Art. XIII.—On Cnemiornis calcitrans, Owen, showing its Affinity to the Lamellirostrate Natatores.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 6, 1873, Page 76

Art. XIII.—On Cnemiornis calcitrans, Owen, showing its Affinity to the Lamellirostrate Natatores. By James Hector, M.D., F.R.S. Plates X.—XIV.A. [Read before the Wellington Philosophical Society, 18th August, 1873.] Since the discovery of the very interesting specimen of the Moa's neck with well-preserved muscular tissue and integuments in the Earnscleugh cave, in the interior of the Province of Otago, * Trans. N.Z. Inst., IV., 111. the locality has been visited several times, and especially last year by the Hon. Captain Fraser, who obtained, besides Moa bones, several belonging to a smaller-sized bird, being part of a skeleton most of which had been previously removed by some gold-diggers. I recognised these to belong to Cnemiornis calcitrans, of Owen, the only difference being that the humerus differed from that described by Professor Owen † Trans. Zool. Soc, V., 399, Pl. 66. in several important characters. ‡ Trans. N.Z. Inst., V., 406. Besides the humerus were the right femur and tarso-metatarsus and the metacarpal bones; the two former agreeing accurately with Professor Owen's description and plates, and the last-mentioned being a new addition to the osteology of the bird. The chief difference in the humerus from that attributed to this bird by Professor Owen is its greater proportional size, it being equal in length to the femur, instead of one-ninth less, and in its having a very distinct pneumatic fossa, closed by a cribriform bony septum. In addition, the tuberosity representing the pectoral ridge is not so wide, and the proximal articular surface is slightly broader and more convex at its middle part than in the typical bone. These characters might lead to the surmise that it belonged to a carinate bird, but the massiveness of the bone was thought sufficient to disprove this. In order to determine this point with some degree of accuracy, I compared the weight with the bulk of the same bone in several species of birds, with the following results:— Weight. Bulk. 1. Cnemiornis (Earnscleugh cave) 10 244 2. Weka (Ocydromus) (non-volant) 10 210 3. Kakapo (Stringops) (non-volant) 10 187 4. Kaka (Nestor) (volant) 10 131 5. Hawk (Hieracidea) (volant) 10 126 A small portion of the shaft was also removed, and the thickness of the bony wall found to be so great that the internal diameter is only two-thirds that of the external. In consequence of the above divergence of character from the humerus described by Professor Owen, I was much interested in obtaining the

Skeleton of Cnemiornis Calcitrans

remainder of the bones of this skeleton, and, after tracing it through several hands, Captain Fraser at last succeeded in obtaining possession of the box containing the bones in the same state in which they had been found, and at once handed them over to me for description. The skeleton is still far from complete, but the following bones are in a very admirable state of preservation:—Skull; vertebræ, 12 cervical, 4 dorsal; sternum; furculum; humerus, right; metacarpal, right; sacrum; femur; tibiæ, both; tarso-metatarsal; ribs, six. They agree perfectly in appearance, colour, peculiar stains, texture, and other external characters, so that there is no reason to doubt that they all belong to one individual, which is further confirmed by the study of their anatomical characters. The structure and form of the skull and sternum shew that this bird belongs to the Lamellirostrate family of the order Natatores, but that the power of flight had become obsolete, and that it differs from most others of the duck kind in its short, lofty head, very solid palate, and in the peculiar character of the tympanic cavity, which is bridged across by a bony process between the mastoid process and the basi-occipital. The great solidity of the skull, and the absence of occipital fontanelles and of all sutures excepting the naso-frontal and the lachrymal, is also remarkable. Every bone of the skeleton, excepting the upper part of the sternum, has the close-grained, reticulated surface which is so characteristic of the bones of Cnemiornis, giving the impression of a very solid, powerful framework, that in the fresh state would contain much oily matter. The absence of the power of flight is evidenced by the rudimentary tubercular ridge that represents the keel, and the small area of attachment for the pectoral muscle on the surface of the sternum. Skull. Pl. XI. What remains of the skull is well preserved, every process being sharp and distinct, as in a freshly-macerated specimen, but unfortunately the following appendages have been lost:—the quadrate, jugal, pterygoid, and lachrymal bones. The shape of the head, including the lower jaw, and allowing the usual proportion for the quadrate, would have been:—Height, two-thirds the length; transverse diameter, one-half the length. The brain-case is short, high, and compressed laterally, its posterior-inferior diameter being greatest. The nasal portion of the skull, which is distinctly shorter than the cranial, is detached; and the mobility of the upper mandible, which in such birds is usually effected by the flexibility of the thin nasal bones, must, if it existed, have been effected by a straight joint with thick, irregular margins, somewhat as in the parrot. In the occipital region the muscular ridges are moderately developed. The

