BIRDS OF NORTH AND MIDDLE AMERICA 149 



d. Black and gray bands on the tail about equal in width. 



e. Underparts barred and marked with rusty brown; size smaller wing 



under 150 mm. (Puerto Rico). A. striatus venator, adult (p. 201) 



ee. Underparts white, with thighs buff (adult), or breast and sides 



faintly striped with pale brown; size larger (wing over 160 mm.) 



(Guatemala and Nicaragua). 



A. chionogaster chionogaster (p. 202) 

 dd. Black bands on tail decidedly wider than the gray interspaces. 

 e. Tibiae uniform. 



/. Underparts white, with thighs uniform rufous. 



A. salvini 89 (extralimital) .*" 

 ff. Underparts mostly rufous (mountains of Venezuela, Colombia, 

 Ecuador, Peru, and northwestern Bolivia). 



A. chionogaster ventralis (extralimital) "' 

 ee. Tibiae banded. 



/. Smaller (wing 127-148 mm.), and darker; upperparts less suf- 

 fused with dark gray (Costa Rica to northern and western 



Colombia) A. superciliosus fontanieri, adult (p. 182) 



ff. Larger (wing 135-164 mm.) and paler; upperparts suflFused with 

 dark gray (Venezuela and the Guianas south to Brazil). 



A. superciliosus superciliosus, adult (extralimital) 



** This is usually regarded as a form distinct from A. chionogaster ventralis, but 

 it may be that the two are only phases of the same form. Their ranges appear 

 to coincide (that is, both are recorded from Ecuador, cf. Chapman, Bull. Amer. 

 Mus Nat. Hist., Iv, 1926, 224, and there are specimens of both in the United 

 States National Museum taken at the same locality in Venezuela). A. salvini 

 agrees with ventralis in having the dark bands on the tail distinctly wider than 

 the gray, and thus differs from chionogaster, but in its white underparts it agrees 

 with chionogaster. A specimen from Valle, Venezuela, taken along with another 

 having rufous underparts, is ■white below, but has traces of rufous, especially on 

 the flanks. It seems, therefore, that chionogaster is a good race characterized by 

 its paler tibiae and white underparts, and that it may be that the only other 

 recognizable race is ventralis (of which salvini would be a synonym), characterized 

 by its more rufous tibiae, and that ventralis has two phases. A. erythronemius 

 seems to belong to the nisus group, rather than here. However, as these birds 

 are extralimital to the present work, they are not of immediate concern and 

 have been kept separate in this key. 



8« Nisus salvini Ridgway, U. S. Geol. and Geogr. Surv. Terr., Bull. 2, 1876, 

 121 (Merida, Venezuela). — Accipiter salvini Sharpe, Hand-list, i, 1899, 253; Bra- 

 bourne and Chubb, Birds South Amer., i, 1912, 65; Swann, Synopt. List Accip., 

 1919, 36. — Accipiter erythronemius salvini Swann, Synop. Accip., pt. 1, 1921, 

 58; Monogr. Birds Prey, i, pt. 5, 1926, 316; Peters, Check-list Birds of World, 

 i, 1931, 221. 



" Accipiter ventralis Sclater, Proc. Zool. Soc. London, 1866, 303 (interior of 

 Colombia); Sharpe, Cat. Birds Brit. Mus., i, 1874, 149; Hand-list, i, 1899, 253; 

 Brabourne and Chubb, Birds South Amer., i, 1912, 65. — Accipiter ventralis 

 ventralis Swann, Synop. Accip., pt. 1, 1921, 58; Monogr. Birds of Prey, i, pt. v, 

 1926, 319. — Accipiter nigroplumbeus Lawrence, Ann. Lye. Nat. Hist. New York, 

 ix, 1869, 270. — Accipiter ventralis nigroplumbeus Swann, Synop. Accip., 1921, 

 59; Monogr. Birds Prey, i, pt. 5, 1926, 320. — Accipiter erythronemius ventralis 

 Peters, Check-list Birds of World, i, 1931, 221. 



