CONTRIBUTIONS FROM THE MUSEUM OF GEOLOGY AND ARCHEOLOGY
OF PRINCETON COLLEGE.

No. 1.

Palæontological Report

OF

The Princeton Scientific Expedition
of 1877.

BY

HENRY F. OSBORN,
WM. B. SCOTT,
FRANCIS SPEIR, Jr.

September 1, 1878.

New York:
S. W. GREEN, PRINTER, Nos. 16 AND 18 JACOB STREET.
1878.

Palæontological Division.

PROF. JOSEPH KARGÈ,
ROLLIN H. LYNDE,
HENRY F. OSBORN,
JOTHAM POTTER,
WM. B. SCOTT,
FRANCIS SPEIR, Jr.

To the Honorable the President and Board of Trustees of
the College of New Jersey:

Gentlemen:

I have the honor to transmit herewith the Palæontological Report of the College Scientific Expedition of 1877.

The fossils collected by the Palæontological party, and deposited by the chief of the expedition in the Geological Museum, consisted of two sets, one numbering some two thousand specimens of fossil plants and insects from the tertiary beds of Central Colorado, the other of a considerable series of fossil vertebrates, mostly mammals, from the tertiary beds of Wyoming Territory, around Fort Bridger. This last collection has been studied and worked out with unabated zeal and diligence by the three post-graduate members of the Palæontological party, Messrs. H. Osborn, W. Scott, and F. Speir, who devoted most of the time of their course to this special work, with what success this Report will show.

It will be a source of gratification to the generous friends of the College, who furnished means for the Scientific Expedition of 1877, that it not only enriched our Museum to so great an extent, but did more still by fostering in our College a thorough study of Palæontology, which could not have been undertaken without such means as these thus placed at the disposal of our students.

The fossil insects and plants have been intrusted to the hands of the best specialists for determination. Dr. S. Scudder, of Cambridge, Mass., has kindly consented to revise the insects. Prof. G. L. Lequereux the plants.

Very Respectfully,
A. GUYOT,
Director of the E. M. Museum of Geology and Archæology.

Princeton, June 1, 1878.

Sir: We transmit herewith our report upon the Palæontological collections made by the Princeton party in the summer of 1877.

The following persons constituted the Palæontological division: Rollin H. Lynde, Henry F. Osborn, Jotham Potter, Wm. B. Scott, Francis Speir, Jr. The division remained in Colorado from the first of July until the first of August, when, under the direction of Professor Kargè, it left the main party, and passed the month of August in Wyoming, returning in the first part of September. The Colorado collections were mostly made in the (probable) Miocene beds near Florissant, and in the beds near the Garden of the Gods, variously referred to the Dakota and Wealden groups. In Wyoming, with Fort Bridger as a base of explorations, the division was wholly occupied in the Bridger series, camping successively on Smith's Fork, Henry's Fork, and Dry Creek, and exploring the beds adjacent.

It has been our endeavor, in confining our attention to the remains of vertebrated animals collected during the trip, not merely to catalogue the direct results, but, by the aid of fresh materials, to supplement the work of others. For, of the descriptions and data of the Bridger Eocene Fauna published up to the present time, we find that even those which have been most accurately prepared are lacking in important details; and that, owing to imperfect materials, large gaps yet remain in our knowledge of genera and species named and classed years ago. Although such supplementary work may appear, at first sight, tedious and ill-directed, we are confident that in the end it will prove of some value to science, and that it is therefore well worthy of our effort. While our field work did not extend beyond a region previously well explored, we obtained material by means of which we are able to add a number of new fossils to the Eocene Fauna of the Bridger group.

In the preparation of this report we have experienced much difficulty in assigning some of our specimens to their proper genera and species. For, while we have desired to respect the classifications made by others, we have in many cases found it impossible to do so, owing to uncharacteristic definition, which, without doubt, has been unavoidable. In all cases of uncertainty, we have adopted the classification which appeared to be the best established. This, in short, has proved the only available course.

The drawings have been executed with much care as to accuracy of proportion and outline. They are, with one exception, the work of a member of the party; and they are inserted simply to illustrate certain parts of the context, which would be unsatisfactory without reference to figures of the kind.

Now that the present work is ready for the press, we are very sensible that it must contain errors which, while they have escaped our notice, will be readily detected by eyes more experienced. These, we trust, will be excused, when it is remembered that we are just entering a field which others have explored for years; and opening a work which Princeton, with her many other lines of study, has never hitherto attempted.

We lake this opportunity to return our most hearty thanks to General Flint, to Judge and Dr. Carter, to Mr. Hamilton, and other officers and residents at Fort Bridger, who, by their good will and liberal assistance, contributed largely to our success. Our gratitude is also due to Professors Leidy and Cope for their generous aid, both in the way of advice and of material put in our hands for comparison.

The following pages do not embrace descriptions of the entire collections made by the Princeton party last summer. The valuable specimens of fossil plants and insects have passed into other hands.

Respectfully submitted,
HENRY F. OSBORN,
WM. B. SCOTT,
FRANCIS SPEIR, Jr.

Dr. Arnold Guyot,
Director of the E. M. Museum.

INTRODUCTORY NOTE

UPON THE

GEOLOGY OF THE BRIDGER BASIN.

Fort Bridger is a government military post, situated on the high southern plateau of western Wyoming Territory, in the midst of one of the most interesting geological regions of the world.

The country on all sides was once the bottom of a great eocene lake, the water of which was probably slightly brackish. Whether this lake district had direct communication with the ocean, is undetermined as yet, but there is a possibility that it had.[1]

[1] Ichthyic fauna of the Green River shales, Hayden's Surveys, vol. iii., No. 4, p. 819.

The tributaries of the Green River, which drain this plateau, render the valleys along the edges of the streams green and wooded. Beyond this fertile strip, wide, barren plains extend, covered by a dense growth of short "sage brush" (Artemisia tridentata).

From the fragmentary debris lying scattered over the surface of the ground, it would seem as if the various streams formerly were of much greater size and volume than they now are, and that long after the eocene lakes had been drained-rivers of considerable size ploughed up the lake bottoms, excavating an immense area. The formation known as "Mauvaises Terres" rises abruptly from the valleys, and extends in a series of plateaus, one above the other, on either side.

The high land shows the effect of violent erosion in two forms; first, the irregular and jagged cones that appear upon the sides of the high benches; and, second, the isolated butte structure, rising directly out of the plain.

