690 
The American Naturalist. [August, 
It is not yet possible to determine the number of digits pres- 

Fig. 3.—Epitherium 
laternarium Amegh. ; 
i posterior foot; 
two-fifths natural size, 
From Ameghi 
‘ent in the greater number of the genera of 
this family.. It is, however, extremely prob- 
able that we have four or five in the Eocene 
genera Anisolophus and Thoatherium. If so, 
we have a successional reduction in later peri- 
_ ods, similar to that which has been shown to 
have occurred in the horse line. In Epitherium 
Amegh. we reach a stage corresponding to 
that of Hippotherium, the lateral digits being 
reduced to the condition ot dew-claws, accord- 
ing to Ameghino (Fig. 3). Proterotherium 
and Anisolophus are the only genera in which 
the dentition is well known, and our knowledge 
is confined to the molars. The premolars 
nearly resemble the true molars in both jaws. 
Two species certainly belong to Proterothe- 
rium, and they are from the “ inferior Oligo- 
cene” of Patagonia. Three other species 
from the Oligocene and Eocene of the same 
region, described under distinct generic names 
by Ameghino, have not been yet clearly distin- 
guished from Proterotherium. Two species 
are referred to Anisolophus, which with the 
single one of Thoatherium are from the Eo- 
cene. But one species of Epitherium (Æ. 
laternarium Amegh.) is yet known. This is 
a most interesting animal, and took the place 
in the Miocene fauna of Argentina (accord- 
ing to Ameghino) of the three-toed horses in the corre- 
sponding age in the Northern Hemisphere. It was about the 
size of a small guanaco, and was digitigrade, like the higher 
Diplarthra. 
In the MacrAUCHENIID& we have a line of modification differ- 
ent from that of either of the other two families, resembling in 
some respects both of them. Thus the superior molars are con- 
structed on the one type in the modification of the external 
"dees: 