condyle (oc) moderate, reniform, flattened, and excavated above, with a mesial notch, slightly excavated beneath, but not laterally. The foramen (Fm) is very large, being one-third the height of the occiput in its vertical diameter, which is one-fourth greater than the transverse. It is rounded above, but has the lateral and inferior margins almost straight. The occipital area is rather square in form, with a blunt mesial ridge (So), having a shallow pit (a) on either side, but no fontanelles. A bold par-occipital process (po) extends downwards and backwards on each side, and forms the extremity of the cranium in that direction, giving rise to the most remarkable feature in its external conformation as viewed laterally. A deep perforated pit (b) separates this process from the basi-occipital, which is very largely developed, and has two inferior lateral processes (l) separated by a wide, smooth, sub-condylar notch (c), and then extends forward as a broad, slightly-concave surface, which occupies a large area at the base of the skull (br). The basi-sphenoid (Bs) has a small share in the base of the skull, and has large oval basi-pterygoid facets (bp) only slightly divergent. The character presented by the tympanic fossa is very remarkable, as it is divided into a posterior and anterior portion by a quadrate ossicle (ms) that connects the tip of the mastoid process with the basi-occipital and with the anterior process of the ex-occipital, thus enclosing a wide canal descending obliquely backwards and outwards, with a sub-circular aperture deeply notched inferiorly. The articular portion of the tympanic fossa, with its two facets, is thus separated from the posterior or auricular portion, a character which appears to be unique. * I have since found it, but less marked, in Cereopsis. The frontal bone (F) is slightly swollen at the vertex and depressed between the orbits, which have strong overhanging orbital processes, on which are rough, deeply-impressed areas (d), which probably gave attachment to a posterior development of the cere of the mandible, these impressions being separated by a smooth groove with only a faint mesial ridge. There appears to have been a deep notch (d′) in the upper part of the orbital border, but the lachrymal bones having been lost this is not very clear. The width of the nasal suture (Fn) is equal to the length of that for the attachment of the lachrymals (Fl), which extends from the glandular groove to the transverse suture. The inter-orbital septum is complete, and there are well-ossified rhinal chambers (Rh). The roof of the orbit is flat, and with a very slight granular groove. The optic foramen is at the posterior and inner angle of the orbit, directly above the front of the basi-pterygoid facet. Behind the post-orbital process is a deep imperforate pit. The brain cavity extends for 6 lines anterior to the optic foramen. The upper mandible has all its elements completely fused; the large nostrils (e) occupy more than half of the superior sloping area, their

