I am sorry that I cannot agree with my friend Mr. Meldola as to the insufficiency of the explanation of the white coloration of Arctic mammals and birds as due to protective adaptation, since it appears to me that there is no important group of facts in natural history of which the explanation is more complete; while on the other hand I venture (though with some hesitation) to question the basis of his counter explanation, as I am not aware of any sufficient proof that colour, per se, affects the radiation of low grade heat. At all events I feel tolerably certain that this cause, if it exists, has had no perceptible influence in determining the white colours of Arctic animals.

    I am not myself aware of there being "many species" possessing the white coloration as to which there is any difficulty in seeing the advantage they may derive from it, and there is certainly a large body of facts showing that colour is, in almost all animals and in every part of the world, more or less protective or adaptive. If the white coloration of Arctic animals stood alone, it might be thought necessary to supplement the protective theory by any available physical explanation, but we have to take account of the parallel cases of the sand-coloured desert animals and the green-coloured denizens of the ever-verdant tropical forests; and though in both these regions there are numerous exceptional cases, we can almost always see the reason of these, either in the absence of the need of protection or in the greater importance of conspicuous colouring. In the Arctic regions these exceptions are particularly instructive because in almost every case the reason of them is obvious. Let me call attention to a few which now occur to me.

    In the Arctic zone the wolf does not turn white like the fox, the reason evidently being that he hunts in packs, and concealment from his prey is not needed. So the musk-sheep and the yak, though both exposed to the extremest cold, are not white, because they are both swift and strong and need no concealment from their enemies. For the same reason neither the moose, the caribou, nor the reindeer are wholly white. Again, the glutton and the sable are dark-coloured, though inhabiting the coldest regions, and this is clearly because they are arboreal, and are better concealed from their prey by a dark than a light colour. If any useful protection from cold were to be obtained by a white coat, we should expect it to appear in such a case as the Esquimaux dogs, exposed for countless generations to the severest climate. But they gained the required warmth by a thickening of the woolly undercoat in winter, as do many other animals; and this suggests the general proposition that it will be always easier and safer to gain warmth in this way than by a modification of colour, which could certainly have but a very small effect, and might often interfere with adaptations of far greater importance. Exactly analogous cases occur among birds. The raven is, perhaps, the extremest Arctic species, but, feeding on carrion, it has no need of concealment in approaching its prey, and thus it keeps its jet black coat in the depths of the Polar winter.

    The physical explanation of melanism in butterflies and some other insects, on the other hand, seems to me to be probably a sound one; but even that requires more evidence and a fuller knowledge of the habits of the species before we can admit it as proved. It may be that the dark colouring is protective, assimilating with the surroundings of the insect when at rest, and this can only be decided by observations specially directed to the point in question.

    But even if, in this case, the dark colour has been produced in order to favour the absorption of the direct rays of the northern sun, it affords no support whatever to the totally different case in which the radiation of the obscure heat from an animal body has to be checked. I may, perhaps, be ignorant on the point, as it is rather out of my line, but I am not aware of any good experiments to determine the influence of colour per se, as distinct from the structure and surface-texture of coloured substances, on the radiation or absorption of heat of a low grade of temperature, and from a dark source. The only authority I have at hand (Ganot's "Physics," eighth edition) seems rather to imply that colour has no effect in such cases, for I find it stated, at p. 338, that the radiating power of lampblack and whitelead are identical, both being given as 100, while Indian ink is only 88. Again, at p. 352, the absorptive power of these two substances is given as 100, the source of heat being copper at 100° C., while that of Indian ink is given as 85. This seems to show that surface-texture or molecular structure is the important point, while colour has no effect whatever.

    In order to determine experimentally whether white fur or feathers are inferior to black as radiators of animal heat, it would not do to employ stained or dyed materials, because the pigments employed might affect the texture of the surface, and produce an effect not at all due to the colour. A fair test would be afforded by two samples of cloth or flannel woven from white and black natural wool respectively, the wool to be obtained from the same breed of sheep, and, if possible, from the same district, while the material must be as nearly as possible identical in weight and texture. I shall be glad to learn from Mr. Meldola, or any other of your readers, whether any experiment of this kind has been made, or whether there is any valid reason for believing that the radiation of animal heat is at all affected by colour alone.