Palæontological Report of the Princeton Scientific Expedition of 1877

HELALETES, Marsh.

Am. Journ. Sc. and Arts, 1872, p. 218.

Additional material enables us to give further characteristics of this genus, which may be thus defined: Mammals, allied to Lophiodon, having a third lobe to the last lower molar, short symphysis, and with the teeth of the lower jaw forming a continuous series. Astragalus has narrow, oblique condyles.

Synopsis of Species.

Helaletes latidens, sp. nov.

Lower Jaw.—The ramus is stout, and resembles in general form that of Hyrachyus, but the alveolar border is straighter than in that genus; the ascending portion forms a right angle with the horizontal portion, and the masseteric fossa is deep. The symphysis is comparatively short, and there is complete bony union between the rami, which do not diverge so much as in Hyrachyus. As in that genus, there are several small mental foramina beneath the premolar series. The dental series forms a nearly continuous line, the only diastema being a very short one between the first and second premolars. The incisors have compressed fangs, and are arranged in a slight curve. The canine is small and subcircular at base. The premolars also are small; they somewhat resemble those of Hyrachyus in shape, but are smaller, and have the posterior lobe not so well developed. The first is implanted by a single fang, the others by two. The molars (except the last) consist of two pyramidal lobes, which form transverse cutting-crests, as in Tapirus and Hyrachyus. The basal ridge is shown on the anterior and internal sides only. Externally the lobes meet at their bases, inclosing the valley on that side. The last molar is by far the longest of the series, and consists of three lobes, of which the anterior is the largest, and the posterior the smallest. The basal ridge at the anterior part of the tooth is very strongly marked.

Upper Jaw.—The molars increase in size from first to last with considerable regularity. The premolars are small, and, as in Hyrachyus, have one single internal cone, which is joined by transverse ridges from the two outer cusps; the anterior accessory tubercle is well developed. The first molar is considerably larger than the last premolar; it is much worn, and does not show anything more than that there are two internal cones. The anterior conical buttress is very largely developed throughout the true molar series. The second and third molars resemble almost precisely the corresponding teeth of Hyrachyus. Their shape, however, is more nearly square than is usual in that genus.

Measurements.

UPPER JAW.

LOWER JAW.

Found at Henry's Fork.

ARTIODACTYLA.

GENERA INCERTÆ SEDIS.

ITHYGRAMMODON, gen. nov.

Our fragments of this fossil show an animal about the size of the llama, and approximating more closely to the Tylopoda than to any living forms.

The genus is established upon the two premaxillary bones, containing the incisors, parts of the maxillaries, the canine and the first premolar; besides fragmentary portions of the palatine plates. As the peculiar features of these remains render them unique, we are warranted in establishing the genus Ithygrammodon.

Generic characteristics.

The premaxillaries are long, narrow, and very straight, bent slightly in on their own axis as in Camelidæ, but expanded laterally at the end of symphysis as in Pecora. The upward processes are sharply defined, with a wide, rounded upper border.

The upper incisors are six in number, and are placed nearly in a straight line fore-and-aft, separated from each other and from the canine by small and subequal diastemas.

The incisors are nearly equal in size, cylindrical in shape, are placed vertically in their alveoli, and are slightly recurved.

The spine of the premaxillaries is long and thin, giving very long anterior palatine foramina, as in Ruminantia. The premaxillaries, in position, are close together, showing a very narrow muzzle.

The maxillaries have two well-developed canines, with long, laterally compressed fangs, recurved and becoming circular in section at the crowns. The diastema between the canine and the first premolar is proportionate to that in Procamelus.

Ithygrammodon cameloides, sp. nov.

Specific characteristics.

The premaxillaries (See Plate X., Figs. 1-4) are long, with a wide, strongly upward curving process; from the premaxillary symphysis to the third incisor, they increase in thickness; then narrow slightly to the maxillary symphysis. The upper margin of the process is smooth and narrow, curving very gradually upwards; while the lower margin runs more abruptly up, giving to it a strongly pointed curvature.

The outer surfaces are smooth and slightly convex; the inner are marked by a wide deep canal running from just behind the premaxillary symphysis into the maxilla.

The great size of this groove may perhaps be a generic characteristic; as it is much larger than in Protolabis (Plate X., Fig. 6), the modern camels, or any of the other ungulates, recent or fossil, that have come under our notice.

