Grmip of the Cplumbid(e, 471 



liarity of economy is almost invariably accompanied by some cor- 

 responding modification of structure, we shall receive considerable 

 assistance by accurately examining such variations. We may note 

 the forms, without being acquainted with their reference to the 

 peculiar habits of the group; and, although our inference, in 

 some cases, may be erroneous, in others we shall not be far from 

 the truth. The Passenger Pigeons, for instance, have their first 

 yuill feather as long as any of the others ; a sure indication of 

 that rapid and long-continued power of flight they are known to 

 possess. The Columbi-GalUnes of M. Le Vaillant* are described 

 as having naked and somewhat lengthened tarsi ; a structure well 

 adapted to those ambulating habits, which bring some of them 

 close to the Phasianidw, Vig., and others to the Crucidce, Vig. 



Another group,+ the Co/o/«6ar5 of M.Le Vaillant (Fewag-o,Cuv., 

 Treron, Vieil.,) have a strong hard bill ; and their short clasping 

 tarsi covered with feathers, leads us to conclude they seldom perch 

 upon the ground; in fact, MM. Le Vaillant andCuvier both assert 

 that these birds are only found in the tropical forests of the Old 

 world. Apparently confined to the same regions, we see another 

 group, wherein the bill partakes of that weak structure observed 

 in the generality of Pigeons, while the tarsi are thickly clothed 

 with feathers, similar to the group last mentioned. These seem 

 to be the principal divisions among the Columbidw. Minor 

 distinctions may be founded on the relative length and structure 

 of the quill feathers, which in some are very peculiar; and, as 

 being connected with the powers of loco-motion, deserve our 

 attention. Some writers have attached more importance to the 

 form of the tail ; and have therefore divided the Columbidw into 

 two great divisions : separating such species as have this organ 

 rounded or lengthened, from those in which the tail is short and 

 even. This plan, however it may help to discover a species, is 

 obviously artificial, and totally inadequate to give us any ideas on 



• Le Colombi-galliiie a Carnail. Oi». d'Af. 6 tab. 279. Le Colombi-Hocco, 

 {il. 280. Le Colombi-perdix a cravnte noire, pi. 28K Le Colonibi-perdix 

 roux-violet, pi. 282. Le Colonibi-caillo, pi. 283. 



+ I^ Colonibar. Oin. d'Af. 6. pi. 276, 277. Columbacapellei. Teni. I'l. 

 Col. liv. 24. lab. 143. &c. 



