FALCONID^ — THE FALCONS. 165 



Atlantic coasts of South America (probably the whole Atlantic coast of 

 tropical and subtropical America), the true sparverius is changed into what 

 Swainson has called " Falco isahellmus," which differs from the former only 

 in having the cinereous of the crown and wings considerably darker (as well 

 as less bluish), approaching plumbeous ; the rufous of the crown is totally 

 absent, or only present in faint touches ; the lower parts are of a deeper 

 ochraceous, and the black spots on the breast and sides sparse. 



Allied to the last in tints of coloration, and apparently a direct offshoot 

 from it, is the dominicensis of Gmelin (based upon description by Brisson), 

 whi(;h inhabits the Lesser Antilles, from Trinidad nortliward to Porto Eico. 

 Although I consider this (v;ir. do7ninicensis) as a modified form of the var. 

 isahellinus, yet it is tlie one of all the varieties referrible to sparverius 

 which deviates most widely from the typical or original style. The charac- 

 ters of this are, tints those of var. isahellinus, but, in addition, the tail has 

 numerous more or less complete black bands, while those of the back and 

 scapulars are very broad and numerous ; also, tlie crown has a decided rufous 

 patch ; the bill, too, is larger than in any other American member of the 

 genus. A style of considerable uniformity spreads over the whole of South 

 America, including both coasts, from Bogota to the Parana, excepting the 

 northeastern coast region, before mentioned as inhabited l)y the var. isahel- 

 linus. It differs from all the other styles, except the einnainow,inus of Chile, 

 in having the lower parts continuously didl white, any ochraceous tinge 

 being scarcely perceptible ; there is seldom a trace of rufous on the crown, 

 which has the light bluish tint seen in ^•ar. sparverius, and the black zone 

 of the tail is scarcely more than half as wide as in the northern races. In 

 size, also, it somewhat exceeds the others. Swainson named this "Falco 

 gracilis" ; but the F. {Tinnunculus) (jracilis of Lesson being a different spe- 

 cies, and the name as applied to it of prior date, I have bestowed upon the 

 present bird tire name var. australis. 



In Chile and Brazil (Western ?) we find a form resembling the last in some 

 respects, but differing in points of almost specific value. It differs from all 

 the other x4.merican members of the genus in having the tail continuously 

 rufous to the extreme tip, the l)lack zone being considerably narrower than 

 the terminal rufous, the lateral tail-feather inmiaculate rufous, etc. The 

 grayish of the head is much darker and more slaty than in tlie var. australis. 

 This is, without doubt, the Falco ciyinaruoriiinus of Swainson, the specimens 

 in the collection corresponding exactly with tlie description by that author. 



The rufous patch on the crown must not be too much considered, as it is 

 of all characters perhaps the most treacherous, though its presence or 

 absence is in a measure characteristic of the several varieties. Neither does 

 the exact number of spots on the lateral tail-feather prove sufficiently con- 

 stant to serve as a character in which the least reliance can be placed, 

 thoufdi Swainson attaches considerable value to it. I have found that, 

 besides varying almost with the individual, in some specimens the feathers 

 of opposite sides did not correspond. 