The premaxillaries do not co-ossify, as the spines are smooth on their inner side. Just outside the spines, and in front of the deep groove, the bones are marked by numerous small pits for the attachment of the muscles of the upper lip.

The portion of the maxilla in which the first premolar is situated, shows that the maxillaries swell considerably on their alveolar border; for the canine then narrow gradually for the first premolar diastema; then swell for its reception, and become very thin in the diastema between the first and second premolars.

The Teeth.—I. ³/_?, C. ¹/_?, Pm. ¹⁺/_?, M. ^?/_?.

The incisors have quite deep fangs, are cylindrical in shape, vertical in position, and have no basal ridge.

The crown of the first is worn perfectly smooth, flat across the top, unlike the wearing exhibited in any modern form that has come under our notice. The others have their crowns broken, but indicate a nearly subequal series, with the last slightly the largest.

The canines have long, recurved compressed fangs, with rather short crowns, which are circular in section. They are much larger than in Procamelus or Protolabis of Pliocene, or the modern camels. (See X., Figs. 3, 5, 6, 7.)

The first premolar is the only one of the molar series preserved; its crown is chipped considerably, but nevertheless shows that it had one fang and no basal ridge. It resembles the incisors in form, but is slightly narrower and longer.

The genus Ithygrammodon has been placed for the present under the genera incertæ sedis; but its chief features point to an unmistakable affinity with the ruminants. It is probable that Ithygrammodon was the representative and the ancestor, in the eocene age, of that type of ungulates of which the camel and llama are the modern forms.

The line of descent of the camels was first indicated in the Proceedings of the Philadelphia Academy, 1875, p. 262; then in the Bulletin No. I, 1874, p. 25, of the U.S. Geological Surveys of the Territories, (Government Report of Explorations of 1873, pp. 498-500); and lastly in U. S. Geographical Surveys of Territories, vol. iv., pp. 341-44.

In the above writings, Professor Cope traces the development of the modern camels from the miocene genus Poëbrotherium, Leidy, showing the modifications found in Procamelus, Leidy, and Protolabis, Cope, of the succeeding age.

From the last work (Wheeler's Survey, vol. iv., p. 342), we will quote at length:

"The evolution of the existing types of Camelidæ is a good illustration of the operation of the laws of acceleration and retardation. In evidence of this we may follow the growth of the foot and dentition of the most specialized, and therefore the terminal genus of the series, the American Auchenia ----. It is well known since the time of Goodsir, that the embryos of ruminants exhibit a series of superior incisor teeth which disappear early. It is probable, but not certain, that in the miocene genus Poëbrotherium, as in various contemporary selenodont artiodactyla, the superior incisors persisted. I have, however, discovered that these teeth persisted in the Loup-Fork genus Protolabis during adult life. I have found that one (the second) of these teeth in Procamelus occidentalis persisted, without being protruded from the alveolus, until nearly adult age.

"In genera (e.g., the bunodont artiodactyla) where the incisors are normally developed, they appear at about the same time with the other teeth, and continue to develop to functional completeness. This development is retarded in Protolabis, since they are not so matured as to remain fixed throughout life in their alveoli.

"In Procamelus, the retardation is still greater, since the first incisor reaches very small dimensions, and is, with its alveolus, early removed; while the second incisor only grows large enough, and for a sufficient time, to occupy a shallow alveolus, without extending beyond it. In the first incisor the process of retardation has reached its necessary termination, i.e., atrophy or extinction; while in the existing Camelidæ the second incisor also has disappeared the same way. In ruminants other than Camelidæ, the third or external incisor has undergone the same process, while in the Bovidæ the canines also have been retarded in development, down to atrophy."

Professor Cope continuing the discussion of the teeth of these types, generalizes: "From these and many analogous cases, the general law may be deduced that identical modifications of structure, constituting evolution of types, have supervenea on distinct lines of descent?"

This summary of Professor Cope's conclusions has been necessary to show clearly in this series the probable place of the eocene genus.

From the shape of the premaxillaries and maxillaries, and the form and position of the teeth the camelline affinities have been indicated. Future research must be relied on to throw light upon the structure of the limbs and the axial skeleton; as it is from these that the conclusive proof must be deduced that Ithygrammodon is the ancestor of the camels, and that Poëbrotherium is the second link in the chain, instead of the progenitor.

Measurements of Ithygrammodon cameloides with Protolabis.

AMBLYPODA.

(Cope. Wheeler's Survey, vol. iv., pp. 178 et ss.)