The bad lands of Cottonwood Creek, Henry's Fork, Dry Creek, etc., are examples of the first, and Bridger Butte the best known example of the second.

Bridger Butte, six miles to the south-west of Fort Bridger, rises to a height of over a hundred feet, and is about two miles long; its sides slope steeply up, and its level top serves as a landmark that can be seen miles away.

The stratification throughout this whole formation is nearly horizontal, and across the valley can be distinctly noted, owing to the color and appearance of the various layers.

No satisfactory explanation has been given of the causes which occasioned the removal of the waters of these lakes, nor of the agencies necessary to accomplish the great excavations that now show the former bottoms of the basins. These and like points future investigation will undoubtedly solve.

It is certain that the level of the lakes varied at different times, and also that great stretches of marshy land surrounded their borders.

The first fact is proved by the characters of the different layers of strata; the second by the fossil remains entombed. A careful study of the formation of the beds of Cottonwood Creek, at a point about fourteen miles south of Fort Bridger, yielded the following result, which will serve as an example illustrative of the regular formation in this section.

Three distinct lines of bluffs are to be noticed, the first rising to a height of one hundred and fifty feet; from these extends a plain, gently sloping south-west to the foot of the second line of cliffs; these seem high, owing to the downward slope of the plain, but they really rise only fifty feet higher than the first.

On the top of the second, but less broad than the first, extends a level plain, with a slight dip to the south-east; at the end of this, the third line rises two hundred feet above the top of the second line of bluffs, making a total height, in the series, of four hundred feet above the level of Smith's Fork.

The strata throughout are nearly horizontal, and are of different color and composition.

Specimens of the rocks and clays from this section were gathered, and submitted for analysis to Professor Cornwall, of the Scientific School at Princeton, who has kindly furnished us with the following notes:

No. 1, very friable, light greenish-gray sedimentary rocks, consisting chiefly of crystalline grains of quartz, orthoclase, and hornblende (this often in slender crystals), with a little dark mica, and irregular fragments of a light-greenish, transparent, not dichroitic mineral. The above are not perceptibly affected by hot hydrochloric acid. The whole is loosely cemented with a calcareous clay, containing considerable phosphoric acid. These rocks might result from the disintegration of a neighboring hornblendic granite.

The greatest mass of the strata is made up of this kind of rock, and it is this which gives the peculiar color to the "Mauvaises Terres."

Above this in places is found a second kind, which is a light gray indurated clay, with a slight greenish tint. It contains much fine crystalline quartz, with considerable carbonate of lime, and a little phosphoric acid.

It appears to be of similar origin with the first, but was deposited in quieter waters. This mineralogical evidence is strengthened by the fact that no remains of mammals were found in strata of this kind, but only shells regularly deposited in layers one above another.

The third kind is found in thin layers, overtopping the highest line of buttes; it consists of very fine-grained dark-brown sandstones, containing a considerable proportion of carbonate and phosphate of lime. They are hard and tough, and are mechanically deposited, and no fossils are found in them.

In No. 2 the indurated clay is often found above a coarser sandstone than No. 1, but of the same general appearance; with the exception that it contains smooth, rounded pebbles, which were deposited either on a beach or in running water. In this stratum the fossils found are separate bones, often showing marks of having been broken before they were silicified. This would prove that the lake level was changing continually.

Several skeletons of animals have been found in a standing position, with their legs slightly stretched out. They were probably mired, and, being unable to extricate themselves, died in an erect position. This fact affords evidence that extensive swamps surrounded the borders of the lake.

The snow-water and the spring rains wear deep gulleys through the lines of cliffs, and wash down fragments of bones into the dried-up water-courses. These pieces can generally be traced up to the spot from which they came, and the rest of the skeleton can so be secured.

The fossils found in the eocene of Wyoming are entirely petrified, presenting a darker appearance than their matrix; but loose bones washed out, and subjected to the influence of the sun and rain, often become bleached so as to resemble in color modern bones.

The state of preservation of the fossils differs according to the matrix in which they are found. Generally speaking, the remains found in the lower lines of buttes have been considerably distorted by pressure; while those from the highest line have suffered very little from this cause.

[MAMMALIA.]

PRIMATES.

TOMITHERIUM, Cope.

Ann. Rep. U. S. Geol. Survey of Terrs. 1872, p. 546.

"Dental formula of the inferior series: I. 2, C. 2, Pm. 4, M. 3. The last molar has an expanded heel. The third premolar consists of a cone with posterior heel. Fourth premolar exhibits, besides its principal cone, an interior lateral one and a large heel. The true molars support two anterior tubercles, of which the inner is represented by two distinct cusps in one or more of them, and the external is crescentoid in section. The posterior part of the crown is wide and concave, and bordered at its posterior angles by an obsolete tubercle on the inner, and an elevated angle on the outer side." (Palæontology, Wheeler's Survey, iv (pt. 2), p. 135.)

Tomitherium rostratum, Cope.

Loc. cit., p. 548.

This species exhibits considerable variation, both in size and proportion of the teeth. We have two specimens of it, one of which agrees exactly with the measurements given by Professor Cope, while the other is stouter, and probably belonged to an old male.

The incisors are too much broken for description. The canine has a long, stout fang, which is subcircular in section; the crown is compressed, and shows a distinct cutting edge posteriorly. It is short, and tapers rapidly; in this respect differing from Notharctus, Leidy, which has a long recurved canine. The first and second premolars are inserted each by a single fang; the third and fourth by two. The third premolar consists of a simple conical crown with a small posterior heel; and the fourth has this heel enlarged, with a small tubercle developed inside and slightly behind the principal lobe. All the premolars have striated enamel, and very feebly marked basal ridges.

The true molars are considerably larger than the premolars; the third is the longest of the series, and the second is the widest. They all seem to be inserted by two fangs. The posterior fang of the last molar is a flat quadrate, of the same size throughout, having a great fore-and-aft diameter, and apparently no nerve cavity. The mandible is strong but shallow, has a curved alveolus and lower margin, and the teeth form a curve with convexity outwards. The symphysis is short and oblique.

