LIFE IN TERTIARY TIMES 285 



the evolution of Birds is wrapped in impenetrable mystery. 

 This evolution must have been rapid, because although Birds 

 are descended from a highly specialized branch of the Reptiles, 

 they differ among themselves only in characters that are really 

 secondary. Already at the end of the Cretaceous era some 

 had lost their wings and the keel of their breast-bone, and being 

 more completely affected by this retrograde evolution than 

 those of our present-day birds unable to fly, cannot be con- 

 sidered as their ancestors. This alone suffices to render 

 suspect the natural character of the order of Ratites. Since, 

 however, it is characterized by this inability to fly, it is the 

 only order which has raised hopes of providing some indication 

 of the ancestral form of Birds. Unfortunately, the faculty 

 of flying can be lost as well as acquired ; and it is often 

 difficult to distinguish retrogressive from initial forms. The 

 earliest known Tertiary Birds are Gastornis, Diatryma, 

 Dasornis, and Remiornis, all Eocene. There is no reason for 

 placing the one that lived in France during the Tertiary 

 Period — Gastornis — at the head of their geneaological tree. 

 Some naturalists make it a Goose, some a Bustard of the size 

 of an Ostrich and unable to fly. The others : Diatryma 

 of New Mexico, Dasornis of the London clay, and Remiornis 

 of the neighbourhood of Rheims, are all incompletely known. 

 Only a metatarsal of the first, skull fragments of the second, 

 and some fragmentary bones of the third have been discovered. 

 It is too little to warrant us in drawing inferences as to the 

 structure of primitive Birds. 



In the Miocene strata at Santa Cruz in Patagonia, 

 Dr. Ameghino dug up a whole series of Birds, which have been 

 grouped by Morenco and Mercerat under the denomination 

 of Stereornithes. But this grouping would seem to be entirely 

 artificial. Of its constituent genera, Mesembriornis seems 

 to be akin to the Nandus which still abound in South America ; 

 Dryomis was a bird of prey related to the Condor; Dicholophus 

 resembled the Cariama ; Phororhachos, with its enormous 

 skull, its upper mandible terminating in a strong hook and 

 the lower bent up over it, remains enigmatic. 



The Ostriches may perhaps be nearer to the initial type 

 than any of these fossil Birds. The digits of their wings 

 approximate more closely to the ordinary digital type than 

 those of all other Birds ; and they present a pubic symphysis, in 



