BIEDS OF NORTH AND MIDDLE AMERICA. 737 



the right one slightly larger than the left. Wing large, with 

 longest primaries decidedly longer than secondaries; eighth, seventh 

 and eighth, or eighth and ninth ^ primaries longest, the tenth (appar- 

 ent outermost) longer than fifth,^ usually equal to or longer than 

 sixth; <= two to four outer primaries with inner webs emarginated or 

 sinuated. Tail about half as long as wing, slightly rounded, the 

 longest under tail-coverts falUng far short of its tip. Tarsus stout, 

 shorter than middle toe with claw, densely covered, all round, with 

 rather long, soft feathers, the toes (except underneath and on 

 terminal phalanges) also densely feathered. Head with conspicuous 

 ear- tufts. 



Coloration (of American species). — Under parts barred (but not 

 striped) with black or dusky and whitish, usually mth a white jugular 

 area; upper parts mottled and vermiculated with grayish brown, 

 dusky, and buffy or tawny or whitish. 



Range. — Northern hemisphere in general, south to India, southern 

 Africa, and southern extremity of South America; wanting in West 

 Indies and Galapagos Islands. (About ten species,*^ with many sub- 

 species, but only one or two species occurring in America.) 



The treatment here given to Bubo virginianus mth reference to its 

 geographic variations is the same as that of Mr. Oberholser in his 

 revision of the group,* the present MTiter having examined the same 

 material. The present case is an extreme example of the difficulty 

 which sometimes attends any attempt to elucidate the variations of 

 an exceedingly variable species of extensive geographic range, and 

 probably very few authors would agree in their treatment of the case. 

 The subdivision of B. virginianus into a considerable number of geo- 

 graphic subspecies seems necessary if any subdivision is to be made 

 at all, for a middle ground between this extreme subdivision and the 

 ignoring of subspecies altogether seems impracticable; at least such 

 a course would leave a vastly greater number of specimens undeter- 



O' Third, third and foiirth, or fourth and fifth from outside, not counting the rudi- 

 mentary (concealed) eleventh (first) primary. 



& Sixth from outside. 



c Fifth from outside. 



d Owing to absence of specimens for examination I am not able to state precisely 

 the number of known species which should be referred to this genus. In his "Hand 

 List of the Genera and Species of Birds" (i, 1899, 282, 283) Dr. R. Bowdler Sharpe 

 names twenty-five species (and three additional fossil species), but of these at least 

 six are merely geographic forms or subspecies, while two others, B. blakistoni Seebohm 

 and B. doerriesi Seebohm, are certainly not members of the genus but belong to an 

 exceedingly distinct one which may or may not have been already characterized and 

 named. 



eA Revision of the American Great Horned Owls. By Harry C. Oberholser, 

 Assistant Ornithologist, U. S. Department of Agriculture. Proc. U. S. Nat. Mus., 

 xxvii, no. 1352, Jan. 22, 1904, pp. 177-192. 

 3622°— Bull. 50. pt 6—14 47 