"Mammalia, with small cerebral hemispheres which leave the olfactory lobes and cerebellum exposed. The feet short and plantigrade, with numerous (in the known genera, five) digits, terminating in flat, hoof-bearing, ungual phalanges. The seven bones of the carpus distinct, the unciform articulating with both lunar and cuneiform. The astragalus flat, without trochlear surface, and attached to the tibia with very little freedom of movement; its distal extremity divided into two facets, one for the navicular, and the other more or less for the cuboid. Molars inserted with enamel, with wide crowns and transverse crests. A post-glenoid process."

This order falls naturally into two sub-orders:

This sub-order, Dinocerata, includes at present three distinct genera, Uintatherium, Leidy, Dinoceras, Marsh, and Loxolophodon, Cope (also probably Megacerops, Leidy). These three genera, in addition to the characters above given, are marked by the possession of two or more osseous projections from the upper surface of the head; and of these the posterior pair, developed from the parietals, are the largest.

Synopsis of genera of

DINOCERATA.

Synopsis of species of

UINTATHERIUM.

UINTATHERIUM.

Skull broader proportionally than in the other genera of the order, ridged and possessing several concavities on the upper surface; zygomas slender and but little curved; temporal fossæ comparatively short and have well-defined superciliary margins. The cervical vertebræ are rather long; the sacrum has four vertebræ; and the tail is quite long, very flat and broad. The tibia has its proximal face divided by a prominent ridge into two parts. The dental formula for the upper jaw is: I. 0, C. 1, M. 6. The molars are small, and increase from first to sixth. The last is much the largest; in it the anterior lobe considerably exceeds the posterior in size, and there may or may not be a tubercle at the entrance of the valley between the lobes; but there is always one developed from the cingulum at the posterior part of the tooth.

Uintatherium leidianum, sp. nov.[1]

Established upon a head and nearly perfect skeleton of one individual, and parts of two more.

In this species, as in most others of the sub-order, the nasals are of immense length and thickness; they overhang the anterior nares, and project considerably beyond the premaxillaries; they form more than half of the entire length of the skull, articulating with the frontals somewhat behind the orbits; the median suture is distinct throughout. Instead of having an expanding shovel-shaped forward projection, as in Loxolophodon, they narrow from the median osseous projections (horn cores?) anteriorly. Above the muzzle they are strongly curved from side to side, on their upper surface forming a continuous arch; on the under surface of each bone is a deep concavity, separated from its fellow by the sutural ridge, which disappears forwards, near the extremity. On the upper surface of each nasal, near the forward end, is a large osseous tuberosity which is directed forwards and outwards; these processes are much smaller than the corresponding ones in the Loxolophodon, while they are larger, of different shape and direction from those in Dinoceras. In this species they are divided throughout by a deep median groove, which anteriorly becomes a fissure, and separates the extremities of the nasals completely. In front of these tuberosities, the nasals taper very rapidly, and end in sharp points which project downwards and forwards; this portion of the bones is shorter, sharper, and projects more decidedly downwards than in Dinoceras. Above and slightly behind the sockets of the cranium, the nasals and maxillaries give rise to the median pair of osseous projections. These are very large, subtrihedral at the base, and project upwards, outwards, and strongly forwards. They differ from the corresponding processes in Dinoceras, in their forward projection, in being longer and more everted, and in their approach to each other at the base. They taper quite regularly from base to tip, but do not come to a point; a section here would be subcircular. They are not so long and are not knobbed at the ends as in Loxolophodon. Their inner curve is convex, the outer concave; converging below at an angle of 90°, they are united by a strong ridge, which is raised decidedly above the surface of the nasals. At their posterior insertion they touch the frontals, lachrymals, and malars. It is probable that the nasals send up processes on the internal and posterior side of these median projections; the sutures between these and the maxillary portions are marked by slight ridges.

Frontals.—In our specimen it is very difficult to determine the exact relations of these bones. The nasals narrow at their posterior end, and articulate with the frontals by a V-shaped suture which thus encloses them on each side. It is probable that the frontals run to some point between the posterior or parietal projections. If we have discovered the suture, they overlap the parietals, sending up V-shaped processes, which join the sides of the posterior cranial projections, and form a deep concavity with them. They constitute a large part of the interior wall of the orbit, but have no postorbital processes; the superciliary ridge is very strong, and gradually rises into a large rounded crest, which joins the posterior projection. The frontal eminences are large, and situated immediately over the lachrymals. The upper surface of the frontals is marked by two high ridges, which probably represent the divided parts of the sagittal crest; they rise from the anterior margin of the bones, and are strongest above the frontal eminences. They converge, and then diverging, together forming an X-like curve (without crossing, however), gradually disappear posteriorly. These various ridges form four deep concavities on the upper surface of the skull: (1) The largest already mentioned, between the parietals and the posterior part of the frontals; (2) that marked by the suture between the nasals and frontal, not so large; (3 and 4) two much smaller lateral ridges between the sagittal and superciliary ridges. This great irregularity of the upper surface of the head is, so far as is yet known, peculiar to Uintatherium.

