380 Bulletin American Museum of Natural History. [Vol. XXXIV, 
however, represent the extremes in color of this sex, the latter being sharply 
streaked with black while the underparts of fernandensis are heavily washed 
with rufous, with darker, obscure shaft streaks. It thus bears some re- 
semblance to three females of ochracea in juvenal plumage. 
I have been permitted to name this interesting island race through the 
courtesy of Dr. L. C. Sanford and Mr. Frederick F. Brewster, whose col- 
lection contains the only specimens of it known to me. 
Specimens examined.— Juan Fernandez, 12 oo’, 11 9 Q. 
Cerchneis sparverius australis (Ridgqw.). 
Falco gracilis (nec Lesson) Swartns., Anim. in Menag., 1838, p. 281 (Bahia, 
Brazil). 
Falco sparverius var. australis Ripaw., Hist. N. A. Birds, III, 1874, p. 166. 
Char. subsp.— Adult male with the underparts largely white and more or less 
thickly marked with round black spots; the back and scapulars usually with broad, 
numerous black bars; average female paler below than in more northern forms. 
Closely resembling C. s. cinnamomina but male usually with no rufous in the tip of 
the tail. 
Range.— From the eastern slopes of the Andes in Peru and Bolivia (and Argen- 
tina) eastward to the Atlantic (north to the Amazon?). 
Remarks.— Assuming that Ridgway’s name of australis was proposed as a 
substitute for gracilis of Swainson, preoccupied, Bahia is the type-locality 
of thisrace. Unfortunately two males from Pernambuco and one from San 
Paulo are the only specimens I have seen from the Brazilian coast. In color 
and markings of the underparts and back, practically the entire range of vari- 
ation shown by the 16 males from the range I ascribe to this form is covered 
by these birds. In one Pernambuco specimen the outer pair of rectrices 
is lost; in the other these feathers are broadly barred with black from base 
toend. In the Rio Grande do Sul specimen the terminal half of this feather 
is barred. Two of Chapada specimens have the outer rectrix with two or 
more bars, while two have only the subterminal bar. Two or more bars 
appear therefore to be the more frequent marking among Brazilian speci- 
mens. 
In size, as the appended table shows, there is much variation but appar- 
ently no constant or racial difference in dimensions. Birds from Chapada 
have the tail comparatively short but they are matched by Pernambuco 
‘specimens and exceeded in size by a male from San Paulo. 
Of Cerchneis s. peruana Cory (Field Mus. Pub. 182, p. 296, Feb. 1915) 
I have seen only two males and a female from Macate, Peru, loaned me by 
Mr. Cory. These specimens can be matched in color and in size by birds 
in our series from Brazil, and I am unable therefore to separate them from 
