1393.] Fossil Mammalia of Southern Patagonia. 443 
ancestral line of the horses, the genera Palaeotherium and 
Hyracotherium, as well as the different types of the sub-order of 
Condylarthra. In my opinion, the horses originate from a 
group of Litopterna, from which sprang, at the same time, the 
Paleotheriidæ and the Hyracotheriidæ, and this ancestral type 
would be closely related to the Proterotheriidæ. 
It would seem that I am right, for the splendid monograph 
of the genus Mesohippus, recently published by M. Scott, shows 
that the calcaneum of this genus still possessed a very small 
articular facet for the fibula, the last vestige of the litoptern 
organization, however, the shape of the astragalus, as a whole, 
is already like that of a Perissodactyl. 
We now have evidence to warrant the statement that the 
horses are descended from a form allied to the Proterotheride, 
and which ought to differ from it above all by its orbits open 
behind, and its complete dentition, in a continuous series. 
This ancestral is, perhaps, the Notohippus or a kindred genus, 
but, be that as it may, it is undoubtedly in the Northern Hem- 
isphere that the group of horses has acquired the characters 
which distinguish it at the present time. 
The genus Astrapotherium is now known, not only by a skull 
almost entire, but by many of the bones of the skeleton. The 
skull, in its anterior part, agrees in a number of points with 
those of the Proboscidians, and could have carried a proboscis 
as fully developed as that of the elephants. The bones of the 
limbs also greatly resemble those of the elephants, and there is 
no doubt that this type represents the nearest relative of the 
Proboscidians discovered up to this time in the older forma- 
tions. The genus Astrapotherium is certainly not the direct 
ancestor of the Proboscidians, but only a collateral branch of 
the trunk from which the latter have sprung. However, some 
of the other genera of the family of the Astrapotheride, Astra- 
podon for example, may well be considered the ancestor of the 
living Proboscidians. 
In any case, the latter have acquired their characters upon 
some other continent, as it is certain that during the epoch of 
the formation of the fossiliferous beds of the Parana, the Pro- 
boscidians had no representatives in our country. These ani; 
