172 BULLETIN 121, UNITED STATES NATIONAL MUSEUM. 



Gary's chicken " has, however, hirgely protected them. Capt. J. W. 

 Collins (1884) thus describes the method formerly in use by fish- 

 ermen desiring these birds for bait : 



The most common and effective way of killing them was with a whip, which 

 was made hy tying several parts of cod line — each part G or 8 feet long — to a 

 staff 5 or 6 feet in length. The petrels were tolled up by throwing out a large 

 piece of codfish liver, and when they had gathered in a dense mass, huddling 

 over the object which attracted them, swish went the thongs of the whip, cut- 

 ting their way through the crowded flock and perhaps killing or maiming a 

 score or more at a single sweep. By the time these were picked up another 

 flock was gathered, and the cruel work went on until, maybe, 400 or 500 birds 

 were killed, though perhaps it was seldom that so great a number was obtained 

 at once. 



Wilson petrels have a certain economic value as foretellers of 

 bad weather, for they are more active just before storms, at least 

 such is their reputation among sailors — and it is possible that on 

 this account they derive their common name, " Mother Gary's chick- 

 ens," from Mater Cara^ the blessed Virgin, A group of petrels sit- 

 ting quietl}^ on the water on the other hand is said to be a sign of 

 calm weather, although the contrary is sometimes affirmed. 



Pall. — The fall migration of the Wilson petrel is, of course, the mi- 

 gration from their breeding place at the end of the Antarctic sum- 

 mer. Their winter habits and haunts in our northern summer have 

 already been described. 



The Wilson petrel resembles the Leach petrel in life, but can be 

 distinguished from it in the following manner : The Wilson petrel is 

 slightly smaller than the Leach and blacker — less rusty — although 

 these differences are very slight and both birds appear to be black 

 with white rumps. The tail of the Wilson petrel is rounded, that 

 of the Leach slightly forked, but this again is a point that is diffi- 

 cult to make out. Murphy (1915) emphasizes the " notable different 

 style of flight of Leach's as distinguished from Wilson's petrel." He 

 says : " Oceanodroma flies with rapid, ' leaping ' strokes, quite unlike 

 the alternations of gliding and synchronous flutters which character- 

 ize the flight of Oceanites?'' The most diagnostic point, however, and 

 the one that Wilson himself was the first to record, although a num- 

 ber of others have noted it independently, is the fact that the feet 

 with their yeUov) webs extend beyond the tail in the Wilson, but the 

 short, black-webbed feet of the Leach petrel are concealed below the 

 tail. 



DISTRIBUTION. 



Breeding range. — Found breeding on Mauritius and Kerguelen 

 Islands in the Indian Ocean; at Gape Horn, on Adelie Land, and 

 on Victoria Land (Gape Adare) ; also on the South Shetland and 