Parietals.—The peculiar feature of these bones is the huge pair of projections (the so-called "posterior horn cores") to which they give rise. The outer margin of these processes is nearly straight; the inner margin for its upper two thirds is also straight, the lower third curving in a high strong ridge to meet its fellow. The anterior face is rounded and produced into the long crest of the frontals, while the posterior face is flattened and produced into a similar but shorter and higher crest which joins the supra-occipital. At the base of these processes their greatest diameter is fore-and-aft, while at the top it is transverse. They project upwards and outwards, but are not curved as in Dinoceras, nor are their upper borders so much arched. Their greatest diameter at top is at right angles to the corresponding measurement in this last-named genus. Behind the ridge which connects these projections, the parietals curve sharply upward to a high occipital crest; between which and the projections they form a deep basin whose floor is raised above that formed by the frontals immediately anterior to it. This arrangement differs from that of both Loxolophodon and Dinoceras, especially from the latter. The temporal fossæ are of great length, deeply concave, but rather low from above downwards, formed almost entirely by the parietals, and bounded posteriorly by a sharp outward curve of these bones. They are of about the same proportionate length, but higher and very much deeper than in Loxolophodon; and are not nearly so long as in Dinoceras, in which genus the parietals expand far behind the horn-like processes. In our specimen the posterior part of the fossa is pierced by numerous small venous foramina, and corresponds precisely in every way, save that of size, to the specimen figured by Dr. Leidy. (Cont. to Extinct Vert. Fauna, Plate XXVI., Fig. 1.)

The Squamosals are large and heavy but encroach little upon the temporal fossæ; they are situated directly below the large parietal projections. The glenoid cavity is transverse and straight in this direction, broad and shallow, with no internal process; and the post-glenoid process is long and massive. The zygomatic process is short, stout, high, and trihedral, with a strongly arched upper margin. The articulation with the malar is by a straight, flat face, and anchylosis of the two never takes place. The anterior termination is pointed, and the outer margin is rounded.

The Malars form none of the face. They are long, slender, curved downwards and backwards, and but little outwards. They are longer, straighter, and less curved outwards than in either Loxolophodon or Dinoceras; they do not present the sharp angle in the lower margin shown in the latter genus, nor are they so extensively overlapped by the zygomatic processes of the squamosal. Posterior to the molar series, the malars are greatly compressed and very slender, but at the junction with the maxillaries they become much wider and thicker. There is no trace of a postorbital process; and the projections from the under surface of the bone at their junction with the squamosals, so prominent in Dinoceras, are here rudimentary or absent. As a whole, the zygomatic arch is very long, slender, simple, curved upwards and very slightly outwards, so slightly that it is completely overhung by the superciliary ridge and frontal crest.

The Lachrymals are unusually large, and form the anterior part of the orbit; they encroach considerably upon the face, and articulate with the superciliary ridges above. The lachrymal foramen is very large.

The Maxillaries are of great length, being nearly as long as the nasals. They extend somewhat beyond the last molar; but the suture between them and the pterygoids is very obscure. There is a very long diastema between the canine and molar series, and the lower margin is arched upwards; above this, between the orbit and the socket of the canine, there is a large, deep fossa, but we can discover no infraorbital foramina. The sockets of the canines are very large and prominent, they curve upwards and backwards to the base of the median cranial projections, but these are not excavated to receive them, as is the case in Dinoceras. The most peculiar feature of the maxillary bones is the pair of large horn-like projections, to which, in conjunction with the nasals, they give rise. These have already been described, and it only remains to add that they correspond in position to those of Dinoceras, and are therefore much further forward than those of Loxolophodon.

The palatine plates of the maxillaries are long and very narrow; they are concave transversely, and are separated from each other by a high median ridge. The posterior palatine foramina are small.

The Palatines are very short, narrow, and concave, and are separated by a ridge. They are considerably excavated on the posterior border, in this respect differing from Loxolophodon.

