FALCONID.E: VULTURES, FALCONS, HAWKS, ETC. 649 



and absolute lengths. The feet are usually strong and efficient instruments of prehension and 

 weapons of offence or defence, with widely separable and strongly contractile toes, cleft to the 

 base or there only united by small movable webs, and generally scabrous underneath with 

 wart-like pads or tylari to prevent slipping, as shown in fig. 46. The claws are developed 

 into large sharp curved talons. The tarsal envelop (podotheca) varies ; sometimes the whole 

 tarsus is feathered, and it is usually so in part ; the horny covering takes the form of scutella, 

 or reticulations, or rugous granulations, and is occasionally fused. The capacious gullet dilates 

 into a crop ; gizzard moderately muscular ; coeca extremely small ; oil-gland tufted; syrinx of 

 ordinary broncho-tracheal form ; ambiens and femorocaudal muscles are present ; accessory femo- 

 rocaudal, semitendinosus, and its accessory are absent. There are good osteological characters : 

 Phalanges of hind toe are more than half as long as those of outer toe ; basal joint of middle or 

 outer toe is longer than next one. No basipterygoid processes. Sternum manubriated, and 

 when not entire behind is single-notched or fenestrate on each side (doubly so in most Ov.-ls). 

 Huxley has called attention to a character of the shoulder-girdle, afterward well elaborated by 

 others (fig. 444) : In certain genera, as Falco, Micrastur, Herpetotheres, and in PolyhorincB, 

 tlie scapular process of the coracoid (fig. 444, c) is prolonged beneath the scapula, d, to meet 

 the clavicle, h; which is not the case in other groups of genera of Falconidce, nor iu Pandi- 

 onidce. This distinction has been made the basis of a primary division of diurnal Aceipitres 

 into two subfamilies, Falconince and Buteonince, the former including Polyhorus and its allies, 

 the latter including Pandion; but some modificatitm of this scheme is advisable, I think. It 

 seems to me that the primary divisi(»n should be made as on p. 619, by excluding PandionidcB 

 as a family distinct from Falconidce proper, on ground of its many peculiarities. This being 

 done, the character of the shoulder-joint may properly be considered in dividing i^aZconjVZfC into 

 subfamilies. I am perfectly willing to approximate Polyhorus to Falco on this technical arround, 

 notwith.standing the great outward dissimilarity of these two forms ; but it is unlikely that or- 

 nithologists will allow the construction of the shoulder-joint to outweigh all other characters 

 combined. The feet are sympelmous, but in two different ways iu the two families, Falconidce 

 and Pandioyiidce. 



Diurnal Birds of Prey abound in all j^arts of the world, holding the relation to the rest of 

 their class that carnivorous beasts do to other mammals. With many exceptions, the sexes 

 are alike in color, but the 9 ^s almost invariably larger than the ^. Changes of plumage 

 with age are great, and render determination of species perplexing — the more so since purely 

 individual, and somewhat climatic, color-variations, and such special conditions as melanism, 

 are very frequent. Modes of nesting are various ; the eggs as a rule are blotched, and not so 

 nearly spherical as those of Owls. The food is exclusively of an animal nature, thougli end- 

 lessly varied; refuse of the stomach is ejected in a pellet by the mouth. The voice is loud and 

 harsh. As a rule, Birds of Prey are not strictly migratory, though many of them change their 

 abode with much regularity. Their mode of life renders them usually non-gregarious, except- 

 ing, however. Vultures and vulture-like Hawks, which congregate where carrion is plentiful, 

 ([iiite like American Cathartides. There are upward of 3.50 species or good geoorraphic.il races, 

 referable to about 75 genera, and divisible into two families — Falconidce and Pa)idionid(C. 



Family FALCONID^ : Vultures. Falcons. Hawks, Eagles, etc. 



Characters as above, exclusive of those markini: the Fi^<h-hawl;s, Pandionidcp, beyond. 

 No unexceptionable division of the family liavin<: been proposed, and the subfamilies being still 

 at i.ssue, it may be best not to materially modify the arramrt'inent presented in the earliest edi- 

 tion of this work, further thau to separate Pandiunidcc from Falcouidtr proper, as was done iu 

 tlje 2d edition. 



