FALCONTD.F- — TTIK FALCONS. 221 



of the chdid of tuo fulmcn, its upper outline greatly luscendinjj liasally ; commissure with 

 a proiniiuiit fcstodii. Superciliary shield very proiuiucnt. Xostril broadly ovate, 

 obliquely imi'i/.diual. Tarsus loiincr than the middle toe, the frontal and posterior series 

 of regular Iraiisverse seulellic very distinct, and continuous, sonietinies fused into a con- 

 tinuous plate (as in tiie TnnUmv!). Outer too longer than the inner; claws strongly 

 curved, very acute. Wing short, nutch rounded, very concave beneath ; third to lifth 

 quills longest; liisl, usually siiortest, never longer than the sixth; outer three to five with 

 inner webs cut (usually sinualed). Tail long, nearly e(pial to VMig, usually rounded, 

 sometimes even, more rarely graduated {.Utiir maerounis) or emarginated (some species 

 of subgemis Kisus). 



Subgenera. 



Less than one third of the upper portion of the tarsus feathered in front, 

 the teatliering widely separated behind ; frontal transverse scutellie of the 

 tarsus an.l toes uninterrupted in the neighborhood of the digito-tarsal joint, 

 but continuous from knees to claws. Tarsal scntella' sometimes fused into 

 a continuous plate A7,s'».s. 



More than one third (about one half) of the ujipcr portion of the tarsus 

 feathered in front, the feathering .scarcely separated behind ; irontal trans- 

 ver.se .scutelhe of the tarsus and toes interrupted in the region of the digito- 

 tarsal joint, where replaced liy irregular snndl scales. Tarsal scutclhc never 

 fused Xxtur. 



The species of this jfonus are e.\ceediii<fly nuinerous, about fifty-seven 

 being the number of nominal "species" recognized at the present date. 

 Among so many species, tliere is, of course, a great range of variation in tlie 

 details of form, so that many generic and subgencvic names have been pro- 

 posed and adopted to cover tlie several groups of species which agree in 

 certain peculiarities of external structure. That too many generti and sub- 

 genera have been recognized is my final conclusion, after critically e.xamin- 

 iug and comparing forty of the fifty-seven species of Gray's catalogtie (Hand 

 List of Birds, 1, 18(59, pp. 29-35). The variation of almost every character 

 ranges l)etween great extremes ; but when all the species aro compared, it is 

 found that, taking each character separately, tliey do not a. <.oj^.ond, and 

 cross and re-cross each other in the series in such a manner that it is almost 

 impossible to arrange tlie species into well-defined groui)S. PVom this genus 

 I exclude Loj)hospiza, Ktiup (type, L. trivirgatus) ; Asturinu, Vieill. (type, A. 

 nitida) ; Ruponiis, Kaup (type, 11. iiuirfnirostris) ; Butcola, Dubus ( ^^ Bnteo, 

 type, B. hraehynra, Vieill.) ; included Ijy Gray under Astur, as subgenera, 

 and Tachi/spiza, Kauj) (type, T. sohcnsis) ; and Scflospiza, Kau]» (type, *S'. 

 J'vancesii) ; whicli are given by Gray as subgenera of Microiiisvs, CJray 

 (type, Accipiter gabar), the species of the typical subgenus of which, as 

 an-anged in Grtiy's Hand List, I refer to Nisus. All these excluded names 

 I consider as representing distinct genera. 



The species of tliis genus are noted for their very predatory disposition, 

 exceeding the Falcons in their daring, and in the quickness of their assault 

 upon their prey, which consists chiefiy of stnall birds. 



