impossible. We may, however, generally gain a clue by comparing all the
species of the same group, if it be a large one; for some of them will
probably retain, at least partially, traces of their former characters.
Instead of entering on tedious details respecting various groups, in
which striking instances of gradation could be given, it seems the best
plan to take one or two strongly marked cases, for instance that of the
peacock, in order to see if light can be thrown on the steps by which
this bird has become so splendidly decorated. The peacock is chiefly
remarkable from the extraordinary length of his tail-coverts; the tail
itself not being much elongated. The barbs along nearly the whole
length of these feathers stand separate or are decomposed; but this is
the case with the feathers of many species, and with some varieties of
the domestic fowl and pigeon. The barbs coalesce towards the extremity
of the shaft forming the oval disc or ocellus, which is certainly one
of the most beautiful objects in the world. It consists of an
iridescent, intensely blue, indented centre, surrounded by a rich green
zone, this by a broad coppery-brown zone, and this by five other narrow
zones of slightly different iridescent shades. A trifling character in
the disc deserves notice; the barbs, for a space along one of the
concentric zones are more or less destitute of their barbules, so that
a part of the disc is surrounded by an almost transparent zone, which
gives it a highly finished aspect. But I have elsewhere described (50.
Variation of Animals and Plants under Domestication, vol. i. p. 254.)
an exactly analogous variation in the hackles of a sub-variety of the
game-cock, in which the tips, having a metallic lustre, are separated
from the lower part of the feather by a symmetrically shaped
transparent zone, composed of the naked portions of the barbs. The
lower margin or base of the dark-blue centre of the ocellus is deeply
indented on the line of the shaft. The surrounding zones likewise shew
traces, as may be seen in the drawing (Fig. 54), of indentations, or
rather breaks. These indentations are common to the Indian and Javan
peacocks (Pavo cristatus and P. muticus); and they seem to deserve
particular attention, as probably connected with the development of the
ocellus; but for a long time I could not conjecture their meaning.
If we admit the principle of gradual evolution, there must formerly
have existed many species which presented every successive step between
the wonderfully elongated tail-coverts of the peacock and the short
tail-coverts of all ordinary birds; and again between the magnificent
ocelli of the former, and the simpler ocelli or mere coloured spots on
453