We therefore proceed to experiments with a subdued light, which is essential to the appearance of colour. Let a small opening be made in the window-shutter of a dark room; let the crossing sun-light which enters, be received on a surface of white paper, and we shall find that the smaller the opening is, the dimmer the light image will be. This is quite obvious, because the paper does not receive light from the whole sun, but partially from single points of its disk. 405. If we look attentively at this dim image of the sun, we find it still dimmer towards the outlines where a yellow border is perceptible. The colour is still more apparent if a vapour or a transparent cloud passes before the sun, thus subduing and dimming its brightness. The halo on the wall, the effect of the decreasing brightness of a light placed near it, is here forced on our recollection. (88.) 406. If we examine the image more accurately, we perceive that this yellow border is not the only appearance of colour; we can see, besides, a bluish circle, if not even a halo-like repetition of the coloured border. If the room is quite dark, we discern that the sky next the sun also has its effect: we see the blue sky, nay, even the whole landscape, on the paper, and are thus again convinced that as far as regards the sun, we have here only to do with a luminous image. 407. If we take a somewhat larger square opening, so large that the image of the sun shining through it does not immediately become round, we may distinctly observe the half-shadows of every edge or side, the junction of these in the corners, and their colours; just as in the above-mentioned appearance with the round opening. 408. We have now subdued a parallactic light by causing it to shine through small apertures, but we have not taken from it its parallactic character; so that it can produce double shadows of bodies, although with diminished power. These double shadows which we have hitherto been describing, follow each other in light and dark, coloured and colourless circles, and produce repeated, nay, almost innumerable halos. These effects have been often represented in drawings and engravings. By placing needles, hairs, and other small bodies, in the subdued light, the numerous halo-like double shadows may be increased; thus observed, they have been ascribed to an alternating flexile action of the light, and the same assumption has been employed to explain the obliteration of the central shadow, and the appearance of a light in the place of the dark. 409. For ourselves, we maintain that these again are parallactic double shadows, which appear edged with coloured borders and halos. 410. After having seen and investigated the foregoing phenomena, we can proceed to the experiments with knife-blades,[1] exhibiting effects which may be referred to the contact and parallactic mutual intersection of the half-shadows and halos already familiar to us. 411.