298. We are thus brought to the third statement, according to which philosophical relation, instead of being an idea acquired upon comparison, is itself the comparison. A comparison of ideas may seem not far removed from the simple sequence of resembling ideas; but if we examine the definition of cause, as stated above, which with Hume corresponds to the view of the relation of cause and effect as a ‘_philosophical_’ one, we find that the relation in question is neither a comparison of the related objects nor an idea which arises upon such comparison. According to his statement a comparison is indeed necessary to give us an idea of the relation--a comparison, however, not of the objects which we reckon severally cause and effect with each other, but _(a)_ of each of the two objects with other like objects, and _(b)_ of the relation of precedency and contiguity between the two objects with that previously observed between the like objects. Now, unless the idea of relation between objects in the way of cause and effect is one that consists in, or is acquired by, comparison _of those objects_, the fact that another sort of comparison is necessary to constitute it does not touch the question of its possibility. However we come to have it, however reducible to impressions the objects may be, it is not only other than the idea of either object taken singly; it is not, as an idea of resemblance might be supposed to be, constituted by the joint presence or immediate sequence upon each other of the objects. Here, then, is an idea which is not taken either from an impression or from a compound of impressions (if such composition be possible), and this idea is ‘the source of all our reasonings concerning matters of fact.’ The comparison is between present and past experience of succession of objects.