Page:The World as Will and Idea - Schopenhauer, tr. Haldane and Kemp - Volume 1.djvu/73

THE WORLD AS IDEA. 31 of a stick dipped in water; the reflections in spherical mirrors, which, when the surface is convex appear somewhat behind it, and when the surface is concave appear a long way in front of it. To this class also belongs the apparently greater extension of the moon at the horizon than at the zenith. This appearance is not optical, for as the micrometre proves, the eye receives the image of the moon at the zenith, at an even greater angle of vision than at the horizon. The mistake is due to the understanding, which assumes that the cause of the feebler light of the moon and of all stars at the horizon is that they are further off, thus treating them as earthly objects, according to the laws of atmospheric perspective, and therefore it takes the moon to be much larger at the horizon than at the zenith, and also regards the vault of heaven as more extended or flattened out at the horizon. The same false application of the laws of atmospheric perspective leads us to suppose that very high mountains, whose summits alone are visible in pure transparent air, are much nearer than they really are, and therefore not so high as they are; for example, Mont Blanc seen from Salenche. All such illusions are immediately present to us as perceptions, and cannot be dispelled by any arguments of the reason. Reason can only prevent error, that is, a judgment on insufficient grounds, by opposing to it a truth; as for example, the abstract knowledge that the cause of the weaker light of the moon and the stars at the horizon is not greater distance, but the denser atmosphere; but in all the cases we have referred to, the illusion remains in spite of every abstract explanation. For the understanding is in itself, even in the case of man, irrational, and is completely and sharply distinguished from the reason, which is a faculty of knowledge that belongs to man alone. The reason can only know; perception remains free from its influence and belongs to the understanding alone.

§ 7. With reference to our exposition up to this point,