84 PTERYLOGRAPHY. 



2. Dendrocolaptes. Pterylosis exactly as in Sitta ; remiges of the same number, but the 

 first larger; third, fourth, and fifth longest. Tarsus behind, with a series of scutes instead of the 

 simple plates. 



3. Certhia. Dilatation of the pectoral tract distinctly separated for some distance ; saddle 

 broader, more acute-angled, posteriorly more obtuse (Plate III, fig. 3). Nineteen remiges, the 

 first very small, the third to the sixth the longest. 



4. Philedon. Much as the species of this group differ in the form of the bill, they present 

 little variation in the pterylosis, which is exactly as in Sitta. The number of remiges is also the 

 same, but the first is usually half the length of the second, and the fourth alone, or the fourth and 

 fifth, usually exceed all the others in length, although but little. I find the saddle broad and 

 rather acute-angled in P. cuciilliyerm LIGHT, and P. JVbves Hottandiee LATH. ; it is very narrow and 

 lanceolate in P. fl/irygiHs (Merops phryyius LATH.). 



5. Campylops hamalus LIGHT. Pterylosis as in Certhia, but the divergent portion of the 

 pectoral tract is short. Wings imperfect in the specimen examined. 



6. Nectarinia. Pterylosis exactly as in the preceding genus. Nineteen remiges, of which 

 the first is small, and the third, fourth, and fifth are the longest, occur in N. scarlatina, N. 

 aurifrons, and their allies. N.flaveda and its allies, on the contrary, have only eighteen, the first 

 small one being deficient. In the latter the saddle also is narrower and lanceolate. 



B. With a space in the saddle of the spinal tract. 



7. Promerops ca/er Cuv. Saddle very broad and acute-angled, completely occupied by the 

 space, so that the rump-band commences with only two converging rows of single feathers ; pec- 

 toral band somewhat divergent at the apex. Nineteen remiges, the first primaries graduated, 

 with an emargination'of the inner half of the vane, which is remarkably large on the fifth. Tarsus 

 behind with small elliptical scales, the two posterior laminae not being in contact. Tail long, 

 forked. 



8. Arachnotheres chrysogenys TEMM., PI. Col., 388, 1. Pterylosis exactly as above, except 

 that the saddle is more elongated anteriorly, and therefore the space is also larger. Nineteen 

 remiges, of which the first is very small, and the fourth the longest. Margins of the upper 

 mandible finely denticulated at the apex. 



9. Dicteum pectorale. Pterylosis exactly as in the Swallows, namely, the saddle broad and 

 deeply bilobate, and the rump-band completely separated from it (Plate III, fig. 14). Eighteen 

 remiges, of which the first, second, and third are nearly of equal length and the longest. 



10. HIRUNDINE.E. 



The single genus Hirundo, which constitutes this group, differs more than any other in its 

 habitus from the general type of the Singing-birds, and in this respect approaches very closely to 

 some Cuculina>, namely, the Cypseli. For this reason I usually place it at the end of the 

 PasserinaB, in the vicinity of the anomalous cuculine form just mentioned, which stands in the 

 same relation to the true Cuckoos as the Swallows to the ordinary Song-birds. However, pterylo- 



