LIFE HISTORIES OF NORTH AMERICAN PETRELS AND PELICANS. 39 



some good evidence to the contrary. Stejneger (1885) suggests 

 that: 



There seems to be a decided difference in tlie geograpliical range of the two 

 phases in botli oceans. It appears that the dark phase in both instances is a 

 particularly western bird, while the light-colored ones seem to have a more 

 eastern distribution. 



If the dark form had occurred breeding in Iceland, where Faber found the 

 white one exceedingly numerous, he could scarcely have escaped mentioning 

 It. Nor does it seem to have been found in Saint Kilda by John INIacgillivray, 

 and the form at present breeding on the Faer Islands seems also to be the 

 unmixed light phase. In the Pacific a similar distribution obtains, the dark 

 form being comparatively scarce on the American side, while it is by far the 

 predominating form on the Asiatic shore, at least as far south as Kamtschatka. 



The dark phase has not been recorded from any of the breeding 

 places in Bering Sea, which is the basis for the belief that the so- 

 called Rodgers fulmar has no dark phase. If the dusky birds repre- 

 sent only a dark phase, it seems strange that they should have such 

 a different breeding range, which is not usually the case in well- 

 known color phases in other species. Furthermore, where the breed- 

 ing ranges of the two phases overlap, they do not intermingle and 

 have never been seen mated together. Stejneger (1885) says: 



The dark phase was found by me on the Commander Islands in countless 

 numbers. In the colonies breeding on Bering Island not a single light bird 

 was to be seen, and the same was the case at the rookeries on the northern 

 part of Copper Island — for example, that close to the village. At Glinka, near 

 the southern extremity of the latter island, were found a few small white 

 colonies, but the percentage of the light-colored birds was quite trifling, as I 

 estimated it to be between 1 and 5 per cent. 



In the light of what evidence we have, it seems to me more logical 

 to recognize a light and a dark species, each perhaps with Atlantic 

 and Pacific subspecies, and to eliminate rodgersi as not separable 

 from the light bird of the Pacific Ocean. 



Nesting. — Very little has been published about the breeding habits 

 of the Pacific fulmar. It is supposed to breed on some of the 

 western Aleutian Islands, but although we cruised as far west as 

 Attu Island we did not even see any of the birds. We were unable 

 to visit Semichi Island where it is said to breed. Stejneger (1885) 

 found it breeding abundantly in the Commander Islands and has 

 given us the following accoinit of it: 



The fulmar is the first one of the nonresident water birds to arrive at the 

 rookeries in early spring, usually in March, the order of arrival being Fulmarus, 

 Vria arra, Lunda cirrhata, Fratercula corniculata. One specimen of the 

 white form was obtained on Bering Island, February 7, which would indicate 

 that the advance guard had already reached the islands by that time, or 

 else, what I am rather inclined to believe, that many of the birds pass the 

 winter on the open ocean not so very far from the shores tliey inhabit in 

 summer. 



