1894 8086 1894] Wortman, Osteology of Patriofelis. 161 incisors in the lower jaw, and the molar dentition is much reduced. In the fore limb the humerus is shorter than the scapula, and provided with a great deltoid crest and an entepicondylar foramen. The scapula is large, with short neck and metacromion process. The ulna has the same powerful olecranon, which, together with the prominent deltoid crest, however, appears to be very common among the Creodonts. The trapezium is enlarged; the feet are broad and spreading, and in the ungual phalanges, the subungual processes are largely developed. In the vertebral column, as already noted, some of the seals (notably Phoca vitulina) show marked traces of the complex articulation of the lumbar zygapophyses. The pubic symphysis of Patriofelis is short, lies behind the acetabulum, and is not anchylosed. The fibula is large and unreduced in both groups; the trochlea of the astragalus is not grooved; the tuber of the calcaneum is short, and the cuboid has a very oblique calcaneal facet and a large contact with the astragalus. The foot is broad and spreading, and the ungual phalanges have perforated subungual processes. PROBABLE HABITS OF PATRIOFELIS. From the structure of the limbs more than any other feature in the osteology of Patriofelis, I am led to conclude that it was aquatic or semi-aquatic in habits. The broad, flat, plantigrade feet, with their spreading toes, suggest at the first glance their use for swimming. The eversion of the feet, together with the general clumsiness of the limbs, point, moreover, to the fact that the animal was not an active runner. Now, if the animal was aquatic, what was the nature of its food? It certainly could not have been fish, for the reason that the remains of fishes are very scarce in the Bridger sediments. If, however, we can form any judgment from their remains, I think that it can be safely stated that the Bridger Lake literally swarmed with turtles, and if Patriofelis frequented the water, it is highly probable that they formed a staple article of its diet. This supposition accords well with the great strength and power of the jaws, together with the robust and much-worn condition of the teeth. There is another fact which may be mentioned in this connection, which has a direct [May, 1894.] 11 bearing upon this conclusion, and that is the existence of coprolites in the Bridger sediments containing fragments of turtle shells. This, while it is not at all conclusive, yet demonstrates that there was an animal living on the borders of the ancient lake, that was accustomed to capture turtles for food, and from what has already been stated, I think that animal was Patriofelis. He was, perhaps, not as expert a swimmer as the seals now are, but was sufficiently active in the water to capture turtles. When the lake disappeared, it can be conjectured that Patriofelis took to the open sea, and finally came to feed upon fish exclusively. It is further conceivable that in their new habitat their swimming power was gradually increased, and, owing to the soft nature of their food, the great strength and power of the jaws were gradually lost, and the teeth became gradually modified into the simple degenerate organs which constitute the dental equipment of the modern Pinnipedia. IV. CLASSIFICATION AND SPECIES OF PATRIO FELIS. It will be seen from what has already been stated that Patriofelis is a member of the Creodonta. Various efforts have been made from time to time to give an exact definition of this group, but these definitions have as yet proven very unsatisfactory. That the Creodonta stand in general antecedent relationship to the Carnivora is now abundantly demonstrated, but whether the Carnivora arose from one or several stems of the Creodonta, is still an open question. It is held by Cope and Scott that all the Fissipedia are descended from the Miacidæ of the Credonta. 1 have expressed a contrary opinion, with Schlosser, in regard to the cats. One of the chief osteological distinctions between the Creodonta and the Carnivora consists in the union of the scaphoid, lunar and centrale in the carpus of the Carnivora, whereas they are free in the Creodonta. There are, moreover, such characters as the fissured ungual phalanges, the complex articulations of the lumbar vertebræ, the relative size and degree of convolution of 13-4 the cerebral hemispheres, and a number of other characters of less importance which serve to distinguish these groups from each other. It must be borne in mind, however, in considering these differences, that if the Carnivora have been derived from the Creodonta, the distinctions between them must have been exceedingly slight at the point where they actually meet, and that any definition which can be given will, according to the very nature of the case, fail. There is considerable evidence to show that wherever the Creodonta continued beyond the Lower Miocene they took on certain characters which now so sharply distinguish the Carnivora. In the Miocene genus Hyanodon, the cerebral hemispheres were almost, if not quite, as well convoluted as their carnivorous cotemporaries, and in the European species of the same genus, according to Scott, the scaphoid, lunar and centrale were united