in such a case is to be sought in utility. The usefulness of color in animate nature as an inspirer and satisfier of our own æsthetic needs and capacities, or of color patterns as means whereby we may distinguish and recognize various sorts of animals and plants, is a usefulness which may be answer enough to the passing poet on the one hand, and to the old-line Linnæan systematist on the other, but it is, of course, no answer to science. Science demands a usefulness to the color-bearing organisms themselves: and a usefulness large and serious enough to be the sufficient cause for so highly specialized and amazing a development.

The explanations of some of the color phenomena of animals are obvious: some uses we recognize quickly as certain, some as probable, some as possible. Some colors are obviously there simply because of the chemical make-up of parts of the insect body. That gold is yellow, cinnabar red, and certain copper ores green or blue, are facts which lead us to no special inquiry after significance: at least, not after significance based on utility. If an insect has part of its body composed of or containing a substance that is by its very chemical and physical constitution always red or blue or green, we may be content with knowing that, and not be too insistent in our demand to the insect to show cause, on a basis of utility, for being partly red or blue or green. And even if this red or blue be disposed with some symmetry, some regularity of repetition, either segmentally or bilaterally, this we may well attribute to the natural segmental and bilaterally symmetrical repetition of similar body parts. Some color and some color pattern, then, may be explicable on the same basis as the color of a mineral specimen or of a tier of bricks.

But no such explanation will for a moment satisfy us as to the presence of and arrangement of colors in the wings of Kallima, the dead leaf butterfly, or in Phyllium, the green leaf phasmid, or in the butterfly fish, Chaetodon, or in the lichen. spider, or in the chameleon with its changing tints, or in any one of a score of other more or less familiar forms whose color pattern makes, even on the casual observer, an insistent demand for rational explanation.

Certain uses of color seem apparent: the colored eye flecks or pigment spots of many of the lower animals presumably serve their possessors as organs by which to distinguish the

presence or absence of light, by virtue of their capacity to absorb light and thus stimulate the specially sensitive cells composing them. And the pigment or absence of it (dark or light color) in the fur and plumage of certain mammals and birds may perhaps serve to absorb or to reflect the sun's rays so as to help keep warm or cool the animals thus colored. But such explanations of animal colors can obviously apply to but few cases. Very plainly color, and especially pattern, has its significance if anywhere in connection with certain special relations of animals to other animals and to the world generally.

So, ever since the days of Darwin, two general categories of such significance or explanation of color and pattern have been in the minds of naturalists. One of these is the significance attributed to color pattern by the theory of sexual selection; the other is that attributed to it by the general theory or group of theories of protective resemblance, recognition, warning, directive, and mimetic coloration, etc. Of these two general explanations, one has steadily lost ground since Darwinian and early post-Darwinian days, while the other has slowly but steadily gained adherents and has been extended to cover more and more cases of animal ornamentation. Of the theory of sexual selection it must be said that it certainly cannot explain the conditions of secondary sexual differences, including colors and patterns, in many groups of animals, and it has really not been proved to explain them in any single group, although in the case of birds and mammals it seems possible that the theory is applicable: at least no other explanation of equal validity has yet been presented. Of the specialization of color and pattern for the sake of protecting the animal by making it so harmonize or fuse with the usual environment as to be indistinguishable, or by making it simulate with sufficient fidelity some particular part of its surroundings as a green or dead leaf, a twig, the dropping of a bird, a bit of lichen or what not, or by making it mimic some other animal notoriously well defended by sting or fangs or ill-tasting body, so that the otherwise defenseless mimicker is mistaken by its enemies for the defended mimicked kind of animal-of this specialization and utility of color and pattern, evidence for its reality is gradually accumulating to convincing amount. And it is of this sort of color and pattern specialization that the brief discussion to follow will be devoted.

