Brisson. 148. Gatuinua rrontata, Wallace, P. Z. S. 1863, p. 35, “ Bouru.” Gallinula hematopus, Temm. Mus. Lugd. ; Schlegel, Mus. Pays-Bas, Ralli, p. 44, “ Celebes.” Hab. Menado (mus. nostr.); Ayer-pannas, 12th of August, adult male, 17th of August, adult female, 26th of August, female, 19th of August, young, one day old; Panybie, 9th of September, female of the year; Limbotto, 4th of September, female of the year, 31st of August, male and female of the year, 28th of August, female (Von Rosenberg); Amboyna (Mus. Lugd.); Bouru (type). - Professor Schlegel, in his admirable list of the birds of the Leyden Museum, the most perfect and practically useful work of its kind ever published, identifies the Celebean bird with that described by Mr. Wallace from Bouru. Temminck’s MS. title of hematopus had never been hitherto used, except by Bonaparte (Comptes Rend. xliii. p. 600, 1856), and then only as a synonym of the nearly allied G. tenebrosa, Gould (P. Z. 8. 1846, p. 20). I have therefore retained Mr. Wallace’s title for the species. 94 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. It must, however, be remembered that no actual comparison appears as yet to have been made between Bouru and Celebean examples. 149. GALLINULA oRIENTALIS, Horsf. Trans. Linn. Soe. xiii. p. 195, “Java” (1822). Gallinula ardosiacea, Vieill. Galerie, ii. p. 173, pl. 268, “ Java” (1825). Hab. Sava (Horsfield); Macassar (Wallace). An example of an adult male Moor-hen, closely resembling the common European species, was collected by Mr. Wallace at Macassar. It differs from G. chloropus in its smaller dimensions, and the size and form of the frontal plate. I have been unable to compare it with Javan individuals; but I shall provisionally adopt the title of the race which inhabits Java. Wing 5, tarsus 1§, bill from anterior side of plate to tip 144, greatest breadth of frontal plate 7%. Erytura, Reichenbach. 150. ErytHra pHa@nicura (Forsten): Zool. Ind. p. 19, pl. 9, “ Ceylon” (1781). Hab. Macassar (Reinwardt) ; Gorontalo, adult male, 20th of April, 17th of July, 1st of August—male with some black spots on sides of head, 26th of May; Negrielama, male in first plumage, 20th of September (Von Rosenberg) ; Gorontalo, male in imper- fect plumage, October (Forsten); Banka, Java, Borneo (Mus. Lugd.); China, summer visitant (?), Formosa (Swinhoe); throughout India (Jerdon); Ceylon (type); Zamboanga (Mindanao) (Von Martens); Malayan peninsula (Eyton). ORTYGOMETRA, Linneeus. 151. Orrycomerra crveREA (Vieillot): Nouv. Dict. d’Hist. Nat. xxviii. p. 29 (1819), ex Java; Pucheran, Rev. et Mag. Zool. 1851, p. 563; Schlegel, Mus. Pays-Bas, Ralli, p. 82. Ortygometra quadristrigata (Horsf.), Finsch et Hartl. Fauna Centralpolyn. p. 164. Hab. Gorontalo, April 21, May 23; Ayer-pannas, August 25, female in first plumage, August 25 (Von Rosenberg). For the geographical distribution of this species and its complete synonymy, exclusive of the title, conf. Finsch and Hartl. 7. c. Those gentlemen seem to have overlooked in this and in one or two other instances Dr. Pucheran’s valuable notices of the types contained in the Paris Museum. According to the learned doctor, Porphyrio cinereus, Vieill., was collected in Java by Labillardiére. This species is included in Mr. Hodgson’s Catalogue of the Birds of Nipaul (J. A. S. Bengal, 1855, p. 381. no, 765) under the title of Zapornia nigrolineata. Mr. G. R. Gray, Cat. B. Mus. Nepal, 1846, p. 143, identified Z. nigrolineata, Hodgs., with Rallus superciliaris, Eyton, ex Malacca, and in the 3rd edition of that catalogue (1863) adopted Eyton’s specific title, Mr. Blyth (Cat. Cale. Mus. p. 339) includes Nipaul within the range of 2. superciliaris, Eyton. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 95 Professor Schlegel (/.c.) has identified Eyton’s species with P. cinereus, Vieill.; and Drs. Finsch & Hartlaub (0. c.) with R. quadristrigatus, Horsf. ‘The species, however, is not included in Dr. Jerdon’s work as an inhabitant of India. HyporanipiA, Reichenbach. 152. HyporHNIDIA CELEBENSIS (Quoy et Gaimard), Voy. Astrol. Zool. i. p. 250, * Celebes,” pl. 24. f. 2; Schlegel, Mus. Pays-Bas, Ralli, p. 22. Hab. Menado (Forsten); Gorontalo, Limbotto (Von Rosenberg). This is a representative form of the Philippine Radius torquatus, Linn. (Schlegel, l.¢.). Von Pelzeln (Novara, Aves, p. 134), with doubt, refers an example of a young Rail from Borneo to the Celebean species. 153. Hyporamnipia striata (Linn.), Syst. Nat. ed. 12, i. p. 262 (1766), ex Brisson, * Philippines.” Rallus gularis, Horsf. Trans. Linn. Soe. xiii. p. 196, “ Java” (1822). Hab. Philippines (type); all India and Ceylon, Burmah (Jerdon); Sumatra (faffles) ; Java (Horsfield); Cochin-china (Diard); Formosa (Swinhoe); China (Mus. Lugd.); Menado (Wallace); Banjarmassing (Sc/later). Mr. Wallace obtained near Menado a female example of a Rail which so well agrees with Brisson’s description of the Philippine bird, that I have little hesitation in making the above identification. It must, however, be noted that, in the specimen referred to, the under tail-coverts are distinctly pale rufous and black, and not white and black. 154. Hyporanipia PHILIPPENSIS (Linn.), Syst. Nat. ed. 12, i. p. 263 (1766), ex Brisson, Orn. v. p. 163, “ Philippines ;” Schlegel, Mus. Pays-Bas, Ralli, p. 25. Rallus pectoralis (Cuvier), Lesson, ap. Finsch et Hartl. Faun. Centralpolyn. p. 157, nec Cuv. ; conf. Pucheran, Rev. et Mag. Zool. 1851, p. 276. Hab. Macassar (mus. nostr.); Tondano, in September (Morsten); Gorontalo, April 17, 24, May; a chick newly hatched, August 4 (Von Rosenberg); Australia (Gould) ; New Caledonia (Verreaux et O. des Murs); Philippines (type). The Celebean bird has the nape rusty as in Australian individuals. In the event of the Philippine species proving distinct, the birds from the other localities above given will require a different title. Messrs. Finsch & Hartlaub (J. ¢.) have adopted Cuvier’s title of pectoralis, copied by Lesson (Tr. p. 536), for this species, although Dr. Pucheran (1. c.) had shown that the type of R. pectoralis, Cuv., was 2. lewinit, Swains. (conf. Hartl. J. fiir Orn. 1855, p. 420). Ratu, Reichenbach. 155. Rauua minawasa, Wallace, P. Z. S. 1862, p. 346, “Sula Islands, Minahasa.” Hab. Minahasa (N. Celebes), Sula Islands (Wallace). 96 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. > 156. RALLINA ISABELLINA (Temm.), Schlegel, Mus. Pays-Bas, Ralli, p. 16 “Celebes’ (1865). Hab. Gorontalo, type (Forsten); Ayer-pannas, Modelido (Von Rosenberg). 157. Ratna (2) ROsSENBERGH, Schlegel, Nederl. Tijdschr. Dierk. 1866, p. 213, “ Kema.” Hab. Kema (Von Rosenberg). SCOLOPACID. Nvumentus, Linnzus. 158. Numenius pHmorvs (Linn.), Syst. Nat. ed. 12, i. p. 243 (1766); Schlegel, Mus. Pays- Bas, Scolopaces, p. 97. Hab. Bonthain, South Celebes, March 7th (S. Miiller); Tondano, North Celebes (Forsten) ; The Old World and Australia. Until the breeding-grounds of the so-called distinct species of Whimbrels are dis- covered it is useless to attempt discriminating between them. Both the Celebean examples in the Leyden Museum possess the characters whereby Mr. Gould has distin- guished his V. wropygialis. 159. Numenrus minutus, Gould, P. Z. S. 1840, p. 176, “ New South Wales;” Schlegel, Nederl. Tijdschr. Dierk. 1866, p. 348. Numenius minor, S. Miller, Verhandl. p. 110, “ Amboyna;” Schlegel, Faun. Japon. Aves, p. 111, pl. 67. Hab. North Celebes, Aru Islands (Schlegel); Japan (Von Siebold); Amboyna (S. Miiller); coasts of China (Swinhoe); New 8. Wales, Port Essington (Gould). Actitis, Illiger. 160. ActITIS GLAREOLA (Gmm.), Syst. Nat. ed. 13, i. p. 677 (1788); Schlegel, Mus. Pays- Bas, Scolopaces, p. 73. Hab. Gorontalo, October 9th (Forsten); Europe, Africa, Asia and its islands. 161. Actitis HYPOLEUCOsS (Linn.), Syst. Nat. ed. 12, i. p. 250 (1766); Schlegel, Mus. Pays-Bas, Scolopaces, p. 83. Hab. Gorontalo, in October (Forsten); Europe, Africa, Australia, Asia and its islands. Toranus, Bechstein. 162. Toranus Gxortis (Linn.), Syst. Nat. ed. 12, i. p. 245 (1766); Schlegel, Mus. Pays- Bas, Scolopaces, p. 63. Hab. Celebes, in winter plumage (Forsten); Bonthain, South Celebes, in March (S. Miller): universal. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. oa 163. Toranus caLipris (Linn.), Syst. Nat. ed. 12, i. p. 245 (1766); Schlegel, Mus. Pays- Bas, Scolopaces, p. 67. Hab. Celebes, winter plumage, November (Forsten); Europe, Asia and its islands, Africa. Liwosa, Brisson. 164. Limosa vropyeranis, Gould, P. Z. 8. 1848, p. 38, “ Australia;” Birds. Austr. vi. pl. 29; Schlegel, Mus. Pays-Bas, Scolopaces, p. 25. Hab. Celebes, in November (Forsten); Gilolo (Bernstein); Java (Van Hasselt) ; Timor (Miller) ; Japan, New Zealand (Mus. Lugd.); Australia (type). (Conf. Finsch & Hartl. Fauna Centralpolyn. p. 177.) Trinea, Linnzus. 165. Trinea minuta, Leisler, Nachtr. Bechst. Naturgesch. Deutschlands, Heft i. p. 74 (1811-15); Schlegel, Mus. Pays-Bas, Scolopaces, p. 46. Hab. Celebes, in November, winter plumage (Fursten); Europe, Africa, Asia, Malay archipelago, New Guinea, Australia (Mus. Lugd.). 166. TrinGa Damacensis (Horsf.), Trans. Linn. Soe. xiii. p. 192, “ Java” (1822); Swinhoe, P. Z. 8. 1863, p. 316; Schlegel, Mus. Pays-Bas, Scolopaces, p.49. Tringa subminuta, Von Middendorf, Sibir. Reise, Végel, p. 222, pl. 19. f. 6 (tarsus). Hab. Tondano, winter plumage; Tondano, male, partly in nuptial plumage, September; Gorontalo, male, winter plumage, October 9th; Celebes, moulting into perfect plumage (Forsten) ; Java (type); Borneo (Schwaner) ; China, Formosa (Swinhoe); Eastern Siberia (Von Middendorf); Amoor river (Schrenck). Losipes, Cuvier. 167. LopiPes HYPERBOREUS (Linn.), Syst. Nat. ed. 12, i. p. 249 (1766); Schlegel, Mus. Pays-Bas, Scolopaces, p. 59. Phalaropus australis, Temm. Mus. Lugd.; Bp. Compt. Rend. xlii. p. 421 (1856), “ Celebes,” fide Schlegel, /. c. Hab. Celebes, winter plumage (Reinwardt); Amboyna, winter plumage (Hoedt); Aru Islands, in winter plumage (Wallace); Madras (Jerdon); Peninsula of Luichow, April 3rd (Swinhoe); high latitudes of northern hemisphere, in summer. Are the Moluccas the only, or at least the principal, winter residence of this species? Its occurrence has only been once observed in India. VOL. VIII.—PART Il. May, 1872, Q 98 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. GaLLinaco, Stephens. 168. Gattinaco MEcaLa, Swinhoe, Ibis, 1861, p. 343, “Amoy;” Schlegel, Mus. Pays- Bas, Scolopaces, p. 12. Hab. Gorontalo (Forsten); Gilolo, Batchian (Bernstein) ; China, Formosa (Swinhoe). ARDEIDZ. ARDEA, Linneus. 169. ARDEA SUMATRANA, Raffles, Trans. Linn. Soc. xiii. p. 325, “Sumatra” (1822); Jerd. Birds Ind. ii. p. 740; Gould, Birds Austr. ii. p. 296. Ardea typhon, Temm. Pl. Col. 475, “river Gambia” (errore), (Sept. 5, 1829) ; Schlegel, Mus. Pays- Bas, Ardea, p. 3. rectirostris, Gould, P. Z. 8. 1842, p. 22, “New South Wales.” fusca, Blyth, Ann. Nat. Hist. xiii. p. 176, “ Arracan” (1844). insignis, Hodgs., Gray’s Zool. Misc. p. 86. Typhon robusta (S. Mill.), Bp. Consp. 11. p. 110, “ Timor.” Ardea goliath, Temm., Bp. l. c., “ Celebes.” Ardeomega nobilis, “ Blyth,” Bp. 1. c., nee Blyth. (Conf. Blyth, Ibis, 1865, p. 36.) Hab. Celebes (Reinw.); “Inde continentale,” type of A. typhon, Temm. (Schlegel) ; Morty Island, Batchian, Toloforo (Gilolo) (Bernstein); Sumatra (Raffles); Coburg Peninsula (Gould); Clarence river (Australia) (Schlegel); Arracan (Blyth); N.-E. Bengal, Nepaul, Sikim, Terai, Assam (Jerdon); Sindh (drawing, Sir A. Burnes); Flores (Wallace). ARDEOLA, Boie. 170. ArpEoLA spEctosa (Horsf.), Trans. Linn. Soc. xiii. p. 188, “Java” (1822); Zool. Res. pl. —. Hab. Java (type); Celebes (Wallace). I include the Javan form of A. leucoptera, Bodd., =A. malaccensis, Gm., on the authority of Mr. Wallace. There appear to be four closely allied Asiatic species of Ardeola; but they yet require to be brought together and closely compared. 1. A. leucoptera, Bodd, ex Pl. Enl. 911, =A. malaccensis, Gm., ex Pl. Enl. 911, “Malacca,” av. juv. Buphus bacchus, Bp., having been described from a Malaccan individual, becomes a synonym of the type species. 2. A. grayi, Sykes, “ India,” =A. leucoptera, Bodd., ap. Jerdon (B. of Ind.). 3. A. speciosa, Horsf., “Java.” Most probably the same as the Malaccan form. Stated by Professor Schlegel to also occur in Sumbawa and Borneo. 4. A. prasinosceles, Swinhoe, ‘“ China.” Not recognizing the fact that Boddaert and Gmelin founded their titles on the same VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 99 plate, Mr. Blyth (Ibis, 1865, p. 38) called the Indian bird leucoptera, Bodd., and that of the Malayan peninsula and Sumatra malaccensis, Gm. Heroptas, Bor. 171. Heropias nicripes (Temm.), Man. d’Orn. 2nd edit. part iv. p. 376, “ L’Archipel des Indes” (1840); Schlegel, Mus. Pays-Bas, Ardew, p. 14, “ Lac de Gorontalo, Celebes.”’ Not possessing a sufficient number of examples of H. garzetta (Linn.) and its allies to attempt an elucidation of its races, their habitats, and synonymy, I have followed Temminck, and given to the Celebean bird the title by which the Dutch zoologist distinguished the little Egret of India, of the Malay archipelago, and of New Guinea, from the European, North Asiatic, and Japanese bird. Professor Schlegel (J. ¢.) does not admit their specific distinction, and includes all under A. garzetta, Linn. To him we owe the important fact that Temminck founded his A. nigripes on examples from Java, Borneo, and Celebes now in the Leyden Museum. We are thus provided with a clue to the maze of confusion into which Prince Bonaparte (Consp. ii.) has thrown the synonymy of the White Egrets (conf. Schlegel, op. cit. p. 19). 172. Heropias EGRETTA (Gm.), Syst. Nat. ed. 13, i. p. 629 (1788), ex Buff. Hist. Nat. Ois. vii. p. 377, “ America;” Schlegel, Mus. Pays-Bas, Ardee, p. 17. Hab. Gorontalo (Forsten). I adopt Professor Schlegel’s determination with reserve, its correctness depending on the identity of the Asiatic with the American bird. The Celebean example is probably the H. alba (L.), ap. Jerd. (Birds of India), =A. modesta, Gray & Hardw., A. alba vera being restricted by Professor Schlegel to Southern Europe, Northern Africa, and Western Asia. The history of the Egrets has yet to be written. Arpetta, G. R. Gray. 173. ARDETTA SINENSIS (Gm.), Syst. Nat. ed. 13, i. p. 642 (1788), ex Lath., “ China.” Ardea melanoptera, Cuvier, Mus. Paris. Puch.; Rev. Mag. Zool. 1851, p. 575. melanophis, Cuy., ap. Less. Tr. p. 578, errore. lepida, Horsf. Trans. Linn. Soc. xiii. p. 190, “ Java.” -— melanotis, Cuv., ap. G. R. Gray, Genera, iii. Append. p. 25, errore. Hab. Menado (mus. nostr.); all India (Jerdon); Java (Horsf.); Ceylon, Arracan (Blyth); China, from Canton to Tientsin, Formosa, in summer (Swinhoe); Borneo, Philippines (Mus. Lugd.); Ladrone or Marian Isles (2) (G. R. Gray). I cannot concur with Mr. Blyth nor with Dr. Jerdon in regarding Ardea nebulosa, Horsf. (/. ¢.), as belonging to this species. Horsfield’s diagnosis applies far better to Ardetta cinnamomea (Gm.). The expression “cauda remigibusque badiis” appears to me conclusive. Q2 100 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Demizcretra, Blyth. 174. DemincReTta sacra (Gm.), Syst. Nat. ed. 13, i. p. 640, ex Latham. Hab. Menado (mus. nostr.). Two examples of an Ashy Egret were received from Menado in the dark ashy phase, but wanting the crest, dorsal trains, and pectoral plumes of the breeding-plumage. Both have a narrow median white line commencing at the chin and descending, with broken intervals, down the throat. No other part of the plumage is white. The wing measures 104 inches; the bill, from the forehead, 3 inches; the tarsus 2? inches ; middle toe, without the nail, 1? inch. The dimensions of the bill, tarsus, and middle toe are much less than those given by Dr. Jerdon of the Indian bird, D. asha (Sykes). Dr. O. Finsch (Centralpolyn. p. 201) has united all the titles given to the numerous named local races of this species under Gmelin’s title of sacra, bestowed by him on the Sacred Heron of Latham, brought by Sir J. Banks from Otaheite. A want of a suf- ficient number of examples prevents me from questioning the correctness of this deduction, and I therefore provisionally adopt Gmelin’s title. For an elaborate essay on the species, conf. Finsch & Hartl. /. c. Nycricorax, Stephens. 175. Nycvicornax GRISEUS (Linn.), Syst. Nat. ed. 12, i. p. 239 (1766); Schlegel, Mus. Pays- Bas, Ardew, p. 58. Hab. Gorontalo (Forsten); Europe, Africa, Asia, America (Ardea gardeni, Gm.). 176. NycTicoRax CALEDONICUS (Gm.), Syst. Nat. ed. 13, i. p. 626, “Nova Caledonia” (1788), ex Lath. Syn. iii. p. 55; Schlegel, Mus. Pays-Bas, Ardew, p. 59. Ardea maculata, Lath. Ind. Orn. Supp. p. lxiv (1801), “ New Holland,” fide Strickl. Ann. Nat. Hist. xi. p. 338, av. juv. Hab. Macassar (S. Miller); Tondano (Forsten); Timor, Gilolo, Morty Island, Am- boyna (Mus. Lugd.) ; New Caledonia (type); Australia (Gould); Cape York (mus. nostr.). Ardea caledonica, Forster, apud Meyen (N. Act. Ac. C. L. C. xvi. Suppl. prim. p. 103), seems to be Nycticorax manilensis, Vigors; and I have therefore omitted the Philippines from the range of Gmelin’s species. Butoripes, Blyth. 177. Burormpes Javanica (Horsf.), Trans. Linn. Soc. xiii. p. 190, “Java” (1822); Schlegel, Mus. Pays-Bas, Ardee, p. 44. Hab. Gorontalo (Forsten); Menado (mus. nostr.). The range of this Heron cannot be accurately stated until its conspecies have been studied and defined. Conf. Finsch & Hartl. Faun. Centralpolynes. p. 207, by whom, however, no specific differences are admitted to exist. My Menado example is in full VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 101 breeding-plumage, and conspicuously differs from Indian and Cingalese examples in having the crown and crest dark green, almost black, instead of a much lighter shade of green. Other differences are to be detected, which may not prove constant. For instance, in a Ceylon example, all the wing-coverts, and the four secondary quills nearest the body, are bordered with bright ochreous yellow, and not with white as in the Menado individual. If the Menado bird agrees with the Javan, Mr. Hodgson appears to have been justified in separating the continental form under the title of chloriceps. CICONIID. MELANOPELARGUS, Reichenbach. 178. MELANOPELARGUS EPIscoPus (Bodd.), Tabl. Pl. Enl. (1783), ex Daubent. Pl. Enl. 906. Ardea leucocephala, Gm. Syst. Nat. ed. 13, i. p. 642, “ Coromandel” (1788), ex Buffon, Hist. Nat. Ois. vii. p. 370; Schlegel, Mus. Pays-Bas, Ciconie, p. 11. Hab. Saoussou, in June (Rosenberg) ; Ceylon (mus. nostr.) ; all India, Burma (Jerdon) ; Java, Borneo (Mus. Lugd.); Tropical Africa (Schlegel). Conf. O. Finsch & Hartl. Vog. Ost-Afr. pp. 722, 723. TANTALID. FALcrInELLvus, Bechstein. 179. FauciveLtus 1enEus (S. G. Gmelin), Nov. Comm. Ac. Scient. Imp. Petropol. xv. p- 460, pl. 18' (1771). Numenius viridis, 8S. G. Gmelin, op. cit. p. 462, pl. 19. Tantalus castaneus, P. L. S. Miiller, Syst. Nat. Suppl. p. 112 (1776), ex Marsigli. Falcinellus peregrinus (S. Miller), Mus. Lugd., Bp. Consp. ii. p. 159, “ Celebes, Java” (1857). Ibis falcinellus, Vieill., Schlegel, Mus. Pays-Bas, Ibis, p. 2. Tantalus falcinellus, Linn. 8. N. ed. 12, 1. p. 241 (1766). Hab. Gorontalo, female, moulted, 30th September—male, in almost perfect plumage, 1st October—female in almost perfect plumage, September—male, moulted 30th Sep- tember; Northern Celebes, male in perfect plumage; Celebes, examples in first plumage (orsten); Macassar, female, moulted, March (S. Miiller). I do not venture on the general distribution of the Glossy Ibis, as it is still an open question whether the European, Asiatic, American, African, and Australian races are identical (conf. Bp. J. c.). S. Miiller’s specimen of Inocotis papillosa (‘Temm.), stated by Prince Bonaparte (op. cit. ii. p. 154) to have been collected in Celebes, came from Borneo (conf. Schlegel, op. cit. p. 10). 1 J. F. Gmelin (S. N. p. 649) quotes the thirteenth plate, thus copying a misprint in 8. G. Gmelin’s text. The thirteenth plate represents Caccabis rufa (Linn.). 102 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. ANSERES. ANATID. QUERQUEDULA, Stephens. 180. QuERQquEDULA circta (Linn.), Syst. Nat. ed. 12, i. p. 204. Anas querquedula, L., Schlegel, Mus. Pays-Bas, Anseres, p. 49. Hab. Limbotto, 8th January, male in imperfect plumage—6th and 13th Jannary, females (Von Rosenberg); Europe, Northern Africa, Asia to Island of Formosa. Q. humeralis, Miiller (Verhandl. p. 159), described from examples obtained on the north shores of Java, is not admitted to be distinct by Professor Schlegel. Mareca, Stephens. 181. Marrca cisperirrons, 8. Miiller, Verhandel. p. 159, “Celebes;” Schlegel, Mus. Pays-Bas, Anseres, p. 58. Anas gracilis, Buller, Ibis, 1869, p. 41, “New Zealand.” Hab. Gorontalo, young bird and an adult male; Menado, adult male; Tondano, male; Pegoiat, female, in November (Forsten); Ayer-pannas, 18th August, male, 13th August, female; Panybie, 13th September, female (Von Rosenberg); Macassar, female (Miller); Timor (Miiller); near Port Essington (Mus. Lugd); Australia (Ver- reaux); near Melbourne, 8. Australia (Ferd. Miiller); New Caledonia (Verreaua) ; Flores (Wallace); New Zealand (Buller). Professor Schlegel (/. c.) remarks that Celebean examples are smaller than those from other localities. It is probably this species that Mr. Gould alludes to (B. Austr. 8vo, 11. p. 366) as one of the races of MM. punctata (Cuvier) found in Australia. Denprocyena, Swainson. 182. Denprocyena currata (Forsten), Mus. Lugd.; Schlegel, Mus. Pays-Bas, Anseres, p. 85, “ Celebes.”’ Dendrocygna guttulata, Temm.; Wallace, P. Z. 8. 1863, p. 36. » Miller; Sclater, P. Z.S. 1864, p. 300. Hab. Limbotto, 1st September, adult male; Panybie, 12th September, adult female ; Kema, 24th August, adult male and female (Von Rosenberg); ‘Ternate, Gilolo (Bern- stein); island of Kelang, Amboyna (Hoedt); Bouru, Ceram (Wallace); Goram, Aru, Little Key (Von Rosenberg). 183. DenpRocyGNA VAGANS, Eyton, MS. ; Fraser, Zool. Typica, pl. 68, “ Manilla” (1849) ; Schlegel, Mus. Pays-Bas, Anseres, p. 88; Sclater, P. Z. S. 1864, p. 300. Hab. Tondano, December, adult female (Forsten); Limbotto, 9th January, adult VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 103 male; Gorontalo, 27th May and 20th July, male and female; Ayer-pannas, 11th, 15th, 17th August, males and females; Limbotto, 13th September, female; Pagouat 29th July, a nestling (Von Rosenberg); Macassar, E. Timor (Wallace); Philippines (Cum- ing); Java (Diard); New Caledonia (Verreaux). Conf. Hartl. & Finsch, Centralpolyn. p. 212. According to Mr. Sclater (J. c.) there appear to be three races of this Tree-Duck— the Philippine, which is the type, the Australian (D. gouldi, Bp.), and the one inhabiting Celebes and Timor. It is true that at a later date (P. Z.S. 1866, p. 149) Mr. Sclater maintained that the three races are not separable. Professor Schlegel mentions that this species only occurs accidentally in Java. LARIDZ. STERNINA. HyprocHe.ipon, Boie. 184. HyprocHELIpon nigra (Linn.), Syst. Nat. ed. 12, i. p. 227 (1766); Schlegel, Mus. Pays-Bas, Sterne, p. 31. Hab. Northern Celebes (Forsten, fide Schlegel); Northern Africa ; Southern Europe ; Northern and Eastern Asia. I include this species on Professor Schlegel’s authority. It has not been recorded from any other island of the Malay archipelago, although found throughout China (Swinhoe). Its occurrence in India rests on the evidence of specimens collected by Dr. L. Stewart (Jerd. B. of India, iv. App. p. 875); in what part of India, is not mentioned. 185. HypRocHELIDON LEUCOPAREIA (Natterer), Temm. Man. d’Orn. 2nd ed. ii. p. 746 (1820), “‘ Hungary ;” Verner, Atlas, pl. Sterna grisea, Horsf. Trans. Linn. Soc. xiii. p. 199, “ Java” (1822). hybrida, Pallas, Zoog. Rosso-Asiatiea, ii. p. 338 (1831), “ Volga; ” Schlegel, Mus. Pays-Bas, Sterne, p. 32. Viralva indica, Stephens, Shaw, Gen. Zool. Aves, xiii. p. 169, “ India” (1825), ex Lath. Gen. Hist. x. p. 103. no. 5. Sterna delamotta, Vieill. Ency. Méth. Orn. Add. i. p. 350, “‘ Europe” (1823). similis, J. E. Gray, Illust. Ind. Zool. i. pl. 70. f. 2, ex India (1832). Hydrochelidon fluviatilis, Gould, P. Z. S. 1842, p. 140, “interior of New South Wales.” delalandii, Bp. Compt. Rend. xlii. p. 773, “Cap. B. Spei” (1856). Hab. Lake of Gorontalo, 8th of October, female passing into winter plumage (orsten) Pontianac, in Borneo (Diard); abundant in India (Jerdon); Ceylon (mus. nostr.); S.-W. Formosa (Swinhoe); interior of Australia (Gould); Cape York (mus. nostr.); Java (Horsfield); South and South eastern Europe; Northern and Western Africa. 104 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. Onycuoprion, Wagler. 186. OnycHoPpRION MELANAUCHEN (Temm.), Pl. Col. 427, “Celebes” (April 25, 1827); Schlegel, Mus. Pays-Bas, Sterne, p. 28; Finsch & Hartl. Centralpolyn. p. 224; S. Miiller, Verhandel. Land- en Volkenk. p. 125. Sterna sumatrana, Raffles, Trans. Linn. Soc. xiii. p. 829, “Sumatra” (1822). Hab. Celebes (Reinwardt); Ternate, Morty, Raou (Bernstein); west coast of New Guinea (8. Miller); common in the straits and bays of the Lobo district (S. Miller); Bay of Bengal (Jerdon); Nicobars (Blyth); Andamans (Walden); New Caledonia (Verr.); Loyalty Islands (G. R. Gray); Viti-Levu, Ovalu, Stewart Islands (Finsch & Hartl.); Cape York (Macgillivray); Sumatra (Raffles). An Andaman and a New-Guinea example in my collection are identical. There can be little doubt that this is the S. swmatrana, Raffles. A title, most inap- propriate, but which has priority. 187. ONYCHOPRION ANZSTHETUS (Scopoli), Del. Fl. Faun. Insub. ii. p. 92. no. 72 (1786), ex Sonnerat, Voy. Nouv. Guin. pl. 84. Sterna panayensis, Gm. Syst. Nat. ed. 13, i. p. 607 (1788), ex Sonnerat, /. c. panaya, Gm.; Finsch & Hartl. Orn. Centralpolyn. p. 228; Vogel Ost-Afr. p. 833. Hab. (Salayer) Celebes (Wallace). An example of this species, in young plumage, was obtained by Mr. Wallace at Salayer. For complete synonymy and distribution conf. O. Finsch & Hartl. (/. ¢.). PELEcANoPUS, Wagler. 188. PrLEcanorus mepDIvS (Horsf.), Trans. Linn. Soc. p. 199, “ Java” (1822). Sterna affinis (Cretzsch.), Riipp. Reise, p. 23, pl. 14, ‘coasts of the Red Sea” (1826) ; Schlegel, Mus. Pays-Bas, Sterne, p. 6. bengalensis, Cuv. Mus. Paris; Less. Tr. p. 621. no. 9, “cdtes de Inde” (1831); Pucheran, Rey. et Mag. Zool. 1850, p. 512; conf. Jerd. Birds Ind. iu. p. 843; Blyth, Ibis, 1865, p. 39; op. cit. (1867) p. 177. Thalasseus torresi, Gould, P. Z. 8S. 1842, p. 140, “ Port Essington.” Hab. Macassar, March (S. Miiller); North Celebes (Forsten). If Mr. Blyth’s identifications of the North-African and South-Asiatic species be correct, the range of this Tern extends from Sicily to Madagascar, the coasts of India, of Northern Australia, and the islands of the Malay archipelago, at least as far as Celebes. It must be remembered, however, that Dr. Pucheran (/. ¢.) has stated that the Abyssinian and Bengal species differs ‘‘ par plus de noir dans les rémiges et par son bec moins courbé, et par cela méme plus droit.’ (Conf. Finsch & Hartl. Vogel Ost-Afrika’s, p. 830.) VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 105 189. Peecanopus cristatus (Stephens), Shaw, Gen. Zool. contin. Aves, xii. p. 146, “ China, many of the south-eastern islands of Asia” (1825). Sterna pelecanoides, King, Surv. Intertrop. Austr. ii. App. Aves, p. 422 (1827) ; Schlegel, Mus. Pays-Bas, Sterne, p. 9. velox, Riipp.; S. Miiller, Verhandel. Land- en Volkenk. p. 125. — bergii, Lichtenst.; Finsch & Hartl. Centralpolyn. p. 216, part. Hab. Celebes, female, winter plumage (Forsten); Batchian, Gilolo, Morty, Oby (Berstein); Ceram (Forsten); Timor, west coast of New Guinea (S. Miller); Flores (Semmelink); coast of W. Australia, Port Essington, Torres Straits (Gould); Mysol (Wallace); Java (S. Miller); mouth of the Hoogly, Madras, Malabar coast (Jerdon) ; Southern China, Formosa (Swinhoe). A race of this species, probably belonging to the Asiatic form, inhabits many of the islands still more to the eastward. Messrs. Finsch & Hartlaub (/. ¢.) have united the large Sea-Terns of Eastern Africa (S. velox, Riipp.) and of South Australia and Van Dieman’s Land (Thalasseus poliocercus, Gould) with the Asiatic and North-Australian species, under the title of S. bergii, Lichtenstein, bestowed on a Cape-of-Good-Hope individual. Professor Schlegel, in his masterly catalogue (J. c.), keeps these representa- tive forms separate, but with much reluctance. With the exception of 7. poliocercus, the facts known favour the opinion that the species are severally permanent residents in the localities they frequent. Dr. S. Miiller, who identified the New-Guinea bird with the Red-Sea S. velox, Riipp., states (/. c.) that it is known to the inhabitants of the straits and bays of the Lobo district by the name of Ressa. Ata subsequent date he appears to have regarded it as a distinct species; for Professor Schlegel cites Sterna ressa, S. Miller, as a synonym. PODOCIPID. Popicrrs, Latham. 190. Popiceps mrvor (Gm.), S. N. ed. 13, i. p. 591 (1788); Schlegel, Mus. Pays-Bas, Urinatores, p. 45. Hab. Panybie (Von Rosenberg). According to Professor Schlegel the Little Grebe of Java and Celebes is identical with that of Europe; and provisionally I refer the Celebean bird to the European species. But judging from the few South-Asiatic examples I have been able to compare with European individuals, I am not prepared to admit their identification as absolute. For instance, the Javan Little Grebe has an exceedingly stout bill which measures seven eighths of an inch in length, the wing being four inches and a quarter. The cheeks, chin, and throat are black ; and a dark ferruginous line starts from behind each eye, and extends down each side of throat. Thus the Javan bird closely resembles the Australian P. gularis, Gould—a species, however, which Professor Schlegel does not admit. VOL. VIII.—PART I. May, 1872. R 106 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. PELECANID. Dysporvs, Iliger. 191. Dysporus suxa (Linn.), Syst. Nat. ed. 12, i. p. 218, “ Pelago indico” (1766). Sula fiber, G. R. Gray ; Schlegel, Mus. Pays-Bas, Pelecani, p. 41. Hab. Celebes (Mus. Lugd.). On the general distribution of this species, conf. Finsch & Hartl. Faun. Centralpolyn. p. 261. PHALAcRocoRAX, Brisson. 192. PHALOCROCORAX MELANOLEuCUS (Vieill.), Nouv. Dict. d'Hist. Nat. vol. viii. p. 88, New Holland (1817); Gould, Birds Austr. vii. pl. 70; Schlegel, Mus. Pays-Bas, Pelecani, p. 15. Hab. Celebes, nuptial plumage (mus. nostr.); Gorontalo, imperfect plumage (Forsten); Gilolo, Timor (Mus. Lugd.); Tasmania, every part of Australia (Gould); Salwati (mus. nostr.). Piotus, Linneus. 193. PLoTus MELANOGASTER (Forster), Zool. Ind. p. 22, pl. xii. “ Java, Ceylon” (1781). Hab. Menado (mus. nostr.); all India, Ceylon, Burma (Jerdon); Java (Mus. Lugd.) ; Australia (Schlegel). A single example of a Plotus in adult male plumage is in my possession, collected near Menado. It does not appear to differ from Indian examples. P. nove-hollandie, Gould, P. Z. 8. 1847, p. 34, is not admitted as distinct from P. melanogaster by Professor Schlegel. Mr. Gould relies on its shorter scapularies and larger size. List of species stated by various authors to occur in Celebes for which there is not sufficient authority :— Urospiza torquata (Cuv.), Hand-list Birds Brit. Mus. no. 327. Scops mantis, J. Miiller, op. cit. no. 477. Halcyon diops, Temm., op. cit. no. 1107. Halcyon funebris, Forsten, op. cit. no. 1126. Philemon moluccensis' (Gm.), op. cit. no. 2074. Philemon inornatus, G. R. Gray, op. cit. no. 2077. Philemon? collaris, Reichenbach, op. cit. no. 2083. The genus Philemon is un- known in Celebes. Climacteris leucophea (Lath.), op. cit. no. 2521. The genus Climacteris is. un- known in Celebes. Dicrurus bimaénsis, Temm. op. cit. no. 4212. Dicrurus atroceruleus, G. R. Gray, op. cit. no. 4220. * This species is a Mimeta (conf. Wallace, P. Z. 8, 1863, p 26). VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 107 Drymophila alecto, Temm. op. cit.no. 4794. The genus Drymophila is unknown in Celebes. Lalage aurea, Temm. op. cit. no. 5114. Calornis metallica, Temm. op. cit. no. 6376. Munia pallida, Wallace, op. cit. no. 6756. Eos cochinsinensis (Lath.), op. cit. no. 8202. The genus Hos is unknown in Celebes. Psittacus cyanicollis, 8. Miiller and Schl. op. cit. no. 8275. Ptilinopus flavicollis, G. R. Gray, op. cit. no. 9125. Ptilinopus xanthogaster (Wagler), op. cit. no. 9136. Pitilinopus hyogaster (Reinw.), op. cit. no. 9144. Macropygia leptogrammica (Temm.), op. cit. no. 9305. Reinwardtena reinwardti (Temm.), op. cit. no. 9310. The genus Retnwardtena is unknown in Celebes. Most of the erroneous habitats enumerated in the above list are transcribed from the older authors. In nearly every instance they have been corrected by more recent writers, especially by S. Miiller, Schlegel, Wallace, and O. Finsch. DESCRIPTION OF THE PLATES. PLATE III. Outline Map of Celebes and the adjoining islands, p. 23. PLATE IV. Trichoglossus meyeri, p. 32. From a specimen in Lord Walden’s collection. PLATE V. Buceros exaratus,1¢,22,p.47. From specimens in Lord Walden’s collection. PLATE VI. Fig. 1. Artamus monachus, p. 67. From a specimen in Lord Walden’s collection. Fig. 2. Geocichla erythronota, p.61. From the typical specimen in Mr. A. R. Wallace’s collection. PLATE VII. Fig. 1. Myialestes helianthea, p. 66. Fig. 2. Hypothymis puella, p. 66. bem specimens in Mr. A. R. Wallace’s collection. Fig. 3. Cyornis rufigula, p. 66. . Volvocivora morio, p. 69. . Lalage leucopygialis, p. 69. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. PLATE VIIL. \ From specimens in Lord Walden’s collection. PLATE IX. . Munia brumneiceps, p. 73. . Zosterops intermedia, p. m=} From specimens in Mr. A. R. Wallace’s collection. . Zosterops atrifrons, p. 72. PLATE X. . Acridotheres cinereus, p. T7. From a specimen in Lord Walden’s collection. . Aigialites peronii, p. 90. From a specimen in Mr. A. R. Wallace’s collection. —— ey, pansLool S00 Lol G IA; W3 Outline Map of @EEEBES and the ADJOINING ISLANDS feenghir Islands Tongicaaa Tare feck Groomnich Gront Lob pe WU GEG M&W Hanhart ump - J Sot lith TRIG AOGLO SSIS MMS Naa ise © phages EA rant Dobe LEV M&.N Hanhart anxp J Smatdith INDAUAIENE, SIS: TRG IIWAING I IeLich 7 ISNA TaNaMons; JP INeLyLWAN CACCMECIN| INS; TUNE (EAUUILvaS Wp J Smat hth | 9 G VOLVO CIVORA MORIO LALAGE LEUCOPYGIALIS M & N Hanhart imp Nw VAS Ey UN IN FE iC HP'S: Ee On) LIN SISNeT TEROPS ATRIFRONS M& N Hanhart imp — [ 109 ] III. Appendix to a List of Birds known to inhabit the Island of Celebes. By AntHUR, Viscount Waupen, F.R.S., President of the Society. Read May 7th, 1872. [Puates XI. to XIII.] I. Additional Observations on the Birds included in the previous List. WHILE the List of Celebean Birds contained in the preceding pages was passing through the press the island of Celebes was being visited and its zoology investigated by a most indefatigable collector and naturalist, Dr. Bernhard Meyer. With the greatest liberality Dr. Meyer has permitted me to examine all the birds collected by him in Celebes; and I avail myself of this opportunity to thank him for his courtesy. The additional materials thus placed at my disposal have enabled me to add to the list several species which had not previously been known to inhabit Celebes, as well as a few more which were altogether new to science. The considerable number of examples, repre- senting the rarer species, collected by Dr. Meyer, has also rendered it possible and desirable to add some supplementary observations. The greater part of the collection was made in North Celebes, and consequently on old ground; yet Dr. Meyer has added twelve species new to the island, and at least four of which were previously undescribed. In the Togian islands a small collection was likewise made, showing that these islands, as we might have naturally supposed, possess a generally Celebean ornis; yet, among the small number of species thus obtained, two’ were new to science, and have not as yet been discovered on the mainland of Celebes. TERASPIZA RHODOGASTRA, anted, p. 33. (Plate XI.) Three stages of immature plumage, hitherto undescribed, are represented by three individuals obtained in North Celebes. One, a male (Pl. XI.), has the head dark brown, mixed with rufous. The back, wings, and tail are bright rufous. The nuchal and dorsal feathers are centred with dark brown. The wing-coverts have each a black subterminal drop. The secondary quills are crossed by five distinct black bands. The basal halves of the primaries are banded with brown and pale rufous alternating; the terminal halves are light brown, obscurely banded with dark brown, rufous replacing the light brown on the outer webs. On the under surface of the quills the dark brown bands are better defined and more conspicuous. ‘The five middle pairs of rectrices have four broad black bands, besides 1 Loriculus quadricolor, Walden, Ann. & Mag. N. Hist. ser. 4, vol. ix. p. 398 ; and Criniger aureus, Walden, 1c. p. 400. VOL. VIII.—PART 11. May, 1872. § 110 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. an obscure brown band at the root of the feathers. The outer pair have seven bands. The plumage of the under surface of the body is fulvous, each feather with a bold brown longitudinal central stripe. The under tail- and shoulder-coverts are unspotted fulvous. The middle toe is very long; and the tail is conspicuously forked. The second example is of a young female, much resembling the male above described, but having bold brown drops on the under shoulder-coverts and axillaries, and the general colouring of the upper surface not quite so bright a chestnut. In it also the tail is not forked, and the outer pair of rectrices are shorter than the middle. The third example is of a young female passing from the chestnut plumage of no. 2 into that of the adult. The nuchal feathers are ashy; and a few similar plumes are interspersed on the throat and upper part of the breast. The breast-feathers and a few on the flanks are pure vinous red. A few of the upper tail-coverts are dark ashy; and one of the long wing-coverts has come in ash-coloured, and with two pure white spots on the inner web. The chestnut colouring of the remainder of the plumage is very dingy and faded. The tail is not forked. Dimensions. | Rectrices. Toes. Wing. |——___—_| Tarsus. Middle.| Outer. Outer. | Middle.| Inner. inches. | inches. | inches. | inches. | inch. inch. inch. 7-50 | 6°75 | 663 | 2°25 | 113 | 162] -75 750 | 650 | 6:12 | 2:13] 1:12} 1:50] -75 6 5:12 | 5:50 | 1:87 ‘81 | 1:13) +50 @. Changing into adult plumage. 2. Immature. 3. Immature. The toes are measured without the nails. The nails of the inner toe and hallux are very large, strong, and equal. Those of the middle and outer toes are slender and short. TACHYSPIZA SOLOENSIS (Horsf.), anted, p. 34. This species also inhabits China, extending at least as far north as Pekin (Swinhoe, (P..Z.S. 1871, p. 342). LIMNAETUS LANCEOLATUS, Bp., anted, p. 34. Four examples from North Celebes have reached me—two (male and female) fully adult, and two (male and female) in the immature plumage already described (/. s. ¢.). The adult pair do not differ; and the example of the immature female only differs from that of the male by having the two pairs of middle rectrices more frequently banded and in a different manner. In the adult birds of both sexes the middle rectrices have a broad, terminal, dark brown band ; then, above, a broader band of pale greyish brown, and then three narrow dark brown bands separated by broad pale bands. This is also VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 111 the character of the banding on the middle rectrices of the immature male. But in the immature female there is no terminal dark brown band, and the middle rectrices are almost evenly divided by seven pale and seven dark-brown bands. Yet in all other respects the immature pair are identical in plumage. PERNIS CELEBENSIS. Pernis ptilorhyncha (Temm.), antea, p. 36. An examination of several examples of the Celebean Honey-Buzzard has convinced me that it is distinct from the Indian and Javan species. In this view I am only con- curring with both Messrs. Gurney and Wallace, and therefore propose the above title for it. The remarkable resemblance of this species to Limnaétus lanceolatus, in adult plumage, has been commented on by Mr. Wallace and Professor Schlegel. Upper surface brown. Chin, throat, and cheeks white, each feather broadly centred with dark brown. Breast pale rufous, some of the feathers with brown central stripes. Abdominal and ventral region, flanks, under wing- and tail-coverts, and the thigh-coverts white, with two, three, or four broad transverse bands. Tail crossed by three broad dark-brown bands, one being terminal; between the terminal band and the next a broad, light greyish-brown band of irregular shading and marking ; between the second. dark-brown band and the third a paler brown band. Yuneipicus TEMMINcKU (Malherbe), anted, p. 41. Dr. Meyer has sent a male as well as several females of this rare species. The male, hitherto unknown, is peculiar in having the sides of the neck blood-red instead of a narrow stripe behind the eyes. In other respects it exactly resembles the female. Notwithstanding Bonaparte’s remark (Consp. i. p. 137, no. 20), this species in no way resembles Y. kisuki. It is an isolated form, readily distinguished by its olive-brown plumage, spotted on the wings with yellowish-white dots, by its fulvous upper tail- coverts and rump, and by all its rectrices being barred rufous and brown. MEROPOGON FoRSTENI (Temm.), anted, p. 42. This species has the first primary half the length of the second, which is a little shorter than the third. The third and fourth are longest, and equal. ‘The fifth is somewhat shorter than the third and fourth, but longer than the second. In the structure of the wing, therefore, it differs from both Merops and Melittophagus', but agrees with Nyctiornis. The grooved culmen of Nyctiornis is not present ; but a shallow channel extends from the base of the maxilla, on both sides of the culmen, for two 1 Prince Bonaparte says (Consp. i. p. 164), “ale Melittophag ;” but in Melittophagus, M. minutus being the type, the third quill is the longest. The African species which most resembles MW. forsteni in the graduation of the quills and the form of the rectrices, the middle pair excepted, is M. bullockoides, Smith. s2 112 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. thirds of its length. This character is not possessed by either Myctiornis, Merops, or Nelittophagus. The rectrices are truncated, as in Nyctiornis; but the middle pair are elongated, as in Merops, and closely resemble in form and proportion those of MW. philip- pensis. ‘The feet are those of the family. ‘The elongated pectoral plumes resemble in character the same feathers in Nyctiornis. Altogether M. forsteni may be regarded as a link uniting Nyctiornis to Merops, but most nearly allied to Nyctiornis. CrycopsiIs FALLAX (Schlegel), antea, p. 45. Several examples were obtained in North Celebes by Dr. Meyer. Tanysiptera riedelii was not obtained in Celebes (conf. P. Z. 8. 1872, p. 1); and as yet there is no evidence that the genus occurs in the island. LyNCORNIS MACROPTERUS, Bp., anted, p. 47. This species, LZ. macrotis, Vigors, and L. temmincki, Gould, are representative forms, closely resembling each other in plumage, but differing in size, the Celebean species being a little smaller than the Philippine. L. cerviniceps, Gould, the giant of the genus, differs considerably in colouring and markings. EUDYNAMIS MELANORHYNCHA, Miill., antea, p. 53. This species also passes through a rufous phase of plumage. The entire upper surface, in one individual, is deep bay, each feather being traversed by broad and per- fectly regular black bands. The chin, throat, and cheeks darker bay, with longitudinal black central streaks. A broad white stripe from the rictus to the neck. Lower plumage fulvous, with narrow, crooked, transverse markings. Edge of shoulder white. CENTROCOCCYX AFFINIS (Horsf.), anted, p. 56. CENTROCOCCYX JAVANENSIS (Dumont), anted, p. 60. I have had an opportunity of examining a large Celebean series of both these species, and find that they do not differ from Javan individuals. BRODERIPUS CELEBENSIS. Broderipus coronatus (Swains.), anted, p. 60. Of fourteen examples of the Celebean Broderipus collected by Dr. Meyer, seven have the black coronal ring complete, and seven incomplete. One of the latter exhibits faint traces of yellow at the tips of the lesser wing-coverts. The remaining thirteen speci- mens are without any indication of a wing-spot. None have the middle pair of rectrices completely black, although in one example they are nearly so; and yet it shows no wing-spot. ‘he series illustrates the progress of the coronal ring before uniting. In one individual the black loral mark of Oriolus galbula extends behind the eye somewhat VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 113 further than what is found in O. kundoo; in others it has extended still further, until it is found encircling the head. Unfortunately Dr. Meyer has not noted the localities of the several individuals, and we are left in doubt whether there are two species or one; but, from the graduations the coronal ring exhibits, it seems likely that there is but one species. This gradual development of the coronal ring has not been observed in any other species of Broderipus, and, taken together with the almost total absence of the wing-spot, separates the Celebean Oriole from all known species. GEOCICHLA ERYTHRONOTA, Sclater, anted, p. 61. - Several examples were collected in North Celebes. TRICHOSTOMA CELEBENSE, Strickl., anted, p. 62. Obtained in North Celebes. ARTAMUS MoNACHUS, Temm., anted, p. 67. Celebean examples are identical with those obtained in the Sula Islands. GRAUCALUS TEMMINCKII, S. Miiller, anted, p. 68. (Plate XII.) A few examples of this rare species were obtained in North Celebes by Dr. Meyer. Corvus encA (Horsf.), antea, p. 74. Two examples sent by Dr. Meyer from Celebes give me the opportunity of comparison with the Javan species. Iam unable to detect any difference, except in the dimensions, the Javan bird being somewhat the largest. C. validus, Temm., as represented at Malacca is a very distinct species. Dimensions. . Bill : Wing. from forehead. Tail. aT rrag inches, inches, inches. C. enca, ex Java.......... 11°75 2-50 6-50 Spe te gece ete ria 11-50 2-50 6:37 >» eX Celebes........ 10-50 2-12 5:50 | is Bee A 11-12 2-18 5:75 CaLORNIs NEGLECTA, Walden, anted, p. 79. Numerous examples of this species were obtained by Dr. Meyer in Celebes, thus fully establishing its Celebean habitat. OsMOTRERON VERNANS (Linn.), anted, p. 81. On examination of a large series of the Celebean form from North Celebes I find 114 VISCOUNT WALDEN ON THE BIRDS OF CELEBES, that the grey cap is paler than in Malaccan examples. The greenish tinge on the throat is common to individuals from both localities. I can find no difference between the amount of lilac on the neck (conf. Wallace, Ibis, 1863, p. 320). Mr. Maingay obtained the male and female of 0. dicincta at Malacca. CHALCOPHAPS STEPHANI, Jacq. & Puch. anted, p. 85. A single example of this rare and well-marked species has been sent from North Celebes by Dr. Meyer. It is in full plumage, and in every respect agrees with the plate and description (.s.¢.). If then the New-Guinea species differs, the origin of the type must have been Celebes, and not New Guinea (west coast) as stated by M. Pucheran (J. c.). CHALCOPHAPS INDICA (Linn.), anted, p. 86. Examples from North Celebes in no way differ from Ceylon, Indian, Burman, Malaccan, and Javan individuals. Judging from the number of specimens obtained by Dr. Meyer, this species cannot be so rare in Celebes as stated by Professor Schlegel ((. s. ¢.). Nycticorax caALEDoNicus (Gm.), anted, p. 100. It may be inferred that this species breeds in Celebes, an example in spotted imma- ture plumage having been obtained in the northern part of the island by Dr. Meyer. Il. List of Species to be added to the Celebean Avifauna. Genus Caprimu.eus, Linn. 1. CaprimuLeus AFFINIS, Horsf. Trans. Linn. Soe. xiii. p. 142, “ Java” (1820). Caprimulgus europeus, L. ap. Raffles, op. cit. p. 315, in part. “ Sumatra.” Hab. Java (type); Sumatra (Raffles); East Timor, Lombock (Wadllace); Celebes (Meyer). Dr. Meyer has sent from Celebes a single example of a Caprimulgus which appears to belong to the species cited above. I have been unable to compare it with a Javan individual; but it perfectly agrees with Lombock and East-Timor specimens. Four pairs of rectrices are missing ; and it is otherwise in indifferent order; I therefore adda short description of the species, taken from a Lombock individual—Horsfield’s account, the only one published, being very meagre. Above, the general aspect of the plumage is iron-grey, somewhat mixed with brown, caused by the feathers being finely dotted or sprinkled with black and grey, here and there with fulvous. In some of the crown-feathers black prevails; but there are no regular stripes on the head. On the sides of the throat are two white spots. The lower breast-feathers are fulvous, with several well-defined brown transverse bars. The VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 115 ventral region and the under tail-coverts are fulvous, without any markings. The major wing-coverts are distinctly banded with alternate rufous and brown. A large white spot on each of the first four primaries. The chin, throat, and upper breast are clothed with feathers finely marked with fulvous-grey points on a brown ground. Many of the upper breast-feathers with bold rufo-fulvous tips. The middle rectrices have the general colouring and marking of the upper plumage, and are traversed by eight or nine more or less distinct irregular black bands. The two outer pairs are pure white throughout their entire length. The wing in six examples averages 6:25 inches, and the tail 4°25. This species, as has been well observed by Mr. Blyth (Cat. Calc. Mus. p. 84, note), is as diminutive as C. monticola, Frankl. 2. CAPRIMULGUS, sp. ? A large dark-coloured Caprimulgus is among the novelties obtained in Celebes by Dr. Meyer. The example is unfortunately in such indifferent order that the inherent difficulties which attend the discrimination of many species of the family are very much increased. Above, this Celebean Goat-Sucker closely resembles Javan examples of C. macrourus, Horsf. Underneath, it is darker in colour, and the transverse barring of the abdominal plumage is less regular and well defined. The principal points in which it differs from a considerable series of C. macrourus are:—the great length of the rictal bristles, which measure a full inch; the greater length of the bill; the smallness of the terminal white spots on the two outer pairs of rectrices; and the peculiar markings on the under surface of all the rectrices except the middle pair. In true C. macrourus the white terminal spot on the outer pair of rectrices measures about two inches; in this individual it measures only seven eighths of an inch. In the Javan bird the under surface of the rectrices is more or less uniform brown, without markings; this Celebean bird has some eleven or twelve distinct narrow rufous bars crossing the rectrices. Its first primary has no white spot, while on each of the three next it is much smaller than in €. macrourus, which has a large white spot on all four primaries. The length of wing is equal to the average length observable in the Javan bird, seven and a quarter inches; nor do the dimensions of the tail (six inches) differ. As there are two Bornean species, C. arundinaceus, Jacq. & Puch., and C. dinotatus, Bp., which have yet to be satisfactorily identified, I refrain from conferrmg a distinct title on this Celebean Nightjar. Cucutus, Linn. 3. CuctuLus canorvs, Linn. S. N. i. p. 168 (1766). A single specimen obtained by Dr. Meyer in North Celebes is not to be distinguished from a British-killed example of the adult Common Cuckoo. The wing alone is shorter, 7-50 against 8°31; the tail is equal. Two other specimens, with the upper 116 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. plumage changing to the adult stage, the transverse, pectoral, and abdominal bands rather broader, and with immaculate buff under tail-coverts, seem to belong to the same species; and a fourth, in bright chestnut and brown plumage, must be referred to it. Without the example in full plumage it would have been difficult to say whether the other three did not belong to C. canoroides, Miiller. If C. canoroides is equal to C. saturatus, Hodgs., =C. himalayanus, ap. Jerd., it is a very distinct form; but I have never met with an Archipelagic Cuckoo in the dark adult plumage of Himalayan C. saturatus. Timor and Amboyna examples of so-called C. canoroides only differ from those of C. canorus by having a shorter wing. But individuals of C. canorus from different parts of the Old World (that is, individuals identical in plumage) vary extremely in the length of wing, as the following table shows :— Win olan digererweperctateteteasrelsverieiciere eterereicy veers 8:31 INVERN so agg onbb 000s DodOdEdOSena aE 9-00 Menado:. Racine wstiete sale tois.< sin lorelstereroke 7-50 | These measurements are taken from DeyrasD oonvewcepereyereereierers te eial-i smite) et 8:00 examples in fully adult plumage, oc Ee oboe Arcee ie ier cas 8°75 and almost identical in colouring Indias cjceact SHAG HMO SORE DE OOO OTOE 9-50 and marking. IMM MIES sasooecesososeaeoedu O90 05 9-25 Un, So pSsoiaeasosoascdednoacadoucat 8-50 Hirrococcrx, S. Miller. 4. Hrerococcyx crasstrostris, Walden, Ann. & Mag. Nat. Hist. ser. 4, vol. ix. p. 305, “ North Celebes” (1st April 1872). (Plate XIII.) This species, in mature plumage, most nearly resembles C. micropterus, Gould. 5, YCACOMANTIS SEPULCRALIS (Miiller), Verhandel. p. 177, not., sp. 2, “Java, Sumatra.” For the present I refer three examples of a Cacomantis obtained by Dr. Meyer in North Celebes to the Javan species, rather than create a new title; for without a large series of individuals inhabiting all parts of the archipelago it is impossible to discriminate the species belonging to this perplexing group. One of the three Celebean examples is in fully mature plumage, and has the chin, cheeks, and throat pale grey, the head iron-grey, the upper plumage deep bronze- green, the breast, abdominal region, flanks, under tail-, and shoulder-coverts deep rufous; the middle pair of rectrices are black, the outer one black-brown tipped with white, and with one or two small white shallow triangular marks on the edge of the inner webs; the quills are traversed by the usual white band. Wing 4:25 inches, tail 5°75. These Celebean individuals differ from all examples of the Javan C. sepulcralis known to me in the much deeper bronze-green of the upper plumage, the much deeper rufous of the under, and in their shorter wings and tail. VISCOUNT WALDEN ON THE BIRDS OF CELEBES. 117 6. CisticoLa @RAyI, Walden, Ann. & Mag. Nat. Hist. ser. 4, vol. ix. p. 400, “ North Celebes” (1st May 1872). CoryDALLA, Vigors. 7. CoryDALLA eustavi (Swinh.), P. Z. S. 1863, p. 90, “ China;” op. cit. 1871, p. 366. Pipastes batchianensis, G. R. Gray, H.-list, i. p. 251. no. 3642, “ Batchian” (1869). Anthus arboreus, var., G. R. Gray, P. Z. S. 1860, p. 350, ‘ Batchian.” Hab. Amoy, China (Swinhoe); Batchian (Wallace); North Celebes (Meyer). A single individual from North Celebes agrees well with the Batchian example contained in the British Museum, and there identified by Mr. Swinhoe. 8. Cyornis BANyUMAS (Horsf.), Trans. Linn. Soc. xiii. p. 146, “ Jaya” (1820); Zool. Res. in Java, pl. —. Muscicapa cantatriz, Temm. Pl. Col. 226, ¢, 2 (1823). ? Muscicapa rufigastra, Raffles, Trans. Linn. Soc. xiii. p. 312, “ Sumatra.” Hab. Java (type); Sumatra? (Raffles); Banjarmassing (mus. nostr.); North Celebes (Meyer). Undistinguishable from Javan examples. HYLoTEeRPE, Cabanis. 9, HYLOTERPE SULFURIVENTRA, Walden, Ann. & Mag. Nat. Hist. ser. 4, vol. ix. p. 399, ** North Celebes” (1st May 1872). Hab. North Celebes (Meyer). Myzometa, Vigors & Horsfield. 10. Myzometa cHLoRopTeRa, Walden, Ann. & Mag. Nat. Hist. ser. 4, vol. ix. p. 399, “ Celebes” (1st May 1872). The discovery of this species adds another Papuan genus to the Celebean fauna. GLAREOLID. GLAREOLA, Brisson. 11. GLAREOLA GRALLARIA, Temm. Man. d’Ornith. vol. ii. p. 503, ‘Austral-Asie” (October 1820); Gould, B. of Austr. vol. vi. pl. 22. Hab. Moreton Bay (Gould); west Coast of New Guinea (Miller); Obi-major (Bernstein); Flores (Semmelink); Borneo (Schwaner); Java (Kuhl and V. Hasselt) ; Celebes (Meyer). Several examples in mature and immature plumage were obtained in Celebes by Dr. Meyer. VOL. viil.—PartT u. May, 1872. T 118 VISCOUNT WALDEN ON THE BIRDS OF CELEBES. STERNULA, Boie. 12. SrerNvLA minvuta (Linn.), S. N. vol. i. p. 228 (1766). Several examples from North Celebes have been sent by Dr. Meyer. They are all in full plumage. ‘Tail pure white. These twelve additional species thus raise the number of authentically recorded Celebean birds to two hundred and five. DESCRIPTION OF THE PLATES. PLATE XI. Teraspiza rhodogastra, p. 109. From a specimen in Lord Walden’s collection. PLATE XII. Graucalus temminckii, p. 113. From a specimen in Lord Walden’s collection. PLATE XIII. Hierococcyx crassirostris, p. 116. From specimens in Lord Walden’s collection. = M2N Hanhart J, Smit.ith . GRAUCALUS TEMMINCKI J. Smt lith mF rN t=) oO (@) [eunoro] IV, On Dinornis (Part XVII.): containing a Description of the Sternum and Pelvis, with an attempted Restoration, of Aptornis defossor, Ow. By Professor Owen, F.RS., F.LS., &e. Read June 6th, 1871, [Puates XIV. to XVI.] THE nearly complete sternum and pelvis of Aptornis defossor about to be described, in addition to the evidences of that bird submitted to the Zoological Society in a former Memoir’, encourage me to attempt a restoration of the skeleton of the extinct species (Pl, XVI.), with which I propose to conclude the present Memoir. The sternum, figured of the natural size in Pl. XIV. figs. 2, 3, 4, presents a close family, if not generic, resemblance to that described and figured in my fourth Memoir on Dinornis’, with regard to which, recognizing in it unequivocal characters of a Ralline type, closely resembling that of the wingless Wekas (Ocydromus), but of much larger size, I had doubts whether to refer the bone to the Notornis (subsequently recognized as a still existent species), or to the still larger Aptornis otidiformis, which is most probably extinct. In Tribonyx ventralis the length of the sternum is 2 inches, that of the femur is 2 inches 3 lines. In Ocydromus australis the length of the sternum is 2 inches, that of the femur is 2 inches 8 lines. In Notornis* the length of the femur is 4 inches 10 lines; in Aptornis otidiformis* the length of the femur is 6 inches 3 lines. The length of the sternum described p. 18, and figured in pl. 4. figs. 5-8, in Zool. Trans. vol. iv., is 4 inches 6 lines, measured from the foremost part of the coracoid groove to the tapering hind end, which is not entire. I do not suppose, however, that the missing portion would exceed 6 lines in length; and it might probably be less, An addition of 4 lines would make the sternum in question equal in length to the femur of Notornis, and 11 inch shorter than the femur of Aptornis otidiformis. I select the femoral bone of the hind leg for this test comparison, because it varies less in proportion to the trunk or general size of the bird than the more distal elements of the limb. A comparison of the metatarsus of Notornis with that of Aptornis instructively illustrates this principle’, The contiguity of the bones of Aptornis defossor here described, when discovered, 2 On Dinornis, part xv., Trans. Zool. Soc. vol. vii, p. 353, 3 Trans. Zool. Soc. vol. iv. p. 17, pl. 4, figs. 5-8, 3 Th. vol. iv. p. 12, pl. 2. fig. 3 $ Tb. vol. iv, p. 10, pl. 3. fig. 3, 5 «On the Archetype and Homologies of the Vertebrate Skeleton,’ 8vo, 1848, p. 200, note, p. 202, VOL. VIII,—PART 11. September, 1872. U 120 PROFESSOR OWEN ON THE GENUS DINORNIS. enables me to refer with confidence the sternum to that species, most probably to the same individual as the pelvis; and the comparison of the sternum with the femur described and figured p. 371, pl. 43. figs. 5, 6, 7, Zool. Trans. vol. vii., gives the fol- lowing proportions of length:—sternum 6 inches 6 lines, femur 7 inches 6 lines. The length of the sternum is here taken from the fore border of the coracoid groove to the tapering hind end, which is all but entire. The inferiority of size of the sternum figured in vol. iv. pl. 4. figs. 5-8 to the sternum of Aptornis defossor is greater than that of the femur of Aptornis otidiformis as com- pared with that of Apt. defossor. But the relative size of the smaller sternum to the femur of Aptornis otidiformis is more like the relative size of the sternum of Aptornis defossor to the femur of that species than is the relative size of the smaller sternum to the femur of Notornis. Seeing therefore, in the two existing Rallines here compared, that as the wings decrease and the legs increase in size the sternum becomes shorter in proportion to the femur, I am inclined to believe that the smaller sternum (Trans. Zool. Soe. vol. iv. pl. 4. figs. 5-8) has belonged to Aptornis otidiformis rather than to Notornis. This view derives further support from the fact that, with the decreasing relative size to the femur, there is a progressive simplification of the sternum in the recent Rallines (Tribonyx, Ocydromus), by which the still more simple type of the bone in Aptornis defossor is approached. The slender lateral processes, for example, in Tribonya ven- tralis come off immediately behind the costal border, diverge and extend backward beyond the body of the bone, with slightly expanded terminations. In Ocydromus australis the slender lateral processes come off at some distance from the costal borders, near the hind part of the body of the sternum, are consequently much shorter, and have no terminal expansion. The keel, which is well developed in Tridonyx ventralis, is much reduced in Ocydromus australis’; it is almost obsolete in Aptornis defossor. As the wings of Notornis are relatively less than those of the Weka, and the legs relatively larger, it is not likely that the sternum would bear the same proportions of length to the femur as the sternum figured in tom. cit. pl. 4. figs. 5-8 bears to the femur figured in tom. cit. pl. 2. fig. 3. Whether, however, the sternum of the smaller species of Aptornis would be shorter in proportion to the femur than it is in Aptornis defossor may be questioned. Demonstration on this point still waits the acquisition, so long desiderated, of the entire skeleton of Notornis. This, at least, is certain, that the larger sternum, belonging to Aptornis defossor, repeats all the generic characters of that of which the reference oscillates between Aptornis otidiformis and Notornis mantelli. As in it, the present larger sternum is of an elongate triangular form, the base anterior and concave, the curve of the emargination being interrupted by the pair of 1 Descriptive Catalogue of the Osteological Series, Mus. Coll. of Surgeons, 4to, 1853, p. 238. no. 1280 (Brachypterye). PROFESSOR OWEN ON THE GENUS DINORNIS. 121 prominences (Pl. XIV. fig. 2, a, a), from which the faint beginnings of the keel converge as they retrograde to the thick low ridge (ib. s, s) representing that part of the sternum in birds of flight. The body of the sternum describes a slight curve lengthwise to the tapering end, with the convexity downward or outward (ib. fig. 4); the general trans- verse lay of the outer surface of the sternum is slightly convex; but between the keel-ridge and the lateral margin the surface is feebly excavated. The ridge (ib. s, s) expands and subsides about an inch and a quarter from the hind end. In the smaller sternum (of Aptornis otidiformis?) the initial ridges converging from the anterior tuberosities to the sternal ridge are better defined by the excavation of the surface outside them than in Aptornis defossor, and the ridge expands into a trans- versely convex prominent tract, which is continued to the broken end (as shown in fig. 5. pl. 4, Zool. Trans. vol. iv.); it consequently extends further back than in Aptornis defossor, apparently not leaving a terminal flattened tract of the outer surface of the bone, such as is shown in fig. 2. Pl. XIV., near the end of the sternum. The costal border in Aptornis defossor forms two fifths of the lateral border of the sternum, and presents articular surfaces (ib. fig. 4, c, c) for five sternal ribs. The foremost surface is a narrow ridge, crossing obliquely the costal surface close to the articular groove for the coracoid. This surface appears to be obliterated by anchylosis of the coracoid to the sternum on the right side (ib. fig. 2, d); and the left side shows a fracture at the part corresponding to the smooth deep coracoid groove in the sternum of Apt. otidiformis ('). The coracoid union with the sternum is restricted to as small an extent, relatively, of the anterior border or base of the bone as in Apt. otidiformis (?). The second, third, and fourth articular surfaces for the sternal ribs are ridges with extensive intervening smooth and imperforate cavities (Pl. XIV. fig. 4,¢,c). The last or hindermost articulation is a small subcircular cavity with a raised border. ‘The costal border contracts from the third ridge backward. The non-articular side-border of the sternum contracts as it recedes to a rather sharp edge, the two sides converging to a nearly pointed end. There is no trace of the lateral fissures and slender processes characterizing the sternum of Ocydromus or Tribonyx’. A very shallow tract (ib. fig. 2,7) for the insertion of muscle is bounded by a feeble ridge (ib. fig. 2, ¢, 7") 8 lines behind the anterior border of the sternum, which tract is less smooth and even than the rest of the outer surface of the bane. At the fore part of the inner or upper surface the sternum is strengthened by a transverse prominence or bar (ib. fig. 3, 4), which expands as it subsides at each end upon the inner part of the costal tract. This bar is further from the anterior border, and consequently more internal in position, than is the corresponding ridge in the sternum of Apt. otidiformis (?), where it seems to form the upper or convex border of a thickened anterior margin of the sternum’. A shallow excavation of the inner surface of the sternum is bounded by a curved border (ib. fig. 3, ¢), concave forward, nearly midway 1 Trans. Zool. Soc. vol. iv. p. 18. 2 Tb. vol. iy. pl. 4. fig. 8. v2 122 PROFESSOR OWEN ON THE GENUS DINORNIS. between the fore and hind sides of the sternum. The corresponding part of the inner sur- face of the sternum of Aptornis otidiformis, or of Notornis (?) is not so defined, but is undu- lated by a rising, convex lengthwise, parallel with the mid third of the costal borders. The figures of the sternum of Aptornis defossor in Pl. XIV., being of the natural size, preclude the necessity of further admeasurements. There is no pneumatic perforation in any part of the sternum, which agrees, in this respect, with that provisionally referred to the smaller species of the genus. The generic distinction between Aptornis and the extinct gigantic Ralline bird called Cnemiornis is strongly expressed by the sternum. In Cnemiornis the ridged repre- sentative of the keel’ subsides much nearer the fore part of the bone; the costal border is relatively longer, and includes articular surfaces for seven sternal ribs; the body of the bone appears to retain its breadth for a greater extent as it recedes than in Aptornis defossor. The portions of sternum picked up with the other bones of Cnemiornis enable the above comparisons to be instituted, but they are not sufficient to give a satisfactory idea of the shape of the entire sternum in that genus. The pelvis of Aptornis defossor, figured of the natural size in Pl. XIV. fig. 1 and Pl. XV., exhibits a slight deformity or departure from the normal symmetrical figure at its fore part, which is somewhat bent to the right—the twist, which chiefly affects the neural spine and ilia, being most conspicuous when the bone is viewed from the front and from below, as in fig. 2. Pl. XV. It is doubtless an individual variety, and indicates the bird from which it was derived to have been aged. The articular surface of the centrum of the first sacral vertebra (Pl. XV. fig. 2, s 1) is transversely extended, measuring in that diameter 1 inch 44 lines, the vertical diameter at the middle being 5 lines; but this extends to 7 lines on each side beyond the neural canal, which is circular, and only 3% lines in diameter at its anterior outlet. The prezygapophyses (Pl. XIV. fig. 1, z) are large, with their outer borders bent strongly forward; the right is the largest, partaking of the asymmetry above noticed. The articular surface, concave transversely, is vertically plane, or rather convex. The fore part of the neural spine shows a rough surface for ligamentous attachment along the lower half of its extent; this surface expands to a breadth of 4 lines, then contracts to a point, beyond which the spine presents a sharp anterior margin to near the summit, where it again expands and coalesces with the contiguous parts of the iliac bones. A deep and narrow ilio-neural fissure extends backward on each side the spinal plate. The neurapophysis, developing the zygapophyses, is coextensive with the centrum, and expands beyond that element into a broad depressed diapophysis, confluent above with the ilium (Pl. XV. fig. 2,7): this process forms at its hinder angle a roundish flat surface (Pl. XIV. fig. 1, d) for the tubercle of the first free sacral rib. The articular surface for the head of that rib is large and flat on the left side (Pl. XIV. fig. 1, pd), but is small on the right side partaking of the asymmetry above noticed. The under 1 Trans. Zool. Soc. vol. v. pl. 63. fig. 9, s, PROFESSOR OWEN ON THE GENUS DINORNIS. 123 surface of the centrum is carinate; the keel beginning 2 lines behind the lower border of the articular surface. The keel runs to the third sacral centrum (Pl. XV. fig. 2, ¢), where it begins to expand, as in Aptornis otidiformis (Zool. Trans. vol. vii. pl. 42. fig. 2). The surface for the head of the second sacral rib is small, subcircular, concave, and produced. The succeeding pleurapophyses (pl) are represented by short thick parapophyses abutting against the lower border of the ilia, to the fifth (seventh, including the moveable ribs) pair, which abuts against the part to which the head of the pubis is anchylosed (Pl. XV. fig. 2,64). There are consequently six pairs of interapophysial vacuities (ib, id. id.) at the antacetabular part of the pelvis. The mid tract beneath the centrums gains a breadth of } inch at the seventh vertebra, beyond which it contracts to a point at the fourteenth. The sacral centrums maintain their breadth to the seventh vertebra, contract at the eighth, but between the acetabula maintain a breadth of 94 lines to the eleventh vertebra, beyond which they contract to the fifteenth, and again expand at the seven- teenth (17) to a breadth of 5 lines, which they retain, below, to the twentieth vertebra. The last three of them (n,n, fig. 1, Pl. XV.) are caudals, which, like the dorsal and lumbar vertebre at the other end of the pelvis, have become “sacral” by anchylosis. In the three interacetabular sacrals (Pl. XV. fig. 2, c’) the parapophyses are, as usual, suppressed ; there is, however, a filamentary representative of one of those processes from the left side of the eighth sacral centrum, The parapophyses reappear at the eleventh sacral (ib. p, 11), where they are long and slender, and combine at their distal ends with those of the twelfth, thirteenth, and fourteenth sacrals to form a plate or screen of bone (ib. w), dividing the interacetabular depression (¢) from the postacetabular or postrenal one (uw). The parapophyses of the fourteenth (ib. fig. 2, 14), fifteenth, and sixteenth sacrals increase in breadth, and bend or arch outward and upward to form the lower and lateral walls of a passage or cavity on each side of the crest formed by the continuous or confluent neural spines of the corresponding vertebre. These “ ectoneural” canals are partially divided above by diapophysial or upper transverse plates, arching from the neuro-spinal crest to the inner surface of the plate or ectoneural side-wall. The civil engineer might study, perhaps with advantage, the disposition of the several buttresses, beams, and arched plates of bone which support the iliac roof. of the pelvis, and strengthen the acetabular walls receiving the pressure of the thigh-bones, in this large and powerful Wood-hen. The unusual depth and width of the excavation at the postacetabular part of the pelvis, the hind part of which excavation is partitioned off from the general pelvic cavity by a deck, as it were, of bone (Pl. XV. fig. 2, v), extending from the ischium and confluent part of the ilium inward or mesiad to join the hinder sacral vertebra (ib. ¢, 17), led me to examine the pelvic viscera in a recent Ralline (Rallus aquaticus) with a view to determine the nature of the contents of the homologous ilio-ischial postacetabular excavation in that bird. 124 PROFESSOR OWEN ON THE GENUS DINORNIS. On a first view of the pelvic viscera, as exposed by removal of the intestines, the kidneys appear to be chiefly developed anteriorly, the broad lobes there extending beyond the ilia with a convex border covering, or lying in contact with, part of the iliac origins of the abdominal muscles. The length of these anterior renal lobes equals that of the antacetabular part of the pelvis in Aptornis defossor; the second lobes, commencing mesiad of the anterior ones, acquire their greatest breadth where those terminate, but are here much narrower than the anterior lobes. Thence the middle lobes contract transversely to a point, underlapping the narrow posterior lobes, which seem to end where the ischia join the sacrum, and to blend with each other; but it is merely close contact, they are distinct’. On each side of the hind part of the kidney is the belly of the “ obturator internus” or postrenal muscle, which underlies the inferior opening of the ilio-ischial excavation. This muscle being removed, the renal substance is found to pass “ neurad,” expanding to a breadth equal or superior to that of the anterior lobes, with much greater depth or thickness in the neuro-hemal or dorso-ventral direction; and this the true posterior renal lobe is shown, by a vertical longitudinal section of the side of the pelvis, to fill the whole of the great posterior ilio-ischial cavity, extending from the partition bounding posteriorly the interacetabular cavity backward to above and beyond the ischio-iliac deck-like plates (Pl. XV. fig. 2, v’). The vertebree numbered 19, 20in Pl. XV. fig. 2 are homologous with the first two free caudals in Rallus aquaticus and Ocydromus australis. 'The parapophyses of the seven- teenth and eighteenth sacrals expand, coalesce (a small foramen intervening), and unite with the “deck” (v'), of which they form the inner beginning. The parapophyses of the first caudal in Radlws abut against the ilia, leaving a small intervening foramen between them and the antecedent parapophyses. The homologous foramina are seen on each side of the vertebra 19 in fig. 2. Pl. XV. From these foramina each ilium extends backward 2 inches. The free terminal ends bend slightly toward each other, leaving an interval of 9 lines; they are obtusely rounded. Externally each is strengthened by a vertical ridge (Pl. XIV. fig. 1, Z). The entire length of the ilium, following the upper curve, of Aptornis defossor is 124 inches, equalling that of the same part of the pelvis in Casuarius bennettii. This length is pretty equally divided by the mesial beginnings of the “gluteal ridges” (Pl. XV. fig. 1, 9). The fore half of the ilium repeats very closely the characters of that part in Aptornis otidiformis*. The upper curve is greater in Aptornis defossor; the gluteal ridge is stronger, runs more outward, and ends by an obtuse process (Pl. XIV. fig. 1,/). The 1 In my ‘Anatomy of Vertebrates,’ the statement (vol. ii. p. 227) that “the kidneys are more or less blended together at their lower extremities in the Coots (Fulica)”’ will probably bear the above explanation. T am indebted to our excellent Secretary, Mr. Sclater, for the subject of the above dissection. 2 Trans. Zool. Soc. vol. vii. p. 366, pl. 42. fig. 1, pl. 43. fig. 1. PROFESSOR OWEN ON THE GENUS DINORNIS. 125 rest of the pelvic disk (Pl. XV. fig. 1,7, 7) is bounded externally by the strongly pro- duced ridge overhanging the vertical postacetabular plate (Pl. XIV. fig. 1, 62, 7) of the ilium, which coalesces with the similarly vertical expanding plate of the ischium (ib. 63). The foramen (ib. m) between the ilium and beginning of the ischium is a full ellipse or oval. The ischium developes downward a ridge, behind the obturator notch (q); but this does not meet the ridge reciprocally directed upward from the pubis (ib. 64): there are impressions of the attachment of a strong fibrous sheet which closed the obturator groove behind; and this sheet becomes a thin plate of bone in old individuals of some existing Rallines (Tribonyx ventralis, Ocydromus australis). The proportions and form of so much of the pubis (64) as is preserved adhere to the ralline type of that bone; but the tubercle, prominent below the fore part of the bone in most existing Rallines, is not developed in Aptornis. I have restored, in dotted outline, the parts of the ischium and pubis broken away from the otherwise complete and truly singular and interesting form of pelvis in Aptornis defossor. It may aid in future comparisons of this most complex of bones to subjoin a list of the parts of the pelvis, conveniently indicated by names, with the symbols used to indicate them in Pl. XIV. fig. 1, and Pl. XV. The specimens of Aptornis defossor above described are from a cavernous fissure at Timaru, Canterbury Settlement, South Island of New Zealand. I am indebted to Dr. D. S. Price, of the Crystal Palace, Sydenham, for the much valued opportunity of describing, comparing, and figuring them. Parts of the Pelvis of Aptornis defossor, Ow. a. Acetabulum (the letter marks, in Pl. XIV. fig. 1, the inner aperture). b. Postacetabular facet. c. Centrum (marking, m Pl. XV. fig. 2, the ridged underparts of the first and second sacral vertebre). c'. Centrum (marking the unridged underparts of succeeding centrums). d. Diapophysis (marking, in Pl. XIV. fig. 1, the diapophysial articular surface). e. Subacetabular fossa. f. f. Antacetabular part of ilium (Pl. XV. fig. 1). g. Gluteal ridge. h. Gluteal process (Pl. XIV. fig. 1). z. Ilium (marking, in Pl. XV. fig. 2, the part anchylosed with and overhanging the foremost diapophysis). id. Interdiapophysial vacuities (marking the six anterior ones). l. Postiliac tuberosity. m. Ischiadic foramen. n. Neurapophysis; ' in Pl. XV. fig. 1, is the fore part, n* the hind part, of the neur- apophysial crest. o. Ilio-neural orifices. p. Parapophyses, aes combine to separate, below, the interdiapophysial vacuities: pl. Pleurapophyses,) in Pl. XIV. fig. 1, p/ marks the articular surface for the head of the first sacral rib. 126 PROFESSOR OWEN ON THE GENUS DINORNIS. g. Obturator notch. r. Pelvic disk. s. Sacral vertebre (s 1, in Pl. XV. fig. 2, marking the anterior articular surface of the first centrum). t. Interacetabular cavity. u. Postrenal cavity. v. Tlio-ischial plate or deck-like production. z. Prezygapophysis (marking, in Pl, XIV. fig. 1, that of the first sacral vertebra). 1-20 indicate the sacral vertebree from before backward. 62. Tlium. 63. Ischium. 64, Pubis. In the restoration of the skeleton (in which I have taken the vertebral formula of Ocydromus) the Aptornis defossor, with the neck bent and head supported as in the ordinary standing position of the bird, would be about 3 feet in a straight line from the top of the head to the soles of the feet; the length, in a straight line, from the end of the beak to that of the tail would be about 2 feet 10 inches. The chief depar- ture from the ordinary form of the Coots is seen in the shorter and more robust proportions of the tibie and metatarsal segments of the legs, in which feature Notornis, amongst recent Rallines, offers the nearest approach to Aptornis. DESCRIPTION OF THE PLATES, PLATE XIV. . Side view of pelvis of Aptornis defossor, nat. size. . Front or under view of sternum of ditto, nat. size. . Back or upper view of sternum of ditto, nat. size. . Side view of sternum of ditto, nat. size, he} _ ee) S m op DS ee PLATE XV. Fig. 1. Upper view of pelvis of Aptornis defossor, nat. size. Fig. 2. Under view of pelvis of ditto, nat. size. PLATE XVI. Fig. 1. Restoration of skeleton of Aptornis defossor. 3th nat, size. Fig. 2. Skeleton of Ocydromus australis, 4th nat, size, M & N-Hanhart imp Front Nature on Stone by J Erxleben. SSOR O APTORNIS DEF Pe ° (fB M & N Hanhart ump FromNawnre on Stone by J Exzleben. APTORNIS DEFOSSOR ‘ MAN Hamhart imp. ROwen del J.Erdeben lith 1. APTORNIS DEFOSSOR. 2 OCYDROMUS AUSTRALIS tia URai HN peIoeh J V. On the Form and Structure of the Manatee (Manatus americanus). By Dr. James Muriz, F.L.S., F.GS., &c. Read November 15th, 1870. [Puates XVII. to XXVI.] I. Exterior ASPECTS AND DIMENSIONS. 1. GENERAL CONTOUR.—Among original investigators of the uncouth aquatic -mammal the Manatee, the painstaking Daubenton, in one of Buffon’s well-known volumes, has given a representation of a foetus in profile, and of the head foreshortened'. These evince only a certain accuracy, being too small to bring out the minor tegu- mentary characters. This remark applies in common to Sir Everard Home’s better executed figure’, copied by Frederick Cuvier’, to J. A. Albers’s foetus‘, and to that of Alexander von Humboldt*®. Dr. Hermann Stannius’ in his larger illustrations of the head of the animal, likewise fails to depict skin-texture, although the peculiarities of the muzzle are effectively rendered. Professor W. Vrolik’ expresses himself thus concerning his own folio plates:—“I suppose that my representation is quite exact, and will give such a just idea of the animal that it is useless to enter into more particulars.” It is therefore in no hypercritical or self-satisfied spirit that I draw attention to pictorial defects in the able treatises of these revered preceptors of comparative anatomy, but rather in vindication of my reproducing work already done. Of the two specimens which have come under my scalpel, female and male, I was content to note that the former did not coincide in certain points with Vrolik’s delineation. Bethinking myself, however, on receipt of the latter, that a good photograph would be a desirable acquisition, I took such views as seemed best to afford evidence of the external affinities of this singular creature, and of the adaptability of its unwieldy carcass to propulsion through a watery element. In museums, it is customary to see * Histoire Naturelle, tome xiii. plate 57. ? Lectures on Comp. Anat. vol. iy. pl. 55. * Pl. 1 of the Hist. Nat. des Cétacés. Paris, 1836. * Tcones ad illustrandam Anat. Comp. fase. ii. tab. iv. Lipsiz, 1818. 5 In Dr. Wiegmann’s ‘ Archiy fiir Naturgeschichte,’ 1838, p. 1, pls. 1 & 2. ®° Beitriige zur Kenntniss der amerikanischen Manatis von Hermann Stannius. Rostock, 1846. Tab. 1. figs. 1 & 2. 7 «Bijdrage tot de Natuur- en Ontleedkundige Kennis van den Manatus americanus,” p. 58 (pls. 1 & 2), in the Memoirs’of the Zool. Soc. Amsterdam, Natura Artis Magistra, 1852, For a copy of this folio monograph I am indebted to the liberality of J. Norrdhoem Heyt, Superintendent of the Zoological Gardens, Amsterdam. VOL. VIII.—PART 111. September, 1872. x 128 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. such bloated over-stuffed specimens, that from them, as well as figures extant, an unfair idea of the configuration is obtained, and one is embarrassed to comprehend the mode of progression of such an awkward form in water as on land. Emendations on the text of the forementioned authors relative to the shape of the animal are less necessary; so that I limit myself to a notice of the points which photo- tography has elucidated in the present case. Pl. XVII. fig. 1, profile view, shows that in the young Manatee the head and body, to as far as the root of the tail, have together a very elongated biconical contour—not so protuberant at the posterior belly part as in Home’s figure, and quite different from the barrel-like aspect of Vrolik’s animal. Seen on the dorsal (P]. X VIII. fig. 3) and on the ventral (Pl. XVII. fig. 2) aspects, the biconical form is less rigid, from the deep skin-sulci being more emarginate; whilst towards the pelvic region there is a sudden rugged contraction, as if behind the ribs a broad band had been tightly lashed round the short axis of the body. Posteriorly to this the outer border-lines obliquely diverge in a very gradual and regular manner, so as to form a comparatively long and very broad, thin, shovel-shaped, caudal organ. The hindermost border, whilst rounded', has a remarkable truncated character, and centrally is incised; or rather there is a short, shallow longitudinal sulcus on its upper surface, which forms a corresponding convexity below. ‘The tail of Home’s specimen gives a three-quarter or tilted view; but this, if compared with the present fig. 1, is too much narrowed at the end. The Beaver-shaped tail attributed to Manatus by some writers is true only to a certain extent, inasmuch as in the latter it broadens greatly compared with the former, and, as Albers’s figure distinctly represents, there is a slight mesial V-shaped indentation or emargination. Regarding the head, Stannius’s figure is by far the most trustworthy; but, as already particularized, the absence of texture detracts from the otherwise characteristic phy- siognomy of the creature. W. Vrolik’s best representation, to my mind, seems to be that depicting the under surface of the lower jaw and muzzle (pl. 2. fig. 5); his front foreshortened view of the head (pl. 2. fig. 4) has far too large, staring eyes; and these are not widely enough apart. The facial expression, as dependent on the eye, is markedly noticeable on comparing Vrolik’s plates and those now given. The accuracy of the present lithographs are vouched for as carefully taken photographs, faithfully and minutely copied by my artists, Messrs. Berjeau and Smit, figure 3 alone having been slightly reduced from the negative impression so as to fit the length of the plate. 2. Admeasurements——Humboldt, Stannius, and Vrolik have each recorded some of the proportional dimensions of the body. On this account it might be deemed super- fluous to institute a fresh series of measurements, were it not that it lends precision to the description and figures of the specimens under immediate consideration. Inches and decimals have been taken as the standard throughout. 1 This peculiarity has evidently caused the adoption of the name “ Round-tailed Trichechus” for the Manatee, vid. Shaw’s ‘ Zoology,’ vol. i. p. 244. il DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. Female. Body generally. inches. Extreme length, from the snout to the tip of the tail . . 69°0 Girth of neck just behind vertex of cranium . . 2 = Girth immediately in front of the pectoral biiednitios fae B25 Girth immediately behind the pectoral extremities . . . 39°5 Girth about 43 inches in front of the umbilicus . . . . 47°0 Girth at the umbilicus . . . ene Suniel Girth at the opening of the penis in the male HSH) Shane Se Girth at about 3 inches in advance of rectum in male, or equivalent to the position of the vulva in female . . . 41-0 Gurthvatithe anusi iy stil xe Wel S ledel epee eA? 9 38° The Head. Extreme length, or distance between the muzzle and occiput 11:3 Girth at the snout, including the lower lip. . . . . . 140 Girth vertically before the eyes, including the lower lip - 18-6 Girth behind the re part of lower lip, vertical to the eyes . . pose be ZO0 Girth, middle of head jie anterior to ssiopeaticn of siadalible 23°0 Oral region, or length of side of the mouth . . . . . . . 2°2 Length of under lip, following tegumentary curve . . . 49 Breadth of chin, following tegumentary curve. . . . . 40 Breadth of chin at the angles of the mouth. . . . . . 5:0 Muzzle and Nostrils. Measurement of arch over muzzle covering nostril . . . 85 Nasal orifices, distance behind free end of muzzle. . . . — Nasal orifices, diameters across and tea when dilated, about. =. « : Bhise ead Nasal orifices, aemcs spe at inner nites about ie: Nasal orifices, distance apart at outer angles . . . . . 17 The Eye. Diameter. . . Jt oc ntireiosgortin de 6 Distance from front - of naa (oxisanhs inpele ishinkalnas BO Distance from front of muzzle (mesially) . . . . . -. 60 Distance from the inner angle of nostril. . . . . . . 45 Distance between them at their centres. . . . . . . 70 Distant from the angle of the mouth. ©. . 2 1. 6 1 om The Ear. Auditory orifice distant from posterior angle ofeye . . . 40 Auditory orifices apart from each other in diameter. . . 7:0 x2 Male. inches. 48:0 24:5 25:5 28:5 29°5 31°5 32°5 28-0 20-0 9-0 12°5 15:0 17-0 20°5 1:35 130 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. Female. Male. Pectoral Extremity. inches. inches. Distant from the muzzle. . . . - - . 2... 2. 2 — 10:0 Distant from the end of tail. . . . . a2 38:0 Width from one tip to the other when suntaeeuchibatt Apis 25:0 Roots apart from each other, following curve of back . . 19-0 — Extreme length of free portion. . . . . . .. . . 100 8:0 Extreme breadth of free portion . . . . .... . 44 3:5 At the narrowest portion its transverse diameter is . . . — 2°5 Girth at the root ARCA SCT Cy SSM His SFO . 10°3 8-0 Girth at about the middle . . . Vie AO a0 8:0 Girth at about the outer or distal third ee eit RASS 8:0 Thickness at the elbow .. . sis i Waa ks Web ogi 1-7 Thickness at the middle of the ee SiS aS 0-6 Tail, or caudal extremity. Extreme length, or from the last loin-wrinkle backwards, measured on the ventral surface. . . . OT eee GEO) 16:0 Extreme length measured on the dorsal suithide Sie Sd sol) 15-0 Distance betwixt first wrinkle and angle of terminal border. — 12:0 Breadth following curvature of terminal border . . . . — 18-0 Extreme breadth following superior curvature of skin . . 16°5 13-0 Girth, about four inches from the tip . . . . . . . 278 26-0 Girth, about ten inches from the tip. . . . . 2... — 20°5 Girth at the root or hindermost loin-wrinkle . . . . . — 17:0 Notch or median terminal incision, in diameter . . . . — 0°5 Generative organs &c. Rectum distant from end of tail . . . . Rrepilese Qa 18°5 Distance from middle of anus to middle of vagina . . . 3:2 — Distance from middle of anus to opening of penis . . . — 9-8 Length of generative outlet, including labia . . . . . 14 1:3 Length of generative outlet, solely the opening . . . . 0°9 0-9 Diameters of anus, about. . . . 2 et 085 0-5 Breadth of each labial elevation potinaing el apeinate _— 1:3 Circumference of both. . . . . a EL a i 4:0 The umbilicus is distant from the nels RIS EE ge 18-0 The umbilicus is distant from the end of tail . . . . . — 30:0 The umbilicus is distant from the middle of penal opening — 2°4 The umbilicus in the male has a cicatrix one inch long by half an inch broad. When studying in the Stuttgart Museum I derived much information from Professor Kyauss, the able Director. Among other things he mentioned that their large stuffed DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 131 specimen of Manatee was the mother of our Society's young male, as attested by Herr Kappler, of Surinam, who transmitted it. The length of the female mounted skin I ascertained to be 122 inches, therefore twice and a half the length of the young animal possibly six or eight months old. Another stuffed male specimen at Stuttgart measures 94 inches. Both the above are doubtless stretched to their fullest extent; still one is justified in assuming the adult Manatus to be from 9 to 10 feet long. Comparing this with Steller’s account of Rhytina, it appears that the extinct northern form exceeded the existing American one in the proportion of two and a half to one, or something equivalent to the difference betwixt the young and old of the latter. 3. Weight.—According to Mr. Greey the entire carcass of the Zoological Society’s female when weighed immediately after death on board ship was 228 lb. ‘That of the young male as ascertained by myself was 61 1b. 4. Colowr.—This has been defined by different writers as grey, bluish grey, and steel- grey. It is indeed difficult to specify the precise tone of colouring, which is a kind of neutral tint, varying according to the condition of the skin. When the epidermis is dry, the colour approaches to a dull iron-grey; but when moist, it appears more of a dull black or sooty hue. It is best compared with that of the Elephant, to which in other respects it offers strong resemblances, as shall afterwards be mentioned. The anterior truncated portion of the muzzle, projecting part of the palate, and lower lips are paler than the body, namely of a dull yellow. IL. Tue InrecuMent, 1rs APPENDAGES AND SUBJACENT TEXTURES. 1. The Skin.