170 THE LESSON OF EVOLUTION 



at present, suggest a use for any one of these 

 characters except the lamelte on the palate, which, 

 no doubt, are useful by retaining in the mouth the 

 small animals which they catch. The points which 

 separate the genera are chiefly the relative lengths 

 of the bones of the wing, leg, and toes; and the 

 shape of the bill and nasal tubes. These may pos- 

 sibly be of some special use to the different animals, 

 although we cannot explain them. The specific 

 characters are, principally, size and the colours of 

 the feathers, none of which have any particular use. 

 Take the largest genus, (Estrelata. Here we find 

 that four or five species are entirely dark brown, 

 while the rest are white below and brown or grey 

 above. Some have the head white, others dark. 

 In some the upper tail-coverts, or the under- wing- 

 coverts are dark, in others they are white ; and there 

 are still smaller differences. These colours cannot 

 be specially useful to each species, for all hunt to- 

 gether for food on the ocean, and all have very 

 similar habits. Neither can they be recognition 

 marks ; for the birds do not recognise each other at 

 sea, but fly singly. Indeed, one of the characters 

 which divide the species is the colour of the lower sur- 

 face of the primary feathers of the wing , which can- 

 not be seen unless the feathers are separated. At the 

 breeding season each species retires to its nesting- 

 place before courtship begins, and so recognition 

 marks cannot be necessary. 



There are also theoretical reasons for supposing 

 that, if a variation arises in a small group of indi- 

 viduals isolated from the rest of their species, the 