This genus presents a close resemblance to the modern Lemur, but at the same time shows several differences. We may give these differences in systematic order: (1) Greater number of premolars, in Tomitherium = 4, in Lemur = 2. It will be observed, however, that the first and second premolars of Tomitherium are very small and single-rooted, and that their disappearance is a comparatively slight change. (2) The canines are subcircular in section, not nearly so much compressed. (3) Greater breadth of the molars in proportion to their length. (4) Two internal cusps on the molar. (5) Much greater size of the last molar. (6) Greater depth and thickness of the jaw. (7) Greater curvature of alveolus and lower margin of ramus. (8) Longer and more oblique symphysis.

The third and fourth premolars of Tomitherium correspond almost exactly to the first and second of Lemur, but they are not quite so high and sharp. The interior tubercle of the second premolar is not so distinct in the latter genus.

In the second specimen the cusps are all low, and the crests prominent, giving the molars something of the appearance of Opisthotomus. This difference is probably sexual.

Measurements.

[1] In measurements of teeth we use the word length to mean antero-posterior diameter, and width to mean transverse diameter.

Specimen No. 1 was found at Cottonwood Creek, and specimen No, 2 at Henry's Fork, Wyoming.

HYOPSODUS, Leidy.

Pr. Ac. Nat. Sc., 1870, p. 110.

Lower teeth: I. 3, C. 1, Pm. 4, M. 3, in uninterrupted succession. Last molar has cusps in opposing pairs; the antero-internal cusp on all the molars is single; the last molar has a heel, and the last premolar has an inner cusp. The true molar cusps are all high and simple.

Hyopsodus paulus, Leidy.

Loc. cit., p. 110.

This is one of the most common fossils found in the Bridger Basin. We have numerous specimens, chiefly from Henry's Fork, exhibiting a large range of individual and sexual variation. As Dr. Leidy has already pointed out, the strength and depth of the lower jaw is extremely variable, increasing with the age of the animal; so that the most worn teeth are associated with the deepest jaws.

In addition to the jaws and teeth, (which have been very accurately described), we have a portion of a pelvis and femur, which are important as tending to confirm the reference of this genus to the Lemurs.

The pelvis is represented by the acetabulum and a small portion of ilium and ischium. They resemble the corresponding parts in the skeleton of Stenops gracilis. The acetabulum is a long oval, not subcircular, deep, and quite narrow from side to side. It appears to be directed nearly straight outwards. The ilium is narrow, has a concave gluteal surface, and a prominent acetabular border. The pubis evidently projected forward, making a right angle with the ilium; while the ischium is slender and nearly in the same plane with the ilium. The femur is thoroughly lemurine in shape. The shaft, (as much of it as is preserved), is straight and sub-cylindrical; it is not flattened even distally, but becomes very thick and trihedral in shape just above the trochlea. The trochlea is long, and rises obliquely upon the shaft; the groove is deep, and the two divisions are asymmetrical, the external somewhat the larger. The condyles are large. They are but slightly convex in either direction and project backwards, and are broadest posteriorly. The internal is the larger. They are separated by a deep but not wide popliteal groove, which does not extend into a popliteal fossa. The position and shape of the condyles are such as show that the femur must have been very oblique to the tibia, as in the other lemurs. The tuberosities, especially the internal, are very large and prominent. The whole distal end has an asymmetrical appearance, owing to the greater size of the internal condyle.

Measurements.

OLIGOTOMUS, Cope.

Ann. Rep. U. S. Geol. Survey of the Terrs., 1872, p. 607.

"Molars constructed much as in Hyopsodus and Lophiotherium, viz., with two external subtrihedral cusps which wear into crescents, the posterior connected by a low oblique ridge with the basis of the anterior cone of the inner side; the latter with two conic cusps. It differs from these genera and from Orotherium in the possession of two premolars; the inferior molars are probably six, leaving four true molars."

Oligotomus cinctus, Cope.

Loc. cit.

Represented in our collection by the penultimate lower molar, and a caudal vertebra, which has a remarkably long, slender, and simple centrum, with rudimentary metapophyses.

OPISTHOTOMUS, Cope.

Wheeler's Survey, Pal. v. iv., pt. 11, p. 152.

"The inferior lower molars do not display a bifid or double anterior cusp; and the crowns exhibit two anterior cones, and an inner cone and outer crescent posteriorly. The posterior crescent is well defined, and is continued on a narrow crest to the anterior inner tubercle. The posterior molar presents the peculiarity of a series of three cusps in one line, the median having another or lateral cusp near it."

This genus has hitherto been found only in the Wahsatch formation; but we have discovered it to be represented in the Bridger series by the species O. astutus, Cope. Our specimen consists of a part of the ramus mandibuli containing a single molar tooth.

[CARNIVORA]

SINOPA, Leidy.

A genus of small carnivorous animals, which Dr. Leidy regards as intermediate between the recent Canis and the extinct Hyænodon. Owing to the fragmentary condition of the remains found, no satisfactory generic definition has been given.

From the portion in our collection, we are able to throw some further light upon the genus, summing up the generic characteristics thus: Small carnivores, which have the last upper premolar as sectorial (thus differing from Hyænodon), the other premolars simple and conical.

The sectorial is shorter, antero-posteriorly, than the preceding tooth; has a short blade of a single lobe, and a large cusp developed from the posterior part; a cingulum surrounds the entire crown. The lower sectorial has the blade of a single lobe, and with a short heel.

Sinopa rapax, Leidy.

Proceedings of Ac. Nat. Sc., 1871, p. 115.

In addition to the molars of the lower jaw, described by Dr. Leidy we have what corresponds to the third and fourth premolars of the fox, their dental formulas being probably the same.

The third premolar is small and pointed; differing from the corresponding tooth in the fox, (1) in its being less compressed, (2) in its shorter antero-posterior diameter, (3) in the straighter and more nearly equal margins, and in (4) the absence of a posterior heel.

The tooth is inserted by two fangs, as in Canis and Hyænodon. The posterior shows a rudiment of a third, which is connate with its entire length above the alveolus. There is an indistinct cingulum around the entire crown.

The fourth premolar has a very curious shape. The blade of this tooth resembles the crown of the third, but is smaller. It is inserted by three fangs, the disposition of which is opposite to that in Canis, the internal, being on the same transverse line as the posterior external, instead of the anterior, as in Canis. From the internal fang arises a sharp cusp, which is nearly as large as the blade of the tooth, the two are connate at base. The anterior face of the crown is much worn, and there is a small anterior heel formed by the basal ridge. The cingulum is complete all around.

