STELGIDOPTERYX. 313 



verging. Frontal feathers soft, and, like chin, without bristles. Tarsi equal 

 to middle toe without claw ; the uppes end covered with feathers all round, 

 none at lower end. Basal joint of middle toe adherent externally nearly to 

 end ; internally, scarcely half. Lateral toes about equal, their claws not 

 reaching beyond base of middle claw. Tail slightly emarginate ; the feathers 

 broad, and obliquely rounded at end. Edge of wing rough to the touch ; the 

 shafts of the fibrillse of outer web of outer primary prolonged and bent at 

 right angles into a short stiff hook. 

 Color dull brown above. 



The great peculiarity in this genus consists in the remarkable 

 roughness of the edge of the wing, said to occur also in Paalido- 

 procne, Cab. The object is uncertain, but is probably to enable the 

 bird to secure a foothold on vertical or inclined rocks, among or 

 on which it makes its nest. A favorite breeding place of S. serri- 

 pennis is in the piers and abutments of bridges, and these hooks 

 might render essential aid in entering into their holes. 



The birds of this genus have usually been referred to Gotyle, 

 which, however, they resemble only in color. The nostrils are ex- 

 posed, instead of being overhung ; the tarsus is bare below, not 

 feathered, and the lateral claws are considerably curved, and not 

 reaching beyond the base of the lateral, as in Gotyle. The structure 

 of the wing is very different. 



According to Cabanis, Psandoprocne (P. cypselina, Cab. of 

 Africa) has the same structure of wing, but it seems to differ in 

 having the tail deeply forked, as in Atticora ; the toes and nails even 

 shorter than in Atticora, not longer ; and in having the outer toe 

 shorter than the inner, instead of equal to it. 



The genus has a wide range, extending from British America to 

 Brazil, and probably Ecuador. 



as his description of the tail and its under coverts, at least, does not apply at 

 all. Brisson's article is evidently copied from Feuille (1725), a very vague 

 author, as likely to call a Cuckoo or Flycatcher "Hirundo" as anything else. 

 This species is much in form like A. cyanoleuca, but considerably larger. 

 I have little doubt that the A. cyanophoea, of Cabanis, is the young bird of this 

 species, as a specimen in the museum of the Philadelphia Academy, from 

 Bogota, agrees exactly with Cabanis's description, and is hardly to be distin- 

 guished, except in its evident immaturity, from the Academy-type of Cassin's 

 Pelrochelidon murina, from Ecuador. It is probably closely related to the S. 

 andecola, of B'Oih. & Lafr. Syn. Av. 1837, 69 (La Paz), but differs somewhat. 



