[Footnote 45: A very remarkable illustration of function in an apparently useless ornament is given by Semper. He says, "It is known that the skin of reptiles encloses the body with scales. These scales are distinguished by very various sculpturings, highly characteristic of the different species. Irrespective of their systematic significance they appear to be of no value in the life of the animal; indeed, they are viewed as ornamental without regard to the fact that they are microscopic and much too delicate to be visible to other animals of their own species. It might, therefore, seem hopeless to show the necessity for their existence on Darwinian principles, and to prove that they are physiologically active organs. Nevertheless, recent investigations on this point have furnished evidence that this is possible. "It is known that many reptiles, and above all the snakes, cast off the whole skin at once, whereas human beings do so by degrees. If by any accident they are prevented doing so, they infallibly die, because the old skin has grown so tough and hard that it hinders the increase in volume which is inseparable from the growth of the animal. The casting of the skin is induced by the formation on the surface of the inner epidermis, of a layer of very fine and equally distributed hairs, which evidently serve the purpose of mechanically raising the old skin by their rigidity and position. These hairs then may be designated as _casting hairs_. That they are destined and calculated for this end is evident to me from the fact established by Dr. Braun, that the casting of the shells of the river crayfish is induced in exactly the same manner by the formation of a coating of hairs which mechanically loosens the old skin or shell from the new. Now the researches of Braun and Cartier have shown that these casting hairs--which serve the same purpose in two groups of animals so far apart in the systematic scale--after the casting, are partly transformed into the concentric stripes, sharp spikes, ridges, or warts which ornament the outer edges of the skin-scales of reptiles or the carapace of crabs."[1] Professor Semper adds that this example, with many others that might be quoted, shows that we need not abandon the hope of explaining morphological characters on Darwinian principles, although their nature is often difficult to understand.