when
she comes to the more definitely terrestrial Amphibia Nature gives up
the attempt to use the gill-chamber as a lung, and creates a new organ,
the true vertebrate lung, specially adapted for breathing air (p. 475).

But whatever means Nature adopts, her aim is always the same--to
specialise, to differentiate, to produce diversity from uniformity.

Differentiation not only raises the level of organisation; it usually
also takes the direction of adaptation to particular habits of life, and
this is perhaps the most fruitful cause of diversity. Everywhere we find
animals specialised in adaptation to their environment--to life in air
or water, or on land--and many of their most striking differences are
due to this cause. But adaptation may also act in reducing diversity,
for there necessarily occur many instances of parallel adaptation or
convergence. So we get the extraordinary parallelism between the
families of marsupials and the orders of placentals,[307] the remarkable
similarity between the respiratory organs of land-crabs and
air-breathing fish--to mention only two out of an immense range of
analogous facts.

The last cause of diversity that Milne-Edwards adduces is what he calls
a "borrowing" of peculiarities of structure from another systematic
group. Thus, "among reptiles, the tortoises seem to have borrowed from
birds some of their characteristic features of organisation; and among
the sauroid fishes the piscine type seems to have been influenced by the
type from which reptiles are derived" (p. 479). So many riddles that, a
little later on, stimulated the ingenuity of the evolutionists!

Such, then, were the factors which Milne-Edwards considered adequate to
explain the rich variety of animal forms. We cannot do better than quote
his own summary of his doctrine:--"To sum up, then, the great
differences introduced by Nature into the constitution of animals seem
to depend essentially upon the existence of a certain number of general
plans or distinct types, upon the perfecting in various degrees either
of the whole or of parts of each of these structural plans, upon the
adaptation of each type to varied conditions of existence, and upon the
secondary imitation of foreign types by certain derivatives of each
particular type" (p. 480).

We have laid stress on the fact that Milne-Edwards put function before
form, for this is the mark of the true Cuvierian. With it goes the
belief that Nature forms new parts t