Osteological Characters in Ornithology. 41 



the species. Thus it is that the form of the sternum is of such 

 importance, since it has no immediate dependence on external 

 form and habits. The Sparrow, the Flycatcher, the Wren, and 

 the Sunbird, all have one characteristic form of sternum ; while 

 between those of the Swallow and the Swift there is the greatest 

 diversity. 



It is evident also that the modifications of form immediately 

 dependent on habits and external conditions are generally to be 

 seen in the skin even better than in the skeleton of a bird. 

 These are principally changes of form, size, and proportion in 

 the bill, the feet, and the wings, which are excellent characters 

 for distinguishing genera and even families ; while for deter- 

 mining the true affinities of isolated groups we must have re- 

 course to those characters which, having no direct dependence 

 on habits, &c., are often persistent in a remarkable degree. Of 

 these, no doubt, the sternum is of the highest value ; but there 

 are many others of almost equal importance. Such are the tex- 

 ture of the plumage ; the form of the feathers and their arrange- 

 ment over the surface of the body ; the form of the nostrils ; the 

 scutellation of the tarsi ; the mode of nidification, with the form, 

 texture, and colour of the eggs ; the covering of the young bird, 

 and its changes of plumage ; peculiarities of food, characteristic 

 habits, and peculiar attitudes and actions. 



As an instance of the value of such apparently trifling charac- 

 ters as the last, I may mention that the first time I saw a Roller 

 [Coracias temmincki) alive, I was at once satisfied it was a Fissi- 

 rostral bird, from a peculiar jerking motion of the head and tail 

 when it alighted, which is common to Kingfishers, Trogons, 

 Bee-eaters, and Motmots, but never seen in the typical Passeres. 

 In like manner the motions of the Eurylcemi convinced me that 

 they were not Fissirostres, but typical Passeres, as mentioned in 

 my paper quoted at the beginning of these remarks (Ann. Nat. 

 Hist. 1856, p. 199). 



Now that true principles of classification are becoming so 

 much better understood, we may, I think, hope that the chaos 

 which has so long existed in ornithology will soon give way to 

 a truly natural system which must obtain general acceptance. 



