The power of flight is an excellent example of adaptation. It has been evolved independently in Pterodactyls, Bats, and Birds. In the two first groups, and to a slight degree in the third, the expanse of the wing is formed by an extension of the skin into a thin membrane, supported by the fore-limbs. It is not necessary to argue in detail that the evolution of this membrane and of the modifications of bones and muscles by which it is supported and moved, can be satisfactorily explained on the theory that modifications due to mechanical and functional stimulation are ultimately inherited. In birds, however, the surface of the wing is supplied chiefly by feathers, and consideration of the matter affords no reason for supposing that the evolution of feathers was due to any external or functional stimulation. It is often stated that the feathers of birds are a modification of the epidermic scales of reptiles, but investigation does not fully confirm this statement. The reptilian scales are retained on the tarso-metatarsal region of the leg in the majority of birds, and it would be expected, if the view just quoted were correct, that a transition from scales to feathers would be visible at the ankle-joint. This, however, is not the case. In fowls some breeds have scaly shanks and others feathered. In those with scaly legs I have found cases in winch, in the chicks, there were two or three very minute feathers, and I have examined these microscopically by means of sections of the skin. The result was to show that the minute feathers were not a prolongation of the tips or edges of the scales, but arose from follicles between the scales. The scale is flat and is a fold of the epidermis not arising from an invaginated follicle. The feather, on the other hand, is a tubular structure arising from a papilla at the base of a deep follicle extending inwards from the surface of the skin. As the feather grows the papilla grows with it. This papilla consists of vascular dermal, _i.e._ mesodermic tissue, and if the feather is pulled out during growth bleeding occurs. The epidermic horny tube splits posteriorly towards the apex of the feather, and is divided into rachis and barbs, and thus the dermal tissue within, by this time dead and dry, is exposed and is shed. Every feather is in fact an open wound, and is perhaps the only other case, in addition to that of the antlers of stags, in which vascular mesodermic tissue is normally shed in such considerable quantities. When the development of the feather is complete, growth gradually ceases, the proximal part of the feather remains tubular and does not split, and the vascular tissue within dies, shrivels, and dries up, forming the pith of the quill When the papilla recommences to grow the old feather is pushed out, and this process causes the moult. It would appear, therefore, that the feather must have been evolved, not by a continuous modification from the scale but by a development of a new kind