can perhaps maintain that we only know our 'ideas,' and not things as in themselves. Locke passes into elaborate distinctions between primary qualities of things, of which he holds exact representations are given, and secondary qualities, which are not in the same position; but the whole difficulty we meet with is summed up in the question whether we really _know_ substance, or whether it is that we can only hope to know ideas, and 'suppose' some substratum of reality outside. Then another difficulty is that we can hardly really know our _selves_. How can we know that the self exists; and if, like Malebranche, we speak of God revealing substance to us, how do we know about God? We cannot form any 'general' impressions, have any 'general' knowledge; only a sort of conglomeration of unrelated or detached bits of knowledge can possibly come home to us. The fact is, that modern philosophy starts with two separate and self-existent substances; that it does not see how they can be combined, and that the 'white-paper' theory is so abstract that we can never arrive at self-consciousness by its means. Berkeley followed out the logical consequences of Locke, though perhaps he hardly knew where these would carry him. He acknowledged that we know nothing but ideas--nothing outside of our mind. But he adds the conception of self, and by analogy the conception of God, who acts as a principle of causation. Whether there is necessary connection in his sensations or not, he does not say. Hume followed with criticism, scathing and merciless. He states that all we know of is the experience we have; and by experience he signifies perceptions. Ideas to him are nothing more than perceptions, and whether they are ideas simply of the mind, or ideas of some object, is to him the same. If we begin to imagine such conceptions as those of universality or necessity, of God or the self, beyond a complex of successive ideas, we are going farther than experience permits. We cannot connect our perceptions with an object, nor can we get beyond what experience allows. Custom merely brings about certain conclusions which are often enough misleading. It connects effect and cause, really different events: it brings about ideas of morality very often deceptive. We have our custom of regarding things, another has his--who can say which is correct? All we can do is, what seems a hopeless task enough--we can try to show how these unrelated particulars seem by repetition to produce an illusionary connection in our minds.