292 UNGULATA 
EXTINCT TRANSITIONAL ARTIODACTYLES. 
In this place it will be convenient to notice briefly a few of 
the extinct types of Tertiary Artiodactyles which connect the 
existing bunodont Suina with the more specialised selenodont 
groups mentioned below so closely as to show that in a strictly 
paleontological classification such groups cannot be maintained. 
It should be mentioned that while some of these extinct forms 
were in all probability actual ancestral links between the bun- 
odonts and selenodonts, others, like the Anoplotheres, died out 
entirely without giving rise to any more specialised descendants. 
Cheropotamide.—In this family the molars are intermediate in 
structure between those of the Suidw and the next family. The 
upper ones have very broad crowns, with the five columns arranged 
as in Anthracotherium , while the premolars are not secant, and may 
be very large. The best known forms are the small Cebochwrus of 
the Phosphorites of Central France ; Cheropotamus of the Upper 
Eocene, the type species of which was of the size of a large Pig, 
with the dental formula 2 3, ¢ 4,p 4, m %, and no distinctly 
selenodont structure in the molars; the much larger Elotherium, 
from the Upper Eocene and Lower Miocene of both the Old and New 
Worlds, which presents the very rare feature of the absence of a third 
lobe to the last. lower molar; and the equally large Tetraconodon of 
the Pliocene of India, in which this third lobe was present and the 
premolars were of enormous size. The remarkable North American 
Eocene genus Achenodon should perhaps also be placed here. 
Anthracothertide.—The genera Anthracotherium and Hyopotamus, 
of the upper Eocene and Miocene, 
have the typical Eutherian dental for- 
mula; the upper molars (Fig. 111) 
, carrying three columns on the anterior 
7 and two on the posterior half of the 
crown, all of which are of a more or 
less decidedly selenodont structure. 
The mandible has a descending flange 
at the angle. The figured tooth (in 
which the antero-internal and antero- 
median columns are imperfect) may be 
compared with the diagram given in 
Fig. 5, p. 32, when the homology of 
Fic. 111.—The imperfect third lett the columns or tubercles will be at 
otamus git e 
pram Homan genee, once epparent, the broken antaro- 
logia Indica.) presenting the proto- 
conule. Some of the species are of 
large size, while others are comparatively small. 
