al
with the slightest blemish or in any degree inferior may be freely
rejected. Hence length of time is an important element of success. Thus,
also, propagation at an early age and at short intervals favours the work.
Facility in pairing animals, or their inhabiting a confined area, is
advantageous as a check to free crossing. Whenever and {248} wherever
selection is not practised, distinct races are not formed. When any one
part of the body or quality is not attended to, it remains either unchanged
or varies in a fluctuating manner, whilst at the same time other parts and
other qualities may become permanently and greatly modified. But from the
tendency to reversion and to continued variability, those parts or organs
which are now undergoing rapid improvement through selection, are likewise
found to vary much. Consequently highly-bred animals, when neglected, soon
degenerate; but we have no reason to believe that the effects of
long-continued selection would, if the conditions of life remained the
same, be soon and completely lost.

Man always tends to go to an extreme point in the selection, whether
methodical or unconscious, of all useful and pleasing qualities. This is an
important principle, as it leads to continued divergence, and in some rare
cases to convergence of character. The possibility of continued divergence
rests on the tendency in each part or organ to go on varying in the same
manner in which it has already varied; and that this occurs, is proved by
the steady and gradual improvement of many animals and plants during
lengthened periods. The principle of divergence of character, combined with
the neglect and final extinction of all previous, less-valued, and
intermediate varieties, explains the amount of difference and the
distinctness of our several races. Although we may have reached the utmost
limit to which certain characters can be modified, yet we are far from
having reached, as we have good reason to believe, the limit in the
majority of cases. Finally, from the difference between selection as
carried on by man and by nature, we can understand how it is that domestic
races often, though by no means always, differ in general aspect from
closely allied natural species.

Throughout this chapter and elsewhere I have spoken of selection as the
paramount power, yet its action absolutely depends on what we in our
ignorance call spontaneous or accidental variability. Let an architect be
compelled to bui