244 Bulletin American Museum of Natural History. [Vol. XXXVI, 



The distribution of this Hawk in Colombia is most puzzling. It occurs 

 in both the Tropical and Subtropical Zones and occupies all the region east 

 of the Central Andes and north of the Cauca Valley, and it is also found on 

 the Pacific coast and southward along the Pacific coast of Ecuador. In 

 general it may be said that in Colombia the range of magnirostris surrounds 

 that of those representatives of ruficavda which inhabit the Cauca Valley 

 and slopes arising from it. 



All the Colombian specimens listed below (except one from Barbacoas) 

 including two from the Atrato Valley, are essentially typical of magniros- 

 tris in color, though those from west of the Eastern Andes are somewhat 

 smaller than those from the region about Villavicencio ; but a bird from 

 Barbacoas and eight from western Ecuador show some slight approach 

 toward ruficauda. The white bars below perhaps average wider and the 

 bars on the thighs may average somewhat more rufous; the tail shows the 

 slightest trace of tawny confined as a rule to the outer pair of feathers ; the 

 under wing-coverts are white barred with brownish or black; the upper tail- 

 coverts are white, in two specimens with a faint buff tint and barred with 

 blackish. Further material and finer discrimination may show that these 

 birds are distinguishable from true magnirostris with which they appear to 

 have no geographic connection; but there can be no doubt that they are 

 referable to that form rather than to ruficavda. 



In addition to the specimens from western Colombia listed below, we 

 have Ecuador specimens from Esmeraldas (4), Santa Rosa (1), Naranjo, 

 Prov. Guayas (2), and Daule (1). The specimens from the Atrato and 

 Dabeiba are essentially typical magnirostris as that species is represented 

 by an excellent series from eastern Colombia and Venezuela. The presence 

 of this form on the Pacific coast is doubtless due to its extension westward 

 from northern Colombia where, as we have seen, magnirostris occurs. 



Inosculation of these forms is indicated by the capture of a specimen at 

 Noanamd which is obviously nearer to ruficavda than to magnirostris; but 

 this makes it difficult to explain the reappearance of magnirostris in Ecuador. 



The further interesting distribution of this species in Colombia is treated 

 under the next form. 



Atrato River, 1; Dabeiba, 1; Barbacoas, 1; Puerto Valdivia, 1; Sta. 

 Elena,!; Barro Blanco, 2; LaPalma, 1; Chicoral, 2; 30 m. west of Honda, 

 3; Villavicencio, 3; Florida, 1. 



(625a) Rupornis magnirostris ruficauda {Scl). 



Asturina ruficauda Scl., P. Z. S., 1869, p. 133 (no definite type-locality desig- 

 nated, of the localities named, I select David, western Panama, as type-locality). 

 ? Asturina magnirostris Cass., Proc. Acad. N. S., Phila., 1860, p. 132 (Turbo). 



