430 The American Naturalist. [May, 
slightly notched, while the first phalanges of the lateral digits 
are without the notch. This is an important fact, and in 
connection with what has just been said of Coryphodon, I think 
that it may be accepted as demonstrated, that the form a- 
tion of this notch is dueto the impact of the 
keel upon the lower border of the phalanx. Up to this stage 
therefore, we have the following conditions, viz: a compara- 
tively low primary keel bounded upon either side by well 
marked lateral grooves whose outer borders are as much 
elevated as the keel. We also note that the grooves and keel 
extend but half-way around the end of the metapodial, and 
that the lower edge of the first phalanx in the middle toe is 
distinctly notched. This may be termed the digitigrade 
stage in the development of the keel. 
In the next or unguligrade stage of this development, 
beginning with Mesohippus,' the foot changed from the digiti- 
grade to the unguligrade position in which the main flexure 
of the foot was transferred to the articulation between the two 
last phalanges, and the two proximal phalanges came to 
occupy a position in line with the long axis of the metapodial. 
It is in this genus therefore, that we find the first distinctive 
evidence of an advance of the metapodial keel around the 
lower surface toward the dorsal aspect of the bone. While 
this advance or extension of the keel is slight, it is interesting 
to note that the notch in the first phalanx, which we see just 
beginning in the tapir, has now been transformed into a 
groove and extended nearly acrossthe articular 
face of the bone. Weare thus able to demonstrate that, 
in the further extension of the keel of the metapodial beyond 
that of the digitigrade stage, the groove was formed in the 
phalanz first. It is to be further noted that the keel has 
become prominent and that the lateral grooves of the meta- 
podials are almost entirely obliterated, at least their outside 
walls have disappeared, leaving two nearly plain articular 
surfaces separated by a now prominent, or secondary 
. median keel. Again, if the bones be placed in position, 
*It is highly probable that the most ponies stages of this process are to be traced 2 
to forms somewhat older than this genus. l 