The Pterygoids and the pterygoid plates of the alisphenoid are compressed; the former join the alveolar borders of the maxillary, which are produced somewhat beyond the last molars. The alisphenoid canal is very large, but rather short.

The Premaxillaries are of very peculiar shape, somewhat like a u, with one side—the lower—the longer. The upper portion articulates with the nasals, running along the narial opening to about three inches from the angle; the free portion is short, slender, and tapering; it is curved downwards and slightly inwards; the premaxillaries do not meet, leaving the incisive foramen unenclosed, and are edentulous. At the end of the upper portion there are prominent processes for the attachment of the muscles of the proboscis, which probably resembled that of the tapir.

The anterior narial opening is very large, but is considerably concealed by the overhanging of the nasals. There is no osseous septum between the nostrils. The posterior nares are much smaller, being especially contracted in width; it is divided above by the Vomer, which is very short, and does not reach the palatines. The nasal cavity thus formed is long, straight, and gradually narrows backwards.

Of the Mandible we have but a small portion of the right ramus, comprising the part opposed to the upper canine, from the dental canal downwards. The jaw at this place has a large downward-projecting process, very similar to that figured by Marsh in his plate of Dinoceras laticeps. This process curves slightly outward, and has its external side convex in both directions, and its internal side convex fore-and-aft, concave from above downwards. The posterior mental foramen, which is the only one preserved in our specimen, corresponds in size and position with Dinoceras laticeps. The only difference between the two is a very slight one: in D. the curve of the posterior margin of the process is convex; in Uintatherium it is concave; the angle which the process makes with the jaw is also greater.

Cranial Measurements.

Teeth.—These are peculiar for their small size compared with the bulk of the animal; for their slender fangs, and for the distance through which these are exposed.

Upper Jaw.—Formula: I. ⁰/₀, C. ¹/₁, Pm. ³/₃, M. ³/₃.

Canines.—The sockets of these teeth have been already described: the fangs are very long, somewhat longer than the crown, and the teeth apparently grew from permanent pulps. The crown is long, compressed and recurved; the posterior margin is sharp, and exhibits some indications of a slight serration.

Premolars.—The first premolar is not preserved in our specimen. The second and third are subequal, and of about the same conformation. They are implanted by three fangs, one internal and two external. The crown is subcircular at the base. The basal ridge is large, and completely surrounds the crown; above this the crown becomes somewhat trihedral: and is composed of two pyramidal lobes, which meet internally and externally, and are divided by a valley. The summits of the lobes form transverse grinding ridges; the anterior one is the lower, and is crescent-shaped, while the posterior is straight; the anterior ridges throughout the series are much the most worn. The basal ridge is not indented on the outer side at the opening of the valley as in Dinoceras.

Molars.—The true molars increase in size regularly backwards. The first molar is larger than the last premolar; it is worn down nearly to the basal ridge, but shows a small internal accessory tubercle on the back part of the posterior lobe. The extremities of the lobes rise into points, and their summits are transversely concave. The number and arrangement of the fangs in the first and second molars is the same as that of the premolars; in the last molar there are but two, which are long and wide, and extend the whole breadth of the crown. This tooth is by far the largest of the molar series; it is ovoidal in shape, with the apex at the outer point of the forward lobe; it is proportionally broader than in Uintatherium robustum. The basal ridge is strong, and extends around the entire tooth, with some irregularities of outline at the sides. This tooth, like the others, consists of two pyramidal lobes separated by a valley, which in this case is wide and deep. The summit of the anterior lobe is the longer, and extends obliquely across the crown, while the posterior is nearly straight. The free ends and junction of the lobes are prolonged into points, which give a tripodal character to the crown. The outer point of the anterior lobe is the longest of the three, but the difference is not nearly so marked as in U. robustum; the other two are of about the same height; but the external is spinous in character, while the internal is stout and obtuse. The anterior slope of each lobe is steep, while the posterior is long and gentle. As in U. robustum, there is a small rounded tubercle on the inner side of the posterior basal ridge; but there is no tubercle occupying the entrance of the triangular valley between the lobes. The enamel of all the molars is smooth.

Teeth Measurements.

