342 TROPICAL NATURE v 
It is true that brilliant tropical birds mostly belong to 
groups which are wholly tropical—as the chatterers, toucans, 
trogons, and pittas; but as there are perhaps an equal num- 
ber of tropical groups which are wholly dull-coloured, while 
others contain dull and bright-coloured species in nearly equal 
proportions, the evidence is by no means strong that tropical 
light and heat have anything to do with the matter. But 
there are other groups in which the cold and temperate zones 
produce finer-coloured species than the tropics. Thus the 
arctic ducks and divers are handsomer than those of the 
tropical zone ; while the king-duck of temperate America and 
the mandarin-duck of North China are the most beautifully 
coloured of the whole family. In the pheasant family we 
have the gorgeous gold and silver pheasants in North China 
and Mongolia, and the superb Impeyan pheasant in the tem- 
perate North-Western Himalayas, as against the peacock and 
fire-backed pheasants of tropical Asia. Then we have the 
curious fact that most of the bright-coloured birds of the 
tropics are denizens of the forests, where they are shaded 
from the direct light of the sun, and that they abound near 
the equator, where cloudy skies are very prevalent ; while, on 
the other hand, places where light and heat are at a maxi- 
mum have often dull-coloured birds. Such are the Sahara 
and other deserts, where almost all the living things are 
sand-coloured ; but the most curious case is that of the Gala- 
pagos islands, situated under the equator, and not far from 
South America, where the most gorgeous colours abound, but 
which are yet characterised by prevailing dull and sombre - 
tints in birds, insects, and flowers, so that they reminded Mr. 
Darwin of the cold and barren plains of Patagonia rather 
than of any tropical country. Insects are wonderfully 
brilliant in tropical countries generally, and any one looking 
over a collection of South American or Malayan butterflies 
would scout the idea of their being no more gaily-coloured 
than the average of European species, and in this he would 
be undoubtedly right. But on examination we should find 
that all the more brilliantly-coloured groups were exclusively 
tropical, and that where a genus has a wide range there is 
little difference in coloration between the species of cold and 
warm countries, Thus the European Vanessides, including 
