270 



Mons. A. Milne-Edwards on the 



within very wide limits, not that one can attach a very great 

 importance to these differences. This may be proved by a 

 glance at the following numbers, which show the proportion of 

 the tarso-metatarsus to the tibia — the length of the latter being 

 taken at 100 : — 



Fossil Metatarsus G7 



Ocydromus avistralis . . »>7'5 



Tribonyx mortieri .... 64 



Aramicles cayennensis . . 72 



Metopidius africanus . . 71 



Porpliyrio madagascariensis 68 



Rallus crex 64 



aquaticus 64 



Gallinula cliloropus 62* 



Fulica atra 57 



Thus in our fossil bird the proportion of the leg to the foot 

 was nearly the same as in Porphjrio, Tribonyx, and the Rails. 



To sum up, we see then that the remains the characters of 

 which I have just been examining belong evidently to the bird 

 which Herr von Frauenfeld has recently figured, and that they 

 are sufficient to indicate clearly the systematic jjosition of this 

 remarkable animal. It evidently was one of the family Rallidce, 

 and there is much less difference between it and Oajdromus 

 than between this last and the (true) Rails. It constitutes 

 in this group one of the transitional forms so remarkable in 

 the animal kingdom, and should be regarded as a Ralline the 

 organization of which was adapted to an essentially terrestrial 

 existence. The feathers of the wings are too slight and offer far 

 too little resistance to have been of use in flight ; and, besides 

 this, the wings themselves are rudimentary. The feet, on the 

 contrary, show considerable strength ; but they are only slightly 

 elevated, and the toes are less elongated than is usual in this 

 family. This last fact gives us reason to think that this species 

 had less aquatic habits than most of the Rullido!. The hind toe, 

 however, is very long, as in birds which haunt muddy places 

 or a soil of little consistency — although in the true Runners it 

 disappears more or less completely, so as to diminish the weight 

 of the arm of the lever formed by the foot. It may be seen by 

 the nature of the feathers that the Poule rouge was still more 

 brevipennate than Notornis; and it is also probable that the 



* [In M. Milne-Edwards's paper (Ann. Sc. Nat. id supra) llie propor- 

 tional length in this species is given as " 72."— Er.] 



