5

Logic II 5

classification. In case we know what the genesis of
objects has really been, they ought to be classified genetically
[Those two were?] If this be true of each of these modes of
classification, they ought to coincide; and if each be truly
understood, they will coincide. In order to test this, let
us imagine a progress in the invention of some useful art, where
we may know both the governing idea and the historical
evolution of an invention. It may be that these appear not
to coincide; for an inventor following out a certain idea
may not be able to attain his purpose, and so may revert
to a former idea, which had for some time been abandoned.
He thus produces an invention governed by quite a
different idea from that which governed the previous invention
of which his new invention was an improvement.
But in this case, that previous invention was only in
a negative way the origin of the new invention: the
true origin of it was the earlier abandoned device. Then,