354 TROPICAL NATURE v 
Mr. Darwin’s Descent of Man, chapters x. to xviii, and to 
chapters iii. iv. and vii. of the first portion of the present 
volume. 
Normal Colours 
The fourth group—of normally coloured animals—includes 
all species which are brilliantly or conspicuously coloured in 
both sexes, and for whose particular colours we can assign no 
function or use.t It comprises an immense number of showy 
birds, such as kingfishers, barbets, toucans, lories, tits, and 
starlings; among insects most of the largest and handsomest 
butterflies, innumerable bright-coloured beetles, locusts, dragon- 
flies, and hymenoptera; a few mammalia, as the zebras; a 
great number of marine fishes; thousands of striped and 
spotted caterpillars ; and abundance of mollusca, star-fish, and 
other marine animals. Among these we have included some 
which, like the gaudy caterpillars, have warning colours ; but 
as that theory does not explain the particular colours or the 
varied patterns with which they are adorned, it is best to 
include them also in this class. It is a suggestive fact that 
all the brightly-coloured birds mentioned above build in holes 
or form covered nests, so that the females do not need that 
protection during the breeding season which I believe to be 
one of the chief causes of the dull colour of female birds when 
their partners are gaily coloured. This subject is fully argued 
in chapter vi. of the present volume. 
Leaving the colours of flowers to be discussed in another 
chapter, we will now consider how the general facts of colour " 
here sketched out can be explained. We have first to inquire 
what is colour, and how it is produced; secondly, what is 
known of the causes of change of colour; and, lastly, what 
theory best accords with the whole assemblage of facts. 
The Nature of Colour 
The sensation of colour is caused by vibrations or undula- 
tions of the ethereal medium of different lengths and velocities. 
The whole body of vibrations caused by the sun is termed 
1 Distinctness of marking for purposes of recognition is probably the use 
in all cases. See p. 367, and Darwinism, p. 217. 