—The coarse, hairy hide of the Manatee is one of those external features which at once arrest attention, and claim for it kindred with Pachyderms rather than with its aquatic congeners the smooth-skinned Cetz. The construction of the extraordinary-looking truncated muzzle having been com- mented on by Baron von Humboldt, and freshly described by Professors Stannius and Vrolik, 1 need therefore but cursorily allude to its dermal minutie. The anterior face, and particularly its under surface, has a very warty-looking, but regular, pitted structure. In some parts the arrangement of the furrows and ridges is of a cross-linear kind; and in the neighbourhood of the bristly projections (afterwards to be noted) the puckered skin assumes a stellate sculpturing. The much smaller-sized lower lip has three or four regular transversely arched ridges and deepish intervening furrows, and in front of these some half a dozen shallow linear grooves. The under surface of the bulging chin is altogether smoother than the muzzle; but yet its dermis has a tessellated superficies. The short throat has deep transverse wrinkles. The various tegumentary folds, though passed over by authors, are worthy of special consideration. On the upper surface of the head, and well nigh obscured neck, several deep transverse wrinkles extend in arches from side to side. The furrow immediately behind the muzzle is much the deepest, and especially so 132 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. towards the angle of the mouth, where it is quite a cleft. Vrolik’s figures well define it, Stannius’s less so. Between it and the eye there is a second notable, but narrower, groove. At the vertex another deep furrow runs round quite to the angles of the lower jaw. Betwixt these there are shorter and shallower grooves, some of them obliquely joining those described. There are no very determinate upper and lower eyelids; but radiating round the palpebral fissure are a series of crooked wrinkles. These, I have been told by those who have seen the living animal, are twisted together in the act of closing the eye. Vrolik’s artist has so circumscribed these ocular radii by an external dark, broad, circular line, as to deceive any ordinary observer by the supposition that the animal has a large patent eyeball. This deception is further heightened by a heavy backwardly overlapping orbital fold, which certainly was not present in either of the specimens examined by me. The text, however, corrects this misapprehension, as the author pointedly alludes to the diminutive eye of the Manatee. Besides those very long grooves and tegumentary areas just mentioned, a striking feature of the head, and notably so on its upper half, are the rough scale-like patches. These circumscribed elevated spots are irregularly shaped, though chiefly roundish, flat on their upper surface, some smooth, others roughened or minutely pitted, as is the rest of the skin. They vary in size from o to (.), on the vertex give a nailed appear- ance, and on the side of the face subdivide the rugged skin into elongated and diamond- shaped corrugations. The pectoral extremity, as has been noted by others, is sunk into a great shoulder-fold. As far as the elbow there are deepish transversely oblique skin-folds; but the remainder of the limb presents only minute wrinkling. The axillary creases are short and decussating. Both on the outer and inner aspects of the flattened limb the surface is studded with the small warty flattened bodies spoken of as existing on the head. In the limb, however, they are much more uniform in outline and size. The body, from the obscurely defined neck backwards to the loin-constriction, possesses multitudinous encircling narrow linear plaits, which run parallel to each other, and frequently obliquely interdigitate. This gives to the skin a kind of velvety structure, increased in semblance by the sombre tint of the derm. Behind the shoulders several massive folds are mapped out rather than project; and these are carried from the back round the chest. Upon the sides and shoulders tuberculated scale-patches, resembling those on the head, are here and there distributed. Numerous short, but irregular, longitudinal wrinkles are met with upon the throat and abdomen. The marked broad but sudden constriction immediately behind the ribs, forming a loin-girdle, consists below of two large hoop-like folds, the one before the other, the anus being set midway between them, and the vulva just anteriorly in the female. As these folds reach the back, their boundary furrows augment and increase the number of the folds, while their height is diminished accordingly. DR. J. MURIE ON TH# FORM AND STRUCTURE OF THE MANATEE. 133 That portion of the subcaudal surface which, as it were, defines the fleshy limits of the tail possesses numerous short sinuous transverse wrinkles; but the remainder of the expanded organ is devoid of these. The upper surface is comparatively smooth. The scaly epidermal patching met with on the fore part of the body is very sparsely distri- buted on the tail—and where present is chiefly at the margins, as small punctate dermal tracery. All round the very posterior edge of the caudal expansion, but on the dorsal surface, there is a smooth cord-like rim one eighth of an inch broad. The thickness of the skin varies with its situation. Near the generative outlet I found it to be 0-4 inch, the epidermis itself 3! of an inch*. 2. Hair and Bristles—The sparsely distributed hairs upon the head, trunk, tail, and extremities of Manatus and Halicore have been mentioned by all observers; I shall but append a supplement to their remarks. Two kinds of dermal appendages have been noted—longish pliant hairs, and short stiff bristles. The former, scattered over the back and belly, have an average length of 13 inch; but many are shorter, though a fair proportion reach and even exceed 2 inches. Each hair is very fine, soft, smooth, and pale-coloured. Upon the limbs the hairs are considerably shorter than above stated, but are closer set together, especially on the palmar aspect. On the upper surface of the head they are likewise curtailed in length compared with those on the body and tail. At each angle of the mouth, partially within the lips, developed on the upper, but still more so on the lower jaw, is a pretty thick bunch of long, somewhat coarse, hairs. Of these Stannius says that they cause the cheeks to appear thickly beset with hair; but this neither his nor Vrolik’s sketches clearly exemplify. These hairs within the mouth are not without interest, as, it may be, they, and not the horny palate, are the homologues of the whalebone or baleen plates of some Cetacea. Quite under the chin, as Vrolik shows, the hairs are stiffer than those described, uniform in calibre, and about a quarter of an inch in length. These, both in outward aspect and texture, are intermediate between the hairs of the body generally and the true bristles. The latter are stout, blunt-pointed, and spring from the pits of the-rough muzzle and ridged lower lip. They vary in total length from 0-3 to 0-4 inch, though not more than half that is free. Towards the nares, where smallest, they project only slightly, but lower down increase in size and rigidity, so that when the hand is passed over the surface it feels like a rasp. 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 are specially pronounced. This spot would seem to possess most tactile delicacy ; for twigs of the infraorbital and facial nerves are abundant thereto. On the semilunar, pale-coloured, tough, lower lip, there are three transverse rows of bristles and trapezoidal ridges. Besides bristles there were many of the long silky hairs scat- tered on the face of the muzzle, they being in greatest plenty, however, circumferentially. 1 For its microscopical composition, consult Professor Paulsen’s observations, and woodcut in Brandt’s Symb. Siren. iii. p. 252, and Leydig’s Lehrb. d. Histol. p. 87. 134 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 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. 3. Fatty Envelope &c.—The two animals differed considerably as regards their bodily condition. The female was fat as a pig, whilst the younger male, though on the whole plump, possessed rather an abundance of areolar and fibrous textures than fatty tissue. In the former, immediately beneath the skin, and enveloping the whole of the body and root of the tail, there existed a layer of remarkably dense fat. This adipose material, under the knife, cut not unlike bacon or solid mutton suet, being rather more greasy, however, than the latter. The pectoral limbs and the anterior portion of the muzzle differed from the body in being almost destitute of fatty clothing, its place being supplied by fibroid tissues. On the back the fat had a thickness of 13 inch, and at one spot, behind the shoulder-blade, where the panniculus muscle becomes aponeurotic, it had a depth of 2 inches. On the abdomen generally it did not exceed 1 inch thick, thinning to }an inch or so towards the vulva and anus. Still further backwards it lessened by degrees, until lost in the interlacing tendinous aponeurosis forming the flat caudal expansion. In front, over the head and lower jaws, the fat likewise diminished gradually, so as to leave the great nasal and mandibular muscles almost superficially free from it. As referred to in my description of the muscles, and mentioned by Stannius in his dissection, there was a layer of softer fat intervening between the panniculus carnosus and the muscles lying beneath. In some places, chiefly the anterior half of the body, this exceeded half an inch in depth; but posteriorly it was considerably less in quantity. A lump of fat covers the deep layer of fascial muscles beneath the infra- orbital proc*ss. Structurally, as Vrolik justly observes, the cutaneous fat is unlike that of Cetacea in possessing little or no free liquid oil; and in consequence it more resembles that of ordinary mammals. I noticed particularly in the abdominal layer a vast number of minute red puncta. These appeared to be the cut extremities of vascular twigs, the continuation, it might be, of the subjacent rete mirabile. I had the curiosity to weigh the fat taken from the outside of the body of the female, and found there were 24 1b. 10 oz. ‘This approximates closely to one ninth of the total weight of the animal. Desirous of judging of the flavour of the flesh of the Manatee, I had several portions of the specimen forwarded by Mr. Latimer cooked. One or two gentlemen partook of it along with myself; and the unanimous opinion expressed was, that it ate excellently. When broiled, the fibre appeared white and delicate, and the flavour was that of a crisp, tender veal cutlet. This is en rapport with the accounts of the natives, travellers, &c., who eat it freely; and, indeed, it is said the Catholic clergy in South America do not object to its being used on fast days, on the supposition of its being allied to the DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 135 fish tribe. Steller’s advice that the Rhytina was good food too was eagerly adopted by the northern mariners, to the annihilation of that remarkable Sirenian race. The same fate awaits the Dugong, since not only is its flesh appreciated in Australia, but the oil, obtained by boiling the fat on the body, according to Dr. Hobbs of Queensland, rivals, if not surpasses, in therapeutical excellence the better-known cod-liver oil. Ill. THe SKELETON AND 11s LigAMENTOUS CONNEXIONS. My annotation concerning the skeleton of the Manatee shall be circumscribed, foras- much as its osteology has heretofore been subjected to careful research at the hands of shrewd, scrupulous, and laborious investigators. The literature on the osteology of the Sirenia, though the order contains but four genera, Halitherium, Rhytina, Halicore, and Manatus, with few species, stands forth prominently on account of the galaxy of talent that has swept the field. Steller’s’ early observations still hold a worthy place. On the Lamatins, living and fossil, the genius of Cuvier, in his ‘Ossemens Fossiles’ (vol. v.), and De Blainville, in his ‘Ostéographie,’ are monuments of masterly gene- ralization. Stannius and Vrolik and Krauss’, in their special monographs on the American species, largely treat of the skeleton; whilst the names of Schlegel*, Owen‘, Gervais’ and Serres, Kaup*, Brandt’, Gray®, Nordmann’, Huxley", and others not a few, individually attest to the assiduous toil bestowed on the above group of Mammals and the excellency of the workers thereon. 1. The Spinal Axis. Notwithstanding what I have said, it is somewhat remarkable that no two authors virtually agree as to the total number of vertebre in Manatus. This, it would seem, may arise from several reasons. Ist. Computation in some instances possibly has been taken from set-up skeletons, incomplete in the terminal caudal elements; 2nd, the number may differ in the very young and adult animals; 5rd, the amount present may bear a relation to the sex; 1 De Bestiis Marinis. 1749-51. 2 « Beitrige zur Osteologie des surinamischen Manatus,” in Mill. Archiv f. Anat. & Phys. 1858; also Archiv, 1862, p. 415, tab. xiii.; on Halitherium, by the same author, N. Jarhb. f. Mineral. 1858 and 1862. 4 Abhandl. &. Leyden, 1841. + Dugong, P. Z. 8. 1838, p. 28. 5 « Sirénides, Mam. Foss. du Midi d. 1. France,” Ann. d, Sci. Nat. 1841, 1846, 1847, 1849, 1850, &c., and Zool. et Paléon. Gén. 1867-69, &e. 5 Halitherium, ‘Beitriige, Darmstadt, 1855-62; N. Jahrb. f. Mineral. 1838, 1856, 1858, &c., with many illustrations. 7 «Symbol Sirenologice,” Mém. de l’Acad. Imp. d. Sci. de St. Pétersb. 1849, 1861, 1868, 1869, and other papers in the Bull. d. l’Acad.: altogether a series of magnificent monographs. 8 P.Z.S.; Ann. & Mag. N. H., various; and B. M. Catalogues. 9 Beitrige, Rhytina, Helsingfors, 1861, and Act. Soc. Sci. Fennice, tom. vi. 10 Hunterian Lectures, reported in ‘ Lancet,’ Feb. 1868. Furthermore, see extended references to literature, Brandt, J. c. fase. ili. pp. 237, 300. VOL. VIII—PaRT 11. September, 1872. Y 136 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 4th, numerical variability may occur in individuals of the same species apart from sex ; 5th, if there are several distinct species, as some hold, the aggregate in each may be different. In reply to the first of these reasons, it is doubtless true that the tiny ossicles termi- nating the caudal region, and each of which represents a vertebral element, occasionally are lost; the total numbers therefore in such cases would be under the maximum. But this only accounts for one kind of deficiency, whereas differences in numbers are attributed to the cervical and dorsal vertebrae, where the same excuse does not hold good. Concerning the second reason, here also non-ossification of the terminal caudals would give rise to the very young animal having a minimum of vertebre in the spine. But in this, as in the last, uniformity of regional numbers seemingly does not obtain. Third, the data extant showing relation of sex to spinal formula, does not prove that male and female possess a constant ratio the one to the other. The fourth proposition, I am of opinion, is the true explanation of the manifold discrepancies, excepting what concerns the cervicals. Professor Krauss, I may affirm, has had more Manatee skeletons pass through his hands than any other savant in Europe; and these have been received all from one locality, and undoubtedly of one species. His observations, most accurately made, are in every way trustworthy; and they go to show that the numbers of dorsal as well as lumbo-caudal vertebre are subject to irregularity. Results coming under my own notice substantiate his data. Fifth, specific distinction yields no very determinate data of spine-formulz, especially as concerns the supposition of distinctive American forms. It may be concluded, therefore, that the vertebral series of Manatus is inconstant within certain limits, and in this respect presents resemblances to those of the Cetacea. Besides differences of opinion respecting the total number of vertebral elements, authorities also disagree as to the numbers and character of vertebre taken regionally. The singular MZanatus has thus afforded a moot case, every ray of light shed upon which brings out fresh features or readings of the facts. As regards the cervical region, two points have excited discussion :—one, whether six or seven was the normal number of bones; the other, which vertebra was the missing one, provided the mammalian law of seven was deviated from. The vantage-ground has latterly been ceded to those who have maintained the presence of but six osseous representatives. Of special observers as to the first point at issue, Sir Everard Home, Alex. v. Humboldt, De Blainville, Leuckart!, and Robert? have enunciated that there are seven cervical elements; whilst Daubenton, the brothers Cuvier, Meckel®, Schlegel, Stannius, A. Wagner‘, Vrolik, Owen, Krauss, Brandt, Flower, and Gray, on more weighty grounds, have recognized only six clearly developed neck-vertebre. In both animals dissected by myself, only six appreciable neck-vertebre obtained. I thought I had detected the rudiment of a seventh in the young male; but a more scrutinizing search failed to justify * Zool. Bemerk, (Stuttgart, 1841), p. 62. 2 Comptes Rendus, 1836, p. 363. 3 Vergl. Anat. 4 Schreb, Siiugethiere, Fortsetz, 1846, pt. 7. p. 106, DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 137 my first impression. I may further say that of very many skeletons, in various stages of growth, examined by me in Continental and English Museums, none exhibited more than six cervicals. As to the second point, recognition of the absent one, De Blainville took up the question very categorically, inasmuch as he maintained that in one of the Leyden speci- mens he counted seven; and he assumes rather than proves that “la sixieme, finit par disparaitre dans son corps; l’arc restant libre dans les chairs, est enlevé avec elles.” This statement has been contradicted by Vrolik, who cites Temminck, Schlegel, and Peters as witnesses in evidence of its absence in the Leyden skeleton in question. In my examination of the same specimen I certainly only found six. Professor Brandt also throws doubts on De Blainville’s assertion ; and he himself, in a study of the Sirenian neck-vertebre, holds, from analogy in the disposition of the cervicals of Halicore and Rhytina, and the way in which the head of the first rib articulates, that the seventh is that which is wanting. The first dorsal, however, or numerically the seventh from the cranium, he is inclined to regard in the light of an anomaly—functionally a dorsal, yet in some way a cervical. Somewhat incongruously I think, while admitting on sound grounds but six for the neck, he would do away with this apparent exception by the less stable assumption of a cervical simulating an undoubted true dorsal. Professor Flower, in a short communication’, very sensibly argues against Brandt’s opinion. Basing his reasoning oi the cervical irregularity extant in the Sloth, as elucidated by Bell and Turner, and on the individual characters of the seven cervical vertebrae of the Dugong compared with those of the Manatee, he concludes that morphologically the sixth is the missing one in the latter animal. For my own part I venture to dissent from the above distinguished authorities; and suggest that it is the usual third cervical of Mammals which is the undeveloped or absent one in Manatus. This conviction I am led to adopt for several reasons. In Cetacea with ankylosed cervicals more generally the third is the least distinct, the fourth, fifth, and sixth by degrees evincing greater separation. In adult Sirenia occasionally the axis and so-called third and fourth are found partially united. In them also the three vertebra succeeding the axis, although subequal in thickness, do show slight successive increment, so that, ceteris paribus, the missing true third one would be most reduced, and its thinned body and lamina more readily coalesced to non- detection with the enlarged axis. Again, in my dissections (vide fig. 29) I found that there is a tiny accessory tendon of the scalenus muscle, which comes from a small triangular fleshy slip alongside of the larger axial division, and is fixed immediately hehind it to the same vertebra. The third nerve passes between them. This diminutive additional tendon, therefore, completes the normal number seven of the cervical attach- ments of the scalenus, notwithstanding there being only six well-developed neck-ver- tebre ; moreover its relation to the third nerve is, I hold, important. Inferentially this 1 Nat. Hist. Rev. 1864, p. 259. y2 138 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. and the other reasons adduced point somewhat weightily to the presumption that the third is the deficient cervical vertebra in the Manatee. Regarding dorsal vertebra, Krauss’s tables show 16 and 17 to be most frequent, though Stannius records 15. I select four examples from my own observations in support of inconstancy in dorsal vertebrae and ribs. In the Stuttgart Museum there is a skeleton of a young animal (received from Herr Kappler, 1864) of which the spinal numbers are 6 cervical, 16 dorsal (with ribs), and 25 lumbo-caudal, =47. In the Zoological Society’s juvenile male there obtained 6 cervical, 17 dorsal (with ribs)’, and 25 lumbo-caudal, =48. In an adolescent skeleton in Heidelberg Anat. Mus. I counted 6 cervical, 17 dorsal (17 ribs on left and 16 on right side), lumbo-caudal 25, but possibly two ossicles deficient, =48 or 50% In the Zoological Society’s female the numbers (with ribs) were, 6 cervical, 18 dorsal, and 27 lumbo-caudal, =51 in all. Excepting the Heidel- berg specimen, the terminal tail-elements were intact, being connected by ligaments. The so-called lumbar vertebre are two or three, according to circumstances. Thus, one thing with the other, it results that the spinal column of the genus presents con- flicting anomalies. 2. The Spinal Ligaments. There is a certain amount of rigidity in the spinal column of the larger-sized Manatus, resultant from the very limited amount of intervertebral substance. In the older female specimen the thickness of the intervertebral cartilage barely exceeded one tenth of an inch, the bones in consequence approximating very closely. This deficiency of the elastic cushion of soft cartilaginous substance is not confined to any one region of the spine, but is met with from the cervical almost to the tip of the caudal vertebre. In the young animal there is much greater flexibility of the spine; as, indeed, one would anticipate, seeing this is the universal rule in the Mammalia. Counterbalancing the deficiency of intervertebral cartilage, there is an ample develop- ment of ligamentum subflavum in all the spine-bearing vertebra. In the neck, as might be expected from the antero-posterior compression of the six bones, it is thin, and accommodated to the wide arches. Between the dorsal spinous processes, especially from the fifth backwards, it is remarkably thick and strong. In some interspaces it measured above an inch broad, and almost as much deep. From the true dorsal ver- tebree backwards it decreases in the ratio of the size of the spines. This ligamentum subflavum is composed chiefly of yellow elastic tissue. The equivalent of the anterior common ligament of the vertebral column, here hori- zontally placed, is only of moderate breadth and strength. From the seventh dorsal segment to the lumbar vertebre the ligament in question is mainly noticeable as existing between the keeled part of the ventral processes. ' | have to regret an error in Plate XXVI. fig. 37, where eighteen ribs are drawn. I bear due share of blame for not detecting in time that my artist had mistaken a fibrous band coming from the cartilaginous tip of the first lumbar as a rib; the eighteen costs in the female doubtless also misled us both at the moment. DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 139 The posterior common ligament within the spinal canal was not examined. I observed in nine or ten of the anterior dorsal vertebre an extra ligament. This passes as a more or less strongish band from the posterior surface or border of the lower portion of the vertebral lamina to the anterior border of the transverse process, and over the articulating process. From the fourth to the seventin vertebra it is well marked, but is less distinct in the succeeding ones. The capsular ligament is divided by a strong interarticular spur from the inter- vertebral. The two synovial cavities are very distinctly separate. Of the true ligamentous bands lashing the cost to the spinal elements, each stellate ligament is only imperfectly divided into two bundles. The anterior costo-transverse ligaments of human anatomists are wanting. But in all the ribs there are developed short, but immensely strong, middle costo- transverse ligaments. These are situated deeply, and pass in an oblique direction between the ribs and the transverse processes. They are covered by a portion of the external intercostal muscles, and partly surrounded by the intervertebral plexus of vessels. Every one of the eighteen posterior costo-transverse ligaments is remarkably broad and strong. Along with the stellate ligament the middle costo-transverse ligament prevents the rib rotating too far forwards. 3. Limbs, Pelvis, and Ligaments. Vrolik’s portraiture (pl. 3) and remarks (7. ¢. p. 69) on the progressive development of the pectoral limb are sufficient for practical purposes. In his larger specimen the ossific centres of the phalanges, three to each and two to the thumb, are not quite so rigidly defined as in that coming under my own observation. The ligaments connecting the limb-segments are very simple, the flattened condition of the bones obviating much differentiation ; and each joint is uncommonly lax. The tendon of the subscapularis pierces and greatly strengthens the shoulder-joint. The pelvic bones of two males at different ages have been delineated in the ‘ Bijdragen’ (pl. 5); and, excepting in greater circularity of the central mass, my young male agreed. In the older female ossification had proceeded further. The figure of the bone is furcate or semilunar, therefore differing from the adult male, where it has an irregular diamond-shape. ‘The extremities of the horns are cartilaginous, the rectus abdominis muscle being inserted between (vide fig. 50); and the posterior concave border has likewise a cartilaginous rim, to which the ischio-coccygeus is affixed. On the inner border the transversus perinei &c. are attached. This edge, therefore, represents the ischium, the tuberosity being that turned rearwards; the anterior cartilages tipping the cornua are respectively pubis and ilium. The surfaces of the bone are smooth, and indeed slightly concave; but all limb-structure is absent. The relative dimensions of the pelvis were :—male, } inch in diameter; female, about 14 inch in long and 1 inch 140 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. in cross diameter. Besides suspensory ligament derived from the transverse process of what may be considered a sacral vertebra, there is another, equally strong, passing inwards from the pubo-ischial region to the vulva, behind which it meets its fellow of the opposite side. This interpelvic bridge appears to be the homologue of a sub- pubic ligament. 4. Cranium and Dentition. So often has the skull of the Manatee been described and figured, that I restrict myself to a short notice of its interior, and to a few remarks on the foetal cranium. The sections chosen to illustrate the cranial cavity are the internal basis as opened horizontally or with calvarium removed, and a longitudinal vertical section to the left of the middle line. The former (fig. 56, Pl. XXV.), of the female specimen, has the dura mater attached on the left side, but the bones have been cleaned on the right ; the latter (fig. 87, Pl. XXVI.), of the younger male, is part and parcel of the sectional view of the body with membranes and organs in situ. The interior basis cranii may be likened in figure to a broad stirrup, being very square across the supraoccipital region, and arching regularly round from the temporal to the frontal region. It is flattish below, as is the vault; but the sides of the entire cavity are steep and but slightly arched mesially. The length and breadth of the cranial cavity are nearly equal; but the height is rather less than either. In the female skull (that depicted in fig. 36) the two horizontal diameters were about 31 inches, the vertical close upon 2? inches. The skull’s walls are very remarkable as regards inequality of thickness. The vertex, to say the least, is as solid a piece of bone as can well be conceived, whilst the sides, especially at the temporal region, are quite the reverse of this namely, a thin plate of bone. To specify, and with reference to our figure, the nearly vertical section of the frontal bone in this female was above three quarters of an inch deep, the temporal and part of the parietal plates little more than a line, the occipital wall less than half an inch at its middle, but thicker at the sides. The structure of the bone also varies. The frontal is, to an extreme, dense and compact, as, indeed, is the osseous consistence generally; only a film of diploé is apparent on the lateral walls, and the occipital segment has a distinct (though finely cancellated) interior, with a thin outer vitreous table. The internal basis cranii examined, as in our figure, with the dura mater retained on one moiety, presents a notable difference on the two sides. In that with the membrane remaining (the left side, but right of fig. 36, as seen in the Plate) there are two subequal- sized oval fossx, divided by a nearly transverse arched membranous ridge. These correspond respectively to the posterior and middle fosse of, say, the human skull. The anterior fossa of Man, in the Manatee is nearly perpendicular, or forms the front cranial wall, and hence is only partially visible on looking directly downwards into the cranial cavity. Those fossee present in Manatus contain the anterior and posterior DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 141 cerebral lobes. An ethmoidal elevation or ridge of bone is well marked; outside it (partly seen in fig. 36) is the broad depression which lodges the olfactory bulb (1); the optic foramen (2) at its posterior border is barely distinguishable. A narrow elliptical slit (5) pierces the outer wing of the arched transverse fibrous septa, and transmits the fifth nerve. In the posterior fossa, behind the septa and partly to the inner side of the slit spoken of, there is a broad hollow (¢), wherein the temporal lobe of the brain lies. About the centre of the cranial floor and within the area of the somewhat triangular inner horn of the transverse fibrous septa from before backwards, are as follows:—a tiny elevation representative of the lesser wing of the sphenoid; outside it a perforation in the membrane corresponding to the sphenoidal fissure (3), transmitting the third and fourth nerves &c.; behind that a shallow pituitary fossa (p); and lastly, posteriorly, an oblique groove and membranous perforation for the carotid artery (ca). About the middle of the posterior fossa, in the valley of the nodular petrous eleva- tion, is the large internal auditory foramen(7). Rearwards of this, at the occipital, is a wide groove running outwards and backwards, the membranous covering of which hides the jugular portion (7) of the foramen lacerum posterius. Outwards from this, in an angular corner of our section, are the great vascular plexus (Px) and vein which fill the deep lateral sinus. The spinal vascular plexus (Px*), which communicates with this, occupies a considerable portion of the foramen magnum. On removal of the dura mater (as the opposite half of the figure shows), the bones of the cranial basis display material distinction from the fossa as covered by the membrane. They agree in most respects with the allied form Halicore, to some extent with the young of Elephas, and certainly resemble these more than they do Cetaceans. The frontal, fenestrated plate, and stout crista galli of the ethmoid form the anterior perpendicular cranial wall, partly assisted below by the fair-sized orbito-sphenoid (0.8). The cribriform ethmoidal lamella is oval, fully half an inch in vertical depth, and freely perforated. A groove leading to the small optic foramen (2*) runs obliquely outwards; and immediately external to this is a large irregular ovoid perforation, separating the orbital from the alisphenoid. ‘This interspace, one inch long by half an inch broad in our female specimen, is partly freely open and partly blocked up by a semiglobular projection of the dental portion of the maxillary bone (Mz), which in this case con- tained the germs of molars. The alisphenoid plate (As), ankylosed with the basisphenoid (Bs), constitute the mid floor; the former constitutes a part of the lateral cranial wall, being wedged in between the squamosal parietal and tip of the orbito-sphenoid. The suture betwixt the basioccipital and basisphenoid is distinct, that between the latter and orbito-sphenoid less so. A pair of aborted nipple-like leaflets of bone appear to represent the anterior clinoid processes; and behind these are a few foramina. Backwards from the basisphenoid is the flat bar of the basioccipital (Bo), which, forking outwards with the inner spur of the exoccipital (£o), circuit the lower margin 142 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE, of the great foramen magnum. But the two more interesting phases of the interior osseous cranial construction are an immense fissure (a continuous foramen lacerum me- dium and posterius) and as remarkable a development of the periotic (Per). The great fissure spoken of forms a considerable segment of a circle, broad and irregularly con- toured in front, and narrowing as it sweeps inwards and then round the periotic. It is bounded laterally, forwards, and internally, respectively, by two divisions of the periotic presently to be mentioned, a tip of the wedge squamo-parietal, the posterior border of the alisphenoid, and by the basioccipital. Its narrow posterior horn, or what corresponds to the jugular portion, dips between the posterior border of the periotic and exoccipital, and communicates with the great inferior basal petrotympanic cavity. The massive and dense periotic within the skull is bicuspid, and occupies nearly half the interior. The anterior smaller division partially constitutes the lateral cranial wall, and abuts upon the squamo-parietal wedge behind the alisphenoid. The posterior larger division (=pars petrosa) juts across the cranial basis, as a thick nodular mass, behind the above-men- tioned foramen lacerum medium. Its upper moiety is swollen, a prominent node marking the semicircular canals (sc*), on the posterior surface of which is a vertical fissure (aqueductus vestibuli?). The lower moiety is separated from the upper by a transverse sulcus, superior petrosal groove, near the anterior end of which is the meatus auditorius internus (7), and above and forwards by two foramina (=hiatus Fallopius and lamina cribrosa 2). The great cranial fissure is ordinarily closed above by the dura mater, as has been shown; and beneath this is a large sac, connected with the Eustachian tube, and com- municating with the tympano-periotic fossa. The lower wall-membrane of this sac reaches from the alisphenoid to the exoccipital and stylo-hyal cartilage, and crosswise from the basiocciput to the tympanic. The youngest M/anatus skeletou which I have had access to is that in the Amsterdam Zoological Gardens, and said by Vrolik, in his memoir, to be that of a fcetus. Each half of the inferior maxillary bone apparently has had three centres of ossifica- tion, at least is suturally divided into three areas (1, 2, 3, fig. 16)—namely, symphysial, angular, and ascending ramal divisions. The sutural lines of demarcation spring triradially from the proximal end of the body of the bone, and are pretty regular in their course, that across the ramus being the longest. The frontal bone (fr) is bilateral, as Vrolik has shown’; and a large fontanelle mesially divides the parietals backwards to the supraoccipital. The coronal suture runs in nearly a straight line across the vertex. A parieto-squamal suture is well defined. The supraoccipital (So) is a single transversely oval-figured bony area, quite separated from the exoccipital by interfibrous material, and laterally bounded by broadish fontanelles ( fo”), which continue backwards and divide the temporal from both. The tympanic (Zy) and squamo-malar 1 Bijdragen, pl. iv. fig. 13—an upper view of skull, but which I have supplemented by two other sketches of the same specimen (Pl, XXII.). DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 143 elements of the latter are very distinct. The exoccipital (Zo) is in two subquadrate halves widely apart, the foramen magnum (fm) being surrounded by membrane and fibroid tissue. The basiocciput (Bo) is free, its basisphenoid articulation, as in older animals, being unossified. Each alisphenoid (As) is disconnected from the basisphe- noid (4s); and behind them the membrane of the considerable-sized Eustachian sac is left intact (fig. 17, Hus). The palatines (Pl), maxille (Mz), premaxille (Pmr), and Jugal bones (Jw) have their lines of approximate union very marked; how many ossific centres each had I did not note. In the adult, at the outer posterior angle of the orbit, a bony process is sent up from the jugum; this is a sesamoid or separate ossific element (s) in the foetal skeleton. > In the above fcetus, on each side, a pair of spaces indicated the future molars in their saccular condition, and a tiny orifice a premaxillary incisor. In our Society’s specimens with difference of age the same conditions obtained, viz. five grinding-teeth in use and a sixth almost erupted, whilst in the cavity behind there was evidence of at least three more in an undeveloped state. Minute denticles representative of a pair of upper and lower incisors I distinctly detected. 1V. THe Muscunar System. To my knowledge Stannius is the only author who has treated of the myology of the Manatee’; and his descriptive remarks are chiefly confined to a very general account of the abdominal and caudal muscles. These he has compared with those of Cetaceans, taking the Common Porpoise as his type. He briefly points to certain resemblances between the tail-muscles of the two, shows that the cutaneous panniculus, the muscles of the abdomen, and the so-called psoas muscles differ in the one form and in the other. But the restricted mauner in which he traces the homologies, and the fact that he has left unnoted the muscularly clad anterior extremity, the extraordinarily developed facial muscles, and the large deep muscles of the otherwise shortened neck, render it desirable that further demonstration of the fleshy structure of this singular mammal should be placed on record. The Manatee’s pseudo-Whale-characters (herbivorous Cete of the Cuviers and others) and Gravigrade tendencies (of Blainville) cause me to compare its myology respectively with Whales and the Elephant. Laurillard’s” superb delineations serve well my purpose for the latter; Stannius’s, Carte and Macalister’s*, and my own dissections of Cetacea abundantly supply me with material for the former. 1. Muscles of the Axial Skeleton. (A) Those connecting the Spinal Column.—Dorsal Aspect.—I shall take the deep fleshy and tendinous bundles upon the dorsal surface of the spine as the starting-point whereon to build up the muscular structure incorporating the soft frame of the 1 Loe. cit. p. 34. ? Recueil de Myologie, plates 272 to 295. * Phil. Trans. 1868, p. 218, VOL. Vill.—PaRT 1. September, 1872. Z 144 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE, Manatee. Here we find that definition into separate or individual muscles is not easy, from the very fused condition of the parts. There may be traced, however, through the length of the dorsal and what constitutes the lumbar region oblique sets of fibres which answer to those of the multifidus spine and semispinales. The levatores costarum, corresponding to the number of ribs less one, are more easily defined than the preceding; but they also have their fibres much intermingled with the long spinal muscles presently to be spoken of. As regards the interspinales, these either are aborted or so masked by the volume of interspinal yellow elastic ligamentum subflavum that their function is supplanted by the latter. From that close adherence of the mass of the erector spine to the tissues beneath, intertransversales muscular slips are chiefly apparent in the lumbar and caudal regions, reference to which shall be made further on. What corresponds to the combined or continuous spinalis dorsi and levator caude internus is a long, narrow, but, in the back, vertically deep muscle, which runs from the neck backwards as far as the end of the tail. Anteriorly, where laterally com- pressed but fleshy, it fills vertically the hollow between the cervical spines and trans- verse processes. Posteriorly it becomes tendinous and aponeurotic, and is fastened to the caudal vertebre superiorly. There is a very massive and in great part fleshy longissimus dorsi, which extends outside the last from the first rib backwards to the very end of the caudal vertebre, thus including what constitutes the Jevator caude externus of most other Mammals. Like the preceding the tail-tendons are interwoven into an aponeurosis, partially fixed to the transverse and to the spinous processes. The well-marked sacro-lumbalis is a rather narrow but thick muscular elongation, lying upon and firmly attached to the whole of the ribs outside their angles. Its outer tendons are short and fixed to the coste along with the fleshy part of the external border of the muscle; the internal tendons are even more imbedded among the mus- cular substance. At the first rib the sacro-lumbalis is very narrow, but is broader towards the middle of the body—in the larger specimen being 2 inches in transverse diameter. At the last rib there is a fusiform muscle, almost like a continuation of the sacro-lumbalis, but which I shall describe along with the subcaudal series. A few fleshy fibres are continued forwards on to the axis, from the sacro-lumbalis, and a still larger amount from the longissimus dorsi; but, as might be expected from the remark- able shortness of the Manatee’s neck, neither of these compressed bundles is of much import functionally. They are individually homologues of the cervicalis ascendens and transversalis colli. There is, moreover, a better representative of the trachelo-mastoid, which is a much longer and distinct muscular band, proceeding forwards from the edge of the trans- yersalis colli, and is inserted into the skull. The cranial attachment is upon the DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 145 exoccipital bone, betwixt the cephalo-humeral and the complexus muscles; the united superior obliquus and the rectus lateralis are situated below and within. The splenius and complexus appear almost a continuation of the long internal spinal muscle; there is, however, a distinct separation, excepting a few of the fibres. Of the two the splenius (Sp, fig. 8) is much the smaller. It diverges, so to speak, from the fleshy fibres of the complexus outwardly, opposite the second rib, and proceeds broadly to the skull, where it is inserted by a short, flat, but very strong tendon into the exoccipital ridge above and behind the cephalo-humeral. ‘The larger, thick and long complexus arises from the outer side of the spinalis dorsi, above the head of the sixth rib. It covers the remaining interspace between the ribs and the compressed anterior portion of the spinalis dorsi forwards to the cervical vertebre, where it spreads out and lies superficial to the short recti and oblique muscles. It is inserted into the whole of the back of the cranium, as far outwards as the paramastoid. The short deep muscles of the back of the neck are well represented, in spite of the diminution of the posterior cervical vertebre; for the atlas and axis are still of fair dimensions. ‘They show no deviation in attachments from those of ordinary mammals. The rectus capitis posticus major and minor have coalesced fleshy fibres, the former being much the larger of the two. The odliquus inferior is well developed, and somewhat fusiform. The obliquus superior and rectus lateralis are closely united, and together form a short fleshy band. These posterior short muscles of the neck Stannius' partly treats along with the semispinales in the Porpoise; but in the Pilot Whale, Macalister? and myself have both found a very large rectus posticus, apparently major and minor. I have also defined obliqui in the same animal. In the Elephant* there are a distinct voluminous rectus capitis major, an obliquus superior, and obliquus inferior obtain. Ventral Aspect.—Before drawing attention to the descriptions and opinions of Rapp and Stannius concerning the presence and homologies of the infralumbar and subcaudal muscles of the Dolphin and Manatee, I deem it preferable to render an account of my own dissection of the parts in question. In Manatus five, or at least four, distinct muscular masses can be traced without difficulty, as superimposed in two broad flattish layers, with an additional lateral or outlying fusiform one. The first and notable muscle is that which in the profile and under-view appears as a great and the only mass filling the interval between the last rib and the caudal extre- mity, and the space between the chevron bones and the tips of the lumbo-caudal transverse processes. ‘This aspect is in some respects deceptive, as the muscle, when manipulated by the scalpel, is found to be only one of two thick and long layers occupying the area in question. The superficial stratum or musculo-tendinous lamella arises from the outer half and inferior surface of the last rib, being here partially overlain by the external oblique and panniculus; thence, with inwardly oblique fibres, 1 Loe. cit. p. 30. 2 P.Z.8. 1867, p. 481. 3 Myologie, pl. 282, lettered M’, N, and N*. Z2 146 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. it is inserted mesially from the third chevron bone, backwards to the termination of the spinal column, and outwardly is fixed to the tips of the transverse processes. Anteriorly the muscle is strong, thick, and very fleshy; but halfway along the tail, and nearly throughout the middle line, it becomes tendinous, by degrees thinner, and towards the end is little else than a dense glistening aponeurotic fascia with coarse tough fibres. These fibres, when unravelled with care, separate into broadish tendons, one to each vertebra, which posteriorly commingle with the great flat tail-aponeurosis. The second or deeper muscular lamella, also taper-shaped, is as a whole much thicker and fleshy, but not quite so broad as the last. Besides a very small slip ante- riorly derived from the last rib, it has firm attachments along the under surfaces of the two lumbar and all the caudal vertebre, filling the interspace betwixt the vertebral bodies, the sides of the chevron bones, and the distal extremities of the transverse osseous elements. This sheet, like the former superficial one, is fleshy anteriorly and tendinous inwardly and behind. Its terminal fasciz or tendons are more cord-like, and with less difficulty resolvable into separate elements. The direction of the fibres of no. 2 are somewhat more backwardly oblique than no. 1. Neither of these two muscles, be it noted, passes underneath the diaphragm, but stops short quite abreast of its posterior surface. The next muscle (or pair of muscles) is very diminutive compared with the foregoing. It lies on the inner and anterior aspect of the deep caudal layer, and partly passes for- wards beneath the diaphragm. ‘The innermost and slightly longer of the two arises by tendinous and fleshy fibres from the sides of the bodies of the last two dorsal vertebree and of the vertebral end of the final rib; narrowing posteriorly it is inserted on the first chevron bone. The outermost is attached in front to the last rib and behind to the outside of the same chevron bone as its neighbour. Nerves apparently representative of the lumbar plexus issue between these two muscles. Lastly, if considered amongst the subcaudal muscles, and not what it to some extent simulates, a continuation of the sacro-lumbalis, we have the lateral or superficial out- lying fusiform muscle intermediate between the dorsal and ventral surfaces of the tail. This numerically fifth infracaudal muscle, narrow, roundish, and tapering, has origin close to the termination of the sacro-lumbalis, from the cartilaginous tip of the trans- verse process of the sacral or first true caudal vertebra, and lies horizontally along the next eight processes. It terminates in a long but strong tendon upon the surface of the subcaudal muscle (Sc), mingling with its fascia. My interpretation of these muscles is, that the superficial great broad layer represents an expanded sacro-coccygeus, in this case extending more than usually forwards, and the caudal tendons (each separate in most quadrupeds) are here coalesced into an apo- neurotic sheet, adapted to the osseous and anomalous fleshy tail-formation. The layer beneath is an enormously developed infracoccygeus, with which muscle it corresponds in position &c. The anterior pair of small slips agree best with the guadratus lum- es DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 147 borum in every respect save size. The last superficial muscle ought possibly to be included among the dorsal series. It is evidently the homologue of a small muscle met with by Mivart and me in Galago crassicaudatus. In that animal it lies outside the origin of the levator caude externus, and comes from the lumbo-iliac fascia, and is inserted by tendons on the side of the root of the tail above the sacro-coccygeus. Doubtfully named by us', I here denominate it in the Manatee lwmbo-caudalis. I do not recognize in Manatus any division of the infralumbar muscles agreeing with psoas and iliacus. If these are present they are indivisibly fused with the infra and sacro- coccygeus, and, besides, can have no limb-attachment. The great inferior loin- and tail-muscle of Cetacea Rapp regarded as a psoas; and he describes as costalis one of the outer dorso-caudal muscles. Meckel and others adopt a similar interpretation; but Stannius, in his myology of Delphinus, Phocena, and the American Manatee, opposes Rapp. He names in the former a sacro-lumbalis superior and sacro-lumbalis inferior, a longissimus superior and inferior, a transversarius superior and inferior, a caudalis inferior, besides intertransversarii. He regards the Sirenian caudal muscles as nearly equivalent, more particularly laying stress on the so-called transverse muscles, these being below, as I presume, my sacro- and infra-coccygeus, and above the levatores. Of the deep muscles of the ventral surface of the neck, the longus colli, which is altogether broad and flat, may be reckoned as consisting of three triangular parts. These, however, are not very readily separable into distinct portions; but the difference in direction of the fibres and attachments sufficiently define them. The first or pos- terior portion, homologous with the inferior oblique portion of higher mammals, covers the under surface of the transverse processes of the last two (fifth and sixth) cervicals and roots of the first two ribs. The second anterior or superior oblique slip of fleshy fibres arises widely from the ventral surface of the atlas, and is inserted narrowly and superficially tendinous into the rudimentary transverse process of the fifth cervical vertebra. The third inner and largest moiety of the longus colli has origin apically from the middle of the body of the atlas, and, widening on the surface of the neck, is attached to the inferior surfaces of the bodies of the succeeding cervical vertebra. The rectus anticus minor is seen on the outside of the rectus anticus major, and is fairly developed as a fleshy band whose origin is behind and beneath it cranially. It is inserted on the outer and under surface of the atlas. Considering the diminished extent of neck, the rectus anticus major is remarkably large and comparatively long. As in the Galagos and other aberrant Lemurs, it extends from the basilar process of the skull backwards to the bodies of the anterior dorsal vertebre. In the thorax it ends in two long, flattened, strong tendons—one to the 1 In Trans. Zool. Soc. yol. vii. pl. 2. fig. 3, lettered Jt.cd, it is regarded as perhaps an anterior prolonga- tion of the intertransyersarii caude ; but in the tail-dissection, pl. 6. fig. 25, where fully exposed, we leave it unlettered. 148 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. middle of the under surface of the third dorsal vertebra, and the other to the under- | side of the head of the second rib. At the cranium the muscle is fixed by fleshy fibre. Portions of each longus colli tertius are exposed between the bellies of the recti antici majores from the atlas backwards. (B) Those of the Skull or Cephalic Segment: Facial or Supracranial._—The three muscles of the face respectively recognized by anthropotomists as the pyramidalis nasi, the compressor nasi, and the dilatator naris, each and all appear to be well developed in the genus IManatus, notwithstanding that their fibres are indefinitely united. In the remarkably deep but narrow hollow intervening betwixt the maxillary bone and the nasal cartilage there lies a strip of muscular fibres, much intermixed, however, with what appears to be fatty tissue. The fibres possess a partly transverse and partly oblique direction. At the upper part the transverse muscular structure is necessarily short, from the configuration of the parts, but forwards from this by degrees lengthens, becomes more oblique, and as a thick bundle fills the bony depression above the zygomatici. Mesially situated, or upon the nasal cartilages, the fibres curve archedly over the nares and meet those of the fellow muscle of the opposite side. I am inclined to regard the upper narrow but deep portion of this combined muscle as homologous with the pyramidalis nasi (see fig. 12, P.n)—those fibres that cross the naris, with the compressor nasi of human anatomy (figs. 10, 11, 12, C.m)—and the most anterior fibres, or those that deeply encircle the aperture of the nose, with the dilatator naris (D.n, fig. 12). These last, moreover, appear to include those diminutive human muscles styled the levator proprius ale nasi posterior and levator proprius ale nasi anterior. The anomalous fibres of Albinus, or nasal rhomboideus of Santorini, may here be represented by a longitudinal slip at the outer border of the above triadherent muscle. The fibres of the said slip arise from the inner aspect and upper surface of the orbit, and, running obliquely inwards and well forwards, mingle with the premaxillary portion of the foregeing. Unless what has been taken as Santorini’s rhomboideus is a displaced zygomaticus minor, then there is present but one well developed zygomaticus muscle. This arises from the deep infraorbital fossa, and is inserted into the anterior portion of the naris, there interblending with the depressor labii superioris aleeque nasi. The infraorbital arteries and nerves, as might have been expected from the vast size of the muzzle, are of large size, and lie alongside of and upon the zygomaticus muscle. A levator labii superioris proprius 1 identify in a broad fan-shaped or triangular layer of muscle, which arises apically from underneath the projecting orbit, and expands upon the sides of the nares, front of the muzzle, and upper lip. It is much shorter than the layer covering it, but is equally fleshy, and rather the thicker of the two; in magnitude it is much greater than the zygomaticus, which it overlies and hides. A few only of the fibres of the levator labii superioris proprius proceed towards and over the DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 149 upper nares, the greater amount going to the lip and and anterior infranarial region. The latter are sent inwards in distinct transverse lines, more particularly the deep layer. The section of this part thus resembles, on a small scale, the trunk of the Elephant when cut across—the tissue intervening between the muscular bundles and fibrille being fatty and fibrous. 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.” The true action of this muscle here may most plausibly be assumed to be a dilater of the nares. Deeper than the last, is a broad and thick plane of muscle, which, issuing from underneath the projecting orbit, proceeds forwards by parallel fleshy fibres, less broad than the preceding. These wind round the anterior portion of the intermaxillary, and lie above the buccinator, being inserted into the incisive fossa. ‘The upper border of this muscle, and its anterior portion, have apparently oblique fibres, which, being difficult to dissect, are readily cut across, and have a coarse aspect. These are what may represent the levator anguli oris, or be part and parcel of the combined levator labii proprius and levator anguli oris—here, however, not clearly separable. I name more than define a depressor labii superioris aleque nasi a thick mass of the deepermost fibres of the muscle just described, and partly continuous with the zygo- matici. These may represent, in a modified manner, the muscle in question. It is not at all clear or distinct, excepting by an alteration in the inner and narial fibres of these muscles. It seems to constitute a muscular layer directly in front of the intermaxillaries, stretching from the gum round the external aperture of the naris. The muscular layer the most superficial of those upon the muzzle, I take to be the equivalent of the-levator labii superioris aleque nasi. It is of considerable thickness and great breadth, and almost throughout fleshy. Trapezoidal in figure, the four unequal sides respectively form the medio-nasal, the orbital, the labial, and the muzzle boun- daries. Fibres arise in a longish peaked manner from the outer side of the nasal cartilage upon the superior maxillary bone; thence they spread downwards and forwards, part winding outwards round the orbit, and part inwards to the nares; but the main body of the muscle has a median plane and covers the entire frontal superficies of the nares and upper lip. Below the nasal orifices, towards the median line, the fibres pass inwards curvilinearly, and are dovetailed with those of the levator labii superioris pro- prius, and partially inserted into the intermaxillary bones. They likewise cross above the nares and decussate with those of the opposite side of the face. Concerning this muscle’s action, the attachments and direction of its fibres show that it is an elevator, retractor, and dilater of the nares. Mandibular Arch and Side of Skull= Infracranial—The muscles clothing the symphysial portion of the mandible greatly increase the remarkable appearance and unusual form of the bone of this region. Of these the levator labit inferioris is * Memoir cited, p. 59. 150 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. represented by a short but fairly developed mass of muscle and fibroid tissue, which covers the anterior inferior blunt point of the mandible and mingles with the submucous tissue of the lower lip. Laterally, covering the outside of the swollen symphysis, and reaching the concavity of the horizontal ramus, is the depressor labii inferioris, a broad, fleshy, and most unusually developed plane of muscle. From its inferior attachment the fibres incline upwards and forwards, and intermix with the submucous tissues of the mouth and under lip. The two muscles of opposite sides traverse round the downwardly pro- jecting symphysis and unite behind; but in front of this union there intervenes a strongish connecting fascia. The homologue of the depressor anguli oris is a thick cap of muscle which projects on the lateral and anterior surface of the symphysial knee and overspreads the depressor labii inferioris. The fibres of the former have quite a different inclination from those of the latter, which cross them at a right angle upwards and backwards. ‘The two depressors of the angle of the mouth are connected by a wide, strong aponeurotic fascia on the under aspect of the bone; and each as it ascends narrows, and is inserted by membrane on the side of the lower lip. The nerves and vessels emitted from the mental foramina are distributed to this and the preceding muscle. . They are very numerous and of considerable size. In Whales neither of these three lower-lip muscles can be differentiated; the mylo- hyoid and fibres of panniculus cover the rami. The moveable lower lip of the Elephant partially derives its power from an extension forwards of the platysmal (facien and cervico-facien of Cuvier’) panniculus; but there is besides a broad and fleshy combined elevator and depressor of the inferior lip’. There is not much difficulty in recognizing the homologue of the orbicularis oris, which, as usual, surrounds the oral fissure. Its fibres are in intimate union with the anterior part of the buccinator and that portion of the fleshy panniculus or platysma which covers the cheeks. It is merely represented by a few indefinite fibres in Cetacea; but is a broad band in the Elephant (J. c. pl. 274. 1), chiefly arching to the upper lip. The buccinator (Bu) is large, long, and very thick. Its line of attachment above is the beak-like process of the maxillary and intermaxillary bones; and below the eye it fills the great vacuity overarched by the projecting orbit. It comports itself to the deep bay formed by the ascending ramus of the mandible, and thence proceeds forwards, by a broad attachment, to near the front of the upper lip. Single, thick, and fleshy in Globiocephalus, wider flat fibres in Elephas as in Manatus. Upon the surface of the buccinator, but inferiorly, there lies a long, ribband-like, but comparatively thick and completely fleshy slip of muscle, which has origin from the fossa on the anterior and inner portion of the ascending ramus. The muscle in question ' Myologie, plate 272. 2. ? Op. cit. pl. 274, n. DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 151 is covered in great part by the facial portion of the panniculus; but its fibres are quite separate, and differ in direction from those of the panniculus. It passes downwards and forwards between the depressor anguli oris and the depressor labii inferioris, termi- nating among the fibres of the former and upon the surface of the latter. It appears to counteract and check the action of the two previously mentioned muscles. Provisionally I name it mandibularis (Md, fig. 11). It may be remarked of the temporalis that, considering the great size of the bones and capacity of the temporal fossa, it is relatively small, and covered with a great mass of fat. The temporal muscle, of fair size in the Elephant’, is upwardly elongate, as is the skull; but in Whales it is the reverse of this, being set obliquely backwards in direction, short and thick. There is a double masseter. The broadest and strongest portion, relatively weak in itself, is that which, fan-shaped, and with an obliquely forward and upward direction, stretches from the outer surface of the broad mandibular angle to the descending process of the malar arch, where it is most strongly tendinous. The narrower deeper portion or layer is attached to the ascending ramus and to the hinder half and lower border of the malar arch. The fibres of this portion run counter to the upper layer; that is, they assume a downward and forward course. Thus the linear arrangement of the upper and lower muscular fibres is contrariwise or X-shaped, the diagonal of the forces between which necessarily acts in an up-and-down direction. The facial artery, nerve, and Stenon’s duct, as usual, cross the masseter, but parallel to each other, and in a nearly horizontal line. These and the muscle are entirely covered by the thick extension forwards of the panniculus carnosus. In Elephas the masseter unequivocally has two layers’; the fibres, however, are more nearly alike, and vertical, than in the Manatee. In my own dissection of Globiceps and Lagenorhynchus 1 have considered this muscle to be single, as does Stannius* in Phocena, with some additional fibres which he terms malaris externus; but in the Pike Whale‘ the masseter is stated to consist of two planes of fibres, superficial and deep. (C) Those of the Costal Arches: Thoracic—The intercostal muscles, seventeen in number on each side, are remarkably strong and fleshy. The external series are by far the thickest of the two. They are oblique in direction, but not so much as the internal series. . The diminished length of the costal cartilages causes the above muscles to stop short wide of the median line of the abdomen; but, with this exceptional circum- stance, they agree with their ordinary situation and attachments. There is an arterial intercostal plexus, as Stannius has noted, betwixt the various ribs. This at first lies within the external intercostal muscle, covered by the pleura costalis, then dips between the internal and external intercostales. 1 Recueil, pls. 276, 287, 5. 2 Recueil, pl. 276, 7, 7’. 5 L.c. pp. 4 & 5. 4 Phil. Trans. 1868, p. 323. VOL. VII.—PaRT 1. September, 1872. 2A > 152 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. From the extraordinary thickness of these respiratory muscles it would follow that they possessed equal increment of power on the movements of the chest and belly. This quality undoubtedly belongs to them; but their contractive efforts must in a great measure depend on the coordinate action of the remarkably placed diaphragm and the potent thoracico-abdominal muscles. The ribs are each and all firmly fixed at their vertebral ends, although comparatively free abdominally. The scalene muscle bounds and is closely adherent to the inner and median side of the serratus magnus. It is a long, strong, fleshy slip, more than an inch broad, and has origin at and partially covers the three anterior ribs from their angles inwards. In the neck it passes along the tips of the transverse processes of the cervical vertebra, and is inserted into all but that of the atlas. Although described as single, there appears to be an imperfect division into an anticus and posticus. The attachments of the inner and smaller portion are the trans- verse processes of the three posterior cervical vertebre and the first rib. The outer, broader and stronger portion has three cervical tendons—namely, two large and an intervening very diminutive one. The anterior large and small tendons, separated by the third cervical nerve, are attached to the outstanding process of the axis; the posterior large tendon is fixed to the succeeding vertebra. ‘The latter portion of the scalenus in question arises from the second and third ribs and interspaces. The rudimentary tiny slip of tendon referred to above, as being inserted into the axis alongside but rather behind the first tendon, is not without interest. Its presence demonstrates that the regular number of tendons exists in spite of the remarkable deficiency of a cervical vertebra, compared with what obtains in other mammals. The inferences derived therefrom have been discussed n connexion with the bones of the neck. A vascular plexus of considerable extent intervenes between the scalene tendons and the cervical plexus of nerves at their issue from the vertebral foramina. Rapp! says that the scalenus is wanting in the Manatee—a statement quite at variance with my dissection. He admits its presence in the Porpoise; so does Stannius*, who names both a scalenus anticus and posticus to the first and second ribs. This agrees with what Macalister? and I have found in the Pilot Whale; but Carte* and he only mention a single scalene in Balewnoptera rostrata. The relations of the nerves and cervical vascular plexus in the Cete bear much resemblance to those of Sirenia. There appears to be a double scalenus in the Elephant°—the superficial one, answer- ing to the serratus anticus, being fleshy, of great size, and spreading upon the chest by digitations as low as the fifth rib. So far as strength and action are concerned, the superficial throat-portion of the panniculus replaces the sterno-mastoid in the Manatee; but the latter muscle, notwith- + ¢Cetacea,’ already quoted, p. 86. 2 Loc. cit. p. 36; and Meckel, vol. vi. p. 156. 2 Pp. Z.8. 1867, p. 481. 4 Phil. Trans. 1868, p. 219. 5 Myologie, pls. 276 and 287. 6, 6°. DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE, 153 standing, is present, though greatly reduced in size. It arises as usual from the manubrium, but in front of the pectoralis major. Hence with a diminutive thin flat muscular belly, 3 inches long and 4 an inch broad, it proceeds forwards anteriorly to about a level with the outer border of the scapula, where it bifurcates (see fig. 9, S¢.m). The outer limb, a thin, strong, round tendon (shown in figs. 29, 30, Sé.m), pierces the substance of the parotid gland, and is inserted into the paramastoid; the inner fork, a broadish aponeurosis, joins the deep cervical fascia of the neck, superficial to the carotid artery, and posterior to the thick fleshy digastric muscle. The shortening of the neck, great vascular rete, &c. give peculiar relations to the parts in the Manatee. From the unusually flat and wide arched form of the ribs, the serratus magnus muscle appears to have a different position from what it has in deep-chested animals, though in reality it departs little, if at all, from its normal situation; it is nevertheless com- paratively short and thin. The costal attachments are from the sixth rib forwards. On the scapula it is fixed underneath the rhomboideus the whole length of the vertebral border and to a triangular corner of the posterior inferior angle for about an inch in extent. Anteriorly in firm connexion with the fleshy scalene muscle it extends forwards, and is inserted into the outer and posterior surface of the enlarged transverse process of the atlas (S.mg, fig. 29). The nuchal portion of the muscle has a twist upon itself, and is overlain by the broad belly and insertions of the scalenus, so that only a small segment of it is seen on the ventral surface of the neck when the parts are examined in position. The serratus in the Elephant! is very massive, and extends backwards to the tenth or eleventh rib. Meckel® says it is very small in Cetaceans, only fixed to at most four ribs; but my own researches agree rather with Carte and Macalister® and partly with Stannius‘, in their being an apparent twisting or duplicity of the muscle and greater costal attachment. The latissimus dorsi, as in the Cetacea, is entirely hidden by the superimposed layer of the panniculus, and it itself overrides in part the costal portion of the serratus magnus. On the surface of the chest it presents a broadish fan-shaped fleshy expausion, relatively short, upon the ribs from the fourth to the eighth; there is a deficiency of tendon or fascia at this costal attachment, the muscular fibres being well defined, and reaching no higher than on a level with the post-inferior scapular angle, below the sacro-lumbalis. It continues fleshy anteriorly to where it joins the teres major, the two muscles (as mentioned further on) being fixed together (La.d, fig. 8) to the middle of the shaft of the humerus. Abdominal.—Previous observers concur in noting the strange reptilian-like lengthening of the lungs and diaphragm, and consequent relatively altered position of the heart and upper or anterior abdominal viscera in Manatus; the same obtains in the two other 1 Pl. 267. 9. ? Vol. vi. p. 246. 3 Memoir, p. 224. 4 Under serratus anticus major, l. c. p. 13, and Rapp. p. 89. 242 154 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. genera of the group, Halicore and Rhytina, examined in the flesh. From the apparent correlation of rib-structure in the ancient Halitherium, doubtless it also was similarly constructed. . This characteristic formation of the thoracico-abdominal parts, alluded to by many, has hitherto not been figured, at least its position to the viscera in situ. The desideratum I have to some extent supplied in the several sketches Pl. X XVI. figs. 44, 37, 49, and 50, and also partially in Pl. XXIV. fig. 30. These demonstrate, as regards the diaphragm, that it extends in a retrogradent tolerably horizontal plane from the first to the last rib, and forms a nearly mesial septum separating the lungs within an upper, and the rest of the viscera within a lower compartment. Altogether it is very tendinous—though Daubenton’ says “ye n'y ai point apergu de centre nerveux,’ meaning, I apprehend, that there is no ordinary free central tendon; but instead, as fig. 44, Pl. XX VI. exhibits, the elongated middle tendinous raphe is fastened to the vertebral bodies, and to the keels from the fourth to the last dorsal vertebra. The fleshy portion of the diaphragm forms a narrow band on each costal margin for its entire length. In the larger (female) specimen the muscular breadth measured one inch anteriorly, by degrees increasing to two or more towards the posterior end. On the inner costal wall the fleshy attachment reaches to the roots of the very shortened forwardly median-directed cartilages (see figs. 37 and 44 respectively). Posteriorly the diaphragm forms two short crura and pillars. The external, broader one of these crosses outwards from the body of the last dorsal to the under surface of the hindmost rib, partially covers the diminutive representative of quadratus lumborum, but debars entrance of the depressores caude, though giving egress to, or rather separated by the abdominal aorta. The diaphragm at this point folds somewhat on itself, and, fastened _ to the three last ribs, but not to their tips, constitutes a pocket, by the intrusion, so to say, of the thick fold of the external oblique muscle. The cesophageal and caval apertures of the diaphragm are far forwards and wide apart from the crura and pockets in question. The deepest layer of the outer wall-muscles of the abdomen, the transversalis, has broader and stronger fleshy fibres than the internal oblique, though less so than those of the external oblique. It has origin in a digitate manner by attachments from the inner surfaces of all the ribs but the second—extending inwards, and joining the diaphragm an inch to the vertebral side of the rib-cartilages, and rather more than that on the five posterior ribs. The muscular substance extends inwards to near the outer third of the rectus, or to within about three inches of the middle line (fig. 9, Tra), that measurement being equivalent to its inner aponeurosis. A cross section of the abdominal wall shows that the fleshy part of the transversalis and of the rectus are coequal ; each muscle thus viewed has a flattened biconvex shape, the 1 Op. cit. p. 428. DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 155 rectus, as aforesaid, slightly overlapping the transversalis. The peritonzeum and a slight layer of fat cover these muscles internally ; and it is noticeable that the fat is thickest at the median line and rib-cartilages. Between the transversalis and the internal oblique muscles is the rete mirabile (Ad. Rete, fig. 9), which lies parallel to the inner fleshy edge of the former. é The superincumbent layer of the internal oblique is narrow and shallow. Like the external oblique, presently to be described, it has a digitate appearance externally, and arises by muscular detachments from the third to the last rib; posteriorly the origins are in close contact with those of the external oblique but anteriorly. The fleshy part of the muscle extends inwards to the edge of the rectus, and ends in a strong glistening tendinous fascia; this passes chiefly over the surface of the rectus to the linea alba, or forms the anterior sheath of that muscle, which joins and is lost in that of the external oblique. The said aponeurosis is here and there arranged in stronger bundles; and the whole has a forward and inward direction contrary to that of the external oblique. The external oblique muscle of the abdomen manifestly differs in appearance from that of ordinary mammalia in not being spread in a thin uniform sheet over the whole abdomen, but rather may be said to be relatively narrow, thick, and composed of a series of elongated digitations clasping the extremities of the ribs. Thus it has origin, in the manner indicated, from the costz above the cartilages, from the third to the last rib inclusive. Its coarse fibres in broad bundles, which mask or imitate separate digitate parts, pass inwards and backwards, and end in or are inserted by strong slips of fascia into the outer border of the superficial surface of the rectus. These tendinous fibrille are moreover continued in broadish strips over the surface of the rectus, parallel to the direction of the fleshy fibres. These latter, it may further be observed, with the oblique direction spoken of, proceed from one rib back towards the second behind it—that is, embrace three ribs. Posteriorly the external oblique tapers in a wedge-shape manner; and this part is inserted upon the surface of the inferior caudal muscle, and besides, by the intermuscular fascia, deeply between the last muscle and the long tapering muscle coming from the median side line of the ribs. Finally, the wedge terminates opposite the second chevron bone. In the Sirenian under consideration the rectus abdominis comports in most respects with the condition of this muscle in the Cetaceans, e.g. Globiceps and the Porpoise. Throughout its entire length, however, it retains its breadth more than in the Whale tribe. This may be partly owing to the body of the latter narrowing more posteriorly, or partly, no doubt, to the muscle itself serving mesially as a chief support of the visceral organs in the land-waddling Manatee, the ribs of the latter being widely apart compared with those of the former group. The rectus has origin forwards by a very strong tendinous fascia from the outer edge of the sternum, from its projection to the ensiform cartilage, and by some sparse muscular fibres from the first, second, and third ribs and their sternal cartilages. At this place the pectoralis minor is immediately 156 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. superficial to it, a slight portion of the pectoralis major only covering its inner edge behind the pectoralis minor. The muscle in question occupies the inner half of the abdomen (in the female specimen being four inches across) between the linea alba and the cartilage-tipped rib-extremities; and posterior to the pectoral muscles, or from the fifth rib, it is overlain on its outer border by the tendinous fibres of the external and internal oblique muscles. Posteriorly it is inserted into the cleft of the pelvic bone by a thickened pyramidal point. The rete mirabile derivative from the internal mammary artery lies underneath the rectus abdominis, as shall be described with the vascular system. 2. Muscles of the Accessory Skeleton. (A) Those of the Shoulder-girdle—The supra- and infraspinatus are each long, narrow, and thick; they are of about equal size. ‘The former is inserted broadly and muscularly into the summit and front of the head of the humerus; the latter, narrower, rounder, and more tendinous, is implanted on the outer tuberosity. The subscapularis occupies the whole of the subscapular fossa excepting the space where the serratus magnus and the rhomboideus are fixed. It is entirely fleshy, and overlaps the lower border of the bone. A short, but strong, tendon forms its insertion into the inner tuberosity of the humerus. The fleshy and strong teres major has origin from the lower dorsal edge and border of the scapula, posterior to the spine. In close union with the latissimus dorsi it is inserted into the inner aspect of the middle of the shaft of the humerus. The latissimus partly overlaps the teres; and besides the approximation of these fibres, there is a distinct fleshy slip which passes from the upper border of the former and penetrates the latter close to where the tendon of insertion commences. No teres minor was observed. A single plane of broad, but short, muscular fibres, and but little tendon, the rhomboideus, arises from the inner surface and vertebral border of the scapula, and extends for an inch or more over the inner surface of the serratus; it forms a shallow, wide muscular arch spread over the sacro-lumbalis and below the posterior end of the limited trapezius. A trapezius is present in the Manatee, the muscular portion of which is in some degree circumscribed ; for although broad and fan-shaped, the shortness of the neck reduces it anteriorly. The fleshy part is attached the whole length of the spine and dorsum of the scapula and the acromion process; therefrom it radiates, and peripherally is bounded by fascia which, with less distinct aponeurotic fibres than commonly obtains in mammals, is lost in that of the dorsal region above and behind the scapula. It is thickest and most muscular in the neck, and is fastened to the occiput. As in Cetacea, notwithstanding the shortness of the neck, a broad, moderately thick, but powerful muscle, homologous with the so-called cephalo-humeral, is present in the DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 157 Manatee. This has a fleshy origin from the entire length of the prominent ridge of the outer occipital, thence is directed nearly vertically outwards and backwards, marked by limiting the anterior flattened surface of the shoulder. It continues obliquely round the head of the humerus immediately in front of the deltoid, at this part resembling the latter in shape, and is inserted on the inner and anterior side of the neck of the humerus by a very short, but strong, tendon. The humeral attachment is on a level with that of the deltoid, above that of the pectoralis major, and close to the front upper edge of the biceps. Although a clavicle is absent, a muscle corresponding to a levator clavicule (*) exists. It is a narrow fleshy slip arising from the paramastoid directly behind the stylo-hyal and origin of the digastric. Coursing downwards and backwards parallel with the anterior border of the cephalo-humeral, it becomes fused with this last at the head of the humerus. There is a slight transverse fibrous line of demarcation where the levator terminates; and this, it may be, is a representative of an osseous clavicle. At the head the muscle is in close relation with the hinder part of the parotid gland; and a small portion of the glandular substance and several vessels separate it from the cephalo- humeral. The two above muscles apparently have a combined action, and drag the pectoral extremity forwards, upwards, and fully rotate it outwards. They doubtless oppose the latissimus dorsi and teres major; but the shortness of their leverage must detract from any great power. The pectoralis minor agrees with the pectoralis major in being relatively long and narrow. It springs from the anterior portion of the rectus abdominis over the fibrous cords or sternal cartilages of the second and third ribs, and from the projecting sharp angle of the sternum, not reaching, however, the median line. As the muscle trends forwards it narrows slightly, and finally is inserted into the head of the humerus. Several large glands lie beneath the muscle; and axillary vessels and nerves pass beneath it, the rete mirabile being closest to its deep surface. The pectoralis major muscle is thick and powerful, though relatively narrow. It has origin from the whole length of the sternum, and from the median line of the abdomen two inches beyond the ensiform cartilage, where it overlaps a small inner portion of the anterior inner edge of the rectus. The muscular fibres take a direction obliquely or sharply forwards and outwards, and are inserted, by a very strong short tendon, into the upper portion of the lower third of the thoracic aspect of the shaft of the humerus. There is only an indistinct division into sternal and manubrial portions, which might be overlooked, except that the former rolls somewhat round, or rather dips beneath, the anterior more fleshy moiety at the axilla. The humeral tendon of the pectoralis major while firmly implanted into the deep pit or continuation of the bicipital groove on the inner and lower aspect of the shaft of the bone, also possesses continuity with the radial segment of the forearm. Adduction of 158 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. the arm towards the chest is mainly effected through the upper attachment ; but increased leverage is gained by means of the more distant fascial fixed point. The continuation into the forearm is bya strong aponeurosis, which stretches in a bridge-like manner from the humeral tendon over the bend of the elbow, and is fastened to the neck of the radius at its tuberosity. The pronator radii teres shoots obliquely over this radial insertion ; and the diminu- tive tendon or second portion of the biceps joins and terminates along with it. On its outer surface the aponeurosis in question is covered by and adherent to the flesh of the brachialis anticus. The thick cord of the rete mirabile and the brachial nerves are lodged securely, and pass downwards beneath the aforesaid aponeurotic bridge ; for as extension of the forearm occurs the expanse of fascia is rendered tense, and thus prevents undue pressure or strain upon the vessels beneath. (B) Those of the Pectoral Limb. Dorsal=Extensors.—As respects the triceps, the long scapular head is fleshy, its origin extending for one and a half inch from the capsular ligament, to which it is firmly attached. The middle head arises the whole length of the back of the neck of the humerus, and has a flat broad belly. The short head occupies the whole of the back of the shaft of the bone; above it is narrowest, and inclines obliquely inwards. The three heads, as usual, unite below, and proceed to a short flat tendon to the olecranon process. No anconeus was observed. The obliquely triangular elevator and retractor of the humerus, the deltoid, is of con- siderable size and moderate thickness. Its fibres are not very coarse or disposed in protuberant bundles as obtains in Man. Origin, the whole length of the short spine of the scapula, its lower border; and it is firmly fixed to the loose acromial cartilage. From this last point its fibres trend downwards, those behind slanting forwards, the front of the humerus projecting beyond its anterior edge The muscle bifurcates slightly above the middle of the humerus, at the position of the here absent deltoid eminence; the insertions, each tendinous, embrace the biceps, and are fixed to about the middle of the shaft; but the posterior portion continues towards the elbow. The supinator longus is fair-sized. Origin, humerus above outer condyle, close to the brachialis anticus. It runs along the outside of the radial shaft, and round the styloid process, being inserted into the trapezium. It is tendinous where passing along the radial groove, and at insertion spreads out aponeurotically. If a supinator brevis obtains, it is with difficulty separated from the brachialis anticus; muscular fibres apparently continuous with the latter cover the orbicular ligament and neighbourhood a little downwards. Extensores carpi radiales longior and brevior appear to be represented by a single muscle with a condylar origin, a flat belly continuing three fourths the length of the radius. The distal tenaon, also flat, lies in the middle radial groove and widens out over the carpus, being fastened into the proximal ends of the second and third digits, DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE, 159 the tendons of the primi internodii pollicis and long supinator obliquely crossing it at the wrist. An extensor communis digitorum has an origin, as usual, from the outer condyle and ulna; its well-developed muscular mass lies mesially in the forearm, filling the broad hollow thereon. It becomes tendinous as it approaches the wrist, and divides into four flat broadish slips, which proceed in an expanded manner to as many of the outer digits. That to the third finger is rather the broadest and strongest; and the two ulnar tendons come off together somewhat higher than the radial ones. Differentiated from the last is an extensor minimi digiti, which, like it, is flat. Springing chiefly from the radio-ulnar ligament of the joint, it runs at first below and then alongside the communis digitorum. Its tendon occupies the groove on the radial side of the ulna, and at the wrist splits into two,—the short, broad, stronger division being inserted into the proximal end of the fifth metacarpal—the second, longer tendon proceeding to the proximal phalanx of the same digit. An extensor carpi ulnaris springs from the back of the outer humeral condyle, at first lying upon the pollicial extensor and then obliquely on the surface of the ulna. Its flat tendon widens as it courses round the outer distal ulnar extremity, and becomes incorporated with the strong carpal aponeurosis. Its fibres, moreover, mingle with those of the carpi radialis, whilst it finally has distinct insertions into the unciforme, os magnum, and proximal end of the fifth metacarpal. There is barely a division between what appears to represent the extensor primi and secundi internodii pollicis. ‘The former may be differentiated as a flat, broadish- based, lanceolate-shaped, fleshy muscle, which has origin in an oblique position along three fourths of the shaft of the ulna. Its tendon, also broad, flat, and strong, passes as usual in the styloid groove, and over that of the long radial extensor to be inserted into the proximal end of the first metacarpal. ‘The latter, secundi internodii pollicis, has also a lanceolate (but much smaller) muscular belly, which fills the deepish interosseous interval, and at the radial groove emits a tendon which joins deeply that of the first-mentioned muscle. The difference in direction of the fleshy fibres dis- tinguishes the above conjoined muscles. Ventral Surface= Flexors.—The biceps has a large, long inner muscular belly, and a comparatively short, diminutive, chiefly tendinous outer belly. The first or main portion of muscle arises by a short, but strong, tendon from the rudimentary coracoid process. Proceeding over the insertions of the subscapularis and the conjoined tendon of the teres major and latissimus dorsi, it is inserted by a flat oblique tendon into the lower third of the inner side of the humeral shaft, on a level with the pectoralis major. The second portion of the biceps springs from the outer or anterior border of the first belly, and, with a short rounded thin muscular development (5), ends in a narrow tendon. The latter is continued to the radial tuberosity, but mingles with the infundibular or VOL. VIII.—PaRT 11. September, 1872. 