from surprising when no one pretends to
define the meaning of the term relation or the ulterior object of all
classification. We shall immediately see on the theory of descent how it
comes that there should be "real" and "analogical" affinities; and why
the former alone should be of value in classification--difficulties
which it would be I believe impossible to explain on the ordinary theory
of separate creations.

{440} In the corresponding passage in the _Origin_, Ed. i. p. 430,
vi. p. 591, the term _general_ is used in place of _generic_, and
seems a better expression. In the margin the author gives
Waterhouse as his authority.

{441} _Origin_, Ed. i. p. 430, vi. p. 591.

_Classification of Races or Varieties._

Let us now for a few moments turn to the classification of the generally
acknowledged varieties and subdivisions of our domestic beings{442}; we
shall find them systematically arranged in groups of higher and higher
value. De Candolle has treated the varieties of the cabbage exactly as
he would have done a natural family with various divisions and
subdivisions. In dogs again we have one main division which may be
called the _family_ of hounds; of these, there are several (we will call
them) _genera_, such as blood-hounds, fox-hounds, and harriers; and of
each of these we have different _species_, as the blood-hound of Cuba
and that of England; and of the latter again we have breeds truly
producing their own kind, which may be called races or varieties. Here
we see a classification practically used which typifies on a lesser
scale that which holds good in nature. But amongst true species in the
natural system and amongst domestic races the number of divisions or
groups, instituted between those most alike and those most unlike, seems
to be quite arbitrary. The number of the forms in both cases seems
practically, whether or not it ought theoretically, to influence the
denomination of groups including them. In both, geographical
distribution has sometimes been used as an aid to classification{443};
amongst varieties, I may instance, the cattle of India or the sheep of
Siberia, which from possessing some characters in common permit a
classification of Indian and European cattle, or Siberian and European
sheep. Amongst domestic varieties we have even something very like the
relations of "analogy" or "adaptation{444}"; thus the common and Swedish
turnip are both artificial varieties w