ORDER CARINAT^. 1 229 



1 1 2-1). Remains of Gastornis have been recorded from Meudon, 

 near Paris, from Rheims, and from Croydon, and have been referred 

 to four species. The huge Diatryma, from the Lower Eocene of 

 North America, appears to be closely allied to, if not identical with, 

 Gastornis. Bird-bones from the Tertiary of South America, de- 

 scribed by Dr Moreno under the name of Mesembriornis, and com- 

 pared to the Anseres, probably indicate a member of this group, 

 since they are fully as large as the corresponding bones of the 

 Ostrich. 



Here also may be noticed an imperfect cranium from the London 

 Clay, described by Sir R. Owen as Dasornis, and regarded by him 

 as pertaining to a Ratite Bird. And it may be also mentioned that 

 an imperfect limb-bone from these deposits, which has been named 

 Megalornis, may perhaps belong to the same genus. 



Order III. Carinatte. — The third order, which is now by far 

 the most numerously represented, is generally characterised by the 

 presence of a median keel to the sternum, and by the long axes of 

 the adjacent portions of the scapula and coracoid forming at their 

 junction an acute or slightly obtuse angle, as well as by the upward 

 extension of the ischium towards the ilium (fig. 1102). In most 

 cases the wings are adapted for flight, but in some instances they 

 have become atrophied, while in others they are modified into 

 swimming organs. 



There is still considerable diversity of views obtaining among 

 Ornithologists as to the classification of this order, and all such 

 schemes must consequently be regarded as more or less provisional. 

 In the present work the classification adopted by Mr P. L. Sclater 

 is followed in the main, although certain modifications suggested by 

 Professor Newton have been incorporated. 



Before noticing those fossil forms which are susceptible of being 

 placed in definite groups it will be advisable to mention briefly 

 certain remains of which the affinity has not yet been determined, 

 although it is probable that at least the majority should find a place 

 in the present order. As to Laopteryx, of the Upper Jurassic of 

 North America, there appears, as already mentioned, to be consider- 

 able doubt whether it is really Avian at all. In the Cretaceous of 

 the same country we have Afiatornis, G?'aculavus, Laornis, Palceo- 

 tringa, and Telmatornis ; many or all of which probably belong to 

 the first suborder. In England the oldest known bird-remains 

 occur in the Cambridge Greensand, and have been named Ena- 

 liornis, although it is quite probable that they may indicate more 

 than one genus. Some of these vertebrae have more or less flat- 

 tened centra ; while the tarso-metatarsus, in which the fusion of the 

 component elements is incomplete, is compared to that of the exist- 

 ing Colymbus and also to that of Ichthyornis, and it is highly prob- 



