rs that they form with the latter one great group--the ichthyopsida of
Professor Huxley--which differs widely from reptiles; while its two
component classes (fishes and amphibians) are difficult to separate from
each other in a thoroughly satisfactory manner.

If we admit the hypothesis of gradual and minute modification, the
succession of organisms on this planet must have been a progress from the
more general to the more special, and no doubt this has been the case in
the majority of instances. Yet it cannot be denied that some of the most
recently formed fossils show a structure singularly more generalized than
any exhibited by older forms; while others are more specialized than are
any allied creatures of the existing creation.

A notable example of the former circumstance is offered by macrauchenia--a
hoofed animal, which was at first supposed to be a kind of great llama
(whence its name)--the llama being a ruminant, which, like all the rest,
has two toes to each foot. Now hoofed animals are divisible into two very
distinct series, according as the number of functional toes on each hind
foot is odd or even. And many other characters are found to go with this
obvious one. Even the very earliest Ungulata show this distinction, which
is completely developed and marked even in the Eocene palaeotherium and
anoplotherium found in Paris by Cuvier. The former of these has the toes
odd (perissodactyle), the other has them even (artiodactyle).

Now, the macrauchenia, from the first relics of it which were found, {110}
was thought to belong, as has been said, to the even-toed division.
Subsequent discoveries, however, seemed to give it an equal claim to rank
amongst the perissodactyle forms. Others again inclined the balance of
probability towards the artiodactyle. Finally, it appears that this very
recently extinct beast presents a highly generalized type of structure,
uniting in one organic form both artiodactyle and perissodactyle
characters, and that in a manner not similarly found in any other known
creature living, or fossil. At the same time the differentiation of
artiodactyle and perissodactyle forms existed as long ago as in the period
of the Eocene ungulata, and that in a marked degree, as has been before
observed.

Again, no armadillo _now living_ presents nearly so remarkable a speciality
of structure as was possessed by the _extinct_ glyptodon. In that singular
animal the spinal column had most of its joints f