warranted in calculating the relation of the gas in the balls to the weight of the atmosphere, and thus to discover the velocity of movement of his machine, although he did not know yet whether he would ever be able to discover a gas sufficiently lighter than air; and Archimedes was able to calculate the machine by which he could move the globe out of its place, although he knew well enough that he could find no place beyond the attraction of the earth from which to operate with his machine. Thus, also, with our science of knowledge. It is not as such something which exists independent of us, but rather something which must first be produced by the freedom of our mind, working in a certain direction, that is, if there is such a freedom, which, of course, can also not be known as yet. Let us determine this direction in advance, and obtain a clear conception of what is to be our work. Whether we can produce it or not will appear from the fact whether we do produce it, but this is not yet our purpose. We at present merely wish to see what it really is we intend to do.
1st. First of all, the described science is to be a science of science generally. Every possible science has one fundamental principle, which can not be proven in it, but must be certain in advance of it. But where, then, is this first principle to be proven? Evidently in that science which is to be the ground of all possible sciences. In this respect the science of knowledge would have a twofold object: Firstly, to show the possibility of fundamental principles
