432 The American Naturalist. [May, 
phalanx went back to its more normal position, that is, in line 
with the long axis of the metapodial, there would then be a 
_ plane surface of bone lying immediately in front of the keel 
which would be opposed to a groove in the phalanx. An 
extension of the keel would necessarily follow as a result of 
moulding of this metapodial surface to fit the groove. Flex- 
ion, therefore, of the phalanx upon the metapodial under 
pressure, bringing the keel already formed to a more and more 
anterior position with reference to the articular surface of the 
phalanx, thus causing it to become grooved, and the subse- 
quent moulding of the keel to fit this groove, I conceive to be 
the complete mechanical explanation of the production of this 
structure. The various stages in the forward extension of the 
keel is to be found in such series as Hyracotherium, Epihippus, 
Mesohippus, Anchitherium, Hipparion and Equus. 
FicurE 5.—Modern Tapir. Middle toe inside view, showing position of the 
phalanges upon the ground. 
In the preceding discussion, only the possibility of the meta- 
podial keels having been produced and extended forward by 
mechanical means, has been shown. Iwill now turn to the 
matter of the application of this reasoning to the somewhat 
broader question of the transmission of acquired characters, 
and see whether or not any argument can be adduced in sup- 
port of the proposition that the development of these structures 
has been due solely to mechanical causes, and that they have 
in time been transmitted to succeeding generations. Upon 
this hypothesis it must be assumed that the changes took 
place first within the lifetime of the individual and that they 
finally came to be inherited. It must likewise be shown that 
wherever a groove or a keel has occurred, there must have 
been pressure, impact or strain, exerted at that particular 
point sufficient to accomplish the result. The only other pos- 
