56 PTERYLOGRAPHY. 



two rows unite only at the caudal pit, and then pass in the form of a broad band to the oil-gland. 

 We found just this tract also in Pandion, and its form is peculiar to these two groups. 



Humeral tracts not very broad, with few, but very large and strongly tubular feathers. 



Lumbar tracts always small, each consisting at the utmost of twenty feathers, sometimes 

 almost abortive, intermixed with powder-down-feathers, and very closely approximated to the 

 spinal tract. 



Crural tracts formed of contour-feathers only on the outside, forming breeches ; internally 

 downy. 



Inferior tract simple as far as the middle of the neck, narrow, separated from the spinal 

 tract by the lateral neck-spaces. Its two limbs likewise narrow, strongly divergent, and in con- 

 sequence united with the axillary tract at the shoulder, but destitute of the interior branch which 

 occurs in most Falcons, and descends from the furcula into the bend of the neck (fig. 3). On 

 the other hand, the exterior branch, equal in breadth to the main stem, is present on the pectoral 

 muscles. The interspace between them is considerable, and greater than usual ; the exterior 

 hook at its apex, which occurs in many Falcons, and leads to the hypopterum, is also wanting. 

 The main stem closely approaches the crest of the sternum anteriorly, and diverges behind, so 

 that the inferior space is much broader, on the ventral part ; the two main stems curve towards the 

 anus, and terminate close to it. 



Remiges twenty-three, sometimes twenty-four or twenty-five, rarely only twenty-one; 

 primaries alway ten. The first more or less abbreviated, the second always the longest ; both, or 

 only the former, with an angular emargination situated very close to the apex. The third some- 

 times longer, sometimes shorter, than the first. 



Rectrices very rigid, usually somewhat longer than the apices of the folded wings ; the 

 outer ones with their tips bent inwards, and at this point completely concealed beneath the inter- 

 mediate ones ; hence the tail appears narrow posteriorly. 



The species examined differ only in the number and form of the remiges, and I therefore 

 describe these ; the latter might be appropriately employed as a specific character. 



R. islandicus. Twenty-five remiges. The first two with an angular emargination ; the 

 third longer than the first, but the fourth shorter. 



R. peregrinus. Twenty-three remiges. The first two with an angular emargination, which, 

 however, is imperfect on the second ; the third shorter than the first. 



R. (Bsalon. Twenty-four remiges. The first two with a complete angular emargination, 

 which is also perceptible on the third ; the first as long as the fourth, or a little shorter. 



R. subbuteo. Twenty-three remiges. The second the longest ; the first but little shorter, 

 longer than the third. 



R. tinnunculus. Twenty -three remiges. First and second with a slight angular emargination ; 

 first nearly as long as the fourth. 



R. tinnunculoides seu cenchris. Twenty-three remiges. The second the longest; first as long 

 as the third, or a little shorter. 



R. rufipes. Number of remiges uncertain ; only the first with an angular emargination, and 

 somewhat shorter than the third. 



R. concolor TEMM. PL Col, 330. Twenty-three remiges. Primaries very narrow and 

 pointed ; the first only with a perceptible angular emargination, and longer than the third ; the 

 fourth reaches to the emargination in the first ; the twentieth is at the elbow, and is the longest 



