306 STUDIES IN BIRD-MIGRATION 



witnessed the act of birds rising on the wing to depart 

 on their nocturnal journeys ^ ; while the observations 

 made at land-stations, which may be considered to bear 

 upon the question, are surrounded by and associated 

 with elements of great uncertainty. At the Eddystone, 

 and other stations situated immediately off the south 

 coast, it seemed possible in the autumn to procure data 

 which might enable one to fix this time of embarkation 

 with some degree of accuracy. To this end I made a 

 series of careful observations on the time of first appear- 

 ance of emigrants at the lighthouse, and found that on 

 a number of occasions in October this ranged from 

 6.50 P.M. to 7.15 P.M. On the dates on which these 

 observations were made, the hour of sunset ranged from 

 5.30 P.M. to 6 P.M., but darkness did not ensue until about 

 6.15 P.M., or a little later. It is fair to assume that 

 these earliest birds to appear had only a short time 

 previously set out from localities contiguous to the shores 

 of the mainland, some 12 miles distant. Taking these 

 facts into account, I have come to the conclusion that 

 when the weather conditions are favourable, the initial 

 movement for crossing the Channel is embarked upon 

 almost immediately after darkness prevails. On no 

 occasion during each major movement witnessed did all 

 the individuals of a species appear simultaneously, though 

 sometimes several kinds arrived in company, and thus 

 the passages were a succession of arrivals of birds 

 previously observed. Here we have evidence, I think, 

 that certain of the individual emigrants had journeyed 

 from districts more or less distant ere the Channel was 

 reached on the voyage southwards. 



* Since my visit to the Eddystone, I have observed it at Fair Isle (see 

 Vol. II., p. 91). 