23 160 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. bridging aponeurosis of the bend of the elbow already described in connexion with the insertion of the pectoralis major. ‘The thick humeral rete mirabile partially over- lies the biceps muscle. The coraco-brachialis is absent, or indivisibly united with the first head of the biceps. Compared with the other muscular structures of the humerus, the brachialis anticus is enormously developed, and, as it is entirely fleshy, causes the lower and outer aspect of the upper limb to have unusual breadth. The deltoid embraces it above; and lower than that the brachialis anticus occupies the outer and anterior surface of the humerus. It moreover covers the anterior outer aspect of the elbow-joint, being inserted muscularly into the upper half of the radius forwardly on its radial side. It is a powerful flexor of the forearm. The pronator radii teres is of fair size; its origin is by a strong flat tendon from the radial side of the inner condyle; and, with only a moderate amount of fleshy belly, it is inserted obliquely upon the inner surface of the middle third of the shaft of the radius. Its power of pronation is very limited. The flewor carpi radialis is long, narrow, and flat, both in tendons and belly. Origin, the inner condyle to the ulnar side of the pronator teres; its distal tendon commences at the lower end of the radius, and is inserted broadly into the proximal extremities of the indicial and pollicial metacarpals. There is no special groove for its reception on the scaphoid and trapezium. The flexor sublimis, profundus, and longus pollicis muscles form a complex mass, whose origin is single and extends from the lower part of the inner condyle down the whole length of the inner flat surface of the shaft of the ulna. The thick fleshy belly,’ without possessing any clear separation, is nevertheless divisible, chiefly by tendon, into an upper and a lower layer. The upper layer, by a semidivided expanse of palmar fascia, supplies the second, third, and fourth digits. The fascia or united tendinous mass lies upon and is fixed to the broad interosseous muscles, and it is continued on to the proximal ends of the first phalanges of the second, third, and fourth digits. The aponeurosis is not finely interwoven and smooth on its surface, but is easily torn up into parallel coarse threads. The lower layer of this united muscle is fleshy for a short _ distance further than the upper layer, and does not become tendinous until reaching the palm, whereas the upper layer is tendinous above the wrist. It is inserted into the third and fourth digits, being closely adherent to the. superficial tendinous mass. From the slightly bent or oblique position assumed by the fifth metacarpal to the ulna, and the very flat condition of bones of the forearm, the palmaris longus and flexor carpi ulnaris appear to lie both on the dorsal and palmar surface of the limb. They together project, as it were, and fill the angle whose two sides re- spectively are the bones above mentioned. Origin by a flat tendon from the surface of the inner condyle immediately above the flexor communis. The bellies commence DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE, 161 above the middle of the shaft of the ulna, and end in long triangular pyramidal fleshy masses inserted broadly into the upper half of the ulnar border of the fifth meta- carpal. At the lower part they are much compressed. They act as powerful flexors and abductors of the manus, and give extraordinary breadth to the wrist. The palmaris enwraps the ulnaris, and is in continuity with the tough superficial fascia of the forearm. The interossei are most extraordinarily well developed, contrary to what might be expected in such an immoveable encased manus as is possessed by the Manatee. Not only are they large, but they are also found in a manifestly double layer; and while true flexores breves, there are also a set of dorsal interossei or extensores breves. After removal of the palmar aponeurosis of the combined sublimis and profundus, they are seen arising by tendon from the wrist, and altogether form a broad, flat, fleshy sheet, which covers the palmar surface of the second, third, fourth, and fifth metacarpals—the small flexor brevis and abductor minimi digiti muscles bounding the ulnar side of the plane. The fibres on the second, third, and fourth digits seem to possess a slight tendency to subdivide or divaricate; but those on the fifth are quite single. It is possible these superior palmar muscles may represent en- larged lumbricales. or be an anomalously developed flexor brevis manus. Beneath this another equally thick muscular layer of better-defined and somewhat double interossei exists; they pass to all the digits, excepting the first, and are shorter than the super- ficial interosseous layer. Part of the fibres between the second and third, third and fourth, and fourth and fifth digits obliquely cross the intermetacarpal spaces, and simulate dorsal interossei. These portions approximate the bones. The dorsal inter- ossei are four, and go to the second, third, fourth, and fifth digits respectively. The flexor brevis and abductor minimi digiti are each represented by thin muscular bands which run parallel to each other, arising individually from the cuneiform. The latter partly covers the fifth digital interosseous muscle; and the former is in close approximation with the insertion of the flexor carpi ulnaris: they are inserted by aponeurosis together, along the ulnar edge of the fifth metacarpal. (C) Those of the Hip-Girdle: Pelvic and Generative.—Kach pelvic bone is suspended abdominally from the first chevron process or that from the fourth lumbar vertebra, by a strong sheet of glistening membrane, and is held in place fore and aft chiefly by two muscles. The anterior one is the rectus abdominis, already described, which terminates in the anterior V-shaped concavity of the female pelvis—and in the male similarly, excepting the difference in the bone’s shape. The posterior one continues as it were the line backwards from the rectus to the chevron bones behind the anus. ‘This post- pelvic muscle is a long, broad band, throughout fleshy, whose origin, in the female as in the male Manatee, is from the posterior border of the pelvic bone. Thence it trends inwards and backwards, passing outside and then behind the rectum towards the middle line, where it is inserted,.or becomes incorporated, with the rearward 232 162 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. extension of the panniculus carnosus. Rapp' mentions, under the head of retractor ischii, or ischio-caudalis, that he found this in the Manatee thin and weak, springing from the hemal spines of the first two caudal vertebre and inserted into the pelvic bones. I apprehend he alludes to the above-described muscle, which, as far as action and attachments are concerned, is rightly named by him. If, however, we seek its homology among the four-limbed and long-tailed Vertebrata, we find it representative of the ischio- or mayhap ‘lio-coccygeus, possibly these two combined. Stannius regards the sphincter ani as double, inasmuch as he refers to the existence of an internal and an external anal sphincter. This I have found so far true that thick, fleshy, circular fibres, besides its ordinary muscular coat, surround the intestine. Above the anus, and at the external orifice, these expand broadly as they become superficial. Virtually, less or more continuous, these may be regarded as outer and inner sphincter from position. The same disposition and unusually developed condition of the anal muscle is met with in Whales. In them, as in the Manatee, the gut is firmly compressed at its outlet and above, leaving in the contracted condition but a very narrow orifice. The feeces in these two groups are consequently of small calibre, and very different from the scybalous masses of the Elephant. The lower gut in it seems altogether more capa- cious; but it is nevertheless provided with a broad muscular sphincter, as Laurillard’s figure shows (pl. 285, Q). I mention in my description of the panniculus that in the female a fleshy slip, about an inch wide, is posteriorly derived from it. This offshoot of the panniculus car- nosus, but most probably representing a separate perineal muscle, diverges from the more backwardly extended caudal fibres about opposite the generative fissure. Di- rected obliquely inwards towards the median line and posteriorly, it is inserted into the fascia beneath the skin of the perineal raphe, between the vaginal and anal_ sphincters. ‘This muscle, although apparently a continuation of or derivative from the panniculus, I regard as the homologue of the levator ani; for, besides the dermal slip in question, an additional short portion comes from the rectal surface near the pelvis and joins the former. In the female a transversus perinei appears to be represented by a longish narrow muscle springing from the inner border of the pelvis and going forwards to the outside | of the sphincter vagine and inside and behind the lesser slip of the erector clitoridis. In the male the levator ani and transversus perinei were more or less united, and with a greater transverse direction of fibre. Other muscles connected with the generative parts in the female as follows:—a sphincter vagine, consisting of a thick layer of fibres surrounding the vagina and vestibule, and which are strongest towards the perineum; an erector clitoridis (£.c), divisible into two slips: the external fleshy fusiform bundle arises from the apex of the inner pelvic horn; the smaller inner slip lies alongside the last, but has no pelvic 1 Op. cit. p. 83. DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 163 attachment ; both pass round the vagina towards the clitoris, partially decussating with the fibres of the sphincter vaginz. In the male, as Vrolik! more particularly has noted and figured, there is a well- defined fleshy ischio-cavernosus and long retractores penis. Moreover I readily recog- nized a bulbo-cavernosus. While fully developed and normal in attachment, these three muscles correspond with the somewhat altered position of the pelvis &c. from that of quadrupedal Mammals. 3. Muscles of the Dermis. The superficial fleshy investment to which the name of panniculus carnosus has been given, is enormously developed in the Manatee. Indeed, as Stannius justly observes, the ventral portion of it represents in a certain degree a supplementary belly-muscle, sup- plying the voluminous entrails with an additional powerful support, which is so much needed owing to the extraordinary shortening of the rib-cartilages. In the female, as also to some extent in the younger male, the depth of muscle varied in different regions. The muscular fibre in the former specimen, at its posterior abdominal or genital portion, had a thickness of 14 inch; but further forwards, towards the middle of the body, it decreased to 0-6 inch, thinning upwards on the back to 0:2 inch or thereabouts. On the side of the neck it equalled the midventral part in volume, thickening considerably, however, on the side of the cheek. Although the panniculus forms a uniform whole, I shall prefer, for several reasons, to describe it piecemeal. As indicated above, the extensive panniculus carnosus forms a great broad fleshy wrapper of longitudinally directed fibres, covering the entire abdominal surface (P.c') from the pectoral limbs backwards to beyond the anus (P.c***). On the side of the posterior half of the body the muscle stops short of the tips of the ribs, where part of the external oblique is exposed; but forwards from this it rises higher, the fleshy part reaching almost to the level of the vertebral border of the scapula. The line of demarcation of this lateral or costal border, however, is not so abrupt, as exhibited in fig. 8, Pl. XXI., but trends, and is lost dorsalwise in a strong aponeurotic fascia. ‘The muscular fibres of the outer shoulder portion pass forwards along the neck and side of the cheek, and partly become inserted by aponeurosis on the broadest surface of the malar bone, and, partly fleshy, are continued onwards and commingle behind the angle of the mouth with a deeper throat-layer of the muscle presently to be described. A few inches behind the axilla the belly portion of the panniculus splits into two segments, the outer or upper fork of which is that already spoken of as covering the outer surface of the shoulder. From this a slip runs towards the humerus (P.c*). The inner sternal fork, at its divarication, possesses an external tongue-shaped corner (P.c**), mesially to which the muscle lies over the sternum (in the female being 1 Bijdragen, p. 77, pls. 5 & 6. figs. 22, 23, 6, d. 164 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 23 inches broad and 13 inch thick at this point); thence it passes forwards and out- wards (P.c?) on the neck, beneath the platysmal cross fibres, ultimately being inserted as a broad and strong tendinous sheet into the malar bone, posteriorly and inferiorly to the first portion of the panniculus, as already detailed. Superficial to the shoulder and sternal segments, and as it were forming a bridge between them in front of the pectoral extremity (P.c’), is what appears as a platysma myoides. This is represented by a coarse thin sheet of muscle interspersed with much fatty tissue. The fibres may be said to arise from the inferior median surface of the neck and mandible for several inches in breadth, intermingling there with those deeper segments of the panniculus mentioned above. Crossing the neck in a transverse direc- tion, the fibres radiate slightly, and are firmly but superficially intertwined with those of the lower border of the outer nuchal panniculus, whilst fibrous tissue and fat connect them with the infraspinatus. A narrow slip of the muscle extends downwards on the anterior border of the limb as low as the middle of the biceps, whence, becoming fibrous, it descends as a cord-like body as far as the base of the metacarpal bone of the first digit (P.c**, fig. 13). At the anal or posterior end the fibres of the panniculus diverge triradially (P.c***), a broad portion curving gently outwards; and the intermediate portion (larger) ends wedge-shaped with inwardly inclined fibres, which, along with its fellow of the opposite side, are inserted by strong aponeurosis between the depressores caude into the sixth chevron bone. In the female there was also observed another fleshy slip, about an inch broad, which was given off from the body of the muscle opposite the generative outlet; and this proceeded downwards and backwards, being likewise inserted into the skin on the median perineal region. Connected with the action of this subdermal muscle it may be remarked that, as in Cetacea, it can have little or no power over the skin itself, not being fastened thereon, a thick coating of fat intervening. On the other hand the attachments of the muscle point to its subserving the bodily force of various parts. For instance, the fixed points to the malar bones must give the advanced segment of the muscle a very long leverage for movements of the head; at the same time the insertion on the facial bones and lips must lend power to the labial muscles, and in some ways act as a dilatator oris, and impart additional strength to the thick muscles of the great muzzle. V. Tue Digestive TRAct. 1. Interior of the Mouth, and Tongue. The very curious structure of both the exterior and interior parts of the mouth has been a favourite topic of those who have contributed memoirs on Manatus. Stannius’s account is, on the whole, the most explicit, though brief. In treating of the mandu- DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 165 catory and lingual apparatus, then, I propose to append to the remarks of Rapp, Vrolik, and Stannius reference to sketches of the palatal and mandibular arches. ‘These I have had illustrated, I believe for the first time, both as viewed in conjunction, in a longi- tudinal vertical section (fig. 37, Pl. XX VI.), and as separated (Pl. XXII. figs. 18 & 19). I shall also offer some observations on the composition of the parts, particularly as respects the so-called horny plates, and their bearings towards those of the edentulous Rhytina stelleri. If the section made lengthwise through the cranium, and slightly to the left of the median vertical line, as given in fig. 37, be examined, the relations of the parts will be easily comprehended. Confining a survey to the mouth-cavity, the lower lip, with its sinuous, bristle-clad, thick epidermis, points forwards. ach bristle springs from a considerable-sized lenticular hair-sac. The fibro-muscular tissues beneath are unusually well developed. The mandibular pad (7.¢. inner lower lip) is composed deeply of a thick gristly or fibrous layer, and a thinner superincumbent epidermis. In front it is separated from the outer lip by a deepish furrow, and behind stops short at the tip of the tongue, though, as afterwards shall be shown, it is continuous with the gums. The tongue is bound down behind the pad, and is incapable of being protruded. Forwards, from the soft membranous uvular curtain, the fleshy palate, to just in front of the molars, is only moderately thick, whence it by degrees increases in thickness and fills the deep concavity of the premaxillary bones, its anterior smoother portion forming the upper inner lip-pad. This latter is separated by a deep furrow from the true bearded lip and truncated muzzle. The palate and the said pad are equally made up of a thick substratum of firm fibro-elastic material overlain by much thinner derm and epiderm. What has been termed the horny plate is alone distinguishable by warty elevations. The lips and cheeks, from opposite the superior pad backwards to the front molar, have a clothing of long stiffish hairs, thickest set along the outer border of the gum; these are chiefly directed downwards and backwards. Turning now to the view of the mandible when removed (fig. 18), the outer true lip is notably dotted with short truncated bristles and longer hairs. The coriaceous pad has a deep and straight longitudinal median groove its whole length. Each moiety of this is again partially divided by a wavy but shallower furrow, which anteriorly and posteriorly curves outwards. The outer raised segments of the pad are only moderately roughened—but the inner ones remarkably so, being studded with short, erect, hard papillz of two sorts. The larger kind are conical, and about 0:5 inch high; the smaller setose sort are nearly as long, and abundantly fill the interstices between the first mentioned. The tongue quite agrees with Rapp’s statement, having long brush-like retroverted filiform papille towards the tip, many irregularly dispersed and different-sized fungi- form papillz, and a very numerous, closely arranged, double set of circumvallate glands situate at the root. The series of circumvallate, as noted by him, also extends linearly 166 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. and laterally forwards to the anterior third of the organ, and they lie in close approxi- mation with the dental portion of the gum. In the younger male the lengths of the tongue and symphysial pad mesially were 3-2 inches and 1-4 respectively, in the female 4 inches and 2; the breadth of the tongue of the latter was 0:8, and the widest portion of the inferior pad 1:3 inch. On the palato-dental arch (fig. 19) the cheeks and lips, as before mentioned, are bestrewed with hair and short stiff bristles. These latter form a scattered row, reaching from the upper external labial clump (seen on the front of the muzzle, figs. 6 & 7) backwards slightly beyond the angle of the mouth. The thick brush of hairs in the deep hollow just outside the palate is well shown in fig. 19, 6h, The long narrow palate is divisible into three portions. The anterior, somewhat horseshoe-shaped, is the smoothish, convex, elastic pad regarded by some as the inner upper lip. The middle portion is the rasping horny plate, which is slightly concave longways and across, and does not extend to the front molar by half an inch. Its surface is very rough and warty-looking, being almost entirely covered by thick, flattish, V-shaped, retroverted elevations. Some of these are rounder than others on the summit, many are acerate, and all are fringed by short sete from base to tip. The intervening palatal spaces have a less rasping surface, but are not altogether smooth. ‘The largest V-shaped papilla in the female measured 0-2 inch long and 0:1 inch in diameter at the base. In front and behind their size diminishes as they merge into the smoother anterior pad and posterior palate. The posterior third portion is equal to the preceding two in length, and is smoother; in it there is a longitudinal mesial and linear elevation, which runs backwards from opposite the anterior molar tooth. Leaving the histological consideration of these buccal appendages and mouth-arma- ture for further inquiry, I shall meantime, in the superficial relations of parts, compare what obtains in the other Sirenian genera and some neighbouring orders of mammals. According to Huxley’, Messrs. Quoy and Gaimard?’ first paid attention to the horny jaw-plates of the Dugong. Be this as it may, from their and subsequent researches it is now known that in //alicore bristles and hairs are found in the mouth almost identical in position with those above described. The bent-down symphysial portion of the mandible and palatal surface of the premaxillaries are also each covered by a coriaceous tuberculated plate ; and the tongue is bound down behind. Both Dugong and Manatee possess « series of molar teeth ; and in both, upper and lower incisors are present. But it is further to be observed that these latter bear a gradated development, inasmuch as in Manatus they are quite rudimentary, only discovered in the foetus, and never protruded, whilst in Halicore they are diminutive and functionless in the female, but two upper ones in the male form powerful tusks. With respect, therefore, to the formation of the lips, mouth-armature, and dentition, the homologous parts coexist in the above two forms. 1 Hunterian Lectures, Lancet, 1866, p. 180, 2 Voyage de l’Astrolabe, 1830, vol. i. p. 146. DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE, 167 Of the fossil genus Halitherium, besides other observers, the valuable researches of Professors Kaup and Krauss prove its being furnished with a full complement of molar teeth and tusks, as in Halicore. From the construction of the palate, intermaxillaries, and symphysial portion of mandible, I think we are justified in believing it was also provided with horny plates akin to those of the living Sirenia. This granted, a hairy and possibly a full, truncate muzzle may likewise have characterized it. On carefully studying Steller’s admirable description of the muzzle and interior of the mouth of the now extinct northern Rhytina, I have been struck with the similitude to that of the Manati examined by myself. Indeed, excepting in size, slight variation of the rasping-plates, and absence of teeth, what he says perfectly accords with the formation of parts in the latter species. It has been reserved for the very able and learned Professor Brandt, of St. Peters- burgh, to correct the otherwise accurate Steller upon an important point, viz. the structure of the palatal and mandibular lamine. ‘These Steller regarded as two osseous plates, not true teeth, but rather as it were supplying the place of these in mastication. But although their function may undoubtedly have been trituration of the food, still Brandt, after a very elaborate microscopic examination, has satisfactorily demonstrated their indurated epithelial character, quite wanting in bony or dental substance. Steller’s figures of them when removed convey but a hazy impression of what they must have appeared when in the mouth. Brandt’s illustration of the palatal plate in situ, however, enables a clearer conception and estimate of it to be made. From this, and what he himself states, the structure in question can be no other than the homo- logue of that found in Manatus and Halicore. It certainly does not appear to me to be the representative of teeth, nor of the baleen plates met with in the true Cetacea (an idea some are disposed to accept). Although Rhytina was edentulous in the adult condi- tion, I strongly suspect that, like the other Sirenian genera, rudimentary teeth may have existed in its earlier stages of growth. Nordmann seems favourably inclined to this opinion. The maxillary alveolar ridges are narrow and quite behind the bruising plate, the latter occupying the intermaxillary and not the maxillary bones. Among Cetacea the toothed and whalebone groups necessarily present differences. As exemplifying the former, Globiocephalus has no hairy bristles on the snout or within the lips—these parts superficially exhibiting a moderately smooth, tough, jet- black membrane. The alveoli are well defined, and the gum-tissues highly ridged betwixt the numerous teeth. The front V-shaped arch formed by the junction of the upper gums is callous, and evidently homologous with the front pad or inner upper lip already alluded to. The anterior third of the membrane of the hard palate is dense and fibrous beneath, and beset with irregular rows of hardened, closely placed excres- cences; posteriorly the roof of the mouth is smoother and of a lighter colour. The anterior part, therefore, in external appearance and structure, is the homologue of the VOL. VIII.—PART I. September, 1872. 2¢ 168 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. Sirenian manducatory plate. The lower masticatory plate, again, may have its homo- logue in the firm membranous portion of the symphysis, which is toothless. In the Rorqual (Physalus antiquorum) bristle-like hairs are met with on the lower lip, and there is a protuberant chin. ‘The mucous membrane of the roof of the mouth is indurated, transversely arched, and about a foot wide behind, but flattened, more callous, and only half the width in front. The baleen-plates and vasculo-fibrous root- matrix spring from outside the palate, and, though in close relation, are not a diffe- rentiated portion of it. The baleen, in fact, at its hindermost end, is little else than a matted tuft of hairs; and quite in front it shortens and resolves itself into isolated patches composed of aggregated clumps of bristles, there being mesially a distinct but small-sized pad’. We learn, moreover, from the extensive researches of MM. Eschricht and Reinhardt?, that in the foetal and very young Balena mysticetus short stiff hairs are distributed on the outer anterior surface of the upper and lower lips; and besides these there are median bald spaces, apparently corresponding to the pads of Manatus. The palate is much narrower than in Balenoptera and Megaptera; nevertheless it is strictly defined by the raised membranous fold or “ wreathband” (Kranzband). This circumstance, and that of the baleen-plates and matrix forming two long strips outside, and not merely occupying the anterior midpalatal space, militate against the baleen being the homologue of the Sirenian horny rasping-plate. The above authorities affirm that in the foetus the subsidiary whalebone-blades “consist each of a fasciculus of hairs agelutinated by the gum;” and as there is no special cortical tissue, this serves as “a sufficient proof that the hairs are the primitive formation of every baleen blade.” Hence the conclusion arises that the homologues of the baleen, in the so-called herbi- vorous Cetacea, are the long hairs and bristles found inside the mouth, and situated, like the baleen, lengthwise outside on each side of the palate. Reverting to the structure of the Ruminant mouth (for instance, that of the Sheep), the palate is smooth behind, transversely ridged (or covered by short, double, somewhat V-shaped arches*) in front of the molars to as far as the terminal well-known semilunar pad. This last, when seen in front, quite resembles the so-called inner lip of Manatus. The fringed part appears as the homologue of the bruising plate of that genus; and the posterior part corresponds in each as the smooth portion of the palate. It is true there are no hairs developed within the buccal region in the Sheep, but, instead, 1 In this species also Professor Flower (P. Z. 8S. 1869, p. 606) has found a fringe of short, stout, coarse fibres or hairs basally and outside the baleen, corresponding with which in the Manatee short hairs and bristles obtain. ? Ray Soc. 1866. Translated from the Danish Roy. Acad. Mem. 1861 and 1862. * Professor Turner (Trans. Roy. Soc. Edinb. vol. xxvi. p. 222) holds that these palatal folds in Ruminants are the equivalent of whalebone. But in the Cetacea we have palatal structures with which they seem more allied ; hence the homologue of the whalebone must be looked for external to them. DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 169 numerous conical papille, elevations of the mucous membrane, which find their counter- part in the trapezoidal ridges of the Manatee, between which the hairs and bristles sprout. Furthermore, from Eschricht and Reinhardt’s statement that the soft portion of the baleen is only an excessively developed condition of mucous membrane and epi- thelial cells, it may be presumed that the ruminant papille and the identical structure in the mouth of Sirenia are the homologous constituents of the Cetacean gum. Con- tinuing the comparison of parts, the pad behind the lower incisors in the Sheep accords with the symphysial pad in the Manatee, and the fringe of roughened skin-texture on the edge of the lower lip of the former with the broader, thicker, semilunar patch in the latter. Both have a hairy muzzle and beard; and the cleft tendency of the upper lip in Ovis recalls the wider, semilunar, truncated muzzle of Manatus. Lastly, if the oral cavity of Pachyderms be considered, homologous parts are dis- cernible. In Elephas the palatal pads are much softer and smaller in proportion than in the Lamatins, the muzzle is elongated into a prehensile trunk, the lower lip is like- wise lengthened. But long bristly hairs are largely developed within the buccal region, especially in the African Elephant. From the facts which have been particularized, I think that with some show of reason the following inferences may be drawn :— Ist. That the upper horny masticating-plate of Manatus is homologous with the roughened, warty, or retroverted papillary portion of the palate of Cetacea, Ruminantia, Pachydermata, &c. 2nd. That the horny baleen plates of Cetacea find their homologue in the Sirenia and some Pachydermata, in those developments of hairs and bristle-strips within the mouth and cheeks, existing either in bunches or as more separate filaments. 3rd. That the folds of mucous membrane within the inner upper lip of the Sirenia are represented by buccal papille in Ruminants and other forms, and that the greater development of similarly constituted mucous membrane in the interior of the mouth of some Cetaceans forms the intermediate substance of the baleen. 2. Alimentary Canal. Beyond the constrictor muscles of the pharynx to within an inch of the stomach, the cesophagus is very narrow, not large, as it is said to be in Rhytina', being no more than half an inch in diameter in the ordinary undistended condition. But its muscular walls are uncommonly thick at its lower end, agreeing in this respect with the Dugong, as mentioned by Owen’. There is an outer layer of longitudinal fleshy fibres, and beneath that the usual decussating oblique layers. For eight inches or more in the female specimen the average thickness of the walls was a little over 0:1 inch; the relative depth of the cuticular and submucous lining to the fleshy fibre lining is as one to two. Near and at the cardiac orifice, however, the muscular covering increased 1 Steller, op. cit. p. 310. ? P.Z.8. 1838, p. 30. 2c2 170 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. amazingly, forming, as Owen has justly said of the same structure in Halicore, a powerful sphincter “to defend the cardia against the pressure of the contents of the stomach.” ‘This latter organ, indeed, is almost gizzard-like as respects its strength, of which more hereafter. The cesophageal tube in the older female measured 10 inches long from the posterior border of the inferior constrictor to the gastric extremity. Its mucous membrane has quite a Cetacean aspect, there being some half a dozen longi- tudinal furrows, and as many flattened ridges throughout its course. There are besides innumerable very minute rug disposed in wavy transverse strie, giving a kind of velvety appearance to the ridged portions of the tunic when in the contracted condition. The inner epithelial lining, as Steller also has noted in Rhytina, is pale-coloured and slightly corneous in texture. I did not observe in Manatus any of those deep glandular pits or crypts in the mucous lining which I have found scattered here and there in the ceso- phagus of Physalus and Globiocephalus. Before reaching me, each Manatee had the stomach cut open and the contents removed. Consequently this very muscular organ had shrunk considerably, and on cursory inspection exhibited any thing but the enormous relative proportion assigned to this viscus in its congener the Great Northern Manatee. ‘To Steller’s surprise, four stout men with difficulty dragged out the stomach of the latter animal examined by him. The entire length of its carcass, however, was 24 feet 8 inches; and the stomach, full of fucus, measured 6 feet long by 5 feet across. But he further notes in the table of dimensions, “ Ventriculus latus seu longus potius” 44 inches. I suppose, therefore, that this diminished capacity applies to the empty and contracted stomach, and to that portion which corresponds to the so-called cardiac cavity of the Dugong and Southern Manatee. If so, the cavity in question in the Rhytina has its long diameter in propor- tion to the length of the body as 149 is to 1000, or rather more than +. In the male Dugong, 6 feet 10} inches long, dissected by Professor Owen in 1838, the first or cardiac cavity was 9 inches in greatest diameter, equivalent, therefore, to one ninth the animal’s length. Taking the same portion of the contracted viscus in the female Manatee as about four inches, this gives but one sixteenth as its proportion scale. Of the compound stomach and bifid cecum Daubenton’s, Home's, and Vrolik’s original figures are suggestive, though not quite in accordance with my inspection ; nor have they shown the interior arrangement of the parts (vide sections, Pl. X-X1II.), which are as remarkable as the exterior configuration. ‘The first cavity is the most capacious, and not unlike the human stomach in shape. Immediately to the left of the esophageal entrance, and bulging upwards to the side and partly behind the gullet, there is a pro- tuberant cul de sac lined interiorly by corrugate mucous ruge. Beneath this is the constricted orifice of the cardiac gland. From here, the narrowest part, the first stomach widens to its middle, beyond the bend narrowing by degrees to the distal end, which is guarded by a sphincter-like narrow passage communicating with the fourth digestive compartment about its upper third. The inner tunic is composed of a pale- DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 171 coloured, thickish, tough material—mouths of gastric glands being promiscuously scattered throughout its surface. A considerable median area has few folds; and this smoother space with increment of muscular walls gives the cavity a gizzard-like cha- racter. The folds of the greater end are chiefly longitudinal, and at the lesser end increase and interdigitate more. The large cardiac diverticular gland is thumb- shaped or cylindrical, and perforated in its long axis by a compressed central channel with side pockets and sacculi, which wend obliquely and irregularly upwards. With the latter there are connected short secondary recesses into which the mouths of innu- merable small flask-shaped glands open. Thus in longitudinal section this secerning apparatus has a dendritic appearance, whilst cut transversely it exhibits radii whose limbs are highly convoluted. The glands secrete abundantly a viscid creamy substance, as in the Rhytina and Dugong; but unlike them, as Steller and Owen mention, the passages contained no parasitic worms. The subequal smaller-sized tubular appendages or cornua, which may be regarded as second and third accessory gastric cavities, are situate above and on either side of the further extremity of the lessser curve of the first stomach. ‘Their parietes are only moderately thick, their internal coat chiefly thrown into longitudinal folds; and the two chambers, by thick-walled passages, end together in a pouting enlargement (com- parable to the os uteri) at the summit of the fourth gastric compartment. The latter, elongate and intestiniform, possesses a series of softish, florid, mucous plications abun- dantly glandular, and sinuously longitudinal and fureate. The pylorus is a firm ring. The duodenum has a moderate expansion and a relatively smooth inner coat for several inches; the pancreatic and common bile-ducts enter wide apart. The empty small intestines have an average diameter of rather over half an inch; and their muscular coat is uncommonly thick. In the female their length is 25 feet, and in the younger male 24 feet 4 inches. Valvule conniventes are absent; but commencing near the duodenal loop in the circumference of the gut are five or six longitudinal mucous ridges, which, with sinuous lines, ‘continue straight on as far as the ileum. Between these are short transverse interdigitations and corresponding depressions. Each short Peyer’s patch is from } to 1 inch apart; and, besides being distributed in an opposite zigzag manner as obtains in Halicore, they follow straight lines in the long furrows. In the ileum the longitudinal rugée frequently fork and pass obliquely to each other ; and the short spurs from these enclose in profusion scattered loculi and glands of Lieberkiihn. The ileo-cecal orifice is guarded by a powerful tumid muscular sphincter ; and there is a pouched ileo-colic agminate gland resembling that of the Giraffe and Hippopotamus. The cecal appendages are thick-walled, ridged, glandular within, and outwardly look like a pair of conical teats. Lengths 1; and 13 inch and 0°65 inch in diameter, and their roots 1 inch from the ileo-cxecal valve. At the commencement of the colon there is a dilatation for a couple of inches or so, with a diameter of above three and a half inches distally, where it is constricted; and then follows a second, but 172 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. narrower, expansion. ‘The first is very glandular, and with convolute ruge; the second, thinner walled, assumes interiorly the character of the rest of the great intestines. These in the female, including czecal appendage, are 17 feet 9 inches long, and in the male 18 feet—the greater muscular contraction of the former probably accounting for the difference. The rugz are very numerous, close-set, and chiefly longitudinal, and obliquely interdigitate, forming shallow elliptical depressions, among which are glan- dular patches. Halfway on the gut the ruge and glands diminish in size and number. To supply a desideratum as regards the abdominal viscera in their natural position, I have given in fig. 20 a reduced copy of a diagrammatic sketch taken from the young male animal. It represents the parts as seen when a median longitudinal section has been made from near the anus forwards to the middle of the sternum, the fleshy walls being dragged outwards. Anteriorly the heart appears to occupy the full breadth of the chest, the severed pericardium stretching across at its bifid apex. Behind is the liver, segmented into four divisions,—a very large triangular portion of the right and another equal-sized portion of the left lobe filling respectively the right and left sides of the cavity; whilst between them, in the triangle bounded by the pericardium and their anterior borders, are two much smaller lobes, the right one of which contains the rather large gall-bladder. No lungs or diaphragm are exposed, the apparent and not real absence of the latter doubtless having deceived Dr. G. A. Perkins! in his examina- tion of Wyman’s” Manatus nasutus. Mesially situated and betwixt the hinder fork of the great liver-masses, a small piece of the stomach and curved appendix are exposed. The remaining posterior half of the abdominal cavity shows only intestinal coils, and partially the urinary bladder when this viscus is distended. When, however, the thoracico-abdominal cavities with the entrails im situ are examined sidewards, a representation of which has been given in the body-section (fig. 37, Pl. XXVI.) with the ribs in place and the intervening tissues removed, a widely different view is obtained. ‘The relations of the parts mentioned (heart, liver, intestines, and bladder), to some extent, remain good. But above them is brought out in relief the enormous lung, which reaches from the first to the last rib, and extends more than midway downwards, just permitting a fringe of the elongated diaphragm to peep through below and be the barrier line betwixt the dorsal pulmonary and ventral cardo-alimentary compartments. 3. Glands concerned in Digestion. Of the secretory apparatus connected with the mouth, the most conspicuous bodies are the parotid glands. As briefly noted by Stannius, these are very large and lie at the sides of the lower jaw. They have a coarse granular texture, are broad and flat, and reach from the insertions of the cephalo-humeral and levator claviculi muscles forwards to beyond the angles of the mandible. In the vertical and horizontal direc- ? Proce. Boston Soc. Nat Hist. (1845-48) vol. ii. p. 198. * Thid. vol. iii. p. 192. ee DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 173 tions they stretch from the temporal part of the malar arch downwards to the digastric muscles and the stylo-hyoid cartilages. A portion of each gland also wraps round the stylo-hyal rod, and dips into the hollow continuous with the foramen lacerum posterius. The internal maxillary artery dips beneath them and forms a vascular network, partly enveloping their deep and superficial surfaces; while behind is the cervical rete mirabile. The narrow tendinous cord of the sterno-mastoid pierces and goes quite through the posterior half of each gland. Superficially the neck and cheek portions of the fleshy panniculus cover and completely hide the parotids as well as their Stenonian ducts. The submaxillary gland is also flattened, and of no mean size. It has a horseshoe- shape, the anterior convexity of which is lodged in the postsymphysial angle, whilst the posterior concavity reaches the transverse bar of the hyoidean arch, and partly covers the thyro-hyoid muscles. But the body of the gland chiefly lies upon the mylo- hyoidei, the two limbs of the crescent filling the deep hollows of the mandible in front of the angles. At this latter point, however, a portion of its substance is in close apposi- tion with the anterior twigs of the inframandibular plexus of vessels, with the facial artery, and with the subjacent pterygoidei. I regard as representatives of sublingual glands a series of partially separate, small, lenticular bodies, which lie near and backwards from the frenum lingue ; but I did not detect their excretory duct. My observations support Rapp’s account of the tonsils, viz. flat elliptical lamine, their numerous orifices having a sieve-like aspect. The velum pendulum palati is a broad membranous fold without appreciable uvular thickening of muscular fibres. The pale-coloured firm pancreas has the usual situation, within the duodenal loop. Its duct opens into the intestine close to the pylorus. The liver of the larger specimen had been hacked in pieces, so that nothing but its weight, 33 lbs., and apparent long diameter, fully 8 inches, could be made out satis- factorily. In the younger male this gland was more intact. Its relative position towards the neighbouring organs has already been mentioned. Vrolik’s description of the Manatee’s liver corresponds more with what I have found than does Daubenton’s: the latter having but examined a foetus may account for this. Jn situ, but still more so when removed, the entire liver has great resemblance in shape to the inflated lungs of an ordinary mammal. Thus the posterior broad surface of the main right and left lobes forms a deep arched hollow, enclosing the stomach and duodenum, and which may be compared to the dome-shaped concavity of the lungs as they rest on the diaphragm. The anterior partially segmented lobules and the gall-bladder simulate the upper pulmonary lobes overlapping the base of the heart. What I gather from Steller’s and Owen’s words concerning the mainly trifid hepatic organs of Rhytina and Halicore leads me to infer that this gland in Manatus differs little, if at all, from them. In the latter the two large somewhat triangular and much separate right and left lobes possess few emarginations. There is a short, shallow notch on the middle of the ventral 174 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. margin of each; and the left possesses a small subtriangular lobule at its posterior spinal corner. The third smaller anterior and semidivided lobe, transversely bitrian- gular or co-shaped, Steller’s anvil-formed and Owen's quadrate-figured portion, may either be regarded as the homologue of the so-called cystic lobe of some mammals, or as representing additional upper or anterior lobules of the right and left lobes, bridged together by a diminutive lobus guadratus. This last, as in the human subject, is that portion bounded dorsally by the transverse fissure, laterally by the gall-bladder on the one side, and the round ligament and short longitudinal fissure on the other. There is a compressed boot-shaped diminutive lobule immediately to the right of the inferior vena cava, and a second rather elongate, but terminally flattened, lobule attached to the left wall of the same vein. Both of these small lobules spring from the root of the main right lobe, and respectively appear to be homologous with the caudate and Spigelian lobes. Owen remarks that the small Spigelian lobulus in the Dugong is continued from the root of the left lobe. This origin, however, according to my observations in those mammals where the liver is deeply cleft, would not precisely correspond with the Spigelian lobule, which arises from the right moiety, and is separated from the left lobe by the ductus venosus. Notwithstanding, it does not militate against the Professor's clear definition that “the homologue of the ‘ Spigelian lobule’ is shown by its relation to the lesser curvature of the stomach ”?. All the hepatic fissures are shallow. The most marked ones, the longitudinal and that of the ductus venosus, being filled up by strong fibrous tissue, covering the vessels therein. As to the ligaments, the suspensorium hepatis is moderately broad, and firmly fixes the organ to the pericardium and the diaphragm. ‘The round ligament, as usual, forms the anterior or ventral one positionally. In the young male it was a narrow cord, nearly impervious, 1 inch from the liver. The two lateral ligaments diverge from the vena cava, and traverse lengthwise the right and left lobe about an inch outside their vertebral margins. The pyriform but forwardly projecting gall-bladder lies superficially on the ventral aspect of the small anterior right lobule. When distended it is 2} inches long and 1 inch in diameter at the fundus. The cystic duct, of considerable calibre, winds in an S-shaped manner, and at about three quarters of an inch distance from the neck of the gall-bladder receives singly the united hepatic duct on its left wall, as Daubenton and Vrolik have recorded. In the Dugong the cervix of the gall-bladder is said to be obliquely pierced by two hepato-cystic ducts, entering, as the ureters do, into the urinary bladder. The ductus communis choledochus in the male Manatee was as thick as the barrel of a goose-quill, and penetrates the intestine about three inches from the pylorus. 1 «The Anatomy of Vertebrates’ (1868), vol. iii. p. 483. See also Owen “ On the Anatomy of the Cheetah,” Trans. Zool. Soc. vol. i. p. 131. DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 175 On section the liver exhibits a fine glandular structure, and not a coarse lobular substance as in some Ruminants. ‘The interior of the gall-bladder is smooth. VI. OrGANS OF CIRCULATION. 1. The Heart. The following tabular arrangement expresses in inches and tenths the several dimen- sions of the heart of the female Manatee when contracted as ordinarily after death, in this case, however, having been very slightly preserved in spirit :— Right auricle. Left auricle. Pixtremie lCUP thoy ete tents tok is oe segs onl st 1:6 ixtreme, Dread tlie wyukicy Slaten wesw aS aes ee 15 Right ventricle. Left ventricle. Extreme length, anteriorly . . . . .. . 2:4 2°8 Extreme length, posteriorly. . . ... . 20 2°8 Extreme breadth at base, anteriorly . . . . 2°7 2°3 Extreme breadth at base, posteriorly. . . . 2-4 2°5 Extreme breadth af apex 3 2-5 s «2, Ll 15 Extreme thickness of walls . . . . . . . 0 15 This organ has been so repeatedly described that I can add nothing material to the statements of previous observers, and agree with them as to its cleft nature. 2. Blood-vessels and Lymphatic Glands. The arterial distribution of the American Manatee has been so lucidly explained in the masterly compendium on the vascular system by Professor Hermann Stannius', that, were it illustrated, I should be content to leave the subject untouched. But the remarkable character of the vessels, splitting up, as they do, in certain parts into multi- farious plexuses and rete mirabile, is a sufficient reason why they should be figured and further commented on. According to the above author, and as my dissection (fig. 30, Pl. XXIV.) demon- strates, a short, wide, innominate trunk springs from the arch of the aorta, and divides into a right subclavian and common carotid. From the summit of the arch the left common carotid is derived ; and, lastly, further beyond is the left subclavian. As regards the carotid and branches, he states only that the common carotids have tolerably long stems, which, at first glance, appear to simulate division into an external and internal carotid, although this does not truly take place. The first main branch of the carotid proceeds inwards, and supplies the larynx, the hyoid apparatus, and the tongue. What I have found obtain is as follows:—The stem of the fair-sized common carotid is branchless until about opposite the deep hollow in front of the shoulder or supra- 1 Op. cit, p. 31 et seg., also Yon Baer, Mém. d. l’Acad. St. Pétersb. 1835, tom. ii. p. 199. VOL. VIII.—PaRT 111. September, 1872. 2D 176 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. scapular region, where an artery (transverse humeral) of small calibre strikes directly outwards, and terminally subdivides into a broad radiate rete mirabile, covering the subscapularis, supraspinatus, &c. The single arterial tube is superficial to the cervical nerves, but is itself enwrapped by portion of the vascular neck-plexus. Beyond and on a level with the cricoid is a division apparently equivalent to internal and external carotids. The former dips in among the rete mirabile at the posterior base of the skull just behind the cranial series of plexuses. Among others, one occupies the poste- rior portion of the great fissure between the occipital and tympano-periotic bones ; and whilst mingling with the cervical and spinal rete, complex branches are lodged within the skull at what corresponds to a groove or recess of the lateral sinus, where also venous channels obtain. The external carotid at the stylo-hyal and under cover of the digastric and parotid gland bifurcates; and plexuses are derived from both of these. The branch agreeing with the facial runs towards the angle of the mandible and at the concavity of the body of the bone turns upwards and is distributed with a plexiform arrangement on the face. From its proximal end, and in fact enwrapping it, are retia, which may be regarded as submaxillary, submental, &c. subdivisions; and these lie within the concavity of the jaw, twigs supplying the muscles and other parts, while some inosculate with their fellows of the opposite side. The other, widest branch of the external carotid ascends behind the mandibular angle, previously supplying plexi to the parotid gland, superficies of the digastric, &c. Other plexi, which were not followed in detail, spread over the tympanic, temporal, and malar areas. At the ptery- goid region there is separation of the external carotid into several thick retial bundles, whereof the inferior dental, lingual, and internal maxillary are most conspicuous by their volume. The vascular network, as it pierces the large vacuity of the lower jaw, has a remarkably open character, resembling a meshwork of fibrous tissue ; and, as in Cetacea, the interstices are partially occupied by fatty tissue and nerves. This vasculo-nervous mass issues at the mental foramen, and supplies the lower labial parts. ‘The numerous capillaries of the internal maxillary division pass on to the pterygo-maxillary fissure, and send inwards superior dental arterioles ; whilst the main mass, lying in the lateral groove of the maxilla, is continued on through the orbit and emerges at the infraorbital foramen, spreading amongst the fleshy and other structures of the face and snout. Ciliary branchlets from the above are given off to the eye &c. Stannius says that the subclavian artery divides into two main branches—a de- scending large internal mammary, and the axillary. Before these divide, a very short twig is sent upwards over the head of the first rib. Both arteries are equally split up into narrow channels, which intertwine among the rete mirabile of the cervical region and thorax. Von Baer likewise shows that the axillary artery becomes broken up into minute vessels which overlie and are partly distributed to the shoulder and partly continued on to the anterior extremity, mingling like the last with the rete cervicale. This is certainly the case, and the complexity and quantity of parallel channels is es ee a DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 177 truly astonishing. Besides an axillary, however, I observed another trunk derivative of the subclavian, and which may be representative of a thyroid axis, terminating in axillary rete. In tracing continuations of the axillary rete I could distinguish circumflex anterior and posterior bundles. The brachial rete diminishes as it reaches the elbow and passing beneath the pronator radii teres, forms an ulnar rete. This keeps close to the bone, under the forearm-muscles, and crosses obliquely to the proximal end of the fifth digit. Here, besides muscular radicles, a bunch goes to the palmar aspect of the wrist, and by intricate partition helps to constitute the palmar arch. ‘The radial rete appears less complicate, but unfortunately its manner of palmar division was not satis- factorily made out. The multitudinous networks, nuchal, spinal, thoracic, and caudal, are sufficiently like those of whales, and have been so frequently referred to by other authors, that I need not dwell on them. As regards the intercostal plexuses, these do differ from those of Cetacea, inasmuch as, instead of great contorted coils lying superficial to the ribs, each intercostal artery dips singly between the ribs, and in the space covered by the pleura and muscles divides arborescently. With all due respect therefore to the accuracy and acumen of our great leader, Professor Owen, I venture to predict the presence of similar rete in the Dugong, where such arterial modification is denied. When unin- jected the closely-packed vessels so simulate coarse muscular fibre as readily to deceive one unless critically inspected. There is an abnormal rete, hitherto unrecorded, in continuance of the internal mam- mary. ‘The latter vessel issues beneath the rectus abdominis, opposite the third costal interspace, and proceeds upon the transversalis as a rete, which posteriorly anastomoses with a returning epigastric series. Anteriorly the rami run outwards in parallel radii like the plume of a pen. The abdominal surface of the posterior half of the diaphragm is supplied with vessels and nerves arranged in a like fashion. I may advert to a retal offshoot of the deep lumbo-caudal mass of vessels, which, for distinction’s sake, may be named either sacral, pelvic, or hypogastric rete. Derived from where the costal channels strike within the chevron bones, it forms a thin but wide sheet of arterial and venous rami, which cover the sacro-coccygeus and parts beneath the urino-genital organs. From it vesical and other supplies are given off, the most characteristic being a hypo- gastric, obliterated beyond the fundus of the bladder, a uterine, and in the male sper- matic plexus, with marked arteria dorsalis penis. Circumstances already mentioned prevented my observance of the giving off of the abdominal aortic trunks or following the visceral distribution. ‘The mesenteric vessels appeared to split into primary, secondary, and tertiary arches without any striking peculiarity in the yasia brevia, and the hemorrhoidal arterioles have inosculations with the hypogastric rete. The veins of the face and head were not followed in detail. Branches, however, were observed to return from the submaxillary region and outer side of the jaw; these con- 2D2 178 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. verge below the parotid gland and join the external jugular opposite the paramastoid. The brachio-cephalic vein comes inwards from above the insertion of the pectoralis minor; and another marked tributary is derived from the vascular plexus covering the inner subscapular region. The course of the external jugular vein is from behind the cranial end of the stylo- hyal and the thereupon attached portion of the digastric muscle, backwards and slightly obliquely outwards, uniting with the internal jugular near the first rib. The external almost equals the internal jugular vein in calibre; and it lies over the tendon of the sterno-mastoid muscle and the suprascapular artery. The internal jugular vein commences at the cranial aperture, foramen lacerum posterius, near to the attachment of the rectus lateralis and cephalo-humeral muscles, where there is a large venous plexus as in Cetacea. As it traverses the neck backwards it lies chiefly to the inner side of the carotid artery, crossing it, however, about the level of the bifurcation of the trachea, posterior to which it converges to the large innominate trunk formed by it, the external jugular, and the subclavian vein. There are several oblique bridging communications between the ecto- and entojugular veins. No valves were observed in the above veins of the neck. The deep cervical glands are very voluminous, and fill the intervening cleft, bounded anteriorly by the cephalo-humeral muscle, externally by the subscapularis, deeply or dorsally by the short neck-muscles, the lateralis and obliquus externus, and within or mesially by the extension forwards of the serratus magnus. The cervical plexus of nerves passes over the glands in question, the latter being enveloped amidst the rete mirabile. VII. Vocan AnD Respiratory APPARATUS. 1. The Air-passages. Stannius has been successful in his exposition of the structure of the larynx. I may refer, however, to two points he and Rapp have failed to notice, viz. the existence of a small recess or pseudo-sacculus laryngis at the anterior extremity of the vocal cord, as in the Dugong; and to the presence of small nodular cornicula laryngis or cartilages of Santorini, surmounting the arytenoid bodies. As these authors and Vrolik state, the epiglottis, unlike the Cetacean, is of the most rudimentary character; the thyroid car- tilages are united anteriorly by a narrow bridge, and anterior and posterior cornua are well developed; the cricoid is a complete ring posteriorly, very broad, and with a marked prominence for the attachment of the thyroidal posterior cornu; the arytenoids are trihedral; the vocal cords are the reverse of prominent, and deficient in inferior excavation. The several ligaments are composed of tough yellow and fibro-elastic tissues; and the diminutive epiglottis, curiously enough, consists of like material, void of a cartilage basis. The dozen tracheal rings, and, as far as I could make out, the bronchial also, are not DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 179 continuously spiral, as obtains in Halicore, but, as Stannius figures in Manatus, there are some which bifurcate and obliquely cross the long axis of the tube. In my female specimen the trachea, # of an inch in diameter, split at about 5 inches distance from the lungs, and each bronchus entered almost at the summit of the pulmonary organ. Within the lung it goes in a straight line to the posterior extremity, lying a little to the inner side of the middle. About a dozen bronchia branch outwards, these again subdividing in the pulmonary substance. There is a narrow and moderate-sized thyroid gland on each side of the upper portion of the trachea. The lungs, their shape and singular relation to the diaphragm, &c. have often been commented on since Daubenton’s original description. My illustrations of the parts in their natural position, figs. 20 and 37, supplant verbal detail. Some two or three inden- tations, } to 1 inch deep, are the only trace of segmentation; but anteriorly they ter- minate in a short rounded lobule (J, fig. 41). In the uninflated state the greatest thick- ness of the lung-substance of the female was 1 inch; extreme length 23 inches; breadth towards the anterior extremity 2? inches, about the middle 3% inches, and rearwards 13 inch, tapering finally to an obtuse termination sunk in a pocket at the lumbo-vertebral end of the diaphragm. 2. Hyoid and the surrounding pharyngo-glossal fleshy parts. The hyoidean arch comprises three bony pieces—to wit, a small, flat, oval basihyal, and a pair of long subcompressed stylohyals. Each of the latter measured 1:7 in the young male, and 2°2 inches in the older female. To the upper narrow extremity of the stylohyal a strip of cartilage an inch long is fixed, by which it is fastened to the inferior tubercle of the exoccipital. Betwixt the other (broader) end of the bone and the basihyal is a >-piece of cartilage representative of ceratohyal. This extends con- tinuously along the outer border of the basihyal, and forms a retrocurrent wing to it on either side; and to these the anterior cornua of the thyroid ale are attached. The thyrohyals or connecting ligaments between the hyoid and larynx are tough thickish membranes, and apparently contain a considerable amount of yellow elastic tissue. I examined the intrinsic muscles of the larynx carefully, and found that, notwith- standing the rudimentary nature of the epiglottis and comparative absence of laryngeal pouch, I could differentiate superior and inferior aryteno-epiglottidei, and even noted fibres equivalent to a thyro-epiglottideus. Indeed, each and all of the laryngeal muscles are relatively well developed. The extrinsic laryngeal muscles maintain a fair size, with attachments of the ordinary kind. The keratic muscles, so notably developed in Cetacea, are feebly represented in Manatus; and the hyoepiglottidei of the former are entirely wanting in the latter, as might have been expected from the condition of the epiglottis. The sterno-hyoid and sterno-thyroid are interblended. An omo-hyoid was not indis- 180 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. putably traced from origin to insertion; the anterior remnant remained, its posterior fibres being lost among the tangled vessels and deep cervical fascia. Stylo-hyoid and stylo-pharyngeus are somewhat adherent, the latter broad and well developed. The digastric, single-bellied and broad anteriorly, fills the hollow at the inflection of the mandible, and thence passes rearwards to the junction of the stylo-hyal with its cranial cartilage. The horseshoe-shaped submaxillary gland abuts on its inner, and the parotid on its outer margin. The stylo-glossus is large, and the constrictores superior and medius are full and fleshy. The thin sheet of fibres of the levator palati cover the Eustachian enlargement, and are spread out and lost in the posterior palate. The tensor palati, better marked, arises near the tympanic bulla, passes round the pterygoid process, and, by a strong flat tendon, widens out on the posterior palatal membrane. The pterygoidei and plexuses lie outside. ‘The palato-glossus is moderately broad, and the palato- pharyngeus fairly developed. Mylo-hyoidei, as a thin fleshy plane, stretch and fill the angle betwixt the ramal bodies. Long and thin genio-hyoidei pass from the basihyal to the concavity of the chin, a vascular plexus existing beneath. A distinct hyo-glossus was not observed ; but genio-hyo-glossi and lingualis are both well represented. VIII. Tue Nervous System. When treating of the interior of the skull I dwelt upon the dura mater as it- lines the bones and the foramina piercing it basally. Its upper surface, when the calvarium is removed, is tolerably smooth, a superior longitudinal sinus being but faintly indicated. On each side the membrane is tucked into a deep Sylvian sulcus, which traverses well across the cerebrum; another, marked but shallow, depression is manifest about the centre of the posterior cerebral division. The dura mater is tough, strong, and rough at the vascular sinuses and plexuses. While it is intact, the cerebrum leaves the bulging cerebellum uncovered to a considerable extent. The encephalon of the younger male was so destroyed as to be unfit for examination. While the membranes surrounded the brain of the female specimen a tolerably accurate idea of the cerebral contour was got; but on raising the dura mater the brain itself was found to be softened, and with difficulty extracted. No measurements or weight were taken, but the whole placed in spirit as rapidly as possible. A cast of the cranial cavity with its enclosed dura mater was subsequently made; and by the help of this cast and the shrunken brain the sketches (Pl. XXV.) were drawn. I may remark, en passant, that the views (figs. 31 & 33) of the upper and under surface of the brain slightly exaggerate the relative breadth of the anterior to the posterior lobes, by the former not being approximate enough at the longitudinal fissure. With regard to dimensions, the drawings are given as nearly as possible of the size of nature ; the cranial interior, its model, and the preserved brain respectively yielding the scale of relations. From above and below, the outline of both cerebral hemispheres is somewhat qua- DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. 181 drangular, but rounded at the corners; and the crescentic cerebellar posterior margin lengthens this behind. The breadth to the cerebral length is absolutely great, but less than in the Delphinide. In profile view the height of the cerebrum is nearly equal to the length, and the figure, as a whole, remarkably Elephantine. The posterior cerebral lobes cover but half of the cerebellum, thus leaving a considerable portion of the latter free at the posterior end of the superior longitudinal fissure. Four lobes may be distinguished, viz. frontal, parietal, occipital, and temporal. The frontal lobe (F) is remarkably deep and perpendicular in direction, but of considerably less diameter antero-posteriorly. Its orbital division fills the anterior fossa of the cranial cavity; and its frontal part abuts against its anterior wall, the large olfactory bulb being situated towards the lower end of the latter. The parietal lobe (P), as defined by cerebral anatomists, may be said in Manatus to consist of two parts—to wit, that anterior and that posterior to the deep transverse Sylvian fissure. The former is a broad coronal band; the latter is a markedly three-sided area, occupying chiefly the vertex, but also partially the lateral surface of the brain, and appears to represent the angular lobule of some authors. The occipital lobe (O) forms chiefly the rounded broad knuckle of the hinder surface of the cerebrum, and is scooped out considerably below and mesially for the reception of the cerebellum. The temporal (7), like the frontal lobe, is very deep, but of smaller antero-posterior diameter than it, and not quite so perpendicular in its long axis. In fact, it forms a thick, somewhat conical, mass which lies obliquely down- wards and forwards, and occupies the wide sunken anterior area of the posterior fossa of the skull’s basis. The so-called central or median lobe, said to be of good size in the Elephant, I could not differentiate in this Manatee, in consequence of the unsound condition of the cerebral substance. Its existence, however, I cannot question. The cerebral mass, as a whole, is fair-sized, full, and with very convex surfaces in all directions. The hemispheres are divided by a deep, widish, great longitudinal fissure, and each, moreover, possesses a most trenchant division into anterior and posterior half by the Sylvian fissure. As regards parts at the base of the encephalon, the pituitary body, when first examined, appeared relatively large, was very vascular, flattish, and of a trefoil figure. It occupied a greater area lengthwise and across than pi, figs. 33, 34, represent. The transverse lozenge-shaped interpeduncular space, bounded by the optic tracts and crura cerebri, is fair-sized. What appears as corpora albicantia, and possibly tubera cinerea, are two antero-posterior, moderately large, oval eminences, situated in the middle of the space, and behind and at the sides of these, respectively, distinct posterior and anterior perforated spaces. Each crus (cr) is long, full, and prominent, and the two diverge rather obliquely, not separating entirely till a short distance from the pons. The pons Varolii (pv) is rather flat-surfaced, as, to a less degree, are the cerebral 182 DR. J. MURIE ON THE FORM AND STRUCTURE OF THE MANATEE. peduncles; it is wide and crescentic, but moderately narrow from before backwards, and the posterior border is slightly concave. The medulla oblongata is not remarkable for size ; at least, like the pons, it is not prominent superficially at the anterior pyramids (ap); and these have a very shallow longitudinal median depression. The olivary and restiform bodies are individually well represented, though imperfectly defined in our drawing, on account of the membrane having partially been left attached to the nerve-roots. Of the cranial nerves I may remark that the olfactory root comes into view at the basal end of the Sylvian fissure as a great, broad flattened tract. This narrows forwards and then expands into a large pyriform bulb (1), which curves upwards and protrudes, as an adpressed mass, against the antero-inferior surface of the frontal lobe. The optic tracts (2), of moderate calibre, approach each other nearly transversely from the inner borders of the so-called central lobes, and form a short, narrow commissure. The third nerves (3) have a usual situation from the crural junction close to the pons. The trochlear (fourth) nerve (4), a particularly fine filament, was but partially traced as it wound round the right peduncle. Relatively and absolutely the fifth (5) nerve is of enor- mous size, and, as it leaves the side of the pons, appears composed of a great number of funiculi; but among these I did not discriminate its sensory and motor roots. The flattened nerve passes sharply outwards and pierces the dura mater at the recess beneath the posterior margin of the alisphenoid transverse band, where the Casserian ganglion is lodged. ‘The sixth nerve (6) appeared large, as certainly was the facial motor branch of the seventh pair (7). The glossopharyngeal and pneumogastric branches of Willis’s eighth pair of nerves (8), as also the hypoglossal or ninth nerve (9), issued as numerous filamentous cords from the side of the medulla oblongata, and, unless from their more anterior situation, undistinguishable from the spinal series. The different direction and deeper situation of the spinal accessory branch (8*) of the eighth enabled it to be distinguished with ease. The interior of the brain I examined in two sections, viz. horizontally and vertically. When the lateral ventricle is exposed, as in the left hemisphere (fig. 35), it is seen to be a large chamber, and altogether spacious. The anterior cornu (ac), a sweeping semi- circle, is both deep and wide. The portion of the corpus striatum (cs) is a considerable ovoid and prominent mass, leaving, however, a large cornual space in front. At its posterior border, between it and the thalamus (th), the teenia semicircularis (¢s), a narrow linear strip, crosses it obliquely outwards and backwards. The middle descending cornu (dc) is a moderate-sized cavity, with a more than usual yertical curve in the usual directions. The eminence of the hippocampus major is very proisinent, convex, and well defined; and continuing on to almost the tip of the tem- poral lobe, it forms a not very distinctly notched pes hippocampi. P.Z.8. 1865, p. 207. * See also a remark thereon by Professor Turner in Balenoptera sibbaldii (1. c. p. 203). ° L.c. pl. 1. figs. 2(¢) and 3(@). 240 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. of the cranium and its superincumbent nodosity. The under surface of the chin deepens towards the throat; and the contour from this point ventrally towards the tail is a nearly uniform, narrowing, sweeping line. The dorsal antero-posterior curve almost corresponds to the abdominal one, excepting where interrupted by the back-fin. To- wards the caudal extremity the perpendicular depth is but moderate. Opposite the dorsal fin the body insensibly narrows, which narrowing steadily increases the nearer it approaches the root of the caudal fin, and becomes laterally compressed. The subjoined measurements of our female give the relative total length in a straight line, and curvilinearly following the dorsal arch, in distances between various points. From these data it follows that the amount of arching of the back and projecting mass of the snout is longer by about one foot and a half than is the body taken in a hori- zontal parallel straight line. ff. in. Length from the snout to the fork of the tail. . . . . . - . . 1010 Distance from the projecting upper lip to the blow-hole, following curve 1 11 From the blow-hole to the anterior root of the dorsal fin . 2 OF Dorsal fin in length at its base : ae 1 94 Distance from the posterior root of dorsal fin | to tail’ s root 4 9 Tail from root to fork 1 0 The dorsal fin is a prominent feature in the profile outline of the animal. It is large. falcate-shaped, and very much laterally compressed near and at its upper free margin. The anterior upper smooth border forms the segment of a wide arch whose summit is perpendicular to the posterior attached root of the fin. Behind, it terminates in an almost hook-like manner. ‘The posterior inferior border is deeply emarginate. This fin is situated in front of the middle of the body. Measured from the centre of the fin itself, the distance between it and the front of the truncated snout is six inches less than from the same point backwards to the root of the tail, while it is eighteen inches less when the measurement is carried to the fork of the tail-fin. The dimensions of the dorsal fin are:—greatest length or that continuous with the upper curved border to the hooked tip, 2 feet 14 inch; Jength of the base in a straight line, 1 foot 94 inches; greatest vertical height, 85 inches. In relation to the body, the low-set position, and the peculiarly narrow, tapering, scythe-shape of the pectoral fins, in ccntrast with the blunt globose snout, render the Pilot Whale at once as remarkable as it is characteristic among the Cetacea. Each pectoral extremity or fin is thin and knife-like on the edges. The length, taken in a straight line, is 1 foot 10$ inches; but the measurement following the curve of the anterior border equals 2 feet 5} inches, whereas the posterior border in its curve is only 1 foot 64 inches. Their position is about the posterior third of the distance between the snout and the anterior end of the dorsal fin. In shape the tail seems uncommonly like that of the Porpoise. It nevertheless DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 241 4 differs, inasmuch as the terminal free borders of the flukes are straighter or set more at a right angle to the long diameter of the spine, as Couch has observed. The tips of the flukes are also relatively more pointed than in the Porpoise. The median incision or fork is but of moderate depth; and in this specimen no overlapping of the adjoining borders occurred, as is occasionally found in Phocena communis. The measurement of the tail gave the subjoined results :— Distance from tip to tip of the flukes The transverse diameter at middle of flange . The transverse diameter at the root Length from the vertebral root to the fork Length of the outer border of each fluke . Length of the inner or posterior border of the same HEHE eH OFRNe? woaACcCw aw If the accompanying illustrations be compared with those of Traill', the Cuviers?, Scoresby *, Bell‘, Jardine *, and Couch®, it will be seen that they do not quite corre- spond to one or other in outline or proportions. As the latter authority finds fault with the figures extant, and professes to correct previous delineations, I may be allowed to point out wherein his own as well as others differ from those views now given. To my mind, indeed, Mr. Couch’s representation does not nearly so well convey an idea of the curious club-like, yet harmonious, symmetrical appearance of the Pilot Whale’s body as do the very authors whom he quotes as having misrepresented it, In his spe- cimen the dorsal fin is far too rounded at its posterior extremity, the pectoral fin too thick at its proximal half, and the caudal extremity of the body at the setting-on of the tail has too great vertical depth in proportion to the dimensions of the figure. What he remarks of Scoresby’s and Bell’s flexion of the tail forwards may be just, though not necessarily so. In but one point does he decidedly agree with my obser- vations, namely the shape of the tail. A pardonable error often fallen into in Cetacean illustrations is too great thickness of the body relative to its length. This happens on close inspection of the animal, which deceives as respects its yastness—considering that, as a whole, there is a certain sym- metrical graceful proportion, 7. ¢, thickness and breadth which decrease in an equiva- lent ratio rearwards. Such is my experience of the coup dil in various Whale genera I have seen in the flesh. In this respect Professor Traill’s figure fails; moreover the dorsal fin is too much rounded, the pectorals proportionally in advance of their true position, and the eye rather high. Of Baron Cuvier’s side view of a male Globiceps, copied by his brother Frederick, I may note that the nasal prominence is less abrupt, and the distance betwixt this and the ' Loc, cit. * Annales, as already quoted; also De I’Hist. Nat. des Cétacés (Paris, 1836), pl. 13. fig. 2. * Op. cit. pl. 13. fig. 1. ‘ British Quadrupeds, p. 483. Naturalist’s Library, loc. cit. p. 212, * Ann. Nat, Hist. vol. ix. (1842), pl. 6. 242 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. dorsal fin greater than in the present fig. 1. Moreover a faint depression is given to the occipital region, whereas in mine there is a gradual elevation at the same point. ‘The depth of the root of the tail is exaggerated; and the pectorals spring quite from the abdomen. If his belly view be contrasted with the present dorsal one (fig. 2), it will be evident that the thoracic region is flatter than I have found it, the caudal root thicker, the terminal emargination wider, and the flukes antero-posteriorly narrower. III. Parts RELATED TO THE SENSES. 1. The Eye and its surroundings.—On first beholding a large Whale, among other things which impress strangeness of aspect to this marine mammal is the seemingly small organ of vision. The narrow elliptical aperture of which and dull-coloured eyeball (at least when the animal is dead and shored) give an odd and sly expression, contrasting unfavourably with those lustrous orbs of most Seals, or even with the wide optics, though lurid hue, of many great fish. John Hunter’, always philosophical in his similes, supposed that their locomotion is not great on this account, and considered them as sea soarers compared with birds. Situated a little higher than the angle of the mouth, from which its anterior canthus was 4 inches distant, the eye of Globiceps, as in all other Cetacea, appears extremely diminutive. The palpebral fissure, a narrow ellipse, and without eyelashes, has an extreme length of 1} inch. The two eyes are 293 inches apart, as measured following the arch of the head. The horizontal dark-coloured pupil is ellipsoidal, and $ of an inch in antero-posterior diameter. When the tegument is removed, an external muscular sphincter is brought into view. This representative of orbicularis palpebrarum is only moderately developed, the fleshy fibres being intermingled with fatty tissue. Rapp® denies the existence of upper fibres in the Porpoise; but my observations both in small and large genera coincide with Stannius*® and Carte and Macalister as to their oval figure round the orbit. In the Globiocephalus killed near Lewchew, the eye is mentioned as having a “sclerotic nearly osseous; iris dark, but not red or orange”*. In one of the specimens stranded at Dundrum Bay, Gulliver says that “around the eyeball was a firm bony plate in the sclerotic coat”*. In the present specimen the sclerotic certainly presented a dense fibrous texture, where thickest simulating cartilage in appearance, as in other Cete, but no true bone obtained. * 2. The Nasal Passages: Homology of the Sacs and adjunct Fleshy Structures.—Dr. Francis Sibson® has commented “on the Blow-hole of the Porpoise ;” and (excepting, it may be)in the number of muscular layers) my observations on that animal corroborate * « Observations on the Structure and (Hconomy of Whales,” Phil. Trans, 1785, vol. xvi. p. 335. * Die Cetaceen zool.-anat. dargestellt, 1837, p. 92. * “ Beschreibung der Muskeln des Tiimmlers,” Miill. Archiv f. Anat. 1849, p. 11. * Chinese Repository, U. ¢. p. 411. 5 ..¢.-p. 66: ® Philos. Trans. 1848, p. 117. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE, 245 his with regard to the structures. He looks upon the whole only in the light of a mechanical apparatus, and shows that the adjustments are such that “ when the outer passage is closed, the posterior pouches can be distended and the anterior are emptied ; and when the passage is open, the anterior pouches can be distended and the posterior are emptied.” He believes the pouches serve to buoy up the head, and considers the spouting of the Whale to be sea-water regurgitated from the stomach. Stannius! alludes to only a single muscle connected with the nasal sacs and spout-hole ; this he calls the nasalis. He apparently disregarded the separation into layers of the muscular constituents surrounding the outer naris and its pouches. He describes, more- over, a musculus cutaneus maxilla superioris—to wit, the long premaxillary muscle covered by the blubber (NJ of my plates); but as this is posteriorly interwoven with the buccinator, he includes it under the cheek-muscles. With the exception of Hunter's lucid physiological deductions as to the use of the nasal sacs, Von Baer’s paper’ is by far the most philosophical description and piece of reasoning treating of this curious Cetacean apparatus. In it he notices the existence of a small single bone near the intermaxillaries, which Camper believed to be a process of the ethmoid, though other anatomists have overlooked it. Von Baer regards this little bone as a rudimentary inferior turbinate. There are cartilages present which he considers represent the blade of the ethmoid and not the vomer. He goes on to say that Cuvier only takes notice of two nasal sacs, Camper and Ray thought there were three, while Blainville speaks of two pairs. He then gives a description of the sacs as they in reality exist in the Porpoise, and of the external layers of muscles. He states these may be divided into as many as six portions, which, however, he is inclined to regard as but one entire layer infolded upon itself. He pointedly observes, moreover, that they have a general resemblance to the nasal muscles of Man. There is, he thinks, no special sphincter to close the blow-hole, as the lips are thick and close together by their own elasticity. According to Blainville the sacs have three pairs of muscles acting upon them. Von Baer criticises Cuvier and others who say the muscles press upon the spout-hole to eject water therefrom, whereas he reasons upon his examination and known data to show that view is not tenable. From smell and inspiration of air having but a single passage, Von Baer believes the former is subdued. He further discusses the number and position of the turbinal bones in various orders of animals, and shows that the anterior one in them has been taken for the upper one in Man, whereas he avers this is not the case. Gurlt, as he mentions, calls it the middle turbinal bone. Von Baer himself admits, however, that the turbinals are not specially organized parts—in other terms, are irregular in construction. He endeavours to show in different genera of Cete that the blow-hole influences the formation of the * L.c.p. 4; and Rapp, p. 106. 2 L. ¢. p. 338. 3 «Der Nase der Cetaceen,” Isis, 1826, p. 811. VOL, VIII.—PART Iv. February. 1873. bo 4 244 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. skull. The main facts, as he well exemplifies, are, that the turbinals alter from the horizontal to the vertical position, and occasionally in some animals are more than three in number, Moreover he upholds that the upper corrugated nasal sacs of the Whales are the homologues of the anterior or maxillary turbinate bones, in them a fibrous and not hard sclerous skeleton; and, again, the posterior sacs may possibly represent the inferior or ethmoidal turbinates. The outer opening he takes to be equivalent to the narial orifices of other animals with muscles attached; the cartilages in the Whales being absent. Olfactory nerves obtain, but very diminished in size, indeed hair-like; but he is dubious as to their capacity in influencing smell, which he believes to be very much modified in the Cetaceans. It seems to me that Sibson loses sight of the significance and homology of the parts in their ulterior specialization as a cranial floating organ—this, in my opinion, being as absurd a proposition as that, from the direction of the first stomach of the Rorqual, water swallowed in plenty (?) is thrown up therefrom. The otherwise correct Stannius must have but glanced at the fleshy narial layers. With Von Baer I agree, not only in surmising, but in recognizing individual nasal muscles, homologues of the occasionally diminutive series connected with the nose and upper lips in higher Mammals. I have little doubt also that odour, in a modified condition from land-breathers, is appreciable to the marine Cete, which notion Hunter had already promulgated, - I am inclined to think the skull’s development quite as much influences the position of the blow-hole than the reverse; but I differ most decidedly from Von Baer in not recognizing in the nasal sacs transformed turbinate bones. In this female Globiceps the outer orifice of the blow-hole, or naso-respiratory opening (figs. 3 & 2), is situate upon the summit of the forehead, vertically rather behind the eye, and, as in Delphinus, a trifle to the side of the median line. It is distant rear- wards 23 inches from the front of the labial prominence, and its outer angles each a foot from the eye. In the ordinary condition of the parts the lips of the aperture are closely approximated, leaving only a wide V-shaped transverse shallow sulcus, with slightly sinuous edges. This slit measures barely 2 inches from corner to corner, and its mid angle is directed backwards. When looked into by pressing forwards with the finger the front margin (as fig. 27, of Lagenorhynchus depicts), part of the anterior and lateral walls are seen to be thrown into innumerable fine striz or black-pigmented cuticular wriggly but parallel rugee, which radiate centrally and forwards, These, the loose surrounding membrane, and deep fatty tissues give great elasticity to the parts, which, furthermore, from two obliquely transverse cushions or backwardly resilient and smooth fibro-cartilaginous oblong bodies, naturally close the orifice while the muscles are relaxed. At first intent I had described the spiracular cavity, and its sacs in detail, of our specimen. But as in my papers on Risso’s Grampus and the White-beaked Bottlenose, 4 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 245 a careful survey of these structures has been instituted, it is unnecessary to pursue the subject here. I may add, though, that the sacs of Globiceps are three on either side, the so-called facial division of the naso-frontal being absent. All three genera, therefore, differ from Phocena in division of the anterior or slipper-shaped premaxillary cavity. The next structures whereby homology of parts can be traced are the cartilages. These, compared with the massive head, I may say, are very small, much interwoven with the fibrous and fatty tissues bordering the narial orifice, but nevertheless traceable. Shooting forwards from in front of and between the nasals as a narrow, short and diminu- tive wedge, is a little stump of cartilage, which I take to be the prevomerine ethmoidal or septal cartilage in a very rudimentary condition’. On each side of the above, partly continuous and partly connected by fibrous material, is another, irregular-shaped but somewhat curvilinear and small fibro-cartilaginous isthmus, but which I could not satisfactorily follow out. From position they would accord with the upper lateral cartilages of land Mammalia. Lastly, forward from these and only connected by fibroid tissues (I speak only as far as my dissection permitted me to observe) are the two much larger and transversely oblique masses, visible in the upper opening of the nares, and which are composed of fibro-cartilage fatty substance and mucous membrane. As bearing relationship to the above mentioned, these bodies, to my reading, are homologous with the alar cartilages and their mucous membranous covering, either in whole or in part. Now, if these fibro-cartilaginous masses collectively be rudiments of what I have said, even though I may have mistaken them individually, they afford a clue by which the muscular layers above can be understood aside from their office of dilators and com- pressors of the sacs and narial orifice. Furthermore the muscles and the cartilages furnish data expressive of what the nasal sacs are themselves. Von Baer builds up his homological theory of the Cetacean nasal sacs being the turbinate bones upon one gratuitous assumption, which if unsupported by other evidence than he has given, the whole structural pile of his reasoning falls baseless. I allude to the fact that he considers the thin fibrous induplicated membrane of certain of the rugose sacs an arrested condition of the osseous twisted lamine of the turbinals. In short, he takes for granted membrane replaces bone. As to this being the case, we have not a shadow of evidence. Indeed it is far more reasonable to suppose the ethmoidal turbinates aborted or entirely absent, than that they should be represented by or be transformed into a sacculate membrane of fibro-areolar consistence. It is undoubted that the Cetacea have singularly modified skulls. Some bones are very diminished in size, some wonderfully increased, and others jammed into most abnormal positions, or are unsymmetrical ; but I think it yet lacks positive proof that a membrane completely replaces an entire cranial bone in the adult cranium. Luckily for me, Von Baer compares the Porpoise’s nasal sacs with those of a calf, 1 As bearing on the subject, consult Dr. Cleland’s well-reasoned paper “On the Relations of the Vomer, Ethmoid, and Intermaxillary Bones,” Philos. Trans. 1863, p. 289, pls. 4 & 5. 2n2 246 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. among other forms; and Professor Huxley’, who appears to support his views, instances the Tapir as having the ethmoidal turbinals posterior to the maxillary ones. Although the latter statement must be admitted, I will nevertheless confine myself to this animal, and more especially toa Ruminant, as excellent types, leading me to believe the Cetacean nasal sacs are not at all modified turbinals. The Ruminant markedly possessing the shortest nasal bones, and therefore Whale- and Tapir-like in this respect, is the Saiga (Saiga tartarica). Moreover this Ruminant, like the Tapir, has a soft narial proboscis—shorter no doubt; but also as in that animal it has well-developed turbinal bones, the maxillary one of which, if not quite in advance of the ethmoidal, is at least nearly so. But both these animals have other nasal structures simulating closely, if not indeed the veritable homologues of, the Cetacean spiracular sinuses and their rudimentary nasal cartilages. Adverting first to the Saiga’, I have found on dissection its protruberant snout to A partial upper view of the face of the male Saiga, with the nares opened on both sides, so as to display the parts which I deem homologous with the spiracular cavity and appendages of Whales. Na, shortened nasal bones, from which runs forwards (2/.c) the upper lateral cartilages comprising also the alar; a narrow fibrous cord (f.) is continued anteriorly into the soft patulous nares; mz¢ and i¢ denote respectively the middle and internal turbinals ; arrows in contrary direc- tions indicate the course of the main narial passage ; nf, sulcus corresponding to the naso-frontal canal of Cete ; m, inner orifice of the maxillary sinus or sac, the dots giving its entire outline; p, palatal recess agreeing with Cetacean premaxillary sacs. be made up of tegument, carneous substance, vessels, and nerves, lined with a delicate mucous membrane. ‘The muscles can be identified with those smaller bands &c. which act on the nose-chamber and upper lip in other and higher Mammalian orders. The nasal cartilages are insignificant, and pass but a little beyond the short nasal bones; yet ' Hunterian Lectures, 1866, Reported in abstract, Med. Times and Gazette, p. 350, March 31 of that year. ? See P. Z. S. 1870, pp. 461, 478, figs. 5 & 8. