erence will be intensified by natural selection as a means of
identification and recognition by members of the same variety or
incipient species. It has also been observed that each differently
coloured variety of wild animals, or of domesticated animals which have
run wild, keep together, and refuse to pair with individuals of the
other colours; and this must of itself act to keep the races separate as
completely as physical isolation.

_On the Advance of Organisation by Natural Selection._

As natural selection acts solely by the preservation of useful
variations, or those which are beneficial to the organism under the
conditions to which it is exposed, the result must necessarily be that
each species or group tends to become more and more improved in relation
to its conditions. Hence we should expect that the larger groups in each
class of animals and plants--those which have persisted and have been
abundant throughout geological ages--would, almost necessarily, have
arrived at a high degree of organisation, both physical and mental.
Illustrations of this are to be seen everywhere. Among mammalia we have
the carnivora, which from Eocene times have been becoming more and more
specialised, till they have culminated in the cat and dog tribes, which
have reached a degree of perfection both in structure and intelligence
fully equal to that of any other animals. In another line of
development, the herbivora have been specialised for living solely on
vegetable food till they have culminated in the sheep, the cattle, the
deer, and the antelopes. The horse tribe, commencing with an early
four-toed ancestor in the Eocene age, has increased in size and in
perfect adaptation of feet and teeth to a life on open plains, and has
reached its highest perfection in the horse, the ass, and the zebra. In
birds, also, we see an advance from the imperfect tooth-billed and
reptile-tailed birds of the secondary epoch, to the wonderfully
developed falcons, crows, and swallows of our time. So, the ferns,
lycopods, conifers, and monocotyledons of the palaeozoic and mesozoic
rocks, have developed into the marvellous wealth of forms of the higher
dicotyledons that now adorn the earth.

But this remarkable advance in the higher and larger groups does not
imply any universal law of progress in organisation, because we have at
the same time numerous examples (as has been already pointed out) of the
persistence of lowly organised forms, and also