PRIMITIVE MAMMALS IN THE LONDON CLAY OF HARWICH. 101
hoofed mammal. There is a complete set of front teeth, and the
size of the antorbital foramen of the skull shows that there
cannot have been any special development of the upper lip.
The teeth, indeed, are adapted for feeding on succulent vegetation,
which would be grasped by the incisors. The humerus is typically
that of a hoofed mammal, and Owen described it as most like
the humerus of a hyrax, with very little power of twisting. The
femur is that of an odd-toed hoofed mammal. The three meta-
tarsal bones preserved are nearly similar in stoutness, indicating
that probably the hind foot comprised five toes. Hyracotherium,
therefore, must have been specially adapted for living in marshes,
like the existing tapirs of Central America and Malaya.
Hyracotherium has hitherto occurred in Europe only in the
London Clay, but fragments of nearly related mammals have
been found in somewhat later Eocene formations on the. Continent.
These are of great interest as showing that in subsequent members
of the same group, the grinding teeth began to be ridged rather
than tuberculate—became, indeed, more efficient for crushing
and tearing vegetables—while the feet were reduced to three
toes, with the middle toe the largest and stoutest. Owen closely
followed these discoveries and tried to interpret them. So long
ago as 1857, in a lecture at the Royal School of Mines,8 he
pointed out that Hipparion of the Miocene period was a link
between the Eocene Palaeotherium and the modern horse (Equus)
in the structure of both the feet and teeth : the ridged teeth
became deepened for grinding hard food, while the three-toed
foot gradually became one-toed. There seemed to be a "filiation
of species," as he termed it. In 1868 9 he recognised other
possible links in this series of ancestral horses, and added :
"Furthermore, in the oldest Eocene (London Clay, etc.) we get
evidence of Ungulates (Pliolophus, Hyracotherium, Coryphodon),
in which the perisso- and artio-dactyle characters were less differ-
entiated than in Palaeotherium and Anoplotherium, affording
additional significant evidence of progressive departure from
generalised type. Thus, the succession in time accords with
the gradation modifications by which Palaeotherium is linked
on to Equus."
Hyracotherium (or Pliolophus), as represented by Mr. Bull's
8 R. Owen, Anatomy of Vertebrates, vol. iii. (1868), p. 791.
9 R. Owen, loc. cit., 1868, p. 792.