106 THE RIDDLE OF MIGRATION 



vagrants, In regions such as those of the trade winds 

 which blow with marked regularity and produce 

 seasonal rains, might well convey the impression of 

 being true migrants. Their movements would take 

 on a suggestive periodicity that might, on final 

 analysis, even prove to be true migration. 



A non-competitive environment, with all require- 

 ments found in a circumscribed area ultimately 

 leads in birds, to the condition furthest removed 

 from migration — loss of the power of flight. 



The great chasm between conditions in the 

 tropics and in the far north is in no respect more 

 emphasised than in breeding dates. Many, perhaps 

 all, tropical species show a breeding rhythm but 

 there is no agreement in dates. A given species 

 may breed from January to July, another from July 

 to December, but birds' eggs may be found at all 

 times of the year. In the far north the breeding 

 season of hundreds of species is synchronous and 

 crowded into a few short weeks. There is no sea- 

 sonal selection in the tropics to compare with that of 

 the north. 



The operation of selection is in no case better in- 

 stanced than among the crepuscular nightjars. The 

 most diurnal of these on the north American conti- 

 nent, the common nighthawk {Chordeiles virginianus) 

 ranges further north than any of the other species. 



