ones--and in such a way, moreover, as to be
afterwards developed into action in the neuters only. And
use-inheritance all the while is being thoroughly overpowered by
impression-inheritance--by the full transmission of that which is merely
seen in others! If such a law prevails, one may feel cold because an
ancestor thought of the frosty Caucasus. None of this absurdity would
arise if it were clearly seen that a parent is only a trustee--that
transmission and development are perfectly distinct--that parental
modifications are irrelevant to those transmitted to offspring.

FOOTNOTES:

[67] _Essays on Heredity_, p. 104. Weismann's theory is clear, simple
and convenient, but incomplete; for, unlike Darwin's theory of
pangenesis, it scarcely attempts any real explanation of the extremely
complex potentialities possessed by the reproductive elements. Perhaps
we might retain Darwin's self-multiplying gemmules without supposing
them to be thrown off by the cells, which will no longer be credited
with _two_ modes of multiplication. These minute germs or gemmules may
have been evolved by natural selection playing upon the sample germs
that achieve development; and they may exist either separately, or
(preferably but perhaps not invariably) in aggregates to form Weismann's
germ-plasm.

[68] _Contemporary Review_, Dec., 1875, p. 88.

[69] _Variation of Animals and Plants under Domestication_, ii. 286.

[70] _Variation of Animals and Plants under Domestication_, ii. 388,
398, 367; _Life and Letters_, iii. 44.

[71] _Contemporary Review_, Dec., 1875, pp. 94, 95.

CONCLUSIONS.

USE-INHERITANCE DISCREDITED AS UNNECESSARY, UNPROVEN, AND IMPROBABLE.

General experience teaches that acquired characters are not usually
inherited; and investigation shows that the apparent exceptions to this
great rule are probably fallacious. Even the alleged instances of
use-inheritance culled by such great and judicious selectors as Darwin
and Spencer break down upon examination; for they can be better
explained without use-inheritance than with it. On the other hand, the
adverse facts and considerations are almost strong enough to prove the
actual non-existence of such a law or tendency. There is no need to
undertake the apparently impossible task of demonstrating an absolute
negative. It will be enough to ask that the Lamarckian factor of
use-inheritance shall be removed from the category of accredited factors
of evolution to that