1893.] Evolution of Metapodial Keels of Diplarthra. 433 
sible explanation of these structures is upon the hypothesis of 
the accumulation of a number of spontaneous variations in 
this direction, entirely independently of any mechanical 
results acquired during the lifetime of the individual. Accord- 
ing to this theory, all mechanical influence must of necessity, 
be rigidly excluded, for the reason that it is possible for it to 
have been exerted, or come into play only during the time - 
that the foot was in use, and therefore, during the lifetime of 
the individual. If mechanical influence is admitted even to 
the slightest extent, then the whole proposition of the trans- 
mission of acquired characters is at once proyen, and who can 
say how much is due to mechanics and how much to natural 
selection ? 
In the production of the completed keel and phalangeal 
groove, it must be shown by those who hold the theory of 
origin purely by natural selection that they are congenit- 
ally correlated structures, otherwise they must, according to the 
very nature of the case, have proven injurious to their 
possessor. In order for a keel to be useful it must be accom- 
panied by a groove into which it is received. In other words 
it must be shown that any variation in the one must have 
affected the other. I shall now attempt to show that the 
keels and grooves are not congenitally cor- 
related. It has been shown in the preceding pages, that 
the phalangeal notch which is the earliest and incipient stage 
of the phalangeal groove is not correlated with the keel. The 
evidence for this is to be seen in Coryphodon, Elephas, Metamy- 
nodon, Titanotherium and others in which the keel exists 
without any vestige of the phalangeal notch. 
Again, in the tapir and rhinoceros, the lateral metapodials 
are provided with distinct keels but the notch is absent. 
In the middle toe of many of the lower Perissodactyla, the 
keel is associated with the notch, but it was not until the foot 
assumed a position whereby it was possible for the keel to 
impinge upon the lower border of the phalanx that the notch 
appeared. It would indeed require a great stretch of the 
imagination to believe that the keel and notch were congenit- 
ally correlated structures in one toe and not in the others of 
