XI COLOUR AND SONG 315 
air, though its power of flight is due solely to the fact 
that it could not live without it; and the horribly 
cruel instinct that leads a cat to play with a mouse 
before killing it may be, perhaps, accounted for in 
the same way. 
Dr. Wallace seeks to explain the facts by an 
entirely different theory. But before entering into 
this, it is necessary to clear the ground a little. In- 
organic things—e.g. gold—have brilliant colours. Mere 
brightness of colour, therefore, requires no theory to 
explain it. In animals the pigments are probably 
waste products derived from their food, and if not 
employed for this purpose would be of no service 
at all. It is easy, then, to account for the bright 
colours of many deep sea animals, where there is no 
light except from the occasional phosphorescent lamps 
borne by some of the fishes. Even the bright colours 
of a butterfly may be no tax upon its strength, for, 
very probably, they can be as easily produced in the 
animal system asa dull neutral tint. The difficulty 
consists partly in the constancy of the colours and 
patterns, but, chiefly, in the long plumes, the annual 
production of which must tax the bird’s strength, and 
which are a source of danger to it. The moulting 
season brings with it an increased rate of mortality 
among the birds at the Zoological Gardens. How 
the Argus Pheasant is over-burdened by its plumes, I 
have already shown. 
With regard to the constancy of colours Dr. Wallace 
has pointed out that much may be explained on the 
principle of protective coloration. The hen-bird is 
exposed to greater danger than the cock-bird, since 
