HA WKS. 



305 



count. As an illustration of the lack of uniformity among systematists with regard 

 to these genera, it may be mentioned that of two prominent authorities who published 

 their views at about the same time (1874), one refers but six species to Astur, while 

 the other includes thirty-one. The latter author, however, allows but twenty-three 

 species to the genus Accipiter, while the former admits forty-five. For our present 

 purposes it makes little difference which we follow in this respect, though there seems 

 to be little doubt now that only a few species ought to be included among the gos- 

 hawks. The word goshawk is evidently only a corruption of goose-hawk, and though 



PIG. 143. Astur palumbarius, goshawk. 



now only applied to birds of the genus Astur, it seems probable, as Professor Newton 

 remarks, that it was originally given to one of the large true falcons, which might 

 reasonably be supposed to prey on geese, as such game is evidently beyond the capacity 

 of Astur. 



The goshawk of the northern United States, Astur atricapillus, is by many be- 

 lieved to be merely a geographical race of the European goshawk, A. palumbarius. 

 If so, it is certainly a larger and handsomer form, and in habits the two are very simi- 

 lar. The adult American bird is one of the handsomest of our birds of prey, the 

 whole top of the head being pure, deep black, the rest of the upper parts pure bluish 

 VOL. iv. 20 



