66 THE MIGRATION OF BIRDS. 



wing is correlated with the habits of the bird is very 

 significantly proved by the fact that in Arctic and 

 temperate regions where migration is most practised, 

 there are few birds with the short rounded concave 

 wings of the sedentary birds of the Tropics, where 

 migration is little practised. It should also be 

 stated that the plumage of migratory birds is 

 generally close and compact ; a loose and fluffy 

 dermal covering is seldom worn by any but the 

 most sedentary birds. We may thus fairly begin 

 with the fact that birds of passage are both in the 

 formation of the wing and in the texture of their 

 plumage the best of all birds adapted for flight. 



We will now advance a step further in our 

 subject. The feathers of a bird are renewed — in 

 other words moulted — as every naturalist knows, 

 at stated periods. These periods vary a good deal 

 according to circum.stances, the only one of which 

 we need stay to discuss being the period which 

 seems correlated with migration. In many migra- 

 tory birds, certainly in all those whose flights are 

 the most extended (as in the Charadriid.e), a 

 complete double moult takes place, in spring and 

 in autumn. In all others a change is undergone in 

 autumn, with the exception of the birds comprising 

 the Laniid.e (Shrikes) and the Hiruxdixid.e 

 (Swallows), in w^iich the plumage is renewed in 

 early spring. These latter exceptions may be easily 

 explained. The birds in both these families appear 

 to belong to the Southern Hemisphere, and have 

 only within comparatively recent times extended 