Lower Jaw.—Dr. Leidy has very kindly sent us some of the lower molars of an undescribed Uintatherium, which may belong to this species. They consist of the anterior lobe of the last molar, and second and third premolars entire. They all have nearly the same conformation, consisting of three acute pyramidal lobes, of which the anterior is very much the highest and broadest; the median lobe is partly separated from the anterior by a valley which opens outwards, while the valley between the posterior and median lobes passes completely across the tooth. The highest point of the crown is the inner end of the anterior lobe; but the difference between this and the outer end, very great in the molars, becomes slight in the premolars. The inner end has an accessory tubercle just at the lip of the anterior lobe, and there is another on the anterior basal ridge. This latter feature would seem to distinguish it from U. robustum.

The basal ridge is distinct in front and behind, indistinct on the external side, and entirely absent from the internal.

Measurements.

(TEETH, LOWER JAW.)

Vertebræ.

Cervical region (Plate VI., Fig. 1).—(Only one preserved, probably fifth or sixth.)

The centrum is short, compared with the dorsals, but is much longer than the cervical centra of the Proboscidea; it is broad and depressed, oval in form, and slightly opisthocœlous. The zygapophyses are developed upon tuberous projections of the pedicles; they are large, flat, and in the same plane with each other. The diapophyses are very slender and short, and but slightly heavier than the parapophyses, with which they unite, enclosing a large vertebraterial canal. At the anterior margin of the parapophysis, a small pointed process projects downward.

The pedicles are low and very heavy, bounding a narrow neural canal.

The epiphyses are not so completely ossified as they are in the dorsal region.

Dorso-lumbar region (Plate VI., Figs. 2, 3, 4, 5),—(Description based upon nine dorsals and two lumbar vertebræ.)

The centra are large, subtriangular, and slightly compressed; they are opisthocœlous, but less so than in the Proboscidea. They increase in size slowly but regularly from before backwards. In the middle dorsal region they are excessively expanded laterally for the posterior-costal attachments; but become less broad and higher as they recede in the series. In the middle of the series the centra are marked by a prominent hypophysial keel.

In the anterior dorsals the costal surfaces are developed almost entirely upon the pedicles, and are two in number. These are very large in about the sixth and seventh, and meet. They decrease in size as they recede. In the middle region the posterior costal surfaces are small, lozenge-shaped facets, and are developed upon thin lateral projections of the centra.

The centrum of the last dorsal is heavy, resembling the centra of the lumbar, from which it is distinguished by a single pair of small costal surfaces, developed, half on the pedicles and half on the centrum.

The neural spines are markedly smaller than those of the Proboscidea and Rhinoceros; in the anterior part of the dorsal region the spines have the same angle as the corresponding ones in Mastodon. The spines of the twelfth and thirteenth (approximately) are much expanded at the ends and bifid; in the last dorsals they are wide, straight, and very thin; in the lumbar region they are short, stout, tuberous, and stand almost at right angles to the axis of the vertebræ. The laminæ in the anterior part of the series are long and thin, decreasing in length but increasing in thickness from before backwards. The zygapophyses in the fore part of the dorsals are mere flat facets on the laminæ; they increase in size and become characteristic in the posterior part of the dorsal region. The pre-zygapophyses of the last lumbar vertebra are very large, deeply concave, and parallel with the axis of the column. The metapophyses appear in the middle of the series, and regularly increase in size to the last lumbar. From their appearance the pre-zygapophyses are developed upon them.

The transverse processes present the most unique feature of the vertebral column. In the anterior region they are long, wide, and rugose, and in the same plane with the laminæ; they send directly out wide, downward-curving projections. In the middle dorsal region the transverse processes lose these thin projections, and become heavy and subtrihedral, with a smooth facet on their lower face for articulation with the tubercle of the rib; their upper face is here a little twisted from the plane of the laminæ. In the posterior dorsal region the transverse processes lose the facet for the articulation with the tubercle of the rib, and become short and very thin, pointing slightly backwards.

The lumbar transverse differ from those of the last dorsal in having a median transverse ridge on their posterior face, and are directed more vertically out.

The pedicles throughout most of the dorsal series are short, heavy, and deeply notched behind. In the posterior dorsal region they become longer and more slender, resembling almost exactly the pedicles of the lumbar region.

Sacral region.—The sacrum is composed of four vertebræ, three true and one pseudo-sacral. The centra are extremely depressed, and rapidly decrease in width and thickness from before backwards. The first is shorter than the last lumbar, but much longer than the other sacrals, which are subequal.

The face of the first is elliptical, and is nearly three times the diameter of the fourth, longitudinally. The metapophyses are exceedingly large and tuberous on the first; with wide, deeply concave pre-zygapophyses developed upon them; in the other three the metapophyses are rudimentary.