Skeleton of Cnemiornis Calcitrans

aperture being directed outwards, forwards, and upwards, rounded in front and angled behind, and they are separated by a smooth bony interspace (f) which is one-fourth the width of the mandible. The tip is rounded, with a tumid area for the attachment of the horny mandible, the length of which is equal to the width. Interiorly the palatal plate is flat, with deeply-incurved borders, notched on each side of the tip, and deeply excavated by a longitudinal groove (h), which is perforated by two well-defined apertures, the one (h′) large and directed upwards, the other (h″) small, directed backward in a line with the groove. The palatines are firmly united with the vomer, the upper surface of which has a slight groove to receive the pre-sphenoid. The lower mandible is stout, but broad and compressed in every part, the rami preserving a lamellate structure throughout, and being united by a broad symphysis (sy), the length of which is equal to one-fourth of the mandible, the anterior half being flat and the posterior excavated. Inferiorly the punctate surface of attachment of the horny mandible covers the whole of the symphysial portion. Pelvis. Pl. XIV. This bone agrees with Professor Owen's description * l. c. p. 397. so far as his imperfect specimen enabled him to fix its characters, but the complete preservation of the bone obtained from the Earnscleugh cave enables me to add the following:— The neck of the ischium (a) is compressed to form the inferior notch, which is 9 lines in diameter, and contracted posteriorly (at b) to 5 lines. The ischium then expands to 8 lines, with a concave external surface, its upper margin being united with the ilium for the last two inches (cd), forming a rhomboidal convex plate with a thin posterior margin that descends obliquely backwards; the inferior margin is produced (e), and has been united by cartilage to the pubic styles for about 9 lines. The latter are attached by a stout compressed process to the inferior fifth of the acetabulum, and thence produced backwards as a narrow, curved bone, flat externally, and with a strong ridge internally, 3 lines wide at its narrowest part, and posteriorly expanding into a flat curved process that descends at an obtuse angle and continues the edge of the posterior pelvic aperture (f). The coccygeal bones are wanting. The first sacral vertebra is anchylosed to the sacrum only by its spinous process. The posterior roof of the pelvis is pierced by eight foramina in almost parallel lines an inch apart, separated by a concave interspace anteriorly and a convex ridge posteriorly; the rhomboidal form of the area being produced by a blunt expansion of the border which, on either side, overhangs the anti-trochanteric process.

Sternum. Pl. XII., fig. 1, and XIII., figs. 1, 2. This bone is almost perfect, having lost only a few lines of its inferior margin, and, though it differs considerably from the characters attributed to it by Professor Owen, this is, without doubt, due to his not having had a connected fragment of the superior portion of the bone, so that its enormous posterior concavity was not appreciated. It is chiefly remarkable for its regular oblong shape, without any irregularities of outline or unossified interspaces. The texture must be cancellated, for, though apparently thick and massive, it is very light in proportion to its size, as will be seen in the appended table. Its general form is scaphoid, the concavity being very marked in the upper half, amounting to one inch in depth measured from a transverse chord, and to one-and-a-half inches in depth if measured from a longitudinal chord; the total length of the latter being 7 inches. The anterior width at the costal processes (a) is 4 inches, and at the posterior end of the costal border 3 inches 6 lines. The costal border (e-e′) occupies half the lateral margin of the bone, the posterior half of the line being only slightly concave interiorly, and exteriorly being flat in the middle and sloping very slightly to the inferior angles. The superior margin is thin, and presents a wide mesial notch (f) and two lateral notches (g), which are bounded exteriorly by the costal processes, which project backwards and upwards for 6 lines. The coracoid grooves (b′) are 1 inch 6 lines in length and 2 lines in depth of anterior border. They are separated by a slight triangular interspace (f′) 5 lines wide, beneath which is a smaller triangular pit (h). The keel (c) commences by two angular ridges bounding this pit posteriorly, and forms a blunt process 3 inches in length (c-c′), expanded anteriorly to a rough tuberculate surface 4 lines in width and 9 in length, and then compressed into a narrow tuberculate ridge that is gradually lost in the smooth convex surface of the bone at less than one-half the distance from its superior margin. The greatest elevation of the keel above the convex surface of the bone is less than 3 lines. The impression (d) for the attachment of the pectoral muscle extends from the exterior angle of the coracoid notch towards the posterior part of the keel, including a triangular area which occupies only one-sixth of the exterior surface of the bone, showing the extremely limited and feeble attachment of the great muscles of flight. Large pneumatic foramina (i) exist in the interior of the bone at the upper angles, and one (i′) on the exterior surface on the left side only. Furculum. Pl. XII., figs. 2 and 3. The clavicles are completely joined into a smooth, slightly-compressed furculum, like that of the goose, except that the antero-posterior curvature is confined to the articular processes, which diverge 1 inch above the general

Cnemiornis Calcitrans.

