moment you asked him. In short every part of the naturalist's belief on this subject has been antipodally wrong, and just for want of asking sight-specialists' (i. e., painters') aid about matters of sight.

Here is the whole indisputable fact in a nut-shell. As all painters know, two or more patterns on one thing tend to pass for so many separate things. All art schools will tell you that it takes a far-advanced pupil to be able to represent the patterns on any decorated object so true in degree of light and darkness as not to "cut to pieces" the object itself, and destroy its reality. Objects show or don't show by silhouetting dark or light or of a different color against a more distant thing-a stick against the ground, a tree against the sky. Among the million details that constitute out-door nature, every smallest detail is only distinguishable by becoming to the spectator a pattern against its background. Consequently when nature paints with marvelous accuracy on some animal a picture of a twig and the ground, the mind inevitably accepts it as twig and

ground. Could nature any further conceal an inhabitant of this scene than by painting superhumanly perfect copies of these objects on each inhabitant?

All the patterns and brilliant colors on the animal kingdom, instead of making their wearers conspicuous, are, on the contrary, pure concealing coloration, being the actual color notes of the scene in which the wearer lives, so that he really is nature's utmost picture of his background.

All colors and designs on animals are pure art, taking the lead, in the purity of their generalizations, of all human performance. Each bird's or beast's costume is pure scenery.

To discover what scene a bird represents, in cases where one plumage lasts all the year, consider what circumstances cause him the greatest need of protection, and you will commonly discover that he is colored in representation of such part of the scene, and such phase of it as the eyes that he most needs to avoid would see him against.