The maxillary does not show the outward bulge at the third premolar, which is so marked in the fox. The alveolus is straighter, and the palatine plates are comparatively thicker and flatter. The infraorbital foramen is oval, and not so much compressed as in the fox, to which it corresponds very nearly in position, though situated slightly forward as in Hyænodon.

Measurements.

These exhibit nearly the same proportionate size as in the gray fox.

Genus ——. Species ——.

Sacrum ([Plate IX.], Fig. 8).—This peculiar sacrum is composed of only one true vertebra; there may have been one or more pseudo-sacrals, but this is not certain.

The centrum is very long, strongly depressed, and straight on the inner margin, not curved as in the sacrum of most mammals. The anterior articular face is much depressed, and is one third larger than the posterior. The neural canal is low and subtriangular, resembling very much that of Canis. The pleuropophysial plates for articulation with the ilia are large and stout. The laminæ are heavy and concave on their upper side, supporting a very long, stout spine, which is retroverted and decidedly tuberous at the end.

The pedicles are deeply notched behind; and on the fore part, just inside the metapophyses, there is a deep fossa.

The chief features of this sacrum are decidedly carnivorous; but to what genus or family it should be referred we are unable to say.

It has some of the characteristics of Canis, but the length and retroversion of the spine, as well as the size of the centrum, prevent this classification. In the general form of the pleuropophysial plates it approximates to the seals; while in its angle and curvature, it partakes of the character of the Ursidæ.

The chief point of interest in this fossil centres in the fact that it was found only a few feet from the brain cast that is described below.

Measurements of Sacrum.

MEGENCEPHALON.

Megencephalon primævus. Gen. et spec. nov.

In close proximity to the pelvis of the Uintatherium Leidianum, in one of the upper beds we found an intracranial cast, separate from the bone which had enclosed it, and in such preservation as to warrant a partial determination, at least, of the type to which it belonged. Wishing to obtain as full information as the nature of the cast permitted, we put it in the hands of Dr. Spitzka, of New York, who kindly undertook an examination, and sent us the following as the result:

"Sir: The specimen submitted to me is the intracranial cast of some species of Placental Mammals. The cranium had been subject to the influences of the atmosphere, etc., for a considerable period preceding the formation of the cast, and therefore the cast reflects the sutural dislocations which occurred in consequence. The base of the brain cast it is not advisable to attempt to expose, on account of the treacherous nature of the material. The convolutions corresponding to the internal aspect of the Os temporale have not been clearly demarcated by the bone surface. The two narrow eminences on it are casts of the grooves of the middle meningeal arteries. The convolutions of the occipital surface had been well marked, but somewhat obliterated through denudation, etc. The important region bordering on each side of the median fissure, and corresponding to the fronto-parietal suture, is unfortunately as good as destroyed; and with this destruction the key to the interpretation of the specimen is lost. However, this much can be stated with absolute certainty, that the frontal region is sufficiently well preserved to state that its convolutions do not correspond to those of the brain of the tapir, rhinoceros, hippopotamus, elephant, pig, horse, hyrax, manatus, or any ruminant or cetacean.

"They also differ in important particulars from those of the Canidæ, differ less from those of the Felidæ, still less from the Ursidæ, although corresponding to none of them. The outline of the cerebral cast is found in two living animals—the marine otter and the seal. But in the seal the gyri show the transverse interrupting series of sulci, characteristic of extreme brachycephaly; and it therefore cannot belong to any animal corresponding to the seal.

"The sea otter's convolutional details are unknown to me, and I believe have not yet been studied. I therefore content myself with stating that the outline of this cast corresponds to the outline of the sea otter's cranium.

"It would help us a great deal if we could decide the existence or non-existence of a bony tentorium. The sutures of this cranium, as far as I can reconstruct them, ran as in the diagram.

"We may state definitely that this was not an ursine, feline, or canine brain, nor the brain of any terrestrial viverrine. It is an open question between an aquatic carnivore and an aquatic pachyderm; and although not placing my conclusion on an exact basis, yet, in view of the general outline, the course of the convolutions, and the course of the sutures, I incline to the former view.

"It certainly corresponds to no known brain of a living creature. In one point I was inclined to suspect it to be a pachyderm, namely, the decided asymmetry of some of the sulci, but this, by itself, is not decisive."

"Dr. Spitzka.

"308 East 123d street."

The interesting letter quoted in full above, contains as near a determination of the character of the animal to which the brain belonged, as the nature of the cast and the materials for comparison would permit. In a later report, by means of more complete comparative material, we hope to be able to reach a more satisfactory conclusion. However, as Dr. Spitzka writes, the general outline, the course of the convolutions, and the line of the sutures offer strong presumptive evidence that the cast belongs to one of the Aquatic carnivores. Not far from the brain was found a sacrum, which is described above as belonging to some carnivore, though further determination was impossible. Whether there was any connection between the two is difficult to state. The presence of an aquatic carnivore in the Bridger eocene is new to science; but, aside from this, the brain is of a much higher order than previous discoveries would lead us to expect in such an early formation.

Professor Marsh's researches have led him to form the opinion that the eocene mammals had brains of a low character; but this specimen shows that this is not true of all, if it is of most of them. The convolutions are not only numerous and well marked, but they are complicated, showing the transverse as well as the longitudinal folds. To such an extent is this true that the brain will bear comparison with the very highest modern carnivorous types.

We hope to be able to give further notes upon this interesting specimen at a later date.

[PERISSODACTYLA.]

ANCHITHERIUM.

Von Meyer, Jahrbuch für Mineralogie, 1844, p. 298.

Anchitherium —— ?

A small calcaneum and astragalus of equine type are provisionally referred to this genus until further material enables us to determine them with certainty.

The astragalus has narrow and very oblique condyles, which are more equal in size than in Orohippus; the neck is very short, the internal condyle reaching to the face for the navicular; the posterior projection of this condyle is much shorter than in that genus. The articular face for the navicular is quadrate in shape and concave; the cuboid face is very narrow. The articulation with the calcaneum is made by a narrow, convex face. When the two are in position the navicular face of the astragalus is in the same horizontal line as the cuboid face of the calcaneum, thus resembling the arrangement of the horse's tarsus rather than that of Orohippus.

