ion,
however, continues to occur in it. Since the organ is now useless,
natural selection will no longer restrain variation in such an organ,
and degeneracy will naturally follow, for of all the variations that
occur in the organ, those tending to loss are more numerous than those
tending to addition. If the embryonic development of a whale's hind leg
be compared to some complicated mechanical process, such as the
manufacture of a typewriter, it will be easier to realize that a trivial
variation which affected one of the first stages of the process would
alter all succeeding stages and ruin the final perfection of the
machine. It appears, then, that progressive degeneration of an organ can
be adequately explained by variation with the removal of natural
selection, and that it is not necessary or desirable to appeal to any
Lamarckian factor of an unexplainable and undemonstrable nature.

The situation remains the same, when purely mental processes, such as
instincts, are considered. Habit often repeated becomes instinctive, it
is said; and then the instinct thus formed by the individual is passed
on to his descendants and becomes in the end a racial instinct. Most
psychologists have now abandoned this view, which receives no support
from investigation. Such prevalence as it still retains seems to be
largely due to a confusion of thought brought about by the use of the
word "instinctive" in two different senses,--first literally and then
figuratively.

A persistent attempt has been made in America during recent years, by
C. L. Redfield, a Chicago engineer, to rehabilitate the theory of the
inheritance of the effects of use and disuse. He has presented it in a
way that, to one ignorant of biology, appears very exact and plausible;
but his evidence is defective and his interpretation of his evidence
fallacious. Because of the widespread publicity, Mr. Redfield's work has
received, we discuss it further in Appendix B.

Since the importance of hormones (internal secretions) in the body
became known, it has often been suggested that their action may furnish
the clue to some sort of an inheritance of modifications. The hormone
might conceivably modify the germ-plasm but if so, it would more likely
be in some wholly different way.

In general, we may confidently say that there is neither theoretical
necessity nor adequate experimental proof for belief that the results of
use and disuse are inherited.

(4) When we come to c