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 247 an upper lateral, alar, and rudiment of sesamoid are there. Quite within the narial passages on each side of the cavity, partially below but vertical with the alar cartilage, is a small globular bag which opens by a fissure into the nares. This sac, which I have named the maxillary sinus, is fibrous, smooth, yet glandular within, and contained sebaceous or ceruminous-like substance. Still higher on each side, and lying between the maxillary, turbinal, and lachrymal bones, and therefore nearer the nasals, is another semilunar depression or shallow sinus $ inch long, nearly vertical, though somewhat crescentic. A third and much better marked cavity exists on the floor of either moiety of the nares. This, a slipper-shaped fossa or sinus-like fold, is sunk forwards, or pro- duces a well defined step about an inch long betwixt the deeper postnarial chamber and the more raised floor of the anterior segment of the interior nares. The two last- mentioned pairs of fossa are each smooth-surfaced and lined by a continuation of the moist mucous membrane. Anteriorly the nares are capacious, and under control of the flabby but muscular parietes. Secondly, as regards the Tapir, the late Mr. H. N. Turner’ has shown, and a dissection on my own part® verifies, that quite within the nose, in the semicircular notches outside the projecting nasal bones, are two long, smooth-lined sacs or naso-frontal maxillary sinuses. Each of these is somewhat f-shaped, the blind and bulbous posterior extremity being curved inwards, the anterior straighter end freely communicating with the interior nares. These elongate sinuses are mainly hollowed out of what Turner supposes to be the lateral nasal cartilages, continuous with the septal. He thinks alar cartilages are absent; in this I do not quite agree. The muscles of the proboscis are arranged after the usual type of the nasal group, with a special pair of long levators. I refer the reader to my illustrations and account of the anterior cranial muscles of Lagenorhynchus albirostris and Grampus rissoanus, and the figures 63 to 67 of the present plates, in lieu of redescription. ‘The latter has one layer less than the former, and the diminutive fasicles connected with the naso-facial canal are not so markedly differentiated. As regards the action of the different layers in Globiceps &c., they are nearly identical in the several forms. The superior layer is a dilator of the blow-hole and compressor of the maxillary sac; the second sheet assists the first. The third set of fibres assimilates to the preceding in its use; but there is an additional mechanism of the parts induced by its upper anterior tendinous slip. This runs quite into the nasal blubber; and the fibres cross well over, so that, while creating tension of the fatty nodosity, a certain amount of backward pressure follows, and aid is lent to the elastic fibrous cushion which usually keeps the commissure of the nasal orifice closed. The small, short, and semi- circular muscle connected with the posterior canal acts as a retractor and compressor to it. What I have termed premaxillary or naso-labialis, while less fleshy in G. melas ' P. Z. 8. 1850, p. 103. 2 Rhinocherus sumatranus, Journ. of Anat. & Phys. vol. vi. p. 138, and figs. 8 & 9, pl. x. 248 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. than in the old Lagenorhynchus, nevertheless offered more appearance of separation it the latter (vide figs. 63, 64, N/). The combined action, however, agrees, viz. protraction and retraction of the lips of the blow-hole, according as the longer anterior or shorter fan-shaped vertical fibres contract, and also, coincidently, compression or dilatation of the premaxillary sac. A dilator naris, retractor ale nasi (sew pyramidalis plus constrictor naris, =com- pressores naris), and depressor ale nasi muscle, in superincumbent planes, besides a vomerine and alar fibro-cartilages, are described by Carte and Macalister as existing in Balenoptera rostrata’. Whether they or I have interpreted the structures correctly future investigators must decide. But as regards the muscles I would note that the attachments and numbers of layers more truly belong to the entire facial set of land Mammals than are only restricted to the homologues of the nasal group. 3. Skin and Subcutaneous Coverings—The skin over the entire head is very thin, hardly exceeding 0:1 of an inch; and the superficial film of cuticle has only a thickness of about 73> of an inch. Beneath the skin proper, and corresponding to the subcu- taneous tissue of other Mammals, is a dense fibro-elastic tissue, composed of innumerable reticulations of white glistening fibres, intermingled with fatty or oily material. Some of these fibres are stronger and more prominent than others, so that, as seen in vertical section (fig. 26), there appears to be an irregular meshwork of thickish and more delicate thread-like fibres laid together, warp-and-woof fashion. Some are directed longitu- dinally, others transversely, and others again entwine obliquely to the cord of the nasal prominence. Here and there between the fibres white puncta manifest themselves, these being nothing other than cross and tangential sections of the fibrille themselves. The entire body, side of the head, and throat in @. melas has an envelope of pale yellow-coloured fat, similar in consistence to that found in the same situation in the Porpoise. This fat thins as it approaches the root of the tail, and upon the caudal expansion is lost in the strong fibrous substance of which that organ is composed. On the body generally the fat has a depth of from 1 to 1} inch. In some parts, for example the throat and the chest between the pectoral limbs, it is even more. Where the latter become free it diminishes; and quite on these appendages it is barely recog- nizable. Indeed, on the limbs strong fibrous, almost gristle-like substance takes its place; and this latter, towards the free extremity, becomes itself so reduced in quantity that at the point only a very thin layer of fibro-membrane intervenes between the upper and lower dark-coloured external tegument. Upon the back, in front of the dorsal fin, the fatty tissue is more interwoven with fibres, and finally, in the fin, resolves itself into dense, firm and elastic texture, truly more cartilage-like than fibrous. Regarding the function of the semirigid dorsal fin in this and some of the allied genera we have not the remotest conception, unless balancing of the body in the watery element has something to do with it. But it does not seem a necessary ' See their chapter on its ‘ External Nares,” /.¢. p. 238. —— DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 249 adjunct to the creature's correct equilibrium; for we find several widely different genera of the Cetacean order wanting this appendage. It is within the range of possibility it may be one of those organs the homologue of which we detect in forms removed ; and to our appreciation of things, functionless, inasmuch as use to a direct purpose is indiscernible. We can, notwithstanding, descry in it, both structurally and positionally, a counterpart to the remarkable dorsal humps in some Bovide and Camelide, e.g. the Zebu and Dromedary, and thus recognize another of those links, besides compound stomachs &c., which entitle us to refer the Cetacea and Ruminantia to some ancient primordial intervening form. The dimensions and shape of the dorsal fin have already been mentioned. Here it needs only be noted that, from being thin and laterally compressed above, it gradually thickens basally and slopes outwards, merging with the general curvature of the back. But this remark, it must be borne in mind, applies only to the fixed anterior moiety ; for the posterior region, partly attached and partly, so to say, free, or where it becomes falcate in outline, alters, and is thicker above than below. (Consult the sections in Pl. XXXII. figs. 17 to 24 inclusive.) The very distinctive character which the Deductor, like the Sperm-Whale, possesses in its prominent blunt muzzle or nasal protuberance, is chiefly caused by a vast deposit of a fatty, semioleaginous material. ‘This is overlain exteriorly by the thin dermis, and beneath that by thick and dense fibrous tissue. The latter decreases inwards, from the fibres being wide-meshed and their permitting a sort of gristly-looking fat to be inter- spersed between them. Proceeding still deeper, the denseness of the fat diminishes gradually until it is succeeded by very soft blubber; and this, again becoming slightly firmer, commingles with the muscular fibres of the so-called premaxillary muscle. The main direction of the interwoven fibres among the blubber is transverse, so that when cut in that direction they appear as layered wavy lines of white glistening fibre, varying much in thickness. The entire mass of the fat and blubber of the massive boss of the frontal region, looked at from above, has a somewhat blunt wedge- or even heart-shaped aspect, in this respect agreeing with the contour of the maxillo-premaxillary region. The apex, or tapering front of the blubber, is so interwoven with the strong fibrous tissue anteriorly, that near the tip of the beak (premaxillaries) it becomes lost or indistinguishable among the fibrille. It is arched from side to side above, and also somewhat convex antero- posteriorly. Inferiorly, as noted, it loses itself among the fibres of the premaxillary muscles; but as the right and left one of these have a deep sulcus between them, this is filled up by a median keel-like portion of the blubber, which latter has thus below a wide V-shape. The greatest length of this deposit of blubber and fat is about eleven inches, the greatest vertical depth six and a half inches. 4. Auditory Appendages.—No external auditory orifice was detected by me before removal of the skin. On this being cut off and the subjacent fatty tissue searched, a 250 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. minute point was noticed on a line nearly parallel with and posterior to the angle of the mouth, the eye being above and its postcanthus nearly equidistant between them. As further dissection was proceeded with, the external auditory canal or meatus became more apparent, and was found to be, as in other Cetacea, a cord-like tube as thick as a goose-quill. It was at first directed horizontally inwards, on a level with, but behind the articular condyle of the mandible. Slightly widening it quickly resumed its calibre, and by three several spiral turns, or winding S-ways, it reached the foramen and mem- brane on the outer side of the inflated tympanic bulla, at this point being situate immediately above and in front of the cranial attachment of the stylo-hyal. The auditory tube, though in one sense free, was nevertheless retained in its position throughout by dense fibrous tissue and the other padding of the outer cranial wall; but at the upper ends of the spiral turns muscular fibres were fixed to it, of which more presently. Among the many writers on Cetacean anatomy, Von Baer alone I find mentions the existence of rudimentary muscular fibres in the Porpoise, appertaming to the external auditory apparatus. In my dissection of this female Caaing Whale, and still better in the fleshy adult male White-beaked Bottlenose (Lagenorh ynchus albirostris), | have been fortunate in discovering, not indeterminate fibres, but three well-developed muscular slips, attached to and acting upon the cartilaginous tube of the meatus. These might either represent the diminutive muscles of the helix and tragus of higher Mammals, or more probably, and as I take them to be, they are the homologues of the auricular muscles, in spite of the absence of an expanded pinna. Accordingly I have already named them as external muscles of the ear, viz. attollens, attrahens, and retrahens’. Concerning the action of these muscles, one would think their functional subservience . to the organs of hearing would be very slight. They have no outer pinna to move; how or in what way therefore do they act? Judging from their position and attach- ments (vid. fig. 29), the partially conjoined attrahens and attollens evidently drag both forwards and upwards the outer portion of the auditory tube, particularly the wider horizontal segment. The retrahens, while elevating, or, it may be, very slightly tugging the tube backwards at its root, would also, at the same time, horizontally straighten it. These movements, though of the most constrained description, may serve a purpose in audition. As a precautionary measure, in case of loss of the small ear-bones in transmission of the skull abroad, they were left attached; I had no opportunity, therefore, to examine thoroughly the organ of hearing. The gifted Hunter and others have already ably explained the auditory mechanico-physiological construction in Whales generally; I confine myself therefore to a few loose notes of the parts in situ (fig. 31). ‘The tympanic bulla, which in this inferior aspect has a long oval outline, with a capitulate pedicle (malleus?) about the middle of its outer margin, lies obliquely for- Linn. Soe. Journ. vol. xi. p. 152. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 251 wards and inwards, in the space betwixt exo- and basi-occipital, squamosal, and pterygoid bones. At the inner extremity it is slung, as it were, by a fibro-cartilaginous attach- ment from what appears as a hamular process of the pterygoid. Below this, and ante- riorly, is a broadish ligament which fastens it to the ali- and orbito-sphenoids. To the commencement of the Eustachian tube it is connected by strong fibrous tissue, such as composes the canal itself; but there is besides, somewhat rearwards and inwards, a padding of cartilage. Behind this is a great vascular plexus reaching to the condyloid foramen. Between these and the tympanic fibroid tissues obtain. Posteriorly, exter- nally and at the narrow end of the bulla the broad strip of the stylo-hyal cartilage arises from the exoccipital, and here is so imbedded as to prevent the tympanic im- pinging upon the latter bone. The facial nerve escapes from the skull by a separate foramen immediately in front of the cartilage; and the auditory tube ends in the recess and tympanic membrane hard by. In the dried skull of this species the petrotympanic bones are all but quite loose; the fibro-cartilagmous parts, above described, therefore act as cushions around them. But, moreover, the vascular network and oily and fatty substances hereafter described supply a soft external casing supplementary to audition. The Eustachian canal, as it leaves the tympanic bulla, has considerable diameter, and retains it more or less uniform as it passes forwards towards the fauces. A tough membrane and rete mirabile lie superficial to it. 5. The Tongue.—This is a fleshy organ, dorsally covered by a thick corium or leathery- like envelope. It is perfectly smooth, and superficially flat, excepting at the root. There it exhibits numerous glandular papille and depressions, probably the representa- tives of papille fungiformes; other larger and much deeper furrows behind may be either circumyallate cavities or simply mucus-glands. The glosso-pharyngeal ruge are narrow, linear, and longitudinal. The tongue’s length is 7} inches as it lies in the recumbent posture; posteriorly it is 4 inches across, narrowing regularly to the tip. About a couple of inches of the apex is free ; and there is a considerable fold of loose membrane beneath, forming a distinctly marked frenum lingue. ‘This is flesh-coloured, with nearly black outer edges. The tongue, however, is apparently not capable of protrusion beyond the mouth. The sublingual membrane is thrown into crescentic folds only moderately raised; and these interdigitate the one with the other, whilst the freenum itself connects these by rather irregular, crenate, transverse plice. I may as well at once refer to several of the fleshy masses of the submandibular region and composing the lingual organ itself. Among these the most conspicuous superficial layer, to wit, the mylo-hyoid, is of great breadth, moderate thickness, and coarse in fibre. The two muscles together spread out from the lower jaws to the hyoid, and have throughout transverse fibres. It acts as a compressor of the inframaxillar pouch in the Piked Whale’. What may be the homologue of a digastric is a relatively 1 Phil. Trans. 1868, p. 220. j VOL. Vill.—ParT tv. February, 1873. 20 252 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. enormous mass which has attachments to the skull behind the articulation to the stylo- hyal, and, partially covering the ramus of the mandible, is therein narrowly inserted at its fore part. This apparently agrees with the depressor maxille inferioris of Carte and Macalister. Beneath the last is what I presume is the representative of stylo-glossus. This like- wise is of a long wedge-shape. Posteriorly broad, it occupies nearly all the antero- inferior edge of the stylohyal, and advancing and narrowing is fixed by a tendon betwixt the side of the mandible and tongue. Essentially similar in other genera of Whales’. The hyo-glossus in G. melas I found partly double. Its smaller outer head arises from about the middle of the stylohyal; its larger broader head springs from the front border of the thyrohyal, deeper than and partly outside the genio-hyoid. The two bellies converge, run side by side, and amalgamate about the middle of the tongue, the fibres dipping into the substance of the latter. Stannius’ says that in the Porpoise the united hyo-glossus and hyo-pharyngeus springs broadly from the anterior part of the body of the hyoid and from the fore border and upper surface of the lower corner. The first thicker portion goes to the tongue’s root along with the stylo-glossus; the other proceeds outwards and upwards to the posterior or upper edge of pharynx. One origin, a round fleshy mass from the great cornu of the hyoid bone, is given to the hyo- glossus by Carte in Balenoptera. The pair of strong genio-hyoidei, which lie parallel to each other, run straight from the basihyal forwards to the mandibular symphysis. The genio-hyo-glossi are an expanded sheet, whose bundles of fibres from the inferior median raphe run upwards, forwards, and outwards in sweeping lines, intermingling with the lingualis or intrinsic muscle of the tongue. Posteriorly the genio-hyo-glossi are fixed to the ceratohyals and partially to the basihyal rostrum. TV. ORGANS SUBSERVIENT TO DEGLUTITION AND DIGESTION. 1. Cavity of the Mouth, Dental Armature, and Pharynx.—The absence of baleen in the cavity of the mouth of the Pilot Whale necessarily gives quite a different aspect to it, compared with the mouth of the Whalebone species. No view of the Toothed Whale’s mouth has hitherto been published; that which I furnish (fig. 5) was drawn from the fresh specimen, and therefore ought to give a fair idea of the buccal cavity. The dimensions of the original were :—Distance from the tip of the mandible to the angle of the mouth 113 inches, and the widest stretch or depth of gape at symphysis _ 63 inches; the roof of the mouth antero-posteriorly is above 7 inches. The soft palate, excepting behind, is of a sooty-black colour, dense or firm to the touch, fibrous in structure, and firmly adherent to the bone. ‘There is a middle flat portion similar to the tongue in shape, or lanceolate; and this has a breadth behind of 3°7 inches, and narrows anteriorly to 0-5 inch. Running round on each side of this, * Rapp, p. 132; Stannius, p.8; and Carte & Macalister, p. 231. = seapos — DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 253 and between it and the gums, there is a smooth shallow gutter an inch broad. The anterior third of the mid palate is coriaceous, rough, and bestrewed with hardened, closely placed papille, which are arranged in irregular linear rows. In the median line, however, this rough patch is longitudinally bisected by a narrow shallow sulcus, which is lost behind in the smoother palate; but anteriorly it sinks deeply between the pre- maxillary bones, and there its edges are firmly adherent to the periosteum. Quite at the back of the palate, or faucial portion, there is a smooth whitish-coloured area narrowest forwards ; and on either side of this are numerous openings of muciparous ducts. as The last left lower tooth was very small. The mandibular teeth increased in size from the first to the sixth, again decreasing in magnitude posteriorly ; nearly the same ratio obtained in those of the upper jaw, where the teeth are less curved. I noted that the sixth tooth of the mandible, the largest in the series, had an elevation of 0-4 inch above the gum, and the final one of the same row but 0:15 inch. The superior maxillary teeth were set with tips recurved, and inwards and downwards; the mandibular series agreed as to postero-inward inclination. ‘The length of the lower dental row was 6 and 64, and of the upper 5+ and 5% inches respectively. ‘The conical teeth were implanted in deep sockets, but nevertheless they possessed a certain amount of mobility when pressed laterally by the hand. This slight looseness, even in adolescence, may account for individuals, both young and old, losing them, as authorities record. The mucous membrane of the gums is continuous with the palatine tissues, but is lighter coloured. It embraces the necks of the teeth circularly, and runs between each dental interspace as a longitudinal ridge. Outside the gums and teeth, or labially, the membrane returns to its dark colour, and intensifies as it approaches the jet-black surface of the skin of the body. The portion which corresponds to the upper lip slopes gradually downwards and outwards, so as to form an eave or firm overhanging arch. This is deepest behind at the angle of the mouth, and becomes by degrees less so, until anteriorly reaching what represents the frenum labii superioris, where it is but very slightly elevated above the palate. The maxillary labial border projects sufficiently to overlap and partially hide the mandibular one, when the mouth is closed. The inner surface of the lower jaw shelves rather abruptly downwards. It is narrow and dark-coloured, towards the teeth inwardly smooth, from that rough as far back as the teeth go. The gingival membrane, like the maxillary arch, is light-hued and rises round the dental necks. So suddenly and perpendicularly does the outer smooth and jet-black mandibular surface rise, that there is no defined lower lip or ridge whatsoever. There is no clearly marked dependent uvula’; but the posterior pillars of the fauces prominently arch over the root of the tongue by the great thick palato-glossus. When the mouth is fully agape, the loose folds of the mucous membrane form lateral segments of arches, the middle or keystone dividing these by short, somewhat longitudinal, * A condition likewise mentioned by Professor Macalister, P. Z. S. 1867, p. 478. 202 The dental formula in this specimen may be expressed by 254 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. plications. As the different plice meet they give rise to unequal-sized trapezoidal - elevations, many of which are perforated by glandular orifices, which doubtless secrete abundantly during the process of deglutition. As Macalister has not failed to note, there are a pair of thick palato-glossi, whose transverse fibres line the roof of the mouth, in my specimen having a breadth of more than six inches. Some few of the anterior fibres continue to the frenum lingue; others posteriorly are lost among those of ‘the palato-pharyngeus and constrictores. That the above can act as a sphincter isthmi faucium, “capable of occluding perfectly the aperture of the pharynx,” I quite acquiesce in; but this is aided to a remarkable extent by the presence of a palato-pharyngeal muscle, likewise recorded by the above author’, and well developed in my specimen. By pharyngo-laryngeal valley, I mean the deep groove running round and within the sphincter which grasps the glottis (vide fig. 15). The inner wall of this sulcus is formed by a dipping inwards of folds of the postpharyngeal region, Its depth is irregular, half an inch at the sides, but shallowing in front and behind, where it is cranially attached. The elongated glottis is firmly grasped by the so-called sphincter, the latter leaving only an elliptical aperture, whose fleshy rim is about half an inch thick when cut through. According as the outer walls are contracted or relaxed, so are the pair of gutters or lateral channels of the floor of the pharynx narrowed or otherwise. The lining membrane of this latter portion of the pharynx is smooth, or, at most, minutely puckered. The postnarial passages possess great numbers of mucous glands. Some of these project slightly; others are embedded in shallow cavities, varying from the size of a hemp-seed to twice that. Some are locular, and the whole give a punctated character to the membrane, contrasting with the smoother superficies of the pharyngeal floor. The posterior narial passages, moreover, possess each a muscular layer, aboye chiefly composed of longitudinal fibres. These, as they descend or proceed backwards, increase in thickness, and the fleshy bundles assume an oblique and spiral direction, ultimately becoming circular as they merge into what is usually entitled the superior pharyngeal constrictor. Some of the fibres of this can be traced downwards and backwards to the surface and angle of the thyroid cartilage. Others, the enormously developed circular ones, constitute chiefly the posterior wall of the pharynx, where they meet from the opposite sides in strong, glistening tendinous fibres, covered by a narrowed portion of of the inferior constrictor. The constrictor superior, besides, has attachments to the posterior border of the palate and to the internal pterygoid plates. The constrictor medius is likewise very strong, broad, and fleshy; attachments, side and lower surface of the thyroid cartilage, and upper surface of the constrictores superioris et inferioris. The last-mentioned muscle is considerably thinner than the others. Its narrow upper point has been spoken of with the superior constrictor; its other attach- ments are the inner surface of the thyroid cartilage and descending process; the posterior fibres mingle with the circular layer of the cesophagus. " See P. Z. 8. 1867, p. 480. yt DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 255 The representative of stylo-pharyngeus arises from the stylohyal close to its cranial articulation, therefore with almost a squamous origin. It widens in its oblique or nearly transverse course as it approaches the pharyngeal constrictors, with which its fibres mingle. A superior constrictor is said to be absent in the Piked Whale’; and Macalister’ regards the postnarial sphincter of Globiceps as a displaced representative of the levator palati muscle. From my examination of several Cetaceans I am convinced that the elevator and sphincter of the pharynx is but a modification of constrictor superior, though the distinctly longitudinal fibres may be worthy of a separate name ; for I could make out, besides, a differentiated levator and tensor palati. ‘The levator muscle covers the interspace of the pterygoid plates and the Eustachian enlargement; it is fleshy forwards on the palate, narrows posteriorly, and is fixed fibro-tendimously near the tympanic region (vide fig. 30, Lagenorhynchus). ‘The circumflex or tensor palati is somewhat mingled with the last. Stannius* points out in Phocena a thyreo-pharyngeus as coming from the inferior horn of the thyroid cartilage and going downwards to the pharynx. Possibly it may be equivalent to Carte and Macalister’s kerato-pharyngeus*. If such a band exists in Globiocephalus, it is evidently part and parcel of the constrictor. To illustrate a characteristic view seldom® given of vertebrates, but one most useful Sere li. A longitudinal vertical section through the body of a male Porpoise (P. communis), slightly to the left of the median line, and with the viscera &c. retained nearly in their natural positions. sy, mandibular symphysis; 0, tongue; an arrow leads in the direction of the oral cavity; pa, palate ; nb, nasal blubber ; sp, spiracular cavity, exhibiting its upper corrugated pouches and the fleshy narial passage, continued by an arrow and dotted line; g/, glottis, thrust upwards through pharyngeal cavity; @, esophagus; ¢r, trachea; ce, cerebrum; c/, cerebellum ; cv, cervical vertebrx ; /y, lung; h, heart; li, liver; sf, stomach; d, duodenum; dp, diaphragm ; om, omentum ; 7, intes- tines; %, kidney; spr, suprarenal body; ¢, testicle; @, anus; wm, uterus masculinus, a style has been passed through the smaller external orifice of the male mammary cleft, above and forwards, from which the rudimentary male uterine vesicle is situated ; }, urinary bladder ; p, penis, within its sheath; ~, umbilicus. Phil. Trans. 1868, p. 245. * P. Z.8. 1867, p. 479. 3 Mill. Archiv, 1849, p. 10. * Memoir, p. 235. § Vrolik has produced a side view of the viscera of Hyperoodon, I. c. pl. xii. fig. 41. In it the ribs are left in place, but in front the section of brain and mouth are not displayed; this remark applies to Burmeister’s Epiodon, pl. 16. fig. 4 (cited in footnote anted, p. 235). In Huxley’s ‘ Lectures on the Elements of Comparative Anatomy,’ 1864, and Rolleston’s ‘Forms of Animal Life,’ 1570, there are some good diagrammatic generalized ideas of longitudinal sections ; but they lack precise anatomical data. 256 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. to comprehend the interior organic relations, I introduce the woodcut, figure 3. Ceteris paribus, it may stand as a type of the Cetacean formation, and is essentially applicable to the preceding and succeeding sections. 2. The Alimentary Canal.—The cesophageal tube is throughout wide, dilating some- what as it approaches the cardiac end. From the faucial aperture to where split by the upright, arytenoid, laryngeal pyramid it measures 8 inches, and thence to the stomach 23 inches. The mucous lining is of a pink hue, and plicated longitudinally. These folds widen and enlarge towards the ventral end; at the cardiac orifice, which is large, the ridges alter and merge gradually into the corrugate ruge and white epithelial lining of the first gastric cavity. At about a foot’s length from the stomach I observed a number of irregularly scattered, little, oval depressions, or openings of cesophageal glands. These (fig. 42, gl) are situated an inch or two apart, and occupy the ridges, but not the interspaces or sulci. A transverse section through the wall of the ceso- phagus (fig. 43) showed moderate outer circular, and inner longitudinal muscular coats, and a much greater lamina of fibro-areolar tissue, or middle coat, capped by a thickish mucous layer. Mr. Gulliver, in his notes on the Dundrum-Bay Whales (Glodiceps), alludes to but two compartments of the stomach, as does Williams in the Chinese species’. Dr. Jackson describes five separate cavities and a subsidiary one, besides a supplementary one connected with the first. Professor Turner assigns to the species five gastric com- partments. My examination of this female leads to me think G. melas possesses but four true digestive cavities, that which has been taken for another being merely an enlarged parietal passage between the second and third compartments. In elucidation of this discrepancy of opinion, I render a full account of my dissection of the parts, the drawings of which are shown in figs. 32, 33. The first gastric cavity (J, 7*) is by far the largest, and in several respects corresponds to the Ruminant Paunch. Its measurements were, 20 inches in extreme length, and 34 inches in widest circumference, at the cardiac end somewhere about 10 inches round, and much less than that at the opposite inferior tapering extremity. The cardiac orifice, not constricted, but a trifle narrower than the cesophagus, leads solely into this first chamber. The mucous plice at this point are slightly puckered together, but imme- diately below enlarge and form serpentine longitudinal folds with short interdigitating cross and oblique offshoots; both these diminish at the apical end of the cavity, which is comparatively smooth. Upon the left wall, 3 inches below the cardiac orifice, is a wide aperture leading into the second gastric chamber. As in the Ruminant’s paunch, below the opening into the second cavity in Glodiceps, there stretches obliquely across and downwards on the posterior wall of the first stomach a large and wide fold of ' Professor Huxley agrees with Dr. Brinton’s description, that the stomach of the Porpoise consists of three cavities (Hunt. Lect., Lancet, 1866, p. 350, and Manual of A, V. A. p. 407)—an opinion at variance with John Hunter’s statement ((icon. of Whales) and Owen’s (Art. Cetacra, Cyclop. of A. and P., and Anat. of Vert. vol. ili. p. 453). DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 257 membrane, or septal semidivision. ‘This gives a median pouched character to the first chamber, and doubtless is what Jackson mistook in his inflated specimen for a supple- mentary cavity of a crescentic form opening largely into the first. The interior of the first stomach is of a white or opaline colour, lined with an epithelial layer of membrane a5 of an inch in thickness; this ends abruptly and ring-like at the entrance into the second stomach. ‘The muscular and serous coats are moderate layers. The second chamber or true digestive cavity (JZ), as Jackson says, is a globose bag, and in our female, from upper to lower border, following the curve, measured 11 inches, and from left to right 8 inches. It lies partially in front and to the left side of the first chamber; and its upper posterior wall is adherent to the cesophagus by strong cellular tissue. Its inferior wall approximates both to the third and fourth chambers. The mucous rug of this second stomach give its great thickness. The folds are large free lappets, which radiate from the aperture of entrance towards its left wall, and in so doing wind in long, parallel, crooked lines, with innumerable subsidiary shorter diagonal intersecting plicee connecting them. In these interdigitations they resemble the cha- racter obtained in the longer second stomach of the Porpoise. The mucous membrane of this second cavity in the Deductor, as in the latter Cetacean, is of a rich florid colour, and yery vascular. The communicating aperture or passage leading from this the second to the third chamber is a constricted narrow tube or canal, 4? inches long, tunnelled between the walls of the second, fourth, and partly the third stomachs. It leaves the second stomach on its right inferior wall, an inch below the wide aperture which connects the first and second, and it enters the third stomach above and behind. There is no true sphincter at either end. The third gastric chamber (/ZZ) is also subglobular when distended, and occupies the lower interspace between the second and fourth cavities. Its diameters or, rather, semi- circular dimensions are 33 inches longitudinally and 4 inches transversely. The walls are thin, and the internal mucous coat nearly smooth and devoid of rug, although it may be noted that to the left side, and below the aperture leading to the fourth cavity, there are faint traces of longitudinal plications. The round, smooth, ring-like orifice between the third and fourth stomachs, is a little over half an inch in diameter. It is an inch distant from the opening into the second and third ; and its direction is nearly at right angles to it. The fourth digestive chamber (ZV) is in the main long and cylindrical. Its first por- tion, or left end, however, is rather wider, 23 inches deep from the orifice, and bulges, in a pouch-like manner, upward and towards the right wall of the second stomach ; between which cul-de-sac and the second stomach the tunnelled canal spoken of as connecting the second and third chambers runs. ‘This sac-like part, viewed externally, seems only a portion of the third stomach, but when opened is perfectly separate from it. The remainder of the chamber narrows as it proceeds to the pylorus. The total length of this compartment is 18 inches; following the curvature of the dilated first portion it is 34 inches in breadth, and the remainder is about 2} inches. Its walls are smooth, or 258 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. with only slight indications of longitudinal rug. ‘The true pyloric orifice is ring-like and thickened, thus approaching a sphincter in its formation, though I doubt if this designation can justly be applied to it. The diameter of the opening is fully half am inch. Comparing my interpretation of the gastric chambers with the observations of the aforesaid writers, it does sound odd that Williams and Gulliver should limit the number to two. This anomaly, I think, receives explanation on the ground that they only consider the large dilated sacs as stomachal receptacles, the narrower chambers but as portion of the intestinal tube. Jackson’s five or six gastric divisions are to me accounted for in this way. He describes (and in outline figures’) the compound stomach in the inflated condition, consequently the partial septal membrane of the first chamber appeared to him as a crescentic supplementary cavity. What I name as the canal of junction between the second and third stomachs he ayers is a distinct division. His subsidiary division, fifth or sixth as may be, is the dilated duodenal commencement. Turner? evidently takes the same view of the nature of the canal between the second and third stomachs as Jackson. Where we differ, then, is whether the burrowing passage between the vascular corrugated and large cavity JJ and the globular (JZJ/, fig. 32 in Plates) is entitled to be regarded as a true digestive division or not. I look upon it only as a communicating canal; because of its diminutive capacity* and diameter ; because it is not at all a free chamber, but, strictly speaking, like the end of the bile- duct, a tunnel burrowing its whole length betwixt the adjoining walls of ZZ and JV; because of its smooth mucous membrane showing few or no traces of digestion taking place therein; because the other four chambers agree with what obtains in Phocena, Grampus, and Balenoptera, and the two latter also offer an incipient structure of a similar kind, and corresponding in situation; and, lastly, because I regard certain of the so-called stomachs of some Cetaceans (//yperoodon for example, with six or Se as only canals between the true digestive chambers, as is shown above. As is not uncommon in Whales, the duodenum of the “ Deductor” commences by a dilatation 44 inches in transverse diameter, which, narrowing somewhat at 9 inches distance from the pylorus, is pierced slit-wise by the combined hepatic and pancreatic duct. This interval of gut has smooth-surfaced mucous membrane, and hence by some has been considered asa division of the stomach; but, as has been shown by others, the duct’s entrance and want of constriction evinces its true nature. From this point onwards to the rectum there is no separation by valvular division, cecal appendage, or sudden 1 Loc. cit. pl. 15. fig. 2. 2 Paper cited, p. 70, woodcut, fig. 2. 3 Thad almost omitted reference to a second contribution of Professor Turner’s (“ Further Observations on the Stomach in the Cetacea,” Journ. of Anat. 1869, p. 117), where he gives measurements of the several gastric com- partments of the dried and inflated stomach of an adult Globiocephalus(?) and a foetus, in support of his opinion of there being five, and not four, true digestive stomachs. The dried, distended condition of his adult specimen, as in Dr. Jackson’s case, still causes me to doubt the soundness of his argument. With reference to his third compartment or stomach (my intercommunicating passage), he notes it was 63 inches long and from 1 to 33 inches in diameter in the adult; but the same chamber cr canal in the foetus was with difficulty recognizable. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 259 change of calibre, into distinct small and large intestine, although the mucous mem- brane regionally changes its character. The total length of the gut is 97 feet 4 inches, or about nine times the length of the animal; but in Jackson’s smaller male the pro- portions were as 8 to 1, and in Williams's 7 to 1. From this I infer that, as in other Mammalia, the intestine bears a relation to the age of the animal. As to the intestinal diameter, the following tabular view expresses gradation of calibre at various points :— inches. Diameter at dilate duodenalend. . . . .... . 4 Diameter about 2 feet from pylorus. . . . . . . . 25 Diameter about 12 feet from pylorus . . . ... . 24 Diameter about 48 feet from pylorus . . ... .. Is Diameter about 90 feet from pylorus . . ..... 1 Diameter about 1 foot from the anus . . ..... 2 I can offer testimony to the general correctness of Jackson’s description of the intes- tinal tract, save his statement ‘no Peyer's nor other glands seen.” On the contrary, the glands are a predominant feature throughout. But first as to the membranous folds, which attract attention by their peculiar disposition. In the lower part of what may represent the duodenum and the upper portion of the ileum, to about 6 or 7 feet from the pylorus, valvule conniventes, in great transverse flaps, are present, just as the above author has noted. These by degrees become smaller, but nevertheless continue uninterruptedly for 36 feet or thereabouts. The mucous folds then become reduced, and there begins a tendency towards development of partial longitudinal and oblique intersecting ruge. The longitudinal plications thence are the most marked, and exist in pairs some little distance apart, running nearly parallel, but sinuous and continuously along the length of the gut. Short transverse folds unite these; and this condition obtains for 40 feet or more. Within about 3 feet or so of the anus the longitudinal parallel ridges have increased in volume and strength, whilst the short transverse and oblique rugz have relatively diminished, leaving wider and deeper interspaces between them. The intestinal glands of this zoophagous Cetacean are most interesting physiologi- cally, in number, disposition, and as evidence of important functional activity carried on over an immense area. I counted in all some twenty-four Peyer’s patches; and, although I made a careful search, it is possible others may have been overlooked. Visually I did not detect any representatives of Brunner’s glands at the duodenal end of the intestine. ‘The first agminate gland was situate ten feet from the pyloric orifice; and then, at less or greater distances, they were found for a great way on. I shall record the general appearance of these Peyer’s patches, which were twofold, and tabulate their lengths and distances apart. The first kind were those distributed amongst the large free valvule conniventes. These were ovate, and from two to three VOL. VIII.—PaRT Iv. February, 1873. 2P 260 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. times as long as they were broad, well defined, and raised prominently on the mucous membrane (vide fig. 44, P.g/). The second sort (fig. 45) were found further on in the intestine, and existed as elongate narrow strips, one half to three quarters of an inch broad, and placed between the parallel pairs of longitudinal folds. ft. in. inches, 1st Peyer’s patch, From pylorus .............-.. LOS On ete se Its length 4 2nd 5 From preceding patch .......... BNO ea on 3 3rd ey a atcone Ns aiehRels Pima see a 3 4th op PE ey be Soot oct i) aeons bp 3 5th cf te We Pie werd Bae insets SG oie n 3 6th. 5 AIR Se Co) gee) Oe ay re ates 10 7th op sy OpPr ul wheres ee Sere Br te i fa 10 8th 5 PN em Seiias soe ot Sen Oa aioe eA 6 9th Ps i lL ebnencs aad Sia (Uo aeene 5A 7 10th a reir onl Baookns mon O) Svobesos to 6 11th 5 Pomme Sth Bi tebaees chante. 3) MON ioe. Ae x 6 12th op Ps Me So ocr Oh OW eete eaters 55 4 13th x Bah a PE See Foereeters GEOL Bs 7 14th > sta -ok 6 Hepes TENOR ete. 6 63 15th on Sask Re Vel ome as ecco Lf aeosen xr 2 16th oD File aside & dasade sae. 2 @) ascoes a 5 17th D Fi Auli sie alah o.c.cirib 0.4566 IL. O-poveos » 4 18th oS Shiemaee Veen: eee ak Tee Vane nee 35 2 19th op By Cn TS seerooos 34 Oiad) ex-feetek #9 5 20thy Sos, cpt Ve rasan oop oad OL TO My. joferae 93 4 21st bp a stelle foes a ae DIGS ar sues op 10 22nd > AGB ed tooo. ON Sa eee ot 3 23rd 5 as Met ecneiet IDSC O, pBNd ope rf 1 24th * Ce OE rT on DUG so asoc cE 23 The haste necessitated by the disagreeable stench which arose on the abdomen being opened did not allow of full detail or study of the relative position of the viscera being made; but the relation of parts was thus noted. The liver occupied the interspace between the diaphragm, the second and fourth stomachs, and the duodenal flexure. The omentum was thin and contained little fat, and only partially covered the intestines. Its visceral attachment was in a semicircular manner across the middle of the first gastric cavity, and slightly over the upper part of the spleen and lower anterior or ventral margin of the pancreas. From the first stomach on the left side it passed on to the lower curvature of the second, proceeding across to the right side along the lower margins of the third and fourth gastric divisions, thence to the duodenum and upper gut (dotted line, fig. 33). The second and third lay nearly across the cardiac end of the first stomach. ‘The intestinal loops appeared to fill the remainder of the abdominal cavity, hiding all other parts, and even portions of the stomachs themselves. 3. Glands accessory to Alimentation.—By some’ authorities the Cetacea are said to 1 Cuvier, Lecons; Meckel, Anat. Comp. vol. viii. p. 375; Huxley, Hunterian Lectures, 1866, and Manual A.V. A. 1871, p. 394. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 261 have no salivary glands; by others’ that they are reduced to the most rudimental con- dition, while a third set of observers” evidently point to their presence as fairly deve- loped. In my own researches among the group, notably the toothed section, I have satisfied myself that two of them obtain, and these not so feebly represented as I had been led to imagine. Their position and relations in Globiocephalus, Grampus, and Lagenorhynchus sufficiently agree for one description to serve. The parotid, firm and thick, is situated behind the auditory canal and the small fleshy slip connected there- with. It occupies, and somewhat deeply, the angle between the insertion of the sterno- mastoid and cephalo-humeral and the anterior nuchal continuations of the long dorsal muscles. Steno’s duct, a fair-sized tube, passes forwards to the cheek. The sub- maxillary gland is flatter and thinner, but with a superficies almost as large as the parotid. It lies between, and partially overlaps, the neighbouring borders of the large muscle representing the digastric and the masseter, inferior to the auditory tube and in front of the cephalo-humeral muscle near its insertion. The facial artery and nerve either partially pierce or lie closely adherent and beneath the upper margin of the submaxillary gland. The cutaneous muscle covers both. The liver (fig. 34), in simplicity, agrees with that of other Cete; but, as considered divisible into a right and left lobe, these were nearly of equal size, and not with a pre- ponderant right, as Dr. Jackson*® found in his specimen, nor with a left enlargement, as the same author describes in the Sperm-Whale. With an average diameter of 18 inches, the organ is rather flat, smooth, and of medium thickness. There is no gall-bladder—as all observers record, excepting Williams‘, who says in G. chinensis it is small. The round ligament is of considerable thickness, and dips into the anterior median incision or umbilical fissure. I found its vessel all but closed; but in the younger Craigie’s-Bridge specimen it was “ pervious, opening freely into the vena porte.” The broad ligament has a nearly mesial line of attachment, and is strong. The coronary and the two lateral ligaments are fused together, and cover only the inferior edge of the right moiety of the gland. The umbilical is the only well-marked fissure; that for the vena cava is broad and shallow; and a mere central indentation marks a transverse fissure where the hepatic and portal vessels enter. A single hepatic duct, half an inch in diameter, is joined by the united double and much narrower pancreatic duct, about two inches from the liver. Thence, three inches further on, it enters the serous coat of the duodenum and forms a dilated bile-reservoir; a narrow passage, five inches long, continues the duct within the intestinal wall; and it pierces the mucous coat nine inches distant from the pylorus. The expanded portion of the duct has alone narrow transverse rug within. 1 Owen’s Anat. of Vertebrates, vol. iii., submaxillary and sublinguals in a diffused form in Whalebone Whales but not present in others; Fréd. Cuvier, Cyclop. art. Cetacea, p. 572; Eschricht, Ueber die nordischen Wallthiere, 1849, p. 108. 2 Carte and Macalister, Memoir, pp. 222, 223. * Loc, cit. p. 163, and p. 144. 4 Chinese Repository, 1838, p. 412. 2P2 262 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. The softened condition of the pancreas prevented my making as satisfactory an exami- nation as I could have wished. It measured 113 inches, had a breath of between 4 and 5 inches, and weighed 84 ounces. It occupied chiefly the interspace between the right border of the first stomach and the duodenal loop, being covered in great part by the omentum and gut. Its excretory duct, as mentioned, joined the hepatic. Drs. Williams, Jackson, and Turner nearly agree as to the proportions of this organ, its breadth under half its length; their specimens were younger than the above. Professor Turner’s observations’ on the lacteal vessels and mesenteric glands bear the stamp of accuracy. I am at one with him regarding Abernethy’s” supposed great bags. These are undoubtedly the product of decomposition ; for I satisfied myself, on studying several transverse sections in my specimen of G. melas, where considerable cavities existed, that these were solely due to disintegration of the interior tissue and not to be confounded with the lymphatic sinuses. I may further say, from long experience in such matters, that the mesenteric glands, next to the blood and brains, soonest spoil and internally decompose. ‘The rectal cluster of glands mentioned by Turner are shown in my figure 73. The spleen, &c., I shall notice in connexion with the blood-reservoirs. V. RESPIRATION AND MACHINERY INVOLVED. 1. Hyoidean and laryngeal Structures —The hyoidean arch (figs. 14, 15, & 16) con- sists of five separate elements, a single and two pairs of bones. The body, or more or less ankylosed thyro- and basihyals, is a broad and thickish crescentiform bone, whose widest diameter, from tip to tip, is 9} inches. From the anterior border there juts forwards a flat rostrum, basihyal, 1-7 inch wide, which terminally forks, and to the extremities of which the ceratohyals are affixed by a fibro-cartilaginous jomt. The latter bones are each some 2 inches long, 58; of an inch in diameter, subcylindrical, and very slightly curved. Another fibro-cartilaginous joint exists between the ceratohyal and the stylohyal bone ; and these are bent at a sharp angle to each other. Individually the stylohyal osseous rods are stout, thickish at the middle third, or 1} inch in diameter at this point, and with a length of 8} inches. The larynx has the common Cetacean formation of an elongate, nearly upright, and slightly efflect, tubular epiglottis and arytenoid cartilages. The latter are rather higher than the former’, and with an emarginate lip front forwards. The former is a broader semilune, the aperture being relatively small, and in the ordinary contracted condition, widest transversely. The tube is narrowest in the middle, and basally wide. ‘The body of the thyroid cartilage is somewhat flat and with large expanded ale. ‘These jut out 1 Pilot Whale, 7. c. p. 76. 2 “Some particulars in the Anatomy of a Whale,” Phil. Trans. 1796, vol. xvii. p. 673. * In the young specimen dissected at Boston, U.S., the arytenoid cartilages are mentioned as not rising quite so high as the epiglottis, /. c. p. 165. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 263 broadly with an anterior protuberant blunt corner superior cornu. Posteriorly the wing sweeps backwards and inwards in a long limb, the inferior cornu, which terminally narrows and is fastened to the side and hinder margin of the cricoid cartilage. Between the inner edge of the posterior or inferior thyroidal cornu is a large oval space, partially occupied by the thyro-arytenoid and crico-arytenoideus lateralis muscles. The posterior surface of the cricoid has a considerable arched expanse, and is a trifle longer than the body of the thyroid cartilage. Behind, it is deeply incised or bifid, as Jackson specifies ; in front, at the junction of the arytenoids, more abruptly transverse. Anteriorly the cricoid, like the thyroid cartilage, sends off a pair of cornua, but which proceed in an opposite direction, pass through the lateral thyroidal spaces, and, rounding the edges of the body of the thyroid cartilage, embrace the trachea in front. These cricoid limbs, or rods, well nigh meet in the middle line, and maintain a fair thickness throughout. A strong crescentiform sheet of ligamentous membrane unites the thyroid body and thyrohyal bone. Another expanded sheet connects the latter, the basi-, cerato-, and stylohyals; both membranes have fleshy fasicles overlying them. Besides the muscles, chiefly protractors, having attachment to the hyoidean apparatus, and which I have incidentally mentioned along with those of the tongue &c., the sterno- hyoidei and thyroidei are two most powerful agents influencing retraction. The former, broad throughout, widen as they are inserted into the basi- and thyrohyals. The latter are not nearly so voluminous, and more ribband-like in figure. Their origin, as Mac- alister’ observes, is from the first costal cartilage as well as sternum; their insertion, side of thyroid cartilage. The sterno-thyroidei are wanting in B. rostrata”, but are cer- tainly present in several other forms of Cete. The thyro-hyoidei broadly cover the under surface of the thyroid cartilage and the thyro-hyoidean membrane. Each crico-thyroideus, fleshy, and of considerable size, is attached to the inner border and anterior surface of the cricoid cartilage and lower border of the lateral horn; thence it reaches the inner border of the inferior thyroidean cornu. Upon the back or deep surface of the cricoid and posterior root of the arytenoid cartilages a sheet of muscular fibres exists (fig. 13). From the devious direction of these, more than actual division, two muscles on each side may be noted, the posterior crico-arytenoid® and lateral crico-arytenoid‘. The former, probably the stronger and larger moiety, overlays the dorsum of the cricoid, and, directed outwards and forwards, is fixed to the root and posterior half of the arytenoid cartilage. The latter lies more to the side, and, besides covering the cricoid laterally, has a partial origin from the inner edge of the inferior thyroid cornu; its fibres converge upwards and forwards to the lateral root of the arytenoid in union with the former muscle. A well-defined, 1 P. Z.8. 1867, p. 480. ? Trans. Roy. Soc. 1868, p. 219. * Balenoptera, |. c. p. 237. * Not mentioned by these authors ibid., but both described by Stannius, ‘ Mill. Archiv,’ 1849. 264 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. short, but strong thyro-arytenoideus' muscle lies in front and obliquely to the crico- arytenoideus lateralis; it passes upwards upon the post-lateral root of the upright ary- tenoid cartilage. The true arytenoidei’, each appropriately named arytenoideus trans- versus by Stannius in the Porpoise, have a similar direction in Globiceps. ‘The short but well-developed pair of muscles lie athwart and partially cover the anterior fibres of the posterior crico-arytenodei. In fact they bridge the hinder base of the adnate arytenoid cartilages, but do not run up its erect portion. As in other Cetaceans, there is a pair of hyo-epiglottic muscles, which approach close to each other mesially in front. Each is powerful, and must considerably influence the movements of the glottis, as taking their fixed point from the hyoid bone. They are each of an elongate wedge-shape, point upwards, and externally have two faces, an anterior and a lateral. The broad basal origin is from the hyoidean arch, namely those parts representing cerato- and stylohyals—the whole upper surface of the former, and about one half of the adjoining inner margin of the latter. Converging from these points upwards, the muscle is fixed to the lower moiety of the erect and firm epiglottis, in front and partially to its side. Three additional muscles, viz. a superior and an inferior aryteno-epiglottideus, and an accessory aryteno-epiglottideus, are found in Balenoptera rostrata*. The wider separation of the epiglottis and arytenoid cartilages, with inter- vening folds, may doubtless in this animal necessitate their presence for approximation of the walls of the glottis. Unless part of the hyo-epiglotticus, as above described, include these additional slips, they appear to be absent in several Cetacean genera. As having continuity with the hyo-epiglotticus, | may here mention a small, but distinct muscle arising cranially from the neighbourhood of the auditory canal, and inserted halfway down the stylohyal, close to the posterior end of the hyo-epiglotticus. This doubtless answers to the occipito-hyoideus of Rapp* and Stannius’ in the Porpoise, and stylo-keratic and squamo-styloid of Macalister®. There is a fleshy sheet in the space intervening between the stylohyal, the cerato-, and the thyrohyal in G. melas, the interhyoideus. Macalister, in his short notice of G. svineval, names it hyo-keratic’, and suggests that it is probably a modified hyo- glossus. This, to my reading, is equivalent to the stylohyoid of Stannius*, He describes it in D. phocena as springing along the entire posterior borders of the two pieces of the anterior horns of the hyoid, but also receiving fascicles from the first piece itself. Its oblique fibres are fastened to the upper surface of the body of the hyoid, and partly to the posterior border of its hinder lower cornua. In short, it fills the interval of the hyoidean elements. The same appears to me to be what Carte and Macalister Carte and Macalister, 7. c. p. 238; Stannius, 7. c. p. 11. 1 2 The A. proprius of Carte and Macalister. 3 Loe. cit. pp. 237, 238. * Op. cit. p. 132. 5 Loc. cit. p. 7. 6 In G. svineval, P. Z. 8. 1867, p. 480; and Trans. Roy. Soc. 1868, p. 235. 7 Tbid., marked p in woodcut, p. 478. 8 Loc. cit. p. 7; and Rapp, J. ¢. p. 132. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 265 differentiate in B. rostrata as superficial hyo-keratic, deep hyo-keratic, and kerato- pharyngeus'. Under the title of occipito-thyroideus, Stannius® has called attention to a partially separate slip of muscle adjoining the preceding. He says that in the Porpoise it arises along and from the sharp borders of the exoccipital, and is inserted on the side of the thyroid cartilage and angle towards the epiglottis. In Globiceps there is a muscle corresponding to this, but partially with a thyrohyal attachment. Macalister* desig- nates it basio-thyro-hyoid, and believes it to be an expanded representative of the cephalo-pharyngeus of Theile. 2. Tracheo-pulmonary parts.—The short trachea bifurcates into short, right and left bronchi, about the top of the upper third of the pulmonary organs; and there is a tertiary bronchus to the right. In this respect Globiceps agrees with such Cetacea as have been dissected, save Balena mysticetus. The lungs, when taken out of the chest, were collapsed, and contained apparently but a very small amount of residual air. This gave them the reverse of a crepitant character, viz. a lax spongy kind of texture under pressure. Moreover their pleural covering is such a strong, tough, fibrous envelope, that they acquire marginally an almost leathery consistence. Having inflated the lungs, on their removal en masse with the heart (reproduced in the sketch, fig. 49), their shape and another most remarkable peculiarity were fully disclosed. Each lung is elongate and, strictly speaking, unilobed. But there is a small anterior or apical emargination or wide shallow cleft which tends to mark off an isthmus or indefinite tongue-shaped corner. This lobule, if one may so term it, passes sternally inwards, towards its fellow of the opposite side, at the root of the heart. The area enclosing the heart, great vessels at its root, and the pericardium, is neces- sarily large; but the basal surface of the lungs, or that fitting upon the diaphragm, is likewise relatively enormous. This diaphragmatic superficies is dome-shaped, and doubtless is chiefly intended to receive the capacious compound stomach when dis- tended. At the superficial diaphragmatic end the two lungs are connected by a bridge of pleural membrane, which is also fastened to the diaphragm. I ascertained the length of the lungs to be, right 23, and the left 25 inches. In the Chinese Glodiceps it is stated “the lungs have two lobes on each side—the two central lobes are broad, flat, and thin, but as long as the lateral lobes, and both are well supplied with bronchia”*. A thoroughly cleft and compound lung is certainly an unusual condition in Cetacea; but I conjecture from Dr. Williams's allusion to the longitudinal nature of the lobes, that he but means the long, thinner, free approximate margins partially covering the heart &c. The most interesting feature connected with the pulmonary parts of the Caaing Whale is the presence of a pair, or more, of large lymphatic glands most prominently 1 Loe. cit. p. 235, and pl. 6. figs. 3h, 18, and f. 2 Loc. cit. p. 9. 3 Papers cited. * Repository already cited, p. 412. 266 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. and oddly situated. At first I thought these were unknown, but found they had been incidentally noticed by Hunter’ and Gulliver?, Williams’, and Jackson‘. The two latter entirely misconceived their true nature, thinking them the product of diseased tuberculous deposits. That they are not of pathological origin I am certain, and, moreover, found them well developed at the same spot in Risso’s Grampus. In the female under present consideration the bodies in question lie at the sterno-ventral corner of each lung, and upon the pleural bridge already spoken of (consult respec- tively figs. 49, 50, & 51). They appear as two raised, oval, or even somewhat reniform, glandular bodies, each about a couple of inches long and an inch or so in widest dia- meter, covered by opaque pleural membrane; yet the surface of the lungs themselves, but more especially their pectoral aspect, shows that they are overspread by a series of parallel sinuous veins and arteries. ‘These converge to the glands at the pleural bridge, send a few twigs into their substance, but almost wholly pass over and underneath them to large vessels close by. The latter I shall take into consideration along with the circulatory organs. Regarding the structure of each glandular body, it is firm, as Gulliver says juicy, and texturally in all respects resembles a lymphatic. It is not, therefore, a blood-reservoir, as I was inclined to deem it; for I convinced myself by injecting the vessels hard by. It was then I became fully satisfied of Mr. Gulliver's mistake, who considered the linear pulmonary tracery to be lymphatic vessels running to the glands. There are many lymphatic vessels in the neighbourhood, it is true; but these are very secondary, compared with the blood-channels, in producing the radial pulmonary lines from the glands, so very obvious even on casual inspection. Besides the more prominent pair of glands, there are, particularly in Grampus rissoanus, some smaller deeper ones adjoining the larger. VI. SANGUIFEROUS DISTRIBUTION. 1. The Heart.—This bulky organ agreed sufficiently with Dr. Williams and Jackson’s notice of its exterior; and the cavities and valves so corresponded with those of other large Fin-Whales’, save in one most unusual circumstance, that I need only call atten- tion to the latter. This certainly very remarkable and abnormal structure was the existence, in the septal segment of the tricuspid valve, of an opening the shape and size of an ordinary bean. The said opening, moreover, I found was guarded by a valve similar in every respect to the ventral valve occurring in the auriculo-ventricular orifice of the left ventricle. In 1 In Delphinus tursio, ‘ Essays and Observations,’ vol. ii. p. 107. ° L. c. p. 65. 3 LZ. c. p. 412. 4 ZL. c. p. 164. 5 Several preparations of the cardiac valves of the Rorqual (Physalus antiquorum=Balenoptera musculus, dissected by me in 1859, P. Z.S. 1865, p. 213) are preserved in the Hunterian Museum (nos. 927 B to Be). These have been fully described by my old colleague Dr. Pettigrew, and need no comment on my part. See his ‘‘ Valves of the Vascular System of Vertebrata,’ Trans. Roy. Soc. Edinb. 1864; and Appendix, R. C, 8. E. Museum Report, 1865. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 267 other words, the valve consisted of two segments, the delicate free margins of the segments being supplied with numerous fine chord tendinee attached to the rudi- mentary musculi papillares of the right ventricle. From this it follows that the supplementary mitral valve within the septal segment of the tricuspid (this valve within a valve) appears to be acted upon by the musculi papillares and by the blood, precisely in the same manner as the mitral and tricuspid valves themselves; 7. e. it is opened during the diastole of the heart, and closed during its systole. By reference to figures 52 & 53 a better conception of the position of the perforation and its relation to the tricuspid will be gathered than by entering into minutiz verbally. I may remark, however, that this ancillary valve, whatever its homological significance, appears to me to be to the valvular system of the heart, what the rete mirabile is to the vascular system generally. One may regard it as a tendency in nature to repeat herself, a something superadded without apparently any special end to be served. It may be imagined to weaken rather than strengthen the action of the tricuspid; on the other hand, it may be a kind of safety-valve’. From its anatomical structure, its physiological influence can only be conjectured. ‘The rarity of such an instance shows that it is not a necessity to Cetacean circulation. Biouk: S\ The arch and great arteries of the Pilot Whale, after Turner, from the Journ. of Anat. ii. p. 67. A, aorta; P, pulmonary artery; D, ductus arteriosus; ¢, carotis cerebralis, the diminution in calibre of this artery in the course of its ascent is not sufficiently represented in the figure; 4, carotis facialis; c, subclavia; d, cervico-occipalis ; e, thoracica posterior dextra; f, transversalis colli; g, mammaria interna; h, thoracica posterior sinistra. * See a paper by T. King “On the Safety-valve Functions in the Right Ventricle of the Heart,” Guy’s Hosp. Reports, vol. ii. : VoL. Vill.—ParT Iv. February, 1873. 2Q 268 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 2. Vascular Channels and Reservoirs.—My notes respecting the great vessels springing from the arch of the aorta were less complete than those of Professor Turner ; I must be content, therefore, to abide by his description—adding that, in his specimen as in mine, three main trunks were given off from the arch (vide fig. 49). Of the cranial vascular distribution, circumstances did not permit me to master it in detail. The more notable observations I could make were chiefly regarding a great plexus situated at the inferior base of the skull, and situated with a rete occupying the proximal infundibular cavity of the mandible. I subsequently had an opportunity of investigating the same in Grampus and Lagenorhynchus, where it likewise obtains’. The internal maxillary artery having passed deeply behind the lower jaw, and made a bend, sends forwards a long inferior dental artery. As this pursues its course it distributes ramuscles among the fatty matters and plexus presently to be mentioned. The mandibular cavity contains a mass of softish, marrow-like substance, held together by a network of fibrous tissue. Moreover the interstices are occupied with a maze of vascular channels partly composed of arterial and partly of venous capillaries, interwoven irregularly. Next the bone the tissue and vessels are firmly adherent to the periosteum. In some Cetaceans, ¢. g. the great Balena mysticetus and Balenoptera musculus, as 1 have been myself a witness to, the cavity in question possesses a perfectly enormous amount of oily material. Even in smaller genera the quantity is by no means sparse ; so that the tissues hereabouts as a whole and on section may be compared to blubber surcharged with blood-vessels. Further on, the internal maxillary gives off large mus- cular branches and others forming pterygo-maxillary divisions. These latter were not followed into the cranium. The inferior base of the skull, from the tympanic bone forwards to the maxillary, internally bounded by the levator or sphincter muscle of the postnares, presents one continuous rete mirabile. This spongy network of vessels lies upon a thick layer of fibroid tissue ; and the vessels anastomose with the aforesaid mandibular rete, whilst they likewise appear to intercommunicate with another venous locular network behind and at the root of the Eustachian tube. The venous capillaries collect into a jugular channel, more or less connected with the rete of the neck. The facial artery, vein, and nerve emerge from beneath the cephalo-humeral muscle and submaxillary gland below the auditory canal, thence traverse the face. Steno’s duct bears them company. There is a vascular plexus behind the ear-tube, and above and overlapping the parotid gland. The multitudinous divisions forming the great rete mirabile of the neck, limb, thorax, and spine bore correspondence to the arrangement of the zoophagous Cete, amply ! Of original observations on Cetacean circulation, after Tyson, Hunter, Meckel, Breschet, and Von Baer, all. well known, those of Stannius commend notice (“ Ueber den Verlauf der Arterien bei Delphinus phocena,” Mill. Arch. 1841, p. 379) as being simple, explicit, yet full of detail. For copious illustration, probably a trifle too diagrammatic, Barkow’s plates stand unrivalled, ‘ Die Blutgefiisse vorziiglich die Schlagadern der Siiugethiere,’ 1864, p. iy. Professor Turner’s recent contributions are highly valuable. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 269 discussed by preceding authorities. In fig. 54 I give an illustration, by a transverse section between two of the lumbar vertebrae, of the appearance of the cut vessels in their recent state. Two immense veins lie on either side of the spinal cord, the remainder of the space above and outside them being filled by a closely packed reticular mass of venous and arterial channels of diminished calibre. The nervous cord itself is small compared with the neural canal and vascular aggregation. The deep lumbar plexus extends outwards from the side of the vertebral body to the tip of the transverse process; and in the intervertebral spaces branches communicate with the spinal rete. A strong flat arch of fascia extends between the points mentioned, and it binds down the retia, whilst superficial to it is the immense inferior lumbo- caudal muscle. The plexuses are essentially composed of arteries and veins, varying in proportion according to the situation. Besides the ordinary large and small vessels, there is a vast congeries of capillaries of uniform calibre. Of the latter a primary series runs in parallel lines, often straight, and frequently looped and contorted. Their diameter may be compared to the thickness of a hog’s bristle. They anastomose freely, bridging and intercommunicating by acute forks, the divisions continuing parallel, either side by side, or twisted like a rope, or occasionally tangentally twirling away from each other. Besides the above capillary network, a subsidiary or secondary series springs from the first. These, even when injected, are only half the diameter of a human hair, and in their branches and ramifications diminish to a far finer tenuity. The secondary arise everywhere from the first series, and often run alongside and between them; so that one of the primary may have several of its satellites spun around it in its course. The finer ramifications, however, are by no means so regularly parallel, but split up into thousands of divisions, branching and divaricating in all manner of ways, forming an intercommunicating network, intertwined with the loops of the primary series. The venous radicals have an arrangement somewhat similar; and their primary capillaries lie in juxtaposition and are spread amongst the arterial series. By Hunter and others it is said that in Cetacea the abdominal aorta does not send off any external iliacs. I have found the following condition obtain in this female G. melas. A little way beyond the inferior mesenteric artery and lumbar branches there arose laterally and at right angles from the aorta two short wide stems, common iliac arteries. Each of these was fully an inch long, and split into two divisions, or what may be termed anterior and posterior. The former or hypogastric artery proceeded forwards partly enwrapped by a fold of the uterine broad ligament. Abdominally to this it went to the apex of the urinary bladder, distributing branches thereabouts, and joining its fellow artery of the opposite side, ultimately constituting an umbilical artery, which cord was nearly impervious. The latter or posterior division of the short aortic stem, at two inches distance from the anterior one, bifurcated. One of these two branches, which I take to be an external iliac, almost immediately beyond its origin 2Q2 270 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. split and sent rami to the parietes of the abdomen and genital parts. The second of the two branches, corresponding to the internal iliac, only proceeded for an inch and a half, when it separated into an artery and minor branches, apparently supplying the fleshy and other deep genito-urinary regions. The other artery of equal calibre I followed to beneath the os innominatum and interpelvie fascia; there it broke up into several diminutive channels: some of these were distributed to the pubo- and ilio-coccygeus muscle &c.; others, with a nerve, pierced the interpelvic fascia at the notch, just behind the anterior capitulum of the bone’. The vena cava inferior, after leaving the deep border of the liver and passing to the diaphragm, received thereon numerous phrenic tributaries. Among the large venous sinuses situated at the pleural bridge, there were many tortuous twigs, already alluded to. The series of these derived from the surface of the lungs advanced towards each pulmonary gland, and passing both above and beneath, joined the larger vessels. The great vein-channels at this part did not appear to have valves*, but instead interiorly showed a compound character. Longitudinal septa existed and divided the vessel into compartments of unequal calibre. Moreover, here and there intercommunication took place between the larger passages by oblique twigs, which ramified within the septal walls, Thus a kind of retial sinus obtains, something like, but less complicated than that within the neural canal. An artery which I took for the superior phrenic branch of the internal mammary accompanied the phrenic nerve. This, besides dividing into branches, supplying the diaphragm, pleura, and adjoining parts, also yielded the offshoots which ramified superficially and deeply upon the lymphatic bodies, and then, as the arterial network, spread in long tortuous parallel radii over the surface of the lungs*. As a vascular reservoir, I here insert remarks on the spleen. It lies upon the right side of the first stomach, below the left extremity of the pancreas, and just free from the omental attachment. It is said by others to be compound. In my specimen it is composed of three lobes, together somewhat quadriform and flattened, but each broader than long. It is moderately firm, and weighed 103 ounces. Altogether it is relatively small, 53 inches long by 4 inches broad. In various parts of the body, but notably in the neck, are firm glands, many, doubt- less, belonging to the lymphatic system, but some which seem rather to partake of the nature of blood-reservoirs. I examined several of these, but I did not inject them, so possibly may be mistaken as to their true relationship with the circulatory apparatus. In one delineated in figs. 40, 41, it is seen that the cervical vascular plexus is in close Consult Professor Struther’s remarks, “ Anat. of a Great Fin-Whale,” Journ. of Anat. 1871, p.110, pl. 8. fig. 3. Barkow, op. cited, pl. 16. fig. 3, shows that imperfect valves do exist in certain of the veins of Balenoptera, “On a Supplementary System of Nutrient Arteries for the Lungs,” Brit. & For. Med. Chir. Review, January 1865. In this paper Professor Turner demonstrates a vascular arrangement in the human body in many ways according with the above. Vide also Dr. Chiene, “ Obliteration of Coeliac and Mesent. Arteries,” Journ. of Anat. 1869, p. 65. 1 3 3 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE, 271 connexion with it exteriorly, and even more so than is shown in the drawing; for a great part has been cleared away to bring out the shape of the gland. Many of the vessels also penetrate the glandular substance, and ramify therein. The centre, somewhat fibrous in texture, is more solid than the cortex; the latter locular, and in section uncommonly like the open mouths of the surrounding rete mirabile. The very great blood-supply which these glands receive, makes one suspect their function to be sub- servient to the extraordinary sanguineous distribution; yet, on the other hand, their constitution, ceteris paribus, agrees with a magnified lymphatic gland. The loculi, often the size of a pin’s head, evidently correspond with the so-called lymphatic sinuses ; but they are diminutive as compared with the cavities of disintegration met with by me in the mesenteric glands. Some novel and important observations of Professor Turner on a large moniliform tube in the mesentery of B. sibbaldii', make me see likeness in it to the above-mentioned glands. Both, taken in conjunction with the superficial pulmonary glands, would seem to support Turner’s views of their being diffusers of the arterial stream. ‘This theory, however, I am not inclined to adopt. I recognize in them intimate association with the absorbents, kinds of lymphatic hearts in their way; and moreover I apprehend they bear a relationship to the so-called coccygeal gland of human anatomy. It is quite conceivable, and, indeed, I think probable, that the vascular and absorbent systems of the Cetacea are in far closer connexion than has hitherto been supposed. Such a view is credible, and supported by those who maintain that there is a direct interchange of material between the lymph- and blood-channels within the ordinary lymphatic glands’. Fig. 5, Diagram illustrating division of the iliac arteries of the female G. melas. ao, abdominal aorta; im, inferior mesenteric; 7, lateral lumbar branch ; c7, ci*, common iliac trunk; hy, hypo- gastric; ez, external iliac division; 7, internal iliac an (1, 2) branches. ' Loe. cit. p. 231. * Hunter, always alert in physiological inferences, preserved a most interesting specimen, no. 862, Cat. Mus. : 272 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. VII. Tue Nervous CEentTRE. I can only offer a disjointed fragment concerning the brain, which in the present instance was scooped out piecemeal through the foramen magnum. ‘The portions thus rudely extracted comprised almost, if not quite, the entire encephalon; but this was softened and decomposed. It weighed altogether 54 lbs. avoirdupois. If the body weighed a ton, or thereabouts, as estimated by the railway officials during transport of the creature, then the brain would approximately be something like zo of the weight of the body. Portions of the cerebrum, somewhat more intact than others, showed that, as in the Porpoise, the brain of the Pilot Whale is highly convoluted, and that likewise there is a very considerable superficial thickness and amount of grey matter covering the white substance. The tentorium is composed of a strong fibrous texture. An accurate and detailed account of the brain of a large Cetacean is a great deside- ratum, but one which as yet is not easy to be provided. I look forward, therefore, with interest to Professor Turner’s promised contribution in this untrodden field. VIII. Fiesuy Motor AGEnts or Bopy anp Liss. When I made my dissection, and had the accompanying plates lithographed, illustra- tions of the kind (i. e. showing the whole body and in layers) were unknown to me. Indeed, in this respect I believe they will still stand ground as original views. The literature on Cetacean myology has since received considerable accessions, especially as regards limb-flexors and extensors. Several authorities thereon have already been quoted in my preliminary note, anted, p. 235. 1. Muscles acting chiefly on the Pectoral Limb.—The superficial layer, that which represents the panniculus carnosus of land Mammals, is distributed over nearly the whole surface of the body, but leaves the tail free. What corresponds to Cuvier’s thoracico-fascien, or may be equivalent to platysma myoides, covers the throat from the angle of the mouth back to the pectoralis major. According to the part of the body so are the fibres of the entire panniculus differently directed; and they taper posteriorly upon the side vertically above the genital outlet. A great expanse seems to act upon the pectoral extremity, both superficially and deeply through an axillary portion (Cuvier’s dermo-humérien) joining the latissimus dorsi. A semirotating motion, there- fore, is imparted to the limbs. The great expanse of this muscle and the accessory offshoot to the fore limb present little deviation among the Whale tribe. the description runs :—*“ A portion of a plexus of absorbent vessels from the head of a Spermaceti-Whale (Physeter macrocephalus, Linn.), filled with spermaceti, which was by their action in process of removal. This most valuable preparation affords a strong argument in favour of the doctrine that the waste and superfluous parts are removed by the absorbents, and not by the veins.” On the contrary, consult Professor Wharton Jones’s “ Caudal Lymphatic of the Eel,” Phil. Trans. 1868, p. 675; where he demonstrates the phenomena attending the propulsion of the lymph from the caudal heart into the caudal vein &c. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 273 What constitutes the pectoralis major in Cetaceans is a moot point. Cuvier, in his Lecons', says, “ Dans les mammiféres qui n’ont point de clavicules parfaites, méme dans le dauphin, il y a une premiére portion sternale qui va perpendiculairement a la ligne dpre, et qui forme avec la portion correspondante de l’autre cété, ce que Yon a appelé le muscle commun aux deux bras; cest lui qui produit l’entre-croisement des jambes de devant.” It is evidently this same portion which Meckel*, with whom Rapp* coincides, refers to as an elongated strong triangle, partially continuous with the panniculus, and inserted broadly into the humerus as far as the antibrachial aponeurosis. Stannius‘, however, views it differently, regarding this only as a thoracic portion of the panniculus; and the true pectoralis major he specially points out to be what Rapp‘ and, possibly, Meckel have considered the pectoralis minor. There may be some good grounds for the conclusion which Stannius has arrived at, that the deeper portion is pectoralis major and not pectoralis minor, if it is allowed, there are two layers of the former, which to some extent obtains in Ruminants and Rodents. On the other hand, the superficial layer ought certainly not to be confounded with the thoracic portions of the panniculus; for neither in the direction of its fibres, at nearly right angles to this last, nor homologically traced in other Mammals, does it positively belong to it. For these reasons I shall therefore describe the pectoralis major in the Pilot Whale as a somewhat broad and rather triangularly shaped muscle, arising upon the superficies of the thorax from opposite the fourth to the first rib, meeting its fellow in the median line over the stemum®. Its fibres are directed outwards and slightly forwards, and are inserted by strong aponeurotic tendon into the breadth of the distal end of the humerus. The fibres of the panniculus, as already mentioned, intermingle with those of its posterior border, while anteriorly the representative of the platysma is similarly related. The description of pectoralis major in B. rostrata’ pretty closely agrees with the above; and in other Cetaceans examined by myself little difference obtained. If the pectoralis minor is not, as hinted at, a deep layer of the pectoralis major, then I may assume that in Glodiceps it has an origin from the ventral surface and median line of the thorax over the sternal cartilage of the fourth rib. Only of moderate breadth, but long, it is inserted into the strong fascia at the inner posterior root of the flipper. A muscle equivalent to the above existed in L. albifrons, but was attached as far as the sixth costal cartilage. In the Porpoise, Stannius® notes the pectoralis minor as short and fleshy, springing from the anterior border of the sternum, behind the ? D’Anat. Comp. 2nd ed. (1835), vol. i. p. 393. 2 Op. cit. p. 269. 3 «Tie Cetaceen,” p. 90. * «Muskeln des Tiimmlers,” 1. c. p. 16. * Op. cit. p. 89. ® Fréd. Cuvier speaks of it as the “ musculus communis,” Cyclop. of Anat. & Phys. vol. i. p. 571. 7 Carte and Macalister, p. 218. * ZL. ¢. p. 14 (Rapp, 89). 274 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. sterno-thyroid and close to the costo-humeralis, with an attachment to the coracoid process. Said to be absent in Balenoptera rostrata (1. ¢. p. 218). Rapp! names a costo-humeralis in Phocena. This is short, fleshy, and passes from the posterior border of the first rib-cartilage to the inner side of the humeral tubercle. It goes under the insertion of the scalenus anticus, where it mingles with fibres of the pectoralis minor. I did not observe this in Globiocephalus, though I have recorded its presence in Lagenorhynchus. Cephalo-humeral’ roundish and strong-bellied. Origin, paramastoid along with the sterno-mastoid; directed backwards and downwards parallel with the sterno-mastoid for part of its course, it is inserted by a strong tendon into the anterior surface of the head of the humerus. I recognize a diminutive levator anguli scapule*, with fascial attachments to the transverse process of the atlas, the neck generally, and to the anterjor angle of the shoulder-blade. It differs but triflingly in other Cete, though occasionally relatively stronger. A monogastric omo-hyoid obtains in Balenoptera, which if differentiated in other genera has been overlooked by me. Though wanting a clavicle, yet I consider there is a homologue of a levator claviculz in G. melas. This I found broad, flat, and chiefly composed of tendinous fibres, which radiately arise upon the supra- and infrascapular muscles, and, rather more muscularly, are inserted into the transverse process of the atlas. It lies between the parotid gland and the transversalis cervicis muscle (vide fig. 63, L.cl). There is but a single rhomboideus, as obtains in the Dolphin‘, Lagenorhynchus*, and B. rostrata®. Stannius’ mentions two, rhomboideus superior and rhomboideus inferior, in the Porpoise; but Meckel’s*, Flower’s®, and my own observations on Phocena agree in its being single. I noted in G. rissoanus” a rhomboideus capitis, or what, indeed, might be a trapezius. The serratus magnus has an attachment to the altoid transverse process, covers the side of the neck, and broadening posteriorly is fixed to the scapula. A narrow portion, moreover, digitally descends to the second and third ribs close to the cartilages; a broader costal and more aponeurotic portion is fixed to the ribs and intercostal spaces, 1 L.c. p. 99, and Stannius, 1. c. p. 16; also well shown in Flower’s drawings (infra) of P. communis, besides a pectoralis minor. ? The humero-mastoideus of Fréd. Cuvier (Cyclop. of Anat. & Physiol. vol. i. p. 571, fig. 256), the occipito- humeralis of Stannius (J. c. p. 15), and masto-humeral of Carte and Macalister (J. c. p. 219). 3 Rapp, p. 88; Stannius, p. 13. * Cuvier, Lecons, vol. i. p. 375. 5 Linn. Soe. Journ. vol. xi. p. 152. ° Carte and Macalister, Phil. Trans, 1868, p. 224. 7 Muller’s Archiv, 1849, p. 13. 5 Anat. Comp. vi. p. 240. ® Unpublished dissection, drawings of which were kindly lent me. 20 Journ. of Anat. 1870, p. 154. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 275 from the fourth backwards. Rapp’ and Stannius’ note a serratus anticus major in the Porpoise. The costal origin of the short, narrow, and thin latissimus dorsi in Globiceps is from the seventh, sixth, and fifth ribs and spaces, about their middle; and the insertion is on the inner side of the neck of the humerus by thick fleshy fibres. In the Piked Whale it appears to come by aponeurotic expansion from the dorsal and some of the lumbar spines. According to Stannius*, its costal attachment in the Porpoise is sixth to fourth ribs. I have noted in the same animal eighth to sixth; in the White-beaked Bottle- nose*, twelfth to sixth. The deltoid presents no feature of importance differing from that of Whales generally. Its scapular attachment might be mistaken for a supraspinatus, owing to the altered relation of the fleshy parts by absence of spine. According to my observation (vide fig. 70) the subscapularis covers the entire inner surface of the scapula with a capitular humeral insertion. Macalister’s® younger specimen showed eight tendinous intersec- tions, B. rostrata’ differing in this respect. We agree as to the capsule of the shoulder not being pierced by its tendon, and an absence of bursa. The supraspinatus in G. melas and L. albirostris answers the description of it in Phocena and Balenoptera as given by several authors. The infraspinatus fills the shallow scapular concavity behind the deltoid, its fibres running in an acute angle to those of the latter. In Globiceps, Risso’s Grampus, and the White-beaked Bottlenose, I have found but a single teres =teres major and teres minor. In one instance I met with duplicity of these muscles in the Porpoise, though Meckel’s*, Rapp’s*, Stannius’s”, and Flower’s dis- sections of this animal show it to be more commonly single. Fréd. Cuvier’s figure" of the shoulder-muscles of the Dolphin demonstrates a teres major and minor. But the former is evidently the infraspinatus, the latter the teres major, and his infraspinatus a portion of the deltoid. Macalister avers of the young (lobiocephalus, ‘“ there was no sign of a teres minor or teres major, which are present in Dolphins.” A teres major is recorded in the Piked Whale. The diminutive triceps has two heads of origin :—one, a narrow slip, from the neck and dorsal surface of the scapular over the teres, and which mingles with the panni- culus; the other, more tendinous, from the head of the humerus. In the same species (G. melas) Macalister'* met with only intersecting threads of fibrous tissue devoid of muscularity ; but he and Carte’ mention a tricipital division as obtaining in Balenoptera. Decidedly a triceps is present, but single, in Phocena, Grampus, and Lagenorhynchus. There is in Globiceps a single, well-developed and fleshy coraco-brachialis, which ’ Cetaceen, pp. 88, 89. 7 L.c. p. 13. 3 L.c. p. 224. * L.c. p. 16. ® L.c. p. 151, 6 P.Z.8. 1867, p. 481. 7 Phil. Trans. 1868, p. 226. 5 Op. cit. pp. 278 & 262. § Op. cit. p. 90. 0 Loe. cit. p. 14. " Cyclop. of Anat. & Physiol. vol. i. p. 571, fig. 256. 2 P. Z.8. 1867, p. 481. * P.Z.8. 1867, p. 481. 4 Trans. Roy. Soc. 1868, p. 227. VOL. VIlI.—PaRT Iv. February, 1873. 2k 276 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. Macalister suggests is the short variety of Wood’. In my specimen it arose from the coracoid process, passed downwards, overlapping partially the supraspinatus and sub- scapularis insertion. Strengthening the capsular ligament of the shoulder-joint, it is fixed into the inner process of the humerus. The remainder of the forearm and manus, on both its surfaces, was clad with glistening, flat, tendinous material, as delineated in figs. 63 & 70. This divided below, supplying the several digits. I did not detect muscular fasciculi proximally, though it is possible these may have existed. Certainly in this specimen, as in Macalister’s, distinct flexors and extensors of the limb were wanting, unless indeed the aponeurotic- like fasciculi represented them. In B. rostrata? and B. musculus* five competent observers record them. In these two Whalebone species an extensor communis, a flexor carpi radialis, a flexor ulnaris, a palmaris longus, a flexor sublimis, and a flexor profundus digitorum are severally mentioned. 