The calcaneum is a short, slender bone, having the upper and lower margins convergent toward the tuberosity, and not parallel as in Orohippus. The tuberosity is especially small. The face for the cuboid is very narrow.

From the articular facets of these two bones we can see that the tarsus resembled very much that of the modern horse, with a broad, short navicular, and a narrow cuboid. The strata in which these remains were found were somewhat higher than those containing the bones of Orohippus.

Measurements.

From Henry's Fork.

OROHIPPUS, Marsh.

Am. Jour. Sc. vol iv., p. 207, third series.

Generic Characteristics.—"The crowns of the upper true molars are composed of a pair of external cusps similar to those of anchitherium. There are two corresponding inner tubercles, from which ridges extend obliquely to the anterior inner margin of the outer cusps; but the anterior ridge is divided so as to form an intermediate anterior tubercle. All the teeth preserved have a distinct basal ridge."

Species Known.—

Orohippus pumilus? Marsh.

Specimen obtained. Penultimate and third superior molars, with part of zygoma.

From Cottonwood Creek.

Orohippus major? Marsh.

Femur ([Plate IX.], Fig. 1).—The femur has a small, nearly hemispherical head, developed on a long and slender neck; the head is but slightly out of the axis of the shaft, and has a large pit for ligamentous insertion. The shaft is long, simple, and curved slightly forward. At the proximal end it is broad and flattened axially; below this it becomes expanded fore-and-aft, but it thickens greatly at the distal end, just above the condyles. The great trochanter is large and retroverted, rising above the head, with two prominences rising from it, one on top, the other back. The digital fossa is wide and deep, penetrating far into the great trochanter. The second trochanter is a small rounded ridge; the third trochanter is large and prominent, curving slightly forward. The condyles are long and narrow, projecting very far back, and are separated by a wide and deep popliteal groove. From the external condyle a low ridge runs obliquely, forming the upper border of the shallow popliteal space. The trochleæ are long, very convex, deeply grooved, and symmetrical.

Measurements of Femur.

The tibia ([Plate IX.], Fig. 3) is very long and heavy, with broad proximal articular face, the inner borders of which are prolonged upward and separated by a groove.

The shaft at the tuberosity is subtriangular, with strongly concave sides. The tuberosity is prominent, with a deep pit on its upper surface for the insertion of the ligament of the patella. The shaft below becomes sub-cylindrical, and decreases regularly in size downward. Its curvature is forward. The distal articular face is divided by a smooth ridge into two deep facets. The malleolus is long.

The fibula ([Plate IX.], Fig. 2) is distinct, straight, and very slender. The distal end is but slightly expanded, and is strongly marked by a vertical groove externally. The proximal end articulates with the overhanging portion of the proximal face of the tibia.

Measurements of Tibia.

TIBIA.

The tarsus.—The astragalus ([Plate IX.], Fig. 5) has the condyles asymmetrical and divided by a deep groove; the head is narrow, with the neck elongate. The face for the cuboid is small, and confined to the external border.

The calcaneum ([Plate IX.], Fig. 4) is long and compressed, with its upper and lower margins straight and nearly parallel; its tuberosity is large. The face for the cuboid is small.

The navicular ([Plate IX.], Fig. 6) is proportionately longer and narrower than it is in the modern horse. The internal and middle cuneiforms were probably separate.

Metatarsals ([Plate IX.], Fig. 7).—Three in number. Are very much shorter proportionately than in the modern horse. In shape they are compressed and arched forward. The distal ends are flattened vertically, arched forward, and deeply grooved in the middle.

The phalanges ([Plate IX.], Fig. 7) are very short, rather stout, and very smooth and convex above.

The ungual phalanges are very thin and crescent shaped.

Measurements.

ASTRAGALUS.

CALCANEUM.

NAVICULAR.

PHALANGES.

PALÆOSYOPS, Leidy.

Hayden's Geological Survey of Montana, 1871.—Proceedings Academy Natural Sciences, Philadelphia, 1871, p. 118.—Limnohyus.—Marsh, American Journal Science and Arts, 1872, p. 124.

Generic characters.—The dentition is full, I. 3, C. 1, Pm. 4, M. 3; the same in lower jaw. The internal cones of the superior molars isolated from the crescentoid crests. One inner tubercle on the last three premolars. One internal cone on the last superior molar. In lower jaw, true molars with four acute tubercles alternating in pairs and connected by oblique crests. The last molar adds a fifth posterior tubercle. The last premolar lacks the posterior inner tubercle. The canines are in continuity with the incisors.

A broad, triangular forehead. A wide zygoma. Long, projecting nasals. Large temporal fossæ. High sagittal crest. Prominent and nearly vertical occiput.

Palæosyops major, Leidy.

Survey of Wyoming, 1871, p. 359.—Limnohyus robustus.—Marsh, American Journal Science and Arts, 1872, p. 124.

Specific characters.—Sagittal crest short and thick. Temporal fossæ not very deep. Frontals diverge rapidly. Occipital condyles wide and low; the same is true of the foramen magnum. Meatus auditorius high and deep. Glenoid cavity shallow; no internal process. As compared with P. Paludosus, post-glenoid process not so much compressed. Occipital region comparatively higher and not so concave. The occipital condyles are more prominent. The zygomas are lighter. Head larger than Paludosus, perhaps not so large as Vallidens.

Description from (i) a head, complete, but distorted, with complete dentition, upper and lower. (2) A head somewhat crushed, and lacking some portions, with full set of upper molars and canines. (3) A great number of fragmentary specimens of different parts of the body.

Dentition.—In the upper jaw the incisors are arranged in a semi-circle as in P. paludosus; they have long fangs and short, conical crowns, with a decided basal ridge, which is very strong in the third. They increase from first to third, which is very large and pointed.

The canines have very large and long fangs (longer than the crowns). The crowns are very stout and pointed, constituting formidable weapons. In section they are nearly circular, with a distinct hinder margin and a slight recurve. A rudimentary anterior margin is sometimes present, and of the faces thus marked off, the inner is much smaller and flatter. The general direction is forwards, downwards, and slightly outwards. There is a diastema of about half an inch between the incisors and canines.

Premolars.—The first is very small, about half an inch from the canine, simple and conical, with an obscurely marked basal ridge, and two heavy fangs.

Further description of the upper dentition is unnecessary, owing to the complete work of Dr. Leidy upon the subject.

