Nor must we in all cases alike demand the reason why; sometimes it is enough if the undemonstrated fact be fairly pointed out, as in the case of the starting-points or principles of a science. Undemonstrated facts always form the first step or starting-point of a science; and these starting-points or principles are arrived at some in one way, some in another⁠—some by induction, others by perception, others again by some kind of training. But in each case we must try to apprehend them in the proper way, and do our best to define them clearly; for they have great influence upon the subsequent course of an inquiry. A good start is more than half the race, I think, and our starting-point or principle, once found, clears up a number of our difficulties.

We must not be satisfied, then, with examining this starting-point or principle of ours as a conclusion from our data, but must also view it in its relation to current opinions on the subject; for all experience harmonizes with a true principle, but a false one is soon found to be incompatible with the facts.