4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 134 
felids, resting them rather largely on his interpretation of M, as es- 
sentially the carnassial tooth of the lower series. 
Osborn (1907, fig. 95, E) reillustrated the teeth in the type of 
A. leidyanum indicating that all were deciduous and belonged to 
Patriofelis, but Matthew, in 1909, demonstrated the synonymous posi- 
tion of Aelurotherium, as well as Limnofelis, Oreocyon, and Pro- 
topsalis, with Patriofelis, correctly identifying the various teeth in 
the types. As to whether all the species that he placed in synonymy 
with P. ferox, however, represent that species and none P. ulta, may 
be doubted. 
In addition to describing the skeleton of Patriofelis ferox, Wort- 
man, in his 1894 paper, pointed out a number of characters that he 
believed indicated that Patriofelis or a closely related form gave rise 
to the pinnipeds. This suggestion was criticized by Osborn in 1900, 
and in 1902 Wortman defended his conclusion, taking a strong stand 
against Osborn’s views, but was again refuted, this time by Matthew 
in 1909. Kellogg, in 1922, discussed at length such evidence as had 
been presented on the origin of the pinnipeds, but deduced that a 
case for the Oxyaenidae had not been demonstrated, nor had Mat- 
thew’s suggestion of an arctoid fissiped ancestry solved the problem. 
Matthew’s (1909) monographic study of the Bridger carnivores 
and insectivores included a review and restudy of the known material 
of Patriofelis, as well as a taxonomic revision. The four species that 
he recognized are indicated as representing four distinct stratigraphic 
units and include as the oldest Patriofelis tigrinus (Cope) from the 
Wind River beds; a species from the Huerfano B beds originally re- 
ferred by Osborn (1897) to P. ulta, but named Patriofelis colora- 
densis by Matthew ; the genotypic species Patriofelis ulta Leidy from 
Bridger B; and Marsh’s Patriofelis ferox from the upper Bridger 
levels. In 1915 he decided that P. coloradensis represented Ambloc- 
tonus rather than Patriofelis. 
Later discussions of Patriofelis included a review by Thorpe 
(1923) in which an attempt was made to revive Patriofelis latidens. 
Thorpe recognized the error in Wortman’s taxonomic procedure in 
arriving at the name Aelurotherium latidens, but seems to have com- 
mitted a similar error in regarding the lower jaws, Y.P.M. No. 10940, 
as the type of Oreocyon latidens, whereas the holotype of Limnofelis 
latidens is No. 10904, a last upper premolar. Although Marsh’s char- 
acterization of the genus Oreocyon was based on referred materials, 
the type species was clearly that which he originally described as 
Limnofelis latidens and hence cannot have a different type specimen. 
In his revision of the ‘Pseudocreodi,” Denison (1938) revived 