The green katydid singing in the tree-top or shrubbery is readily known to be there by its music, but just which bit of green that we see is katydid, and which is leaf, is a matter to be decided only by unusually discriminating eyes. The clacking locust, beating its

black wings in the air, is conspicuous enough; but after it has alighted on the ground it is invisible, or, rather, visible but indistinguishable : its gray and brown mottled color pat

FIG. 248.-Katydid, Cyrtophyllis crepitans, from the West Indies, with green body and wings resembling the leaves among which it lives. (After Sharp.)

tern is simply continuous with that of the soil. The green larvæ of the Pierid butterflies lying longitudinally along green grasses simply merge into the color scheme of their environment. The gray moths rest unperceived on the bark of the

tree trunk. Hosts of insect kinds do really harmonize with the color pattern of their usual environment, and by this correspondence in shade and marking, are difficult to perceive for what they are. Now if the eyes that survey the green foliage or run over the gray bark are those of a preying bird, lizard, or other enemy, it is quite certain our reason tells us so insistently-that this possession by the insect of color and pattern tending to make it indistinguishable from its immediate environment is advantageous to it-advantageous to the degree often of saving its life. Now such a use of color and pattern is obviously one which can be widespread through the insect class, and may be, to many species which lead lives exposed to the attacks of insectivorous animals, of large-even of life and death-importance. And naturalists,

most of them at least, believe that this kind of usefulness is real, and that it is the principal clew to the chief significance of color and pattern-and this not alone in the case of insects, but of most other animals as well.

From this point of view, namely, that color patterns may be of advantage in the struggle for existence, just as strength, swiftness, and other capacities and conditions are, the specialization and refinement, all the wide modification and variety of colors and patterns are explicable by the hypothesis of their gradual development in time through the natural selection of fortuitous advantageous variations. On this basis, such special instances of resemblance to particular parts of the environment, as that shown by Kallima in its likeness to a dead leaf, and Diapheromera in its simulation of a dry, leafless twig, are simply the logical extremes of such a line of specialization.

But the nature observer may be inclined to ask how such brilliant and bizarre colors as those of the swallowtail butterflies and the tiger-banded caterpillars of Anosia can be included in any category of "protective resemblance" patterns. They are not so included, but are explained ingeniously by an added hypothesis called that of "warning colors," while for the striking similarities of pattern often noted between two unrelated conspicuously colored species still another hypothesis is proposed. In these cases it is not concealment that the color pattern effects, but indeed just the opposite. Since the pioneer studies of Bates and Wallace and Belt, naturalists have been • observing and experimenting and pondering these exposing, as well as these concealing, conditions of color and pattern, and they have proposed several theories or hypotheses explanatory of the various conditions. These hypotheses are plausible; but they are much more than that: they are each more or less well backed up by observation and experiment, and some of them have gained a large acceptance among naturalists. Both the reasoning and observed facts on which these hypotheses rest are based on the usefulness of the colors and patterns to the animals in their relation to the outside world. And the influence of advantage and natural selection is given the chief credit for determining the present-day conditions of these colors and patterns.

Before, however, we take up these hypotheses, defining

them and looking over some of the evidence adduced for their support, as well as some of the criticism leveled at them, we may advisedly look to the actual physical causation of color in animals. Whatever the use or significance of color, our understanding of this use must be based on a knowledge of the method or modes of its actual production.

Color in organisms is produced as color in inorganic nature is. Certain substances have the capacity of selective absorption of light rays, so that when white light falls on them, certain colors (light waves of certain length) are absorbed, while certain others (light waves of certain other lengths) are reflected. An object is red because the substance of which it is (superficially) composed, reflects the red rays and absorbs the others. Certain other objects or substances may produce color (be colored) because of their physical rather than their chemical constitution; their surface may be composed of superposed lamellæ, or it may be so striated or scaled that the various component rays of white light are reflected, refracted, and diffracted in such varying manner (at different angles and from different depths) that complex interference effects are produced, resulting in the practical extinguishing of certain colors (waves of certain length) or the reflection of some at angles so as not to fall on the eye of the observer, and so on. Such colors will change with changes in the angle of observation, and are the so-called metallic or iridescent colors. These two categories of color have been aptly called chemical and physical: chemical color depending on the chemical make-up of the body, physical on its structural or physical make-up. As a matter of fact we shall find that most animal colors are due to a combination of these two kinds.

(Substances that produce color by virtue of their capacity to absorb certain colors, and reflect only certain others, we may call, in our discussion of color production, "pigments "; and "pigmental" may be used as practically synonymous with "chemical" in referring to colors thus produced, while "structural" may be used as synonymous with "physical" in referring to colors dependent on superficial structural character of the insect body. For colors produced by the coöperation of both pigment and structure, "combination" or "chemicophysical" may be used as a defining name.)

Now in all animals, color depends on the presence and ar