and there, so that the creature must develop in one way or another to
fill it?

Another set of cases is the production of similar functional results by
most diverse means, as in the case of flying animals, birds,
pterodactyles, and bats; here there is a widely different modification
of the fore-arm and other bones, all for the same purpose. The reader
will do well to refer to Mr. Mivart's book on this subject.

Again, the question of types seems to be pointed to in the curious fact
of what I may call the double development of birds from reptiles. Mr.
Mivart says, "If one set of birds sprang from one set of reptiles and
another set from another set of reptiles, the two sets could never by
'natural selection' only have grown into such perfect similarity." Yet
we can trace the _Struthious_ birds (those that, like ostriches, do not
fly) through the Dinosaurs and _Dinornis_, and the flying Carinate birds
though pterodactyles, _Archaeopteryx_, and _Icthyornis_, &c.

It might well be added to this part of the subject, that granted that
developmental changes were often small, that progress was attained
little by little, this does not appear to have been always the case.

The discoveries of the fossil species of horse,[1] _Eohippus,
Hipparion_, and so forth, clearly establish a developmental series, and
the ancient forms are claimed as the ancestor of the modern horse; but
these (Professor Owen tells us) differed more from one another than the
ass and the zebra (for instance) differ from the horse. Still, of course
it may be that there are still undiscovered intermediate forms; and in
any case there need be no desire to detract from the value of the
series, as really pointing towards a gradual perfection of the horse