266 THE MIGRATION OF BRITISH BIRDS 



computed that the migration of the Goldcrest into our 

 area extended over a period of ninety-two days, com- 

 mencing in August ! It must, however, be remarked 

 that the earlier arrivals were from the north-east, and 

 that in computing the duration of the passage of this 

 tiny species this fact is almost invariably overlooked. 

 None the less remarkable are the autumn passages of 

 Sky Larks and Starlings — vast waves of avian life that 

 only spend themselves in the remote western areas of 

 Ireland ! I have already given many instances of this 

 grand migratory movement in autumn in the Migration 

 of Birds (pp. 255-258), to which volume I would refer 

 the reader anxious for greater details. 



It has been said that this East to West migration in 

 autumn sometimes approaches us from points south of 

 east. If such is actually the case the movement is 

 entirely abnormal, as no bird whatever migrates in a 

 northerly direction in autumn. Adverse winds or bad 

 weather may have driven these migrants a little south 

 of their normal course, but the reader may rest assured, 

 if the passage is being undertaken with suitable wind 

 and weather, the birds will never appear by any chance 

 from points south of east, nor in a very marked manner 

 from points north of east. The normal trend of this 

 movement is east. 



There is also just the same cross migration in progress 

 in autumn as we have already found to be the case in 

 spring. Birds coasting south across the British Islands 

 pass almost at right angles the stream that is pouring in 

 from the east across the North Sea. Every movement 

 that is observed in spring is again repeated in autumn, 

 only the directions are exactly reversed. While all this 

 stream of migration from the cast is in progress there is 



