142 
sexual selection: birds. 
ParI 1 
0$ 
w 
ball-and-socket effect, seems as inci’edible, as that 
of Raphael’s Madonnas should have been formed 
the selection of chance daubs of paint made 
long succession of young artists, not one of whom 
tended at first to draw the human figure. In ord® 
discover how the ocelli have been developed, we 6 
c tiV 
not look to a long line of progenitors, nor to vai'* 
closely-allied forms, for such do not now exist. 
ciif 
put 
fortunately the several feathers on the wing 3i y) 
to give us a clue to the problem, and they pro' 6 
demonstration that a gradation is at least possible & 
a mere spot to a finished ball-and-socket ocellus. 
The wing-feathers, bearing the ocelli, are covered v 
dark stripes or rows of dark spots, each stripe or 1 ^ 
running obliquely down the outer side of the shaft ,, 
an ocellus. The spots are generally elongated . 
transverse line to the row in which they stand, J, 
often become confluent, either in the line of the r°' 
and then they form a longitudinal stripe — or t Jl1 ^ 
versely, that is, with the spots in the adjoining 
and then they form transverse stripes. A spot s °; ^ 
times breaks up into smaller spots, which still stau c 
their proper places. ]). 
It will be convenient first to describe a perfect b^j, 
and-socket ocellus. This consists of an intensely b jjv 
circular ring, surrounding a space shaded so as e%a . C , c \i 
to resemble a ball. The figure here given has 1 
cut cannot exhibit the exquisite shading of the or'r* 1 ^ 
The ring is almost always slightly broken or intern 1 !’ 
(see fig. 56) at a point in the upper half, a little h 1 ^ 
right of and above the white shade on the eacl ° 0 v 
ball ; it is also sometimes broken towards the baS e t 
the right hand. These little breaks have an imp 01 '^ 
meaning. The ring is always much thickened, with 
edges ill-defined towards the left-hand upper c °’ 
admirably drawn by Dir. Ford, and engraved, but : 
