In fact here we come upon an important point. It is sometimes said that, granting the above arguments and the partiality and defectiveness of the laws of Science, still they are approximations to the truth, and as each fresh fact is introduced the consequent modification of the old law brings us _nearer and nearer_ to a limit of rigorous exactness which we shall reach at last if we only have patience enough. But is this so? What kind of rigorous statement shall we reach when we have got _all_ the facts in? Remembering that Nature is _one_, and that if we try to get a rigorous statement for one set of phenomena (as say the lunar theory) by isolating them from the rest, we are thereby condemning ourselves beforehand to a false conclusion, is it not evident that our limit is at all times infinitely far off? If one knew all the facts relating to a given inquiry except two or three, one might reasonably suppose that one was near a limit of exactness in one's knowledge; but seeing that in our investigation of Nature we only know two or three, so to speak, out of a million, it is obvious that at any moment the fresh law arising from increased experience may completely upset our former calculations. There is a difference between approximating to a wall and approximating to the North Star. In the one case you are tending to a speedy conclusion of your labours, in the other case you are only _going in a certain direction_. The theories of Science generally belong under the second head. They mark the direction which the human mind is taking at the moment in question, but they mark no limits. At each point the _appearance_ of a limit is introduced--which becomes, like a mirage in the desert, an object of keen pursuit; but the limit is not really there--it is only an effect of the standpoint, and disappears again after a time as the observer moves. In the case of gravitation there is for the moment an appearance of finality in the law of the inverse square of the distance, but this arises probably from the fact that the law is derived from a limited area of observation only, namely the movements (at great distances from each other) of some of the heavenly bodies.[20] The Cavendish and Schehallien experiments do not show more than that the law at ordinary distances on the earth's surface does not vary _very_ much from the above; while the so-called molecular forces compel us (unless we make the very artificial assumption that a variety of attractions and repulsions co-exist in matter alongside of, and yet totally distinct from, the attraction of gravitation) to suppose very _great_ modifications of the law for small distances. In fact, as we saw of Boyle's law before--the Newtonian law is probably metaphysical--true under certain limited conditions--and the appearance of finality has been given to it by the fact that our observations have been made under such or similar conditions. When we extend our observation into quite