two strong arguments against this theory. We have already seen how generally bright coloration is wanting in desert animals, yet here heat and light are both at a maximum, and if these alone were the agents in the production of colour, desert animals should be the most brilliant. Again, all naturalists who have lived in tropical regions know that the proportion of bright to dull coloured species is little if any greater there than in the temperate zone, while there are many tropical groups in which bright colours are almost entirely unknown. No part of the world presents so many brilliant birds as South America, yet there are extensive families, containing many hundreds of species, which are as plainly coloured as our average temperate birds. Such are the families of the bush-shrikes and ant-thrushes (Formicariidæ), the tyrant-shrikes (Tyrannidæ), the American creepers (Dendrocolaptidæ), together with a large proportion of the wood-warblers (Mniotiltidæ), the finches, the wrens, and some other groups. In the eastern hemisphere, also, we have the babbling-thrushes (Timaliidæ), the cuckoo-shrikes (Campephagidæ), the honey-suckers (Meliphagidæ), and several other smaller groups which are certainly not coloured above the average standard of temperate birds.
Again, there are many families of birds which spread over the whole world, temperate and tropical, and among these the tropical species rarely present any exceptional brilliancy of colour. Such are the thrushes, goatsuckers, hawks, plovers, and ducks; and in the last-named group it is the temperate and arctic zones that afford the most brilliant coloration.
The same general facts are found to prevail among insects. Although tropical insects present some of the most gorgeous coloration in the whole realm of nature, yet there are thousands and tens of thousands of species which are as dull coloured as any in our cloudy land. The extensive family of the carnivorous ground-beetles (Carabidæ) attains its greatest brilliancy in the temperate zone; while by far the larger proportion of the great families of the longicorns and the weevils, are of obscure colours even in the tropics. In butterflies, there is undoubtedly a larger proportion of brilliant colour in the tropics; but if we compare families which are almost equally developed over the globe—as the Pieridæ or