The transverse processes are long and wide in the first three; widest in the first but thickest in the second; long and thin in the fourth. The pleuropophysial segments of the true sacrals are very heavy.

The foramina enclosed by the transverse processes are large; and have, on the internal side, their long diameter obliquely transverse to the axis of the sacrum.

The inferior faces of the centra are slightly concave in the first three. The first and fourth have slight hypophysial keels.

The neural canal is very wide and depressed in the first, but decreases rapidly backwards, becoming extremely small in the last.

Caudal region (Plate VI., Fig. 6).—(Description based upon the first four.)

The centra of the caudal vertebræ are rather long, narrow, and greatly depressed in the middle; they decrease in size gradually backward. The pedicles and laminæ are short and thin, enclosing a small neural canal. The neural spines point directly back, being almost parallel with the axis of the centrum. They are slender and tuberous at the extremity. The transverse processes are very long, wide, and thin, thickened somewhat at the ends, and project directly out; they decrease in size backwards. From the persistence of the neural canal, and from its comparative size, it would seem as if the tail was considerably larger than that of the elephant.

Measurements of Vertebræ.

Cervical region (Plate VI., Fig. 11).

Dorsal region (Plate VI., Figs. 2 and 3).

Lumbar region—Last lumbar (Plate VI., Figs. 4 and 5).

Sacral region.

Caudal region (Fig. 6, Plate VI.).

Ribs.—Description based upon two perfect ones, and parts of four more.

The ribs, as in Dinoceras, "resemble very much those of the Mastodon." The capitulum has two convex facets, separated by a narrow groove. The sternal end in one is very much expanded, but less so in the other. The tubercle is small, and is situated upon the prolongation of the lamelliform process for the intercostal muscles. The angle is much sharper than in the corresponding ribs of Mastodon.

Measurements of Ribs.

Scapula (Plate VIII., Fig. 1).—The scapula is subtriangular, with the pre-scapular border as base, and the apex half way up the postscapular.

On the external side the pre-scapular fossa is concave antero-posteriorly, and plane in the direction of its length; it is very thin in the middle, but becomes heavier at the lateral border. The postscapular fossa is larger and less concave; both fossæ, as they approach the upper border, gain greatly in thickness.

The spine rises from the supra-scapular border, and extends to within an inch from the glenoid cavity. It is decidedly antroverted; thickest and highest near glenoid cavity; thinnest and lowest in the middle, expanding again at its upper extremity; its acromio-scapular notch is long and shallow. The acromion is rudimentary.

The coracoid process is a low, rugose tuberosity; the coraco-scapular notch is short and low.

The internal surface of the scapula has a large, smooth median ridge, extending the whole length of the bone, separating it into slightly concave fossæ. About two thirds up, the median ridge sends obliquely up two branches, forming a V, superimposed upon the main ridge; but these disappear before they reach the upper extremity. The whole inner surface is curved outward, presenting a concave appearance throughout.

The glenoid cavity is deep, ovoid in form, with its greatest diameter fore-and-aft, and its smaller end behind.

The resemblance between the scapula of Uintatherium and that of the Proboscidea is more closely marked than in any other corresponding bones. The chief points of similarity are: first, the subtriangular shape; second, the same relative proportions between the fossæ; third, the antroversion of the spine; fourth, the glenoid cavity looking directly down. The marked points of difference may be summed up as, first, in the dissimilar proportion of the glenoid cavity; second, the great thickening of the spine at its upper and lower extremities; third, in the high, shallow acromio-scapular notch; fourth, in the longitudinal concavity of the internal surface.

Measurements.

The Humerus (Plate VII., Fig. 1).—The humerus is short, but excessively stout, twisted slightly on its axis; it decreases gradually in size downward, with the smallest diameter about two inches below the end of the deltoid ridge. The head is large, hemispherical, and sessile, projecting very slightly out of the axis of its shaft. The great tuberosity is heavy, but not high, and is separated from the low lesser tuberosity by a shallow bicipital groove.

The trochleæ are very nearly equal in size, directed obliquely to the axis of the shaft, and are separated by a narrow groove, which runs from the supra-trochlear fossa down and in, then back and up to the anconeal fossa. The condylar tuberosities are large and rugose; the external is the greater, and is directed antero-posteriorly. The deltoid ridge is long and heavy, and extends nearly two thirds down the shaft, branching out into two forks near its end. The supinator ridge is short and rudimentary, differing in this respect entirely from the great development found on the humerus of the Proboscidea.