2. Muscles acting on the Trunk and Tail—Although there is considerable difficulty in pointing out the lines of demarcation of the several muscles composing the erector spine, I shall nevertheless treat them seriatim. Lying upon the tips of the transverse processes of the dorsal vertebra, and covering the heads of the ribs, the sacro-lumbalis has muscular attachments to all the former, and tendino-fleshy slips to the whole of the latter. Continued posterior to the ribs, it ends pointedly at the side of the lumbo-caudal region, in a line with the genital fissure. The anterior continuation of the sacro-lumbalis has been termed cervicalis ascendens, or descendens in other Mammals. Here in the Caaing Whale a part of the former suffi- ciently merits the latter appellation. It is the more fleshy of the two. It has attach- ments to the transverse processes of the more or less coalesced cervical vertebra, and specially to the atlas; moreover it proceeds with the nuchal continuation of the longis- simus dorsi to the cranium, and is fastened partly by tendon to the upper part of the paramastoid. Longissimus dorsi and spinalis dorsi are most intimately bound up together in the dorsal region, forming a long but enormous fleshy mass, interwoven spinally and costally with tendinous fascia. That which may be considered equivalent to a trans- versalis cervicis commences by a short, strong tendon at the paramastoid. Immediately becoming fleshy and thick, it ascends posteriorly on the side of the neck to the anterior dorsal region, and is lost in the combined longissimus and spinalis dorsi. Where the body posteriorly begins to narrow, a division of the two latter is perceptible. Here- abouts a superficial tendon passes obliquely upwards and backwards from the outer longissimus to the inner spinalis. A little way behind, another bridge of two oblique tendons similarly crosses, and immediately posterior to this five more, which together unite into a strong cord, wrapt one within the other (vide diagrams figs. 59, 60, & 63). Meanwhile from each muscle there is continued posteriorly, quite to the end of the 1 Paper cited. ? Carte and Macalister, Memoir, p, 228; Perrin, P. Z.8. 1870, p. 814, figs. 1 & 2. 5 Flower, P. Z.S. 1865, p. 705; and Struthers, Journ. of Anat. 1871, p. 110. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 277 spinal column, a single, thick, massive tendon. Besides the foregoing, both longissimus and spinalis possess a deep series of long, narrow tendons, one to each vertebra, but mingled or bound together by interstitial fleshy fibres. It results that these dorsal muscles act upon every vertebra independently, whilst at the same time the motor power of the fibro-cartilaginous tail is derived from the lengthened and more powerful cords; for from these there extends backwards a firm, glistening fascia, spread over and incorporated with the deep tail-substance. ‘The latter, composed of closely packed parallel fasciculi, something of the consistence of the dorsal fin, traverse the thickness of the tail, and here and there have slits for the passage of the nutrient vessels. Supracaudal. The single muscle (or compound muscle, if so regarded) to which I give this appellation lies external to the last, along the narrow portion of the caudal vertebra, and on the upper side of the transverse process. Narrow in front, where fleshy, it widens somewhat and forms a tolerably thick fusiform belly, which again flattens and becomes tendinous. In its course it is attached partly to the vertebral bodies and partly to the transverse processes, sending off a special tendon to each of the latter. Posteriorly the flattened tendon lies against the sides of the bodies of the ter- minal vertebrae, and ultimately is lost in the general expansion of the upper surface of the tail-flukes. For further remarks, see infra. -Coming under the denomination of multifidus spine and rotatores spin, because of their position, origins, and insertions, are a great number of musculo-tendinous bundles, very apparent and well marked, but difficult individually to separate and define. These are still more numerous and closely packed together in Lagenorhynchus than in Globio- cephalus, in consequence of the number and approximation of the vertebre in the former. Stannius recognizes such a deep set of muscles in the Porpoise; and I can corroborate his observation in that genus. ‘Their general arrangement is by tendons from the dorsal metapophyses, and trending forwards and inwards are attached muscu- larly to the sides of the roots of the spinous process in advance of their origin. The most anterior one is fixed to the atlas. But there are besides a deeper layer of fascicles springing tendinously from the spines and dorsal arches, and these becoming fleshy are inserted into the transverse processes of the same vertebree—doubtless semispinales, as Stannius' names them in Phocena. He alludes, moreover, to another set of fasciculi, close to the last, and connected with the vertebral processes, but he has not named them. In the four-limbed mammals generally there are three, or at most four, muscles described as occupying the iliac region, viz. the psoas major, psoas minor, iliacus, and quadratus lumborum. But in Cetaceans, as most writers state, there is only one enor- mously large inferior lumbo-caudal muscle, which, at first sight, might be supposed either to represent the psoas magnus alone, or the psoas minor, iliacus, and quadratus lumborum incorporated along with it. Whatever relation exists, division at least is 1 « Muskeln des Tiimmlers,” Miill. Archiv, 1849, p. 30. 2R2 278 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. inappreciable in G. melas. This enormously developed sacro-coccygeus muscle is long and fusiform. On each side it occupies the lateral and inferior surfaces of the vertebre and their transverse processes from the ninth dorsal vertebra backwards; and as the transverse processes of the caudal elements are lost, it still continues upon them in the shape of a bundle of tendons continued on to the very end of the spinal column. ‘The volume of its solid fleshy fibre may best be comprehended in the fact that it ranges in our specimen of Globiocephalus from one foot to six inches in transverse diameter, and with a corresponding thickness or depth. Further to particularize attachments and relations—it passes beneath the diaphragm, has the kidneys &c. lying upon it, and narrowing behind the rectum sends off, downwards and backwards, superficially, a series of flat tendons. ‘These are so connected together as to constitute a very strong tendino- aponeurotic sheath, which spreads out and is continued on to the inferior surface of the broad fibrous tail. The main body of the fleshy mass meanwhile terminates in a single strong tendon, which passes direct along the spine, and is fixed to the very last vertebra. Moreover there is an appreciable flat layer of fleshy fibres, which come from the sides of the vertebra and spread over part of the aforesaid tendinous sheath. This muscular layer appears to be a kind of reduplication of the body of the muscle itself. A muscle the exact counterpart of the supracaudal lies on the underside of the transverse processes of the caudal vertebrae, and it bears the same relation to the sacro- coccygeus that the supracaudal does to the longissimus dorsi, save the fact of inversion of position. I distinguish it as the infracaudal. The long spinal muscles of Cetacea have received different names and significations from successive anatomists, though the descriptions, save that of Stannius, tally. Meckel’ demonstrates the parts in the Narwal (J/onodon monoceros) and the Dolphin (Phocena communis’). His text appears to me to imply that he considers present and less or more differentiated :—1, an equivalent of the spinalis dorsi, biventer cervicis, and complexus, a longissimus dorsi, trachelo-mastoid, and splenius capitis; 2, a sacro- lumbalis, with cervicalis ascendens anteriorly (“ trachélo-mastoidien, ou lintertransver- ? saire du cou” of his translators); 3, flexor caude lateralis; 4, depressor caudee =qua- dratus lumborum, psoas, and iliacus; 5, an inferior depressor caude. Frederick Cuvier’ speaks of a levator caude, evidently No. 3 above. Rapp* and Stannius* coincide that there obtains:—a splenius capitis, longissimus and spinalis dorsi, sacro-lumbalis, and transversarius superior and inferior. ‘The former thinks the great lower loin- muscle a psoas major; to the latter it implies more. Stannius, moreover, describes a caudalis superior, a caudalis inferior, a longissimus inferior, a sacro-lumbalis inferior, and a set of caudal muscles unnamed by him. He also traces the short, deep spinal muscles, of which more hereafter. Carte and Macalister, in the Piked Whale®, have * Anat. Comp. vol. vi. p. 128 et seq. ? Art. Cetacea, Cyclop. of Anat. and Physiol. vol. i. p. 569. * Op. cit. p. 83 et seq. * Mull. Archiy, 1849, pp. 22-32. * Loc. cit. p. 22d. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 279 noticed a trachelo-mastoid, a longissimus dorsi, a sacro-lumbalis, with a slip supposed to be the homologue of splenius capitis, a levator caudz, a depressor caude major, and depressor caudz minor. Notwithstanding amplitude of nomenclature, and recognition of two or more en masse or separate, the anterior divisions of the various observers present a certain harmony ; but there is less concord of opinion regarding the posterior tendinous parts and infero- lumbar region. Rapp and Stannius differentiate as transversarius superior the com- pound tendinous enwrapping sheath of the longissimus and spinalis as described by me. But the latter, moreover, unites it with the anterior fleshy belly of my supracaudal, and iraces it forwards to the ribs, thorax, and neck, 7. e. includes part of what more strictly is sacro-lumbalis and cervicalis ascendens. Carte and Macalister’s levator caude agrees partially with Rapp’s transversarius, and partially with Stannius’s caudalis superior. The latter muscle, again, is equivalent to Meckel’s flexor caudz lateralis and F. Cuvier’s levator caudz, one and the same with my supracaudal. None suggest the superior superficial terminal tendons, or aggregate fibrous investing-sheaths of the longissimus and spinalis dorsi, as the homologues of the levatores caude externus and internus of other mammals. Yet in every sense they are undoubtedly such, continuity with the dorsal fleshy masses being the only special deviation from their usual con- dition. The Cetacean supracaudal, again, offers homology in its posterior short slips with the intertransversarii caude of quadrupeds, its longer-bellied and more fleshy anterior moiety being occasionally in mammals almost separate from the intertrans- versarii caude, though not specially recognized as a distinct muscle. In Manatus, however, it is uncommonly well developed, and has been named by me lumbo-caudalis. The inferior depressor caude of Meckel, depressor caudee minor of Carte and Macalister, caudalis inferior of Stannius, and his unnamed musculo-tendinous caudal bundles, cor- respond with the present infracaudal. As regards the depressor caudee of Cuvier and Meckel, the depressor caude major of Carte and Macalister, this undoubtedly is Rapp’s psoas major &c. Stannius viewed it as composed of three divisions, equivalent to the dorsal muscles, and named by him respectively longissimus inferior, sacro-lumbalis inferior, and transversarius inferior. So far I agree with the latter, and therefore differ from Rapp, that the great sublumbo- caudal Cetacean muscle is not purely an ilio-psoas. This latter, I believe, as in Manatus, is all but aborted, certainly not recognizable. The homologue of the Cetacean sublumbar muscle, then, with its tendons and investing sheath, seems to me to be the sacro-coccygeus, whatever its significance as to the dorsal series. My infra- caudal may represent partly inferior intertransversarii caude or perhaps include infra- coccygeus. In default of being unable to determine with accuracy spinal insertions in Globiceps, I was more fortunate in Zagenorhynchus. In this genus the rectus abdominis tapers to a point at the fortieth vertebra, behind this intermingling with the caudal fascia. The 280 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. pubo-coccygeus goes to the chevron bones as far as the sixtieth vertebra. Sacro- coccygeus, muscular to forty-fifth, tendons to sixtieth; between these points the secon- dary tendons which form the sheath, emerge. Supracaudal from fortieth to sixty-sixth vertebra ; the infracaudal is from two to three vertebre shorter. Longissimus dorsi &c. narrows at sixtieth; two oblique tendons given off at thirty-seventh ; the others behind, ere producing aponeurotic sheath. The spinalis dorsi &c., its final tendons inserted from the sixty-fourth to the seventieth vertebral diapophyses. Fig. 6. Spd+Lei Ida+Lce Manner in which the fleshy imbedded inner deep tendons of the longissimus, &c., are distributed to the spine in Cetacea. Spd+Lei, conjoined spinalis dorsi and levator caudx internus, hooked upwards just posterior to the oblique bridging tendons from (d+ Lce) the longissimus dorsi and levator caudx externus; Se, sacro-coccygeus, part of its constituent subsheathing tendon also dragged apart. Sketched from a dissection of Phocena communis. A series of levatores costarum, of moderate strength, and passing from the transverse processes to the ribs, exists in all the species of Whales I have dissected’. In the lumbar region of G. melas the intertransversales* are powerful; they diminish in strength forwards, and can barely be detected in the most anterior dorsals and cervicals. In ZL. albirostris, whilst fleshy, they are shorter, owing to the close approxi- mation of the very numerous and long divergent transverse processes. In P. communis caudally they are tendinous; in the lumbar region, semitendinous and fleshy, a superior and inferior division is noticeable. According to the development of the neural spines, cervical, dorsal, lumbar, and caudal, so are the interspinales* strong or weak. But as a series of muscular bundles they are, I believe, present in every Cetacean. They have been met with by me in five genera. Both Rapp* and Stannius’ have described in the Porpoise a set of muscles linking together the chevron bones. They name these M. interspinales inferiores. They are ‘ Described also by the oft-quoted German authorities. ? The intertransversarii of the foregoing. * The m. interspinales superiores of the preceding writers. * Op. cit. p. 83. * L.c. p. 40. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 281 distinctly marked in Glodiceps, Grampus, and the White-beaked Bottlenose and Rorqual. They undoubtedly resemble the interspinales superiores of these authors, but pass from one chevron hemo-spinal element to the adjoining. I prefer to designate them as interhzemo-spinales, this term being more in accordance with morphological anatomy. Stannius likewise differentiates, and names as M. interaccessorii a number of tendino- fleshy fascicles which intervene between the one and the other accessory spinous processes of the lumbar and dorsal vertebrz, in a longitudinal direction. These have not been observed by me; but I am inclined to regard them as intermetapophysales. 3. Muscles acting on the Sterno-costal Framework.—The rectus abdominis in the “Deductor” is striking on account of its vast thickness in the middle of the thoracic region. Springing by a strong aponeurotic tendon from the manubrium, becoming fleshy opposite the hinder border of the pectoral limb, it spreads out and covers a con- siderable share of the hinder surface of the chest. By degrees narrowing, it tapers and apparently ends near the generative aperture, though a portion is continued on towards the innominate bone, and by a tendon is fastened to the infracaudal fascia. The attachments of the rectus in the Porpoise are nearly similar. Stannius' notes a connexion with a tolerably strong aponeurosis, which proceeds from the pelvic region to the transverse process of the nineteenth lumbar vertebra and general caudal fascia investment. In the Piked Whale’ penniform insertions into the second to sixth ribs are recorded. In Risso’s Grampus°* the muscular belly stops at the second rib, aponeurosis continuing to first rib and sternum; posteriorly it agrees with that of G. melas. The obliquus externus, internus, and transversalis are relatively weaker than the preceding, though each is a great fleshy sheet. The former is attached to all the ribs about their middle, save the three anterior ones, where the fibres spring near their sternal ends. There are a few digitations with the serratus magnus; posteriorly the external oblique does not reach the pelvic bones. Both external and internal intercostals seem to possess considerable motor power over the ribs, judging from their full development. More particularly is this the case with the first external muscle, duly commented on by Stannius in the Porpoise. I also specially noted the presence of strong fleshy fibres corresponding to the triangularis sterni in G. melas (vide fig. 69) and L. albirostris. Under the head of musculi ossium sterni costalium, Stannius* describes bundles in the Porpoise which appear to me identical with the triangularis. The Rostock professor likewise alludes to a sterno-costalis® in the same animal. This double slip, with obliquely set fibres, has one part attached to the sternum and first rib and its cartilage, another between the first and second cost and sternal cartilages. I presume it agrees with what is now better known as the supracostal. Meckel® describes, but does not name, a muscle identical with the above in Cetacea, which he regards as assisting respiration in the diving animals. elaceups LO: * Phil. Trans. 1868, p. 224. ? Journ. of Anat. vol. v. p. 134. SUT yey * L.c. p. 37. § L.c. p. 190. 282 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. I distinguished two scaleni, a scalenus anticus and scalenus posticus. The former springs from the first rib and its sternal cartilage, where it is fleshy and broad ; passing forwards it was inserted by a powerful tendon into the basiocciput, outside the rectus anticus major. The latter has a first costal origin, but not quite to the cartilage; anteriorly it is attached to the transverse process of the axis. I found precisely the same arrangement in Lagenorhynchus, both divisions being strong. Macalister records in G. svineval’ a scalenus anticus attached to first rib and upper cervical vertebre, and a scalenus medius and scalenus posticus conjoined, from first and second ribs to upper cervical transverse and spinous processes. But he suggests that one of the latter “* was probably the germ of the serratus posticus superior, which otherwise was not visible.” He and his colleague mention only a scalenus anticus in Balewnoptera. Meckel?, Rapp, and Stannius agree in there being two scaleni in Phocwna. Sterno-mastoid, origin inner end of manubrium, outside the large sterno-thyroid ; insertion by a strong tendon into the paramastoid along with the cephalo-humeral. Although in Glodiceps it appears but a single muscle, yet there is a tendency to duplicity, inasmuch as the anterior portion rolls round posteriorly, and with what seems almost a separate deep tendon fixed to the manubrium. A double head, viz. from the sternum and cartilages of two coste, obtains in the Piked Whale*. Cuvier* and Meckel? allude also to a cleido-mastoid in the Porpoise. 4, Muscles connected with Neck and Head.—What doubtless answers to the splenius, although it may include complexus, I find, in the Caaing Whale, to be a muscle of a most powerful character and of enormous size. The dorsal attachment is from the eleventh or twelfth vertebra ; continuing thick and fleshy and widening, the muscles of opposite sides are fastened cranially the whole breadth of the exoccipitals. In the Porpoise® it springs aponeurotically from the first dorsal spine, and terminates in a squamal tendon. A trachelo-mastoid, according to my reading, obtains in a narrow longitudinal bundle arising from the transverse processes of the four anterior dorsals, and, running for- wards, is fused anteriorly with the short oblique. Macalister gives the first cervical to the junction of the exoccipital and paramastoid as its attachments in his specimen ot Globiocephalus’. In Balenoptera it is double-headed; one from the first dorsal, the other from the sides of three or four posterior cervicals, a vascular plexus divides these ; cranially inserted into masto-squamoid. Although Meckel names such a Cetacean muscle, he confounds it with splenius capitis. Not mentioned by Rapp and Stannius. The ordinarily small-sized, short, deep muscles of the neck in this Whale are of inordinate proportions, save the longus colli. The two pairs of recti postici are consi- derably interwoven. Occipitally they cover the whole surface of the bone below the inferior curved line, and thence extend to atlas® and axis. Stannius® unites under one ' P. Z. 8. 1867, p. 481, and woodcut, p. 478. 2 Anat. Comp. vi. p. 158. 3 ZL. c. p. 218. * Legons, vol. i. p. 259. 5 Op. cit. p. 161. « Splenius capitis,” Stannius, J. c. p. 21. 7 I. ¢. p. 481, and woodcut. 8 Macalister, P. Z. S. 1867, p. 481. 9 Lc ps 208 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 283 heading the semispinalis and rectus capitis posticus in the Porpoise, the former stretching from the tenth dorsal spine to the atlas; the latter (a bundle derived therefrom) goes to the occiput. An obliquus superior springs from the lateral atloid bony ring in Globiceps as far as the root of the transverse process, and goes to the exoccipital. Beneath is another muscle from the axial process to the skull, apparently an obliquus inferior. Ventrally placed to the last, with similar attachments, is what I presume to be a rectus lateralis. Rectus capitis anticus minor, broader than the last, lies mesially to it, and is overlain by a great rectus capitis anticus major. ‘This latter is fastened to the basi- occciput, to the ventral surface of conjoined atlas and axis, and partly cervicals beyond. The short and highly tendinous longus colli is forwardly fastened to the body of the axis, and, covering the posterior cervicals, goes backwards on the dorsal bodies as far as opposite the fifth rib. In the Porpoise it stretches over three dorsal bodies and all the cervicals'. The great development of these muscles, in spite of shortness and coa- lescence of the neck-bones in Cetacea, is an interesting fact. Masseter short, narrow, but thick. Origin, angle of mandible; insertion, area below the eye; its fibres have but slight downward obliquity. Above it and somewhat superficial, observed by me both in Globiocephalus and Lagenorhynchus, was another muscular plane, relatively sparse in fibre, which occupied the interspace between the last and the orbicularis palpebrarum. This latter, I opine, corresponds to the masseter malaris externus of Professor Stannius* in the Porpoise. He describes it as springing from the malar process (Jochfortsatze), sloping forwards to the outer angle of the orbit (Augenwinkel), and spreading out over fasciculi of the masseter orbicularis palpebrarum. A clearly double masseter is defined in Balewnoptera’, the deeper layer arising tendinous from the margin of the glenoid cavity. The narrow, fleshy, and thick temporalis fills the somewhat infundibular fossa back to the crescentic ridge, and has an insertion into the mandibular coronoid process as far as its root. There is a well-developed buccinator, whose posterior moiety is thick. Its broad maxillary origin is from the bony area anteriorly and inferiorly to the eye; its narrower insertion, the muscle being of an elongate triangular form, is the mandibular surface opposite the hindermost teeth. Weak muscular fibres, representing an orbicularis oris, appear to be derived from, and almost are a continuation of the upper part of the buccinator. They thus proceed from the angle of the mouth and buccinator in a scattered manner, and curve forwards and upwards, thence partly interweave with the strong glistening tendino-fibrous material which goes to form the overhanging buccal arch or projecting upper lip, and merge into the blubber-like fat and the premaxillary muscle. The single-bellied repre- sentative of digastric (antea, p. 251) is little else than an inferior facial muscle. * Rectus anticus major bound up with the longus colli in Phocena, according to Rapp, J. c. p. 82, and Stannius, 1. c. p. 30. ? L.c. p. 4. * Carte and Macalister, /. c. p. 223. VOL. Vill.—PaRT Iv. February, 1873. 2s 284 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. In all the Cetaceans cut up by me I have observed well-developed and separate pterygoidei. The external, flat, broad, fleshy, and of a quadrilateral shape, is fixed to the outer surface of the pterygoid plate, and, with a downward and forward plane, goes to the inner surface of the lower jaw, chiefly to the upper margin of the bone. The internal muscle arises from the superficies of the prominent portion of the pterygo- palatine, passes backwards and downwards to the mandible. Inferiorly and at its posterior border the latter muscle sends off tendinous fibres, which join those of the articulating condylar process of the inferior maxilla. A pterygoid internus and externus are described both by Rapp! and Stannius? in the Porpoise; but Carte and Macalister® state that in the Piked Whale no internal pterygoid was found. TX. URtNo-GENERATIVE ORGANS AND PELVIC-APPENDAGE HOMOLOGUES. 1. Genitalia.—The kidneys, as in other Whales, are conglomerate, and lie upon the great lumbo-caudal muscle. In this female the renal organ (fig. 36) upon the right side measured one foot long by about six inches across at its widest, and was of a longish oval figure. The left one was about an inch smaller, and rather narrower in propor- tion. ‘The lobuli vary in size and shape, characteristically tending to the hexagonal form; their diameter ranges from 0°3 to 0°6 of an inch. They are further aggregated together in groups of from three to a dozen in number, and, when the kidney as a whole is slightly unravelled from the loosely connected but strong fibrous tissue, have an appearance not unlike the acini of a bunch of grapes. Each lobule (fig. 38) is composed of a thickish cortical layer and smooth tubular or medullary substance, which has a single broad hilus and papille, but no marked Malpighian pyramids’. As Mr. Gulliver® has remarked, “the bloodvessels enter the kidney, not near its middle, but at its fore and inner or atlanto-mesial side.” The renal artery has a nearly straight course in the longitudinal direction of the gland, and sends off branches dichotomously. The blood is returned by the veins in a similarly dichotomous manner, which enter two large branches placed on each side of the artery; these join into a main vessel which lies to the outer side of the artery. The ureters leave the kidney posterior to the entrance of the bloodvessels, and pass on to the urinary bladder somewhat apart and behind the neck. The bladder is small and pyriform, in this case containing a limited amount of dark coloured foetid-smelling urine’. The urethra, 7 inches in length and posteriorly not well defined from the neck of the bladder, is adherent to the abdominal parietes of the vagina, and runs forwards 1 Op. cit. p. 85. ? Zc. p. 6. 8 ZL. c. p. 222. 4 Turner, who made an injection, says (J.c. p.75) that ‘ the pyramidal substance was almost non-vascular.” 5 Loc. cit. p. 65; also noticed by Turner. 6 Turner, ibid., suggests that only “a small quantity of urine could collect in that organ prior to its excretion.” DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 285 narrowing by degrees to the meatus urinarius, which opens at the deep surface of the clitoris. The female external generative aperture in Globiceps, as in all true Cetaceans, is an elliptical fissure or median sulcus situate at the hinder end of the body, where com- mences the caudal narrowing, or near the middle of the third fourth of the animal’s length, and inferiorly about vertical to the termination of the panniculus. The tegument around is dark-coloured, and thrown into a good number of minute parallel and wavy wrinklings chiefly transverse in direction. The opening of the vulva is 3-3 inches long, and it is less than a third of that across at its widest part. The anal orifice, not over half an inch in diameter, is distant 2°7 inches behind the posterior pudendal commissure. Immediately within each lip of the vulva, or what may be con- sidered the labia majora, are some twenty or more short, but deep, folds of membrane, in the recesses of which are crypts, the openings of sebaceous glands. The homologues of nymphe or labia minora are two prominent folds of the mucous membrane, each an inch in length, which lie within the anterior pudendal commissure, and slightly converge as they pass backwards. Between these and with lateral plicate sulci is a median ridge 1:3 inch long, which ends in a small but distinctly pronounced conical-shaped clitoris. The vagina, 9 inches in length, whose mouth is about the middle of the external uro-genital fissure, sweeps diagonally towards the abdominal cavity. In the two latter points, as in the uterus entire, Glodiceps agrees pretty well with the Whale type as described by John Hunter and others in specific forms. The lower vaginal half is widish and smooth, or with only fine longitudinal plications; the upper half, on the contrary, is narrower, and has a very uneven surface. This roughening depends on a numerous series of transverse ruge or puckerings of the membrane, some four of which are extremely prominent. In alluding to these valvular folds, Hunter aptly compares them to a succession of ores tincarum. ‘They are composed of thick induplications of the fibroid tissue of the wall of the vagina (fig. 74), inwardly lined by narrow longitudinal mucous rugz, which fringe their free edges. The fold nearest to the os, and only a good thumb’s breadth from it, has a thickness of 345 of an inch. The true os uteri is only distinguishable from the preceding folds by its narrower and somewhat firmer ring-like aperture. The cavity of the uterus above in this specimen is only 2 inches in length, and then divides into right and left cornua. There is a circular membranous fold about half an inch above the os; but the mucous coat of uterus and cornua is longitudinal and wavy. The broad ligament and the fimbrie of the Fallopian tubes form a delicate arched covering or pavilion which overarches the ovary. Each narrow oblong ovarian body is about two inches in extreme diameter, and is of firm consistence. The mammary glands, as might be expected in a young animal, were but of moderate size, namely, about 4 inches long, and their glandular structure in consequence feebly developed. ‘These organs, as in other Cetacea, lie upon each side of the vulva, and 282 286 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. open outwardly by a small slit containing a retracted nipple, in this case apparent, but only imperfectly pronounced. The glands themselves have a tough fibrous external coat, and are surrounded by a strong constrictor muscle. The male organs of generation of Globiceps' have been described by Jackson; they resemble those of the Dolphin. He says, ‘connected with the vasa deferentia, and forming undoubtedly a part of the genital system, was a canal, 23 inches long, opening on the verumontanum, just in front of the vasa, and so closely connected with them as not distinctly to appear, except for about one third of an inch just before the vasa came together. At this last part, where it terminated in a cul-de-sac, it appeared as large as the vasa, but was thinner, and the cavity larger, the inner surface being white and rugous. It afterwards became smaller; but before termination the cavity enlarged to five or six lines. No glands were seen on the inner surface”*. An organ resembling the above was found by the same author in the Sperm-Whale, the Dolphin, and the Porpoise*. I take it to be a uterus masculinus; and as I have myself seen it in two genera, one may infer its presence in most, if not in all the genera of the family (see woodcut antea, fig. 3). No sectional representation of the Cetacean penis, to my knowledge, is published, although the structures themselves have long ago received elucidation by Hunter* &c. I incorporate in the Plates of this memoir, therefore, a drawing (fig. 56) of that of a large Fin-Whale’®, typical in part of what obtains in the genera of smaller size. 2. Pelvic Bones, Ligament, and Muscles—TYhe pelvis in Globiocephalus is represented by a pair of short styliform bones (ossa innominata), which, at a moderate distance from each other, float free from the spine among the muscular and other tissues near the genital outlet. In this female each of these smooth tapering bones measured 4 inches in length, had a strong resemblance, in shape, to a first rib, concavity inwards, or looking towards its fellow. Anteriorly there was an intumed capitulum, as in Orca, just behind the neck of which the bone enlarged somewhat, and then tapered gently towards the narrow and pointed hind extremity. At the thickest belly-part of the bone the diameter was 3% inch, proportionally less in front and behind. There was a 1 In Dr. Williams’s account of the generative parts of the Lewchew Whale he says, “ there was an anomalous connexion between the urethra and rectum, near their termination, of a firm, round, imperforate band of a tendinous nature, half an inch in diameter, the use of which was not at all apparent,” l. c. p. 412. Does he allude to the retractores? or the male parts representative of the mammary glands? 2 L. c. p. 166. > See also a notice of the presence of a uterus masculinus in Phocena by Huxley, Hunt. Lect. and Lancet, 1866, p. 381. 4 Phil. Trans. 1787, p. 441, and his original preparation, Physiol. Cat. Coll. Surg. no. 2527. 5 Physalus antiquorum=B. musculus, specimen no. 2524 x, Coll. Surg. Mus., and same referred to in P. Z.S. 1865, p. 215. I hazily remember having seen a lithograph of a view similar to that here given in Dupuytren’s Museum in Paris. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 287 nodule of cartilage tipping either end, that rearwards was very distinct. No osseous or ligamentous cross bar, representing rudimentary femur, existed, such as is met with in Balena and Balenoptera. The smooth or only slightly roughened bone rather resembles the pelvis of the northern Orca. Uniting the opposite pelvic elements was a strong x-shaped glistening fascia, fixed within a trifle of the whole length of the inner concave border of each bone. A number of muscles were attached, partly to this and to the bones themselves, and bloodvessels penetrated the two latter, which I have already referred to. The muscular structures attached to the pelvic bones, and surrounding the generative outlet of Cetaceans, have heretofore been but imperfectly studied and figured. The pains-taking Stannius, in his “‘ Myology of the Porpoise”’, admits of anal and pelvic muscles the following:—a sphincter ani, an ischio-cayernosus, and a retractor ischi or ischio-caudalis. I believe no other author has stated there are more, though others are inferred to exist”. In my researches on this female Deductor and other Whales, I have been fortunate in meeting with many more than the above; and these I shall describe in layers, from the superficial to the deep, as dissected (vide figs. 76 & 78 in my last Plate). In a study of the parts it is well to keep in mind the sex. I have already (antea, p. 272) alluded to the inferior and posterior (pudendal) portion of the panniculus carnosus, shearing outwards and leaving the tail free; but there is, at the same time, a certain continuity with it of a thin superficial layer of transverse fibres. These of opposite sides form a long ellipse, whose apex anteriorly meets the divergent angle of the panniculi, partially cover the rectus abdominis and other genital and mammary gland-muscles, whilst posteriorly it reaches the perineal raphe and anal sphincter. It acts as a dilator of the vulva, whilst the portion continued to the anus serves as an anal protractor. ‘This muscular sheet altogether appears to be the homo- logue of levator ani and probably combined superficial transversus perinei of human anatomy. I may as well mention that in the perineal region of Risso’s Grampus, beneath the cross fibres of the levator ani, I observed a narrow oblique band of muscle which passed from the surface of the pubo-ilio-coccygeus to the perineal raphe. This slip doubtless would represent a superficial transversus perinei. Moreover, in the same female specimen, I noticed another short, broadish, fleshy band, which ran obliquely from the surface of the rectus abdominis, at the commencement of its posterior tendon, inwards towards the fore part and side of the clitoris. Its superficial position and inguinal relations suggest its being a homologue of a cremaster, though such an anomaly in the female sounds strange. 1 Miiller’s Archiv, 1849. 2 Professor Van Beneden has been kind enough to put in my hands a short paper of his, “ De la Com- position du Bassin des Cétacés,” Bull. 2 ser. t. xxv. Therein he faithfully figures and describes the generative organs of Delphinus tursio, noting two muscles, viz. ischio-cayernosus and retractor of penis. 288 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. The sphincter ani consists of a circlet of coarse, strong, fleshy fibres, the true external sphincter; but there are, besides, circular fibres more deeply placed around the gut, which would represent a sphincter ani internus (vide fig. 48). The rectus abdominis, which I have already described, partly mingles with the generative muscles, inasmuch as its posterior narrowed extremity and terminal tendon enclasp the deeper fleshy structures of the vulva and winds round each innominate bone, finally being inserted in the neighbourhood of the chevron bones. What I assume to be the counterpart of an ischio-coccygeus is a thick and strong muscle which, with its fellow of the opposite moiety, constitutes a crescent-shaped perineal mass. Each muscle arises by narrow but somewhat strong tendon from the surface of the middle of the os innominatum. Passing backwards and inwards, its fleshy fibres expand and are inserted into the perineal raphe, and join those of the levator and sphincter ani. It is a retractor of the pelvic bone when the perineal attachment is the fixed point; reversely it may assist the levator ani. There is a likelihood also that after pregnancy, when the mammary gland is enlarged, it may exert pressure on that organ during lactation. The erector clitoridis, or ischio-cavernosus, is a large diagonally placed fleshy band of muscle, of nearly equal breadth throughout. It lies partially beneath the last and between it and the rectus abdominis, with the compressor mamme and gland to its inner sides. Before the positions of the parts are disturbed this muscle has a more cur- vilinear aspect than the deeper dissections (figs. 77, 78) warrant. It has an extensive origin from the surface of the posterior moiety of the immominatum bone, at least 2 inches, as also from the fibro-cartilage and broad pelvic fascia. With a direction forwards and inwards, it lies upon the sphincter vaginee, and is ultimately inserted into the median line. In male Cetaceans the muscle corresponding to this is a very powerful one, with a great pelvic attachment, and stretching on to the side of the root of the penis. Alongside and mesially from the erector clitoridis is a somewhat fusiform, moderate- sized, or, indeed, compared with the last, small-sized muscle. Its office I take to be chiefly a compressor of the mammary gland, as such I name it ; for it constitutes a semi- circle round and partially over that organ. Its attachment posteriorly is beneath the ischio-coccygeus, from the strong perineal fascia in front of the rectum. From this the muscle, sweeping outwards, forwards, and then inwards, or embracing the mammary glands as a semilune, narrows anteriorly, and is fixed upon the strong fascia before and partially at the side of the mamme’. Besides the two foregoing, two other muscles lie between them and the border of the rectus abdominis. ‘The first, in proximity to the erector clitoridis, is a short and laterally compressed slip. Inwardly it is fastened to the median strong fascia of the 1 Geoff. St.-Hilaire, ‘ Annales,’ 1834, p. 183, alludes to a muscular covering of the mamme, which, he thinks, presses upon the gland like the pouch of the Kangaroo. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 289 vagina, and thence directed outwards and slightly backwards is attached to the fascia in front of the os innominatum. ‘The second fleshy band is situate immediately in front of the last, the two being in close union. Anteriorly it has the same derivation, but posteriorly proceeds along the outer margin of the os innominatum almost to the middle. This pair of muscles serves much the same office, viz. seem to draw inwards, or are protractors of the pelvis, and, through continuity by fascia across the urethra, may constrict this canal. As regards their homology, they bring to mind the two portions of Wilson’s and Guthrie’s muscles, the so-called compressores urethre, or transverse and perpendicular constrictores urethre. Possibly they may include the circular fibres of Santorini, =stratum internum circulare of Miiller, if we take into account along with them their fellows of the opposite side. In studying the parts, I imagined for a time that the larger of the said muscles might be the representative of a pyramidalis. On the other hand there is more reason to believe them very fully developed urethral constrictores. Hence it is most interesting to find in Cetacea parts so manifest which n Man are mere rudimentary structures. Deeper than the four preceding there is another muscle, which, without doubt, is a true sphincter vaginee, =bulbo-cavernosus in the male. It is exceedingly large, fleshy, and powerful. It is attached to the vagina near the anterior end of the pelvic bone, and, as a broad fleshy mass, encircles the generative canal, beneath the mammary gland. Broadest in front, as it narrows behind it becomes inserted on the strong median fascia posterior to the vulva. I may next call attention to a small slip of muscle which bridges between the pos- terior corner of the os innominatum and the recto-pelvic fascia. Springing from the inner deep margin of the bone, it crosses inwards and slightly forwards towards its narrower insertion. It lies upon the broad belly of the combined lumbo- and ilio- coccygeus, and is covered by the erector clitoridis and compressor mamme. ‘The action of this diminutive muscle evidently is to pull each pelvic bone inwards from behind or, contrariwise, to render tense the pelvic fascia. It appears to me to be the homologue of a deep transversus perinel. Finally, I shall refer to a long, elliptical, fleshy and powerful perineo-caudal muscle. This in Globiocephalus, as in Grampus (?), is double-headed; or if considered two muscles, these are conjoined behind. It or they best comply with the pubo- and ilio- coccygeus of other mammals, which I am convinced is the true homology. If I under- stand aright, Stannius terms it retractor ischi in Phocena, and Rapp ischio-caudalis. In the female Caaing Whale the narrow outer portion springs from the outer border of the posterior third of the os innominatum, thence goes backwards as a fleshy riband to the outside of the next. This much more bulky but also fleshy portion commences from the under surface of the deep pelvic fascia, and trends straight backwards along the middle of the narrowing portion of the caudal region. It has the rectum between 290 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. it and its fellow of the other side; but posteriorly the two meet mesially, and as they diminish in calibre are fixed to the chevron bones as far as the vertebre. X. REFLECTIONS, ZOOLOGICAL AND PHYSIOLOGICAL. The skeleton, as I have intimated, is the standard whereby Cetacean generic segre- gation and affinities are best tested; consequently such parts of the organization of Globiceps as have been here treated, afford narrow limits in their applicability to taxonomy. One’s eyes, however, must not be shut to the fact that among the Del- phinide Glodiocephalus seems to be established generically by its outward configuration alone, unless Dr. Gray’s Sphwrocephalus' assimilates; but the exterior aspect of this form is unknown. The truncate, globose, prenarial nodosity, fairly defined from the upper labial rim, the unusually long, narrow, tenuous pectoral limbs, and the low, falcate dorsal fin are very marked features in the genus under present consideration. The head of Orca and its ally Pseudorca, whilst rounded, is proportionally low; and their limbs and dorsal appendage sensibly differ. Grampus approaches our genus; but the facial prominence veers towards the outline of Phocena, or with a gradual arched declivity, rather than the bold perpendicular fulness and labial emargination of Globio- cephalus; besides, its dorsal fin is relatively high, and its limbs shorter and broader. Phocena and the Dolphins (Delphinus &c.) are yet further removed in the points above indicated. Such differences, I grant, are gradational; but so are all osteological data. Within the limits of just comparison, however, they are not fanciful, but visually true, and have their special worth in the outward generic recognition of a group uncommonly like each other in their tout ensemble. Eight species of Globiocephalus are recognized in the British-Museum Catalogue, and other synonyms &c. enumerated. My worthy friend Dr. Gray’s method of recording scattered writers’ indifferently determined species is decidedly useful, but not without objections. He himself allows that such handy lists require constant supervision. His G. affinis, he surmises, is probably a young G. svineval (=G. melas). This opinion one can readily admit; for, unless in non-agreement of dental formula, there is no special characteristic to assign a separation. Now the numbers of the teeth in different specimens of Globiceps are most irregular, simply from the reason that they are so loosely implanted in their sockets that in early life, adolescence, and old age they not unfrequently drop out. But what is the Grampus affinis, p. 300, Cat.? The presumed anatomical distinctions of the American Globiocephalus intermedius rest on treacherous footing. To it fifty-eight vertebre are assigned against fifty-five to the European form ; yet in G. melas there obtain fifty-eight and fifty-nine, as authenticated by Professor Flower (Trans. Zool. Soc. vi. p. 349). Too much stress, I think, is laid on the cervical ossification of the former; and as to difference in dentition, the above remarks apply. G. edwardsii has no solid basis as a species. Regarding the South-Sea Blackfish + P.Z.S. 1864, p. 244, and Cat. of Seals and Whales, 1866, p. 323. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 291 (G. macrorhynchus), the G. indicus, the G. sieboldii, and G. chinensis, some confusion evidently reigns. Dr. Gray says Mr. Blyth’s species (G. indicus) is “perhaps a Neomeris.” But his own trenchant definition of the latter genus, viz. “dorsal fin none,” does not harmonize with Blyth’s recorded dorsal fin 2} feet in length. I have alluded frequently to the anatomy of Dr. Williams’s Lewchew specimen, which agrees with the British G@. melas; yet it is as likely as not to be the G. indicus, Blyth. As far as I can draw a conclusion from what has to me been satisfactorily demon- strated by others concerning Globiceps, there is but one (or two?) well-determined species inhabiting the basin of the Atlantic’. This is manifestly a migratory animal. That in the Pacific and Indian Oceans the species are more numerous is possible; but how much separation or identity of specimens from different localities are entitled to weight requires more research and comparison than the subject has hitherto received. As far as outward figure and colour are concerned, there appears considerable unifor- mity; but paucity of trustworthy observation may have something to do with this, if Couch’s? assertion is well founded. The precise position of the Cetacean dorsal fin would seem to be no sure specific test; for between foetus and mother there is no unanimity: in other words, its position depends pari passu on the age of the animal. Whilst the entire pectoral limb of the Whale is reduced to act in a watery mobile element, it acquires unusual power of a semirotatory kind. Professor Huxley* remarked of the Porpoise, “it is not, however, by any motion of those fins that locomotion is effected, this being almost exclusively produced by the sculling action of the tail.” True! still it seems to me that every movement must to some extent be influenced by these appendages. In nautical parlance, crank vessels require floats or outward balances. Now this is precisely the office of the semirigid extremity. It rotates, moreover, on its own axis, through the well-nigh ball-and-socket joint and fore-and-aft levers. Has subservience to function, preservation of type, selective power, or hereditary transmission had the most enduring influence in the transference of such a complex fleshy arrangement of the naso-facial muscles? Huddled up beneath a mass of blubber, ? This opinion may seem very offhand as weighed against such authorities as Gray, Gervais, and van Beneden ; but the wide geographical range of G. melas makes one hesitate to accept distinction unless well founded. I regret the cessation of the text of the ‘Ostéographie des Cétacés, which, in spite of pls. 51 & 52, leaves a loophole of uncertainty. Burmeister’s Globiocephalus grayi (Anales, as cited, p. 367, pl. 21), as far as I can judge from a study of his text and plates, seems specifically identified more by abnormality in dentition than by strikingly trenchant cranial distinctions; and that, as I have intimated above, is a point of weakness in diagnosis. Regarding Malm’s Globiocephalus propinquus (Svenska Kongl. Akad. 1871) I have not had an opportunity of mastering his differentiations. If, however, G. melas, as proved, ranges from the Polar seas to the Mediterranean, and possibly south of that, is it not probable the Gottenburg specimen may eventually turn out to be a variety of the common species? ‘The skeletal changes from youth to age have yet to be worked out in Globiceps before species can be soundly established. ? Corn. Fauna, p. 10, quoted by Gray. * Hunterian Lectures, 1856. VOL. Vill.—ParT iv. February, 1873. 27 292 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. as they are, they yet retain a physiological importance of no mean value. In them can rest no power of expression whatsoever; withal, what a development of parts which in higher mammalia are devoted mainly to such a faculty! It would seem as if there were no end to utilitarian purposes. The same tool which in the Cetacean is perforce adjunct to the respiratory act, is transformed in the Ruminantia to a grasper of herbage, in the Suid to a digging-instrument, in the Elephantide to a tactile flexible limb, and sO on. Again, in the fish-like form of body of Whales, there rules similarity of spinal tendino-muscular distribution. It is easy to trace the homologues of the great lateral piscine muscles and even their segments (myocommas). For whether studied in the longitudinal direction, or in cross section, there can be recognized counterparts of fibre, tendon, and flesh as found in the predacious Shark and others of the finny tribe. There is one piece of organic mechanism in which, so far as Glodiceps is concerned, I do not quite follow the descriptive evidence of previous anatomists. I allude to the motions of the larynx. As I find it in this form, it is difficult to conceive how it is thrust upwards and downwards, and firmly constricted at every expiration and inspira- tion. The chief directions it seems capable of moving in are fore-and-aft, with a limited power of elevation and moderate dilatation and approximation of the aryteno- epiglottic funnel. It appears to me that the upper portion of the funnel at all times rests in the posterior nares, upright, and with an obliquity corresponding to the direc- tion of the inferior nasal channel. ‘This leaves one of the pharyngeal passages wider than the other, along which the greatest bulk of the food must pass. It is during deglutition, not respiration, that the strongest action of a grasping and elevating nature takes place, in this presenting a certain agreement with the second phase of the act in ourselves. At such times perfect closure of the funnel by compression of the circular fibres of the superior constrictor, and elevation through contraction of the longitudinal ones, blockades the postnares and sends the food on to the cesophageal sphincters in the rear. When expiration and inspiration are about to take place, it is quite obvious that were the naso-laryngeal sphincter forcibly to grasp the aryteno-epiglottic tube, and powerfully drag it upwards, as some assert is the case, the consequence would be closure of the breathing-tube itself. I conceive that in the performance of respiration there is a firm but moderate contraction of the naso-pharyngeal parts, sufficient only to steady the aryteno-epiglottic funnel, and prevent passage of air into the gullet. At the same time those laryngeal muscles which act as dilators open the cartilaginous tube, and give free vent to the outgoing and incoming air. Meanwhile also the muscular and resilient cartilaginous parts at the spiracular orifice play in active unison. The function of the retia mirabilia, with which Whales above all other vertebrata are copiously furnished, is still a matter of conjecture. Such men as Hunter, Cuvier, and Breschet have looked upon them in the light of reservoirs, to store up an excess of arterial blood, needful during the animal’s long submersion—a kind of supplementary DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 293 oxygenated residuum in lieu of more frequent respiratory acts. With reference more particularly to their distribution in the limbs, Von Baer conceived the manifold sub- division as concurrent with paucity of movement of the members. Besides playing the part of great lagoons (reciprocal recipients of a circulatory overflow, as some put forward), Professor Turner avers, as Milne-Edwards' had already advanced, and reason- ably, that such subdivision distributes, equalizes, and retards the blood’s force ere reaching the sensitive nervous centres &c. I believe they are designed to execute a highly important vito-physiological process, which may be combined with some subsidiary mechanical adaptation as has been asserted. ‘Their office is equivalent to modified blood-glands, in some way related to pabulum or nutrient fluid. The retia mirabilia in Cetacea and many other Mammals are not confined to the cerebro-spinal tract and neighbourhood of the respiratory apparatus, but principally follow the lines in body and limbs where the lymphatics and absorbents are known to obtain in the greatest profusion. Moreover, in Cetacea, Sirenia, Phocidz, and other forms where retia are very manifest, even some birds, the lymphatic glands are unusually abundant and of large size; so that their intimate connexion with the vascular plexuses is a most presumptive conclusion. I apprehend that the countless divisions, subdivisions, and minute vascular osculations, by coming in close contact with the lymphatic system, conduces to an interchange or exudation of their constituents’. What further physiological process takes place I am not prepared to demonstrate, though inferentially I would adduce multiplication of the lymph-corpuscles ; a view maintained by some as respects the office of the lymphatic glands. Such a pro- position is applicable to many varied physiological phenomena of absorption and nutrition in divers animals well known to possess fully developed retia mirabilia. + Lecons, Physiol. Anat. Comp. 1859, vol. iv. p. 260. ? I conceive, as already mentioned, that there is a certain functional homology between the so-called caudal hearts of lower vertebrates and the Cetacean mesenteric moniliform tube of Turner. In like manner I regard the human so-called coccygeal gland as strictly homologous with the inferior caudal plexus of Whales &c. Moreover, as the vascular retia distributed throughout the body and limbs are essentially similar in constitu- tion, it follows they may all serve one office. As there is in the first case a manifest intermixture of plasma (Jones, 7. c.), so in the second, by absorption or otherwise, may a phase of nutrition be subserved. Abnormal developments of the lymphatic and vascular systems are pari passu pronounced, though not restricted to the aqueous and amphibious vertebrates and those that hibernate ; in them, though respiration may be checked or subdued, nutrition must perforce go on. Do the retia and lymph-sacs, then, supplant the necessity for frequent respiration, or substitute by subsidiary function a reserve force where depuration or nutritive quality of the bluod is interfered with ? 272 294 5. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. DESCRIPTION OF THE PLATES. PLATE XXX. . Drawing from nature of the young female Pilot Whale, Glodiocephalus melas, Traill. To a scale of 3; of the natural proportions. . Dorsal aspect of the same Whale. . Its blow-hole, of nat. dimen.: @, anterior, and p, posterior lip. . The eye, about nat. size. PLATE XXXI. Foreshortened view of the head, looking into the opened mouth of the same female Globiocephalus. Considerably reduced in size. T,tongue; P, palate ; f, faucial membrane and folds; a, angle of the mouth. . Segment of the roof of the mouth, showing palatal rugosities, glands, and pre- maxillary teeth, the left series of which are complete: g, gutter; a, angle. . A left premaxillary tooth, nat. size, with outline of the gum (9): the ridges seen in cross section. . A longitudinal mesial section of the same tooth: p, pulp. . Left posterior mandibular tooth, seen from behind, and shown with gum-ridges as in fig. 7, also nat. size. . 10. A small portion of the faucial part of the palate, showing mucous crypts, ¢, or fissures and glands, g/; fy, fungiform papille. ig. 11. Underview of the tongue, larynx, and part of the trachea, prepared chiefly to show the muscular structures. The superficial layer remains on the left, but a deeper dissection is made on the right side. f, frenum lingue; Gh, Gh*, genio-hyoid, entire and cut tendinous origin ; S¢.g, stylo- glossus; G.h.g, genio-hyo-glossus ; Hg’, Hg’, hyo-glossus, first and second heads; Thy, interhyoideus; S.4-Mh, insertions of sterno- and mylo-hyoidei: 7h, thyroideus ; S.th, part of sterno-thyroid; C.th, crico-thyroid; Th.g/, thyroid gland; O.thy, occipito- thyroid; th, thyro-hyal; 7, thyroid cartilage; J, thyro-hyoid ligament; tr, trachea; sln, superior laryngeal nerve ; 8’, hypoglossal nerve. Fig. 12. Upper or palatal aspect of the same parts. Arytenoid conjoined with epiglottic cartilages are seen protruding upwards, surrounded by their sphincter. T, tongue; P.g, palato-glossus; P.ph, palato-pharyngeus, cut across; Cs, Cm, & Ci, constrictores superior, medius, et inferior; ep, epiglottis; A, arytenoid cartilage; da, lingual (ranine) artery; hp.n, hypoglossal nerve; Jhy, interhyoideus; Sph, stylo- pharyngeus; @, cesophagus; tr, trachea; Shy, stylohyoideus; Stg, styloglossus; 7; frenum lingue. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 295 Fig. 13. A deep dissection of the hyo-laryngeal region seen laterally. On the left moiety, that towards the observer, the mucous and subjacent tissues have been cut away, bringing several of the small muscles into sight. An arrow, partly in dotted line, denotes the floor of the pharynx, leading to the ceso- phagus, @; two other short arrows indicate the canal of the glottis to the trachea, cut short; A, arytenoid cartilage; ep, epiglottis; phv, pharyngeal valley and ridge; He, hyo-epiglottic muscle ; Sgs, squamo-styloid; Ar, arytenoideus; Cap, crico-aryte- noideus posticus; Cal, crico-arytenoideus lateralis; Tar, thyro-arytenoideus; ic, inferior cornu of thyroid cartilage ; J, thyro-hyoid ligament ; Zh, interhyoideus; sh, stylo-hyal bone. Fig. 14. Semidiagrammatic and much reduced sketch of the hyoidean arch, and the larynx from below, of the Pilot Whale. sh, stylohyal; ch, ceratohyal; bh, basihyal; th, thyrohyal; /, the thyro-hyoid liga- ment, anterior et lateralis; 7’, thyroid cartilage: the latter is placed between the body and right ala; 7.c, its inferior cornu; C*, cricoid cornu; tr, trachea. Fig. 15. Upper or pharyngeal surface of the same, and showing the oblique inclination of (ep and A) the aryteno-epiglottid projection; c, broad portion of cricoid cartilage ; 1 & ic, as in fig. 14. Fig. 16. An outline of the thyroid bones in profile. Letters as in fig. 14. PLATE XXXII. Figures 17-24 inclusive are a series of perpendicular sections of the dorsal fin, cut at points corresponding to the dotted lines, nos. 1-8, in fig. 25. Fig. 17. Perpendicular and transverse section of the anterior portion of the dorsal fin, viz. 19 inches distant from its tip. Fig. 18. Another segment, nearer middle, 13 inches from point. Fig. 19. Another segment, 7°5 inches from tip: a, artery. Fig. 20. Another segment, 6:2 inches from tip. Fig. 21. Perpendicular section of the same fin, close to the posterior concavity, and 4°5 inches in front of the hinder point of fin. Fig. 22. Section, free part of fin, 4 inches in advance of tip. Fig. 23. Another slice, 3:2 inches in advance of tip. Fig. 24. Another slice, 1-25 inch in advance of tip. Fig. 25. Miniature outline of the fin of the back, representing by numerals (1-8) and dotted lines the positions where the slices, figs. 17-24, have been cut from. Fig. 26. Skull of Globiocephalus melas, mandible attached. Also the right half of the great nasal protuberance, exhibiting its structure in mesial section and part of the narial passage and lower pouch. Bi, blubber ; fa, fibro-areolar glistening tissue; p, right premaxillary sinus, partially hidden by the fatty mass; arrow directed into blow-hole. 296 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. Fig. 27. The blow-hole of the White-beaked Bottlenose (Lagenorhynchus albirostris) opened by retroverting the lips. It shows the alar protuberances, smooth- surfaced posterior wall, and antero-lateral corrugated membrane. 4a, anterior, and p, posterior lips; sp, septum narium; ¢, cartilaginous cushion, the arrow being directed into the maxillary pouch; 7, cuticular ridges of the spiracular cavity. Fig. 28. Diagram of the narial pouches and spout-hole of Globiceps. B, blow-hole; spc, spiracular cavity; p, premaxillary sac; m, maxillary, and nf, naso- frontal sacs. Fig. 29. Sketch of the right orbito-auricular region of the male L. albirostris, showing a dissection of the ear-muscles &c. ac, auditory canal; Ret, retrahens aurem; Afr, attrahens; Atl, attollens; Te, tem- poralis ; fav x std, facial vessels, nerve, and Steno’s duct. Fig. 30. An underview of portion of the left mandible and pterygo-maxillary region of the same Lagenorhynchus. Dissected to show the vascular rete inter- vening between the lower jaw and basis cranii, and its relation to the pterygoid muscles &c. Mn, mandible; z, zygomatic bar; Jpt, internal pterygoideus ; Ept, external pterygoid, severed; ete, of the basis cranii and mandible; ima, internal maxillary artery; id, inferior dental artery ; a, artery, and v, vein, pterygo-maxillary branches; pum, postnarial muscle. Fig. 31. Segment of the base of the skull of the Caaing Whale, which displays from below the tympanic region, the Eustachian tube, its enlargement, and the extensive rete mirabile covering its membranous walls and the adjoining depressions. c, left condyle; oc, basioccipital; Hoc, exoccipital; ty, tympanic bulla; ma, malar; 2, zygoma; pt, pterygoid; P.na, posterior nares; ac, auditory canal; Hus, Eustachian tube partially opened; ete, basicranial plexus; wp, venous plexus; ef, condyloid foramen; fn, facial nerve ; she, stylohyal cartilage. PLATE XXXIII. Fig. 32. Portion of cesophagus, the compound stomach, and part of the duodenum of the Pilot Whale. All more or less opened to show their internal walls and connecting passages. Their relative positions are necessarily displaced, to exhibit interior structure. @, esophagus; gl, esophageal gland; J, J*, LJ, III, & IV, the four separate gastric cavities or stomachs; nos. 1, 2, 3, 4, 5, 6, lie over the pointers which indicate the several passages from the cesophagus to the commencement of the small intestine; DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 297 no. 6 is the pyloric orifice, and 3, 4 the narrow openings at either extremity of the passage or diminutive cavity (*), through which a style is passed; p.d, pancreatic duct joining; d.ch, the ductus communis choledochus. The latter enlarges within the intestinal wall, shown by the dotted line, and pierces the inner coat of the gut at no. 7; duo, duodenum, smooth above, and with valvule conniventes below. Fig. 33. Semidiagrammatic view of the liver, stomach, spleen, and part of the mesen- tery in their natural positions after removal from the abdomen, the stomachs being distended. Part of the lettering applies as in fig. 32. The dotted lines upon the surface of the gastric cavities give the places of attachment of the great omentum, which has been removed. I, liver; ¢./, coronary ligament; j, a modicum of the jejunum; Me, mesentery; gl, glands thereon ; Sp, spleen; Pa, pancreas. Fig. 34. Greatly reduced view of the under or postabdominal surface of the liver: r.J, round ligament; v.c, vena cava; h.a, hepatic artery ; h.v, hepatic vein. Fig. 35. Spleen, seen on its free or ventral surface. Fig. 36. Right kidney partially dissected, showing the distribution of the vessels on its atlanto-mesial side. Renal artery, a, and vein, v. Fig. 37. A few of the lobular acini of the kidney, of their natural dimensions. The connecting fibrous envelope has been removed. Fig. 38. A renal lobule bisected, to exhibit its internal structure. Nat.size. c, cortex; m, mammilla; h, hilus. Fig. 39. Suprarenal body, reduced in size to correspond with the kidney, fig. 36, and placed somewhat in the natural relation it bears towards it. PLATE XXXIV. Fig. 40. A side view of one of the vascular glands of the neck, drawn } nat. size. Px, cut plexus of vessels. Fig. 41. A mesial section of the same, the letters agreeing. Fig. 42. Small portion of the esophagus, displaying the longitudinal ruge and a glandular crypt, g/, of nat. size. Fig. 43. A vertical transverse section of the cesophageal wall, and cut through one of the glandular fossee. The relative natural height of the ridges, depth of the furrows, and thickness of the muscular and mucous coats are preserved. gl, glandular crypt; m, mucous and submucous coat; /, longitudinal layer of muscular fibres; ¢, cross fibres or circular layer. Fig. 44. Portion of the duodenum with agminate patch: P.g/, Peyer’s glands; V.co, val- vulz conniventes. Fig. 45. Another piece of the jejunum, where Peyer's glands (P.g/) are elongate and narrow: J.f, the longitudinal folds. 298 Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. 46. 47. 48. 49. . 50. 51. 53. 54. 55. 56. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. Piece of the colon, displaying a pouched character: S.g/, some of the solitary glands. A segment of the unopened bowel, intended to display the comparative narrow diameter of the gut as it approaches the rectum, and the appearance of the mucous folds and surrounding intestinal wall. A portion of the rectum, cut open, showing the rugz, the narrow anal aper- ture and thick sphincter ani muscle (Sp/) in transverse section. PLATE XXXV. Heart and commencement of great vessels, partially enveloped by the inflated lungs. A bridge of fibrous membrane, conjoined to the diaphragm (J), unites the rear sternal angles of the lungs, and it encloses a pair of glands (gl): a small strip of the pericardium (pe) is seen above. pa, pulmonary artery ; a0, arch of aorta; 7, 7*, right and left innominate trunks ; m, mammary artery; 7, pulmonary lobule. Superficial view. Segment of sterno-ventral corner of left lung, and special gland, &c. LL, lung covered by (p) the pleura, which has been removed from the semilunar line towards (g/) gland, and exposing a series of convergent vessels (these, a+, are partly serpentine in course); phv, phrenic vein; sph, superior phrenic vessels; ph, phrenic nerve; pe, portion of pericardium attached to (D) diaphragm. Deep or posterior view of the parts figured in fig. 50; the internal mammary vessels, however, are not shown. 6, bronchus. Other letters as in the pre- ceding figure. Upper view of the septal segment of the bicuspid valve of the heart of the female Globiceps, exhibiting an auxiliary valve: s, septal valve; * and the arrow point to the abnormal perforation. Interior view of the same, auriculo-ventricular opening, the under surface of the septal valve, and its fleshy cords; also the supplementary orifice and its tendinous cords. One arrow denotes the blood’s usual flow towards the right ventricle. The second starred (*) arfow passes through the orifice in the septal segment (S). cc, columna carnea; ct, chorda tendinea, and cé*, supplementary chorda tendinea belonging to the secondary valve. A lumbar vertebra of the Pilot Whale, to show (Sp.px) Section of the spinal plexus of vessels; the large orifice is a vein, the smaller chiefly arteries. sp.c, spinal cord; L.pa, lumbar arterial plexus or rete mirabile; c, centrum of the vertebra; ¢, transverse, and s, spinous process; Zz, zygapophysis. Section of intercommunicating venous network. A transverse section of the penis of the Razorback Whale (Physalus anti- quorum, Gray, =Balenoptera musculus). DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 299 u, urethral passage, surrounded by (cs) corpus spongiosum; ¢.c, corpus cavernosum ; t, trabecular glistening fibrous bands, and (f) fibrous tissue; v.v*, vene dorsales. PLATE XXXVI. Fig. 57. Body of the female G. melas, from which the skin and subcutaneous layer of fat have been removed, excepting on the tail, the pectoral fin, and an edging round the lips. Glistening fascia and fibrous tissue fill-in the areas unoccu- pied by the fleshy coat. ac, orifice of auditory canal; Pe, panniculus carmosus. Fig. 58. Abdominal aspect of same specimen. The lower (left) half displays the super- ficial layer, the upper (right) half a deeper dissection. Pe, Pe, panniculus, belly- and throat-portions; P.ma, pect. major; P.mi, pect. minor ; R.ab, rectus abdominis; E.0, external oblique; Sc, sacro‘coccygeus; Ee+ Isc, erector clitoridis and ischio-coccygeus ; La.d, latissimus dorsi; S, subscapularis; Ch, cephalo- humeral; S¢.m, sterno-mastoid ; Shy, sterno-hyoid; Sth, sterno-thyroid; Ih, mylo- hyoid; Gh, genio-hyoid: Stg, styloglossus; sk, skin; V, vulva; A, anus; v, caudal vein. Fig. 59. Diagram to illustrate the superficial oblique bridging tendons of the tail of Globiceps, and how each superficial long dorsal muscle terminates in a single, thick, apparently cordiform tendon. Sp.d+ Lci, spinalis dorsi and homologue of levator caude internus; Zd+Lce, longissimus dorsi and levator caude externus. Fig. 60. A transverse section of the terminal caudal tendons, and their fibrous (almost cartilaginous) investment, of the Great Northern Rorqual (Balwnoptera mus- culus, =Physalus antiquorum). Prep. No. 45 8, Coll. of Surgeons. The tendons (), as in fig. 59, are packed or overlap one another, so as to seem but composed of a single one when viewed longways, and with their fibrous coat in situ. The (f) investing mass or cushion has interlacing glistening fibres, here and there penetrated by bloodvessels (a) &c. Fig. 61. Diagram showing the manner in which the tendons of the great inferior loin- muscle end in the Caaing Whale. Sc, sacro-coccygeal, secondary tendons dragged out; finally they enwrap (as above) in a sheath (sh) the thick com- pound terminal caudal tendon, whilst presenting continuity with the super- ficial fibrous layer of the tail. Fig. 62. Semidiagrammatic representation of a transverse vertical section of part of the caudal keel of the Caaing Whale. It shows (¢) the lateral compound or vaginate tendons of the sacro-coccygeus; a, arterial channel beneath, and piercing (f)) the firm fibrous keel, overlain by (sk) the skin. VOL. VilI.—PpartT iv. February, 1873. 20 300 DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. PLATE XXXVII. Fig. 63. Body in profile, exhibiting the myology &c. when the panniculus has been taken away. The right moiety of the rostral blubber remains in place. Letters other than those used in fig. 58 are:—JZ.d, longissimus dorsi, +L.c.e, levator caude externus; Sp.d, spinalis dorsi, +Z.c.7, lev. caud. internus ; S.cd, supracaudal ; T.cd, infracaudal; S.J, sacrolumbalis; S.mg, serratus magnus; Rh, rhomboideus ; C.a, cervicalis ascendens; Z.a.s, levator anguli scapule: 7c, transversalis cervicis ; Sp, splenius; J.sp, infraspimatus ; L.cl, levator clavicule; D, deltoid; T.ma, teres major; Di, digastric; Ma, masseter; *, extra cheek-fibres ; Bu, buccinator; Te, tem- poral; Op, orbicularis palpebrarum ; WJ, naso-labialis or premaxillary; Of, occipito- frontalis; J/san, levator labii superioris aleque nasi; B, Blow-hole, and arrow leading therein; /, hook pulling out tendons and superincumbent fascia of sacro- coccygeus. Fig. 64. An upper view of the skull, in which the superior muscles of the blow-hole are dissected on the right side. Fig. 65. Hinder part of skull, showing second layer of muscles acting on the blow-hole. Fig. 66. Third layer of the same, also in profile. Fig. 67. Short deep muscles of the neck, on their lateral aspect. The lettering of these four figures runs:—B, blow-hole ; spe, spiracular cavity; p, pre- maxillary sac; m, maxillary sac; nf, naso-facial sac; N/ dé N/»?, naso-labialis; Of, occipito-frontalis; Lisan, lev. 1.s. al. nasi; Lsp, levator superioris proprius; Z, zygo- maticus &c.; Py, pyramidalis &c.; Te, temporal ; Rp, recti postici; Os and Oz, obliquus superior and inferior; Zms, trachelo-mastoid; J¢s, interspinales, anterior dorsal region; Ch, Stm, Lel, and Las, cut insertions of ceph.-hum., sterno-mastoid, lev. clay., and lev. ang. scap.; 2, transverse process of axis. Fig. 68. Partial view of under surface of the skull, the neck, and thorax, the ribs of the latter being taken away on the right side. Sca, scaleni; Lc, longus colli; Ra.mi, rectus anticus minor; Ra.ma, rectus anticus major, severed; £2, rectus lateralis; Sph.n, sphincter of post-nares; P#, pterygoideus. Fig. 69. Sternum and sternal ribs, from the inside: St, osseous sternum; Ts, triangu- laris sterni. Fig. 70. Inner aspect of the left pectoral limb. S, subscapularis; Smg, serratus magnus, insertion cut; Ssp, supraspinatus ; Tma, teres major; Lad, Pc, & Ch, lat. dors., panniculus, and ceph. hum. insertions; 7, triceps; Cb, coraco-brachialis ; Pt, palmar tendons. Fig. 71. The right scapula, from within: cp, coracoid process, and Ssp, the supra- spinatus, in its groove. Fig. 72. Caudal moiety, minus flukes, of the White-beaked Bottlenose (Lagenorhynchus albirostris), dissected, and showing correspondence with what obtains in Fig. 77. Fig. 78. DR. J. MURIE ON THE ORGANIZATION OF THE CAAING WHALE. 301 Globiceps; compare fig. 63, with which references agree, save Spc and Ifcd to supra- and infra-caudals. The Nos. 40, 45, 60, denote the vertebre, counting from the skull. PLATE XXXVIII. . A posterior or deep view of the rectum, uterus, &c. of G. melas. . The uterus, its cornua, and the urinary bladder, opened, and from above. . Pelvic bones, and their ligamentous bridge &c., deep surface. . Dissection of the external genital region of the same animal, to show situation of the mammary gland and the muscles acting on it and connected with the genitals. An outline of the bladder and uterus are left in place. The super- ficial muscular layer remains on the right side; and on the left a deeper layer is displayed. i The left moiety of the same region, exhibiting a still deeper muscular layer. Another view of fig. 77, but seen from the side, and with some of the muscles removed. All the above sketches are reductions from nature ; the lettering corresponds through- out. Whilst some of the muscles in one figure are shown entire, their origin or inser- tion only is displayed in the succeeding layer. Ut, uterus; os, its os tince; va, vagina; V, vulva, and cl, clitoris; c, cornu, and cc, same opened on right side; 0, ovary, and-o*, same split open; pav, ovarian pavilion; B, urinary bladder; uw, urethra; we, ureter; A, anus; 7, rectum, v, hemorrhoidal vein ; ii, internal iliac artery; Mg, mammary gland; g/, rectal glands; Pi, pelvis, * repre- senting an anterior, and ** a posterior cartilaginous nodule; Pf, interpelvic fascia ; P.c, panniculus carnosus; L.a, levator ani; Sp.a, sphincter ani; Hab &c., rectus abdominis &c.; E.c, erector clitoridis; Cm, compressor mamme; C.wr, constrictores urethre ; Spv, sphincter vagine ; Dtp, deep transversus perinei; Isc, ischio-coccygeus ; P & Ic, pubo- and ilio-coccygeus; a Xn, artery and nerve transmitted through aper- ture in interpelvic fascia (probably the homologues of obturator artery and nerve in quadrupeds). p LOR or > j Tait 52 Sha JA ree Per eh ist “f ar’ eo lect 3) io aia eee Bio ols o Pe ea i “4 . 4 oP. ™ r. . “em ~ . A Gt OM OT i : 2 Baars UG Cee im en iy Soh airy Gxtrsaios bo tuts ealamstic oui (88 Je Ott ep 7 [ elie 75 See ET + Wes Deri foe Perky ‘Ss - UMTS wlMmaRpRetos Bit f3 ges wiiishals onntk acsioy ee 7 2k 10 gO. Sasi copuly fil erties tweaks “i bel 4 t sis eisfee aile es fii Hiee ieee e oe «* =. ies 7204 As A}, Wes 3 AMR iby - hee te lutosy atl Con ledh oer Cae ox AS ee TUTE 5 ; mi ues El for qareyure yy Ne D010 TS UT}? Pp neshw~ags5 Of ULE AUO"R FG Tange me J Mure %y 04 CBeryeau lth MAN Hanhart imp MOUTH CAVITY & HYO-LARYNCEAL APPAR >» JI.Murie reliitlali cic pgeiien M &N.Hanhart ump C.Berjeau del. etlith 1 r Fig 32 4 ‘ ; M&NHarhart imp C Berjeau delet hth ime OBerjeau lithad nat NDS ALIMENTARY bLA LDS & FO hg 49 Fig 56 yoaua.del et Lith. M&N Hanhart imp PULMO-SANGUIFEROUS STRUCTURES & PENIS - Ee dur 9.reuwey N7 WN TIVL SNOGNaL ®? SUAAVT TVIDIZHAdNS “ADO TOAWN py weoliaa'g gC ‘pig 7] pds? 1g by IROE IV 0°, GGG l UP — 3 t TOH M( a +g o cl OTE) SNS qWi SUAAVT dada ras \ a IG OAW (HLA? [ep meelaegg 99 CIPI OL hz Ww eo bus W > ‘S rere OT a /C25°97307,0 LG oy Se By, ZOOP PLL DA y) M& N Hanh ach imp on GENERATIVE ORGANS # PELVIC BONES & MU [ 303 ] VIII. A Description of the Madreporaria dredged up during the Expeditions of H.MLS. ‘ Porcupine’ in 1869 and 1870. By Professor P. Martin Duncan, MB. (Lond.), F.R.S., F.G.S., Professor of Geology to King’s College, London, &c. Read May 16th, 1871. [PLates XXXIX. to XLIX.] ConTENTs. I. Introduction, p. 303. III. Descriptions of and Remarks upon the Species, p. 309. II. Classification of the Species, p. 306. IV. Tables of Localities, &c., p. 338. I. INTRODUCTION. ZOOPHYTOLOGISTS have long known that some simple Madreporaria, or Stony Corals, live at a depth of from 80 to 150 fathoms on the floor of the Atlantic and Pacific Oceans, in areas remote from coral reefs. The existence of Madreporaria at corresponding depths in the Mediterranean Sea has also been known practically to those interested in the red-coral fisheries. Nevertheless it was until lately tacitly admitted that all corals obeyed the laws which regulate the bathymetrical distribution of reef- building forms. But it is only during the last four years’ that a fine and peculiar’ coral fauna has been proved to exist not only at great depths but also in temperatures ranging from below freezing-point to 55° Fahrenheit. Although Foraminifera and Echinodermata have been brought up from profound depths by the sounding-apparatus, no trace of a stony coral had been observed; and it was not until the United-States Coast Survey dredged between Key West and Havana in 1867, that any proot of the existence of Madreporaria at a great depth was obtained. On May 24th, 1867°, stony corals were dredged up alive from the depth of 270 fathoms; and on May 25th the same species was obtained from 350 fathoms. In the following year, 1868*, many species were dredged up under the supervision of M. de Pourtales, as naturalist, off the Florida reef, in the course of the Gulf Stream, from 43 to 324 fathoms. The expedition of H.M.S. ‘ Lightning,’ in the same year, conducted under the super- 1 The researches of Sars, MacAndrew, Norman, and J. Gwyn Jeffreys prepared the way for those of the United States Expeditions and of H.M.SS. ‘ Lightning’ and ‘ Porcupine,’ under Fourtales and A. Agassiz, and Wyville Thomson, Carpenter, and Gwyn Jeffreys respectively. 2 « (oral Faunas of Western Europe,” Quart. Journ. Geol. Soc. 1870, P. M. Duncan. 3 «Contributions to the Fauna of the Gulf Stream at great Depths,” by L. F. de Pourtales, No. 6 Bulletin of the Harvard Museum of Comparative Zoology, Ist series. * Op. cit. 2nd series. You. vilt.—Part v. March, 1873. bo 4 304 PROFESSOR P. M. DUNCAN ON THE vision of Professor Wyville Thomson, Dr. Carpenter, and Mr. J. Gwyn Jeffreys, in the North Atlantic, afforded satisfactory evidence of the existence of a coral fauna in the deep sea, and in water of a very different temperature from that usually considered necessary for the Madreporaria; but it was not until 1869 that, under the same able direction, the systematic dredgings of H.M.S. ‘ Porcupine’ proved the existence of coral life at a depth of 705 fathoms’. A description of this dredging expedition was read before the Royal Society in 1870, by Messrs. Carpenter, Wyville Thomson, and J. Gwyn Jeffreys’, and a report on the corals*® obtained by those naturalists, and intrusted to me for examination and descrip- tion, was read before the same Society March 24th, 1870. When the second expedition was preparing to start in 1870, under the same able guidance, particular requests were made to me to advise concerning the dredgings, so far as the corals were concerned. The employment of the “hempen tangles,” instead of the crushing dredge, had already commended itself to the superintendents of the dredging operations; and it was a satisfaction to find that by using these simple means a fine collection of specimens was obtained off the west and south-west coast of the Spanish Peninsula and along the Mediterranean coast of Africa, from depths which reached to 1095 fathoms. . All the information I have required has been freely given me by the three naturalists who were responsible for the dredgings, and also by M. Lindahl their assistant. Professor A. Agassiz and Count Pourtales were of great assistance to me by sending me their reports and also a collection of the specimens dredged up by them, so that I have been enabled to compare the North-Atlantic forms with the West-Indian and Floridan types. I have also had the advantage of examining the results of the dredgings conducted by Mr. Kent, of the British Museum, and of his assistance in comparing specimens. The majority of the specimens dredged up and entangled during the two expeditions of the ‘Porcupine’ were alive and in good condition; afew had been dead for some time and were covered with Sponges, Serpule, and Polyzoa. Specimens were not invariably obtained at every dredging; and it is evident that corals live here and there in patches, and that they mostly frequent rocky ground, although some species live in the Globigerine ooze. Fine muddy sediment appears to be unfavourable to coral life; and many of the dead specimens which came up in the dredge were filled with it. It is somewhat remarkable that so many well-known species, especially of the Medi- * Proc. Royal Society, 1870. * W. B. Carpenter, F.R.S., Wyville Thomson, F.R.S., and J. Gwyn Jeffreys, F.R.S., « Deep-Sea Researches,” Proc. Roy. Soc. 1870, vol. xviii. pp. 397-492. W. B. Carpenter, F.R.S., and J. Gwyn Jeffreys, F.R.S., “Deep-Sea Researches,” Proc. Roy. Soc. 1870, vol. xix. pp. 146-221. * P. M. Duncan, ‘ Porcupine’-Expedition Madreporaria, Proc. Roy. Soc. 1870, vol. xviii. pp. 289-301, MADREPORARIA OF THE DEEP SEA. 305 terranean coral fauna, should not have been obtained by the dredge and tangles. Some of them are dwellers in moderately deep water, and have been noticed by Forbes and Milne-Edwards and Jules Haime. On the other hand it is satisfactory to have obtained so many specimens of certain rare corals that the doubts about their classificatory position could be solved by subjecting a few to transverse and longitudinal section. The results of this procedure, with respect to Lophohelia prolifera, Pallas, sp., and Amphihelia oculata, Linnus, are very damaging to the integrity of the family of the Oculinide ; for, as the corallites of these species do not fill up from within, they can no longer be separated from the Astreide. Moreover longitudinal sections of the corallum of specimens of the first-named species show dissepiments and large tabule; and thus the propriety of establishing a section of the Madreporaria which shall be differentiated by these perfect transverse floors of endotheca is strongly contra-indicated. The specimens obtained in both of the expeditions of the ‘ Porcupine’ can be arranged into forty-eight groups, consisting of species and varieties—i. e. twenty-seven species and twenty-one varieties. ‘There are, amongst these, fourteen new species, twelve species already described, and one incertw sedis. The varieties consist. of four the typical forms of which are not present in the collection, and of seventeen which accompany the specific types also. Owing to the extraordinary variability of some of the corals, I have been able, by comparing them with the descriptions and types of closely allied recognized species, to absorb several specific forms, and in two instances to treat genera so. Thus I have absorbed the genus Ceratocyathus in the older genus Caryophyllia, and Thecopsammia in Balanophyllia. The species Caryophyllia borealis, Fleming, and C. smith, Stokes, I have made varieties of the older type Caryophyllia clavus, Scacchi. The Sphenotrochus intermedius, Ed. & H., of the Crag and Tangier Bay, and Sphenotrochus milletianus, Defrance, are varieties of the same type. Two species of Desmophyllum are absorbed into the species crista- galli; and its range, therefore, is enlarged. No less than seven species and two genera must now be aggregated in the Amphihelia ramea of Miiller; and three if not four species of Lophohelie should be associated with the species prolifera and its variety gracilis. At least fourteen old species have been thus absorbed, and, it is trusted, not to the detriment of truth or of a true classificatory philosophy. The following is the list of the Madreporaria dredged up in the two expeditions of the ‘ Porcupine’ :— 2x2 306 PROFESSOR P. M. DUNCAN ON THE II. CLAssIFICATION OF THE SPECIES. ZOANTHARIA SCLERODERMATA’. Section APOROSA. Group Turbinoliide. Subfamily CaRYOPHYLLIINA. Genus CARYOPHYLLIA. Species 1. Caryophyllia clavus, Scacchi, sp. ——, variety w. elongata. , variety B. exserta. , variety y. borealis. , variety 5. smithit. , variety ¢. epithecata. cyathus, Ellis and Solander. A variety. arcuata, Kd. & H. — — _ Recent, no. 30. 9 1 1, and 4 allied. | Résumé. 1. That the deep-sea corals differ from the reef-building forms in not possessing certain important coenenchymal structures. 2. That deep-sea corals have been dredged up from a depth of 1094 fathoms. 3. That deep-sea corals live in very different temperatures, from 29°'9 to 56°3 Fahrenheit. 4. That the growth of deep-sea corals is very vigorous at great depths and in low temperatures. 5. That the variability of deep-sea species is very great. 6. That genera and species exist at great depths which are unknown elsewhere. 7. That the horizontal distribution of many deep-sea forms is great. 8. That the present deep-sea coral faunas contain species and varieties of species of 3 B2 338 PROFESSOR P. M. DUNCAN ON THE the Pliocene, Miocene, and Cretaceous deep-sea faunas, and offshoots of the Paleozoic fauna. 9. That, omitting varieties, and counting species only, there are thirty species in the ‘Porcupine’ dredgings, and of these eleven are known as fossil forms also. Moreover four others are closely allied to fossil forms. IV. Tasues or Locauitiss, &c. First Expedition of the ‘ Porcupine.’ sation. te Tenge Depth. RE Fs A - F fathoms. 5 Bo obocoe Bil Bf banat UO) 2B loon 30-40 ...... 52:0 3 pe ome Gil BL Goiwso5 IY Ey. “Beosoe en) 9 Socduc aoa 3 gooone BeOS. 4a oasis 12) 2S Wete rete Slizg sos onee 48-8 GO — saabad 0) Wi) Sowsdc OD OO) “wevose Qe moovies 00°0 Ue os can 53 42 ...... IS} Gis) 4 serdeise ZOOS ewer ara 49°6 ay soos EEy eee) woomoar 16) WG" Goode WG) sabes 49-6 Wy apace Bes Gy Soo caod Lie fene® Myer ce opbeaonD 92 M CHAS, GMINTIE GS Soi bo nih 0 5. Cor eB enIREen ns > Obs Ure Deena 40-100 Mediterranean coast)of Africa. ... 5... .ccesesnenct cccerecves M Cape Sagras and Tangier Bay... 2.205.602 see ve ese ve ne ne 35 Seven miles from Rinaldo’s Chair ..............-000 20 ee eee 60-160 Caxtag eras barycrrajsttors ots ears a\ro\ stgiere's arnt spat auora eieks| so fosislaje ts M. Malia NeleeraphiCables sire cicpale- th eye. snacaie ate eqateliwisie's tisha tee « M 1 M. Mediterranean dredgings. 340 PROFESSOR P. M. DUNCAN ON THE SPECIES. No. 57. Lophohelia prolifera, Amphihelia oculata, Amphihelia ramea, Caryophyllia clavus, var. a, Caryophyllia clavus. 67. Amphihelia oculata, var. A.B. Guynia annulata, Paracyathus striatus, Caryophyllia clavus, var. C. T. Caryophyllia clavus. M. c. of A. Paracyathus striatus, Caryophyllia clavus, var. borealis, Caryophyllia clavus, var. smithi, Caryophyllia clavus, var. elongata, Desmophyllia crista-galli. Cape Sagras and Tangier Bay. Sphenotrochus intermedius. Seven m. R. C. Caryophyllia clavus, var. borealis and var. 8, Paracyathus striatus. Carth. Bay. Caryophyllia clavus, var. /3. M. T. C. Caryophyllia clayus, var. smithi. DESCRIPTION OF THE PLATES. PLATE XXXIX. Figs. 1-13. Flabellum distinctum, Ed. & H. Fig. 1. Part of the calice, magnified. Fig. 2. Front view of corallum. Fig. 3. Side view. Fig. 4. Calice. Figs. 5, 6, 7, 8, 9, 10, 12, 13. Front and upper views of specimens, showing variation in form. Fig. 11. Septa of Mlabellum laciniatum. Figs. 14, 15. Calicular end, nat. size and magnified, of Flabellum laciniatum. Figs. 16-18. Front view, nat. size and magnified. PLATE XL. Figs. 1-4. Caryophyllia abyssorum. ‘Coste, magnified. ico] a a) (08 bo oe Corallum, nat. size. Calice, magnified. Coste, magnified, of Caryophyllia cylindracea. Coste, lower end of, . Corallum Calice, magnified, - . Coste, magnified, of a variety of Caryophyllia cylindracea. Fig. 10. Lower end 7 a Fig. 11. Corallum, nat. size, % 93 22 Fig. 12. Calice, magnified, 7" ” 99 ky a eS 0 1 > GT OP ” Fig. 13. Corallum, nat. size, of Caryophyllia vermiformis. Fig. 14. Corallum, magnified, a Fig. 15. Coste, magnified, Fig. 16. Calice, magnified, a PLATE XLI. Fig. 1. Corallum, nat. size, of Sphenotrochus intermedius. Fig. 2. Corallum, magnified, - Fig. 3. Corallum, magnified, variety ; Fig. 4. Corallum, nat. size, variety is Fig. 5. Calice, magnified, S Fig. 6. Upper view of corallum of Sabinotrochus apertus. Fig. 7. Side view of corallum - Fig. 8. Calice, magnified, Fig. 9. Base, magnified, - Fig. 10. Fig. 11. a a Views of Desmophyllum crista-galli. Fig. 14. Fig. 15. Fig. 16. Section of the peduncle of a small specimen, magnified. PLATE XLII. Fig. 1. Lophohelia prolifera, var. gracilis. Fig. 2. Calices, magnified. Fig. 3. Coste, magnified, of Caryophyllia pourtalesi. Fig. 4. Corallum (young), nat. size, 3 Fig. 5. Same, magnified, s Fig. 6. Calice, magnified, 7 Fig. 7. Very young calice, magnified, _,, Fig. 8. Corallum, nat. size, i. Fig. 9. Coste, magnified, _ Fig. 10. Calice, magnified, 3 Fig. 11. Portion of corallum of Solenosmilia variabilis. Fig. 12. Calice, magnified, 3 Fig. 13. Costal ends, magnified, as Fig. 14. Calice, magnified, rf Fig. 15. Corallum * Fig. 16. Cost, magnified, - MADREPORARIA OF THE DEEP SEA. 341 342 PROFESSOR P. M. DUNCAN ON THE Fig. 17. Calice, magnified, of Solenosmilia variabilis. Fig. 18. Corallum (old) a PLATE XLII. Fig. 1. Corallum of Caryophyllia arcuata, variety. Fig. 2. Calice, magnified. Fig. 3. Coste, magnified. Fig. 4. Ornamentation of the septa, magnified. Fig. 5. Corallum of Paracyathus agassiz. Fig. 6. Its calice, magnified. Fig. 7. The costa, magnified. Fig. 8. Septum and pali, magnified. 9. Corallum of Paracyathus striatus. Fig. 10. Its calice, magnified. Fig. 11. The calicular termination, showing the tall pali. Fig. 12. A septum, magnified. Fig. 13. A corallum growing from the calice of a parent. rahe. ». | Corallites of Balanophyllia socialis. Fig. 15. Calicular view. Fig. 16. Calicular view, magnified. Fig. 17. View of fractured peduncle. Fig. 18. The cost and epitheca, magnified. PLATE XLIV. Fig. 1. Corallum of variety of Amphihelia ramea. Fig. 3. Another specimen, slightly magnified. Fig. 2. A calice, magnified. Fig. 4. Corallum of Caryophillia sequenze. Fig. 5. Its coste, magnified. Fig. 6. The calice, magnified. Figs. 7 & 8. Side views of unusual shapes of buds of Lophohelia prolifera. Fig. 9. Magnified stem, showing the increase of wall-thickness outside the visceral cavity. Fig. 10. Section of a corallite, showing tabule. Fig. 11. Magnified view of a calice. PLATE XLV. Fig. 1. Corallum of Amphihelia oculata. Fig. 2. Calicular ends, magnified. MADREPORARTIA OF THE DEEP SEA, 343 Fig. 3. Calice, magnified. Fig. 4. Amphihelia ramea. Fig. 5. Ornamentation of the corallites, magnified. Fig. 6. Calice magnified. PLATE XLVI. Fig. 1 Fig. 3. Fig. 6. Fig. 8 Fig. 11. Fig. 14. Fig. 17.) The other figures represent magnified views of branches, of ornamentation and calices. | emt of varieties of Amphihelia ramea. PLATE XLVII. Figs. 1 & 5. View of corallum of a species the genus of which is doubtful. Fig. 2. Side view, magnified. Fig. 3. Calicular surface, magnified. Fig. 4. Base, magnified. Figs. 6 & 7. Base and calice of another specimen, magnified. Fig. 8. Coste, magnified. Fig. 9. Fig. 11. } Corallites of Guynia annulata. Figs. 11 & 13. On shells. Fig. 13. ) lpi : Fig. 12. Side view, magnified. Fig. 14. Corallum adherent by its side, magnified. Fig. 15. Fig. 1 oy Calice, magnified. Fig. 17. Group of Rhizotrochus affinis. Fig. 18. Corallum, slightly magnified. Fig. 19. Calice, magnified. PLATE XLVIUII. Fig. 1. Corallum of Bathycyathus atlanticus. Fig. 2. Calice, magnified. Fig. 3. Corallum of Caryophyllia cyathus, variety «. Fig. 4. Calice, magnified. Fig. 5. Corallum of Caryophyllia clavus, var. exserta. Fig. 6. = yar. borealis. 