3

nobody will dispute that such determination must
depend upon our reflections concerning the nature of classification on the one hand,
and concerning the nature of science on the other; nor that the reflec-
tions concerning classification should come first.

Classification must be a topic to be systematically considered
in its proper place in this treatise. But long before we are
prepared for such a systematic study of it, we shall have
to make many classifications. I will here call attention to
such detached truths about it as many can be tolerably
well established at the outset and are, at the same time, of such a kind that they may
afford us some help in making the classifications that
we shall have to make. A classification is usually guided
by one of four kinds of considerations, as follows: Firstly, it
may be based upon the consideration of the purposes, or govern-
ing ideas, of the objects classified, and upon the different
manners in which these purposes are sought to be attained.
Secondly, it may be based upon the genealogy, and history of