advances, the edgings of these feathers in both sexes wear off by abrasion.
A change somewhat similar to this is found in the Great Northern Diver (Colymbus glacialis), and probably in other species as well. It is comparatively difficult to procure an individual of C. glacialis in full winter plumage (as described by Mr. Cecil Smith in his 'Birds of Somerset,' p. 540), which resembles somewhat that of the immature birds; for as soon as the feather is formed, or even while growing, it commences, by a change of colour in the feather itself, to assume the distinguishing features of the breeding plumage for the following year; it is therefore possible to find in the same individual worn feathers of last year's breeding-dress, new feathers that have not yet changed but are of a uniform bluish grey, and those which have already by change of colour assumed the white spots and dark ground of the next year's breeding dress.
Returning to the Corncrake, the similarity of the moult in both sexes, with the subsequent although practically simultaneous assumption of the breeding plumage by the male, tends to show that, in this instance, the reason of the moult is not for the assumption of the breeding dress, for in that case it ought only to be undergone by the male. The evidence before us seems to point to the fact that the moult has no connection with the colour-change, but is due to some ulterior cause of which we are at present ignorant.
The method by which this change of colour is brought about is very hard to determine, but the following notes of a microscopical examination of feathers in process of change, taken from a living bird a few minutes before examination, may be of interest.
If a feather be taken which is fully grown and in process of changing, and be microscopically examined by transmitted light, with a low power of about 66 diameters (with a higher power it was difficult to get satisfactory results by reflected light), we find that, whether the part of the feather be blue or brown, there is absolutely no difference in the arrangement of the rami or radii, and that both are opaque and show no colour. If these same parts be then examined by reflected light, the brownish part appears dull, the rami and radii being both of a light-brown
