Petrels from the North-East Pacific Ocean. 587 



there are none at present known it would Ipe very unwise to 

 dogmatise, as the few birds here studied upset most of the 

 theories advanced by previous workers while enlarging 

 our ignorance (we had nearly written knowledge) of the 

 breeding-habits of these birds beyond previous anticipation. 

 Before going any further, we might indicate the greatest 

 discovery in connection with this collection. In the 

 American Ornithologists' Union Check-List, 3rd ed. 1910, 

 p. 53, we have given as the range of 



Pujfinus griseus (Gmelin) : 

 " Oceans of Southern Hemisphere : occurs in summer on 

 the Pacific coast from southern Alaska to Lower California, 

 and on the Atlantic coast from the Gulf of St. Lawrence to 

 South Carolina : accidental in Alabama : probably breeds in 

 the South Pacific.''' 



And on p. 54 : — ■ 



PuJfiMUs tenuirostr-is (Temminck) : 



" Breeds in Southern Hemisphere : migrates north along 

 both coasts of the North Pacific to Kotzebue Sound, 

 Alaska.'' 



Pufjinus cuneatus Salviu : 

 " North Pacific Ocean. Breeds on Hawaiian Islands, and 

 islands ofi' the coast of western Mexico : occurs in migration 

 north to Bonin Islands, and Lower California." 



In the ' Birds of Australia,' vol. ii. p. 98, Mathews noted : 

 "This name {chilensis Bonaparte) must be accepted at the 

 present time in preference of N. amaurosoma Coues (1864, 

 p. 124), though later this latter name may have to be used 

 for a north Pacific breeding-form (of P. griseus), the types 

 of Coues's species having been obtained at Cape St. Lucas, 

 Lower California. I am not at all certain that the birds 

 met with in such numbers at the extremity of South America 

 are the same as those which occur off" the coast of California. 

 There always seem to be discrepancies in the dates that need 

 adjustment, and the recent discoveries of Petrels breeding 

 in the north Pacific seem to point to many yet to be made." 



SER. X. VOL. III. 2 K 



