390 YEARBOOK OF THE DEPARTMENT OF AGRICULTURE. 



Gulf of Mexico, while the late birds in southern Texas probably 

 travel by a land route through Mexico. Similarly the dates of spring 

 arrival are earlier in northern Georgia than in southeastern Georgia 

 and northeastern Florida, indicating that the earliest migrants across 

 the Gulf of Mexico fly far inland before alighting. 



The summer warbler arrives at Edmonton, Alberta, earlier than 

 at central Montana, 400 miles south. Evidently the Edmonton birds 

 do not come from the south, neither are they from the southeast, for 

 migration is no earlier in southern Manitoba than it is in central 

 Alberta. Hence they must come from the southwest, though this 

 necessitates their crossing the main range of the Eocky Mountains, 

 which at this season is still cold and partly covered with snow. 



CONCLUSION. 



The foregoing facts show conclusively that weather conditions 

 are not the cause of the migration of birds, but that the weather, by 

 influencing the food supply, is the chief factor which determines 

 the average date of arrival at the breeding grounds. Migration is 

 undertaken in response to physiological changes in birds, and the 

 date of starting, in the case of most species, bears no relation what- 

 ever to the local weather conditions in the winter home. The weather 

 encountered en route influences migration in a subordinate way, re- 

 tarding or accelerating the birds' advance by only a few days and 

 having slight relation to the date of arrival at the nesting site. 



Local weather conditions on the day of arrival at any given locality 

 are minor factors in determining the appearance of a species at that 

 place and time. The major factors in the problem are the weather 

 conditions far to the southward, where the night's flight began, and 

 the relation which that place and time bear to the average position 

 of the bird under normal weather conditions. Many, if not most, 

 instances of arrivals of birds under adverse weather conditions are 

 probably explainable by the supposition that the flight was begun 

 under favorable auspices and that late in the night the weather 

 changed. Spring migration usually occurs with a rising temperature 

 and the movements of autumn with a falling temperature. In each 

 case the change seems to be a more potent factor than the absolute 

 degree of cold. 



The direction and force of the wind except as they are occasionally 

 intimately connected with sudden and extreme variations in tempera- 

 ture seem to have only a slight influence on migration. 



Another conclusion equally apparent is that neither the time of 

 migration, the route, nor the speed of one species can be deduced 

 from records of other species, even though closely related ; in other 

 words, each species and even each group of individuals of a species 

 is a law unto itself. 



