76 PTERYLOGRAPHY. 



2. Epimachus. This genus, which is usually placed in the neighbourhood of Upupa, is 

 undoubtedly a member of the Passerine group, and stands in the same relation to Paradisea as 

 Corvus yraculus to the true Ravens and Crows. The two species examined by me, E. supcrbus 

 and E. regius, differ somewhat in their pterylosis : the former approaches Paradisea more closely in 

 so far that the dilatation of the pectoral tract is not separated at the end ; but the dorsal tract has 

 a very narrow, scarcely perceptible saddle. In E. regius the dilatation of the pectoral tract is very 

 distinctly separated at the end, and the saddle is perfectly cordate and bilobate behind. Between 

 the two lobes, in the emargination, the rump-band originates with two rows of feathers, which 

 separate a little, and increase on the caudal pit to three, and further on to four rows. Both 

 species have twenty remiges, of which ten are on the hand ; the first graduated ; the third to the 

 fifth the longest in E. regius, remarkably broad and emarginate at the apex in the male. 



3. Gracula religiosa (Eulabes Cuv.). Saddle as in Paradisea; dilatation of the pectoral 

 tract separated at the end. Nineteen remiges, the fourth the longest. 



4. Kitta tlialassina (Plate III, fig. 6). Saddle rhombic, obtuse-angled, emitting, by the side 

 of the rump-band, two rows of single contour-feathers. Remiges incomplete in the specimen 

 examined. 



B. With a large ephippial space. 



5. Ptilorhynchus holosericeus TEMM., PL Col., 395. All the tracts narrow; the dilatation 

 of the pectoral tract separated at the end. Saddle rhombic, elongated, with an equally elongated, 

 lanceolate space. Twenty-three, or even twenty-four, remiges ; the fifth and sixth the longest. 



3. 



a. With -a large central space in the saddle of the spinal tract. 



(All the members of this group are destitute of the posterior lamina? of the tarsi, instead of 

 which elliptical warts make their appearance ; these, however, are wanting in most of the large 

 species of Pipra, leaving the tarsi quite nak e d behind. BURM.) 



1. Coracina. Twenty or twenty-one remiges ; the first six graduated. Saddle acute-angled 

 on the sides. C. calva, C. scutata, and C. rubricottis. 



2. Cep/iahpterus. Number of remiges not exactly ascertained ; tracts as shown in the 

 figures (Plate III, figs. 9 and 10). The chief distinction consists in the elongated, laterally obtuse 

 form of the saddle. The separation of the inferior tract of the neck from the pectoral tract 

 behind the gular sac is also remarkable. The male alone has the beautiful gular tuft of feathers ; 

 and the female has a short, erect frontal hood, instead of the larger structure of the other sex. 



3. Chasmarliyncliiis nudicollis. Twenty or twenty-one remiges ; the third and fourth longest. 

 Saddle acute-angled at the sides and broad, as in Coracina. 



4. Eurylaimus. Eighteen or nineteen remiges; the third, or, with the latter number, the fourth, 

 the longest. This occurs only in E. corydon, which also has a more elongated and laterally obtuse 

 saddle. I find a shorter saddle, more acutely angled at the sides, and eighteen remiges, of which 



