THE EVOLUTION OF MIGRATIONS 105 



the compass since none of them is involved as one 

 of the selective factors but the fundamental prin- 

 ciple of natural selection is still the organizing force. 

 Migrations within the tropics are equally subject 

 to the sifting of selection, but the factors involved 

 are again different. Here, where the environment 

 is favorable the year round, wandering is both fea- 

 sible and likely on a magnificent scale impossible 

 elsewhere. Movements may depend on the rainy 

 season, or the dry, the windy or the calm, but they 

 need not necessarily amount to anything more 

 than wandering. There are probably few eliminat- 

 ing factors climatic in nature and in many cases 

 it will make no difference to the welfare of the race 

 if a species breeds in a given locality a few hun- 

 dred miles north, south, east, or west. Hence a 

 random following of the food supply may become 

 the rule. The struggle for existence may be keen 

 and competition particularly severe in country 

 climatically equable the year round. Both would 

 be factors in reducing the rate of increase and de- 

 laying, perhaps even eliminating, the expansion 

 that might ultimately push the species far enough 

 north or south to induce true migration. Omnivo- 

 rous feeders might well, under such circumstances, 

 provide entirely sedentary races, while specialised 

 feeders would produce wandering forms. Such 



