between the overhead light which is their ceiling, and the underfoot earth which is their floor. The higher you mount from the earth, and consequently the nearer you get to the outside of this zone of detached objects, the more do all things that you have to look at come between you and the brown earth—and the more these things resemble the earth in color the less you notice them. Contrariwise (let the world for the first time notice this Contrariwise, which would have saved Roosevelt and others so much erroneous writing), contrariwise, the nearer you get to the bottom of this layer of inhabitants of the atmosphere, i. e., close to the earth's surface, the more you have only things above you to look at; and the more these things are bright like the sky the less you see them, and the darker they are the more you see them. This means that the nearer to earth's surface an animal lives, the larger will be the proportion of neighbors he detects by seeing them dark against the sky. Only such neighbors as are no taller than he will escape coming between him and his sky.
After years of trying to bring home to naturalists this great fact, which wholly suffices to account for white top-patterns in general, I know to my sorrow that nothing short of seizing them all and binding them to the ground, so that their eyes will be as low as a mouse's, will ever cause the truth to spring into vital existence for them! I have demonstrated to many audiences on both sides of the Atlantic where scarcely one man would consent actually to go down on the turf and let this immense fact rush upon his consciousness.
Everywhere, in every situation, it is the rule that animals are colored like the background that most concerns their feeding and escaping attack. Sea-birds, in general, are either all ocean-and sky-colored, or this with the addition of the color of the rocks they breed and roost on. Animals living between bare snow and sky are white. (I have shown by ocular demonstration the wonderful aid that this white background-matching gets from the small black marks commonly worn by these white species.) Sand-dwellers, on land or in the sea, are sand-colored. One can find pictured on a locust the peculiarities of the special type of ground he lives on, be it rock, or sand, or meadow. Everywhere, in every case where the animal's background is most unvarying, as in the above-cited instances, the animal's colors are at least amply accounted for by their matching of this background.
Now why is it that even after people intellectually perceive that a terrestrial mouse or lizard sees almost all sizable neighbors against the light, and detects them least when they look lightest—why is it that these people go on failing tothat white upper slopes on all species which need to avoid the sight of these mice and lizards are just as necessary to their wearers as is the brownness of the mouse or lizard, whose enemy looks at him from above downward and consequently sees