260 STUDIES IN BIRD-MIGRATION 



been traced along the south coast of England and across 

 the Channel. The birds of passage of northern origin 

 proceed southwards by both the east and west coast- 

 lines (including the Hebrides), but more especially the 

 former, and finally depart as emigrants, crossing the 

 Channel at various points between Kent and Scilly. As 

 already stated, it is possible that some of our British 

 Starlings may also participate in these emigrations by 

 joining the ranks of the Continental birds and departing 

 with them for the south. 



Some idea of the magnitude of these movements 

 may be gathered from the fact that on the night of 

 1 2th and the early morning of 13th October 1901, vast 

 numbers of Starlings, evidently of Continental origin, 

 passed the Eddystone, going southwards, for ten hours 

 and a half without a break. Sixty-seven perished at 

 the lantern, and great numbers, after striking, fell 

 over into the sea and were drowned. Some of these 

 autumnal visitors belong to a race which is characterised 

 by having a purple head and throat and green ear- 

 coverts. This form occurs on our south-eastern and 

 southern coasts, and, as I have failed to match them with 

 British and Scandinavian specimens obtained at the 

 same season, I think it is probable that these birds come 

 to us from the east.^ 



During the autumnal migratory movements Starlings 

 sometimes considerably overshoot our western limits, 

 and are observed far out in the Atlantic. At the end of 

 October 1870 a large flock was encountered 300 miles 

 west of Scilly,^ and on 23rd October 1876 one alighted 



' I captured a bird of this race on the Kentish Knock lightship. It 

 came on board from the east in an exhausted state. 

 2 Rodd, Bh-ds of Cornwall, p. 292. 



