I cannot blame the cautious reader at all, if, in the course of this book, he has begun to feel reservations about the method which the author has thought proper to follow. For by placing the dogmatic part of the work before the historical part, and thus reasons before experience, I must have created the suspicion that I was proceeding in a cunning fashion. For, although I might perhaps already have had the story in my mind, I nonetheless proceeded as if I knew nothing apart from the pure, abstract observations, my purpose being to end by surprising the completely unsuspecting reader with a welcome confirmation derived from experience. And, indeed, this is a stratagem which philosophers have very successfully deployed on a number of occasions. For it is not to be forgotten, that all knowledge has two ends by which it can be caught; an a priori end, and one which is a posteriori. Various modern students of nature, it is true, have declared that one must start with the a posteriori end; they think that the eel of science can be caught by the tail, their view being that, if enough empirical cognitions are acquired, they can then gradually ascend to higher general concepts. Whether or not this is a prudent procedure, it is far from being sufficiently learned or philosophical, for this manner of proceeding soon leads to a Why? to which no answer can be given. And this is about as creditable to a philosopher as it would be to a merchant who, when requested by a client to settle a bill of exchange, politely requested the creditor to call again some other time. Thus, to avoid this difficulty, men of penetrating understanding have started from the opposite extremity, namely, from the pinnacle of metaphysics. But this approach involves a new difficulty: one starts, I know not whence, and arrives, I know not where; the advance of the arguments refuses to correspond to experience.