RENDAHL, BIRDS FROM THE PEARL ISLANDS. 29 



of the bird really having been obtained in the Pearl Islands 

 and say: «It is possible there may ha ve been some error in 

 the locaHty recorded». 



Thayer and Bångs presume, that it either deals with 

 an accidental occurrence of Temminck's Döve in the Pearl 

 Islands, or that the döve in question never came from this 

 locality, which seems a more likely theory, as the birds collec- 

 ted by Kellett and Wood were put into open tubes of paper 

 with the data written on the tubes and many of the skins 

 got mixed up. 



37. Chaemepelia rufipennis ruflpennis (Bonaparte). 



Columhigallina rufipennis Bångs, Auk. XVIII, 1901, 25 (San Miguel). 



Chanicepelia rufipennis Salvin and Godman, Biol, Centr.-Am. Aves III, 

 1902, 253 (ref.). 



Columhigallina rufipennis rufipennis Thayer and Bångs, BuU. Mus. Comp. 

 Zoöl. XLVI, 1905, 148 (San Mignel; Saboga). 



Chcemepelia rufipennis rufipennis Todd, Ann. Carn. Mus. VIII, 1913, 586 

 and 602 (lit. ; in list of specimens examined: San Miguel and Pearl Is- 

 lands, coll. Mus. Comp. Zoöl.) — Ridgway, Birds North Middle Am. 

 VII, 1916, 424 (San Miguel; Saboga). 



The Ruddy Ground Döve was not coUected by Dr. Bo- 

 VALLius in the Pearl Islands. 



Thayer and Bångs say, that they were unable to find 

 the slightest difference between the birds of the Pearl Islands 

 and the continent. On the eontrary Mr. Todd states the 

 females from the Pearl Islands to be much browner below than 

 the average of specimens from other localities, while no dif- 

 ference in the males is evident. He expresses his surprise, 

 that this fact was overlooked by Messrs. Thayer and Bångs, 

 and if additional material should show that the peculiarities 

 are constant, he proposes for the Ruddy Ground Döve of 

 the Pearl Islands the subspecific name of G. r. nesophila. To 

 this Mr. Ridgway adds (1. c. page 428, foot-note): «The speci- 

 mens upon which Mr. Todd based the above tentative name 

 are, as he remarks, conspicuously different from ordinary 

 females of C. r. rufipennis; but they are so closely similar to 

 immature males of that form, that I strongly suspect error in 

 the determination of sex by the coUector. An adult female 

 from Saboga Island (the reddish-tinted specimens being from 

 San Miguel Island) is precisely like adult females from the 

 mainland ». 



