340 Mr. D. A. Bannerman on the [IWs, 



correctly including Columha jurwnicB [C laurivora auctorum] 

 from tlie Canary Arcliipelago, confused it with the Madeiran 

 bird. 1)elieving both to be the same species. This may have 

 led Cabrera to have included C. trocaz in liis list, altliougli 

 he includes correctly the three resident ])igeons of the 

 Canary Islands. 



It is, of course, quite within the bounds of possibility tliat 

 a specimen of C. trocaz might find its way to the Canaries 

 from Madeira, but that it arrives even "on rare occasions" 

 I cannot believe. Its occurrence is best looked upon with 

 considerable suspicion until a genuine example is obtained; 

 Cabrera had not a skin in liis collection. 



Bunye. C, trocaz is confined to the island of Madeira. 



Streptopelia senegalensis. Senegal Turtle-Dove. 



Turtur senegalens'ts Linn. Syst. Nat. 12th ed. 17GG, p. 283 

 — Type locality : Senegal. 



The Senegal Turtle-Dove is first recorded by BoUe 

 (J. f. O. 1857, p. 33.2) from the Canaries. He includes 

 Turtur seneyalensis from Fuerteventura, and as breeding in 

 that island. 



Cabrera notes (Catalogo, p. 53) tha^ it is of accidental 

 occurrence (" Viajera accidental""), only niet with in the 

 chestnut woods of Santa Ursula and Victoria, in Tenerife. 



Cal)rera had no skins in his collection, and I feel very 

 douljtful about the correct identification of this species. 



liavge. The typical Senegal Turtle-Dove (S. senegulensis 

 senecjalensis) inhaluts tropical Africa from Senegambia to 

 Sierra Leone, eastwards to northern Nigeria : the exact 

 extent of its range is not known. The races of the Dove 

 and their distribution have been fully dealt with by 

 Dr. Hartert (Nov. Zool. xxiii. pp. 81-83). Since then 

 Messrs. Sclater & iMackworth-Praed have still further split 

 up this group (Ibis, 19.20). Both these papers should be 

 consulted. 