Cnemiornis Calcitrans.

plane, and are expanded with a large pneumatic foramen (a) on the external surface, overhanging which is a triangular articular surface (b). The coracoids have not been found, but must have been stout triangular bones 2 inches in length and 1 ½ inches wide at the sternal attachment. The scapula also is missing. Vertebræ. The following vertebræ have been preserved:—The 2nd, 3rd, 5th, 6th, 7th, 8th, 10th, 11th, 13th, and 14th cervical; 2nd, 4th, 5th, and 6th ribbearing or dorsal; the last having no hypapophysis may be termed the first of two pre-sacrals, the second of which is wanting. The total number of vertebræ seems to have been:— Cervical 14 Dorsal 5 Pre-sacral 2 Sacral 17 The total length from the tip of the beak to the coccyx would be about 35 inches. Ribs. Pl. XIII., figs. 3–7 and 7′. The 3rd, 4th, and 5th ribs of the right side; the 6th and 7th of the left, and also the 7th sternal rib of the same side, have been preserved. The first six ribs have well pronounced uncinate processes (a) with a broad attachment one inch in length, that on the 6th rib (a′), being bent backwards so as to be almost parallel with the bone, and having a blunt expanded tip. The ribs are much compressed, their margins having wavy irregular outlines. They are broad in the middle and taper off towards each extremity, when viewed laterally. The last two ribs articulated with sacral segments, the total number having probably been nine. The number of sternal ribs was seven. Humerus. Pl. XII.; figs. 4 and 5. The humerus has already been described, so far as it differs from the bone attributed to Cnemiornis by Professor Owen, and the femur, tibia, and tarso-metatarsal have been figured in Pl. XIV. A for convenience of reference by collectors. They agree so perfectly with Professor Owen's description as to require no further notice, except to point out that the external articular process of the metatarse, instead of being obliquely reflexed, as in the goose, swan, and other swimming birds, is straight, as in true cursorial birds, indicating that the habit of the bird was rather to walk on land than swim in water. The metacarpal (figs. 6, 7) is made up of the first and second digits, which are completely fused at both extremities, leaving a narrow interspace (a) for less than half the length of the bone. It resembles closely the corresponding bone

in the weka (Ocydromus), and bears almost the same proportion as in that bird to the length of the humerus, or about two-fifths of the length of that bone. I should state that this bone has been found in several instances in Canterbury by Dr. Haast associated with fragments of a similar humerus, and rightly assigned by him to Cnemiornis. A fragmentary skull, in which the basal, posterior, and nasal portions are wanting, and several leg bones, in the Colonial Museum—Some found by Mr. W. D. Murison, in Otago, and others from the Wairarapa, in Wellington—must also be referred to this species, and prove that it was widely dispersed over both Islands of New Zealand. As the leg bones of Cnemiornis are not infrequent in collections, especially from the most recent turbary and cave deposits, this bird must have been of common occurrence, and the foregoing details afford conclusive evidence that it was a gigantic bird, probably allied to and of similar habits to the Cape Barren Goose of Australia (Cereopsis* Having procured a skeleton of this species for comparison, through the kindness of Professor M'Coy, I am able to confirm this surmise. Among the chief structural differences, I notice the presence in Cnemiornis of an extra pre-sacral vertebra, so that two, instead of three, ribs articulate with the sacrum, and an elevated pent-roof arrangement of the ossa innominata, which indicate more decided cursorial habits.), but in which the power of flight had become obsolete. From the size of its pelvis, ribs, and sternum, the bulk of its body must have greatly exceeded in proportion any of its existing congeners, while its lower extremities were not less remarkable for their massive development. The height of its back above the ground exceeded 2 feet, and the length of the body from the beak to the tail was at least 32 inches. Table of Admeasurements, in Inches. Skull. Weight, 535 grains (with lower jaw). Length 4.5 Breadth across paroccipitals 2.0 " " post-frontals 2.0 " " temporal fossa 1.5 " " middle of upper mandible 1.5 " " tip " " 1.0 Length from condyle to pre-sphenoid 2.1 " of palatines 1.4 " of pre-maxillary 1.5 "from point of external nostril to end of pre-maxillary 1.0 " of nasal aperture 0.8 Width of nasal aperture 0.4 " of internasal septum—anterior 0.4 Supra-occipital tuberosity to post-nasal suture, following the curve 2.6 Supra-occipital tuberosity to external basilar process 1.6 Length of vertical basilar area 0.3 " horizontal " 0.5 Width of " " 1.3

Cnemiornis Calcitrans.