The lower incisors are placed in a semi-circular row, and are somewhat procumbent, though hardly as much so as in the tapir. They are much as in P. paludosus, but are relatively smaller, and have not such distinct basal ridges; features which indicate a departure from the carnivorous type, and a nearer approach to the herbivorous type. The lateral incisor, though the largest of the series, is not so large nor so pointed as in P. paludosus. There is no diastema.

The canines are about equal in size to those of the upper jaw, growing from stout fangs and permanent pulps. The margins of the crowns are more decided, and are smooth, leaving a very narrow and flat inner face. A feeble basal ridge. When the jaw was closed, the lower canines passed inside and in front of the upper.

The first premolar stands immediately behind the canine, is longer than that of the upper jaw, consisting of a single pointed lobe implanted by one fang, slightly recurved with a flat inner and convex outer face. Passing in front of the first upper premolar, it leaves a considerable diastema between this and the second premolar.

The mandible approximates in form that of the tapir; the lower border is less curved fore and aft, the alveolar border is slightly concave antero-posteriorly. The molars converge in front, presenting a very different arrangement from that in Sus. The ramus is heavy, contracting in depth forward, and very slightly in section. Below the alveolus, on the interior side, the ramus is greatly swollen for two thirds of its depth, to accommodate the very long and strong molar fangs.

The rami converge to the second premolar, where they expand laterally to the canine alveolus. The symphysis is shorter relatively than in P. paludosus, and the chin is regularly rounded in front. The mental foramen is below the second premolar. The dental foramen is small, very high up, and far behind the molars. The alveolar border ascends rapidly behind the last molar, expanding laterally into two ridges.

The coronoid is slender and recurved. The condyle is broad, flat behind, inclines forward internally; there is scarcely any depression between the condyle and the coronoid. The masseteric fossa is very wide and deep. The angle of the ramus is broken, but evidently has no such posterior extension as in P. paludosus. It thins out rapidly behind.

Measurements.

Among other specimens obtained at the divide between Henry's Fork and Cottonwood Creek, was a nearly complete but somewhat shattered skeleton of a Palæosyops major, found contiguous to the head of the same, parts of which are just described. Portions of these are figured on a one fourth scale in [Plate II]. The description is from a nearly perfect atlas and axis, several cervical, dorsal, and lumbar vertebræ, more or less complete, and a portion of the sacrum and pelvis, in addition to several bones of the limbs.

The atlas has a broad inferior arch, contracted antero-posteriorly, and deeply notched for the prominent odontoid of the axis. The heavy superior arch, inclosing a large and depressed neural canal, is capped by a low tuberosity. It slopes into a flat, and very broad transverse process, which thickens backwards, and is perforated by the vertebraterial canal. Forwards the transverse process is notched for the exit of the first spinal nerve. The anterior or condylar faces are continuous, deeply concave from above downwards; and slightly so from side to side; while the posterior faces are oval, nearly flat, and directed backwards and inwards.

The axis is proportionately small, with a long opisthocœlous centrum, depressed, and produced forwards into a stout conical odontoid process, and marked below by a strong hypapophysial keel, which, developed on the latter half, thickens backwards. The wide and thin pedicles inclose a high neural canal, support the posterior zygapophyses and a prominent and peculiar neural spine. The latter is flattened, broad, and recurved posteriorly, thinning rapidly forwards into a prow-shaped recurved edge. The transverse processes arising from the latter half of the centrum are undersized and widely perforated at base. The anterior faces are very broad, directed outwards, and slightly rounded from above downwards; they expand as they diverge (see Fig. 2). The post-zygapophyses are small convex faces, projecting at the base of the neural spine.

The remaining cervicals, five in number, are short, all carinate, except the seventh, and opisthocœlous; with the faces expanding only slightly beyond the body of the centrum. A peculiar feature is a small pit upon the convex anterior face, indicating either a ligamentous attachment with the antecedent vertebra, or a remnant of the notochord. This is a feature we have not noticed elsewhere. A long and heavy transverse process supports a large and widely-perforated inferior lamella. The pedicles are quite wide at base, inclosing a large neural canal. The entire upper part of the arch is unfortunately wanting.

The dorsal centra are smaller than the cervical, slightly opisthocœlous, and carinate. They are sub-cylindrical anteriorly, with decided costal surfaces before and behind; approaching the lumbar region they become subtriangular. A high neural arch supports a stout, but never very high neural spine; this projects backwards, keeled in front, expanding and deeply grooved behind. The zygapophyses are small and nearly vertical. There is a short and thick transverse process.

The lumbar vertebræ are long, decidedly opisthocœlous, becoming wider and more depressed as they approach the sacrum. The first sacral vertebra presents the same characteristics as the last lumbar, only the body of the centrum is slightly shorter. It has a very broad pleuropophysial plate. The remaining sacral vertebræ are broad, and very greatly depressed, rapidly decreasing in size. The transverse processes are slender. The very low neural spines anchylose into a long ridge. The number cannot be ascertained, owing to the fragmentary state of the sacrum. The caudals indicate a tail of not very great length; as the neural canal is small and persists in only a few of the anterior vertebræ.

Measurements of Vertebræ.

The ribs, of which great quantities of fragments remain, were slender and not of very great width.

The femur, which is figured in [Plate I.], has a small head, supported by a short neck. The shaft, very broad below the head, supports the third trochanter one third of the way down; below which it takes a cylindrical form, and expands slightly above the trochlea. The two condyles, separated by a wide and deep popliteal groove, are nearly subequal in size—the internal somewhat the larger, while the trochlea is long, narrow, and symmetrical. The great trochanter rises slightly above the head, and overhangs a long and quite deep digital fossa. The second trochanter is small. The popliteal space is slightly concave.

Tibia.—The proximal end of the tibia is very large, with massive rugosities for muscular attachment. The two proximal faces are subequal, slightly convex, and separated by a prominent spine, which is grooved at the top. The shaft is long, straight, and compressed antero-posteriorly; the anterior ridge on upper third is very high, and expands into the tuberosity, which is enormous; while on the posterior face is a deep fossa just below the head. The distal end of the shaft is remarkably small, with two concave faces—the internal the smallest—with a low oblique ridge dividing them. There is a prominent malleolus.

The fibula (which is a distinct bone) has a small proximal end, and expands slightly below, with a large external malleolus.

Measurements of Limbs.