The supra-condylar fossa is small, subcircular in form, and very deep, it has the peculiarity of being placed above the external condyle alone. The anconeal fossa is median in position and quite deep.

Measurements of Humerus.

The Ulna (Plate VII.. Fig. 2).—The ulna is long, heavy at both ends, with a slender trihedral shaft that curves forward, and decreases in size as it approaches the distal end. The proximal end has its articular face for the humerus divided into three facets, the vertical being long and extremely convex; the horizontal has the pre-axial face the longer. The olecranon is very massive and rugose, sending up on its external side a high, pointed projection; while on the internal side, the olecranon becomes compressed and projects inward; these tuberosities are separated by a wide, shallow groove for the tendon. The distal articular face is large, single, and is concave antero-posteriorly, and convex laterally. The styloid process is very heavy.

Measurements of Ulna.

Pelvis.—The ilia are greatly expanded laterally, with the iliac surface concave and the gluteal surface nearly flat; thin in the middle, they increase in thickness near the borders. The crests curve regularly, and project above and beyond the sacrum, but do not bend over the acetabula. The acetabular borders are only slightly concave; the prominence for the attachment of the rectus muscle (extensor) is low and V-shaped. The direct internal surfaces of the ilia, comprised between the pubic and ischiatic borders, become deeply concave below the ischiatic portion of the acetabulum. The sacral surfaces are wide and triangular in shape; above they project beyond the sacral spines.

The ilio-lumbar angle is about 110°. The ischium is, short; a section of it, as it leaves the acetabulum, is subtriangular; but immediately it becomes flattened fore-and-aft throughout its plane of 90° to the axis of ilium; then verging toward the pubic symphysis it becomes small. The tuberosity of the ischium is small and directed up.

The pubis, as it leaves the acetabulum, is sub-cylindrical, after this it is flattened in the same plane with the ischium. The bone as a whole is short, thin, and slightly curved on its own axis; its smallest part makes up the pubic symphysis, which is short. The thyroid foramen is a large oval, with its long diameter parallel to the axis of the ischium.

The acetabulum is large, subcircular, and deep, with prominent borders; especially the iliac, which is produced on its external extremity into a point; the ischiatic is deeply notched. From the wide ligamentous pit in the centre there runs a deep groove part way down the antero-external side of the ischium.

The anterior opening of the pelvis is a wide oval, with its longest diameter transverse.

Measurements of Pelvis.

The Femur (Plate VIII., Fig. 4).—-The femur is short, with a small oval head, strongly compressed fore-and-aft. It is less out of the axis of shaft than in Proboscidea, and has no pit for the ligamentum teres. The shaft is straight and simple, much compressed transversely at the proximal extremity, becoming sub-cylindrical below.

The great trochanter is heavy, rugose, and strongly recurved; it is separated from the shaft by a wide and deep digital fossa. The second trochanter is a mere rudimental tuberosity. The condyles are nearly of an equal size, very convex, and are divided by a deep popliteal groove.

The condylar tuberosities are low, the internal sends obliquely a ridge three inches long, up and across the axis of the shaft at an angle of 45°, that forms the upper boundary of the popliteal space.

The front part of the trochlear faces for the patella are gone in our specimen.

Measurements of Femur.

The Tibia (Plate VIII., Fig. 2).—The tibia is short, straight, and simple, slender in the middle, but much expanded at the extremities. The proximal end is especially massive, with deeply concave articular faces; the internal cotylus is the larger, and has its greatest diameter fore-and-aft, directly at right angles to the greatest diameter of the external.

The cotyli are separated by a smooth ridge, that is highest at its posterior termination. The tuberosity is high and massive, with a wide depression on its top for the reception of the ligament of the patella; the sides are deeply pitted by venous foramina. A section of the shaft at the lower part of the tuberosity would be subtriangular; while below it becomes more cylindrical, as it becomes smaller, reaching its shortest diameter about three fifths down the shaft. Below this it expands and forms the large subcircular distal end. The articular face is concave, with a slight, smooth ridge running fore-and-aft on its internal side. The malleolus is broad and low.

The fibula is distinct, but very slender; its proximal end has a small circular face, which articulates with the tibia on a facet developed on the bottom of a projection of the tuberosity; the distal end is much flattened, but enters into the ankle-joint.

Measurements of the Tibia.