BIRDS 1591 



tory. Those which breed in the equatorial regions are 

 the chief exceptions, and even they pass from hill to 

 valley and back again. Forms, too, which seem to 

 be constant residents of a non-tropical country are 

 in many cases known to exhibit a partial or a very 

 local migration. This is true, for instance, of the 

 common wren and the red grouse in the north of 

 Scotland. All birds breed in the colder regions of 

 their migration. Changes in food supply and the 

 temperature are the most important conditions im- 

 pelling them to shift their habitats. The general 

 trend of migration is always, as indicated, tow- 

 ard the equator in autumn, from the equator in 

 spring; but the investigations of the British Mi- 

 gration Committee have clearly shown that the 

 courses often come to be circular. The flight is the 

 more universal in a country the more marked the 

 contrast between summer and winter. The annual 

 migration from breeding areas too cold for winter 

 residence and food supply to warmer subsistence 

 areas can not be understood apart from the history 

 of climates. When the European climate was more 

 equable, it was virtually indifferent to the birds where 

 they went. As it grew colder the birds had to fly fur- 

 ther and further south every few winters. Migra- 

 tion has become an inherited habit, for they set about 

 it before the impelling conditions are directly pres- 

 ent. According to Wallace, natural selection has 

 played an important part in confirming this habit. 

 Many facts about migration are still utterly obscure. 

 The power birds have of flying straight and of re- 

 turning to the same locality is very marvelous. It 



VOL. IV. 