3” VOL. VIII.—PART VY. March, 1873. 3c 344 ON THE MADREPORARIA OF THE DEEP SEA. Fig. 7. Corallum of Caryophyllia clavus, var. elongata. Fig. 8. Calice magnified. Fig. 9. Corallum of Caryophyllia clavus. Fig. 10. Calice, magnified. Fig. 11. Corallum of Caryophyllia clavus, var. smithit. Fig. 12. Calice (worn), magnified. Fig. a ee Fig 14 Corallum of Caryophyllia clavus, var. epithecata. = Fig. 15. Corallum of an aberrant specimen of var. epithecata. Fig. 16. Calice, showing a circular outline. PLATE XLIX. Figs. 1-6. Corallites and details of Stylaster gemmascens. Figs. 8-10. 3 3 af. Figs. 13-16. __,, $3 5 Fig. 7. Pliobothrus symmetricus. Fig. 11. Balanophyllia cellulosa, corallum of. Fig. 12. Corallum, greatly magnified. Fig. 16. Fig. 17. Fig. 18. The calice magnified. Fig. 19. Coste with synapticule, magnified. }corattum of Fungia symmetrica. From nat on Stone by J. Erzleben M&N Hanhart inyp . DEEP SEA CORALS, 4 2 if Frommat.on Stone by J Erxleben ML&.N Hanhart imp DEEP SEA CORALS De Wilde hith ial SEA M&N.Hanhart ump M&.N.Hanhart rap From nat on Stone byJ Exdeben er From nat on Stone by J Exxdeb: MENEashart sep De Wilde ith. SEA CORALS DEEP fromraton Stone byJ.Erdeben M&NHanharb imp DISIEI2 Sisva (C(OMRYAMES) From nat. on Stone by J Ercleben DEI SEA cess ae Seb PORTERS Pa, MAN. Hanhart laip , Py De Wilde & bee eo, eee lith.. 5 EEP SEA GORALS. M&N-Hanhart amp M&N lamhart + ip De Wilde lith M&N Hanha : ae > hi: + ~- 7 ” - ! P] - i] y t ' #- by: ‘ i) = fh ; ‘ = ui ra From nat on Stone by J. Ercleben [$845] IX. On the Osteology of the Marsupialia. (Part III.) Modifications of the Skeleton in the Species of Phascolomys, By Professor OwEn, F.R.S., F.Z.8., &e. Read 21st November, 1871. [Puates L. to LVIT.} § 1. Introduction. MY first Paper on this subject was devoted to the osteological characteristics of the Marsupialia generally’; the second to the specific characters afforded by the skeleton, as illustrated by certain cranial ones in the genus Phascolomys*. But subsequent acqui- sitions of fossil remains of Australian marsupials have impressed me with the inade- quacy of the facts and figures communicated in these papers to the requirements of the paleontologist. Fellow-labourers in the field of recent osteology, such as my esteemed colleague Dr. Gray, F.R.S., Prof. McCoy of the Melbourne University, and more especially Dr. James Murie, F.Z.S., F.LS., have contributed valuable observations on the osteology of the Wombats, whereby three existing species, Phascolomys vombatus, Geof.*, P. platyrhinus, Ow.*‘, and P. latifrons, Ow.*, seem to be well established by cranial and dental characters. But the illustrations of these—the more requisite through the difficulty of acquisi- tion of skeletons of the rarer marsupial quadrupeds of Australia and Tasmania, espe- cially by foreign museums—are very much below the needs of this limited field of scientific research. Dr. Murie repeats, by original drawings from later acquired specimens, the upper views of the cranium of the Tasmanian and latifront Wombats, given in my second Memoir, with the addition of a similar view of the cranium of the platyrhine species.” But these three figures are reduced to half the natural size; and experience of much perplexity and loss of time in the attempt to apply such reduced views to the deter- mination of fragmentary fossils has led me more than once to denounce that seeming economy, where the size of originals is not such as to preclude representation in full. I am therefore led to believe, and, indeed, encouraged by the uniform liberal recep- tion by this Society of my contributions to the science for the promotion of which the Society was founded, that the illustrations accompaying the present Paper or ‘ Part,’ may be permitted to appear in the ‘Transactions.’ That they will be acceptable to 1 Trans. Zool. Soc. vol. ii. 4to (1841), p. 379. > Tbid. vol. iii. (1845), p. 303. * Syn. Didelphys ursina, Shaw. * Syn. Phascolomys angasii, Gray ; Ph. niger, Gould; Ph. setosus (?), Gray. 5 Syn. Phascolomys lasiorhinus, Gould ; Lasiorhinus M‘Coyi, Gray. ® Proceedings of the Zoological Society, 1865, p. 844. (See this excellent paper for other synonyms.) VOL. VIII.—PaRT VI. May, 1873. 3D 346 PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. zoologists generally is the more probable since the grand ‘Ostéographie’ of De Blainville was interrupted by the regretted demise of the indefatigable author before it had reached his subclass DipELPHYs'’. § 2. Cranial Characters of Phascolomys. The characters of the skull of Phascolomys platyrhinus, briefly defined in the ‘Descriptive Catalogue of the Osteological Series in the Museum of the Royal College of Surgeons’’, have subsequently been determined by Prof. M‘Coy*® and Dr. Murie? to be those of the continental or Victorian bare-nosed, large, brown or black Wombat. A skull of this species, figured of the natural size in Plate LI. fig. 1, gives a length, from the hindmost ridge of the occiput to the front border of the incisor-alveoli, of 8 inches. Another skull in the British-Museum collection exceeds this only by one line ; a third is in the same slight degree smaller. The specimen submitted to me by Dr. McBain of Edinburgh, in 1855, yielded a length of 8 inches; two other specimens have the length of 7 inches 8 lines (PI. L. fig. 1), and 7 inches 6 lines (PI. LIL. fig. 1). Dr. Murie gives minor dimensions of some, probably female, specimens’. The occiput (Plate L. fig. 1, 2, 3, & fig. 2) rises vertically from the foramen magnum at the median line, but curves a little backward laterally, where it forms the sides of the broad superoccipital (ib. fig 2,3). The lower and lateral parts of the occiput are formed by the exoccipitals (ib. fig. 2, 2), the mastoids (8 ), and squamosals (27). The occiput is higher in proportion to its basal breadth than in Phascolomys vombatus (ib. fig. 3); it is more quadrate in form ; it does not curve upward and inward so regularly from the mastoid processes (8) to the summit, asin Phascolomys latifrons (ib. fig. 4). The composition of the occipital region agrees with that illustrated in Pl. Ixxi. fig. 6 (Trans. Zool. Soc. vol. ii.), in Phascolomys vombatus. But the portions con- tributed by the squamosals (Pl. L. fig. 2, 27,27), do not reach so low down upon the mastoid process as in Phascolomys vombatus (figs. 3, 6, 27,8). The basioccipital (Pl. LIL. fig. 1,1) contributes about half an inch of the thick lower border of the foramen magnum. The exoccipitals (Pl. L. fig. 2,2) form the lateral borders, de- veloping there the condyles; and the superoccipital completes the middle of the upper border, which is sharp; and as ossification of the latter element does not usually extend so low down as to fill up the whole interspace left by the exoccipitals, the 1 The ‘ Prospectus’ of this work was issued after the reading of my first memoir on the osteology of the Marsupialia; see Zool. Trans. vol. ii. p. 379. The character and form of the work, especially the richness and beauty of its illustrations, exemplified in the Part which appeared after the communication of my second memoir, led me to abandon that subject, as being likely to meet with more complete illustration in the ‘ Ostéographie du Squelette et du Systéme Dentaire des cing Classes d’Animaux Vetébrés récents et fossiles,’ 4to, Paris, 1840-1860. 2 4to, 1853, vol. i, p. 334. no. 1841. + Transactions of the Royal Society of Victoria, vol. viii. (1868), p. 267. * Proceedings of the Zoological Society of London, Dec. 1865, p. 838. * Thid. p. 845, PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS, 347 foramen presents a somewhat trilobate or trefoil figure, as noticed in Phascolomys vom- batus, in my first memoir. Since the date of its communication specimens have come to hand which show that this character is not constant in either species, although the exceptions are rare. In fig. 5, Pl. L. is shown one of these exceptions in Phascolomys platyrhinus, and in fig. 6 another in Phascolomys vombatus. The foramen magnum has always a greater trans- verse than vertical diameter; and the exceptions to the trefoil figure show it to be transversely elliptical, as in figs. 5 & 6. In Phascolomys latifrons this is the common form of the foramen magnum, with the ellipse more depressed, as in fig. 4. But of this species I have received the skull of a young, though nearly full-grown, animal, in which the sutures between the ex- and superoccipitals (Pl. L. fig. 7, 2,3) are not obliterated, and in which an unossified space or emargination exists between the exoccipitals below the superoccipital. In two skulls of Phascolomys latifrons the occipital part of the mastoid (8 fig. 7) ascends above the process (8'), of nearly uniform breadth, to the superoccipital (3 ), and separates the exoccipital (2) from the squamosal (27). In a third this character obtains on one side; on the other side it is obscure. In Phascolomys platyrhinus (ib. fig. 2) and Phas- colomys vombatus (ib. fig. 6) the upper part of the exoccipital (2) usually articulates more extensively with the squamosal (27). The characters afforded by the upper surface of the skull differentiate the platyrhine as strongly as they do the Tasmanian Wombat from the latifront species’; the differences shown in this respect between the platyrhine (Pl. LI. fig. 1) and Tasmanian Wombats are less easily seized. In all the skulls I have yet seen of both species the temporal fossee (ib. fig. 1, 7,27) approach each other more nearly, absolutely as well as relatively, in the larger continental species”; yet this character may be shown to be exceptionally affected in an aged male Tasmanian Wombat. Size, of course, is of itself a character of variety. The upper third of the fossa is formed by a longitudinal strip of parietal (PI. LI. fig. 1 & Pl. LIT. fig. 1,7), and this bends down from the temporal ridge more abruptly in Phascolomys vombatus than in Phascolomys platyrhinus; while in the latter a slight rising or ridge is developed from the line of the parieto-squamosal suture. In Phascolomys latifrons the temporal ridges are less marked, and the parietals incline therefrom more outwardly to join the squamosals. The feeble indication of the postorbital process (ib. 12), and the well-defined lacrymal tubercle (ib. 73) defining the fore and upper part of the orbit, are common to both the bare-nosed species, and distinguish them from Phascolomys latifrons. We come next to compare the nasal bones (PI. LI. figs. 1, 3, 4, 5, 15) in regard to shape, size, and connexions. 1 Compare Trans. Zool. Soc. vol. iii. pl. xxxvii. fig. 1 (Phascolomys vombatus) and fig. 4 (Phascolomys latifrons) with fig. 1, Pl. LI. of the present paper. ? Compare fig. 1, tom. cit. with fig. 1 in Pl. LI. of the present paper. 3D2 348 PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. In an old male Tasmanian Wombat their basal breadth equals three fourths of their length (Pl. LI. fig. 4)'. The outer angles of the base are divided from the lacrymal tubercle (73) by a fronto-maxiilary suture (11-21) 3 lines in breadth. The sides of the pair of nasals converge forward at the hinder third, then run parallel, gently curving inward, and finally run to the margin of the nostril, with a slight curve outward. Thus the course of each lateral border of the nasals is undulate. Their tips extend forward about 3 lines in advance of the naso-premaxillary suture (15-22), and are bevelled off to an obtuse point from without obliquely inward and forward. Together they form the middle third of the upper border of the bony nostril. The frontals (ib. 11) make a slight projection into the middle of the fronto-nasal suture (11-15), which from this shallow indent runs outward and a little forward to the nasal process of the maxillary (21). The naso-maxillary suture (15-21) forms the hind fifth part of the lateral border of the nasals ; the naso-premaxillary suture (15-22) runs along the rest of the extent of the nasal bones, i.e. to the beginning of their free ends, which are short and subobtuse. In a second Wombat the nasals differ in their breadth being equal to three fourths of their length, or as 75 to 100, in the absence of any median indent of the fronto-nasal suture, and in the sharper convergence of the hinder fourth part of the lateral margins. These margins describe a similar wavy course, convex outward along the middle, or a little in advance of it. The apices overhanging the nostril are sharper and more pro- minent than in the last specimen. In a third, somewhat younger Phascolomys vombatus (Pl. LI. fig. 3), the lateral margins converge more gradually, and in an almost straight line from the base to the anterior fourth of the nasals, where the margins extend nearly straight to the nostril. The middle sixth part of the fronto-nasal suture (11-15) is nearly straight or transverse ; the rest extends outward and more obliquely forward than in the two preceding spe- cimens. The fronto-maxillary suture (11-21) is 4 lines in extent. The nasal apices projecting beyond the premaxillo-nasal sutures are sharp and form one fifth the length of the whole lateral margin. The basal breadth bears a greater proportion to the length of the nasals than in the first-cited skull. Phascolomys platyrhinus (P1. LI. fig. 1), in the shortness of the naso-maxillary suture (ib. 15-21) and the deep emargination of the fore part of the nasal process of the pre- maxillary (ib. 22), is more nearly allied to Phascolomys vombatus than either of these are to Phascolomys latifrons®. But the nasal bones of Phascolomys platyrhinus are relatively broader than in Phascolomys vombatus; the outer basal angles approach nearer to the 1 This proportion is expressed as follows by Dr. Murie in describing his specimen of Phascolomys vombatus. “The proportional breadth of the two nasal bones at their hinder ends is to their length, 68 to 100,.”— Proc. Zool. Soc. June 27, 1867, p. 802. ? This relation is well pointed out by Dr. Murie, who remarks that ‘‘ Phascolomys latifrons shears off from the common form of Wombat, and reverts to the true marsupial type in several particulars.”—Loc. cit. p. 800. He does not, however, exemplify his ideal type: the postorbital processes are almost peculiar to Ph. latifrons among Marsupials. PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. 349 lacrymal tubercles, with a greater relative breadth of the skull at that part. In two skulls the lateral borders of the nasals have the same undulatory course as in Phasco- lomys vombatus, but more feebly marked. ‘There isa narrow and irregular intrusion of the frontal at the middle of the fronto-nasal suture (11), sometimes at the expense of the right, sometimes of the left nasal bone. The breadth of the base equals five sixths of the length of the nasals in two specimens, and four fifths in a third. The apices pro- jecting anterior to the naso-premaxillary suture are short and blunt, as usual in Phasco- lomys vombatus. The width of the nasals at their base, or fronto-nasal suture, begin to diminish at once as they advance by the converging course of the naso-maxillary and naso-premaxillary sutures. In not one of the three specimens before me is “ the width of the nasals continued forward beyond their middles”*. In Phascolomys latifrons, where the breadth of the fore part of the frontals is made to contrast with the narrowness of the rest of the bones by the outward extension of the postorbital processes, the nasals (Pl. LI. fig. 5, 15) present a more regular triangular form through the prevailing transverse course of the fronto-nasal suture (11-15), and the more regular convergence of the lateral margins of the nasals to the fore ends of the naso-premaxillary sutures (15-22). Beyond these the lateral margins converge more rapidly to the apices, which extend freely forward further than in the two preceding species. The breadth of the nasals at the base of their free extremities is greater than in the other two Wombats, and the upper surface is flatter. In one of four skulls before me of Phascolomys latifrons the left frontal breaks the transverse course of the fronto-nasal suture by a short pointed process or wedge between the two nasals; in a second skull the right frontal sends forward, in the same way, a more obtuse triangular process; in the type skull (Zool. Trans. vol. iii. pl. xxxvii. fig. 4), both frontals contribute equal shares to the wedge; in two later-acquired spe- cimens the suture is uninterrupted. The transverse course of the fronto-nasal suture is constant and characteristic of Phascolomys latifrons. Moreover the suture (11-15) is not separated from the lacrymal (73) as it is in Phascolomys platyrhinus and Phascolomys vombatus, by the maxillary (21); or if, as in an exceptional skull, or on one side, the maxillary goes back as far as the frontal, it is by a very narrow strip. The extent of the naso-maxillary suture (ib. fig. 5, 15-21) equals, in Phascolomys lati- frons, that of the naso-premaxillary suture (15-22). These differences in the connexions of the nasals are more significant of specific distinction than is the shape of the bones. The naso-maxillo-premaxillary suture is very slightly concave outwardly; the free border of the nasal, beyond the suture, is of greater extent in the hairy-nosed than in the bare-nosed Wombats. There would be no doubt in determining Phascolomys latifrons by the naso-maxillo- premaxillary part of the skull, at least as being distinct from the other two known recent species, if even the still more characteristic part of the frontal bones were want- ing. There might be more difficulty in pronouncing from the same part of the skull 1 Murie, loc. cit. at p. 803. 390 PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. of a bare-nosed Wombat, as to whether it belonged to Phascolomys platyrhinus or Phas- colomys vombatus. The basal view of the skull of Phascolomys platyrhinus (Pl. LII. fig. 1) shows one of the degrees of variation to which the basioccipital is subject in the depth and breadth of the emargination of the part contributing to the lower border of the foramen magnum. In this skull the emargination is narrow and deep; in a second skull of like size it is less deep and wider. The basioccipital is subject to the same variety in Phascolomys vombatus. In two skulls of Phascolomys latifrons the emargination of the lower border of the foramen magnum is wider, as in the subject of fig. 1, Pl. LIII., than in the two first- named species. In all the three species, the under surface of the basioccipital has a median longitudinal ridge, which slightly varies in its depth and sharpness; there is a shallow vacuity on each side of the ridge. Each exoccipital, where it coalesces with the basioccipital, develops a tubercle, which, in the platyrhine (Pl. LII. fig. 1, 6) and Tasmanian Wombats, abuts against the petrosal. In Phascolomys latifrons (Pl. LIII. fig. 1, 6) the corresponding (exoccipital) tubercles are more prominent and project freely below the petrosals (16), resembling the pterapophyses of the basisphenoid in birds. The exoccipital is perforated anterior to the condyle by, commonly, two hypoglossal foramina; these are more equal in size in Phascolomys latifrons than in the other two species. There is usually a small vascular foramen external to the upper end of the condyle. The wedge-shaped petrosal (ib. fig. 1, 16) abuts against the side of the basioccipital, with the thin end directed forward. The squamosal (ib. 27) expands at the inner side of the mandibular articular surface (g) to form a tympanic cell or ‘ bulla,’ which is large and widely open backward, receiving the inner orifice of the tympanic (ib. 28) in Phascolomys platyrhinus and P. vombatus. In Phascolomys latifrons this pre- or antetympanic cell of the squamosal (Pl. LIII. fig. 1, 27) is smaller than in the Tasmanian Wombat, much smaller than in Phascolomys platyrhinus. External to this cell the squamosal develops, in the bare-nosed Wombats, a vertical ridge, which is wedged into a groove of the tympanic; it is scarcely marked in the hairy-nosed Wombat. In Phascolomys latifrons (Pl. LIII. fig. 4) the articular bar of the squamosal (g) is relatively shorter than in the platyrhine or Tasmanian Wombat; and its inner end is notched posteriorly, which receives and is reciprocally received by a notch in the fore and outer part of the tympanic (28). This bone sends forward a thick triangular plate, contracting to the part which is notched for the squamosal, in a way which offers a close and interesting analogy to the ‘ gomphosis’ of 28 with 27 in Birds. The more marked division of the supratympanic cell in Phascolomys latifrons (ib. fig. 4, 7, m), and the greater size and depth of the anterior portion or cavity (ib. /), are noted in the cranial characters of the species, and have been found in all the skulls examined since the date of my second Memoir’. ‘ Trans. Zool. Soe. vol. iii. (1845), p. 303. The “ enormously excavated supratympanic cells ” of Ph. latifrons, are noticed by Dr. Murie, loc. cit. (1865), p. 842. PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. 301 In Ph. platyrhinus (Pl. LIII. fig. 6) and Ph. vombatus the two divisions of the supra- tympanic cell are more equal, and the ‘ gomphosis’ of the squamosal with the tympanic is less marked. The malar part (Pl. L. fig. 1, 26) of the zygoma defines the orbit pos- teriorly by an angular process in both the bare-nosed Wombats; it is not developed in the hairy-nosed species, where the hind boundary of the orbit is indicated by the postfrontal process (Pl. LIII. fig. 1, 12) which is not present or is rudimental in the others. The numerous irregular venous foramina of the squamosal (Pls. L., LI. 27) are notable in all the species of Wombat. The superior maxillary sends outward and backward a process (Pl. LII. fig. 1, 21*) which contributes to and strengthens the anterior pier (ib. 26) of the zygomatic arch; it projects above the alveolus of the third and fourth molar teeth, at a greater height above the molars in Ph. latifrons (Pl. LVI. fig. 5, 21*) than in Ph. platyrhinus (ib. fig. 4) or in Ph. vombatus. Ina skull of the larger bare-nosed species, with an upper molar series 2 inches 2 lines in extent, the process rises 7 lines above the outlet of the third molar alveolus. In the skull of a Ph. latifrons with a molar series 1 inch 11 lines in extent, the process rises 10 lines above that outlet. In other words, the outer alveolar plate of the maxillary is deeper, below the zygoma, in the hairy-nosed than in the bare- nosed Wombats: it is, also, more nearly vertical, less inclined mesiad as it descends to the outlets of the sockets. In my first memoir I state that “Phascolomys resembles Phascolarctus and Hypsi- prymnus in having the posterior palatal openings large and situated entirely in the palatal bones,” and that “ posterior and external to these are two small perforations” !. In the other two species (Phascolomys platyrhinus and Ph. latifrons), determined since the date of that remark, which was made on the skull of Ph. vombatus, the generic cha- racters of the postpalatal apertures are repeated. In the skull of Ph. vombatus, figured in Pl. lxxi. fig. 6, of the 2nd volume of the ‘ Zoological Transactions,’ these apertures are oval, the base, which is behind, being rounded; but the small end of the oval is so nearly pointed as to suggest the term ‘triangular.’ In two skulls of full-grown Tasma- nian Wombats, since compared, these foramina present the same shape and proportions (Pl. LII. fig. 4, 6); in two smaller and younger skulls of Ph. vombatus they are rela- tively smaller, and are rather elliptical than oval. In two skulls of Ph. platyrhinus, in the British Museum, the postpalatal apertures are longitudinally elliptical or oblong in one (ib. fig. 1, 6) and are triangular in the other: in both, the apertures extend more forward and come near to the transverse parallel of the middle of the hindmost socket (ib. ib. m3). In three skulls of Ph. latifrons the postpalatal foramina (PI. LII. fig. 5; Pl. LIII. fig. 1, 4) are relatively larger, especially longer, than in either the Tasmanian or platyrhine Wombat, and are rounded anteriorly but less broad there than behind: they advance nearer to the fore part of the last molar alveolus, or reach * “On the Osteology of the Marsupialia,” Trans. Zool. Soc. vol. ii. p. 389. 352 PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. that transverse parallel. Dr. Murie also notes the larger size of these foramina in the skulls of Ph. latifrons which he compared with those of the platyrhine species’. The intermolar part of the bony palate is more contracted anteriorly in the bare-nosed (Pl. LIT. figs. 1, 4, 21,21) than in the hairy-nosed Wombats (ib. fig. 5 21; Pl. LIT. fig. 1, 21); the narrowest part is between the right and left second molars in Ph. platyrhinus and Ph. vombatus; but in the Ph. latifrons there is little difference in the interval between the second (d4) and that between the first (d3) molars. The palate is slightly arched transversely, and is almost flat in some platyrhine Wombats. In the latifront species it is always more vaulted, or deeper when viewed from below, at the diastemal part (Pl. LIII. fig. 1, 21), especially at the prepalatal openings (ib. a, a) than in the bare-nosed species; and the diastemal tract is more sharply defined, laterally, in the hairy-nosed Wombat. The antorbital vacuity is wider, less slit-shaped, in Ph. lati- frons than in Ph. platyrhinus (Pl. L. fig. 1, 21) and Ph. vombatus. The cerebral cavity in Ph. latifrons (Pl. LIII. fig. 5) is, in length, 3 inches 5 lines, and equals one half the length of the cranium, which is 6 inches 9 lines in the specimen figured. The cavity is divided by two vertical ridges into the epencephalic (ep), prosencephalic (pr), and rhinencephalic (7/) compartments. The petrosal (16) contributes a vertical sur- ace to the side of the epencephalic compartment, which is overarched by the superocci- pital and parietal. The acoustic foramen or fossa is subcircular below the horizontal ridge bisecting the cranial surface of the petrosal. The cerebellar pit above the ridge is wider and more shallow than in Thylacinus and most other Marsupials’. The ‘ vagal’ foramen (ov) is between the petrosal and the exoccipital. Behind this, over the two or three inner orifices of the precondyloid canals (p), the foramen ovale (ov), 33 lines in long diameter, perforates the alisphenoid below the tentorial ridge (0'), and opens out- wardly anterior to the excavation forming the fore part of the tympanic cavity. A groove runs forward from the foramen ovale to the foramen rotundum about 8 lines in advance: at the inner side of this groove the basisphenoid is perforated by the ento- carotid. From the inner orifice of this canal a groove runs forward to the common prelacerate and optic fissure (op). The rhinencephalic fossa (rh) is 7 lines in vertical diameter, 5 lines in longitudinal extent. It has a large perforation at its lower and back part leading to the common orbito-temporal fossa; the floor and fore part of the fossa are perforated by the smaller foramina transmitting the olfactory nerves to the ethmo-turbinals. ‘The nasal cavity is divided by a vertical septum extending from the rhinencephalic cavities to 8 lines behind the tips of the nasals. The anterior border of the septum is vertical, 1 inch in depth. In Ph. platyrhinus the bony septum terminates 14 lines behind the tips of the nasals. The alisphenoid tentorial ridge is less developed in that species. The ethmo-turbinals (¢) form a mass about an inch in antero-posterior and less in vertical diameter. The mid turbinal is elongate, developed from a ridge de- scending from the roof of the nasal cavity, a short distance external to the septum, and * Proc, Zool. Soe. 1865, p. 848. 2 Loe, cit. p. 390. PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. 353 defining a longitudinal canal traversing the upper part of the nasal meatus. The pre- turbinal rises from the side of the floor of the chamber. The frontal is excavated by three sinuses above the rhinencephalic chamber and the hind part of the rhinal or nasal one. ‘The roof of the cranium, at the section, has a close cancellous structure 5 lines thick above the prosencephalon, 11 lines thick above the epencephalon. § 3. Mandibular Characters of Phascolomys. In differentiating from cranial characters the species of Wombat called Phascolomys latifrons, I noted, in comparing it with Phascolomys vombatus, that “The curve of the lower border of the lower jaw is deeper” (Pl. LIV. fig. 3); “the inner angle of the condyle” (ib. fig. 5, 6) “is less produced ; the coronoid process” (ib. fig. 3, ¢) “is higher and narrower, and the postsymphysial depression is almost obsolete”. With the exception of the latter particular, which is variable in both species, subsequently acquired skulls have confirmed the constancy of the above characters. They likewise serve to differentiate the mandible of Ph. latifrons from that of Ph. platyrhinus, save that the coronoid process rises higher in the platyrhine species (Pl. LIV. fig. 2, c) than in the Tasmanian Wombat (ib. fig. 1, c); but the broader proportion of the process, as compared with that in the hairy-nosed Wombat (ib. fig. 3, c) is retained. The deeper curve described by the lower contour of the jaw from the neck of the condyle to the incisive alveoli is a constant and well-marked character of Ph. latifrons (ib. fig. 3). So, likewise, is the less-produced inner angle of the condyle (Pl. LIV. fig. 5, 6, and Pl. LVIL. fig. 3,4). In both the Tasmanian (Pl. LV. fig. 5, 6) and platyrhine (Pl. LIV. fig. 4, 3) Wombats this angle is more produced and deflected. The diastemal part (Pls. LIV., LV., & LVIL., /, s') of the long symphysis (ib. s, s') is subject to some variety in existing Wombats. In two mandibles of Ph. platyrhinus, in which the length of the series of molar alveoli is 2 inches 3 lines, that of the interval between the first alveolus and the foremost angle of the symphysis is, in one skull 1 inch 7} lines (Pl. LVII. fig. 1, s’), in the other 1 inch 6} lines; the breadth of the diastema midway is the same in both, viz. 10 lines. In a mandible of Ph. latifrons with the molar series of alveoli 2 inches in extent (Pl. LVII. fig. 3), the diastema (J, s'), taken as above to the foremost point at the interspace of the incisors, is 1 inch 3 lines. In a second mandible, with the molar alveoli 1 inch 104 lines in extent, that of the diastema is also 1 inch 3 lines; the breadth of the diastema midway is, in the first mandible 8 lines, in the second 7 lines. In the two mandibles of the platyrhine Wombat compared, the diastema is slightly convex both lengthwise and across; it is traversed by a pair of shallow longitudinal grooves (ib. fig. 1), and is not sharply defined from the sides of the symphysis. In a third mandible of the same species the defining ridges are better marked, the trans- * «On the Osteology of the Marsupialia,”—Part II. (1845) in Trans. Zool. Soc. vol. iii. p. 304, pl. xxxvii. figs. 2 & 5. VOL. VII.—PART VI. May, 1873. 35 354 PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. verse convexity is less so, and this part of the symphysis is rather longer and narrower than in the other two mandibles. In these respects the third mandible approaches nearer to Ph. latifrons. But it differs, as do the other mandibles of the same species, as well as those of Ph. vombatus (ib. fig. 2, 7,5), in the larger, especially broader incisive alveoli, and the oblique course of their upper margins from the mid line of the sym- physis outward and backward. The fore end of the symphysis of Ph. latifrons is at once recognizable by the narrower outlets of the incisive alveoli, and the more trans- verse course of their upper border (Pl. LVII. fig. 3). The lateral borders of the outlets are also more nearly vertical (Pl. LIV. fig. 3, s'), and do not slope backward as they descend, like those of the incisor alveolar outlets in Ph. platyrhinus (ib. fig. 2) and Ph. vombatus (ib. fig. 1)*. With the narrower alveoli associated with the more compressed form of the incisors of Ph. latifrons one may predicate a generally narrower diastemal part of the sym- physis, the upper surface of which, with a median canal towards the end and the two parallel longitudinal grooves obsolete or nearly so, is better defined from the sides of this part of the symphysis (Pl. LVII. fig. 3, 7, s'). In one jaw of Ph. latifrons the defining ridges are sharp, and the intervening upper surface is concave transversely to near the incisive outlets, where the defining ridges subside. I may note that the ante- rior outlet (v) of the dental canal, in three mandibles of Ph. platyrhinus (P|. LIV. fig. 2), is 1 inch 4 lines, or 1 inch 5 lines, behind the foremost point of the symphysis. In one mandible of Ph. latifrons (ib. fig. 3) it is 1 inch behind the fore end of the sym- physis; in another mandible it is 10 lines from the same part. The foramen is more anteriorly situated in the broad-fronted or hairy-nosed species*. I may further note that, in the mandibles of two individuals, examined since describing that of the type skull of Ph. latifrons, the intercommunicating foramen from the entry of the dental canal to the outer surface of the base of the coronoid is smaller in one (PI. LV. fig. 4, p), as in the type mandible, than in the platyrhine and Tasmanian Wombats; while in another mandible of Ph. latifrons it did not exist. It is interesting to find this variety, because, in the great Diprotodon and Notothere, with some affinities to Phascolomys, the absence of the perforation of the base of the coronoid process is the rule, as in the Marsupialia generally, and this supports Dr. Murie’s view of the hairy-nosed Wombat. The mandible of the Tasmanian Wombat has the singular proportions of being as broad as it is long. In the specimen figured (Pl. LIV. fig. 1) the length is 4 inches 6 lines. In one jaw of Ph. platyrhinus the length is 6 inches, the breadth 5 inches 1 This latter character, differentiating Phascolomys vombatus from Ph. latifrons, is shown in figs. 3, ¢ & 7 ¢ of pl. xxxyii. tom. cit. 2 This character is shown in the figures of the mandible of the Tasmanian and continental broad-fronted Wombats in pl. xxxvii. of my “Second Memoir” (tom. cit.). But I could not then, as now, depend upon the constancy of such character, in which the platyrhine continental Wombat, as usual, resembles the Tasmanian species. PROFESSOR OWEN ON THE SPECIES OF PHASOOLOMYS. 855 9 lines. In a second specimen, with a length of 5 inches 9 lines the breadth is 5 inches 3 lines. In a Ph. datifrons with a mandible 4 inches 10 lines in length, the extreme breadth is 4 inches 4 lines. In a second specimen the mandible, with a length of 5 inches, has a breadth of 4 inches 7 lines (Pl. LVI. fig. 3)*. The subsymphysial foramina (Pl. LVI., 7, 7) are usually closer together in Phascolomys platyrhinus (fig. 1) than in Phascolomys latifrons (fig. 3). § 4. Dental Characters of Phascolomys. The dentition of Phascolomys is unique in the mammalian class; its formula is well known ; every tooth enjoys unceasing growth, is consequently rootless, and is curved in the segment of a circle, it has also a partial investment of enamel. This is not the case with the molars of any rodent, nor of any other marsupial save the Wombats. In the upper molars (Pl. LII. fig. 5, e, Pl. LIII. fig 3, e) the enamel is laid upon the inner, in the lower molars (Pl. LVII. figs. 6 & 7) on the outer side. It is coated, with the rest of the tooth’s surface, by cement. The upper incisors in Phascolomys platyrhinus (Pl. LII. fig. 1, 7, figs. 2, 2') are sub- compressed, with a narrow subelliptic transverse section (fig. 2'), of which the long axis is directed from before rather outward and backward. Those of Phascolomys lati- frons (Pl. LIII. fig. 1, 7, figs. 2, 2'), with a fuller subelliptic section, have the long axis directed from before more outwardly as it extends backwards. ‘The pair of teeth present a broader surface forward; in Phascolomys platyrhinus they present a broader surface outward,—the surface in both species being the convex enamelled one; the enamel in both, also, is longitudinally striate. A larger proportion of the upper incisor is coated with enamel in Phascolomys platyrhinus (P\. LII. fig. 2') than in Phascolomys latifrons (Pl. LIII. fig. 2’). Commencing in the former near the hind border, it extends along the outside, over the front border, and along the inner side to a longitudinal indent, three fourths of the way toward the hind border. The enamel in Phascolomys latifrons, commencing at the hind border, which is, rather, the outer one, extends over the fore part and upon the inner surface to where it bends to form the hind surface. The extent of this hind part of the tooth, coated only with cement (Pl. LIII. fig. 2’, ¢), is greater in Phascolomys latifrons than the unenamelled tract in Phascolomys platyrhinus. The first upper molar in Phascolomys platyrhinus (P1. LII. fig. 1, & fig. 3, d 3) presents a subtriangular transverse section or working-surface, the base being backward, the apex forward ; the inner side is shorter than the outer side, and is indented near the apex by a groove traversing the tooth lengthwise; the outer angle of the base is sometimes, by oblique attrition, produced. The enamel begins anteriorly at the outer part of the apex (fig. 3, ¢), opposite the groove, is continued inward and backward, and upon the base two thirds of the way toward the outer angle, which, with the outer side of the tooth to near the anterior angle or apex, is coated only by cement (ib. c). 1 In the type specimen (tom. cit. pl. Ixxi. fig. 6, d) it is 4 inches 8 lines. 3E2 396 PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. In Phascolomys latifrons the worn surface of the first upper molar (Pl. LIII. figs. 1 & fig. 8, d 3) is subquadrate; or, if viewed as triangular, the antero-internal surface forms the base, and the postero-external angle the obtusely truncated apex. ‘There is a feeble indication of a longitudinal groove, representing that in Phascolomys platy- rhinus (Pl. LIT. fig. 3, g); but it does not mark off an anterior production of the tooth in the bare-nosed species, it simply indents the base or shorter side of the triangle. The enamel extends from the base upon the fore and outer part (Pl. LIII. fig 3, e) and upon the hind part of the tooth, leaving about the same extent of the outer and hinder part unenamelled and coated with cement (c) as in Phascolomys platyrhinus. The chief distinction is the deeper antero-internal longitudinal groove (Pl. LI. fig. 3, g) marking off a more definite anterior angle or lobule of the grinding surface of d 3 in both the bare-nosed Wombats; by which character a detached tooth might be deter- mined as not belonging to the hairy-nosed species. ‘There is only some difference in size between d 3, upper jaw, of Phascolomys platyrhinus and that of Phascolomys vombatus. The second molar (d 4) assumes a greater proportional size to the first (d 3) in Phas- colomys platyrhinus than in Phascolomys latifrons ; and the succeeding molars repeat the same degree of superiority of size. The longitudinal extent of the upper molar series in Phascolomys platyrhinus averages 2 inches 1 line; in Phascolomys latifrons it averages 1 inch 10 lines: the admeasurements are here taken to include the alveolar outlets. In both species there is a slight decrease of size as the teeth recede in position, and chiefly in the hind lobe or division of the last molar ; and in both species the second molar (d 4) is distinguished by the marked increase of size, especially transversely, of the hindmost lobe or division. The apex of the front lobe of the third molar (m 1) does not extend so far inward as that of the contiguous lobe of the second molar. The fourth molar (m 2) has a similar relative position to the third, so that the inner contour of the three mid molars is zig-zag, the palate gaining width between each as they recede in position. This character is better marked in the bare-nosed than in the hairy-nosed Wombat. The outer alveolar wall in all the species is deeper than the inner one, and is nearly on a level with the worn or working surfaces of the teeth. The enamel does not extend upon the outer surface, and thins off before it quite attains the angles between the outer and the fore and hind surfaces (Pl. LII. fig. 8, d 4, e¢; and Pl. LIII. fig. 3, d 4, e). The right and left series of the upper molars, as may be inferred from the palatal characters, diverge from the second to the last, in a greater degree in Phascolomys pla- tyrhinus (P1. LIL. fig. 1, d 4, m 3) than in Phascolomys latifrons (Pl. LIII. fig. 1, d 4, m 3). In a not full-grown latifront Wombat, with a skull 6 inches 2 lines in length, the upper molar series has the same longitudinal extent (1 inch 10 lines) as in a full- grown individual with a skull 6 inches 9 lines in length. The teeth, including the in- cisors, have acquired their full size. This fact bears serviceably on the interpretation of fossil Wombats with markedly smaller molar teeth in upper and lower jaws not exceeding in size those of the young Wombat above compared. PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. 307 The lower incisors of Phascolomys latifrons are more distinct in size and shape from those of Phascolomys platyrhinus than are the upper ones. The vertical diameter of the transverse section (Pl. LVII. fig. 5) is the longest, not the transverse diameter (ib. fig. 4). The outer surface (fig. 5, a), vertical, and slightly channelled, is divided by a well-marked angle from the lower surface (6), which is slightly and transversely convex. ‘The enamel covering the lower surface bends over the angle dividing it from the outer surface and there stops (at a). In Phascolomys platyrhinus (ib. fig. 4) the lower enamelled surface (ib. 6) bends up upon the outer (a) to near the upper surface, terminating there at an angle or ridge. A narrow longitudinal groove representing the wider outer channel in the hairy-nosed Wombat, divides the enamelled outer angle from the flat upper surface. The transverse section of the incisor may be called trian- gular in both species; but the base is internal and the apex external in Phascolomys platyrhinus, while the base is inferior and the apex superior in Phascolomys latifrons. The lower incisors are likewise smaller relatively to the jaw and to the molar teeth in the hairy-nosed than in the bare-nosed Wombat; and this character is more strongly marked in the large extinct Wombats indicated in my second Memoir on the osteology of the Marsupialia’. The first lower molar, on the other hand, is as large in Phascolomys latifrons (Pl. LVI. fig. 7,d 3) as in Phascolomys platyrhinus (ib. fig. 6, d 3); it is consequently larger in pro- portion to the size of the species, and in proportion to the other molar teeth; it has also a different form. In Phascolomys platyrhinus the transverse section and working surface of d 3, (fig. 6) is usually a full ellipse, with the long axis nearly parallel with that of the jaw. In Phascolomys latifrons the section is subquadrate (fig. 7). The anterior surface (e) usually shows a feeble longitudinal groove; the outer surface is rather narrower than the other three. The enamel covering it extends a short way upon the front surface, and then, after an interruption, is resumed upon the antero-internal angle: the outer enamel extends uninterruptedly over two thirds of the hinder surface. The other bilobed or biprismatic molars show little more than the difference of size, the four (d 4, m 1,2,3, fig. 3) equalling in longitudinal extent three and a half of those in Phascolomys platyrhinus (fig. 1). In all the species the enamel is wanting on the inner side of the tooth, which is nearly on a level with the inner wall of the alveolus; the outer wall is lower and exposes more of the tooth; the curves of the positions of the prismatic surfaces are reversed in the upper and lower molars. Mr. Waterhouse, in his instructive paper on the Dentition of the Flying Opossums’, pointed out two subgenera as having four true molar teeth on each side of both jaws, and a third subgenus as having three true molars on each side of both jaws; but the obser- vations on marsupial modifications of dentition were not carried further in that paper. In January 1839 I communicated my paper on the classification of the Marsupialia’, 1 Trans. Zool. Soc. vol. iii. (1845), p. 306. ? Proc. Zool. Soc. 1838, p. 149. 3 Trans. Zool. Soc. vol. ii. p. 315. 358 PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. which contained the generalized expression of observations on the dentition and other systems, carried out through all the materials then at my command. This was chiefly in relation to the actual phase of Mammalian taxonomy in reference to Cuvier’s order “Marsupiaux”!. W.S. Macleay and others opposed such ordinal group or union of the pouched quadrupeds. The learned Vice-Secretary of the Zoological Society had published reasons for rejecting the Marsupialia as a distinct group, and for distributing them among different placental orders according to their supposed closer affinities. The contrary views set forth by Cuvier and De Blainville? were defective in that kind of evi- dence which could alone render them convincing; accordingly Mr. Bennett asks, in 1831, ‘‘ What is there of importance in the structure of the Wombat, except this solitary character of the marsupium, to separate it from the Rodent order?” Amongst the structures shown to be both common and peculiar to the Marsupialia, I adduced the number of the true molar teeth, as characterized by size and shape, remarking that ‘in the dental system itself, the varieties of which have been chiefly appealed to as sanctioning the dispartition of the Marsupial order, we find an important peculiarity, by which the carnivorous, omnivorous, and strictly vegetable-feeding genera alike agree with each other, and differ from the corresponding placental Mammalia. In the ordinary Fere, for example, in the Quadrumana and in the Rodentia, as like- wise in the Pachydermata and Ruminantia, the number of grinders developed on each side of each jaw, which are not subject to vertical displacement and succession, is never more than three, while in the corresponding groups of Marsupialia it is always four”®. Since the date of this paper (1839) the associated group of Marsupialia has not been sought to be dissevered. It received the valuable sanction of Mr. Waterhouse in his ‘ Natural History of the Mammalia;’ and the generalization as to the number of true molars is given amongst the characters of the order in Part I. of that work, which was issued in 1845. 1 Régne Animal, vol. i. p. 172 (ed. 1829). > Although De Blainville was able to anticipate the appearance of the concluding volume of the ‘ Régne Animal,’ by a few months, in the issue of his Tabular Sheets of Classification, the priority of the proposition of a Marsupial series distinct from, but paralleling, the higher Mammalian orders, is due to Cuvier (“ Préface de la premiére édition, Octobre 1816”). 3 Trans. Zool. Soc. vol. ii. p. 332. I noted here the exception, previously pointed out by Mr. Waterhouse, in Petaurus (Acrobates) pygmeus, also that of the similarly minute Phalangista gliriformis, pp. 825-333. Fig. Fig. Fig. Fig. AO oO Re OF De PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. DESCRIPTION OF THE PLATES. PLATE L. . Side view of the skull of Phascolomys platyrhinus. . Back view of ditto ditto. . Back view of ditto of Phascolomys vombatus. . Ditto ditto of Phascolomys latifrons. . Foramen ovale, var. of Phascolomys platyrhinus. . Ditto and part of occiput, var. of Phascolomys vombatus. . Ditto and part of occiput of Phascolomys latifrons. PLATE LI. 359 Fig. 1. Top view of skull of Phascolomys platyrhinus. Fig. 2. Fore end of cranium of Phascolomys platyrhinus. Fig. 3. Nasal bones, var. of Phascolomys vombatus. Fig. 4. Ditto, var. 2, of ditto. Fig. 5. Ditto of Phascolomys latifrons. PLATE LII. Fig. 1. Base view of skull of Phascolomys platyrhinus. Fig. 2. Upper incisor, side view; and 2’, section, ditto. Fig. 3. First and second upper molars, magnified 2 diameters, ditto. Fig. 4. Palate, var. of Phascolomys vombatus. Fig. 5. Ditto, var of Phascolomys latifrons. PLATE LIII. Fig. 1. Base view of skull, Phascolomys latifrons. Fig. 2. Upper incisor, side view; and 2’, section c—e, ditto. Fig. 3. First and second upper molars, magnified 2 diameters. Fig. 4. Supratympanic cells and tympano-squamosal gomphosis, ditto. Fig. 5. Vertical longitudinal section of the skull, ditto. Fig. 6. Supratympanic cells and tympano-squamosal suture, Phascolomys platyrhinus. PLATE LIV. Fig. 1. Outside view of mandibular ramus, Phascolomys vombatus. Fig. 2. Ditto ditto, Phascolomys platyrhinus. Fig. 3. Ditto ditto, Phascolomys latifrons. Fig. 4. Hind view of ascending branch of ditto, Phascolomys platyrhinus. Fig. 5. Ditto ditto Phascolomys latifrons. OH oo Do So oe 0F Dre PROFESSOR OWEN ON THE SPECIES OF PHASCOLOMYS. PLATE LV. . Inside view of mandibular ramus, Phascolomys vombatus. Ditto ditto, Phascolomys platyrhinus. Ditto ditto, Phascolomys latifrons. . Outside view of ascending branch of ditto, var. of Phascolomys latifrons. . Hind view of ascending branch of ditto, Phascolomys vombatus. PLATE LVI. Under view of mandible, Phascolomys platyrhinus. Ditto ditto, Phascolomys vombatus. Ditto ditto, Phascolomys latifrons. . Right maxillary with front pier of zygoma, Phascolomys platyrhinus. . Ditto ditto, Phascolomys latifrons. PLATE LVII. . Upper view of mandible and mandibular dentition, Phascolomys platyrhinus. . Ditto ditto, Phascolomys vombatus. Ditto os sett: Phascolomys latifrons. . Section of lower incisor, Phascolomys platyrhinus. . Ditto ditto, Phascolomys latifrons. . First and second lower molars, right side, magn., Phascolomys platyrhinus, showing the partial coat of enamel (the outer lobes of d 4 ought to be more angular). . First and second lower molars, ib., magn., Phascolomys latifrons. All the figures are of the natural size where not otherwise expressed. ~ tte * a M &N.Hanharb imp Figs 12.5 PHASCOLOMYS PLATYRHINUS- figs 3 6. PHASC.VOMBATUS Figs 4.7 PHASC, LATIFRONS J Smit. ith M.&.N,Hanhart imp fag 1.& 2.PHASC,PLATYRHINUS. 3&4.PHASC.VOMBATUS: 5. PHASC .LATIFRONS. é - . i ea 5 bs + Z = tsa ‘ J Smit. Ith M &NHanharb mp J Smit lith ct ~ D y I S figs 1.5. PHASCOLOMYS LATIFRONS: Ag 6. PHASC. PLATYRHINU — J. Smit lith 1.PHASGOLOMYS VOMBATUS Figs aoe PAS PHATY! MAN Hanhart mp RHINUS Figs. 3 & 5. PHASC. LATIFRONS '. Smut th . M&N Hanhart imp figs 1&5 PHASCOLOMYS VOMBATUS fig 2 PHASC. PLATYRHINUS: Ay 3 &4 PHASC . LATIFRONS iy J. Smit lith M2&N Hanhart imp Fy 14 PHASCOLOMYS PLATYRHINUS. Fig 2. PHASCOLOMIS VOMBATUS. Fy 3.5.PHASCOLOMYS LATIFRONS. SL, a g C SOME. L004 / ay & > ——A ( i eo \ 3 af y iy Smt, hth M XN Hanhar ime Figi4.6 PHASCOLOMYS PLATYRHIUNUS . Fig 2 PHASC. VOMBATUS fig. 3 5.7 PHASC.LATIFRONS, [ 361 ] X. On Divornis (Part XVIII.): containing a Description of the Pelvis and Bones of the Leg of Dinornis gravis, By Professor Owrn, F.R.S., F.Z.S., &e. Read May 7th, 1872, [Puates LVIII, to LXI.] Iv is with feelings akin to compunction that I come again before the Society with claims to place the record of a fifteenth species of Dinornis in a volume of its Trans- actions. I have hesitated for some years in completing this step, tentatively ventured in 1869". But I have no alternative; for I do not see my way, on present experience of the value of characters from leg-bones, to attach thereto other interpretation of those about to be described than that which has led me, in the British-Museum Lists, to refer them to a Dinornis gravis. The fourteen extinct species of terrestrial or wingless birds proposed in preceding Memoirs being characterized by bones only, and chiefly, or mostly, by those of the hind limbs, might well be deemed by ornithologists to require subsequent confirmation before they received general acceptance. Remarks by some esteemed friends, eminent in that branch of natural history, significant of their reticence or expectant attitude, have not surprised me, and ought not. I look forward with equal interest, at least, to the result of further acquisitions of these remarkable and unexpected evidences of avian life in the old times of New Zea- land; and I believe myself ready and willing to yield up any of my species, should intermediate sizes of femur, tibia, and metatarsus, without distinct and well-marked modifications of form or proportion, prove the Dinornis struthioides, e.g., to be an immature Dinornis ingens, or this to show merely a stage of growth of Dinornis giganteus. I can only say that I prepared myself to grapple with the problem of these bones by a comparison of all accessible materials showing age-characters in the femur, tibia, and metatarsus of the known existing struthious birds, and have noted’ and represented® some of the characters by which the leg-bones, say of an immature Ostrich, might be detected and discriminated from bones of the same size of the Emu, Rhea, or Cassowary. It was not until I had satisfied myself that characters as distinctive as those observable in the leg-bones of the existing large species of Cursores or Struthionide were present 1 Trans. Zool. Soc. vol. vii. part 2, p. 141. ? Th. vol. iii. p. 243 (1843). 3 Tb. vol. iii. pl. 28, figs. 1 & 2. VOL, VII.—PART VI. May, 1873. 3F 362 PROFESSOR OWEN ON THE GENUS DINORNIS. in the bones of the six kinds of New-Zealand wingless birds defined in my first Memoir’, that I presumed to express such conclusion in the usual way of specific denominations. On the same grounds were defined four other species in the second Memoir’. As I have already noted, I had had the satisfaction of seeing the characters repeated in more than one specimen of femur, tibia, and metatarsus of eight out of those ten species, in most of them by several bones. Of Dinornis casuarinus, for example, in the year 1846, I had had under inspection ten femora, eleven tibixe, and six metatarsi. I had thus been able, in acquisitions received subsequently to the date of the above-cited Memoirs, to refer bones to species therein characterized and figured. If in time to come other observers and collectors of avian remains in New Zealand be able to match the bones with those which I have described and figured as the types of extinct species of Dinornis, in the way and degree, ¢.g., in which it has been done by the acute and experienced naturalist and Government Geologist of Canterbury Province in the case of the rich depository of dinornithic remains in Glenmark Swamp*, all reasonable scepticism will in the end give way. But I work strongly impressed with the duty of making due and suitable return for the opportunities liberally afforded me of examining and comparing the specimens collected in New Zealand, by giving figures of the natural size of all such as typify species. This is the essential foundation of the work of recognizing the already defined species, and of differentiating additional kinds of the extinct birds of New Zealand. On this ground I proceed to describe a pelvis, femur, tibia, and metatarsus of a Dinornis which comes under the latter category, and to ascribe it to a Dinornis gravis, from the weight of the bird relatively to its bulk, as indicated by the proportions of the bones of the hind limbs. I begin with the metatarsus (Pl. LVIII.), as this bone usually yields the best charac- teristics of the kind of Moa to which it belonged. In length it comes nearest to the metatarsus of Dinornis casswarinus*, in breadth to that of Dinornis crassus®; it is, however, shorter by half an inch than the former, and broader by five lines than the latter; and as Dinornis crassus was differentiated from Din. casuarinus by the greater relative breadth of the metatarsus, this differential cha- racter applies still more strongly to the present species, inasmuch as the entire bone is shorter than that of Dinornis casuarinus, instead of being longer as is the metatarsus of Dinornis crassus. The length of the metatarsus in Dinornis gravis is 7 inches 9 lines, the least breadth of the shaft is 2 inches 1 line, the breadth of the proximal end is 3 inches 2 lines or 3 inches 3 lines, that of the distal end is 4 inches 2 lines, the thickness or antero- posterior diameter of the middle of the shaft is 1 inch, its circumference is 5 inches. Trans. Zool. Soc. vol. iii. p. 235. 2 Tom. cit. p. 344 (1846). ? Haast, “On the Measurements of Dinornis Bones from a Swamp at Glenmark,” in Transactions of the New-Zealand Institute, p. 80. 8vo. Wellington, 1869. * Trans. Zool. Soc. vol. iii. pl. 48. fig. 3. ® Tom. cit. pl. 48. fig. 4. PROFESSOR OWEN ON THE GENUS DINORNIS. 363 A comparison of these dimensions with those of the two species of Dinornis to which the present comes nearest, in the Table of Admeasurements, p. 371, will exemplify the chief metatarsal characteristics of Dinornis gravis. In general configuration, as in robustness of proportion, though exceeding in the latter respect, the present metatarsus most resembles that of Dinornis crassus. I note the following differences, which are repeated in specimens of the metatarsus from two individuals of Din. gravis from remote localities in the South Island of New Zealand. The intercondylar rising (Pl. LVIII. figs. 1, 3, ¢) is but slightly developed anteriorly in any Dinornis, but it has more claims to be considered a “ process” in Din. gravis than in Din. crassus; the entometatarsal tuberosity (ib. figs. 1 & 3,¢) is more pro- minent, and is longer. In the antinterosseal depression (ib. fig. 1, 7) the upper common anterior orifice (ib. 7) of the interosseal canals is less speedily divided into the entinterosseal (ib. fig. 2, 7) and ectinterosseal (ib. m) canals than in Dinornis crassus; the rough depression below for the tibialis anticus (ib. fig. 1, 0) is deeper and better-defined for its size in Dinornis gravis, The calcaneal groove (ib. figs. 2, 3, ~) is narrower and deeper in Din. gravis; and the entocalcaneal process (ib. ib. 7) is rather less prominent. The outer depression on the ectocalcaneal process (ib. s) is better marked, and in one metatarsal specimen of Din. gravis appears as a shallow longitudinal groove. The posterior orifice of the entinterosseal canal (7) does not open into a depression of the bone; that into which the ectinterosseal canal (m) opens, is wide and shallow. The interval between these two orifices in one of the specimens is greater in Din. gravis than in the longer metatarsus of Dinornis crassus; in the other it is as great. As in Dinornis crassus, the places of insertion of the strong ligamentous aponeurosis formed by the confluence of the tendons of the gastrocnemius internus and gastro- cnemius externus are less longitudinally extended, less ridge-like, and are more concen- trated, broader, and better marked than in Apteryx australis’. The entogastrocnemial surface (Pl. LVIII. fig. 2, 7) begins behind the entometatarsal tuberosity, and runs downward and backward to near the mid length of the bone; it becomes more defined as it descends, retaining a breadth of about 3 lines, before it suddenly expands into a rough tract of a semioval shape, about 2 inches long by 6 or 7 lines in extreme breadth. This surface terminates about an inch and a half above the hind end of the articular surface of the inner condyle (11). The lower part of this surface indicates the position (ib. fig. 2,1) of the ligaments which would attach the metatarsal rudiment of a back toe to the main bone, if such toe existed in Dinornis gravis. The ectogastrocnemial tract (Pl. LVIII. fig. 2, 7) is at the outer part of the hind surface, and on the lower or distal half of the shaft, of the metatarsal. It is of a semi- ’ Trans. Zool. Soc. vol. iii. p. 277, plates 31 & 32, R. 362 364 PROFESSOR OWEN ON THE GENUS DINORNIS. oval form, 1 inch 9 lines in length, 7 lines in extreme breadth, and is divided from the back part of the articular surface of the outer trochlea (1v) by a smooth tract about 4 lines in breadth. A narrow, oblong, rough tract at the proximal part of the shaft behind the ectometatarsal ridge (ib. fig. 2, 4) may serve also for gastrocnemial insertion ; but it is divided from the lower ectogastrocnemial tract by a smooth surface of the mid part of the metatarsal, about an inch and a half in extent in one instance, and two inches in the specimen figured in Pl. LVIII. The narrow tract from the entogastro- cnemial tuberosity (gy) is more feebly marked, if it be discernible, in Dinornis crassus. The ectogastrocnemial surface is also less defined, and is continued upwards as a more or less conspicuous ridge to within a few lines of the ectometatarsal rough surface in Dinornis crassus. The fore part of the entotrochlea (fig. 1, 11) is broader relatively to the hind part in Dinornis gravis than it is in Dinornis crassus. The outer side of the fore part of the ectotrochlea (Iv) is more convex or tuberous, and is less defined from the ectotrochlear fossa than in Dinornis crassus. The interspaces of the trochlee are narrower in Dinornis gravis; and this character is the more easily seized, inasmuch as the breadth of the three trochlez is almost the same in the two species, notwithstanding the difference in the length of the metatarsi. Dinornis gravis had a stronger and stouter foot, relatively, than Dinornis crassus ; and the muscular force working it was more powerful, as is indicated by the insertional ridges and tracts (9, y, 2). In a metatarsus of Dinornis crassus, 8 inches 6 lines in length, the greatest breadth of the mid trochlea is 1 inch § lines; in a metatarsus of Dinornis gravis 7 inches 9 lines in length the greatest breadth of the mid trochlea is 1 inch 10 lines. The general characters of the bone, with the disposition and aspects of the distal trochlea, are much alike in the two species; but the differences above defined impress me with the conviction that ornithologists would find the birds to which the metatarsi of Din. crassus and Din. gravis belonged, if they had them entire to compare, to be distinct species. The number of the living species of Caswarius which have of late years been dis- covered in detached remnants of the great Australasian continent, show much more striking differences in plumage and dermal appendages than could have been suspected from any differences which are discernible in the bones of the legs; and these differences, when most distinct, are less marked than those above demonstrated in the metatarsals of the species of Dinornis which least differ in general size. The tibia of Dinornis gravis (P|. LIX.), in comparison with that of Dinornis crassus, which it most resembles, has a stronger or thicker shaft in proportion to its length’. The character of the metatarsal bone of Dinornis gravis is here repeated, but in a minor ’ See Table of Admeasurements, p. 371. PROFESSOR OWEN ON THE GENUS DINORNIS. 365 degree. For the rest, the modifications of the dinornithic character of the tibia which the stouter-legged species (D. robustus, D. elephantopus, D. crassus) present, are closely repeated in the tibia of D. gravis. The articular surface (Pl. LIX. fig. 3, a), adapted to that of the inner condyle of the femur, is large, shallow, semioval in shape, with the small end turned forward. The ectocondylar surface of the tibia (ib. 6) is comparatively small, in the form of a tuber- osity, the outer and hinder half of which is applied to the inner side of the ectocondylar ridge which divides the tibial from the fibular part of the articular surface at the distal expansion of the femur. The intercondylar channel (c) is wide and shallow, and slightly expands as it curves from behind, forward and outward, to the ectocnemial cavity. Anteriorly it is bounded by the low, rough, intercondylar eminence (ib. d) for the attachment of the crucial ligaments. The epicnemial channel (jf) is smooth, broad and shallow. The rotular or epicnemial ridge (e) bounds the anterior and outer half of the proximal expansion of the tibia. The upper end of the procnemial ridge (ib. fig. 1, g) forms the low obtuse angle of the epicnemial ridge; the outer continuation of this ridge forms the upper border of the ectocnemial process (/). The suprafibular facet (ib. fig. 2, 7) is triangular, rough, almost flat. A smooth tract, one inch in extent, divides it from the fibular ridge (m), which is four inches in extent and terminates nearly eight inches below the summit of the epicnemial ridge. The procnemial ridge (ib. fig. 1, g) is continued uninterruptedly down the fore part of the shaft with as much inward inclination as makes it, at the lower third of the bone, the inner boundary (g') of the “extensor groove” (ib. p). The ectocnemial process or ridge (ib. 4) is much shorter and thicker, subsiding four inches below the summit of the epicnemial ridge, but extending outward two inches from the suprafibular facet (fig. 2, 2), and having a thick, smooth border curving to the shaft below its pointed end. ‘he intercnemial surface (ib. fig. 1, £), slightly concave across at its upper part, becomes almost flat below. The ectocnemial surface (ib. fig. 2,7) is uniformly and more deeply concave across; it is 14 inch in width. There are (pneumatic?) foramina (ib. fig. 2, 7,7) behind and below the intercondylar surface. A low longitudinal ridge bounds internally the posterior flat tract of the upper halt of the shaft, and opposite the fibular ridge inclines forward and to the inner side of the shaft. The medullarterial orifice (ib. fig. 2, n) is on the inner side of the fibular ridge, one inch above its end. As the shaft descends the flatness of its back part gradually becomes convex across, and loses the rough reticulate shallow markings indicative of muscular origin. The ridge (g') continued from the procnemial process, longitudinally and pretty equally bisects a great part of the anterior surface of the shaft, and gives the three-sided character to that part of the bone. On the inner and anterior part of the tibia, 366 PROFESSOR OWEN ON THE GENUS DINORNIS. 3 inches from the proximal end, is a rough low ridge and surface (ib. fig. 1, 0), answering to the stronger prominence in the Cnemiornis (marked g in fig. 1, pl. 66. Zool. Trans. vol. v.). The extensor canal (Pl. LIX. fig. 1, p), its bridge (¢), and lower outlet (7) repeat or retain the dinornithic characters'. The tuberosity from the outer pier of the bridge is strongly developed. A flat, rough surface, 33 inches in extent, pointed above and broadening to 10 lines below, marks the inner side of the distal end of the shaft. On the other side of that part is a rough, narrow surface, suggesting a third attachment of the pointed end of a long and slender fibula. The configuration of the distal trochlear articulation closely adheres to the dinornithic pattern. The shallow pit receiving the ectocondylar convexity (Pl. LIX.) is well defined. The epicnemial process is developed as an epiphysis ; it retains this character in the tibia, 14 inches in length, of a young Emu (Dromaius nove-hollandie), and includes the commencement of the procnemial and ectocnemial ridges; it might pass for a distinct bone (the patella), as the distal epiphysis (ib. fig. 4, a, v) has been conjectured to represent an astragalus; but the cartilaginous homologue of the patella in Dromaius, which plays upon the rotular groove of the femur, is ossified in some birds, notably in the Penguins and Loons (Colyméus), in which the development of the epicnemial process is in excess”, and with which the true patella coexists. The distal epiphysis (Pl. LIX. fig. 4) has effected a closer union with the shaft of this tibia, agreeably with the law of its relation to the course or direction of the medullary artery; the portion of the shaft developed epiphysially (ib. a) with the trochlear articulations (4) ascends nearly three inches up the fore part of the diaphysis, and develops the groove and process for the ligamentous bridge of the “ tibialis anticus” tendon. Femur.—In no specimen that has hitherto reached me are the dinornithic modifica- tions of the femur more definitely and strongly marked than in the present example (Pls. LX. & LXI.), in which the effects of the usual posthumous abrasion are limited to the articular prominence of the head, and to parts of the margin of the trochanter and hinder projections of the outer condyle. I propose, therefore, to take the opportunity of noting the characters more in detail than has been done in previously figured, less complete, and commonly larger specimens. The head (a) of the femur, in all the species of Dinornis, is less sessile than in birds generally ; but the part supporting it has more the character of a neck in the present species than in Dinornis giganteus*, in Dinornis casuarinus*, or in Dinornis maximus’. The constriction affects the entire circumference, but is deepest, as usual, anteriorly, ! These are shown, in contrast with those in the tibia of Gastornis parisiensis, in the ‘ Quarterly Journal of the Geological Society,’ 1856, pl. iii. figs. 1 & 2, p, 204, > Owen, ‘ Anatomy of Vertebrates,’ vol. ii. p, 83. fig, 34, 7. > Trans, Zool, Soc, vol. iii. pl. 44, fig. 1, ‘ Tb. pl, 46, figs. 1, 2, 5 Tb, vol. vi, pl. 89, fig. 1. PROFESSOR OWEN ON THE GENUS DINORNIS. 367 the head being turned rather forward as well as projecting inward ; the axis of the head and neck also slightly inclines upward. The head forms more than a hemisphere. From the neck (d), or constriction, the bone rapidly expands to the great trochanter (f) and shaft. The articular surface is continued from the head upon the anterior two thirds of the upper part of the trochanter (Pl. LXI. fig. 2, ¢), and is defined by a linear boundary from the non-articular part. Three surfaces may be noted in the enormous trochanterian enlargement of this femur—the upper (epitrochanterian) subdivided as above, the anterior, and the postero-external. The anterior, or “ pretrochanterian,” surface (Pl. LX. fig. 1, 7) is bisected by a low linear ridge (h), which, rising about an inch below the angular summit, passes down- ward and inward, and subsides (at h’) on the inner side of the shaft above the expansion of the inner condyle. The pre- meets the post-trochanterian tract at an acute angle (7); the latter surface is traversed vertically by a rough, broad tract, commencing about an inch from the summit of the trochanter, and gradually approaching the anterior angle as it descends, below which the ridge bends forward, and terminates in the ectotrochanterian tube- rosity (Pl. LX. fig. 1, 7). The rough tract defines a narrow ectotrochanterian surface from the broader post- trochanterian surface (Pl. LXI. fig. 1,m). On this surface are two rough oval shallow depressions for muscular insertions; the upper one (Pl. LXI. fig. 1, ”, for the “ abductor femoris”) is 10 lines below the epitrochanterian ridge, and measures 15 lines by 10 lines: half an inch below and rather in advance of this is the second depression, of rather smaller size, but with a more irregular surface (ib. ib. 0, for the ‘“ quadratus femoris”). From this surface several longitudinal striz descend vertically, and are continued by one principal linear ridge down the outer side of the shaft to within an inch of the ectocondylar fossa (ib. #’). A rough tract is continued from the lower gluteal surface obliquely downward and backward, contracting to the strongly marked ridge (Pl. LXI. fig. 1, p). To this ridge converges an inner less prominent oblique ridge (ib. g) commencing at the inner side of the shaft, one inch below the neck. The two oblique ridges are divided by a smooth interspace 4 lines in breadth, below which is the orifice of the medullary artery (ib. 7). On each side of this orifice the linee aspere are, as it were, resumed, and descend slightly diverging,—the outer one (ib. p') gradually subsiding near the ectocondylar pit (“); the inner one (ib. g') descends an inch and a half below the arterial orifice to form the tuberosity, q'. Internal to this rises a second rough tuberosity (ib. g"), continued by a ridge to within an inch of the hind angle of the inner condyle (ib. 1). From the ectotrochanterian tuberosity (Pl. LX. fig. 1,7) the strong ridge bisecting the fore part of the mid third of the shaft is continued down near the mid line of that surface. Six inches below the upper end of the femur this ridge, which divided the attachments of the “ vastus externus” and “vastus internus” muscles, bifurcates, its 368 PROFESSOR OWEN ON THE GENUS DINORNIS. divisions diverging to the upper end of the condylar risings (ib. uv, y), but subsiding before attaining these. At the lower part of the pretrochanterian surface, midway between the head and the ectotrochanterian tuberosity, is the rough surface, partly prominent, partly depressed (Pl. LX. fig. 1, s), which seems to represent the small trochanter in Mammals. The rotular cavity or channel (ib. fig. 1,¢), 2 inches across and 14 inch in depth, shows no trace of the rising marked in that part of the femur of Dinornis giganteus’. The anterior intercondylar ridge dividing the rotular fossa (¢) from the intercondylar one (Pl. LXI. fig. 1, v, v') is sharp. The intercondylar fossa is partially divided into two depressions, the inner one (ib. v) being the deepest, the outer one (v') the largest, These are divided from the popliteal cavity (Pl. LXI. fig. 1, z) by the post-intercondylar ridge (ib. w), which is well marked, but shorter, thicker, and more rounded than the anterior one (Pl. LX. fig. 2, 7’), The entocondylar articular surface (ib. w’) has the usual relative size and shape®, The ectocondylar surface for the tibia (ib, 2) is comparatively small, measuring 1 inch 9 lines by 1 inch in extent; it is continued over the ridge-like posterior pro- jection of this part of the condyle to that on the concavity or groove for the head of the fibula (Pl. LXJ. fig. 1, y), which groove is feebly divided into an upper (y') and lower (7’') tract, The popliteal space (Pl. LXI. fig. 1, z) has the usual dinornithic depth, shape, size, and oblique direction. It is rugous; and some small foramina at its deepest part are the sole representatives of a pneumatic system, though probably related only to the transmission of vessels. The contrast is striking, in placing by the side of the above-described bone the femur of any of the species of large existing Struthious birds, in regard to every indication of the strength and vigour of application of the hind limbs. The chief results of the comparison of the femur of Dinornis with that of Struthio were recorded in the original Memoir on the first received bones from New Zealand®. But the femur of the Emu (Dromaius), though still smaller than that of the Ostrich in comparison with the gigantic species of Dinornis, is less different in shape. The shaft of the bone is rounder than in the Ostrich; but the antero-posterior diameter of that part is less than in Dinornis, There is no trace of the bifurcate ridge on the fore part of the shaft, and very feeble indications of “lines aspere” on the back part; of the tube- rosities there developed in Déinornis no rudiment eyen is present in existing Stru- thionide. The medullarterial canal is very minute in Dromaius, as in other pneumatic femora; and the associated large air-hole at the back part of the upper end of the femur significantly differentiates Dromaius, as it does the other large existing Stru- thionide, from Dinornis. The head is sessile; one cannot predicate a cervix in the 1 Trans. Zool. Soc. vol. iii. pl. 44. fig. 2, 7. 2 Th, ib, pl. 46. fig. 3, ¢ (Dinornis casuarinus), 5 Th. ib. (1843) p. 249. PROFESSOR OWEN ON THE GENUS DINORNIS. 369 femur of Dromaius; the trochanter hardly rises above the level of the head: the back of the trochanter is scarcely at all accentuated, chiefly shows a smooth, feeble concavity; there are no gluteal rugosities, no trace of a lesser trochanterian place of muscular attachment. The popliteal cavity is a shallow groove, not bounded by any post-intercondylar ridge from the intercondylar space. The distal expansion is rela- tively much less than in Dinornis; the inner condyle is much narrower. The tibial part of the outer condyle has relatively more longitudinal extent in Dromaius than in Dinornis; it rises well above the fibular division, which is relatively shorter than in Dinornis, where it equals in that dimension the tibial prominence. But the fibular division projects more outwardly in Dromaius, is broader in proportion to its length, and more generally convex. There is no trace of the rough pit for ligamentous or muscular attachment above the fibular division of the outer condyle which so markedly distinguishes the femur of Dinornis. The antero-posterior extent of the outer condyle is much greater than that of the inner condyle in Dromaius; the difference is less in Dinornis giganteus', Din. casua- rinus’, and Din. didiformis*. The antero-posterior dimension of the outer and of the inner condyle are nearly the same in Dinornis gravis. The pelvis of Dinornis gravis is characteristically massive and ponderous, and accords in shape with those figured in plates 19 & 20 of the first Memoir‘. The upper and outer bony wall of the hinder expansion, beyond the gluteal ridges®, is better preserved than in figure 3, plate 20° Eight coalesced vertebrae with combined par- and pleur-apophyses precede the three interacetabular vertebra, in which those processes are wanting. The bodies of these are broader and flatter below than in the subject of figure 2, pl. 19. vol. iii. Trans. Zool. Soc. After the above eleven sacrals follow six vertebre with par- and pleur- apophyses again abutting against the iliac walls. The first sacral has, on each side the centrum, a circular cup for the head of a free rib, behind which cup is a large pneumatic foramen. The ribs of the seven succeeding sacrals are anchylosed and short, abutting against and coalescing with the closely grasping plates of the antacetabular parts of the ilia. The interpleural vacuities of the eight anterior sacrals rapidly decrease in size to the fifth, and again slightly expand in the last two. The first three pairs of anchylosed ribs incline forward; the next three pairs are transverse; the last of this series curves slightly backward, commencing that curve which is carried out by the proximal ends of the ischia. Both ischia and pubes in the present specimen are broken away from their origins at the acetabulum. The following are the dimensions of this pelvis:— ? Trans. Zool. Soc. vol. iv. pl. 44. fig. 2. * Tb. ib. pl. 46. fig. 3. * Tb. ib. pl. 24. fig. 3, * Zool. Trans. tom. cit. (1843). * Ib. vol. vii. p. 367. * Tb. vol. iii. VOL. VIII.—ParT vi. May, 1873. 3G 370 PROFESSOR OWEN ON THE GENUS DINORNIS. in. Length (fore ends and hind ends of ilia broken off). . . . . . . 14 Breadth (behind acetabula, where the prominences are abraded) . . 8 Height of first sacral (summit of spine broken off) . Breadth of first sacral . . . . . Suns nek, oye aiae f Diameter of circular orifice of neural canal of first sacral . 0 From fore part of first sacral to postacetabular angle of ilium 5 8 Breadth: of “pelvicidisk Ave ve hs) Me Gch o Cae its Ete ee Vertical diameter of acetabulum . 9) Length of first eight sacral vertebre 6 Length of last nine sacral vertebre .. - 0. - + 1 ee es we 7 onworROOm OF It may be convenient for future comparers of the hind-limb bones of Dinornis to have, in a tabular form, the principal dimensions of the femur, tibia, and metatarsus of the species now named. In this table are given references to the figures of the type specimens affording the dimensions ; and there are added remarks and dimensions of bones obtained by Dr. Haast, F.R.S., from the swamp of Glenmark, and which are referable to the above species of Dinornis. ' Trans, Zool. Soc, vol, vii. p. 368. TABLE OF ADMEASUREMENTS OF THE BONES OF THE LEG OF THE KNOWN SPECIES OF DINORNIS. The letter H. refers to the admeasurements given by Dr. Haast in the “ Memoir” cited, note 2, p. 362. Dimensions of the Femur. Din. curtus. nares Din. didiformis. Din. dromioides. aut Din. gracilis. Din. gravis. Din. casuarinus.| 4 6 9 ise ae 15 is +3 f4 £8. fi. f17. | f16. f Taylor. H. H. I, |i, = AS | in. lin, in. lin.) in. Jin. fin. lin. in. lin. in. linJjin. lin. in. lin} in. lin. jin. lin. in. lin.in. lin, in. lin. in. Rar in. lin. in. lin.| RMN cdsavcens>-asecasss 60 So 6 3 8 o Ee) Ss Or G GC 7\ G © be Ge soe Aly @ Tey eR 3\9 6* 10 4] ith of proximal end (in the | | axis of the neck) ........-s:++.+++- 23 20 2 3 2 10 Ey a) a Sica 9 APO airbases, Ay 3 508 | Ge ey ia | Breadth (transverse) of distal end|2 5 Zi 2 9 a 2 aa A ON 7 YONA Bl eo eG) Ae 3.49 4 4 Cireumference of middle............ 29 2 7 ez! 4 0 40°60 43 0\"4ao 3 10k] 3 9 14 8 5 8 60 5 3 5 ° 4 9 | * Perhaps not quite enough allowed for mutilated extremities. - Dimensions of the Tibia. leet 10 14 16 19 . t3. t8. eaey H. H, | ReMi itaesssnocscsvavacceecsseras ° 14 0 TS Om r Si Aue yee 4a | Zito peaee) HAGeROnei2 3 OU M2 aun Outs se RG az or eno! 7, ens | I z3 © 4 5 conte, 9 | 2eanto ch} Srceto 4 6 oO oreaco feb naps | agaose GUO) arone: 10 Zz 9 0h gecena cocoon 2 3 oaacde Py CO |) FQ). ppncn0 3) oneore = good BT ‘onentn GAL -ocboce FETS ere crc Avs Ol MER octass fy @ | CSG censos 49 53 o LE ZS) Sree GN) peceno Gin Oe ee or OC) @ dass Meee Woy OQ opens GB casaca — gaseeo Sin On meesteees Dimensions of the Metatarsus 5 8 i 12 17 | 20 4. | it || | Tromp ...sscceceseseceessetesee cence: 6 60 7 0 (Sty waance ITOmAS i Naess © beh © 1 OF © 7 9 7) AN Stein al cessee Circumf, at the middle of the shaft Gil eee aq ey) ceesen GiGi eed AAG WEL cerns SiON SOMES: we eat t5i|aueae e aceoae Breadth (transverse) of distal end 5 I 6 ne MN casedes AG) ES Bera hg seccc: Gy PA By cece AM Tyre Mae teks, che ich metros Breadth of middle of shaft ......... II Co) 4G LOG Fiat ah crce ry’ ie Fares: TRIOns | STingie este Bis Oben Beets ille 24-02 1) L7e = ee Thickness or antero-posterior dia- | meter of shaft ............s.20020++ ° 6 aoeses ° 9 MeO Nicest |fOrrom Werner: 590) liso cee Io 0 10 Breadth (transy.) of proximal end| 1 11 9 BEY NP cconna Petey icaeeco ear hele, Wecncsce ty WeP ect | obec 3.1 3.0 Dimensions of the Femur. Din. crassus. | Din. elephantopus. | Din. struthoides. Din. ingens. Din. giganteus Din. robustus. Din. maximus. 22 25 28 31 d 33 i. f12. ff Colenso. f2. H. H. H. | in. lin. in. lin. in. lin.) in. lin. in. lin.) in. lin. in. lin.Jin. lin. = in. din. sin. lin} in. lin. in. lin.| in. lin. in. lin, edancsses Iz 0 13,0 rz 8 | tz 6% 32 0} 13 9 13 0/16 o WO Geckos 14 2 14 6) 18 3 18 4) Pett of proximal nal ‘Gn ‘the BPMEIMICCK) ....20+.-.-n20000+ 1h ats) lO Se Meese: « 4, 2 yi hall situ Cb TWN Me Yeh edge ascaaé OB leeaae GG iene. Pea —.o. of distal end cd Fit ON Sects 408 Ae Aa xO ANG) acted Pine eee an 7 NOM coereae Circumference of middle............ 6 8 Folie recs 5 6 Pea | 5 aio Cie Heaeep Gye Mteedesse | A Rdeont Geel arse 9 6 9 6 * Perhaps not quite enough allowed for mutilated extremities. Dimensions of the Tibia. 23 26 29 t2. tEl.| 71. tH. 19 0 24 0 21 6|22 0 21 2 |\'29) 10 28 9/35 0 35 0 32 3 6 2 op Gp) in| ae Ge Z|) Oh 3 6 6 | 4 G 7 0 qe hanes 4 2 4 0} 3 2 Ba tale ey 4 0/4 0 4 8 44 45 6 5 O 51 8.6 4 8] 5 3 3 ||_0. @ 6 6 6 9 Io 0 It 6 II o Io O ZOO) Iz 90 12 0/13 0 13 0 Iz oO Dimensions of the Metatarsus. 21 24 27 0 31 | 35 m3. mE. | mH. Colenso.| m1. m EH. H. 8 6 9 8 gg) |12- 0 ir 6| 14 6 13059)|/55-6 18 0 18 9] 15 20 0 21 6) i 4 8 4 8 Aon ONAN) 4h 2lee5 sO Ae 864) gy 16 Ge OMe aan 5 8 4 6 4 (transverse) of distal end 4.0 5 6 BL AG A BR G A | BY i CNH Aare 5 6 9 7 3 of middle of shaft ......... I 10 2s 3G) Gh i 5G) ree?) eens SPI irate 2 3.0 DG) s or antero-posterior dia- | a TQ |) oobeccs eases Io. © ot Teme ah ke (6) T 6... fecseslpeneee: PLES fee cocr an EGS | (transy.) of proximal end 3 34 4 8 Ze 3 s\) asoone Gans 3t 3, 61 0 © At: Orcas 4 6 ol § 6 Gaal e of which a side view given in pl. 20. 4, and there tarsal of Din, casuarinus. 1 Trans, Zool Soc. iii. pl. 47. igs. 3,4, & 5. | ¢3' Tbidem. 2 Ibid. vii. pl. 44. fig. 8 (com- | ire with the first indication of S species, Trans, Zool. Soc. iii. 1. 48. fig, 6), _ * Perhaps not enough allowed for water-worn margins of trochlex. 3 Trans. Zool. Soc, vii. pl. 44. fig. 7. 4 Ibid. v. pl. 65. figs. 5, 6. 5 Ibid. pl. 67. figs. 5, 6. 6 Ibid. iii. pl. 24. 7 Ibid. pl. 26. figs. 3, 4. 8 Ibid. pl. 27. fig. 3. 9 Ibid. pl. 22. 10 Ibid. pl. 47. fig. 1. 11 Ibid. pl. 48. fig. 2. 12 Ibid. iv. pl. 3. 13 Trans. Zool. Soe. iv. pl. 41. fig. 1. 14 Ibid. pl. 42. fig. 2. 15 Ibid. viii. pl. 58. 16 Ibid. pl. 59. 17 Ibid. pls. 60, 61. 18 Ibid. iii. pl. 46. 19 Ibid. pl. 47. 20 Ibid. pl. 48. fig. 3, (See also pl. 20. a, fig. 2, wrongly ascribed to Din. didiformis.) figs. 4 & 5. 22 Ibid. iv. pl. 43, fig. 1. 23 Ibid. figs. 4 & 5. 24 Ibid. pl. 44. fig. 1. 25 Ibid. pl. 41. fig. 2. 96 Ibid. pl. 42. fig. 1. 27 Ibid. pl. 41. 28 Ibid. iii. pl. fig. 4. 21. fig. 1. + With a ridge at the middle of inner condyle at proximal end. 21 Trans. Zool. Soe. iii. pl. 48. fig. 1, pl. 26. fig. 1. 30 Ibid. pl. 48. fig. 1. 31 Ibid. pl. 27. fig. 1. 32 Ibid. iv. pl. 1. fig. 1. 33 Ibid. vi. pl. 89. fig, 1. 34 Ibid. pl. 90. 35 Ibid. pl. 89. fig. 3. 362 29 Trans. Zool. Soe. iii. pl. 25. PROFESSOR OWEN ON THE GENUS DINORNIS. (Sv) +] bo Addition to Note 5, p. 371. The dimensions of the tibia and metatarse of the “smallest size” of Dinornis didiformis in Dr. Haast’s list of bones from Glenmark swamp agree with those of the present species (D. geranoides) ; but the femur therewith associated has the length (9 inches) of a large variety of D. didiformis.—Trans. New-Zealand Institute, vol. i. 1869, p. 83. Addition to Note 8, p. 371. Tibis and metatarsi identical, or nearly so, in size and shape with those of the present species are referred thereto by Dr. Haast (Joc. cit.) ; but the femur associated therewith in the list of specimens from the Glenmark swamp is an inch longer than that of D. didiformis from the North Island. Does this indicate an established variety in the South Island? Addition to Note 9, p. 371. Unless the femur, 9 inches 6 lines in length, referred to “ No. 4, Din. didiformis, largest size,” in Dr. Haast’s list (Joc. cit. p. 83), belongs to a Dinornis dromivides, that species has not yet been met with out of the North Island. The tibia, 15 inches 6 lines in length, and the metatarsus, 7 inches 5 lines in length, which Dr. Haast associates with this femur, would rightly come under the head of Din. didiformis. Addition to Note 14, p. 371. “No. 10. Dinornis gracilis, Owen. Of this elegant species three more or less complete specimens were found amongst the exhumed bones, which agree in every respect with Professor Owen’s figures. For comparison, I shall give the measurements of the best-preserved specimen :— PERT Girth of Girth of shaft, Girth of Migs H. | Panini proximal end, | thinnest part. distal end. | inches. inches. inches. inches. coMetatarsus) .y sa. < 1. all 12:6 10:3 4-6 12:0 i vie vec yeraentans Oke ortecnees | 23-0 13:9 4:7 11:3 IB OMUBD savy cerennea veers ens 11-4 113 5:4 Wh Haast, loc. cit. p. 84. Addition to Note 17, p. 371. To this species I refer ‘No. 17, Dinornis .... (?),” in Dr. Haast’s List (loc. cit. p. 87); on which he re- marks ;—* A species smaller than No. 16, but partaking still of the same character. « Measurements. eee : Girth of Girth of shaft Girth of Bkeb 10 TeeUHAS Gea, proximal end. thinnest part. distal end. inches. inches. inches. inches, GINARCMEDE Godeacbons 7:4 75 4:5 10-3 MEDIAS gtsabatiacey st es 15:7 14:4 4:4 9:5 Loy? 5 hain ae ois Sera 9-3 10:5 53 11-4” The specimens from the Glenmark swamp, referred to under “ No, 16, Dinornis crassus,” agree best with Dinornis gravis. Dr, Haast remarks :—‘ Of this species we obtained ten more or less complete specimens, 80 that I had ample material to assure myself of the correctness of its specific character. *¢ Measurements. | | isin WE *, Girth of Girth of shaft, Girth of | Tah Sh Tength of bone. proximal end. thinnest part. distal end. | inches, inches. inches. inches, |* Metatarsus .......... 79 9-0 4:8 11-1 psn OE Meee men ceteenc acres 16:7 14:2 5-0 10-6 Hemury.). eee 10:8 12:9 | 6:0 15:3 PROFESSOR OWEN ON THE GENUS DINORNIS. 373 ‘The metatarsus is shorter than Professor Owen’s crassus, the circumference larger, but the proximal end is somewhat smaller. “The tibia is smaller, but again thicker than Professor Owen’s crassus. “The femur is also shorter, whilst the circumference is the same as Professor Owen’s species, An examina- tion of the general character shows that it is a somewhat smaller but stouter bird than Professor Owen’s crassus.”—Loc. cit. p. 86. Addition to Note 21, p. 371. To this species I should refer the specimens of which Dr. Haast gives the following average size of the leg- bones of fifteen specimens obtained from the Glenmark swamp :— i Girth of Girth of shaft, Girth of - Ag We Length of bone. proximal end. thinnest part. distal end. inches. inches, inches. inches. SOMeGtALAYSHS) = <\j-)2a:1s ) en at from Nature on stone by JT Excdeben from Nat.on Stone by JE DINORNIS GRAVIS ZN Hanharl omp [ 381 ] XI. On Divornis (Part XTX.): containing a Description of a Femur indicative of a new Genus of large Wingless Bird (Dromornis' australis, Owen) from a post-tertiary deposit in Queensland, Australia. By Professor Owen, F.R.S., F.L.S., &e. Read June 4th, 1872. [Puates LXII. and LXIII.] IN 1836 Sir Thomas Mitchell, F.G.S., Surveyor-General of Australia, discovered in the breccia-cave of Wellington Valley a femur, wanting the lower end, mutilated and in- crusted with the red stalagmite of the cave, which I determined to belong to a large bird, probably, from its size, struthious or brevipennate, but not presenting characters which, at that time, justified me in suggesting closer affinities. Three views of this fossil, of rather less than half the natural size, formed the subject of pl. 52. figs. 12, 13, of my “ Paleontological Appendix” to Mitchell’s work’. The length of this fossil was 13 inches, the breadth of the middle of the shaft was not quite 3 inches. In 1869 the Rev. W. B. Clarke, F.G.S., Government Geologist of the Province of New South Wales, made known the interesting discovery of a femur, nearly 12 inches in length, during the digging of a well at Peak Downs, in Queensland*. The well was sunk through 30 feet of the black trappean alluvial soil common in that part of Australia, and then through 150 feet of drift pebbles and boulders, on one of which boulders (“at that depth,” 150 feet?) rested a short, thick femur, so filled with mineral matter (calc spar and iron pyrites) as to give the internal structure more the appearance of a reptilian than an ornithic bone’. Mr. Clarke submitted this fossil to the able Curator of the Australian Museum, Sydney, and states that “Mr. Krefft has compared it with a collection sent over from New Zealand by Dr. Haast, and has been enabled to determine it to be a bone belonging to Dinornis.” The communication is accordingly headed “ Dinornis, an Australian genus.” So exceptional an extension of New-Zealand forms of life to the Australian continent greatly added to my desire of further and more intimate acquaintance with this second evidence of a large extinct Australian bird—more especially as the femora of Dinornis received from New Zealand subsequently to the publication of Mitchell’s work led me to perceive, from the antero-posterior compression of the shaft and the sessile position of the head in the femur from the Wellington-Valley cavern, that it resembled that bone in the Emu rather than in the Dinornis. ; My wishes on this point, as on others connected with the paleontology of Australia, 1 Spduos, cursus, dpyis, avis. 7” Three Expeditions into the Interior of Eastern Australia, vol. ii. 8vo, 1838. 3 « Dinornis, an Australian Genus,” Geological Magazine, vol. vi. (1869) p. 383. * Loe. cit. p. 383. 