Sternum. Weight, 1009 grains. Extreme length of side 7.0 Extreme width at costal process 4.2 " " middle 3.7 " " posterior margin 3.6 Costal margins—length of each 3.0 " width at middle 0.4 Coracoid grooves—length of each 1.5 " interspace—inter-coracoid 0.6 Keel—length 2.5 Supra-carinal fossa—length 0.6 " " width 0.6 Height of arc 2.2 Furculum. Weight, 81 grains. Vertical chord 2.8 Transverse chord 2.3 Total exterior length 7.0 Average diameter 0.3 Diameter of articular process 0.5 Humerus. Weight, 412 grains. Length 6.2 Extreme breadth, proximal end 1.5 " " distal end 1.1 Circumference, middle of shaft 1.6 Metacarpal. Weight, 85 grains. Length 1.5 Femur. Weight, 1021 grains. Length 6.0 Tibia. Weight, 1789 grains Length 12.1 Tarso-metatarse. Weight, 787 grains. Length 5.6 Ribs. 3rd—length 5.0 4th " 5.5 5th " 6.0 6th " 6.4 7th " 6.8 7th sternal—length 4.6 VertebrŒ. 2nd Cervical. 14th Cervical. 2nd Dorsal. 1st Pre-sacral. Transverse diameter of centrum 0.25 0.6 0.8 0.6 Breadth through transverse process 0.8 1.7 1.8 1.9 Antero-posterior diameter, or height of centrum 0.2 0.4 0.4 0.6 Total height 0.7 1.3 1.6 2.4 Length of centrum 0.9 0.8 0.7 0.7 Pelvis. Weight, 2400 grains. Total length 11.3 Antacetabular 4.1 Acetabulum 1.1

Height through acetabulum 2.5 Antitrochanteric width 3.1 Mesial iliac suture 3.7 Post-sacral area—length 4.7 " " width between anterior lateral foramina 1.2 " " width between posterior lateral foramina 1.1 Ischiatic notch—breadth 1.2 " " length 3.0 Pubic style—length 7.5 " breadth, greatest 0.5 " " at middle 0.3 Superior posterior iliac interspace 0.8 Inferior " " 2.2 " " inter-ischiatic space 3.8 Posterior pubic interspace 2.0 Ilium—anterior width 2.0 " posterior width at middle 1.0 1st sacral vertebra—height of centrum 0.7 " " " neural canal and spine 2.0 " " width of centrum 0.6 " " " transverse processes 2.0 Length of 8 anterior sacrals 3.5 Description of Plates X.—XIV. A. (Cnemiornis calcitrans.) Plate X.—Restored skeleton one-fifth nat. size, from a photograph. XI.—Skull, nat. size. Fig.1. Side. Fig. 2. Inferior. Fig. 3. Posterior. XII.—Fig. 1. Sternum, front view. 2–3. Furculum. 4–5. Humerus. 6–7. Metacarpus. XIII.—Fig. 1. Sternum, side view. 2.Inner side of right costal process. 3–5. 3rd, 4th, and 5th ribs of right side. 6–7. 6th and 7th ribs of left side. 7′. 7th sternal rib of left side. XIV.—Figs. 1–3. Pelvis. XIV.A.—Figs. 1–2. Femur. 3–4. Tibia. 5–6. Tarso-metatarse. Erratum.—The words “Weight” and “Bulk” in the table on page 76 should be transposed, the bulk of the bone being reduced to the same number in each case for comparison. The bulk was ascertained by immersing the bone in water after coating it with a film of wax, and ascertaining the displacement by weight.

Cnemiornis Calcitrans.