The pelvis is described from the right and left ilia, which are broken as they expand towards the crest; the acetabulum is fragmentary, but all the parts remain. The ischium and pubis are only represented by fragments. As figured in [Plate V.], the acetabulum is restored from a somewhat smaller specimen. The most striking feature of the ilium is the long and somewhat constricted neck. The acetabular border is long and curved forward, thinning gradually as it approaches the crest; and marked on the iliac surface, near the acetabulum, by a slight rugosity for the rectus muscle.

The ischial border, while less arched, is probably longer; and the sacral surface, distinctly defined, indicates that the rounded upper border of the ilium rarely reached above the sacral spines. There is a deep groove between the ischial and pubic borders, the latter disappearing about half way up the iliac surface. The gluteal surface near the acetabulum is much depressed; above it expands into a broad, flat, thin plate.

The ischium has a stout neck and thick expansion below; the upper border is very heavy. From what remains of the ischium and pubis we can infer a large elliptical obturator foramen, a short pubic symphysis, a narrow and rather slight pubis, with a small nearly cylindrical neck. The acetabulum is deep, with a thick well-raised border, and a large, oblong, and very deep pit for the ligamentum teres.

Measurements of Pelvis.

The tarsus is arranged in the usual Perissodactyle order. The calcaneum is stout, of good length, tuberous, but not expanding much at the extremity; presenting two faces for the astragalus, of which the internal is the largest, and a large distal face for the cuboid. The astragalus is wide, with less asymmetry in the two articular facets than is common. It articulates with both the cuboid and navicular.

Measurements of Tarsus.

All the above were found at Henry's Fork Divide. In another locality, in connection with a fine head of P. major, was found the lower half of a humerus.

The shaft is cylindrical and twisted, expanding widely at the distal end. Of the two condyles the external is the largest; and, to quote from Dr. Leidy, "a deep supra-condylar fossa occupies the front of the humerus, opposed by a deeper and more capacious anconeal fossa." The greatest breadth between the supracondyloid eminences is ·091 m.

Palæosyops paludosus, Leidy.

Cont. to Ext. Vert. Faun, P. lævidens Cope. U. S. Geol. Survey of Terrs, 1872, p. 591.

Specific characters.—Second superior molar has but one outer tubercle. The cones are low as compared with P. major. The cingula are much less developed. The angle of the lower ramus is much longer. The zygoma is massive and wide. An internal process on the glenoid cavity. Temporal fossæ very deep. Crest of occiput nearly in same vertical line as occipital condyles.

The full details of P. paludosus, which follow, may seem somewhat unnecessary, to one familiar with the comprehensive work of Dr. Leidy upon the subject. Care has been taken not to retrace any of his steps; and the descriptions given below are of those parts of the animal which he did not possess at the time of writing. They include (1) a cranium perfect posterior to the orbits; (2) a right ramus mandibuli, with full dentition, except the first premolar; (3) portions of the fore-limbs, scapula, and pelvis, pes and manus, and many fragments of other parts.

Among the first discoveries on Cottonwood Creek, in a stratum of fine green sand, was a head of P. paludosus, complete posteriorly, and broken off just behind the orbit. A rear view of this has been admirably figured in [Plate I].

General appearance.—The base of the cranium is of great width; the occiput is high, inclined very slightly backwards, and deeply concave from side to side. The forehead is triangular and narrow. The temporal fossa is of immense size and depth, leaving a small intracranial cavity. The zygomas are heavy, and arching widely outwards give this part of the head a strong resemblance to the Felidæ.

Description in detail.—The narrow but prominent basioccipital segment is broadest posteriorly and tapers forwards; divided by a median ridge, which expands anteriorly into a large tuberosity; a feature also characteristic of the tapir. About half an inch in advance of the condyles are the condylar foramina. The condyles are formed of the exoccipitals, which are low and of great lateral expansion. Their junction with the mastoids is marked by a large foramen. The par-occipitals are short and styliform. The supra-occipital region is very large, high, and deeply concave from side to side, much wider than in the tapir, with a marked interparietal suture. The condyles are wide, but not very deep, approaching each other very closely below. The basisphenoids are long and narrow, tapering forwards. The alisphenoids are mutilated, but indicate large vertical ridges joining the parietals, and heavy pterygoid processes, perforated at the base by the alisphenoid canal. They are again perforated by the foramen ovale three fourths of an inch behind this. This completes the base of the skull.

The parietals are very large, they form nearly the whole of the temporal fossil; which, deep and wide, enclosing a small cranial cavity, contrast strongly with the long and shallow temporal fossæ of the tapir. The parietal crest is very broad, and grooved at the top. This high crest and deep adjacent temporal fossæ we at first mistook as pointing to an exclusively carnivorous type. The forehead has a triangular appearance, from the divergence of the two side ridges of the sagittal crest. The postorbital processes are very large, but do not reach the opposing processes of the malar. The orbit is thus left incomplete posteriorly, while in form it greatly resembles that of the Sus. The squamosal encroaches considerably upon the temporal; and sends outwards and downwards a great zygomatic process, which arches outwards from the skull as in the Felidæ, and is more powerful than in any living carnivore (Leidy.) A strong downward direction is especially characteristic, the whole describing a sigmoid curve. The mastoids are of great size vertically, and transversely they are confluent with the par-occipitals. A low, thick process on the internal side of the glenoid cavity prevents lateral motion.

The nasals are long, broad, and thick, convex from side to side, narrowing slightly anteriorly. They are straight, as in Sus, which they resemble more than they do either tapir or rhinoceros. The anterior borders are rounded, and do not reach as far forward as the symphysis of the premaxillary. The malar is broad and thick, probably forming but little of the face, being directed downwards and backwards to meet the zygoma. The postorbital process is short, and rather larger than Dr. Leidy has indicated. The maxillaries, smaller proportionately than in Sus, form posteriorly the floor of the orbit; while the infraorbital foramen is situated over the last premolar. The premaxillaries fail to reach the nasals.

Comparative Measurements of Head.

Lower jaw of P. paludosus described from another specimen.

The peculiar feature of this jaw is its remarkable extension back of the molar series (a feature which has not been noticed heretofore), the distance from the last tooth to the angle being greater than the length of the entire molar series. From beneath the last molar, the lower margin curves gently up to the symphysis. The alveolar border is but slightly curved. Behind the last molar the lower margin forms a sigmoid curve, much more decided than in P. major (first upwards and then downwards), and the ramus thins out rapidly to the angle, where it has a slightly raised border; at the diastema the ramus curves outwards so as to throw the canines out of the line of the molars. The symphysis is long.

