xra PHYLUM CHORDATA 431 



*t 



no evidence that any member of that order was arboreal, or showed 

 the least tendency on the part of the fore-limb to assume the 

 wing-form. Nevertheless the skull and brain of Ornithosauria 

 and the pelvis and hind-limb of many Dinosauria show such 

 approximation to avian characters as can hardly be without 

 significance. 



Probably the earliest Birds were all, in the etymological sense, 

 Carinatse, i.e., had the sternum provided with a keel for the attach- 

 ment of the pectoral muscles. Probably, also, they all possessed 

 teeth, and had diverged into well-marked orders before those organs 

 were lost. The Odontolcae, for instance, have their nearest allies 

 in the Divers (Pygopodes), while the Ichthyornithes resemble the 

 Terns, members of the widely separated order Gaviae. 



In several existing types of Carinatae the power of flight is want- 

 ing, and in all such cases it is practically certain that Sightlessness 

 is due to the degeneration of the wings : in other words, that the 

 ancestors of the Penguins, Great Auk, Dodo, Weka (Ocydromus), 

 Kakapo (Stringops), &c., were ordinary flying Birds. In the 

 Penguins and the Great Auk the wings have simply undergone 

 a change of function, being converted into paddles, and con- 

 sequently the only parts of them which have degenerated are the 

 feathers ; but in the other forms referred to the wing has become 

 more or less functionless, and hence has diminished in size, while 

 the partial atrophy of the muscles has resulted in a more or less 

 complete reduction of the carina sterni and furculum and an increase 

 of the coraco-scapular angle. Now it is by an exaggeration of these 

 peculiarities that the Ratitse are distinguished from the Carinatse, 

 and there is every reason for thinking that they also are the descen- 

 dants of flying Birds, and that their distinctive characters absence 

 of locking apparatus in the feathers* flat sternum, wide coraco- 

 scapular angle, &c. are all due to degeneration correlated with 

 disuse of the wings. From the fact that the dromaeognathous skull 

 is more reptilian than any other type, it would seem that the 

 Ratitse diverged early from the carinate stock. From the fact 

 that, in the structure of the skull and pelvis, the Ostrich and Rhea 

 are widely separated both from one another and from the Austral- 

 asian Ratitaa, it seems probable that the three orders of Ratitse 

 arose independently from primitive Carinatse, and that the entire 

 division is to be looked upon as a convergent or polypliylctic group, 

 owing its distinctive characters, not to descent from a common 

 ancestor, but to the independent acquisition of similar characters 

 under the influence of like surroundings. 



The question of the phylogeny of the orders of Carinatae is far too 

 complex to be discussed here. Suffice it to say that the Ichthy- 

 ornithes, Odontolcae, Impennes, Pygopodes, and Crypturi are to be 

 looked upon as the lowest or most generalised orders, while the 

 highest or most specialised are the Psi.ttaci, the Accipitres, the 



