* * *

The next difficulty to which I shall allude is that of explaining by the
theory of natural selection the preservation of the first beginnings of
structures which are then useless, though afterwards, when more fully
developed, they become useful. For it belongs to the very essence of the
theory of natural selection, that a structure must be supposed already
useful before it can come under the influence of natural selection:
therefore the theory seems incapable of explaining the origin and
conservation of _incipient_ organs, or organs which are not yet
sufficiently developed to be of any service to the organisms presenting
them.

This objection is one that has been advanced by all the critics of
Darwinism; but has been presented with most ability and force by the
Duke of Argyll. I will therefore state it in his words.

If the doctrine of evolution be true--that is to say, if all
organic creatures have been developed by ordinary generation from
parents--then it follows of necessity that the primaeval germs must
have contained potentially the whole succeeding series. Moreover,
if that series has been developed gradually and very slowly, it
follows, also as a matter of necessity, that every modification of
structure must have been functionless at first, when it began to
appear.... Things cannot be selected until they have first been
produced. Nor can any structure be selected by utility in the
struggle for existence until it has not only been produced, but has
been so far perfected as to actually be used.

The Duke proceeds to argue that all adaptive structures must therefore
originally have been due to special design: in the earlier stages of
their development they must all have been what he calls "prophetic
germs." Not yet themselves of any use, and therefore not yet capable of
being improved by natural selection, both in their origin and in the
first stages (at all events) of their development, they must be
regarded as intentionally preparatory to the various uses which they
subsequently acquire.

Now this argument, forcible as it appears at first sight, is really at
fault both in its premiss and in its conclusion. By which I mean that,
in the first place the premiss is not true, and, in the next place, that
even if it were, the conclusion would not necessarily follow. The
premiss is, "that every modification of structure must have been