Let us inquire how it is produced. Some authorities say that there are certain drops of water that transmit light, while some are too compact to be translucent. Thus the brightness is the effect of the former; the shadow, of the latter; by the intermingling of the two is formed the rainbow, part of which is bright, to wit, that which admits sunlight, part darker, namely, that which has shut out the light and cast a shadow from itself over the objects nearest it. Others again deny that 4 this is so. Shade and light, they say, might be the cause if the rainbow had only two colours, and thus was made up of light and shade. But now, though there gleam a thousand diverse hues, Their changes withal elude the eyes that behold. The hues that touch seem actually one, yet the edges are quite different. In it sight detects something that is red, something that is orange, something that is blue; and there are other colours too, laid on in finest lines just like a skilful painting, so that, as the poet remarks above, it is impossible to discover whether the colours differ from one another until the last of them is compared with the first. The junction 5 of colour with colour deceives the sight: with such marvellous skill does nature starting from what is like end in what is totally unlike. What good, then, do the two alleged colours, light and shade, do in a case of this kind, when the presence of an endless variety must be accounted for? Again, certain authorities are of opinion that the following is the method of formation of the rainbow: In the quarter of the sky where rain is falling, they say, the drops of falling rain are so many mirrors; from each mirror, therefore, is reflected an image of the sun. By and by, many, in fact, countless, images, descending and crossing abruptly, are all blended together. Therefore the rainbow is just a blending of a great number of images of the sun. They appeal to the following argument in proof of this: On a clear day, say they, 6 set out a thousand basins, and they will all contain images of the sun. Or arrange single drops of water on single leaves; they will each have an image of the sun. On the other hand, an immense pond will have no more than one image. Why so? Just because every smooth surface that is fenced off, and surrounded by its own boundaries, is a mirror. Again, divide a pond of very large size into several small ponds by inserting partition walls; it will show as many images of the sun as it has divisions. Leave it as it was, spreading out to its full extent, and it will show but one reflection of him. The small extent of the liquid or 7 pond makes no manner of difference. If the surface is circumscribed, it forms a mirror. Well then, those countless drops, which are carried down by a falling shower, are so many mirrors, and contain so many reflections of the sun. To an observer right in front of them they present the appearance of being mixed up: the intervals which part