ROUGH-LEGGED HAWK. 89 



seen it attack water fowl, or in fact any other bird. Among those 

 questioned on the subject are a number of eminent ornithologists, some 

 of whom have published the statement that the Bough-leg feeds on 

 wounded ducks. The error arose from their taking information second- 

 hand from gunners who probably mistook either the Duck Hawk or 

 Prairie Falcon for the bird under consideration When hard pressed by 

 hunger it may feed on dead water fowl as well as the carcasses of other 

 animals. 



Those Hawks that remain farthest north in winter are more often 

 forced by circumstances to feed on refuse. Maynard says that in Mas- 

 sachusetts they feed upon fish and the dead animals cast up by the sea, 

 and Mr. Vernon Bailey, writing from Elk Eiver, Minnesota, says: a A 

 few years ago, probably in 1883, 1 was trapping for muskrats and minks 

 late in the fall. As the place was over a mile from home, and I was 

 catching a good many rats, I skinned them and left the carcasses where 

 caught. Soon I noticed that these were often eaten or gone. Rough- 

 legged Hawks were unusually numerous that fall, and stayed nearly all 

 winter. One morning I came suddenly to the top of a hill and saw a 

 hawk fly away from a half-eaten rat on the other side. It was promptly 

 shot, and proved to be a fine dark Rough-leg. This was the only time 

 that I caught the hawk in the act, but from the manner in Avhich the 

 rats were eaten, the number of hawks present, and after snow came 

 the tracks seen around the remains of the rats, I had no doubt that the 

 hawks had eaten them." 



The few specimens which have been secured in summer south of the 

 northern boundary of the United States are those which have failed 

 to migrate, presumably on account of disease or from inability or indis- 

 position to make long flights after the receipt and healing of gunshot 

 wounds. In all these cases the genitalia are undeveloped. The nest 

 and eggs mentioned in the History of North American Birds as coming 

 from Wiscasset, Me., near the mouth of the Kennebec River, if not an 

 error in identification, is probably the only record of the breeding of 

 this species in eastern United States. Dr. Warren informs the writer 

 that the instance mentioned in his " Birds of Pennsylvania" of the 

 Rough-leg breeding in Pennsylvania was a case of misidentification. 

 Dr. Gr. S. Agersborg gives it as breeding once in southeastern Dakota 

 (Auk, vol. II, p. 285), but does not inform us whether it was found nest- 

 ing or merely that the presence of the bird in summer was taken as 

 evidence. 



It breeds sparingly in Labrador and the southern portions of Canada 

 west of Manitoba. In the vicinity of Winnipeg, Seton gives it as a 

 migrant only. (Auk, vol. in, 1886, p. 154.) Farther north, even far 

 within the Arctic circle, it is an abundant breeder. Although common 

 in the Hudson Bay and Anderson River regions in northern Alaska, 

 the Old World form seems to replace it not only on the seacoast, but 

 a ] ong the entire length of the Yukon. 



The nesting site is more or less varied, most of the nests being 



