518 BULLETIN 50, UNITED STATES NATIONAL MUSEUM. 



with a prominent angle near the base, anterior to which the edge is 

 clecidedl}" concave; maxillary tomia with a decided notch near the 

 base, immediately above the mandibuhir angle; gonys very long (about 

 equal to the exposed portion of the culmen). Nostrils exposed, small, 

 longitudinal, occupying less than the lower half of the nasal fossas. 

 Tarsus a little longer than the middle toe and claw, divided into about 

 six segments, but these entirely fused on the outer side, except the lower 

 one, which is distinct; lateral toes equal, the points of their claws fall- 

 ing short of the base of the middle claw; inner toe entire!}' separated 

 at the base from the middle toe, and outer with onl}- its first phalanx 

 united; hallux about equal in length to the lateral toes, but much 

 stronger, its claw decidedly larger than that of the middle toe. AVing- 

 decidedly longer than the tail, very concave beneath, the seventh, 

 sixth, and fifth primaries nearly equal and longest, the eighth very 

 little shorter, the ninth about equal to the third. Tail about equal to 

 the wing measured to the ends of the secondaries, nearly even or very 

 slightly rounded, consisting of twelve rectrices, which are acute, but 

 not stiffened at the points; the outer web of the intermediw hrostdev at 

 the base than the inner web. 



As characters additional to those given in the above diagnosis, it 

 may be stated that the bill is slightl}' notched near the tip, that the 

 rictal bristles are so mmute as to be hardly perceptible, and that the 

 posterior face of the tarsus, on both sides, is entirely undivided. 



^«W(7(^. ^Highlands of Costa Rica. (Monot^'pic.) 



The adult male of Acanthidops hairdi resembles very closely in col- 

 oration that of TIapIospiza unicoloi\ of Brazil, but is darker and less 

 of a bluish cast, both above and })elow, and has the mandible chiefly 

 light colored. Not having a specimen of the Mexican H. unifwniis^ 

 I am unable to point out wdt^^ exactness the differences from that spe- 

 cies; but judging from the description and remarks given by Messrs. 

 Salvin and Godman ^ it seems to come luuch closer to that species, if 

 it be not actually the same. It would appear, however, to have decid- 

 edly a shorter wing and longer tail, and longer tarsus than that bird. 

 Whether or not it is the same species or congeneric with H. xmiforinis^ 

 there can be no doubt that it is generically distinct from H. unicoloi\ 

 which has the bill much shorter and more typically Fringillinc, the 

 tarsi shorter, and the nostrils more exposed. There can, I thiuK, be 

 no question that Acanthidops comes very near to Hcq^lo.Kpisa^ and I am 

 inclined to believe that the so-called Jl. uniformis belongs to Acan- 

 tJddops and not to Haplospiza^ since Messrs. Salvin and Godman* say 

 that its chief differences from H. unicolor consist in "a larger bill and 

 longer wings and stronger feet," as well as " rather darker" coloration. 

 If this view of its relationship proves correct, we would then have a 

 ]\Iiddle American genus, Acanthidops^ and a related Brazilian genus, 



^ Biologia Centrali-Americana, Aves, pp. 366, 367. 



