1393.] Evolution of Metapodial Keels of Diplarthra. 427 
belongs to the higher forms of both the odd and the even-toed 
ungulates. Now if we attempt to trace the history of the 
metapodial keels in the evolution of such a form as the horse, 
no one will deny, it seems to me, that we have here an 
unusually complete phylum represented, which clearly indi- 
cates the different stages in their development. In the earlier 
members of this series, such for example as Hyracotherium, the 
keels are confined to the plantar and palmar surfaces of the ` 
metapodials, and the animal was subdigitigrade. In such a 
form as Mesohippus, on the other hand, we pass from the digi- 
tigrade position of the foot to that of the unguligrade, wherein 
the weight of the body was supported upon the terminal 
phalanx or coffin bone. Just what induced this change is 
unknown, but it is more than probable that it was in some 
way connected with the reduction of the lateral toes. 
We do not yet know with certainty the ancestors of the 
horse series beyond Hyracotherium, but there can be little doubt 
that the feet were pentadactyle and plantigrade. So far at 
least as the metapodials are concerned, we have such a condi- 
tion in Coryphodon, in which the keels are but faintly indica- 
ted, and the two lateral grooves of which the primary keel 
forms the common median wall, are in the first stages of 
development. There is every reason to believe that this ani- 
mal was plantigrade or subplantigrade in gait, and that the 
failure to develop the keels was due to the more equal distri- 
bution of the strain upon all the digits. This istrue of the 
foot of the elephant in which we observe another structure 
which has certainly had much to do with the retardation of 
keel development viz: a very thick plantar pad. The posses- 
sion of this pad not only tends to distribute the strain and 
equalize the pressure but at the same time limits the amount 
of flexion and extension of the phalanges. It is a fact worthy 
of especial note, and one which I will make use of later on, 
that in all these forms wherein the lateral grooves 
andkeelsare verylittledeveloped,the plantar 
borderofthearcticularsurface ofthe first pha- 
lanx is not at all notched. It is also to be noted 
that the feet are short and spreading. 
