One of the first points that strikes even a casual observer is that amongst the sea-anemonies the colouration is extremely variable, even in the same species and in the same locality. This is in strong contrast to what we generally find amongst the higher organisms, such as insects and birds; for though considerable variation is found in them, it does not run riot as in the anemonies. It would almost appear as if the actual colour itself was of minor importance, and only the pattern essential; the precise hue is not fixed, is not important, but the necessity of colour of some sort properly arranged is the object to be attained. Whether this idea has a germ of truth in it or not, it is hard to say, but when we take the fact in connection with its occurrence just where opacity begins, connecting this with the transparency of the lower organisms, and the application of vivid colour to their internal organs, one seems to associate the instability of the anemony's colouring with the transference of colour from the interior to the exterior. Certain it is, that vivid colour never exists in the interior of opaque animals; it is always developed under the influence of light. The white bones, nerves and cartilages, and the uniform red of mammalian muscles, are not cases of true decorative colouring in our sense of the term, for all bodies must have some colour. All bone is practically white, all mammalian muscle red, but for these colours to be truly decorative, it would be necessary for muscles of apparently the same character often to be differently tinted, just as the apparently similar hairs on a mammal, and scales on an insect, are variously painted. This we do not find, for the shaft-bones and plate-bones, and even such odd bones as the hyoid are all one colour; and no one would undertake to tell, by its hue, a piece of striped from a piece of unstriped muscle. Decorative colouring _must_ be external in an opaque animal; it _may_ be internal in a transparent one. The connection thus shown between decoration and transparency seems to suggest that hypodermal colour is the original, and epidermal the newer scheme: that the latter was derived from the former. This agrees with Haagen's shrewd hint that all mimetic colour was originally hypodermal. Certain it is that the protective colour that is still under personal control, as in the chameleon, &c., is always hypodermal.