propositions to be true without needing to examine all the individual instances.  I can know for certain that in every triangle the greater angle is subtended by the greater side, or that every event has a definite cause among earlier events, though I cannot examine all triangles or all events one by one.  This is because the postulates of geometry and the law of causality are types of order which my mind puts into the data of its knowledge in the very act of attending to them, and it is therefore certain that I shall never perceive or think anything which does not conform to these types.

I give Kant’s answer to the problem of Criticism not because I believe it to be the correct one, but to show what important consequences follow from our acceptance of a solution of this problem.  If it is true that one of the constituent elements of every piece of knowledge is a lump of crude sensation, it follows that we can have no knowledge about our own minds or souls, and still less about God, since, if there are such beings as my soul and God, at any rate neither furnishes me with sense-data.  Hence a great part of Kant’s famous Critique of Pure Reason is taken up by an elaborate attempt to show that psychology and theology contain no real knowledge.  We cannot even know whether there is any probability for or against the existence of the soul or of God, though Kant was very anxious to show that it is our duty on moral grounds to believe very firmly in both.  Now if Kant is right about this, his result is tremendously important.  If psychology and theology are wholly devoid of scientific value, it is most desirable that we should know this, not only that we may not waste time in studying them, but because it may reasonably make a very great difference to the practical ordering of our lives.  If Kant can be proved wrong, it is equally important to be convinced that he is wrong.  We may have been led by belief in his teaching to neglect the acquisition of a great deal of knowledge of high intrinsic interest, and may even have been betrayed into basing the conduct of life on wrong principles.  If, for example, we can really know something about the soul, it may be possible to know whether it is immortal or not, and it is not unreasonable to hold that certain knowledge, or even probable belief, on such a point ought to make a great difference to our choice between rival aims in life.  There is clearly much less to be said for the recommendation to ’eat and drink for to-morrow we die’ if we have reason to believe our souls immortal than if we have not, and some of us do not share Mr. Russell’s view that Philosophy is called upon to abdicate what the Greeks thought her sovereign function, the regulation of life.  It is true that Kant convinced himself that it is a moral duty to act as if we knew the truth of doctrines for or against which we cannot detect the slightest balance of probability.  But the logically sound inference from Kant’s premisses would be that,