224 Remarks on the variation of affined species. [No. 3. 



the trachea in the female alone of Rhynchea australis, which pecu- 

 liarity does not occur in either sex of Rh. bengalensis. 



Following up this enumeration of the variety of modes of differing 

 among closely affined races of birds, it may next be remarked that a 

 great difference of voice and of habits may be only indicated in the 

 structure by minute variations in the form of particular feathers ; e. g, 

 Corvus corone and C. americanus : — Pernis cristata is only distin- 

 guished from P. apivora by an occipital tuft of lengthened feathers 

 more or less developed, in addition to its different habitat ; and in 

 Spizaetus limnaetus and Sp. cristatellus, the last named has a similar 

 occipital crest generally much more developed, this being again the 

 chief distinction besides that of geographical distribution, and that 

 the former race assumes an ultimate phase of plumage which is never 

 (so far as we can learn) seen in the other. The very different form of 

 the crest and adjacent plumage is again the only distinction we are 

 acquainted with between the larger Indian Pelican (Pelicanus javanicus) 

 and the closely affined African species (P. onocrotalus). In many 

 other instances the distinction is best shewn in the varying relative 

 proportions of the wing-primaries, or even in that of a single primary, 

 as exemplified by Acrocephalus arundinaceus, (L., vel Sylvia tur- 

 doides, Tern.,) of Europe, and Acr. brunnescens, (Jerdon,) of India.* 

 Pycnonotns jocosus of Burma and Penang has always a shorter and 

 more intensely crimson ear-tuft than P. jocosus of India, and we have 

 been assured that the voices also differ. The Irena puella of India, 

 and also of Arakan and the Tenasserim provinces, differs constantly 

 from that of the Malay countries by having shorter tail-coverts. 



Then we have cases in which sundry of the foregoing differences are 

 variously combined. In Loxia himalayensis, L. curvirostra, and L. 

 pytiopsittacus, the size is successively larger, with a successively more 

 robust conformation. So likewise in Gracula javanensis and Gr. 

 intermedia. The restricted Edolii differ slightly in size only, except that 

 the larger have successively the frontal crest proportionally more 

 developed. In Cannabis linaria (Fringilla linaria, L., v. Linaria 

 canescens, Gould), as compared with C. minor, a difference of size is 

 combined with a very slight one of plumage, and the song-notes are 

 here again distinguishable. In Pratincola atrata, nobis, of the high- 