382 PROFESSOR OWEN ON THE GENUS DINORNIS. met with a prompt and hearty response. The Trustees of the Australian Museum directed the unique bird’s bone to be moulded, and they forwarded to me a plaster cast. Mr. Krefft was so good as to have three photographs taken of the fossil :—one showing the back view of the bone, three fifths the natural size; the two others the front views of the proximal and distal halves of the bone, of very nearly the natural size. With these evidences a satisfactory comparison can be made of the Australian fossil with the femora of other large wingless birds, both recent and extinct. The bone is the right femur (Pls. LXII. & LXIII. fig. 1). It measures 11 inches 6 lines; and there may be an inch more of this dimension lost by the abrasion to which both ends have been subject. The middle third of the shaft is entire, and shows its natural form and surface; the breadth of this part is 2 inches 6 lines; the antero- posterior thickness does not exceed 1 inch 7 lines (Pl. LXIII. fig. 2). The extreme breadth of the upper end is 5 inches 3 lines, that of the lower end is 5 inches; but these latter dimensions fall short, probably by half an inch, of those which the un- abraded or entire femur would have yielded. Of the femora of Dinornis I have selected that of Din. elephantopus', as nearest to the present fossil in regard to length (13 inches) ; the breadth of the shaft is the same, or, in the largest examples of D. elephantopus, exceeds only by 2 lines that of the Australian femur. But the shaft of the bone in Dromornis is compressed from before backward ; its trans- verse section is a narrow oval (Pl. LXIII. fig. 2), while that of the Dinornis is a fuller and less regular oval (ib. fig. 3) from the greater proportion of fore-and-aft breadth of the shaft. The back part of the shaft of Dromornis australis, besides being less convex transversely, is devoid of the strong ridges and tuberosities which characterize that part in all the species of Dinornis; in this respect, as in the shape of the transverse section of the femoral shaft, Dromornis resembles more that bone in the Emu (Dromaius ater). The bifurcate anterior muscular (“intervastal”) ridge which characterizes the fore part of the femoral shaft in Dinornis elephantopus (vol. vii. pl. 43. fig. 1), as in other species of that genus, is not defined on that part of the femur of Dromornis (Pl. LXII. fig. 1). The longitudinal ridge, descending from the pretrochanterian ridge to the ectocondylar expansion, is traceable in the cast, but is less strongly marked than in Dinornis. The mutilation of the prominent parts at the proximal end of the femur begets a reticence in drawing conclusions from apparent differences ; but some were evidently inherent in the original when entire. The periphery of the head of the femur (d) is not constricted so as give the appearance of a “neck,” as it is in Dinornis. The trochanterian part of the articular surface (¢) is more horizontal, does not ascend as it recedes from the head, in Dromornis. So far as the trochanter (f) is preserved in the cast, and appears in the photographs, it does not rise above the level of the head (a) of the femur, and seems not to have risen, when entire, so much above it as in Dinornis; the lay of the trochanterian articular tract agrees with these indications of the remain- 1 Trans. Zool. Soc. vol. iy. (1856) p. 149, pl. 43. fig. 1. PROFESSOR OWEN ON THE GENUS DINORNIS. 3838 ing epitrochanterian ridge (f). In the above differences Dromornis more resembles Dromaius. The ectotrochanterian surface (Pl. LXII. fig. 2) is slightly concave, bounded above by a low arched ridge, from which the rough convex epitrochanterian part of the surface ascended to the crowning ridge. In this character Dromornis resembles Dromaius, and differs from Dinornis. It resembles Dinornis, and differs from Dromaius, in the absence of the pneumatic foramen at the hind part of the upper expansion of the femur. This expansion is also relatively greater than in Dromaius, and recalls rather that of Dinornis; but the breadth of the ectotrochanterian tract (Pl. LXII. fig. 2) is relatively less than in Dinornis gravis, and still less than in the exceptionally thick and massive femur of Dinornis elephan- topus. The fore part of the upper femoral expansion has had its outer wall crushed in ; but, in both the cast and the photograph, there is an indication of a rough subcircular tract, answering to that which is conspicuous in Dinornis (loc. cit. pl. 43. fig. 1,7), but which is not present in Dromaius. The outer crust of the femoral wall has been crushed inwards at the distal third of the fore part of the shaft (Pl. LXII. fig. 1, 4); but the rotular surface seems to have been broad and shallow. In the prominence and thick convexity of the fore part of the expansion of the outer condyle Dromornis resembles Dinornis rather than Dromaius. The transverse extent of the distal end, in proportion to the size of the shaft of the femur, is less than in Dinornis, but is greater than in Dromaius. The popliteal cavity (Pl. LXIII. fig. 1, z) is oblique, and is deeper and better-defined, especially above, than in Dinornis; it is divided from the intercondylar cavity (v) by a ridge (w) similar to that in Dinornis, and which I do not find in Dromaius. The inter- condylar cavity or pit (v) is deep, and smaller than in Dinornis gravis; it is deeper, but much smaller, than in Dinornis elephantopus (loc. cit. pl. 43. fig. 3). There is a rough “gluteal” depression (Pl. LXIII. fig. 1, x), less deep than in Dinornis gravis, and situated nearer the popliteal cavity, and with a more posterior aspect than in Dinornis elephantopus. The mutilation of the prominent parts of both femoral condyles precludes further profitable comparisons of the fossil under consideration. But from those for which it affords sufficient grounds, I infer that in its essential characters this femur resembles more that bone in the Emu than in the Moa, and that the characters in which it more resembles Dinornis are concomitant with, and related to, the more general strength and robustness of the bone—from which we may infer that the species manifested dinornithic strength and proportions of the hind limbs, combined with characters of closer affinity to the existing smaller, more slender-limbed, and swifter wingless bird peculiar to the Australian continent’, 1 T can now, in 1872, repeat with more confidence the remark in my Memoir of 1846:—‘ No remnant of a Dinornis has yet been found in any of the contiguous islands; and I have in vain searched for such in the col- lections of post-pliocene fossils from Australia.” —Trans. Zool. Soc. vol. iii. p. 366. 384 PROFESSOR OWEN ON THE GENUS DINORNIS. From the proportions of the femur of Dromornis I infer also that those of the tibia and metatarse would be longer and more slender than in Dinornis elephantopus, and in a greater degree than is the case with the femur. Consequently the stature of Dro- mornis would be greater in proportion to the solitary bone by which we now know it than is that of the Dinornis elephantopus. We may therefore have a comfortable assurance that it indicates the former existence in Australia of a bird nearly of the stature of the Ostrich, but with relatively shorter and stronger hind limbs. The period at which this large wingless bird trod that singular land was that at which the elephantine Marsupial (Diprotodon) flourished. I have received remains of both this genus and the somewhat smaller pachydermal Marsupial (Nototherium) from the mass of drift and boulder deposit when this had been reached, at depths equal to that yielding the bird’s fossil at Peak Downs, in the sinking of wells in Queensland. The mineralized condition of these herbivorous mammalian fossils has suggested a comparison of them with the fossil remains of Saurian Reptiles from Oolitic and even older Mesozoic beds in England. Yet the Mollusca which have left their shells with the petrified Australian bones are of the same species as those still living in the fresh waters of the Condamine and its tributary creeks, in the bed of which so many evidences of extinct Marsupial life have been discovered. From the general analogy, not unfrequently pointed out, between the recent animal and vegetable forms of the Australian continent and the extinct ones of the European Oolitic beds, together with the massive mineralized condition of the ornithic and mam- malian fossils found deep in the enormous superficial accumulations of drift and trappean alluvium, we are led to surmise that Australia, or parts of that continent, have not been subject to the frequent movements by which the earth’s crust has been modified in the European continent, but that it may have been subject exclusively to the subaerial conditions of change from the period of the Oolitic deposits in our hemisphere. Thus: the Dromornis of Queensland may have been contemporary with the impressors of the ornithicnites of Connecticut. DESCRIPTION OF THE PLATES. PLATE LXII. Dromornis australis. Right femur: nat. size. Fig. 1. Front view. Fig. 2. Ectotrochanterian surface of femur. PLATE LXIII. 1. Back view of the same femur. Fig. 2. Form of transverse section of middle of the shaft. 3. Form of transverse section of the same part of the femur of Din. elephantopus. PholoanStone by sben open Erxle oT Ston wo) On a. Pho’ 2 a ake al XII. On the Axial Skeleton of the Ostrich (Struthio camelus). By Sv. Grorce Mivart, F.R.S. Received June 18th, 1872. RECENT investigations haying made it probable that the line of affinity between Birds and Reptiles passes through the Struthious members of the first of these classes, I have deemed it advisable to commence a study of the axial skeleton of the Sawropsida by a detailed examination of that of the Ostrich, as of the most generalized type. By kind permission of the authorities of the Royal College of Surgeons, I have been enabled to make use of the rich resources of that institution, not only for examination, but for the purposes of illustration, all the figures being from specimens in that Museum. Bearing in mind the varying posture which the axial skeleton assumes in different Sauropsidans, I think it better, generally, to employ the term preagial to denote that relation which in a vertical spinal column would be called “superior,” and in a hori- zontal one ‘‘ anterior.” Similarly I use the word postawxial for what under the circum- stances mentioned would be either “inferior” or “posterior.” In the same way the terms dorsal and ventral stand for “posterior” or “superior,” and for “anterior” or “inferior ” respectively. After describing the various vertebree throughout the spine one after another, I pro- pose to describe the pelvis as a whole, then the vertebral and sternal ribs, and the sternum, concluding with a recapitulation of the serial modifications the several parts and pro- cesses undergo as we proceed postaxially from the atlas to the coccyx. There are seventeen cervical vertebre, which, in the adult, have either no rib-like processes or only styliform and anchylosed ones (fig. 1, c). The next three vertebree bear longer ribs, generally articulated movably with their vertebre and not directly connected with sternal ribs. They may be called cervico- dorsal vertebre (fig. 1, cD). The next five vertebre (twenty-first to twenty-fifth inclusive) support long ribs, which unite distally with sternal ribs articulated to the sternum, and are therefore true dorsal vertebre ; these vertebre do not anchylose together or with the sacrum (fig. 1, D). There are two vertebre after these (twenty-sixth and twenty-seventh) which bear freely-ending * ribs or rib-like processes, and which normally anchylose with the sacrum in the adult ; these can be distinguished as dorso-lumbar vertebre. Sometimes’ there may be an extra ? As in the mounted skeleton in the Bird Gallery of the British Museum. VOL. VIII.—PART vil. March, 1874. 31 386 MR. ST. GEORGE MIVART ON THE Fig. 1. RW sherwih Fig. 1. Presacral part of axial skeleton ; ©, cervical vertebrae ; CD, cervico-dorsal vertebrae ; D, dorsal vertebrae. Fig. 1*. Coceygeal vertebre. AXIAL SKELETON OF THE OSTRICH. 387 dorso-lumbar vertebra, there being three such with ribs, the first of which is connected by a styliform bone with the side of the sternal rib of the last true dorsal vertebra. The next eight vertebre (twenty-eighth to thirty-fifth) do not support long rib-like processes and are constantly anchylosed into one mass in adults; they may be called lumbar vertebre. Next follow three vertebre with long distally expanded rib-like processes abutting against the ilia. It will be at least convenient to call these sacral vertebre. At the distal end of the sacral mass we have in the adult eight vertebre (thirty-ninth to forty-sixth), which may be termed sacro-caudal vertebre. Finally there are ten postsacral vertebree normally free in the adult, except the last two; these ten are the true caudal vertebra (fig. 1*). The number of these vertebree may sometimes, however, be reduced to eight’. There are thus normally fifty-six vertebrz from the atlas to the coccyx inclusively. In some skeletons, however, there may be one or two vertebre short or a vertebra in ex- cess ; and when such divergences exist, the differential characters of all the various ver- tebre are correspondingly modified ; and this should be borne in mind when the descrip- tion here given is compared with such skeletons. THE PRESACRAL VERTEBR. THE CERVICAL VERTEBR. Tue AtiAs.—The atlas of the Ostrich presents an extreme contrast to the same bone in all mammals, even the lowest, in that it is so small a bone, being little more than an osseous ring, ventrally thickened with three short postaxial projections, and not being more than a quarter the bulk of the axis. Nevertheless, though this vertebra as a whole is relatively so small compared with the atlas of mammals, yet that part of it which is median and ventral (7. e. that hypapo- physial ossification which holds the place of a “centrum”) is relatively much larger than in any mammal. ‘This might perhaps be anticipated from the articulation of the vertebral column with the skull being median in birds, through a single condyle, instead of lateral as in mammals through a pair of condyles. The atlas of the Ostrich consists of this guast body and two neural lamine, which meet together dorsally, but do not develop a neural spine. The whole vertebra in the adult consists of one bone, no trace remaining of the primitive separation between the neural laminz and the median ventral portion. This latter (the quasi-body) apart from its junctions with the neural arch, may be said to have four surfaces—one ventral, one dorsal (or neural), one preaxial, and one postaxial ; and these four surfaces are divided by four corresponding margins. The preasial surface of the centrum, which articulates with the occipital condyle, ’ As in the mounted specimen, No. 1362, in the Museum of the College of Surgeons. ai 2 388 MR. ST. GEORGE MIVART ON THE presents a crescentic, transversely extended concavity, as it were a cup, with a consider- able median dorsal portion cut away for the reception of the odontoid process of the axis vertebra (fig. 2, ac). ATLAS VERTEBRA (natural size). Fig. 3. Aspects. Fig. 2, preaxial ; 3, postaxial; 4, lateral; 5, dorsal; 6, ventral; 7, preaxial of immature bone. ac, Preaxial articular surface of centrum ; ac’, portion of such surface formed by neural lamina ; d, diapophysis or tubercular process ; kp, hyperapophysis ; hy, hypapophysis ; pc, postaxial articular surface of centrum; pe’ lateral prolongation of the same surface ; v, canal for vertebral artery (the spiculum of bone which encloses this is seen to be wanting on one side); 2, a rounded tubercle representing a lateral hypapophysial process ; z, postzygapophysis. The dorsal margin of this preaxial surface forms thus a strongly concave line; and its ventral margin is bounded by a longer line of generally similar curvature (the horns of the crescentic articular surface being as broad as its middle); but the median portion of the ventral margin may occasionally be slightly produced (as in the specimen 1362 8), so as to cause that margin to present a rounded prominence. The postaxial surface, which is applied to the centrum of the axis ventrad of the odon- toid process, presents a crescentic transversely extended articular surface (fig. 3, pc, pc’). This surface is gently convex dorso-ventrally, but straight and flat transversely. Its upper margin is strongly concave, and is produced preaxially on each side, the postaxial articular surface of the quasi-body thus extending on each side somewhat on to the dorsal (neural) surface of the same part, and there articulating with the sides of the ventral surface of the root of the odontoid process. The ventral margin of the postaxial surface of the atlas is more or less convex, with its median portion drawn out into a postaxially projecting hypapophysial process (figs. 3, 4, & 6, hy). The neural surface of the quasi-body is gently concave transversely, with a marked pit (fig. 5) in its midst, no doubt for a ligament proceeding dorsad to the pit at the root of the odontoid process on its ventral aspect. AXTAL SKELETON OF THE OSTRICH. 389 Its pre- and postaxial margins would be nearly parallel, but that the latter is en- croached on by the lateral extensions preaxiad of the postaxial articular surface, as before mentioned. Its preaxial margin is similarly, though to a less extent, encroached on, on each side, by the postaxiad lateral extensions of the preaxial articular surface. The little lateral facets thus formed on each side of each end of the dorsal (or neural) surface articulate with the sides of the odontoid process of the axis (fig. 5, p ¢’). The ventral surface of the quasi-body is gently convex from side to side, and slightly concave antero-posteriorly in its middle. Its preaxial margin is generally slightly concave, though it may be (as in specimen 1362 B) slightly convex; its postaxial margin is nearly straight, except for the sudden production postaxiad of the median hypapophysial process before referred to (fig. 6, hy). This process has a width equal to about a third of that of the postaxial surface of the quasi-centrum ; and the length of its free projection may nearly equal or may slightly exceed its breadth. A small rounded tubercle or lateral hypapophysial process (figs. 4, 6, 7) may be developed on each side of the root of the median hypapophysial process, close to the postaxial margin of the ventral surface of the quasi-body ; and a slight ridge may run obliquely outwards and preaxially from each of these tubercles nearly to the preaxial margin of such ventral surface. The hypapophysial process is directed postaxiad, and but slightly ventrad; indeed its free extremity (as in specimen 1362 8B) may have a slightly dorsal curvature. The dorsal surface of this process articulates with the middle of the front part of the ventral surface of the centrum of the axis vertebra. The neural arch may be considered as consisting of two lateral halves dorsally united in the middle line. Each lateral half springs from the side of the quasi-centrum by a narrow portion or pedicle, expanding dorsally into a neural lamina. A narrow band or a mere spiculum of bone (figs. 2 & 3) may spring from the more postaxial part of the side of the quasi-centrum, and may fuse above with the posterior part of the pedicle, forming a foramen or short bony canal for the vertebral artery. The preaxial margin of the pedicle is strongly concave, but develops no process of any kind. Its postaxial margin is also strongly concave, and ends above in a decided postzyga- pophysis ; but before reaching that postzygapophysis it develops a more or less marked rounded prominence, which is a rudimentary tubercular process or diapophysis’ (figs. 2, 3, & 5, d). The little band or spiculum of bone before noticed, wherever it is developed, fuses ’ J use the word diapophysis to denote all parts which are homotypes of the process articulating with the tubercle of the rib when this exists. Similarly I employ the term parapophysis to denote the part with which the head of the rib articulates. 390 MR. ST. GEORGE MIVART ON THE dorsally with this diapophysial tubercle. It is thus essentially the “ neck” of an undeve- loped rib; and its ventral junction with the quasi-centrum is essentially a parapophysis or capitular process. Sometimes the postaxial margin of this diapophysial tubercle is slightly produced postaxially, thus affording a faint indication of a prominence much more developed in succeeding vertebra. The postzygapophysis projects strongly postaxiad, and has on its inner side a small rounded articular surface, which looks inwards and also slightly ventrad (figs. 3, 6, z). This surface is slightly concave dorso-ventrally, and nearly straight antero-posteriorly. On its dorsal surface each postzygapophysis bears a more or less marked rounded tubercle, which is a hyperapophysis* (figs. 3, 4, & 5, hp). From each postzygapophysis the neural lamina extends preaxially and somewhat dorsad, meeting its fellow of the opposite side without (as before said) developing any neural spine. The conjoined neural laminz thus form a low flat arch of bone, the pre- axial margin (fig. 5) of which is slightly convex, but the postaxial margin very strongly concave, on account of the strong postaxiad projection of the two postzygapophyses (fig. 3, 6, z). No foramina are developed, other than those for the vertebral arteries before described. The immature condition of the atlas vertebra (as seen in the specimen 1397 < of the College-of-Surgeons Museum) shows that the neural arches meet together dorsally, so as to form a straight, though short, antero-posteriorly directed line of junction. Similarly each pedicle joins the quasi-centrum by a line of junction which is nearly straight or slightly concave dorsally. The ventral ends of the neural laminz furnish the dorsal ends of the crescentic arti- cular surface for the occipital condyle, also the articular facets for the side of the odontoid process and for the ventral surface of its root (fig. 7, ac’). Tue AXIs. The axis of the Ostrich exceeds the atlas in size far more than is the case in the axial skeleton of any mammal. It also differs from a mammalian axis in that it is, for all its predominance over the atlas, insignificant im size when compared with the more pos- terior cervical vertebre. In the small size of the odontoid process when compared with the centrum, the Ostrich’s axis vertebra also differs from that of the highest vertebrate class. As compared with the atlas of the Ostrich, the axis differs most markedly in the great excess of its antero-posterior development over that in other directions. The axis consists of a centrum, odontoid process, and neural arch, with zygapophyses and other outgrowths or annexations. 1 This process is well seen in the cervical vertebre of the Dingo, where it was first described. See Proc. Zool Soe. 1865, pp. 574 & 579, figs. 5, 7, & 9, h. AXTAL SKELETON OF THE OSTRICH. agl AXIS VERTEBRA (natural size), Aspects. Fig. 8, lateral; 9, preaxial ; 10, postaxial; 11, dorsal ; 12, ventral: 13, preaxial; 14, lateral of immature bone. Letters as before, except az, prezygapophysis; 7s, neural spine; 0, odontoid process; pi, rib-like or pleurapophysial lamella; pz, postzygapophysis ; 7, running backwards and outwards to parapopbysis ; f, fossa (for ligament) on postaxial aspect of neural spine. The whole vertebra in the adult consists of one bone. The centrum may be considered as regards its four surfaces :— The preaxial surface, which joins the postaxial surface of the quasi-body of the atlas, presents a crescentic, transversely extended articular surface, more or less strongly concaye dorso-ventrally, but nearly straight transversely. Its ventral margin is slightly convex, generally flattened, and never produced me- dianly. Its dorsal margin is not easily defined, as the preaxial articular surface of the centrum is prolonged dorsally on to each side of the ventral aspect of the root of the odontoid process. Its dorsal margin, however, if we count such prolongations as belonging to the preaxial surface of the centrum, must be described as strongly concave (figs. 8, 9, a). The postaaial surface differs greatly in shape from the preaxial surface. Its ventral part is prolonged ventrad into a hypapophysial process; but apart from this it exhibits an articular surface the outline of which approximates slightly to that of an hour-glass (fig. 10, pe). This articular surface is strongly concave dorso-yentrally and strongly convex transversely. Its dorsal and ventral margins are both very decidedly concave, 392 MR. ST. GEORGE MIVART ON THE and its two outer margins very slightly so; the transverse extent of the dorsal margin of this surface considerably exceeds that of its ventral margin. The ventral margin of the whole centrum is of course not only convex, but much produced ventrad by reason of the hypapophysial process. The extreme width of the articular surface of the postaxial side of the centrum exceeds its dorso-ventral extent only slightly. The ventral lip of the articular surface extends much further postaxially than does its dorsal margin (fig. 8, pe); but this extension is actually less, and relatively very far less, than the postaxial extension of the hypapophysial process of the atlas. The neural surface, even apart from the dorsal surface of the odontoid process, which continues it on preaxiad, is exceedingly prolonged compared with that of the atlas. Nearly straight antero-posteriorly, it is slightly concave transversely. Certain small irregular foramina may lead into the substance of the centrum from nearly the postaxial extremity of this neural surface, the postaxial margin of which is slightly convex. The ventral surface of the centrum is nowhere concave, but is greatly produced ven- trad medianly, except at and towards its preaxial end. Its preaxial margin is nearly straight, but may be very slightly convex or very slightly concave ; its postaxial margin is narrower, but is very strongly convex (fig. 12). The median production before referred to is a strong hypapophysial (figs. 8, 12, hy) ridge, which, beginning a little preaxiad of the antero-posterior middle of the ventral surface of the centrum, rapidly increases in prominence postaxially till near the post- axial margin, when it yet more suddenly subsides. Its ventral margin is nowhere pointed, but presents a curve, the preaxial slope of which is more gradual than the postaxial one. As the hypapophysial ridge increases and decreases in dorso-ventral extent, it also, but more gradually, increases and decreases in its transverse dimension. Its ventral surface, however, is not flattened out. Two ridges (fig. 12, 7) traverse obliquely the preaxial part of the ventral surface of the centrum, diverging postaxially from a little behind its preaxial margin to the para- pophysial part hereafter to be noticed. These ridges seem to be the homotypes of those which run preaxiad from the lateral hypapophysial tubercles on the ventral surface of the atlas vertebra. The odontoid process is between three and four times the size of the hypapophysial process of the atlas vertebra. Its preaxial margin is rounded; its upper surface is flat or slightly concave, continuing on preaxially the dorsal (or neural) surface of the centrum. Its ventral aspect presents three surfaces. ‘The most postaxial of these three is the con- tinuation preaxiad of the preaxial articular surface of the centrum; and it is this part which articulates with the little articular surfaces prolonged preaxiad on to the post- axial part of the dorsal surface of the quasi-centrum of the atlas. The median surface of the three is irregularly excavated and probably gives attach- AXTAL SKELETON OF THE OSTRICH. 398 ment to a ligament going ventrad to the pit on the middle of the dorsal surface of the quasi-centrum of the atlas. The most preaxial of the three surfaces is smooth and articular, and in part joins with the preaxial articular facets of the dorsal surface of the quasi-centrum of the atlas; in part it joins the occipital condyle (fig. 9,0). It completes, as it were, the cup mainly formed by the preaxial concavity of that quasi-centrum, the dorsal excavation of which it fills up, occupying as it does the space between the horns of the preaxial arti- cular surface of the atlas. The neural arch is very much more extended antero-posteriorly, more solid, and fur- nished with more annexed processes than is the neural arch of the axis. The pedicle, or part ventral to the zygapoyhyses, is so little extended dorsally as to be insignificant. The postaxial margin of the pedicle is very concave (fig. 8). A flat band of bone ascends obliquely on each side (figs. 8, 9, p/) from the antero-lateral side of the centrum (where the transverse oblique ridge runs into the parapophysis) to between the prezygapophysis and the postzygapophysis, the band expanding dorsally and forming a sort of antero-posterior ridge or plate, connecting these zygapophyses. This band shelters beneath it the vertebral artery. On the side of the centrum and neural arch, within the band, are irregular foramina leading into the substance of the vertebra. The oblique band of bone is essentially the neck of a rib, and corresponds with the spiculum before noticed as sometimes existing beside the atlas. The dorsal expansion of the band is the diapophysis; its inferior origin is the para- pophysis. The preaxial margin of the oblique band of bone is more or less concave. A slight angular process projects postaxially from its postaxial margin (fig. 8, p/). Each prezygapophysis is small and rounded, looks outwards and slightly dorsad, is a little convex dorso-ventrally, and nearly straight antero-posteriorly (figs. 8, 9 & 11, az). The postzygapophysis is between two and four times the size of the prezygapophysis ; its antero-posterior decidedly exceeds its transverse diameter. It looks mainly ventrad, but slightly postaxiad also (figs. 10 & 12, pz). On its dorsal surface each postzygapophysis bears a very prominent and conspicuous hyperapophysis (figs. 8, 10, & 12, hp). The neural lamine at their point of union dorsally, develop a low but strong and stout neural spine or, rather, ridge (figs. 8, 9, 10, & 11, ms). This neural spine gradually rises as we proceed from the preaxial margin of the neural arch postaxially. It attains its greatest elevation at the postaxial end of the middle third of the neural arch. From that point it descends rather sharply to the postaxial border of the neural arch, presenting a margin which is slightly concave postaxially. VOL. VIII.—PART vil. March, 1874. 3K 394 MR. ST. GEORGE MIVART ON THE The neural spine is excavated by a fossa (for a ligament) on its postaxial surface (fig. 10, f); but preaxially it dips gradually to the preaxial border of the neural arch, with- out either excavation or process of any kind. The neural arch encloses a space which is smaller than that enclosed by the neural arch of the atlas; and the transverse diameter of this space is especially diminished, the arch thus appearing less low relatively. When viewed in its dorsal aspect (fig. 11), the neural arch is seen to expand greatly postaxiad, its transverse diameter between the postzygapophyses being double that between the prezypapophyses. Its preaxial margin is concave, with a median preaxial projection; its postaxial margin is generally more or less slightly concave. No foramina are developed other than those already noted. The immature condition of the axis vertebra (as seen in the specimen 1597 4 of the College-of-Surgeons Museum) shows that the neural lamin become perfectly anchy- losed together and to the centrum at a time when the neural lamine of the atlas remain altogether distinct, as also the odontoid process (or rather “ bone ”) itself. The odontoid bone forms a very large part of the so-called centrum of the axis, as well as the whole of the odontoid process itself (fig. 14, ac). All the articular surfaces of the axis for the quasi-body of the atlas are formed by this odontoid bone, except the little surfaces at the sides of the ventral surface of the root of the apparent odon- toid process of the adult. These are seen to be formed by the preaxial ventral ends of the neural laminz themselves (fig. 13, ac’). At this early stage, the lateral perforations of the vertebra (one just postaxial to the diapophysis, and one at the side of the centrum and placed more ventrally) are relatively larger and much more conspicuous. Tue THIRD VERTEBRA. The third vertebra exeeds the axis in size very much less indeed than the axis exceeds the atlas. Its most striking difference in appearance from the axis depends on the absence of the odontoid process and the larger development of the prezygapophyses, which give the bone a quadrate appearance when viewed dorsally, instead of that preaxial tapering which marks the axis (fig. 15). The vertebra consists of the same parts as the axis, except the odontoid process; and these are similarly fused into one solid and complex bone. As to the centrum, its preavial surface, which joins the postaxial surface of the centrum of the axis, presents a transversely extended articular surface, very concaye from side to side and convex dorso-ventrally; both the dorsal and ventral margins of this surface are strongly concave, while its lateral margins are slightly convex. The whole articular surface looks more ventrad than preaxiad (figs. 16, 17, ac). The postaxial surface does not differ from its preaxial surface in shape nearly so AXTAL SKELETON OF THE OSTRICH. 395 THIRD VERTEBRA (natural size). Fig. 16. Aspects. Fig. 15, dorsal; 16, ventral; 17, preaxial; 18, postaxial ; 19, lateral. Letters as before; and besides :—m, metapophysis; and ps, parapophysial root of pleurapophysial lamella. much, nor in the same way, as the postaxial surface of the centrum of the axis differs from that vertebra’s preaxial surface (fig. 18, pc). The postaxial surface of the body of the third vertebra closely resembles the corre- sponding surface of the axis, while it presents a corresponding but inverted relation to its own preaxial central surface, being strongly convex transversely and concave dorso- ventrally, the whole surface looking dorsad as well as postaxiad (figs. 15 & 19, pc). This surface differs from the corresponding surface of the axis in that its ventral part is not prolonged ventrad into a hypapophysial process, in that its extreme dorso-ventral diameter is less in proportion to its extreme transverse dimension, and in that the trans- verse extent of the dorsal margin exceeds that of the ventral margin to a less degree. Stili its ventral margin is sometimes convex (not coinciding with the ventral margin of the central articular surface) ; but its dorsal and lateral margins seem always concave. The neural surface of the centrum, but for its greater length, is very like that of the axis, abstraction being made of the (here absent) odontoid process. The preaxial margin of the neural surface is strongly concave; its postaxial margin is slightly convex (fig. 15). Bik 2 396 MR. ST. GEORGE MIVART ON THE The ventral surface of the centrum is strongly concave from side to side at its most preaxial fourth; and its preaxial margin is concave. A little preaxiad of the middle of the ventral surface a median hypapophysial process (figs. 16 & 19, hy) arises, which, for a little, gradually increases in depth postaxiad, and then proceeds directly postaxiad, coinciding with the ventral surface of the centrum—the postaxial part of the hypapo- physial process being flattened on its ventral surface, so that this may be said to have two lateral margins which diverge postaxially. The hinder margin of the hypapophysis is at the same time the hinder margin of the centrum and is strongly convex. The hypapophysis does not extend nearly so suddenly and prominently ventrad as does the hypapophysis of the axis. No oblique ridges diverge postaxially from near the preaxial margin of the third ver- tebra, as they do in that of the axis; but there is a great prominence on each side near the preaxial margin, and it is their projection ventrad which makes the preaxial part of the ventral surface concave. The neural arch is much like that of the axis, except that its prezygapophysial and pleurapophysial parts (fig. 19, az & pl) are much more developed, while its neural spine is less so. The pedicle is similarly low, with a concave postaxial margin. Its preaxial margin is more extended dorso-ventrally and more concave than is that of the pedicle of the axis vertebra. The flat pleurapophysial band of bone (for the vertebral artery) has greatly increased in antero-posterior extent, compared with that of the axis vertebra, being about equal to half the extent of the centrum in this dimension. It extends oblique dorsad from the parapophysis and expands till it merges into, or, rather, comes to constitute an interzyga- pophysial ridge, the anterior end of which may appear as a slight prominence (metapo- physial) outside and beneath the prezygapophysis (as in specimen 1562 8 on the left side). The preaxial margin of the pleurapophysial lamella is concave. Its postaxial margin is irregular from defective ossification; but it may (as in 1362 a, right side) develop a slight median, postaxial, bluntly pointed prominence (fig. 19). At the middle of the side of the neural arch, near the dorsal postaxial end of the plurapophysial lamella, is a depression with irregular foramina, leading into the sub- stance of the bone. Kach prezygapophysis presents an oblong articular surface, which looks preaxially and dorsally, and is nearly, if not quite, twice as long antero-posteriorly as it is broad. It is slightly convex in both directions, but more so antero-posteriorly (figs. 15, 17, & 19, az). Each postzygapophysis (figs. 16, 18, & 19, pz) is about the same size as the prezyga- pophysis, but is slightly less developed in the antero-posterior direction. Its antero- posterior diameter very much exceeds its transverse diameter, as much as, or more so, than in the axis vertebra. The articular surface is slightly concave transversely and nearly flat antero-posteriorly. AXIAL SKELETON OF THE OSTRICH. 397 It looks mainly ventrad, but slightly postaxiad also. On its dorsal surface each postzygapophysis bears a prominent hyperapophysis, which, however, is considerably smaller than in the axis vertebra (figs. 15 & 19 hp). The neural lamin develop a low neural spine, which, however, extends less both dorsally and antero-posteriorly than does the neurapophysis of the axis. Instead of rising gradually from the preaxial margin of the neural arch, it springs up suddenly, at some distance postaxiad from that preaxial margin (figs. 15 & 19, ms). It subsides less suddenly, at about the same distance from the postaxial margin of the neural arch as its origin is distant from the preaxial margin of that arch. It occupies rather less than the middle three fifths of the dorsal surface of the neural arch. The neural spine is excavated medianly on its postaxial surface (a fossa for a liga- ment thus being formed) in the same way that the same part of the axis is excavated (fig. 18, f). The preaxial surface of the neural spine of this third vertebra, however, is also exca- vated, and in such a way as to cause that spine to bifurcate preaxially (figs. 15, 17, & 19,ns), the two preaxial margins of the depression projecting preaxiad at their dorsal ends, and making the short preaxial margin of the neural spine concave. The neural arch encloses a space of very little different width from that enclosed by the neural arch of the axis (fig. 17). When viewed above. the neural arch is subquadrate, the transverse diameter of the prezygapophyses about equalling that of the postzygapophyses. Its preaxial margin is much more strongly concave than is the corresponding margin of the axis ; its posterior margin may be slightly concave or nearly straight (fig. 15). Tue FourtH VERTEBRA. The fourth vertebra exceeds the third about as much as the third exceeds the second (counting the odontoid process); but in general appearance and arrangement and development of parts, the fourth vertebra very much more nearly agrees with the third than does the third with the second. As to the centrum, its preavial surface is quite like that of the third vertebra, except that it is rather more extended transversely and looks slightly less ventrad (figs. 21 & 24, ac). The postaxial surface only differs from the preaxial surface in the same antithetical way that the two corresponding surfaces of the third vertebra differ from each other (figs. 20 & 22, pe). It differs from the postaxial surface of the third in that its ventral margin is more strongly concave and nearly coincides with the concave ventral margin of the articular surface, and in that the transverse extent of the dorsal part of that surface is nearly equalled by the transverse extent of its ventral part. The neural surface of the centrum closely resembles that of the third vertebra, except that, of course, its absolute length is greater. 398 MR. ST. GEORGE MIVART ON THE The ventral surface of the centrum agrees with that of the third vertebra, except that the median hypapophysis is much less developed (fig. 20, hy), being only in the form of a slight ridge extending antero-posteriorly along the middle third of the centrum. FOURTH VERTEBRA (natural size). Aspects. Fig. 20, lateral; 21, preaxial; 22, postaxial ; 23, dorsal; 24, ventral. Letters as before, except that p denotes parapophysis, and ps the rib-like, freely projecting, pleurapophysial process. It also differs from the third vertebra in that its most postaxial part is decidedly, though slightly, concave transversely, and in that the parapophyses, project a little more sharply and strongly ventrad. The neural arch is very like that of the third vertebra, except that its pleurapo- physial part is more developed, while its neurapophysis, being actually about the same size, is relatively less. The pleurapophysial band is rather more extended antero-posteriorly, though it does not quite attain the postzygapophysis as it does in the third vertebra; while from the ventral end of its postaxial margin (7. ¢. from the parapophysis) a long, tapering, styliform, rib-like process projects in nearly a straight line postaxially (fig. 20, ps). AXIAL SKELETON OF THE OSTRICH. 399 The length of the free part of this process is less than half the antero-posterior extent of the centrum, though its extremity may project beyond the antero-posterior median point of the centrum. The postaxial margin of the pleurapophysial band is irregular, often more or less slightly concave, and proceeds obliquely dorsad and postaxiad. At the anterior part of this band a slight metapophysial prominence may appear between the preaxial end of the parapophysis and the prezygapophysis. The zygapophyses closely resemble those of the third vertebra. The hyperapophyses on the postzygapophyses are absolutely, and still more relatively, smaller than in the third vertebra, though they are still noticeable prominences (figs. 20, & 23, hp). The neural spine and the neural arch generally agree with those of the third vertebra, except that the latter is less quadrate when viewed from above. ‘This is owing to the transverse diameter of the preaxial part of the vertebra (measured across the prezygapophyses) considerably exceeding that of its postaxial part (measured across the postzygapophyses), instead of these two dimensions being equal as in the third vertebra. The postaxial margin of the neural arch is also always decidedly concave (fig. 23), instead of being nearly straight as sometimes is the case in the third vertebra. Tue Firta VERTEBRA. The fifth vertebra, though again a large bone antero-posteriorly, more closely resembles in shape the fourth vertebra than does the fourth the third (fig. 1, C, v). The preaxial surface of its centrum quite agrees with that of the fourth vertebra, while its postaxial surface differs only in having its articular surface relatively broader and narrower dorso-ventrally in its middle (the two sides of the hour-glass coming there nearer together), and in having the ventral margin of that surface very consider- ably more extended than its dorsal margin, and, concomitantly, its lateral margins rather more concave. The neural surface is similar to that of the centrum of the fourth vertebra, except that it is perhaps rather more exposed at its preaxial end, through the slightly greater cutting away, as it were, of the more preaxial part of the neural arch. The ventral surface of the centrum agrees with that of the fourth vertebra, except that its postaxial part is more decidedly and extensively concave transversely, and that the median hypapophysial ridge is still less developed. The neural arch is, of course, longer than that of the fourth vertebra, but is other- wise very like it. The pleurapophysial band more decidedly fails to attain the post- zygapophysis ; while at the dorsal end of its slightly concave, oblique, postaxial margin a slight pointed process projects postaxiad, which process is the homotype of the slight projection noticed as extending from the postaxial margin of the band in the axis 400 MR. ST. GEORGE MIVART ON THE vertebra. It has here mounted more dorsally. The styliform rib-like process is rather longer in this fifth vertebra, but otherwise like that of the fourth. The metapophysis is rather more swollen and prominent; but the hyperapophysis is still more reduced. The prezygapophysis projects preaxially beyond the parapophysis to a less degree than is the case in the fourth vertebra. The neural spine is relatively less than that of the fourth vertebra, but otherwise resembles it, except that the two margins of the preaxial excavation do not stand out preaxially on processes with near so much distinctness. The neural arch when viewed from above is still less quadrate than in the fourth vertebra, because the transverse extent of the preaxial end of the fifth vertebra exceeds that of the postaxial part of the same vertebra more than the one exceeds the other in the fourth vertebra. The postaxial margin of the neural arch is also much more decidedly and sharply concave. THE SrIxtH VERTEBRA. The sixth vertebra, though a larger bone than the fifth, resembles the latter in form and proportions, even more than the fifth vertebra resembles the fourth. Both the VENTRAL ASPECT OF SIXTH VERTEBRA (uatural size). Fig. 25. ac Letters as in last figures, and in addition c, catapophysis. pre- and postaxial surfaces of the centrum agree with those of the fifth vertebra respectively ; and the neural surface only differs from that of its serial predecessor in being slightly more exposed preaxially. The ventral surface exhibits no noticeable differences, except that the hypapophysial ridge is still more obsolete, and that the postero-inner part of each parapophysis begins AXIAL SKELETON OF THE OSTRICH. 401 to exhibit a special though slight prominence (directed medianly inwards and ventrad), which is the beginning of the differentiation of the catapophysis’ (fig. 25, c). The metapophysis may be again slightly more prominent; while the hyperapophysis is so reduced that it would perhaps escape notice were it not traced from the more preaxial vertebre. The prezygapophysis projects beyond the parapophysis still less than in the fifth vertebra, sometimes scarcely or not at all so. Its articular surface is even also relatively longer and narrower than in the bone last described. The neural spine is excavated, both pre- and postaxially, more obliquely and ex- tensively than in the fifth vertebra; so that its median unexcavated part is relatively, and may be absolutely, shorter. The lateral margins of the preaxial excavation do not at all develop preaxial processes. Both the pre- and postaxial margins of the neural arch are rather more concave than in the fifth vertebra. THE SEVENTH VERTEBRA. The seventh vertebra is rather larger than the sixth vertebra, but in other respects is so complete a repetition of the latter as hardly to need distinct notice except in the following few points. The catapophyses and metapophyses are rather more prominent; and the neural spine has (at least sometimes) a greater development dorsad. The prezygapophyses do not project forwards quite so much as do the parapophyses, instead of projecting, in a greater or less degree, more preaxially, as has always been the case in the vertebre hitherto described ; their articular surfaces are directed rather more inwards than are those of the sixth vertebra. Tue Erenutu, Nintu, AND TENTH VERTEBR&. These vertebre slightly increase in length antero-posteriorly, though the rate of increase diminishes as we proceed postaxially. In these three vertebre the catapo- physial projection gradually becomes more marked; and thereby the preaxial part of the ventral surface of the centrum becomes more concave, as also the ventral margin of its preaxial surface. The metapophyses remain as marked (fig. 28, m) or become rather more so, while the hyperapophyses disappear. ! From xara, down. These processes are, as it were, merely transitional processes, sometimes merging with the parapophysis dorsally, and sometimes with the hypapophysis yentrally. By giving them therefore a distinct name I by no means wish to ignore their transition; but as these parts are often so distinct and conspicuous as to require description and distinct reference, it is convenient to be able to speak of them substantively and adjectively by a distinct term. By catapophyses I mean parts which are distinguishable from, but intermediate between, parapophyses and hypapophyses, and which are placed on the ventral side of the former. VOL. VIII.—PaART vil. March, 1874. Bib 402 MR. ST. GEORGE MIVART ON THE The interval between the pre- and postaxial excavations of the neural spine pro- gressively increases. In the tenth vertebra the transverse dimension of the postaxial part of the neural arch (measured from outer margin to outer margin of the two postzygapophyses) begins to be slightly less in defect compared with the same dimension of the preaxial part of that arch. PREAXIAL ASPECT OF EIGHTH VERTEBRA. VENTRAL ASPECT OF EIGHTH VERTEBRA. Immature condition (natural size). Immature condition (natural size). Fig. 26. Fig. 27. The rib-like part, ps, is removed on the left side ; Letters as before, except /, catapophysis, and d, diapophysis. +, parapophysis. NINTH VERTEBRA (size 3 of nature). Fig. 28. Dorsal aspect. Letters as before. In the tenth vertebra also the prezygapophysis projects about as far preaxially as does the parapophysis. THE ELEVENTH VERTEBRA. The eleventh vertebra is of very nearly the same length as the tenth, especially as estimated by the antero-posterior extent between the preaxial end of the prezygapo- physis and the postaxial end of the postzygapophysis. This vertebra, however, is very slightly more massive than the preceding one. AXIAL SKELETON OF THE OSTRICH. 405 The postaxial articular surface of the centrum has the transverse extent of its ventral margin very little in excess of that of its dorsal margin. The articular surfaces of the prezygapophyses look more directly dorsad and less inwardly; those of the postzyya- pophyses are broader. The median unexcavated part of the neural spine is again more or less longer antero- posteriorly. The postaxial part of the neural arch still more nearly approaches in width the preaxial part of that arch than in the tenth vertebra; and its postaxial margin is less concave. THe Twetrtd VERTEBRA. The twelfth vertebra carries on the progressive modifications indicated in the eleventh vertebra, and it is slightly longer and more massive than that vertebra. The catapophyses may here first project ventrally as sharply marked — processes (figs. 30 & 32, ¢). TWELFTH VERTEBRA (2 natural size). Te iN Aspects. Fig. 29, lateral ; 30, postaxial; 31, dorsal; and 32, ventral. Letters as before; c, catapophysis. The transverse extent of the postaxial part of the neural arch nearly equals that of its preaxial part. The styliform rib here generally attains its maximum of length. 312 404 MR. ST. GEORGE MIVART ON THE Tur THIRTEENTH VERTEBRA. This vertebra slightly exceeds the twelfth in length, but still more in breadth and stoutness. Here, once more, the dorso-yentral extent of the lateral margin of the post- axial surface of the centrum equals the transverse extent of the dorsal margin of that surface. The neural arch here attains about its maximum of antero-posterior extent, measured pre- and postaxially in the middle line. The median unexcavated part of the neural spine is slightly shorter antero-posteriorly than in the twelfth vertebra. The free projection of the styliform rib may be for the first time less in extent than in the vertebra next (preaxially) to it; but it is at the same time thicker. The transverse extent of the postaxial part of the neural arch nearly equals that of its preaxial part; and concomitantly with this greater projection outwards of the post- zygapophyses, the lateral margins of the neural laminz (viewed dorsally) become more concave. The articular surfaces of both zygapophyses, but especially of the postzygapophyses, are broader in proportion to their length. Tue FouRTEENTH VERTEBRA. The fourteenth vertebra scarcely exceeds the thirteenth in length, though it does so very decidedly in breadth. Indeed the antero-posterior length of the neural arch in the dorsal middle line is absolutely, though very slightly, less than in the thirteenth vertebra. The styliform ribs are thicker, the catapophyses more projecting, the zygapophysial articular surfaces broader, while the four margins of the postaxial articular surface of the centrum have become about equal. The canal for the vertebral artery has also become rather more capacious. Tue FirreentH VERTEBRA. With this vertebra the absolute antero-posterior dimension has begun slightly to decrease; but the transverse development continues to progress. ‘The styliform ribs are still stouter, the zygapophysial surfaces still broader, the neural spine thicker and shorter, and the lateral margins of the neural arch more concave. The postaxial excavation of the neural spine is enlarged into a considerable fossa. The metapophysis is rather prominent beneath a prezygapophysis, which may here once again begin to decidedly project, preaxiad, beyond the parapophyses. THE SIXTEENTH VERTEBRA. The sixteenth vertebra is, again, like the fifteenth, shorter than its preaxial predecessor ; and even the transverse dimension, measured across the postzygapophyses, is scarcely if at all greater, though the preaxial part of the vertebra is enlarged transversely, and therefore relatively to the more postaxial part as well as absolutely. The posterior end AXIAL SKELETON OF THE OSTRICH. 405 of the ventral surface of the centrum has notably increased in width. ‘The styliform ribs are no longer styliform, but stout obtusely pointed processes, projecting, however, postaxiad and slightly ventrad as usual (fig. 54, ps). The catapophyses (figs. 33 & 34, c) are very largely developed, projecting not only ventrad but somewhat proximad also; they are so extended inwards that the interval between their ventral ends is decidedly less than the breadth of the middle of the ventral surface of the centrum behind them, a condition which we have not yet met with. The middle of the ventral surface of the centrum may still show a trace of the antero-posteriorly extended hypapophysial ridge (fig. 34, hy). SIXTEENTH VERTEBRA (2 natural size). Fig. 33. Fig. 34. Aspects. Fig. 33, lateral; 34, ventral. Letters as before. The pleurapophysial lamella is beginning decidedly to diminish in antero-posterior extent; but the metapophysis is largely prominent beneath a prezygapophysis, which may project very considerably preaxiad beyond the parapophysis (figs. 33 & 34, 7). The low neural spine is as thick as in the fifteenth vertebra and also shorter antero- posteriorly ; its posterior excavation is enlarged into a deep conical fossa, the dorso- lateral margins of which diverge to the postzygapophyses—without, however, sheltering it or covering it in, except to a minute extent on each side. The articular surfaces of the postzygapophyses are nearly as wide transversely as long pre- and postaxially. Tur SEVENTEENTH VERTEBRA. The seventeenth vertebra differs very considerably from the sixteenth, and more than the latter differs from the fifteenth. It is shorter and at the same time broader, while the breadth of its preaxial part is still more in excess of that of its postaxial part (fig. 38) than was the case in the sixteenth vertebra. Thus we have a return to a predominance which existed in more preaxially situated cervical vertebre (fig. 25). 406 MR. ST. GEORGE MIVART ON THE As regards the centrum, its preaxial surface has the shape of its ventral margin changed by the appearance of two small hypapophysial processes (fig. 36, hy). Its postazial surface is larger both actually and relatively; while its ventral margin has once more become more extended transversely than its dorsal margin, and at the same time is less concave than in the vertebra last described (fig. 37, pc). The neural surface of the centrum may be more exposed by the further cutting away, as it were, of the preaxial part of the neural arch. The ventral surface of the centrum presents somewhat of a return to the proportions of the third vertebra, if abstraction is made of the hypapophyses (fig. 39). SEVENTEENTH VERTEBRA (2 natural size). Fig. 35. Fig. 36. Aspects. Fig. 35, lateral ; 36, preaxial; 37, postaxial; 38, dorsal; and 39, ventral. Letters as before, These latter processes (which may conveniently be said to represent and take the place of absent catapophyses) project as a pair of short processes, springing from beneath about the middle (both antero-posteriorly and transversely) of the centrum. They extend ventrally and preaxially, and diverge from each other towards their apices, instead of converging like the catapophyses of the sixteenth vertebra. The hypapophyses may be separated by a notch which extends dorsad to the general level of the ventral surface of the centrum, or may be only imperfectly divided from one another as in the specimen figured. AXIAL SKELETON OF THE OSTRICH. 407 In front of these hypapophyses the ventral surface of the centrum presents a wide and rather shallow transverse concavity, bounded by two diverging ridges, which proceed respectively from the root of each hypapophysis to the adjacent parapophysis. Behind the hypapophyses there is no trace whatever of a median hypapophysial ridge, but the centrum widens rapidly backwards into a nearly flat slightly concave surface with a rounded postaxial margin. In the place of styliform ribs there are conical prominences so short and stumpy that their nature would hardly be suspected but for the conditions presented by more pre- axial vertebre ; their direction, however, is the same as heretofore (figs. 35 & 36, ps). The pleurapophysial bands of bone are still less extended antero-posteriorly, though they enclose a canal for the vertebral artery, which has now become so large as nearly to equal the neural canal in diameter (fig. 37, v). The metapophysis (figs. 35, 36 & 38m) is very largely developed; but the prezygapo- physis extends less decidedly preaxiad of the parapophysis than in the sixteenth vertebra. The postzygapophysis also projects postaxially a little beyond the centrum. The prezygapophysial articular surfaces look slightly more inwards, and the postzyga- pophysial ones decidedly more outwards, than in the sixteenth or preceding vertebre. The neural spine has become so wide as (it may be for the first time) to exceed in transverse extent the dorsal surface of the neural lamine on either side of it. The anterior excavation of the neural spine (fig. 36, ns) has become a considerable and open fossa, while its posterior excavation (fig. 37, /) has assumed very large proportions, with a vertical preaxial wall, and laterally more or less sheltered and overshadowed by the projecting margins of the ridges, which diverge postaxially from the neural spine to the postzygapophyses. The neural spine itself is very little extended antero-posteriorly, less than in any of the nine preceding vertebre; but it has a greater extension dorsad. One or two conspicuous foramina lead from the side of the centrum or neural arch into the substance of the bone. CERVICO-DORSAL VERTEBR. Tue EIGHTEENTH VERTEBRA. The eighteenth vertebra differs as much from the seventeenth as does the latter from the sixteenth. It is a larger bone than the seventeenth vertebra, inasmuch as, though the centrum is shorter, the extreme antero-posterior extent of the whole vertebra is not less, while its transverse and its dorso-ventral dimensions are augmented. The excess of the transverse diameter of its preaxial part over that of its postaxial part is greater than even in the seventeenth vertebra (fig. 41). This vertebra bears the first rib. 408 MR. ST. GEORGE MIVART ON THE The centrum has its preaxial articular surface with its dorsal margin more concave and its ventral margin less concave than in the seventeenth vertebra. ‘The outline of the inferior margin of the whole centrum is modified by the hypapophysial process. The postaxial surface of the centrum is relatively more extended transversely, with its ventral margin wider and more in excess of its dorsal margin than in the seventeenth vertebra, while at the same time the ventral margin is still less concave. EIGHTEENTH VERTEBRA (3 natural size). Fig. 40. Fig. 41. kd S Fig. 40, lateral aspect; 41, ventral aspect. Letters the same, except that d denotes the diapophysis. The ventral surface of the centrum is wide and but very slightly concave at either its preaxial or its postaxial part. There is a very thick median hypapophysis extend- ing from the middle of the ventral surface, and more or less bifurcating ventrally into two short, rounded, diverging processes (fig. 41, hy). No diverging or other ridges connect the hypapophysis with the parapophyses; but a ridge may run postaxially from each parapophysis to the postaxial margin of the ventral surface of the centrum, the two ridges bounding that surface laterally. The postaxial margin of the ventral surface is less convex than in the seventeenth vertebra. If there is no free rib, the pleurapophysial band of bone representing it is extremely short antero-posteriorly. Each margin of it is concave, while from the ventral end of its postaxial margin a triangular blunt prominence (the last rudiment of the styliform rib of more preaxial vertebre) projects postaxially, but at the same time in a more ventral direction than that taken by any of the styliform ribs before described. Such is its condition in a highly ossified skeleton; but very often the pleurapophysial band appears as a separate rib, and in this free condition it is described below as the first rib (fig. 75, 1). This osseous band, when it is anchylosed to the vertebra, bounds externally a bony canal for the vertebral artery, which is so large that it exceeds in diameter the neural canal itself. This arterial canal is bounded internally by the neural lamina and inferiorly by the AXTAL SKELETON OF THE OSTRICH. 409 great lateral projection of the parapophysis, on to which here, as in the more preaxial vertebrie, the preaxial articular surface of the centrum extends itself. Superiorly the canal for the artery is bounded by a process of bone, which extends outwards from beside and beneath the prezygapophysis (being in fact a diapophysis, or tubercular process)', on the outer side of which is to be seen a prominence still representing the metapophysis (figs. 40 & 41, m). These parts exist in the same way in preceding vertebree; but they are most con- veniently noticed here on account of the large size they attain on the dorsal and ventral sides respectively of the great canal for the vertebral artery. The articular surfaces of the prezygapophyses look more inwardly, and those of the postzygapophyses more outwardly, than do the corresponding surfaces of the seventeenth vertebra; and the postzygapophyses also extend postaxially beyond the centrum to a greater degree (fig. 40, pz). The neural spine may be longer antero-posteriorly, and is wider transversely as well as more extended dorsad than is the neural spine of the preceding vertebra. Its anterior excavation is larger and has a nearly vertical postaxial wall, with two prominent ridges, which bound it laterally and proceed preaxially and nearly parallel to near the preaxial margin of the neural arch. The posterior excavation of the neural spine is at about its maximum of development in this vertebra, and is overshadowed and protected by the postaxial extent of its dorsal lateral margins, which diverge from the neural spine to the postzygapophyses. The pre- and postaxial margins of the neural arch are strongly concave, though the former is not so much so as is the corresponding part in the seventeenth vertebra. Sometimes two conspicuous foramina lead into the substance of the bone on either side. One of these is placed a little on the ventral side of the interzygapophysial ridge, rather nearer to the postaxial margin of the pedicle of the neural arch than to its preaxial border; the other may be sometimes found at the side of the middle of the centrum, directly ventrad to the first foramen. THE NINETEENTH VERTEBRA. The nineteenth vertebra is very like the eighteenth in size and general shape, although it has never any pleurapophysial osseous band, but always an articulated rib (the second rib) instead. This absence of course produces a striking difference in the general appearance of the vertebra when it is viewed preaxially. The centrum presents a preaxial surface which ends in an obtusely pointed process on each side; and both the dorsal and ventral margins of its articular surface are less concave than are the corresponding margins of the eighteenth vertebra (fig. 43, ae). 1 The diapophysis is that part to which the tubercle of the rib is attached, while its head joins the parapophysis. VOL. vilt.— PART vu. March, 1874. 3M 410 MR. ST. GEORGE MIVART ON THE The postaaxial aspect of the centrum shows an articular surface which may be more quadrate, its ventral margin often less preponderating over its dorsal margin than in the preceding vertebra. The latter is also less concave, while the former is nearly straight and may even be very slightly convex. The ventral surface has its preaxial margin more expanded and less sharply concave, and its postaxial margin often less convex than is the case with the eighteenth vertebra. The hypapophysis is a single rounded, obtuse tubercle. It extends rather more ventrad and sometimes less preaxiad than does that of the preceding vertebra. The under surface of the centrum may be concave antero-posteriorly in the middle line behind the hypapophysis. NINETEENTH VERTEBRA (4 natural size). Fig. 43. Aspects. Fig. 42, lateral; 48, preaxial; 44, dorsal. Letters as before. The ventral aspect of the hypapophysis exhibits, as it were, a slight lingering tendency to bifurcate. Behind the hypapophysis the surface of the centrum begins to be slightly convex transversely instead of being concave. Not only is there, as has been said, no pleurapophysial band of bone, but no ridge runs postaxially from the parapophysis, the side of the centrum being continued unin- terruptedly dorsad into the side of the neural arch, the whole forming one concavo- convex expanse. The parapophysis extends preaxially and slightly outwards, and has at the outside of it a small, rather deep, concavity for the head of the rib, and on its inner side the lateral continuation of the preaxial central articular surface (fig. 42, p). The diapophysis projects slightly more outwards from the prezygapophysis, is more antero-posteriorly extended; and the dorso-ventral distance between it and the parapo- physis is greater than heretofore, extending to the preaxial margin of the neural lamina. From its ventral outer side, near its preaxial end, the diapophysis develops a rounded prominence (fig. 42, d) to receive the tubercle of the rib. Beneath it the preaxial surface of the neural lamina is deeply and irregularly excavated, and another AXIAL SKELETON OF THE OSTRICH. 411 smaller excavation (either several small openings or one large foramen) is placed a little preaxially of the postaxial margin of the neural lamina, on the ventral side of the interzygapophysial ridge. The neural spine has its dorsal margin more antero-posteriorly extended than in the eighteenth vertebra; and consequently the margins of the ridges running from it to the postzygapophyses, instead of sloping gradually, run at first a little preaxially, thus forming a slightly acute angle. The postaxial excavation of the neural spine is much as in the eighteenth vertebra, though somewhat less deep. The preaxial surface of the neural spine is vertical, but cannot always be said to be medianly excavated; rather there is sometimes a slight depression on each side of it on the neural lamina. The prezygapophyses are broader than in the eighteenth vertebra, and look rather more inwards. The postzygapophyses are smaller and look even more outwardly. For the first time their transverse (here actually dorso-ventral) diameter may exceed their antero-posterior development; and they are so produced that their articular surfaces may be altogether postaxial to the postaxial margin of the centrum. ‘The preaxial margin of the neural arch begins to show a little median preaxially directed process (fig. 44). Tue TWENTIETH VERTEBRA. The twentieth vertebra closely resembles the nineteenth, with the centrum, diapo- physes, and neural spines slightly enlarged. The centrum has its preavial articular surface relatively deeper dorso-ventrally ; and its lateral parapophysial terminations are not so much laterally produced. TWENTIETH VERTEBRA (3 natural size). Fig. 45, lateral aspect; 46, postaxial aspect. Letters as before. The pre- and postaxial diapophysial excavations of the neural lamina are the same as in the nineteenth vertebra; only the postaxial excavation is relatively the more extensive. The postaxial surface has its ventral margin more conyex, but relatively shorter compared with its dorsal margin, than in the nineteenth vertebra; the whole articular surface begins also to be less convex transversely and less concave dorso-ventrally. 3M2 412 MR. ST. GEORGE MIVART ON THE The neural surface of the centrum is still more exposed than heretofore by the cutting away, as it were, of the preaxial part of the neural arch, which has a more marked median process. The ventral surface is nowhere concave transversely, though somewhat so antero- posteriorly; not, however, to the same degree as in the nineteenth vertebra. The preaxial margin is also somewhat less concave, and the postaxial margin less convex. There is a median hypapophysis which may or may not be somewhat expanded laterally towards its tip, and which may or may not project more strongly preaxiad than in the nineteenth vertebra, and may have its dorsal surface concave transversely. The ventral surface of its apex may be marked with a short, faint, antero-posterior groove, the last trace of the bifurcation of the process in more preaxial vertebre. The side of the centrum is continuous with that of the neural arch, exhibiting an undulating expanse (concavo-convex dorso-ventrally), which is smooth, save that there may be a scarcely perceptible indication of a ridge running obliquely ventrad from the capitular process to the postaxial margin of the centrum. The parapophysis extends less both preaxially and outwards than in the nineteenth vertebra; but the articular fossa for the head of the rib is larger, though nearly, if not quite, as deep (fig. 45, p). The diapophysis again projects more outwards, is more extended antero-posteriorly, and more remote from the parapophysis than in the preceding vertebra. It bears on its ventral side, more close to its external margin, a convex articular surface (fig. 45, d) for the tubercle of the rib, which surface is rather more elongated and less rounded than it is in the nineteenth vertebra. It is placed about as near to the preaxial margin of the diapophysis as to its postaxial margin, or rather nearer; and therefore more postaxially with relation to the whole, more extended, diapophysis than in the more preaxial vertebra. Ventral and preaxial to this surface the preaxial aspect of the neural lamina (pedicle) is deeply, widely, and irregularly excavated ; and another excavation is placed near the postaxial margin of the pedicle, in the same position as in the nineteenth vertebra, but larger in size. The neural spine is more extended both dorsally and antero-posteriorly, having a straight dorsal margin, which about equals in length its preaxial margin and also the transverse diameter of the postaxial articular surface of the centrum. The postaxial excavation of the spine extends less dorsally, the dorsal part of the neural spine presenting postaxially a single margin, instead of a flat surface bounded by two ridges as in the nineteenth vertebra. From the side of the neural spine two laminz of bone proceed postaxially to the postzygapophyses, bounding externally that fossa behind the neural spine which repre- sents the postaxial excavation of that spine, which we have seen to exist in vertebra nearer the skull. AXIAL SKELETON OF THE OSTRICH. 413 The dorsal margins of these lamine form nearly a right angle with the postaxial margin of the neural spine. The articular surfaces of the zygapophyses are still smaller; and the postaxial ones pro- ject still more postaxially beyond the centrum than in the nineteenth vertebra (fig. 45, pz). The twentienth vertebra supports a long rib, which ends freely and does not come into contact with any sternal rib, though it is related to the first of these (fig. 75, 111). THE DORSAL VERTEBRA. THe TWENTY-FIRST VERTEBRA. The twenty-first vertebra is so much like the twentieth that little need be said in its description. The preaxial articular surface of the centrum is deeper dorso-ventrally. The hypapophysis is generally smaller, especially narrower from side to side, and its preaxial margin is more concave; it may show a rather more marked tendency to distal bifurcation (fig. 48, hy). No oblique ridge traverses the side of the centrum; but there may be some defects of ossification dorsally, just ventrad of the diapophysis. The parapophysis extends less ; but its articular fossa (p) is larger. TWENTY-FIRST VERTEBRA (3 natural size.) Fig. 47. Fig. 48. q as Fig. 47, lateral aspect; 48, ventral surface. Letters as before. The diapophysial excavations are plainly to be seen; the postaxial one has the form of a single foramen or fossa. Besides these, there is a defect of ossification in the side of the centrum, just below the middle of the diapophysis. The diapophysis projects obliquely postaxiad and slightly dorsad, as well as outwards, and is much more remote from the parapophysis than in the twentieth vertebra. The tubercular convexity (d) is less marked, and is situated almost as near the post- axial border of the diapophysis as it is near its preaxial margin. 414 MR. ST. GEORGE MIVART ON THE Excavations occur preaxially, as in the preceding vertebra; but postaxially they may exist close to the postaxial margin of the diapophysis, extending inwards to the base of the neural spine. The neural spine is more extended dorsally and slightly so antero-posteriorly ; its dorsal margin is rather convex, and bifurcates slightly at each end. That concavity which in the preceding vertebra exists on each side of the preaxial root of the neural spine is here deeper. The postaxial fossa is smaller; and the margins of the laminz which laterally bound it form rather less than a right angle with the postaxial margin of the neural spine. The zygapophysial articular surfaces are again slightly diminished. The twenty-first vertebra supports a long rib, which articulates by its distal end with the second sternal rib. Tur TWENTY-SECOND VERTEBRA. This vertebra is of about the same size as the twenty-first, in some respects less deve- loped, in others more so. Its centrum has its preaaial surface entirely occupied by the articular surface, and the ventral margins of both coincide, as there is no hypapophysial production. The surface is deeper dorso-ventrally, is less laterally produced, and may have its dorsal and ventral margins more concave and convex respectively. TWENTY-SECOND VERTEBRA (3 natural size). Fig. 49. PAO NI, Fig. 49, lateral aspect ; 50, dorsal aspect; 51, preaxial aspect. Letters as before, except v, excavation on preaxial side of diapophysis. Excavations are to be seen on the preaxial, postaxial, and ventral aspects of the diapophysis, the preaxial one being very much the larger. The postaxial surface of the centrum is rather larger and flatter, but very similar in figure to that of the twentieth vertebra; its dorsal margin, however, is rather wide and less concave. The neural surface of the centrum is more concave transversely at its preaxial part than in the vertebra last described. AXIAL SKELETON OF THE OSTRICH. 415 The ventral surface is destitute of any hypapophysis; but its nearly straight median portion is in the form of a slightly marked antero-posterior ridge. On each side of this the centrum rounds off, so that its ventral surface is convex transversely throughout ; its lateral margins are, when viewed dorsally (fig. 50), strongly concave, owing to the ex- pansion of the pre- and postaxial ends of the centrum. ‘The concavity of the preaxial margin is much as in the twenty-first vertebra; but the postaxial margin is less convex. The sides of the centrum and neural arch are continuous, without any trace even of a ridge; but there may be defects of ossification beneath the diapophysis. The parapophysis extends out less even than in the twenty-first vertebra; and its arti- cular fossa is larger (especially dorso-ventrally), but shallower. The diapophysis is much longer and very much produced obliquely, postaxiad and slightly dorsad; its remoteness from the parapophysis is again augmented in passing to this vertebra from the twentieth. The convexity for the tubercle of this rib is, like that of the twenty-first vertebra, smaller and less marked than in the twentieth vertebra, and may be nearer to the post- axial margin of the diapophysis that it is to its preaxial margin. The distal part of the diapophysis is more antero-posteriorly extended than the more proximal part. Numerous irregular preaxial diapophysial excavations are placed beneath the ridge running from the prezygapophysis to the diapophysis ; and other large exavations appear behind the diapophysis and on each side of the postzygapophysis, extending close to the ventral side of the postaxial margin of the diapophysis, which margin runs inwards to the base of the neural spine. | The neural spine is more extended dorsally, but not antero-posteriorly ; it has a slightly convex dorsal margin, and bifurcates again at each end (fig. 50, ns). That concavity which in the two preceding vertebre exists on each side of the preaxial root of the neural spine is here yet further deepened. The postaxial fossa of the neural spine is smaller; and the margins of the lamin which laterally bound it form an obtuse angle with the postaxial margin of that spine. The articular surfaces of the zygapophyses are again still smaller, and look almost entirely inwards and outwards respectively. Almost the whole surface of the postzyga- pophysis projects postaxially beyond the postaxial surface of the centrum. The twenty-second vertebra supports a long rib, which articulates by its distal end with the third sternal rib. THE TWENTY-THIRD VERTEBRA. This vertebra is so like its preaxial predecessor that little need be said in its descrip- tion. 4 The postaxial surface of the centrum is relatively deeper dorso-ventrally, and its ven- tral margin is slightly concave. 416 MR. ST. GEORGE MIVART ON THE The parapophysial surface for the head of the rib is much as in the twenty-second vertebra (fig. 52, p). The diapophysis extends somewhat less postaxially, and is less antero-posteriorly extended at its distal end; and the surface for the tubercle is some- what less remote from the parapophysis. The neural spine is more extended dorsally, being longer in this direction than antero-posteriorly. The prezygapophysial articular surfaces are smaller than those-of the twenty-second vertebra; but the same cannot be said of the postzygapophysial ones. Great excavations appear on each side of the vertebra, in front of, ventrad to, and behind the diapophysis. Thus we have a side fossa, partially filled up with little lamell and spicula of bone, TWENTY-THIRD VERTEBRA (} natural size). Fig. 52. Lateral aspect. Letters as before. on the preaxial aspect of the root of the diapophysis. The fossa is bounded dorsally by the ridge running from the dorsal side of the diapophysis to the prezygapophysis, on the ventral side by the ridge running from the ventral side of the diapophysis to the parapophysis. Again, there is a great depression at the side of the neural lamina, just ventrad to the diapophysis; and the same may exist in the two preceding vertebre. On the dorsum of the neural arch we see that the lateral depressions on each side of the preaxial part of the neural spine have much extended. The postaxial median exca- vation of the neural spine is more open, while the margins of the lamine which laterally bound it form a very obtuse angle with the postaxial margin of that spine. Those postaxial excavations which in the two preceding vertebree were mentioned as existing on each side immediately ventrad to the postaxial margin of the diapophysis may be here more conspicuous; they may be so, in part, on account of the less develop- ment postaxially of the dorsum of the diapophysis, so that these lateral postaxial exca- vations may appear on the dorsum of the vertebra and as but slightly separated from the preaxial lateral depressions on each side. The postzygapophyses either do not project at all postaxially beyond the centrum, or but very slightly so (fig. 52, pz). AXIAL SKELETON OF THE OSTRICH. 417 The twenty-third vertebra supports a long rib (the sixth), which articulates at its ventral end with the fourth sternal rib, Tue TWENTY-FOURTH VERTEBRA. The twenty-fourth vertebra differs from the twenty-third principally in the much greater length (dorsally) of the neural spine. The centrum has its postaxial surface more concave dorso-ventrally, as also is its yentral margin, which at the same time is more everted postaxially. The ventral surface is slightly flatter. The neural canal, which has been diminishing in the last few vertebre, is here decidely smaller. The parapophysis is much as in the twenty-third vertebra. The diapophysis is somewhat less produced preaxially; and its distal end is more THE TWENTY-FOURTH VERTEBRA (2 natural size). Fig. 53. he Lateral aspect. Letters as before. rounded. The surface for the tubercle is again less remote from that for the head of the rib (fig. 53, d). The lateral excavation on each preaxial side of the root of the newral spine is still larger, and separated only by a vertical lamella of bone from the postaxial lateral excavation, which is here greatly augmented in size. Thus, when the neural arch is viewed dorsally, five radiating lamelle are seen to separate five subequal fosse. The two preaxial lateral excavations are separated by the preaxial part of the root of the neural spine. The two postaxial lateral excavations are each separated from the adjacent preaxial lateral excavation by a vertical lamella of bone running from the middle of the side of the neural spine outwards and postaxially to the postaxial part of the diapophysis. The two postaxial lateral excavations are separated from the remain- ing median postaxial excavation by the two lamelle which laterally bound the median VOL. vill.—part vil. March, 1874. : 3N 418 MR. ST. GEORGE MIVART ON THE postaxial excavation, and which diverge from the hinder part of the neural spine to the postzygapophyses. This fifth fossa (the median postaxial fossa) shows signs of subdivision at its fundus through a slight postaxial projection of the postaxial margin of the root of the neural spine. The apex of the neural spine is somewhat swollen, and may be much extended antero- posteriorly. The postzygapophysial surfaces are here, again, somewhat larger, and look much more ventrally and less outwards. : The twenty-fourth vertebra supports a long rib (the seventh), which articulates by its ventral end with the fifth sternal rib. Tue TWENTY-FIFTH VERTEBRA. This last of the free vertebra differs greatly from any yet described, and considerably even from the twenty-fourth vertebra, principally in the diminution of the centrum and great augmentation of the neural spine. The centrum has its preaxial surface less extended dorso-ventrally, while its ventral margin is concave as well as its dorsal one (fig. 55, ac). The postaxial surface of the centrum may be different from any thing we have hitherto seen; its transverse extent may be nearly three times its dorso-ventral dimen- sion. There may be scarcely a trace of dorso-ventral concavity, while the whole surface may be but very feebly convex from side to side’; on the other hand, there may be little difference between it and the postaxial surface of the twenty-fourth vertebra’. The neural canal is even more contracted, especially in its dorso-ventral extent at its postaxial end, than in the twenty-fourth vertebra. The ventral surface of the centrum presents a quadrangular, subequilateral, much flattened surface, such as we have not yet met with; its lateral margins, however, are still concave, while its pre- and postaxial margins are respectively rather more concave and convex than in the twenty-fourth vertebra. The parapophysis is much as in the twenty-fourth vertebra, but extends less outwards. The diapophysis is more slender, and projects directly outwards and somewhat dorsally ; its distal half, at the least, is rather pre- and postaxially compressed than dorso-ventrally depressed. The surface for the tubercle is again less remote from that of the head of the rib. The prezygapophyses are larger than in the twenty-fourth vertebra, and look more dorsad ; the postzygapophysial articular surfaces are still smaller than in the preceding vertebra, and look still more ventrad. The lateral excavations on the ventral and preaxial sides of the diapophysis have more ’ As in the specimen 13624 in the College-of-Surgeons Museum. * As in the specimen 1317 in the College-of-Surgeons Museum. AXIAL SKELETON OF THE OSTRICH. 419 or less coalesced by the abortion of the lamella running from the root of the diapophysis to the parapophysis. The neural spine is swollen at its distal end, and is in length about twice the antero- posterior extent of the whole vertebra. When the neural arch is viewed from above, there may be sometimes seen six fossee (fig. 57, f°, f°, f°), owing, when this is the case, to the complete subdivision of the median postaxial fossa into two by the greater postaxial development of the postaxial margin of the neural spine, which may extend to the postaxial margin of the neural arch. TWENTY-FIFTH VERTEBRA (} natural size). Fig. 54. Fig. 55. Aspects. Fig. 54, lateral; 55, preaxial; 56, postaxial; 57, dorsal (the neural spine being cut short off). Letters as before, except f’, antero-lateral fossa ; f?, lateral fossa; f°, posterior fossa of neural arch. The degree of special development of this vertebra varies, however, with the more or less preaxial extension of the ilia, also with the total number of vertebre of which the spinal column consists; for there may be one less than the number herein given. The twenty-fifth vertebra supports a long rib (the eighth), which articulates with the sixth sternal rib. 3Nn2 420 MR. ST. GEORGE MIVART ON THE THE LUMBO-SACRO-CAUDAL VERTEBRA. These vertebre are normally twenty in number, and are, in the adult, anchylosed together and also with the innominate bones into one great osseous mass (figs. 70-73). The various individual vertebre cannot, therefore, be described from the adult; but the Museum of the College of Surgeons fortunately possesses a preparation of the sacral vertebree (figs. 58, 59, 60, 61, & 62) of a young Ostrich in an unanchylosed condition, which enables the serial description of individual vertebre to be completed. THE DORSO-LUMBAR VERTEBRZ. Tue TWENTY-SIXTH VERTEBRA. This vertebra has, of course, its preaxial surface formed to correspond with the post- axial surface of the twenty-fifth vertebra. It may be greatly or not quite completely overlapped by the iliac bones. Its parapophysis is smaller than that of the preceding vertebra, with a smaller articular surface for the capitulum. Its diapophysis is shorter, more compressed in an oblique pre- and postaxial direction, so as to have its dorso-ventral diameter widest ventrad and preaxiad; its articular sur- face is also elongated in the same direction. Its spinous process is yet higher than in the twenty-fifth vertebra. This vertebra is mostly, but not always, confluent with the sacrum in the adult. It supports a distinct rib (the ninth), which, however, generally ends freely at its ventral extremity without joiing any sternal rib; indeed there is no true sternal rib corre- sponding to it, though it may be connected by a styliform bone with the sternal rib of the preceding vertebra. THE TWENTY-SEVENTH VERTEBRA. This vertebra is always, in the adult, confluent with the sacrum; and its rib (the tenth) is almost always an anchylosed process (but not always so, as é. g. in the specimen 13628). Its spine is again higher. Its diapophysis is still more compressed, appearing as a lamella, the greatest breadth of which is inclined still more preaxiad and ventrad than in the twenty-sixth vertebra. The parapophysis is much smaller, and is directed as much dorsad as ventrad, if not more so. The anchylosis of the transverse processes and rib results in the formation of a per- forated transverse process, with its greatly produced extremity (the shaft of the rib) extending ventrally, slightly arched convex preaxially, greatly flattened from within outwards, and with sharp, pre- and postaxial margins (figs. 70, 71, 72, & 73, p/). AXIAL SKELETON OF THE OSTRICH. 421 THE LUMBAR VERTEBRZ. Tue TWENTY-EIGHTH VERTEBRA. This is hardly to be distinguished in the adult, save by its transverse processes, which are strongly inclined postaxiad and dorsad, and are the first, since the axis, which present no indication of a rib. The under surface of its centrum is transversely concave (fig. 73). From the immature condition it may be seen that the spinous process is again higher, that the diapophysis inclines outwards and dorsad and is ventrally continuous with the parapophysis, that there are no postzygapophyses, but that its neural arch joins that of the twenty-ninth vertebra by suture (fig. 58, 28). ' LATERAL ASPECT OF TWENTY-SEVENTH AND TWENTY-EIGHTH VERTEBRA. Letters as before. Tue TWENTY-NINTH VERTEBRA. This vertebra (figs. 59, 60, & 61, 1) like the preceding, but with a slightly higher spine, more vertically extended transverse process, and no zygayophyses. Tue THIRTIETH VERTEBRA. This vertebra (figs. 59, 60, and 61, 2), as indicated by the young condition, has its spinous process yet higher; its diapophysis is also more dorsally produced, so that with the conjoined parapophysis there is presented an elongated vertical surface for the ilium. Without zygapophyses its neural arch is united to its centrum only by suture, and does not extend postaxially so far as the latter, thus leaving a large intervertebral opening between it and the ventral part of the neural arch of the next vertebra (fig. 59). Tue Tuirty-First VERTEBRA. This vertebra (figs. 59, 60, & 61, 3) has a spinous process slightly more elevated, and even in the immature condition (fig. 59, 3) has this process anchylosed with that of the next postaxial vertebra. 422 MR. ST. GEORGE MIVART ON THE Its neural arch is united by suture to its fellows and to the centrum, and is antero- posteriorly contracted below, so as to have a large intervertebral opening both pre- and postaxial of it. The most preaxial part slightly rests on the postaxial extremity of the thirtieth vertebra. LATERAL ASPECT OF LUMBAR AND SACRAL VERTEBRA OF AN IMMATURE SPECIMEN (¥ natural size). 1, Twenty-ninth (or second lumbar) vertebra; d', its diapophysis; s’, its neural spine. 2. Thirtieth vertebra. 3. Thirty-first vertebra; p”, its posterior parapophysial projeetion, which concurs with the anterior (and smaller) parapophysial projection (p*) of the next vertebra to form an articular surface for the lium. 4. Thirty-second vertebra ; p* and p*, anterior (preaxial) and posterior (postaxial) parapophysial projections ; the latter concurs with the anterior parapophysial projection (p*) of the succeeding vertebra to form an articular surface. 5. Thirty-third vertebra; p® and p”, its parapophysial projections. 6. Thirty-fourth i vertebra. 7. Thirty-fifth vertebra. 8. Thirty-sixth (or first sacral) vertebra; d°, its diapophysis; p®?*, its rib or capitulum, which has coalesced distally with the same parts of the two succeeding vertebree. 9. Thirty-seventh vertebra. 10. Thirty-eighth vertebra. 11. Thirty-ninth (or first sacro-caudal) ver- tebra; dp", its conjoined dia- and parapophysis. 12. Fortieth (or second sacro-caudal) vertebra. The diapophysis is inclined like that of the last vertebra, but has a flattened surface at its extremity ; it is quite separate from the parapophysis, which is formed by the ventral part of the neural arch, and offers a rounded, flat, articular surface (fig. 59) to the ium. This surface is very slightly reinforced by minute adjacent portions of the centra.of this and the preceding vertebra. The side of the centrum at its postaxial part exhibits a small neural foramen near its dorsal border and in front of a large parapophysial surface (fig. 59, p*), which this part of the centrum contributes towards the proper parapophysial surface of the next vertebra. Tue THirty-SEcOND VERTEBRA. This vertebra (figs. 59, 60, & 61, 4) has a spinous process at its maximum of deve- lopment and anchylosed to the adjacent spines. AXIAL SKELETON OF THE OSTRICH. 423 Its centrum is shorter pre- and postaxially, and its ventral surface is concave trans- versely and very strongly so at each lateral margin. Its neural arch rests half on the postaxial part of the preceding centrum, half on the preaxial part of its own centrum ; it has a large intervertebral opening both pre- and postaxial of its ventral portion. The diapophysis is elongated and slender, ascends dorsal and slightly preaxiad and outwards, with a small external flattened facet at its extremity (fig. 59, 4). The parapophysis is formed in a minute degree by the most ventral point of the neural arch, but mainly and subequally by the adjacent portions of the centra which support the arch (fig. 59, p” & p*). The neural foramen of the centrum is larger. TuE THIRTY-THIRD VERTEBRA. This vertebra (figs. 59, 60, & 61, 5) is similar to that last described; but the diapo- physis is shorter, the neural arch more antero-posteriorly extended at its ventral part. the intervertebral opening postaxial to it being much smaller. VENTRAL ASPECT OF LUMBAR AND SACRAL VERTEBRA OF AN IMMATURE SPECIMEN. (¥ natural size). Fig. 60. 