gy that we say that the sun is the "cause" of daylight. The
rival theory to Locke's is that of Hume (_Treatise_, bk. i.), who
derives the conception from the unaided operation of custom. When one
object, A, has been noticed frequently to precede another object, B, an
association between A and B is generated; and by virtue of this
association, according to Hume, we say that A is the cause of B. The
weakness of this account is that many invariable successions, such as
day and night, do not make us regard the earlier members of the
successions as causing the later; while in numberless cases we assert a
causal connexion between two objects from a single experience of them.

We may proceed now to consider the validity of the conception of
causation, which has been attacked from two sides. From the side of
absolute idealism it is argued that the conception of cause, as
involving a transition in time, cannot be ultimately valid, since the
time-relation is not ultimately real. Upon this view (ably stated in
Professor Bosanquet's _Logic_, bk. i. ch. 6) the more we know of causes
and effects the less relevant becomes the time-relation and the nearer
does the conception of cause and effect approach to another conception
which is truly valid, the conception of ground and consequence. This
means that, viewed from the standpoint of science, a draught of alcohol
_causes_ intoxication in no other sense than the triangularity of a
triangle causes the interior angles to be equal to two right angles.
This argument ceases to have cogency so soon as we deny its fundamental
proposition that the time-relation is not ultimately real, but is
irrelevant from the standpoint of science. This is a sheer assertion,
contrary to all ordinary experience, which we have as much right to deny
as the absolute idealists to affirm. It is only plausible to those who
are committed to the Hegelian view of reality as consisting of a static
system of universals, a view which has long been discredited in Germany,
its native land, and is fast losing ground in England. Against the
Hegelians we must maintain that the common distinction between "ground"
and "cause" is perfectly justifiable. Whereas "ground" is an appropriate
term for the relations within a static, simultaneous system, "cause" is
appropriate to the relations within a dynamic, successive system.

From the other side the validity of causation has been attacked in the
interests of the naturalism of the me