THE CLASSIFICATION OF BIRDS 41 



CORACIIFOEMES. 



The Coraciiformes may be divided into two Suborders : 

 The Caprimulgi are aquincubital ; the spinal-feather tract is 

 well defined on the neck and forked on the upper back ; some 

 down is retained on the bare spaces in the adult ; the oil-gland is 

 not tufted (naked or absent) ; the arrangement of the palatines 

 is not segithognathous (schizognathous or desmognathous). 



The Picarice are mostly quincubital, but a few species are said 

 to be aquincubital ; no down is retained in the adult plumage 

 except in the Cypselidce, where it is confined to the bare spaces, 

 and in the Alccdinidce, where it occurs on the feather- tracts and 

 sparingly on the bare spaces ; they have no basipterygoid pro- 

 cesses ; the oil-gland is present (naked or tufted) ; the arrange- 

 ment of the palatines is not schizognathous (segithognathous or 

 desmogn athou s). 



17. CAPRIMULGI. 



The Caprimulgi consist of three families. The Caprimulgidcv 

 and the Steatornithidce have basipterygoid processes and oil-glands 

 (naked), both of which are wanting in the Podargldce'. The 

 Steatornithidce and the FodurgidoB are desmognathous, but the 

 Caprimulgidce are schizognathous. 



The occurrence of aquincubital species as well as quincubital 

 species in the Cypsdidce and Alcedinidce, a statement which, if 



ing in trees. It is obvious that the ambiens muscle could be independently 

 acquired (or lost, if Garrod's view be adopted), as there are at least eight sharply 

 defined groups which contain some species with an ambiens muscle and others 

 without. These are the Columbae, Psittaci, Steganopodes, Herodiones, Tubinares, 

 Gavise, Pygopodes, and Casuarii. 



It is not easy to explain the number of Orders which contain some families 

 with the ambiens muscle and others without. There seems to be some Orders in 

 which it is exceptionally absent, and others in which it is exceptionally present. 

 The birds which are contained in the latter orders are for the most part inhabi- 

 tants of the air, perching freely in trees ; those in the former live principally on 

 the ground or on the water. The ancestral birds probably lived in trees, and, 

 we may presume, rapidly developed the muscles of their fore limbs at the expense 

 of those of their hind ones. It is not improbable, however, that before the 

 ambiens muscle had been completely lost, 6ne or two groups of birds so far 

 returned to the terrestrial habits of their preavian ancestor that the suppression 

 of the ambiens muscle was arrested, and it became advantageous to some of them 

 to re-develop the partially lost muscle. This explanation seems to be more 

 probable than the assumption that the muscle was independently developed in 

 widely distant groups. 



