38 BULLETIN 121, UNITED STATES NATIONAL MUSEUM. 



Josef Land, April 24. Leaves Georges Bank, Massachusetts, about 

 middle of March. 



Fall migration. — Early dates of arrival : Labrador, Cape Harrison, 

 July 19, and Ragged Islands, August 9. Late dates of departure: 

 Wellington Channel, September 2; Hudson Straits, September 15; 

 Cumberland Sound, middle of October; Franz Josef Land, October 

 28; Spitsbergen, October 6. Arrives on Georges Bank, Massachu- 

 setts, in October or earlier. 



Casual records. — Massachusetts (Chatham, September 23, 1912). 

 Connecticut (Stony Creek, October 10, 1907). New Jersey (Ridgc- 

 wood, December, 1891). Accidental in Madeira. 



Egg dates.. — St. Kilda Island: Twenty-eight records. May 6 to 

 June 15 ; fourteen records. May 22 to June 5. Iceland : Nine records, 

 May 14 to July 1 ; five records, May 30 to June 15. Greenland : One 

 record, July 2. 



FULMARUS GLACIALIS GLUPISCHA Stejnegcr. 

 PACIFIC FULMAR. 



HABITS. 



The relationships of the three fulmars found in the northern parts 

 of the Atlantic and Pacific Oceans are none too well understood, 

 and I very much doubt if their relationships are correctly designated 

 in our present classification of the three forms. 



The Pacific fulmar, with its light and its dark phases, is now 

 regarded as a subspecies of the Atlantic fulmar, with its two similar 

 phases, and the Rodgers fulmar, Avith no dark phase, is recognized as 

 a distinct species. Por reasons which I have briefly stated, under 

 that species. I doubt if the Rodgers fulmar will eventually prove to 

 be even subspecifically distinct from the Pacific fulmar, as the 

 characters on which the former species is based can be accounted for 

 by age, seasonal or individual variations. This fact is beautifully 

 illustrated in the magnificent series of fulmars which Mr. Loomis 

 has accumulated in the collection of the California Academy of 

 Sciences. 



The Pacific fulmar may be correctly regarded as a subspecies of the 

 Atlantic bird, although one of the principal characters which sepa- 

 rates the two birds, the color of the bill, is very variable. Although 

 I have never seen the theory advanced, there are several good rea- 

 sons for regarding the dusky birds, of both oceans, as a distinct 

 species, rather than regarding them as dark phases. The color- 

 phase theory has always been a convenient method for disposing 

 of a problem which we could not otherwise solve, but I believe that 

 it should be used only when definitely proven. In this case we have 



