, as in that of
_Branchipus stagnalis_, quoted by Mr. Wallace, or in that of the New
Zealand Kea whose skin, I was assured by the late Sir Julius von Haast,
has already been modified as a consequence of its change of food. Here
we can show that in even a few generations structure is modified under
changed conditions of existence, but as we believe these cases to occur
comparatively rarely, so it is still more rarely that they occur when and
where we can watch them. Nature is eminently conservative, and fixity of
type, even under considerable change of conditions, is surely more
important for the well-being of any species than an over-ready power of
adaptation to, it may be, passing changes. There could be no steady
progress if each generation were not mainly bound by the traditions of
those that have gone before it. It is evolution and not incessant
revolution that both parties are upholding; and this being so, rapid
visible modification must be the exception, not the rule. I have quoted
direct evidence adduced by competent observers, which is, I believe,
sufficient to establish the fact that offspring can be and is sometimes
modified by the acquired habits of a progenitor. I will now proceed to
the still more, as it appears to me, cogent proof afforded by general
considerations.

What, let me ask, are the principal phenomena of heredity? There must be
physical continuity between parent, or parents, and offspring, so that
the offspring is, as Erasmus Darwin well said, a kind of elongation of
the life of the parent.

Erasmus Darwin put the matter so well that I may as well give his words
in full; he wrote:--

"Owing to the imperfection of language the offspring is termed a new
animal, but is in truth a branch or elongation of the parent, since a
part of the embryon animal is, or was, a part of the parent, and
therefore, in strict language, cannot be said to be entirely new at the
time of its production; and therefore it may retain some of the habits of
the parent system.

"At the earliest period of its existence the embryon would seem to
consist of a living filament with certain capabilities of irritation,
sensation, volition, and association, and also with some acquired habits
or propensities peculiar to the parent; the former of these are in common
with other animals; the latter seem to distinguish or produce the kind of
animal, whether man or quadruped, with the similarity of feature or form
to the parent.