Dentition.—The incisors, three in number, from length and shape indicate a semi-circular arrangement, as in most Perissodactyles. They increase from first to third. The crown of the first resembles very much that of the ruminants, but has a straight posterior basal ridge. Its forward edge is worn so as to expose a small tract of dentine. The second is larger, but very much like the first in conformation; while the third has an acute conical crown with a strong basal ridge.

Of the premolars the first is wanting in this specimen. The second is bilobed, the anterior lobe much the larger. It has a small accessory tubercle on its anterior slope. The outer face is rounded, the inner nearly flat. The posterior lobe is very small and obtuse, and with the inner face slightly grooved. In third and fourth the valley between the lobes and the groove on their faces enlarge; while the anterior lobe becomes relatively smaller.

Measurements of Lower Jaw.

Among the first discoveries on Cottonwood Creek, in connection with the lower jaw of P. paludosus just described, were an ulna and radius, and part of the manus and scapula of the same. These have been figured, together with other fragments belonging to the same species, in [Plate III.], and form an interesting complement to the nearly complete hind limb of P. major previously described.

The scapula has a shallow oval glenoid cavity, which is concave longitudinally. Separated from it by a narrow notch is a strongly recurved coracoid, placed on the inner side of the bone. Above the coracoid the border is thin, and arches inwards and then forwards; while the glenoid border is much straighter, rising from a slight tuberosity. The spine rises gradually from the glenoid, and does not bear any indication of an acromion.

The humerus is represented by proximal and distal extremities. The latter presents the same characters on a smaller scale as in P. major. Upon the proximal end there is a deep bicipital groove. The greater tuberosity has a hooked anterior projection, and runs back into a ridge behind. (See Fig 7, [Plate III].)

The proximal end of the ulna (of median breadth anteriorly) narrows into a prominent ridge behind. This gives a trihedral character to the shaft. This is persistent, but is less marked as the fore-and-aft diameter decreases. The remaining features of note are a high tuberous olecranon expanding behind; a shallow sigmoid, divided by a deep fossa into two long divergent articular faces; a trihedral shaft tapering slightly, but distinct from the radius; a narrow distal extremity, with a small irregular face for the cuneiform.

The radius has a rather short shaft, curved forwards, expanding below, and placed immediately in front of the ulna. A strong ridge runs from the external tuberosity up the posterior face of the shaft, disappearing two inches from the top. The proximal end, fitting against the ulna by two small faces, forms the greater part of the elbow-joint, a strong median ridge dividing the proximal face into two subequal surfaces with well-raised borders. The distal end of the shaft is rugose, and more prominent than the proximal. It has an oblong transverse face, which is slightly concave fore and aft. The position of the radius, immediately in front of the ulna, and its manner of articulation, remove any possibility of rotation, a feature in strong contrast with what we should expect from the character of the head.

Measurements of Ulna and Radius.

The manus is described from the third and fourth metatarsals, with three phalanges. The metatarsals are short, with a wide, stout shaft, flat in front and slightly arched forwards behind. The proximal ends unite and form a continuous articular face, concave from side to side; while upon the exposed sides are smooth facets, indicating the presence of two additional toes of less size. (See [Plate III].) The convex distal faces are marked behind by a slight groove.

The phalanges are very short and broad, the lower facets marked by a shallow median groove. The ungual phalanx is wide and short, with a crescentic outline.

Measurements.

Palæosyops vallidens, Cope.

Proceedings Am. Phil. Society, 1872, p. 487; Hayden's Survey, 1872, p. 572.

Specific characters.—Founded on details of dental structure. Superior size. In the upper molars two strong transverse ridges connect the inner tubercle with the outer crescents, inclosing a pit between them. In the premolars the outer crescents fuse almost into a single ridge. These united crescents are relatively narrower. The inner molars are also narrower, and the posterior tubercle of the last is an elevated cone.

Fragment of a lower jaw containing the two posterior molars. The measurements of the teeth are identical with those given by Professor Cope. The depth of the ramus below the last molar is three and one third inches, and it is slightly heavier below than in P. major.

Restoration of Palæosyops.—The elevation of Palæosyops major, which is taken as a type, was approximately the same as that of Palæotherium magnum as restored by Cuvier.

The nearly complete ilium, femur, tibia, and tarsus of P. major in our possession, enable us to make the following estimate of the elevation of the hind quarters.

It is probable that the fore-shoulders were of the same height. From the atlas, axis, and from other cervicals, several dorso-lumbar vertebræ, and part of the sacrum, a rude estimate places the length of the animal at sixty-seven inches, or nearly six feet, including the head, and excluding the tail.

In general features it strongly resembled the tapir, with stout body, slender tail, and very short neck, compensated by a proboscis of considerable length. In comparing the heads of the ancient and modern representatives of this class of Perissodactyles, the points of contrast are the wide stout zygomas, the deep temporal fossæ, the protruding nasals, and the narrow gaps in the dental series of the older type.

There are also strong points of resemblance in the structure of the Palæosyops and Palæothere. Some of the cranial homologues have been detailed by Dr. Leidy. The femur of the Palæothere is shorter, the tibia longer; they are both much stouter. The tarsus of the Palæosyops is narrower, while the remainder of the Pes is very similar. The pelvis, and particularly the ilium and acetabulum of Palæosyops, is more palæotheroid than tapiroid. The similarity of the fore-shoulder in the two types is also striking. The neck of the Palæosyops was probably shorter.

In the descending series of Palæotheres terminating in the diminutive P. minus, and characteristic of the upper eocene of France, we have an interesting counterpart in the large family of Palæosyops, of equal diversity of size and characteristic of our lower eocene, and it is hoped that future discoveries will render a complete discussion of this interesting coincidence possible.

LIMNOHYUS, Leidy.

Proceedings Academy Natural Sciences, Phil., 1872, p. 242; Palæosyops; Marsh, American Journal Sci. and Arts, 1872, p. 122; Hayden's Survey, 1872, p. 592, Cope.

Generic characters.—Resembles Palæosyops in general features. Distinction founded on the possession of two conic tubercles, of the inner series, on the last superior molar instead of one.