1-12. Centra of vertebre from twenty-ninth to fortieth inclusive; d‘, diapophysis of twenty-ninth vertebra ; d*, diapophysis of thirty-sixth vertebra; dp‘, diapophysis of thirty-ninth vertebra conjoined with its parapophysis; p*, anterior parapophysial projection of thirty-second vertebra; p* posterior parapophysial projection of the same vertebra; p°, posterior parapophysial projection of thirty-third vertebra; p°, ante- rior parapophysial projection of thirty-fourth vertebra ; p*°", conjoined parapophyses of thirty-sixth, thirty- seventh, and thirty-eighth vertebre. The last of these three is anchylosed to its supporting centrum (10) ; but each of the two preceding capitula is united by suture, with a pair of slight parapophysial projections contributed by contiguous vertebree. There is a considerable defect of ossification in the neural arch (fig. 61,5) on each side of the spine, oval in shape and extending nearly from the spinous process to the diapophysis. The ventral surface of its centrum (fig. 60, 5) is slightly narrower antero-posteriorly, 424 MR. ST. GEORGE MIVART ON THE with each lateral margin extremely concave, through the two parapophysial processes which the centrum throws out. The preaxial one of these is the more considerable. The neural foramen of the centrum is again rather larger. THE THIRTY-FOURTH VERTEBRA. This vertebra (figs. 59, 60, & 61,6) has again its neural arch supported on two centra; but it rests rather more on its own. It has a larger dorsal defect of ossification (fig. 61, 6) than in the thirty-third vertebra, extending quite from the diapophysis to the neural spine. The neural spine is slightly less extended, though still anchylosed both pre-and post- axially, even in the young. Its diapophysis is rather longer (fig. 61, d°) than in the last vertebra, and inclined more preaxiad. Both the pre- and postaxial intervertebral openings are smaller. Its centrum, the ventral surface of which is exceedingly short antero-posteriorly, forms half of a large parapophysial surface (fig. 60, p*) at its preaxial part, and the greater part of a very much smaller parapophysial surface at its postaxial part. The neural foramen of the centrum is again large; it is here almost at its maximum (fig. 61, 6). THE THIRTY-FIFTH VERTEBRA. This vertebra (figs. 59, 60, & 61, 7) rests mainly on its own centrum. The postaxial intervertebral opening is bounded below by its neural arch. The defect of ossification of this neural arch is at its maximum (fig. 61,7), and more dorso-ventrally extended than in the thirty-fourth vertebra. The diapophysis is longer, almost as slender as in the last vertebra, and inclined more preaxially (fig. 61, 7). The parapophysial surface is much smaller, and is formed less by its own centrum than by that of the more preaxial vertebra. The postvertebral interspace is again rather smaller. The neural foramen of the centrum is smaller. THE SACRAL VERTEBR. Tue Tuirty-sixtH VERTEBRA. This vertebra (figs. 59, 60, & 61, 8) is distinguishable in the adult as that which fur- nishes (from the preaxial end of its centrum) the first of those three parapophysial roots which ascend and meet together to form the large bony plate which abuts against the ilium just behind the acetabulum. In the young it is easily distinguishable as the most preaxial one which develops from the preaxial end of its centrum a suture-united parapophysial process or capitulum of a rib (fig. 60, p’). AXIAL SKELETON OF THE OSTRICH. 425 Compared with the thirty-fifth vertebra, its neural arch is more antero-posteriorly extended, and shows a less defect of ossification dorsally (fig. 61,8). Ventrally the neural lamina extends postaxially over the preaxial half of the root of the capitulum of the next vertebra, and thence ascends so as completely to enclose the postvertebral per- foration, so that the neural arch may be said to be rather perforated than postaxially notched. The postvertebral interspace or perforation is considerably larger than in the three preceding instances. The diapophysis (figs. 59, 60, & 61, d*) is remote from the parapophysis, is stouter than that of the preceding vertebra, and extends more outwards and less preaxially. The parapophysis is a scarcely noticeable prominence from the side of the preaxial part of the ventral side of the centrum; and this concurs with a process from the post- axial part of the thirty-fifth' vertebra (fig. 60,7&8) to form a surface for the capi- tulum. This capitulum arises thence by an expanded base with a somewhat hexagonal outline, rapidly contracts, then rapidly expands, ascending dorsally and postaxially to anchylose with the similar expansion of the capitulum next behind, and contributing to form the dorsally and postaxially extended surface (fig. 59, p*°") which abuts against the postacetabular part of the ilium. THE THIRTY-SEVENTH VERTEBRA. This vertebra (figs. 59, 60, & 61, 9) has a lower neural spine, which, in the young, is free both pre- and postaxially. Its centrum, but not its neural arch, is more extended antero-posteriorly than in the last vertebra. DORSAL ASPECT OF LUMBAR AND SACRAL VERTEBRA OF AN IMMATURE SPECIMEN (¥ natural size). Fig. 61. 1-12. Neural arches of vertebre from the twenty-ninth to the fortieth inclusive; d'—d"", diapophyses of these vertebre to the thirty-eighth inclusive; dp", conjoined di- and parapophysis of the thirty-ninth vertebra ; p**", conjoined capitula of sacral vertebra; s', neural spine of twenty-ninth vertebra. In the neural arches of the vertebre from the thirty-third to the thirty-sixth, considerable defects of ossification are to be seen. VOL. VI.—PART vil. March, 1874. 30 426 MR, ST. GEORGE MIVART ON THE The defect of ossification in the neural arch is very small and behind the diapophysis (fig. 61, 9). The diapophysis (figs. 59 & 61, d°) is much larger, more expanded dorso-ventrally, and is connected with the root of the spinous process by a ridge ; it projects outwards and dorsad and very slightly postaxiad. The neural lamina is scarcely notched behind for the spinal nerves; and the post- vertebral interspace is less than in the two preceding vertebre. The capitulum is again united by suture, but is supported more by the preaxial parapophysial surface of its own centrum than by the postaxial central projection of the thirty-sixth vertebra (fig. 60). The capitulum, though it arises from a similarly expanded base, does not contract so much as in the preceding vertebra; it extends and expands to join its serial predecessor and successor ; but it projects slightly preaxially instead of projecting postaxially as does that of the thirty-sixth vertebra. Tue THIRTY-EIGHTH VERTEBRA. This vertebra (figs. 59, 60, & 61,10) has its centrum a little more antero-posteriorly extended (fig. 60, 10) and considerably more convex transversely than the centrum of the preceding vertebra. Its neural spine is quite free, less high, and more slender. Its diapophysis (figs. 59 & 61, d"°) closely resembles that of the thirty-seventh vertebra ; but its capitulum is anchylosed (fig. 60, p') at a time when the two preceding capitula are still united by suture to the centra. It is almost entirely supported by its own centrum, is still less contracted above its origin, and, projecting decidedly preaxiad as well as dorsad, comes into contact with its own diapophysis, as well as anchyloses with the two capitula preceding. There is no lateral defect of ossification in the neural arch; but it is slightly exca- vated in the middle line postaxially to the neural spine. There is no postvertebral interspace or perforation, except what is quite minute. THE SACRO-CAUDAL VERTEBR. Tue THirty-NINTH VERTEBRA. This vertebra (figs. 59, 60, & 61,11) is slightly more antero-posteriorly extended than is the preceding vertebra; its neural spine is lower and more slender; and its neural arch is medianly notched at its postaxial border, very decidedly (fig. 61, 11) and slightly so at the middle of its opposite margin. Diapophysis, parapophysis, and capitulum all combine in one anchylosed transverse process, which projects much, as does the diapophysis of the thirty-eighth vertebra, but is never dorso-ventrally extended. Its most ventral point of origin does not extend so far ventrad as does that of the capitulum of the preceding vertebra. AXIAL SKELETON OF THE OSTRICH. 427 Tue FortietH VERTEBRA. This has the transverse and spinous processes less developed, and the former have a rather more dorsal origin (figs. 59, 60, & 61, 12). The median notch of the neural arch is more marked (fig. 61, 12). The centrum is more excavated, causing each lateral margin of its ventral surface to be more concave. Tue Forry-First VERTEBRA. The differences just noted are here intensified (fig. 62, 41). The under surface of the centrum begins to exhibit antero-posteriorly directed exca- vations of its substance. LATERAL ASPECT OF THE VERTEBRA FROM THE FORTIETH TO THE FORTY-SIXTH INCLUSIVE IN AN IMMATURE CONDITION (natural size). Fig. 62. d, diapophyses ; ns, neural spines; p, parapophyses; ¢, transverse processes, including the undifferentiated parapophyses and diapophyses. Tue Forty-SECOND VERTEBRA. Again we find an intensification of the same characters; but, in addition, the root of each transverse process sends out a minute process on its ventral origin, 7.é. a para- pophysis (fig. 62, 42, p). Tuer Forty-THIRD AND Forty FOURTH VERTEBRA. With progressive diminution of the other process, the parapophysis is here rather more prominent (fig. 62, 43 « 44). The ventral surface of the centrum is more excavated. Tuer Forty-FIFTH AND ForRry-SIXTH VERTEBRE. These vertebre are like the two last noticed, but are smaller, and have the diapo- physis projecting preaxially and more decidedly separated from the more marked para- pophyses, which latter project outwards, one from each lateral margin of each centrum, while the ventral surface of each centrum is deeply excavated antero-posteriorly on each side of a median antero-posterior lamella (fig. 62, 45 & 48). 302 MR. ST. GEORGE MIVART ON THE pw to THE CAUDAL VERTEBRE. Tue Forry-SEVENTH VERTEBRA. This vertebra remains normally unanchylosed with the preceding bone in the adult bird. It has a massive centrum and a neural arch small in circuit and very little developed antero-posteriorly, with a wide and massive neural spine and a moderate transverse process. The centrum has its postaxial surface nearly flat or slightly concave or convex. Its dorsal margin is slightly concave, its ventral margin very slightly convex. The preavial surface of its centrum is slightly concave, and its ventral margin nearly straight. THE FORTY-SEVENTH, ox FIRST CAUDAL VERTEBRA (natural size). Fig. 63. Fig. 64. Fig, 63, preaxial aspect; 64, lateral aspect (left side), Letters as before. The ventral surface of the centrum shows the subparallel antero-posterior excavations separated by a considerable interval. The excavations, however, are less extensive than are those in more preaxial vertebre. The diapophysis (figs, 63 & 64, d) projects dorsad, outwards and very slightly post- axiad; it is subconical, rather flattened pre- and postaxially, and with a somewhat rounded extremity. A very small parapophysial process projects from or ventrad of the ventral side of its root, extending slightly outwards from the lateral margin of the ventral surface of the centrum at or postaxiad of the antero-posterior middle point of that margin (figs. 63 & 64, p). The neural lamina of each side has its postaxial margin nearly vertical; but its pre- axial margin is inclined preaxiad and dorsad. The neural spine is much more developed transversely than antero-posteriorly, and shows a tendency to bifurcate, its lateral extremities extending more or less outwards and preaxially. There are no zygapophyses. AXIAL SKELETON OF THE OSTRICH. 429 Tue Forry-EIGHTH VERTEBRA. This vertebra is like the last, except that there are no subcentral excavations. The diapophyses extend less dorsad ; and the parapophysial projection is smaller on each side. Tue Forty-NINTH VERTEBRA. This vertebra differs from the preceding chiefly in the sudden reappearance of a parapophysis of considerable size. It is, however, continuous with the diapophyses, the two appearing as a transverse process, which projects outwards and slightly post- axiad from the postaxial part of the side of the centrum, and expands dorso-ventrally towards its distal end (fig. 65, d, p). THE FORTY-NINTH, or THIRD CAUDAL VERTEBRA (natural size). Fig. 65. Fig. 65, preaxial aspect. Letters as before, except that ¢ denotes the “ transverse process,” which probably contains both a diapophysial (@), and a parapophysial (p) element. Its free cuter margin is dorso-ventrally concave, the prominence at the dorsal side of the concavity representing the rounded end of the diapophysis, while the sometimes less marked prominence on the ventral side of the concavity represents the rounded end of the parapophysis. ‘The preaxial surface of the transverse process is excavated and dorso-ventrally concave. Tue Firtierh VERTEBRA. The fiftieth vertebra is slightly smaller than the forty-ninth, which it greatly resembles, except that the parapophysial (ventral) extremity of the transverse process is decidedly more produced than the diapophysial one. The neural canal is also smaller, and the lateral summits of the laterally bifurcating neural spine more divergent. Tue Firty-Frirst VERTEBRA. Here we find an intensification of the differences noticed in describing the fiftieth vertebra ; and the outer margin of the centrum also more nearly approaches a circle. The preaxial surface of the centrum is decidedly concave. 430 MR. ST. GEORGE MIVART ON THE Tue Firty-SecoND VERTEBRA. This vertebra closely resembles its serial predecessor, but is smaller in size (fig. 66). THE FIFTY-SECOND, on SIXTH CAUDAL VERTEBRA (natural size). Fig. 66. Fig. 66, preaxial aspect. Letters as before. THe Firty-THIRD VERTEBRA. Here the transverse processes project more postaxiad and ventrad, and the diapo- physial part has greatly decreased relatively to the parapophysis. The spinous process shows a more or less marked tendency to become trifid transversely at its apex. THE Firry-rourTH VERTEBRA. This vertebra (fig. 67) is much smaller than its predecessor. The transverse pro- cesses are very much smaller, but the apex of the neural spine is very distinctly trifid transversely. The preaxial surface of the centrum is decidedly concave, as in all the vertebrz since the fiftieth. THE FIFTY-FOURTH, or THE FIFTY-FIFTH snp FIFTY-SIXTH, or EIGHTH CAUDAL VERTEBRA NINTH AND TENTH CAUDAL VERTEBRA (natural size). (natural size). Fig. 67. Fig. 68. Fig. 67, preaxial aspect (letters as before); 68, preaxial aspect of ninth caudal vertebra; ns*, neural spine of tenth caudal vertebra; 69, lateral aspect; d, osseous bridge connecting the portions. THE Firty-FIFTH VERTEBRA. This vertebra is devoid of transverse processes; or there are at most but minute faint AXIAL SKELETON OF THE OSTRICH. 431 traces of diapophysial and parapophysial prominences. The neural spine is more or less trifid at its distal end (fig. 68, ms). The postaxial surface of the centrum is decidedly concave. The vertebra often becomes anchylosed with the next and last, both by its centrum and the distal portion of its spine. Tue Firry-sixtH VERTEBRA. This vertebra has no neural arch and no transverse process (fig. 69). It is a dorso-ventrally extended lamina of bone, extremely compressed laterally, with a very irregular, generally more or less rounded, margin dorsally, ventrally, and post- axially. It often anchyloses with the preceding vertebra at the dorsal and ventral parts of its preaxial surface, or rather margin, thus producing a foramen which looks from side to side. The existence of another (but minute) foramen placed postaxially to that just described may indicate that this bone really consists of two or more vertebre fused and anchylosed into one osseous mass. Sometimes a little bridge of bone (fig. 69, d) connects its centrum, laterally, with the centrum of the fifty-fifth, or ninth caudal, vertebra. THE PELVIS. This enormous bone consists of no less than twenty-two vertebre, more or less com- pletely anchylosed together (in the adult), and with the two ossa innominata, which latter thus cannot conveniently be excluded from the description of the axial skeleton. Preaxially the sacrum exhibits the preaxial surface of the first sacral vertebra (with characters corresponding with the vertebra naturally preaxiad to it) roofed over by the PREAXIAL ASPECT OF PELVIS (J natural size). Fig. 70. at, antitrochanteric process; 7 and 7’, ischium; iJ, ilium; p, pubis; pl, rib; s, spinous process of postanterior vertebra ; sy, pubic symphysis. two iliac bones (/), which meet together over the dorsal end of its spine and diverge ventrad at an angle of about 63°—the two.diverging lines being carried on by the two anchylosed ribs (pl) of the first sacral vertebra. 432 MR. ST. GEORGE MIVART ON THE Within the line of these ribs is to be seen a very large ventral arch, each side or lateral half of which is bent at an angle into a dorsal and ventral division or limb, which, when viewed preaxially, appears as follows :— The dorsal division, or limb, consists mainly of the pubic bone (p), which diverges from its fellow of the opposite side at an angle of about 66°. The external margin of this limb is, in the main, concave externally; but the outline is interrupted by the more distant jutting out of the posterior part (7) of the ischium. The internal margin of the dorsal limb is convex, but with its outline interrupted, and the convexity exaggerated by the projecting inwards (7’) of the more anterior part of the ischium. The ventral division or limb is bent inwards on the dorsal limb (of the same half of the great ventral arch) at an angle of about 115°, and consists of the pubis, which ter- minates ventrally by meeting its fellow in (sy) a ventral symphysis. The external margin of this ventral division is concave; its internal margin is convex. When the pelvis is viewed postaxially the same great ventral arch is seen to be con- nected dorsally with a pentagonal mass, one angle of which is dorsad, and which has in its midst the postaxial surface of the small twenty-second sacral vertebra, 7. ¢. the forty- seventh vertebra of the whole spinal column. The two dorsal sides of the pentagon meet at an angle of about 118°; and the margin of each, from the point of junction outwards, is slightly concave, then more strongly convex, then still more sharply concave, the sharp concavity being produced by the pro- minence of the antitrochanteric process. Each of these dorsal sides is formed by an ilium. From the tip of this last-mentioned process each lateral margin of the pentagon pro- ceeds ventrad, forming with the adjacent dorsal side an angle of about 90°; its margin is, for its greater part, gently concave, and is formed by the ischium; it forms with the ventral side of the pentagon an angle of about 120°. This ventral side of the pentagon is formed by the pubes, and is more or less horizontal. When the pelvis is viewed laterally (fig. 71), the sacrum being horizontal, we have a dorsal elongated mass (made up of the ilium and sacrum) something like the skull of a bird, with the tip of the beak turned postaxially, from which two long bars of bone ( & p) diverge ventrally and proceed postaxiad to join and end in a great recurved process (s 7). These bars proceed from beneath the acetabulum. The acetabulum is placed on the ventral side of the ilio-sacral mass, so that its postaxial margin is on the preaxial side of the middle point of that elongated mass. Within the acetabulum are to be seen the four slender, dorsally and preaxially extended diapophyses of the thirty-second, thirty-third, thirty-fourth, and thirty-fifth vertebre. The dorsal margin of the preacetabular part of the ilium forms the roof of the cranial AXTAL SKELETON OF THE OSTRICH. 433 part of what may be fancifully compared to a bird’s skull, while the same margin of the postacetabular part of the ilium forms the apparent upper margin of the beak of such a skull, the whole (straight) inferior margin being formed by the ventral surface of the sacrum. The ventrally proceeding osseous bars (the pubis and ischium) join in an expansion (7) situated ventrally to and opposite the postaxial end of the sacrum. Beyond this point the pubis curves backwards, and then sharply first downwards (sy) and then forwards, its preaxial extremity advancing a little preaxiad of the extreme point to which the whole sacrum extends postaxiad. The obturator foramen is very elongated, and is divided into a very small anterior and a very larger posterior portion by a process (ps) of the ischium which joins the pubis a little behind the acetabulum. The pubis, at its junction with the ilium, developes a strongly projecting, curved and pointed ilio-pectineal process (/p). Sometimes a small process projects from the middle of the ventral margin of the pubis. LATERAL ASPECT OF PELVIS (+ natural size). Fig. 71. Letters as last, except Zp, ilio-pectineal process, and ps, process of ischium dividing off the small anterior part of the elongated obturator foramen. The extreme pre- and postaxial extent of the pelvis is almost double its greatest dorso ventral dimension. When the pelvis is viewed dorsally (fig. 72) we see in the middle of its preaxial third the dorsal iliac median ridge, and on each side the concave external margins of the ilia, with the two sacral ribs (p/) protruding, one on each side, from the more preaxial part of those margins. At about the middle third of the total pre- and postaxial diameter of the pelvis the antitrochanteric processes (at) stand sharply outwards. Behind a transverse line joining these processes we have five elongated pieces of bone, the middle VOL. VIII.—PART VII. March, 1874. 3P 434 MR. ST. GEORGE MIVART ON THE one of which proceeds directly postaxiad, while on each side of it two pieces proceed postaxiad and obliquely outwards till they reach a transverse line nearly coinciding with the postaxial end of the median piece, after which they bend inwards, at an angle of about 90°, to meet in the middle line. The median piece consists of the postacetabular parts of the two ilia with the post- acetabular sacral vertebra; its lateral margins are gently concave; and it terminates postaxially in two diverging horns, which consist of the two postaxial ends of the two ilia projecting postaxiad beyond the last sacral vertebra. Between the antitrochanteric processes each ilium presents a rather wide convex surface; but this very rapidly narrows postaxially till at about the last sacral vertebra but two, whence it again slightly expands laterally to about the postaxial margin of the last sacral vertebra, after which it gently contracts to a bluntly pointed termination directed postaxiad, ventrad, and slightly outward. The two postacetabular dorsal expansions of the ilia are separated from the conjoined expansions of the spines of the sacral vertebre (cs) by two very elongated narrow fissures, which run pre- and postaxiad nearly parallel to each other. The primitive distinctness of those conjoined sacral spines is indicated by successive foramina or defects of ossification. The two lateral diverging pieces (i & p) on each side of the median postacetabular ilio-sacral piece are, of course, the ischium and pubis on each side. The two ischia (7) seem to diverge at an angle of about 20°, the two pubes (p) at an angle of about nearly 30°; but the latter curve inwards and join the ischia at the lateral expansion before mentioned. Preaxiad of this junction the outer margin of the pubis is convex, and its inner margin concave, while the outer margin of the ischium is concave and its inner margin convex. DORSAL ASPECT OF PELVIS (+ natural size). Fig. 72. Letters as last, except cs, median crest formed by the coalesced spinous processes of the sacral vertebre. Postaxiad of the lateral expansions the two pubes converge and meet in the sym- physis (sy), and then project sharply preaxiad. Each pubis, therefore, postaxially to the lateral expansion, has a postaxial margin which is convex, and a preaxial margin which is concave. AXIAL SKELETON OF THE OSTRICH. 435 The greatest transverse diameter of the entire pelvis is across the pubes, at about the middle of their antero-posterior extent. The greatest transverse diameter of the ilio-sacral mass is between the antitro- chanteric processes. The pelvis when viewed ventrally (fig. 73) presents, of course, relations corresponding with those already noticed as shown by the dorsal view. Here, however, the two ilio- pectineal eminences (/p) are very conspicuous, diverging, as they do, preaxiad from the preaxial end of each pubis. The rough surface for the attachment of a small bone, or else a bony process, is also noticeable at about the middle (pre- and postaxially) of the ventral surface of each pubis. The ventral surface of the sacrum is visible throughout, forming one elongated bone in the middle line (in the adult), except that sometimes the last (twenty-second) sacral vertebra shows the line of junction of its centrum with that of the twenty-first sacral vertebra. This median bone, then (formed of the coalesced centra), is widest in front of the acetabula, where it is concave transversely, and finally (between the two postaxial halves of the postacetabular parts of the ilia) it is flat, with successive pairs of antero- posteriorly directed excavations with a bony median ridge between each such pair of excavations. The sacrum tapers backwards very gradually indeed. The lateral processes of the adult sacrum are shown by immature specimens to have yarious natures. Thus we have on each side at first (at the twenty-sixth vertebra) a truly articulated rib, Next an anchylosed rib, Next a transverse process arising entirely from the neural arch, and much inclined © postaxiad (fig. 58, 28, d), Next one similar, but smaller (fig. 59, d*), Then a shorter, formed by a preaxial process of the centrum and by the parapophysis of the thirtieth vertebra, VENTRAL ASPECT OF PELVIS (+ natural size). Fig. 73. Letters as before. 436 MR. ST. GEORGE MIVART ON THE Next a still smaller process formed by the adjacent parapophysial processes of the centra (thirtieth and thirty-first), together with the neural parapophysis of the thirty- first vertebra (fig. 59 & 60). After this follows a larger process on a lower (more yentral) level, abutting directly against the proximal end of the ischium, and formed almost exclusively by the adjacent parapophysial (central) processes of the thirty-first and thirty-second vertebre (fig. 59, p3 & p4). Next follows a smaller and even shorter process, formed by the adjacent parapophysial (central) processes of the thirty-second and thirty-third vertebre (fig. 59, p* & p’). Then we have a closely approximated smaller process, similarly formed by the thirty- third and thirty-fourth vertebra, and, after, a scarcely noticeable prominence formed in the same way by the thirty-fourth and thirty-fifth vertebre (fig. 60, 6, 7). Next follow three conspicuous capitula, uniting distally to form an expanded plate, and belonging respectively to the thirty-sixth, thirty-seventh, and thirty-eighth vertebre (figs. 59, 60, & 61, p*?”), Eight transverse processes then serially succeed, belonging to the eight next vertebre ; and the prominent lateral margins of the centra of the five most posterior sacral vertebree may be regarded as rudimentary parapophyses. Tue Ittum (figs. 70-74). This bone extends preaxiad to roof the twenty-sixth or even the twenty-fifth vertebra ; its dorsal margin is gently convex; its preaxial margin is irregular, more or less produced at its ventral part. The ventral margin of the preacetabular part of the ilium is concave, being bent more or less strongly ventrad preaxially to the acetabulum. At its ventral end it here gives rise to a sharp-pointed ilio-pectineal process (Jp) or spine, which projects strongly preaxially outwards and somewhat ventrad. The ilium forms the preaxial and dorsal border of the acetabulum and the dorsal half of the much-projecting antitrochanteric process (fig. 74). The postacetabular part of the ilium is longer than the preacetabular part; but the excess is less in the young than in the adult. Unlike the preacetabular part, it is separated, even in the adult, from the spines (fig. 72, cs) of the caudo-sacral vertebre ; and, in the young, even from those of the true sacral vertebre. It tapers gradually backwards, the two borders (dorsal and ventral) being nearly straight and approaching each other postaxiad at an angle of about 13°. The free postaxial extremity curves somewhat outwards in the adult (fig. 72). The dorsum of the ilium is marked by a gluteal line or ridge, which curves outwards and postaxially (fig. 72) to above the antitrochanteric process; thence it continues on postaxiad, at first curving inwards, but ultimately (at about the middle of the post- acetabular part of the ilium) curving outwards to the posterior extremity of the bone. ~I AXIAL SKELETON OF THE OSTRICH. 438 Tue Pousis. The pubis is an exceedingly long and narrow bone, extending postaxiad and ventrad from the acetabulum to considerably beyond. the postaxial extremity of the ilium, yet not so much s0 relatively in the young as in the adult. The pubis first curves slightly downwards, as well as strongly postaxiad, then some- what dorsally and outwards, then inwards and ventrad, and curves sharply round directly ventrad, and afterwards and finally preaxiad, at the same time meeting its fellow of the opposite side in a postaxial and ventral pubic sympyhsis. In the adult the pubis anchyloses with the ischium (fig. 71, 7). The pubis forms a small ventral portion of the margin of the acetabulum (fig. 74, p). LATERAL ASPECT OF THE ACETABULAR REGION OF AN IMMATURE PELVIS GZ natural size). Fig. 74. a, acetabulum ; 2, ilium ; 7s, ischium ; p, pubis. It develops a ridge on its dorsal aspect a little behind the acetabulum, which ridge makes a projection dorsad in the ventral margin of the obturator foramen. The pubis does not seem to take any share in forming the ilio-pectineal spine. At its ventral convexity the ventral margin of the pubis may, at about its middle, develop a bony process. This process is generally, however, only represented by a slight thickening of the margin. To this point a small distinct bone’ is sometimes attached, which, when it becomes anchylosed, constitutes the process referred to as occasionally present. Tue IscHivM. The ischium is the shortest of the three pelvic bones. Narrow for the greater part of its length, it expands at each end. Preaxially it anchyloses with the ilium dorsally, and with the pubis ventrally. Postaxially it anchyloses with the pubis (fig. 71, 7), but not at all with the ilium. ’ It forms about a third (the postaxial ventral third) of the margin of the acetabulum, and the ventral half of the antitrochanteric process (fig. 74, 7s). From the outer postaxial margin of this process a sharp ridge continues ventrad upon 1 This bone has been described and figured by Mr. A. H. Garrod (in the Proc. Zool. Soc. for March 1872, p. 359), who has suggested the possibility of its representing the marsupial bone. 438 MR. ST. GEORGE MIVART ON THE a ventrally extending prominence (figs. 71 & 73, ps), which approaches the pubis very closely, and nearly divides a small oval preaxial part of the obturator foramen from a very elongated and large postaxial portion. The shaft of the ischium is divided by three longitudinal ridges:—one well marked and dorsal in position; another ventral, and only marked at the more postaxial part of the shaft ; a third external and very marked, extending from the process approaching the pubis to the postaxial extremity of the bone. At its postaxial end the ischium in the adult expands into a subquadrate dorso- ventrally expanded plate of bone (fig. 71,2) more or less convex externally for the greater part of its extent. The ischium extends postaxially beyond the ilium, relatively more in the young condition than in the adult bird, Tue Riss. The ribs consist of two series, the vertebral and the sternal ribs, the former series being the more developed both in number and size. Tue VERTEBRAL Rips. Of these there may be said to be normally ten on each side, though the first and last of these very generally appear in the form of anchylosed transverse processes of their respective vertebra. Five of the vertebral ribs articulate with sternal ribs, which themselves join the sternum directly by distinct articular surfaces (fig. 1, 1v, Vv, v1, vil, & VII). Behind these there are normally two ribs, which end freely at their ventral extremities ; but there may be three such ribs. In the latter case the first of these may be continued on ventrally by a distinct, curved, rod-like bone, which bone does not reach the sternum, but is applied to the ventral side of the sternal rib next in advance. The First Rib. This little rib (fig. 75, 1) is attached to the eighteenth vertebra, and very often anchyloses with it. It is a flat bone, subtriangular in shape, with its three margins concave, but especially the preaxial margin. Its length only exceeds its breadth by one quarter, if so much, of the latter, and is less than half that of the centrum of the vertebra to which it is attached. It projects freely postaxially at its distal end; and preaxially it presents a head and neck, or capitulum (h), and a tubercle or tuberculum (¢), subsequently developed to articulate respectively with the para- and diapophysis. Both capitulum and tuberculum are broad and rather flattened pre- and postaxially, being less slender and less rounded than the distal end of the rib. The capitulum exhibits scarcely any distinct rounded articular surface; but the tuberculum shows a flattened, slightly excavated facet at its extremity. AXIAL SKELETON OF THE OSTRICH. 439 There is a small pneumatic foramen in the vertebral margin near the head of the rib. The dorsal (or more postaxial) margin of the rib is rather shorter than the ventral one, owing to the head and neck being slightly longer than the tubercle. THE VERTEBRAL RIBS OF RIGHT SIDE (size of figures I-X, } natural size; figures IV’, V’, VII’, 4 natural size). Fig. 75. 1. first rib: its preaxial margin is between / and ¢; its vertebral margin extends downwards from ¢ to the apex of its ventral margin. II. second rib. III. third rib, IV, fourth rib; IV’. proximal end of fourth rib, showing the pneumatic foramem (f), the cup-like surface for the diapophysis (¢), and the convex tubercle (h) for the parapophysis. V. fifth rib; V’. its proximal end (the small but sharp prominence near the letter f is produced by the ridge which passes outside the root of the capitulum). VI. sixth rib. VII. seventh rib; VII’. its proximal end, showing the great articular surface for the diapophysis and the very large pneumatic foramen. VIII. eighth rib, showing the small opening or defect of ossification on the tuberculum, near its junction with the capitulum. IX. ninth rib, showing a defect of ossification larger than that in the eighth rib. X. tenth rib, with a defect of ossification relatively yet larger. In all the figures, / denotes the capitulum, and ¢ the tuberculum ; w, the uncinate process. The Second Rib. : This still absolutely small rib (fig. 75, 11) is relatively much elongated compared with the first rib, its length being about three times its greatest breadth, and three times the 440 MR. ST. GEORGE MIVART ON THE length of the first rib, or even more. Nevertheless its length is to the extreme length of the nineteenth vertebra (to which it is attached) only as about seven to six. It projects freely ventrad (fig. 1, 1), and is in the form of a Y with very short arms and with a curved stem. ‘The stem is, on the whole, rather concave outwards, but with a slightly sigmoid flexure. The head and neck of the rib (h) may be more slender than any part except the distal end of the rib; more slender absolutely, and very much more so relatively than in the first rib. The head itself has a subcircular circumference. The tuberculum (¢) is slightly shorter than the neck, but considerably broader, much flattened pre- and postaxially, but with a rounded articular concavity at its tip. It is very little larger absolutely, and very much less so relatively, than in the first rib. The vertebral margin (between the head and tubercle) is strongly concave, very much more so than in the first rib. It presents an exceedingly large pneumatic foramen (f). The ventral or preaxial margin of the rib is more rounded than the dorsal or post- axial one. Proceeding distad from the capitulum, this margin is at first gently concave and then gently convex. The postaxial margin may develop a low, long prominence, with a rounded outline, a little distad of the tuberculum. On this account this margin, proceeding distad from the tuberculum, is at first concave, then convex, and then again gently concave. The Third Rib. The third rib (fig. 75, 111) is a little more than twice the length of the second rib; but neither the capitulum nor the tuberculum is quite twice the length of the capitulum and tuberculum of the latter. This third rib bears a proportion in length to the twentieth vertebra (to which it is attached) about as 33 to 7. It projects freely ventrad and slightly postaxiad; and the tail of its ¥ has a similar, though less marked, curvature to that of the second rib (fig. 1, mm). The capitulum is relatively rather longer and more slender in proportion to the tuberculum than in the second rib; it is the slenderest part, except close to the actual extremity of the free distal end. The head itself presents a rounded, convex, articular surface for the parapophysis; that on the tubercle (for the diapophysis) is nearly twice the size of the corresponding surface on the second rib. Its dorso-ventral diameter slightly but decidedly exceeds its pre- and postaxial extent. The vertebral margin (between the capitulum and tuberculum) is much more sharply concave than in the second rib, forming an acute angle of about 60°; but its pneumatic foramen is little, if any, larger. The tuberculum seems to continue on in the main and general direction of the rib, and the capitulum to diverge obliquely inwards from this. AXIAL SKELETON OF THE OSTRICH. 44] The preaxial or ventral margin of the rib is generally rounded, the dorsal margin sharp. There is no sudden inflection in the rib which can be called an “angle ;” nor does its dorsal margin develop a rounded prominence as was the case in the second rib. The entire preaxial margin is at first (¢.e. proximally) concave, and afterwards very gently convex. The entire dorsal margin is at first very slightly convex, and then concave in a little more marked degree. The Fourth Rib. Though the increase in length of the fourth rib (fig. 75, 1v & Iv’) is much less than we meet with in proceeding from the second to the third, yet it exceeds the last- mentioned very considerably in bulk and massiveness, though it is less, if not much less, than half as long again (fig. 1, 1v). The fourth does not end freely, but articulates distally by an expanded concave surface, with the second sternal rib ; it is attached dorsally to the twenty-first vertebra. The increase in length of the capitulum of the rib (as compared with that of the third rib) is considerably greater than that of the tuberculum. Concomitantly with this, it seems rather to be the capitulum which continues on the general curve of the rib, from which the tuberculum diverges obliquely dorsad. The stem or tail of the Y, 7.¢. the shaft or body of the rib, presents no sigmoid flexure, but bends rather suddenly a little distad of the junction of the capitulum and tuberculum. Through this bending, the rib may be said to have a sort of “ angle.” The capitulum is actually longer than that of the third rib, and relatively longer as compared with the tuberculum, though not as compared with the whole rib. It is actually the slenderest part. The head itself (fig. 75, tv’, h) presents a convex articular surface, which is pre- ponderatingly developed dorso-ventrally, being nearly twice as long in this direction as transversely. The vertebral margin (between the capitulum and tuberculum) is less sharply concave, its margins forming an angle of about 70°. The pneumatic foramen (f) is actually larger, but relatively smaller. The tuberculum and its articular surface for the diapophysis (¢) are much more extended antero-posteriorly, the latter being thus twice as long as broad. The preaxial margin of the body of the rib is sharper than in the third rib; in fact a rounded ridge near the preaxial margin (on the visceral surface of the rib) seems to answer to the actual ventral margin of the third rib, as, like the latter, it continues on the prominence of the capitulum. If so, then the actual preaxial margin of the fourth rib, at its proximal part, would seem to be formed from a lamella of bone being developed on the ventral side of what was, in the third rib, the actual preaxial margin. The whole preaxial margin is nearly on one level vertically. VOL. ViIL—PaRT vil. March, 1874. 3Q 442 MR. ST. GEORGE MIVART ON THE The postaxial margin is sharp, and becomes more prominent at the angle, causing this part of the margin to be convex. At abcut the dorsal end of the distal third of the postaxial margin of the rib a large lamelliform uncinate process is thrown out. It stands out at its origin nearly at right angles to the course of the rib, and then turns sharply dorsad and tapers to a point, running nearly parallel to the course of the rib; the uncinate process is thus bent at right angles. The Fifth Rib. This rib (fig. 75, v & v’) which is attached dorsally to the twenty-second vertebra, closely resembles the fourth: it is, however, larger; and its distal part is more slender, expanding, nevertheless, at the distal end to form a concave surface for junction with the third sternal rib (fig. 1, v). Both the capitulum and tuberculum are longer and diverge at an angle of about 75°, having a large pneumatic foramen (fig. 75, v’, f) at the vertebral margin, where they bifurcate. The length of the capitulum, as compared with that of the tuberculum, has also somewhat increased. The shaft of the rib is more bent on the head and tubercle, and has a rather more decided “angle ;” but its margins are in the main similar to those of the third rib. The uncinate process (w) is placed more proximally, springing from nearer the middle of the length of the shaft of the rib; it extends out rather less, and its recurved portion is more slender. The preaxial margin of the shaft is continued into that of the tuberculum by a ridge (fig. 75, v’, near /’) passing outside the root of the capitulum, 7.¢. preaxially to that root. The Siath Rib. This rib (fig. 75, v1) is attached to the twenty-third vertebra, and very closely resembles the fifth, which it about equals in length; its distal half, however, is more slender and more rounded, though its distal end may, on the contrary be more expanded. It articulates distally with the fourth sternal rib (fig. 1, v1). The pneumatic foramen is larger; and the capitulum and tuberculum diverge at an angle of about 80°. The proximal part of the shaft of the rib is rather more bent. The uncinate process (w) springs from the dorsal margin, at about the middle point between the distal end of the rib and the root of the tuberculum. It is very much smaller, and, though rather broader at its root, much blunter, and may be not half the length of the recurved part of that of the fifth rib. The articular surface for the diapophysis is rather larger and more concave. The ridge passing outside the root of the capitulum is more marked than in the fifth rib. AXIAL SKELETON OF THE OSTRICH. 445 The Seventh Rib. The seventh rib (fig. 75, vii & vu’) is attached to the twenty-fourth vertebra. In spite of its being much more bent, it is decidedly shorter (by nearly the length of its tuberculum) than the sixth rib, though it is longer than the fourth rib when measured along its convexity (fig. 1, vit). The capitulum and tuberculum diverge at an angle of about 80°. The pneumatic foramen (vu1', f) is still larger, forming quite a chasm. The articular surface for the diapophysis is also rather larger (vi1', ¢). The shaft of the rib is more rounded and slender generally, especially towards its distal part, which, however, expands towards its extremity so largely as to about equal in size the distal end of the sixth rib. It articulates at its distal end with the fifth sternal rib. The shaft is also bent, with a decided or slight sigmoid flexure, convex postaxially at its more proximal part, and convex preaxially at its more distal part; but it may only show a preaxial concavity and a postaxial convexity. The ridge traversing the outside of the root of the capitulum is still more marked. There appears to be generally no rudiment even of an uncinate process, but it may be developed, though always to a less extent than in the seventh rib. The Eighth Rib. This rib (attached to the twenty-fifth vertebra) is smaller in all dimensions (fig. 75, v1i1) than the seventh rib, and in length is intermediate between the third and the fourth ribs. It closely resembles in shape the seventh rib; but the capitulum is rather less long compared with the tuberculum; the pneumatic foramen is smaller; but there are small openings on the postaxial or inner surface of the tuberculum. The articular surface for the diapophysis is smaller and nearly flat. The shaft of the bone is more or less bent ; it may exhibit a rudiment of an uncinate process. It articulates at its ventral end with the sixth sternal rib (fig. 1, viii). The Ninth Rib. The ninth rib (attached to the twenty-sixth vertebra) is much shorter, about the same length as the third, and, like the latter, ends freely at its distal end (fig. 75, 1x). It is much more slender than any other rib. Compared with the eighth rib, the capitulum is again shorter relatively, exceeding the tuberculum less in proportion. The pneumatic foramen on the margin of the neck is minute; but there is a large aperture on the post- axial aspect of the tuberculum, where the capitulum joins it. The shaft of the bone is again straighter (fig. 1, 1x). The ninth rib may have at its ventral end an extremely long, slender, curved bone applied by its other extremity to the outer or ventral side of the sixth eternal rib. 3Q2 444 MR. ST. GEORGE MIVART ON THE The Tenth Rib. This rib (normally anchylosed in the adult with the sacrum and pelvis) generally appears as an elongated ventrally produced process (figs. 70, 71, 72, & 73, pl) of the twenty-seventh vertebra, as therein noticed. The rib is compressed from within outwards (fig. 1, x). The tenth is only about from 2 to 2 as long as the ninth rib, and may be somewhat shorter than the second rib when this latter is measured from the root of the tuberculum distad. Its inner (here actually postaxial) surface between the tuberculum and capitulum is but very slightly concave ; its preaxial margin is convex, and its pustaxial one concave. On the postaxial side of the expanded proximal part of the rib is a large pneumatic opening. The ridge, which in more preaxial ribs ascended the outer surface of the shaft to the interval of the head and tubercle and went on ultimately to the tubercle, is here pro- duced into a sharp crest, which is the actual outer and preaxial margin of the shaft of the rib, and which lies close to that margin all the way to the dorsal extremity of the tuberculum. An Eleventh Rib may be developed, which then closely corresponds with the conditions here attributed to the tenth rib. When this development occurs, the tenth rib approximates in its form very nearly to what has here been described as the ninth rib. THE STERNAL Riss. These six ribs increase gradually in length postaxially from the first to the sixth; so that the fourth is rather more than twice the length of the second, and the sixth is from two and a half to three times the length of the second; the third is more than half the length of the sixth (fig. 1, 1, 11, Iv, V, V1). The sternal ribs are greatly expanded dorso-ventrally at their distal (sternal) ends, but do not bifurcate and divide into two processes like the tuberculum and the capitulum of the vertebral ribs. All these ribs, except the first, unite with one or other of the vertebral ribs by a distinct joint. The angle formed by the second sternal rib with its vertebral rib is only slightly obtuse, viz. about 129°; the angle formed by the third sternal rib with its vertebral rib is more obtuse, viz. about 140°. The fourth sternal rib continues on the curve of its vertebral rib, while the fifth sternal rib cannot be said to form any angle other than 180° with its vertebral rib. The First Sternal Rib. This rib (fig. 76, 1 & 1’), of which I have not been able to see an adult specimen, seems (from the immature skeleton) to be a subcylindrical bone but little expanded at AXIAL SKELETON OF THE OSTRICH. 445 its sternal end, and only a little more than half the length of the second sternal rib. It is articulated distally in front of the fossa, which is preaxial to the first vertical septum of the pleurosteon. It only seems to be connected with the third vertebral rib by soft tissues intervening between them. STERNAL RIBS OF RIGHT SIDE (1-11, 3 natural size ; 1’, the first rib of an immature specimen, natural size; mm’, m1’, ty’, v’, and vi', 4 natural size). Fig. 76. 1-v1. Lateral aspect of the six sternal ribs; m'—v1'. The articular surfaces of the second, third, fourth, fifth, and sixth sternal ribs. In all the figures s represents the concavity which articulates with the dorsal conyexity of each pleurosteal septum, and i represents the concave articular surface for the yentral articular conyexity of each such septum. The Second Sternal Rib. _ The second sternal rib (fig. 76, 1 & 11’) is of about the same length as the second vertebral rib, though it articulates with the fourth vertebral rib. It is irregularly rounded at its proximal end; but its distal end is very greatly expanded dorso-ventrally, and compressed pre- and postaxilly, with a pneumatic foramen on its postaxial surface at its expanded part. The ventral end of the rib presents two superimposed elongated concavities (s & 7) to fit on to the two superimposed articular convexities of the first septum of the pleurosteon of the sternum (fig. 79, 7,s). The bone is slightly curved, with the convexity dorsad. The Third Sternal Rib. This rib (fig. 76, 11 & 11’) is rather less than half as long again as the second sternal rib, and is more curved. 446 MR. ST. GEORGE MIVART ON THE Its proximal end is rounded, but more antero-posteriorly compressed than in the second sternal rib. Its distal end is absolutely more but relatively less expanded; and the two concave articular surfaces are no longer subequal in size, but the ventral one (¢) much exceeds the dorsal one. The postaxial surface of the distal expansion of the bone exhibits large irregular excavations. The postaxial surface of the shaft of this rib is flattened; but its preaxial surface exhibits a prominent ascending ridge. The Fourth Sternal Rib. The fourth sternal rib (fig. 76, 1v & Iv’) is slightly more than twice the length of the second, and is more curved, the concavity, of course, being also dorsal. The preaxial end presents a wider articular surface for the vertebral rib. Its distal end is only very slightly more expanded dorso-ventrally, and is therefore relatively less so expanded. The disproportion between its two concave distal articular surfaces is still greater, the dorsal surface (s) being relatively still smaller. The postaxial surface of the distal expansion is still more excavated. The postaxial surface of the shaft is again flattened; but the ridge on its preaxial surface is much less marked. The Fifth Sternal Rib. The fifth sternal rib (fig. 76, v & v') is longer and slightly more curved than the fourth, being about once and a half the length of the third sternal rib. Its proximal surface (for articulation with the seventh vertebral rib) is broader and flatter than the corresponding surface of the fourth sternal rib, is absolutely rather less dorso-ventrally expanded, and therefore very decidedly so relatively. Its articular surfaces (s & 7) are still more unequal. Its shaft is more rounded post- axially as well as preaxially, and there is no ascending ridge on either surface. The Siath Sternal Rib. This rib (fig. 76, v1 & vi’) is longer again than the fifth one, but slightly less curved; its length is decidedly more than two and a half times the length of the second sternal rib. Its proximal end is still flatter, and the antero-posterior dimension of its articular surface is less inferior to its transverse extent. The sternal end of the rib is less extended dorso-ventrally than in any other sternal rib, except the first, and its articular surfaces have again approached each other. The inner, or rather preaxial, margin of the shaft is sharper than in the other sternal ribs. AXIAL SKELETON OF THE OSTRICH. 447 Occasionally a long delicate ossicle, a seventh sternal rib (though it does not join the sternum ), may be developed, extending ventrad and preaxiad from the ventral end of that rib which is postaxial to the one which distally unites with the sixth sternal rib. It is applied externally and ventrad to the side of the sixth sternal rib. It is, in fact, the ossified cartilage of a “ false” rib. Tue STERNUM. This is a wide sheet of bone with four margins and two surfaces (figs. 77 & 78). Its external (inferior or ventral) surface is convex, but irregularly undulating. Though there is no true keel, yet there is an oval elevated and flattened tract placed in the middle line at the postaxial half of the bone (fig. 77, f). The extreme antero- posterior length of the sternum is to its transverse dimension as about 5 to 3. Medianly and preaxially from the flattened tract a very low ridge may be developed forwards and dorsad. The internal (superior or visceral) surface of the sternum is strongly concave in both directions. At the bottom of the concavity there may be small openings into the substance of the bone. THE STERNUM (figs. 77 & 78, } natural size ; fig. 79, 4 natural size). Fig. 77, outer aspect; fig. 78, inner aspect; cc, coracoid grooves; f, flattened tract; ca ca, costal angles; la la, lateral xiphoid processes ; ma, median xiphoid process. Fig. 79, lateral aspect, showing the five excavations of the pleurosteon separated by five septa, each septum with two articular convexities for one of the sternal ribs; i, one of the ventral articular convexities ; s, one of the dorsal articular convexities. 448 MR. ST. GEORGE MIVART ON THE The preaxial margin shows two elongated coracoid grooves, which form with each other an angle of about 120°. These two grooves nearly meet together in the middle line (fig. 77, ¢c). The ventral (inferior) margin of the grooves does not extend so far preaxially as does the dorsal (superior) margin, especially towards the outer ends of the grooves. At the middle of the dorsal margin there may be a wide notch or a slight median prominence. Between this notch or prominence and the outer ends of the dorsal lip of the preaxial margin that lip is concave, owing to the production preaxially and outwards of the costal angles (fig. 77, ¢a). The postaxial margin of the sternum is deeply excavated, owing to the prominence of two external xiphoid processes (fig. 77, /z). In the middle of the postaxial margin there is a prominence, the median xiphoid, which, however, does not extend far (fig. 77, mx). Between this and each lateral xiphiod there is another rather irregular and less marked prominence. Each lateral margin of the sternum is slightly concave, owing to the angle (about 155°) formed by the external margin of the lateral xiphoid with that of the pleurosteon. The pleurosteon has a very wide and excavated external margin to receive the dorso- ventrally expanded ends of the sternal ribs: it is divided by five more or less complete bony septa, between which are, in the dry skeleton, five deep excavations (fig. 79). The edges of these septa proceed obliquely ventrad and postaxiad; and each develops two convex articular surfaces (one at the dorsal and one at the ventral part of each septum) for the corresponding articular concayities of the sternal ribs. The first septum has these articular convexities most prominent and most approximated (fig. 79, 7s). The fifth septum has them but very slightly marked. VERTEBRAL PARTS AND PROCESSES. It remains now, in conclusion, to consider the several parts and processes of the vertebra, as regards the varying condition of the same parts in different regions of the spinal column. Centra. With the exception of that of the atlas, each centrum is connected by suture or anchylosis with the rest of the vertebra to which it belongs. In absolute size the centra of the true dorsal vertebre excel. Except the preaxial surface of the axis (and apart from the atlas) all the cervical and true dorsal vertebre have their centra concavo-conyex at each axial end. The centra may, however, be slightly biconcave, as in some of the caudal vertebre. Adjacent centra may be anchylosed together from the twenty-sixth to the forty-sixth vertebra inclusive. AXIAL SKELETON OF THE OSTRICH. 449 The centra may or may not develop parapophysial, catapophysial, or hypapophysial processes. The pre- and postaxial diameter of centra may greatly exceed their transverse dimension, as in the cervical vertebre. The reverse condition may obtain, as in the thirty-first to thirty-seventh vertebre. The ventral surface of the centrum may be much excavated antero-posteriorly, as in the caudal vertebre. Neural Lamine. The neural lamine are attached to their own centra only, except in the thirty-second to the thirty-sixth vertebre. In the thirty-second, thirty-third, and thirty-fourth vertebre the neural arch rests on part of the centrum of the adjacent preaxial vertebra; but the arch may, as in the thirty-first vertebra, partly rest on the centrum postaxial to its own. The neural arches are highest relatively to their transverse extent in the lumbar vertebre ; they are most pre- and postaxially developed in the cervical vertebrae, and least so in the caudal vertebre. Adjacent neural arches may join each other by suture, as in the lumbar and sacral vertebre in the young condition—by anchylosis, as in the sacral region of the adult—by articular processes, as in most parts of the axial skeleton—or in none of these ways, as in the caudal region. The neural laminz almost always develop diapophyses. Neural Spines. These parts attain their maximum height in the lumbar region, where (in the adult) they unite together by anchylosis. They are most pre- and postaxially extended in the cervical region, and most trans- versely extended, relatively, in the caudal region. A neural spine may be trifid, as in the eighth caudal. Prezygapophyses. These surfaces may be wanting when the postzygapophyses exist in the same vertebra, as in the atlas; they may be wanting as well as the postzygapophyses, as in the post- dorsal vertebre. Very much longer (pre- and postaxially) than broad, they may broaden considerably, as in the dorsal vertebre. Having become larger they may again diminish in size, as in the posterior dorsals. Strongly convex pre- and postaxially, as in the cervical region, they may become nearly flat, as in the dorsal region. VoL. Viit.—PART vil. March, 1874. 3R 450 MR. ST. GEORGE MIVART ON THE Looking mainly dorsad in the cervical region, they come to look mainly inwards in the dorsal region. Postzygapophyses. These may exist without prezygapophyses, as in the atlas. They undergo changes of form and direction corresponding with those of the prezy- gapophyses, but they never become quite so small as do the latter in the dorsal region. Metapophyses. These processes would escape notice were it not for their recognition through their more developed homologues in other animals. They are more or less to be distinguished outside and ventrad of the prezygapophyses from the fourth vertebra to about the eighteenth, after which they seem to merge in the wider diapophysial expansion. Hyperapophyses. These are only conspicuous in the axis and the third and fourth cervical vertebre, where they are situated above the postzygapophyses. Rudiments of them are to be found on the atlas, and on vertebre postaxial to the fourth, till perhaps the tenth vertebra. Paraxial Parts. By paraxial parts‘ 1 mean those portions of the skeleton which diverge from the centra and neural arches laterally, and tend to surround the visceral cavity. They include:—1, upper transverse processes or diapophyses; 2, lower transverse _ processes or parapophyses; 3, pleurapophysial parts, 7. ¢. the ribs, with their capitular and tubercular portions, sternal ribs, and sternum, or parts representing the whole or portions of each pair of capitula and tubercula, with only a rudiment, or without any rudiment, of more distal pleurapophysial elements. These parts considered as one whole are, of course, as to size, most developed and most differentiated in the true dorsal vertebra. They are least differentiated in the true caudal region, where they stand out laterally as simple imperforate ‘‘ transverse processes.” Diapophyses. These are, with the neural spines, the most constant of all the processes, appearing even in the lumbar, dorsal, and caudal regions, where there are no zygapophyses. More or less antero-posteriorly extended in the cervical region, they are much so in the dorsal one. In the lumbar region they are long and slender, singularly remote from their respective parapophyses, and widely diverging from the latter. In the first four postsacral vertebra the diapophyses quite coincide with the para- 1 See P. Z.S. 1870, p. 260, note; and see also Trans. of Linnean Society for April 21, 1870. AXIAL SKELETON OF THE OSTRICH. 451 pophyses, as also in the posterior caudal vertebr, though in the latter a tendency to diverge is shown by the projecting extremities of each transverse process. Parapophyses. These are very constant structures, existing either as lower transverse processes or as articular surfaces for the capitula of the ribs. In the lumbar region their place of development singularly descends, being there placed quite at the ventral edge of the sides of the centra. They ascend through the three true sacral vertebre, till in the most anterior caudal (or anterior sacro-caudal) vertebree they reassume the same position they occupied in the posterior dorsal region. In the more postaxial vertebre their place of origin again descends, and occupies the side of quite the ventral surface of each centrum. The parapophysis of the thirty-first vertebra is almost entirely formed by the neural arch; but in the three succeeding vertebre each parapophysial prominence is produced by the concurrence of processes from the centra of adjoining vertebre, the thirty-second, thirty-third, and thirty-fourth vertebre having each such a projection from each end of each side of each centrum. These low complex parapophysial projections abut against the ischium and ventral margin of the acetabulum. Pleurapophyses. By pleurapophyses I mean ribs and all elements of the paraxial system which are serially homologous with ribs, including their capitula and tubercula. Consequently when the diapophyses and parapophyses are respectively connected by an osseous bridge, such bridge is pleurapophysial. To the thirty-sixth, thirty-seventh, and thirty-eighth vertebre three capitula of large size are attached; they expand as they extend postaxiad and dorsad to abut against the ilium. Hypapophyses. These are developed as twofold or azygos processes from beneath certain vertebre, generally from the seventeenth to the twenty-first inclusive, as already described. Catapophyses. These have been already noticed as developed from the more postaxial of the true cervical vertebra, generally anterior to the seventeenth vertebra. [ 453 ] XIII. On the Osteology and Dentition of Hylomys. By Joun Anverson, V.D., Curator of the Indian Museum, and Professor of Comparative Anatomy in the Medical College, Calcutta. Read June 4th, 1872. [Puate LXIV.] THIS most interesting form, originally described by S. Miiller and Schlegel’ from two individuals, one from the island of Sumatra and the other from Java, was afterwards discovered by Blyth? in a collection of Mammals forwarded to him by Major Berdmore, from Shuay Gyeen, in the valley of the Sitang, in the Tenasserim Provinces. Blyth writes of the specimens (which were two in number, an adult male and female) that “they so nearly resembled the H. suzllus of the Archipelago, figured and described by Dr. S. Miiller, that I should have considered them identical were it not for the greater development of the tail.” Having, however, removed the skull of one of Blyth’s spe- cimens, and finding it agree with the figure of the skull of H. suillus, S. Miller and Schlegel, I am inclined to regard them as of one species, although at the same time I propose to retain Blyth’s term until the question can be determined by the comparison of Peguan with Sumatran specimens. Dr. Gray’, in describing the genus Ptilocercus, which seems to be closely allied to Tupaia, incidentally mentions Hylomys, and states that the geographical range of the Tupaiine appears to be confined to the Asiatic islands, and that Borneo may be regarded as their more proper home, as possessing all the genera, viz. Tupaia, Hylomys, and Ptilo- cercus ; but he does not mention his authority for extending the distribution of Hylomys to this island. Now, however, it has been ascertained that Tupaia and Hylomys occur in Java, Sumatra, and Pegu, and that they are associated in these two islands and in the Malayan peninsula with Gymnura. The next notice of Hylomys that I am aware of is by Wagner*, who, however, added nothing to what had been already recorded about the characters of the genus. Mivart, in 1867, in his most valuable review of the osteology and dentition of the Insectivora’®, reproduced all that was known regarding these points ' Verhand. over. d. Nat. Gesch. d. Nederl. Mamm. 1843, pp. 153-158, pl. xxvi. fig. 1, pl. xxv. figs. 4-7, skull. ? Journ. As. Soc. Bengal, vol. xxviii. 1859, p. 294; ibid. p. 286. ? Proc. Zool. Soc. Lond. 1848, p. 23. * Schreber’s Siugeth. Suppl. ii. 1841, p. 554; ibid. Suppl. v. p. 855, pp. 529, 530, tab. 36. 5 Journ. Anat. & Phys. vol. i. pp. 301, 302. VOL. VIII.—ParT vil. April, 1874. 38 454 PROFESSOR J. ANDERSON ON THE in the structure of Hylomys; and again in 1868, in a further contribution to the same subject’, gave a short synopsis of the dentition and the general characters of the skull. The foregoing appears to be the extent of the literature on this most interesting form. In 1868, I obtained a specimen which may be provisionally referred to H. pequensis, Blyth. I found it lying dead on a path on a thickly wooded hill-side, in the Kakhyen hills to the east of Bhamé, in Upper Burmah, at an elevation of 3000 feet. A skeleton was made of this specimen, and I also used one of Blyth’s types for the same purpose ; and the result of my examination of these materials is as follows :— The rather long snout (Pl. LXIV. fig. 1) and the general form of the animal, if we exclude its large and rounded ear and short tail, confer on it the aspect of a Shrew; and when I saw my specimen lying on the road, before I picked it up, I thought it was a Shrew that had lost the greater part of its tail. The snout is slightly depressed, moderately attenuated and pointed. The bare fleshy portion at the tip is prolonged backwards on to the upper surface, for some distance behind the nostril; and in front it is marked by a longitudinal groove, which is continued downwards and backwards between the first incisors, with a fleshy ridge on either side of it, each of which termi- nates at the base of the incisor tooth of its own side. The upper and posterior extremity of the groove ends in a line with the anterior margin of the nostril; and behind this it is embraced by a triangular groove, the apex of which is directed back- wards. The two sides of the triangle spring from the anterior margin of the nostril. The posterior margin of the nostril has a convoluted fringe of about twelve fleshy papille. The eye is of moderate size, smaller than in Tupaia. The ears are large and oval, and might be designated nude, as they are only sparsely covered with pale, very short hairs, with the exception of a pencil of hair on the free margin of the tragus. The orifice of the ear is very wide and triangular, the apex of the triangle being directed downwards. The fore limbs are rather short, as are also the fingers; and from a short way above the wrist the anterior surface of the feet is almost nude, having only a slight covering of short, white and brownish, rather bristly hairs, through which the skin is quite apparent. The under surface of the manus is quite nude. Each finger bears on its under aspect from three to six transverse tubercles, that give it a resemblance to the finger of a Gecko. There is a small rounded tubercle at the base of the first finger, a larger quadrangular tubercle, or pad, at the base of the second, a still larger protuberance of the same kind at the bases of the third and fourth, and a rounded tubercle at the base of the fifth. A prominent oval, anteriorly pointed pad covers the carpus, the palm being smooth, but longitudinally folded; and the first and fifth fingers so approximate, when the limb is at rest, that they have to be pulled asunder to display the palm, which is of proportionate size to the body. ‘The claws are rather stout, of moderate length, and very little curved. The hind limb is likewise * Journ. Anat. & Phys. vol. ii. (2nd series, vol. i.) p. 146. OSTEOLOGY AND DENTITION OF HYLOMYS. 455 rather short, and is bare downwards from the middle of the tibia and fibula, being only sparsely covered with short hairs, as in the corresponding part of the fore limb. The whole under surface of the heel, however, has a covering of short, somewhat rigid, dark brown hairs, directed backwards; but the sole is quite naked, and has tubercles distributed over it as on the palm, and over the toes as on the fingers. The claws are more curved than those of the fingers. The tail is mouse-like, short, only one sixth the length of the trunk and head; it is almost nude, having only a sparse covering of short brown hairs, as on the tail of a rat, and the skin is thrown into little eminences, which are arranged in an annular manner. Above the root of the tail (dorsal aspect) there is a rather large surface, which is all but bare, having only a few minute hairs. This surface is covered with a brownish substance, probably an exudation from the skin; it is continuous with a similar but tumid surface surrounding the anus, and reaching forwards to the genitalia. Imme- diately before the anus, and in the female occupying the narrow interval between the opening of the vagina and anus, there are two nipple-like structures side by side, with a slight eminence before them and an obscure rounded eminence between and behind them. Each of the nipple-like bodies has an apical pore, with a fine bristle protruding from it. From the external margin of the vaginal surface a faint fold or linear eminence passes backwards to the root of the tail, where it is continuous with another slight fold that runs over the base of that organ. The former bears from four to six pores in a longitudinal series. I can observe only one pair of inguinal, and one pair of thoracic mamme. Length from tip of snout to vent 4 inches 11 lines; vent to tip of tail 11 lines. Anterior angle of eye to tip of snout 10 lines; posterior angle of eye to anterior margin of ear 6 lines; greatest length of ear 7 lines; greatest breadth of ear 7 lines; length of fore foot 8 lines; length of hind foot 1 inch. The general colour above is rusty brown, and on the underparts a pale yellowish white, the seminude portions of the limbs and tail being brownish yellow. The fur consists of three kinds of hairs :—Ist, a very delicate fine hair, the basal half of which is either succeeded by blackish or yellowish tipped with black; 2nd, stronger erect hairs, their dark basal halves succeeded by a broad rufous yellow band, which may be either terminal or tipped with black; and 3rd, a still stronger kind of almost bristly hair, with the terminal halves wholly black. The fur of the under surface is much shorter than that of the back and sides, and is mouse-coloured in its basal half, termi- nating in pale yellowish white. . The whiskers are black, and not very long. The skull (figs. 2, 3, & 4) differs from that of Tupaia in being less tapered both anteriorly and posteriorly, but more especially in the latter direction. The facial portion (snout) of the skull, although long, is not so attenuated as in Tupaia; and the nasal bones are not so much arched downwards as in the species of that genus, 382 456 PROFESSOR J. ANDERSON ON THE but are straight and directed forwards and downwards, the upper surface of the snout being flat and not convex transversely as in Tupaia. The nasals are slightly dilated in the middle, beyond which they contract posteriorly to a point, contracting also ante- riorly, but again feebly dilating at their anterior extremities. ‘They reach back to nearly in a line with the anterior margin of the orbit; and their posterior fourth is invested externally by the frontals. The lateral concavity of the snout between the second pre- molar and lateral incisor of Tupaia is absent in Hylomys, the sides of its snout being almost straight. The maxille reach a long way posterior to the nasals; and in this the skull differs both from Tupaia and Erinaceus. The premaxille do not reach to the frontal, but their inferior margin is on a line with the first premolar. In Tupaia the skull from the parietal eminence slopes markedly downwards and backwards to the lambdoidal crest, the two sides of which meet in a point, giving a triangular outline to that portion of the skull; and there is a semicircular interparietal. In Hylomys there is only a slight depression between the vertex and the middle line of the lambdoidal crest, which is almost on a level with the vertex, and not considerably below it as in Tupaia. The lambdoidal crest is much truncated above, forming an obtuse angle with its sides. This portion of the cranium, therefore, of Hylomys is much broader than in Tupaia; and the occipital region is slightly fuller, and, instead of sloping downwards and forwards as in the latter genus, it is directed downwards and backwards, the foramen ovale looking more backwards than downwards. In the character of the hinder part of the cranium, Hylomys approaches more to Erinaceus’. The space between the orbits is nearly flat, and is not so broad as in Twpaia; neither is the orbit enclosed by bone behind, nor the zygomatic arch perforated; and in this it resembles Erinaceus. The malar is a small spicule of bone applied to the outside of the perfect zygomatic arch. It also approaches Erinaceus rather than Tupaia in its large infra- orbital foramen and shorter canal, and in the prominent ridge at the anterior margin of the orbit. The temporal fossa is large. The postorbital process, which, however, is very minute, is more distinctly marked than in Erinaceus ; and immediately behind there is a slight contraction of the skull, so that that portion of the skull is narrower than the narrowest part between the orbits; and in this respect also it resembles Hrinaceus, and differs from Tupaia. It is like the latter genus, however, in having the ridge from the postorbital process, reaching backwards within a short way of the lambdoidal crest ; but the sagittal ridge, which results from the union of the ridges of either side, is very feebly developed, and before it joins the lambdoidal ridge it forms a flat triangular surface, with the apex directed forwards. In Erinaceus these postorbital ridges are far removed from the lambdoidal, by nearly half the length of the cranium. In Hylomys the skull attains its greatest breadth at the posterior roots of the zygomata. The palatal surface, unlike Tupaia and Erinaceus, has not any imperfections; but a faintly marked median ridge occurs as in Hrinaceus. It is deeper than in Tupaia, and of more ' J regret that there is no skull of Gymnura in this Museum with which to compare Hylomys. OSTEOLOY AND DENTITION OF HYLOMYS. 457 equal breadth throughout, it being widest between the first and second molars. Its posterior margin is slightly thickened, but not nearly so strongly as in the last-mentioned genus. ‘The palatal suture terminates posteriorly in a very minute spine, the equivalent of that structure which is so strongly developed in Hrinaceus; and there are indications of the transverse bony plate which is so well defined at the posterior margin of the palate of the Hedgehog. The pterygoid fossa is deep and well developed, and reaches forwards, even anterior to the posterior margin of the palate; and the outer plate is decidedly larger than the internal, and is directed outwards and slightly downwards, being prolonged backwards to the tympanic bulla, from which it is only separated by an extremely short ridge, internal to the foramen ovale. The mesopterygoid fossa is rather deep and narrow from before backwards, more so than in T'upaia, and less so than in Hrinaceus. The ectopterygoid plate is perforated externally, at its base, by two prominent foramina, one before the other, with an intervening arch of bone, producing the appearance of an incomplete canal, which is intensified by the grooved surface before the foremost foramen and which forms the outer margin of the sphenoidal fissure. The two foramina and the foramen ovale are in one line, and only separated from each other by narrow bony septa. One or two minute foramina open almost into the posterior border of the hindmost of the two foramina, which is slightly below, but immediately anterior to, the foramen ovale. Are these two foramina external ali- sphenoid canals? or is the more anterior of them the foramen rotundum? What leads me to believe they are the former is their direction ; for a fine wire can pass straight through them from before backwards from the outer surface of the cranium. If this opinion is correct, the foramen rotundum is merged in an enlarged sphenoidal fissure, which is large, and has its floor formed by the ectopterygoid and pterygoid plate of the palatine bone. The optic foramen is round, and separated from the inner and anterior margin of the sphenoidal fissure by a thin plate of bone. Below, and slightly posterior to the optic, is the suboptic foramen, which is even larger than the former. A fine wire passed through it comes out at a small foramen on the inner side of the internal opening of the optic foramen of the opposite side. The two foramina between the optic foramina, on the inner aspect of the skull, are transversely oval, rather well-marked apertures, much more developed than in Erinaceus; and in looking through the optic foramina from without they are seen beyond it. There is a minute foramen on the anterior margin of the optic foramen, and three venous foramina a short way above and anterior to it, two of which are directly above each other and posterior to the third, and lead at once into the cavity of the skull. Another and larger, slightly above and posterior to the two, and on the anterior orbital margin of the parietal and between it and the orbito- sphenoid, leads backwards to a short canal that opens on the internal aspect of the parietal, traversing it and the squamosal in the groove of the lateral venous sinus that communicates with the supraglenoid foramen. It opens between the parietal and squamosal slightly posterior to the postglenoid foramen, which is large. Anterior to 458 PROFESSOR J. ANDERSON ON THE this foramen and the three previously mentioned foramina, and at a considerably higher level than it, there is a well-marked foramen below and slightly anterior to the post- orbital process. It is directed upwards and rather backwards, apparently leading into the frontal sinuses; at least a fine wire inserted along it does not appear in the cranial cavity. The spheno-palatine canal is of moderate size, and opens at a short distance anterior to the sphenoidal fissure, and rather on a line external to it. Its lower border is formed by the pterygoid plate of the palatine. The posterior palatine canal is very short, and opens anterior to the external angle of the slightly thickened posterior margin of the palate. The anterior palatine foramina are bounded in front, within, and externally by the premaxille, and posteriorly by the maxille. The lachrymal foramen, which opens into the orbit above and rather external to the infraorbital foramen, is marked at its anterior border by a strong process that projects outwards and slightly backwards, being con- tinued upwards along the margin of the orbit to near the postorbital process as a distinct ridge as in Hrinaceus; and, as in that genus, the prelachrymal process itself is perforated by a minute foramen. Externally the lachrymal canal appears as a distinct ridge, arching forwards and downwards to the upper margin of the external orifice of the infraorbital foramen. Immediately above the infraorbital foramen there is a small deep lachrymal pit, which also occurs in a less-marked degree in Erinaceus, but is not observed in Zupaia. The infraorbital foramen is separated from this pit by a thin plate of bone. It is a large opening leading into a wide and moderately long canal, that opens by a large orifice immediately above the interior fang of the fourth premolar. To return to the base of the skull (fig. 2), the mesopterygoid fossa terminates in a true excavation, but not of the same marked character as in Hrinaceus. In Hylomys the tympanic bulle are more posteriorly divergent than in Hrinaceus, and in this they more approach Tupaia; but the area between them is excavated into two shallow, elongated, oval troughs lying side by side at the extremity of the basisphenoid, but separated from each other by a well-marked, low, sharp, longitudinal ridge, and defined on the sides by the expansions of the basisphenoid, which go to assist in the formation of the tympanic bulla. At the base of these processes there are one or two small foramina. The glenoid surface is nearly flat externally, but slightly concave internally. The condyles of the occipital are divided, or nearly so, into two articular facets, of which the lower is almost circular ; but the skulls do not appear to be young. ‘There is a rather large precondyloid foramen on each side, immediately below the condylar constriction; and anterior to it is a jugular foramen. There is a well-marked paroccipital process, anterior to which are two rather obscure processes in the mastoid region, and behind the glenoid surface a postglenoid foramen, but no process. The tympanic is ring-like; but there is a vacuity between it and the tympanic process of the basisphenoid, and its anterior extremity is anchylosed to the process of the alisphenoid immediately below the foramen ovale, which is enclosed by the last-mentioned bone. The mandible is proportionally shorter than that of Tupaia, and in general form it OSTEOLOGY AND DENTITION OF HYLOMYS. 459 approaches more to that of Hrinaceus, which it also resembles in its stouter horizontal ramus, and in the more erect position and form of.its coronoid. The condylar articular surface has not a great transverse extension. The ridge (internal) running from the con- dyle to the ramus is very prominent. The anterior orifice of the dental canal is below the anterior margin of the last premolar. The process of the angle is hook-like and pointed. The dental formula is :—i. an c. it. pm. itp m. stg 44 (figs. 5, 6, & 7). The first incisor is conical and curved, with a very minute posterior cusp at the base of the crown. The second incisor, separated by a short interval from the first, is con- siderably smaller than that tooth, and is not so much curved; but it has also a minute cusp at its base posteriorly. It has about the same length as the canine. The third incisor is placed close to the second, and is about half its size, but differs from it in having a small, laterally compressed, almost triangular crown, with a constriction above it, the posterior basal angle of the crown showing the indication of a rudi- mentary cusp. The canine is immediately behind the premaxillary suture, separated by a short interval from the third incisor. The anterior half of its crown resembles the second incisor; but the hinder half bears a minute, almost hook-shaped cusp at its base. The first premolar is immediately behind the canine, and is followed in close succession by the second and third premolars, all having the cingulum more or less developed internally and externally. They are all small teeth, with about the same downward extension. The second is slightly larger than the tooth on either side of it; but all are of about the same length as the third incisor. Each has a central conical crown, very faintly curved, with a small cusp at its base anteriorly and posteriorly, a development of the cingulum. ‘The anterior cusp is most developed in the second ; and the third tooth has the most pronouncedly marked posterior cusp, with the anterior cusp hardly visible. The crown of this last-mentioned tooth is more truncated than that of the others. These premolars have a much stronger resemblance to the first, small anterior premolars of Hrinaceus than to the corresponding teeth of Tupaia. The fourth premolar is a large tricuspidate tooth, with an external cingulum and large pointedly conical crown, which has the greatest vertical extension of all the teeth in the upper jaw. It has two internal, pointed, conical cusps, the anterior of which is much the longer and larger. Besides these, however, the cingulum is so much developed posterior to the crown that it almost produces another cusp. This tooth is very closely allied by its structure to the last premolar of Erinaceus, and is very dif- ferent from the same tooth of Tupaia. It is almost the exact fellow of the former. The fourth molar is the largest of the teeth in the upper jaw, and consists of four cusps, two external and two internal; but the cingulum which encircles the tooth is very strongly developed externally, and behind the postero-external cusps it almost forms a fifth cusp. The anterior of the external cusps, which are conical with a greater vertical extent than the internal cusps, has a less vertical extent than its fellow. The internal cusps are the same as in Hrinaceus. Situated, internal to the postero- 460 PROFESSOR J. ANDERSON ON THE external cusp, and between the two internal cusps, is a small conical tubercle (the ridge of Huxley and Mivart in Erinaceus), that for its size and characters merits to be regarded as acusp. It is not a ridge, but a distinct fifth cusp, situated in the locality I have indicated; and in a skull of an Indian Erinaceus before me it forms a well- marked cusp in the same situation. The second molar is the same as the first, only smaller, and without the median cusp (ridge), which, however, is represented by a ridge connecting the postero-external to the antero-internal cusp. The third molar is still smaller than the former, with the cingulum less developed externally, and only one internal cusp, =tricuspidate. The antero-external cusp is the same as in the pre- ceding molars; but the postero-external cusp is externally broad, and about twice the breadth of its fellow before it, and its apex shows an obscure notch, which would seem to indicate that it has resulted from union with the postero-internal cusp. The solitary internal cusp has all the characters of the antero-internal cusps of the two pre- ceding molars. That the premolars and other molars of this curious insectivorous mammal should have such a remarkable resemblance to the corresponding teeth of Erinaceus, and that its last molar should differ so much from the same tooth of that form, is very remarkable. The first incisors of the lower jaw are widely different in form from those of the upper jaw. They have cylindrical fangs, terminating in laterally expanded, shovel-like crowns, compressed from before backwards, slightly convex ex- ternally and concave internally. The second incisor has a strong resemblance to the first, but the crown has not quite the lateral expansion of that tooth. The posterior margin, a little above the neck of the tooth, is marked by a very minute process, which becomes intensified in the same position in the tooth posterior to it, where it is evi- dently the product of the cingulum. The crown of the third incisor is also laterally compressed, somewhat oval in outline, and is set on unequally to the fang, projecting anteriorly more in advance of the shaft of the tooth than posteriorly. The margin of the crown posteriorly above the neck bears a well-marked process. The canine partakes of the characters of the lower incisors in its compressed character from without inwards, and general form. The crown above the neck is slightly contracted; but it expands and terminates in a more truncated point than the first incisor. Its surface, along the posterior margin, is markedly concave. The first, second, and third premolars are the smallest teeth in the lower jaw, and they have all one form, the middle tooth being the smallest of the three. They are more or less compressed laterally; and the first is set obliquely in the jaw like the teeth preceding it; but the second is less so, and the third is nearly vertical. ‘The crown of the first is small and conical, with a cusp nearly as large as itself projecting forwards from its anterior margin, and another, and smaller, posteriorly, the product of the cingulum. In the second the crown is merged in the anterior cusp, so that the external outline of the crown is rounded, with only the posterior cusp developed. The third has a slightly recurved, small, short, conical crown, with an anterior and a posterior cusp. The fourth premolar is the most verti- OSTEOLOGY AND DENTITION OF HYLOMYS. 461 cally extended tooth of the whole lower jaw. The crown is long, conical; and slightly recurved; and the cingulum encircling it forms a process anteriorly and posteriorly at its base, but no distinct cusp. ‘The first and second molars have five cusps, of the same form and arrangement as in Erinaceus, with the cingulum very strongly marked externally. The third molar, which is only a little more than half the size of the first, has also five cusps, the anterior of which, however, is even less developed than it is in the second molar. The basihyal is a flat, transversely extended, bony plate, consisting of two halves placed side by side, with a posterior cornu directed outwards and backwards, to which the slender, slightly upwardly curved thyrohyal is attached. The ceratohyal is a short, rather stout, cylindrical bone, constricted in the middle, with a head at either end; and it is succeeded by another, slender, laterally compressed bone, bent at its upper half— which leads me to believe that it is the stylohyal. But as the hyoid apparatus before me is imperfect, I cannot speak with any degree of confidence as to the number of pieces in the anterior arch. Measurements of the Skull of Hylomys peguensis, Blyth. in, lin. in. lin. Greatest length of skull to tip of intermaxillaries . -1 44$tol 48 Greatest breadth across zygomatic arch UTD 0 ce Greatest breadth of squamosal . . . HOP Lev.) ORG? Breadth above lachrymal process - 0 52, 0° oF Lachrymal process (upper margin) to tip of ae Ole stares yh Ol Ory, Breadth in front of canines + O22 5) 0822 Breadth in front of fourth paanalas -0 4 ,0 4 Breadth (external) at first incisors . OM Zi nO se Breadth behind supraorbital process -0 4 ,0 32 Least breadth between orbits . 0) Ag pOne4s Posterior palatine margin to tip of intertheccillsiis 09 , 0 82 Greatest breadth between alveolar internal margins (between second and third molars) 0 33,0 3 Length of alveolar border . OM Sie OSE Breadth behind origin of zygomatic sci, infexiond dipect of skull . ai On Tulip One6y Distance between arnt bulla fartteria eatrenity) Oks es Ooi: Distance between tympanic bulle (posterior extremity) 0 42,,0 42 Depth of skull at anterior extremity of nasals . 0 13,0 12 Depth of skull at posterior extremity of nasals . 0 3 ,0 22 Depth through posterior margin of palate 0 42,0 42 Depth through highest point of parietal . 0 4% ,,0 42 VOL. VIII.—PART vill. April, 1874. 3T 462 PROFESSOR J. ANDERSON ON THE in. lin. in. lin. Anterior extremity of PaEABTG to extremity of process of lower jaw . . oe oR Sean eg Ele towels Anterior extremity ee STMTRFEE to aomigie Se eee er tad en) erica ay ol) Anterior extremity of symphysis to coronoid process . . . 0 113 ,, 0 11 Tenethi of, alveolar sumace oo asst atte it atime ek aes Aa Depth through coronoid process. . . + th ovat eee, OF At Ouse Depth from base of corono-condyloid natch