1920.] Birds of the Canary Islands. 561 



Warbler, Egyptian Vulture^ Osprey, Courser, and Mi<;rat()ry 

 Quail. Ethiopian types prevail in the Cape Verde Islands, 

 as is only to be expected, and this Archipelago cannot by 

 any stretch of imagination be included in the Palsearctic 

 region. 



The curious distribution of the Shearwaters and Petrels 

 in the Atlantic islands was the subject of a paper which 

 I published in ' The Ibis ' in 1914, pp. 438-494, and I then 

 pointed out that although the same species were in certain 

 cases fouTul breeding in islands as widely separated as the 

 Cajjc Verdes and the Azores, yet this did not necessarily 

 indicate that the birds from one colony had any connexion 

 with the birds of other breeding stations. The fact that in 

 the Cape Verde Islands is found a form of Puffinus kiihli 

 \_P. k. edwardsi] perfectly distinct from the race inhabiting 

 the more northern Archipelagos, strengthens this view, as 

 also does the fact that Oceanodroma castro castro breeds in 

 all the jSTorth Atlantic islands except the Canaries, which 

 islands lie in the centre of its breeding range and form the 

 missing link in an otherwise connected chain of Atlantic 

 breeding stations. There are other instances, but tliese will 

 suffice to emphasize my point. 



The Birds of Passage. 



In the J. f. O. 1890, Koenig makes the astounding state- 

 ment "I declare plainly that the Canaries are visited quite 

 by chance by Palsearctic birds on their flight, and that there 

 can be no question of regular appearances of migratory 

 birds there ^^ ! The inaccuracy of this remark is obvious to 

 anyone who has passed even one spring and one autumn in 

 the islands. 



No fewer than 33 Palaearctic species pass regularly 

 through the Canaries in spring and autumn (probably a great 

 many more). These are the birds which follow the coast 

 of Spain and Portugal and take the extreme westerly course 

 passing far out to sea via Madeira, the Salvages, and the 



