I went to Cambridge in the middle sixties with a very small equipment of either philosophy or science, but a very keen desire to discover _what_ I ought to think of the world, and _why_. For the history of speculation I cared not a jot. Dead systems seemed to me of no more interest than abandoned fashions. My business was with the ground-work of living beliefs; in particular, with the ground-work of that scientific knowledge whose recent developments had so profoundly moved mankind. And surely there was nothing perverse in asking modern philosophers to provide us with a theory of modern science! I was referred to Mill; and the shock of disillusionment remains with me to the present hour. Mill possessed at that time an authority in the English Universities, and, for anything I know to the contrary, in the Scotch Universities also, comparable to that wielded forty years earlier by Hegel in Germany and in the Middle Ages by Aristotle. Precisely the kind of questions which I wished to put, his Logic was deemed qualified to answer. He was supposed to have done for scientific inference what Bacon tried to do, and failed. He had provided science with a philosophy. I could have forgiven the claims then made for him by his admirers; I could have forgiven, though young and intolerant, what seemed to me the futility of his philosophic system, if he had ever displayed any serious misgiving as to the scope and validity of his empirical methods. If he had admitted, for example, that, when all had been done that could be done to systematise our ordinary modes of experimental inference, the underlying problem of knowledge still remained unsolved. But he seemed to hold, in common with the whole empirical school of which, in English-speaking countries, he was the head, that the fundamental difficulties of knowledge do not begin till the frontier is crossed which divides physics from metaphysics, the natural from the supernatural, the world of “phenomena” from the world of “noumena,” “positive” experiences from religious dreams. It may be urged that, if these be errors, they are errors shared by ninety-nine out of every hundred persons educated in the atmosphere of Western civilisation, whatever be their theological views: and I admit that it has sunk deep into our ordinary habits of thought. Apologetics are saturated with it, not less than agnosticism or infidelity. But, for my own part, I feel now, as I felt in the early days of which I am speaking, that the problem of knowledge cannot properly be sundered in this fashion. Its difficulties begin with the convictions of common sense, not with remote, or subtle, or otherworldly speculations; and if we could solve the problem in respect of the beliefs which, roughly speaking, everybody shares, we might see our way more clearly in respect of the beliefs on which many people are profoundly divided.