as well. In a like manner many of the Miocene Carnivora show marked traces of their Creodont ancestry. This is especially seen in the flat astragalus and the remains of the suture uniting the scaphoid and the lunar, as well as the simpler type and less convoluted cerebral hemispheres. It would appear, therefore, that these groups, being incapable of exact definition, have lost much of their original significance, and are now to be regarded as mere matters of convenience in classification. The same may be said of the Insectivora with relation to the Creodonta. The systematic position of Patriofelis within the Creodonta is not difficult to discover. Its general skeletal structure is so much like that of Oxyana that, notwithstanding the differences in the teeth, they must be placed in the same family. Oxyana is the older form, and has the more primitive dentition, but the differences are not greater than we would be led to anticipate in the ancestral genus. I think that it can be accepted as demonstrated that Patriofelis is the direct descendant of Oxyana, which may have likewise given off a branch which terminated in the modern seals. It is somewhat doubtful whether this branch leads through Patriofelis. Regarding the relationship of Patriofelis to Hyanodon I have spoken on a former page. I do not think that they can be consistently associated in the same family. The family definition may now be stated as follows: Oxyanidæ.-Muzzle short and truncate; interorbital region constricted and elongated; sagittal crest extended well in advance of the brain-case. A preglenoid process and no postglenoid foramen; an alisphenoid canal and prominent mastoid. Lachrymal bone not extended out upon the face. Two pairs of lower incisors. Trapezium enlarged; pubic symphysis not anchylosed; fibula unreduced; calcaneo-cuboidal facet very oblique; cuboid having large contact with astragalus. Fibula not articulating with calcaneum. Oxyana Cope.-Premolars in the lower jaw 4, molars 2. Last superior molar transverse. Patriofelis Leidy.-Premolars in lower jaw 3; molars 2, Last superior molar longitudinal. The species are not numerous, and it is indeed questionable whether more than three should be referred to Patriofelis. Leidy's type species P. ulta is easily distinguished by its small size; it is scarcely more than half as large as P. ferox. I have chosen to regard Cope's species P. tigrinus as distinct chiefly on account of its having come from an older formation, although there is no character observable in the fragmentary specimen of P. tigrinus to warrant such a belief. When we have better specimens of it, however, it will doubtless show a nearer relationship to Oxyana than to the Bridger species. There is yet another species which was described by the writer' under the name of Patriofelis leidyanus." It was stated at the time that its reference to the genus Patriofelis is doubtful. I am now convinced that it does not belong here, but is probably a forerunner of the Miocene Nimravidæ. Until more of it is known it is impossible to give to it a generic definition, and I therefore refrain from proposing a new name. 1 Fossil Mammals of the Wahsatch, Bull. Amer. Mus., Vol. IV, 1891, p. 98. 2 This species, together with the figure of it, is erroneously attributed by Zittell, in his 'Handbook of Palæontology,' to Osborn. Article VI. ON THE MAMMALS OF ARANSAS COUNTY, TEXAS, WITH DESCRIPTIONS OF NEW FORMS OF LEPUS AND ORYZOMYS. By J. A. ALLEN. The following paper on the mammals of Aransas County, Texas, is based on a collection made by Mr. H. P. Attwater during the years 1892 and 1893. The collection numbers about 300 specimens, representing 24 species, in most instances by good series of both young and old. About three-fourths of the specimens have been purchased by the Museum, the rest being reserved by Mr. Attwater for his private collection. They have all been kindly forwarded to the Museum for study, and I am further indebted to Mr. Attwater for the valuable field notes presented in the following pages. The present collection is especially interesting from having been gathered from a very limited area, for the most part within a radius of ten miles of the town of Rockport. As shown on the accompanying map, Aransas County consists of a number of small islands and of several irregularly shaped peninsulas formed by the extension inland of various bays. The specimens were collected partly on the mainland, on marshy ground in the immediate vicinity of Rockport, and partly on the adjoining small islands numbered 2, 3 and 4 on the accompanying map. They doubtless fairly represent the mammalian fauna of this very limited area, but probably a number of additional species occur in other parts of the county. These would probably include a number of additional species of Bats, perhaps one or two Shrews, and a Harvest Mouse (Reithrodontomys). No species of Kangaroo Rat was met with, and Mr. Attwater is very confident that none occurs within the area covered by his explorations. As marked changes have already occurred in the mammalian fauna of Aransas County since its settlement, it has seemed desirable to include in the list a few species that have already become nearly or quite extirpated. The total number of species |