THE EVOLUTION OF MIGRATIONS 99 



to go far enough south in the fall to escape the 

 selective elements. We are familiar today with 

 the fact that many species of birds wander in the 

 fall to all points of the compass. Gulls and herons 

 provide examples on the most striking scale though 

 such everyday occurrences as the crowding of birds 

 to a suddenly available source of food supply indicate 

 extensive random wandering on the part of many 

 species and at almost all seasons. Every winter a 

 certain percentage of longspurs must have wandered 

 south and survived while others moved north, east 

 and west and perished. Winter would find the 

 southern suvivors in a region more or less crowded. 

 With the reawakening of sexual impulses in the 

 spring, spreading and dispersal would again be im- 

 posed and some individuals would inevitably return 

 to the north. From this fortuitous battledore and 

 shuttlecock state of affairs a north-and-south swing, 

 synchronous with the seasons, must sooner or later 

 have established itself. That it involved the loss 

 of incredible numbers of birds or took hundreds or 

 even thousands of years to set up is immaterial. 

 Millions of birds and millions of years have been 

 available. 



The argument assumes that neither intelligence, 

 intention nor conscious understanding of the factors 

 of the environment were involved on the part of the 



