18 



has been carefully studied. These lines of flight follow the courses 

 of rivers, valleys, mountain chains, or coast lines, which, to the eyes 

 of the migrants, flying high in the air, are distinctly outlined 

 for many miles, provided the night be clear, and thus form easily 

 recognizable landmarks. 



It follows from this that observers situated on the regular lines of 

 flight, as, for instance, in a large river valley, are liable to see many 

 more migrants than will be found at points on either side of them. 

 Some transient species, indeed, will likely occur regularly at the 

 former locality which will seldom or never be seen at the latter. 

 The summer resident birds also arrive in the lowlands (along lines 

 of flight) sooner than in the elevated regions of the same latitude, 

 pushing up the river valleys where the signs of spring first appear, 

 and spreading out into the hills and mountains as the climate there 

 becomes congenial. 



The influence of meteorology upon bird migration is a matter of 

 great importance, but which requires much careful investigation. In 

 general, it may be said that the bulk of migration occurs on clear 

 nights, and that birds do not start to migrate in stormy or cloudy 

 weather, though they are frequently overtaken by storms while on 

 the wing, and it is at such times that they become bewildered and fly 

 against the lighthouses, where so many annually kill themselves. 



In the spring, warm weather hastens migration and cold weather 

 retards it, while in the fall tlie reverse is the case, tliough the migra- 

 tion is more irregular and lagging and lasts for a longer period than 

 in spring (^. e., so far as each species is concerned). 



During migrations the flight of birds is by no means continuous, 

 and although some migration probably occurs on every clear night, 

 the numbers which start during or immediately after a hot spell («". e., 

 in spring) are enormous compared with those migrating in cooler 

 weather. Thus, instead of a uniform fliglit of birds we have a series 

 of " waves" or " rushes" with corresponding rests or lulls between 

 them. 



During these waves, which seem to occur on the night following 

 the sudden change of temperature and generally continue on the fol- 

 lowing night, the increase of birds at a given station is clearly 

 apparent, and to record the time of occurrence and duration of the 

 waves is an easy matter. Wliile each species has a certain average 

 time of arrival at a given station, the exact date, especially when it 

 first arrives in abundance, is dependent upon the weather. Some- 



