tiful affs and collars. The tail-feathers are frequently increased in length; as we see in the tail-coverts of the peacock, and in the tail itself of the Argus pheasant. With the peacock even the bones of the tail have been modified to support the heavy tail-coverts." The body of the Argus is not larger than that of a fowl; yet the length from the end of the beak to the extremity of the tail is no less than five feet three inches," and that of the beautifully ocellated secondary wing-feathers nearly three feet. In a small African night-jar (Cosmetornis vexillarius) one of the primary wing-feathers, during the breeding season, attains a length of twenty-six inches, while the bird itself is only ten inches in length. In another closely allied genus of night-jars, the shafts of the elongated wing-feathers are naked, except at the extremity, where there is a disk." Again, in another genus of night-jars, the tail-feathers are even still more prodigiously developed. In general the feathers of the tail are more often elongated than those of the wings, as any great elongation of the latter impedes flight. We thus see that in closely allied birds ornaments of the same kind have been gained by the males through the development of widely different feathers.

It is a curious fact that the feathers of species belonging to very distinct groups have been modified in almost exactly the same peculiar manner. Thus the wing-feathers in one. of the above-mentioned night-jars are bare along the shaft, and terminate in a disk; or are, as they are sometimes called, spoon or racket-shaped. Feathers of this kind occur in the tail of a motmot (Eumomota superciliaris), of a kingfisher, finch, humming-bird, parrot, several Indian drongos (Dicrurus and Edolius, in one of which the disk stands vertically), and in the tail of certain birds of paradise. In these latter birds, similar feathers, beautifully

66 Dr. W. Marshall, "Ueber den Vogelschwanz," ibid., B. I. Heft 2, 1872. 67 Jardine's "Naturalist Library: Birds," vol. xiv. p. 166.

68 Sclater, in the "Ibis," vol. vi., 1864, p. 114. Livingstone, "Expedition to the Zambesi," 1865, p. 66.

ocellated, ornament the head, as is likewise the case with some gallinaceous birds. In an Indian bustard (Sypheotides auritus) the feathers forming the ear-tufts, which are about four inches in length, also terminate in disks." It is a most singular fact that the motmots, as Mr. Salvin has clearly shown, give to their tail-feathers the racket-shape by biting off the barbs, and, further, that this continued mutilation has produced a certain amount of inherited effect.

70

Again, the barbs of the feathers in various widely distinct birds are filamentous or plumose, as with some herons, ibises, birds of paradise, and Gallinaceæ. In other cases the barbs disappear, leaving the shafts bare from end to end; and these in the tail of the Paradisea apoda attain a length of thirty-four inches;" in P. Papuana (Fig. 47) they are much shorter and thin. Smaller feathers when thus denuded appear like bristles, as on the breast of the turkeycock. As any fleeting fashion in dress comes to be admired by man, so with birds a change of almost any kind in the structure or coloring of the feathers in the male appears to have been admired by the female. The fact of the feathers in widely distinct groups having been modified in an analogous manner, no doubt depends primarily on all the feathers having nearly the same structure and manner of development, and consequently tending to vary in the same manner. We often see a tendency to analogous variability in the plumage of our domestic breeds belonging to distinct species. Thus topknots have appeared in several species. In an extinct variety of the turkey, the topknot consisted of bare quills surmounted with plumes of down, so that they somewhat resembled the racket-shaped feathers above described. In certain breeds of the pigeon and fowl the feathers are plumose, with some tendency in the shafts to be naked. In the Sebastopol goose the scapular feathers

69 Jerdon, "Birds of India," vol. iii. p. 620.

70 "Proc. Zoolog. Soc.," 1873, p. 429.

71 Wallace, in "Annals and Mag. of Nat. Hist.," vol. xx., in his "Malay Archipelago," vol. ii., 1869, p. 390.

are greatly elongated, curled, or even spirally twisted, with the margins plumose.

In regard to color hardly anything need here be said, for every one knows how splendid are the tints of many birds,

19 See my work on "The Variation of Animals and Plants Under Domestica tion," vol. i. pp. 289, 293.

and how harmoniously they are combined. The colors are often metallic and iridescent. Circular spots are sometimes surrounded by one or more differently shaded zones, and are thus converted into ocelli. Nor need much be said on the wonderful difference between the sexes of many birds. The common peacock offers a striking instance. Female birds of paradise are obscurely colored and destitute of all orna ments, while the males are probably the most highly decorated of all birds, and in so many different ways, that they must be seen to be appreciated. The elongated and golden orange plumes which spring from beneath the wings of the Paradisea apoda, when vertically erected and made to vibrate, are described as forming a sort of halo, in the centre of which the head "looks like a little emerald sun with its rays formed by the two plumes."" In another most beau tiful species the head is bald, “and of a rich cobalt blue, crossed by several lines of black velvety feathers.” “

Male humming-birds (Figs. 48 and 49) almost vie with birds of paradise in their beauty, as every one will admit who has seen Mr. Gould's splendid volumes, or his rich collection. It is very remarkable in how many different ways these birds are ornamented. Almost every part of their plumage has been taken advantage of, and modified; and the modifications have been carried, as Mr. Gould showed me, to a wonderful extreme in some species belonging to nearly every sub-group. Such cases are curiously like those which we see in our fancy breeds, reared by man for the sake of ornament: certain individuals orig. inally varied in one character, and other individuals of the same species in other characters; and these have been seized on by man and much augmented-as shown by the tail of the fan-tail pigeon, the hood of the jacobin, the beak and wattle of the carrier, and so forth. The sole difference be

18 Quoted from M. de Lafresnaye, in “Annals and Magazine of Natural History," vol. xiii., 1854, p. 157; see also Mr. Wallace's much fuller account in vol. xx., 1857, p. 412, and in his "Malay Archipelago.

14 Wallace, “The Malay Archipelago,” vol. ii., 1869, p. 405.

tween the cases is that in the one the result is due to man's selection, while in the other, as with humming-birds, birds

FIG. 48.-Lophornis ornatus, male and female (from Brehm).

of paradise, etc., it is due to the selection by the females of the more beautiful males.

I will mention only one other bird, remarkable from the extreme contrast in color between the sexes, namely, the famous bell-bird (Chasmorhynchus niveus) of South America,

Descent-VOL. II.-4