FALCONID^E—CIRCIN.E: HARRIERS. 651 



the end by the 1st and 3d ; the nicking of the quills confined to a few, if occurring on more than 

 one, and situated near the tip. Such a wing is as potent in its feathers as in the construction 

 of its shoulder-joint, and indicates the acme of raptorial power in its possessor, a Falcon being 

 able to dash down upon its quarry with almost incredible velocity and violence. The mark- 

 ings of a Falcon's wing are no less characteristic, consisting of clean-cut, distinct spots of light 

 cok)r on both webs of primaries and secondaries, throughout their whole extent, or almost so. 

 (b) Any true "Hawk," as an Astur or Accijnter, has a rounded concavo-convex wing, confer- 

 ring a rapid, almost whirring, tlight, like that of a Partridge at full speed ; and such a bird 

 captures its prey by chasing after it with a wonderful impetuosity, but not at a single plunge 

 like a Falcon. Such a wing has more primaries cut, farther from their ends, and the markings 

 are pretty regular and distinct hais. (c) Any " Buzzard," as a Buteo, a heavy and compara- 

 tively slow or even lumbering bird in flight, taking its prey by surprise and merely dropping 

 on it without special address, has many or most of the primaries cut, far from their ends, and 

 tlie tendency of the markings is to fuse and blend in large irregular masses of color, the sharp 

 markings of Falco or Accipiter he'mg thua done away with. Of course there are exceptions, 

 as well as every possible gradation, in the case ; but if one will compare the wing of Circus or 

 Archibuteo with that oi Aecipiter or Falco, he cannot fail to perceive the point I raise. The 

 tail is in somewhat like case. In the most noble Birds of Prey it is very stiff and strong, with 

 almost lance(jlate feathers, sharply spotted as a rule; in a Hawk, longer and weaker, still regu- 

 larly barred ; in a Buzzard general! 1/ (there are marked exceptions) of medium length and strength, 

 with the markings tending to merge in large areas of color, just as those of the wings do. Fur- 

 thermore, in large and difficult genera, as Buteo for example, the best specific characters may 

 be aSorded by the markings of the tail. These are usually quite different in young and old 

 birds; but are amc^ng a Hawk's most specific credentials, after the mature plumage is assumed, 

 even when the rest of the plumage varies greatly, or is subject to melanism, erythrism, etc In 

 fine, many Hawks are best known by their tails. Melanism is frequent in Falconidce; erythrism 

 is not (just the reverse of the case of Strigidce). The further generalization may be made, that 

 the coloration of under parts of Falconidce is more distinctive of species than that of upper parts ; 

 and that when these parts are barred crosswise in the adult they are streaked lengthwise in the 

 young. Sexual differences are rather in size than in color, such a case as that of Circus being 

 exceptional. 



Analysis of Subfamilies. 



Scapular process of coracoid reaching clavicle. 



Upper mandible toothed, lower mandible notched Falcon'is.e 



Mandibles without tooth or notch PoLTBORiNi: 



Scapular process of coracoid not reacliing clavicle. 



Face with a ruff somewhat as in Owls C1RCIN.E 



Face without ruff. 



Tarsus approximately equal to tibia in length ; rounded wings little longer than tail .... Accipitrin.k 

 Tarsus decidedly shorter than tibia. 



Tail forked, or much sliorter than the long pointed wings Milvin'.« 



Tail not forked, moderately shorter tlian tlie obtuse wings Bitteoxik.e 



Subfamily CIRCIN/E : Harriers. 



Face surrounded with an incomplete ruff (as in most Owls;) orifice of ear about as large 

 as eye, and in some cases at least with a decided of>nch (fig. 446). Bill rather weak, not 

 toothed or notched. Legs lengthened ; tarsus approximately equalling tibia in length (as in 

 Accijiitrinfr). Wings and tail lenytliened. Form light and litlie; plumage loose; general 

 organization of the buteonine rather than of the falconine division of the family. Thus, 

 scapular process of coracoid not produced to clavicle ; no nirdiaii ridir'' <'n jialatc anteriorly; 

 septum nasi less complete than in Fnlco. and nostrils not circular with a central tubercle. 



