1 883.] 



ANATOMY OF PHCENICOITERUS, 



651 



long flexor of the hallux are almost the only points in which it differs 

 from a Stork. 



In conclusion I may point out an osteological comparison of some 

 importance, which so far as I can learn has not been noticed. 



In the group of Storks the number of dorsal vertebrae is very 

 constantly y?2;e, of which cue is covered by the ilium, the "sacral" 

 vertebrae (all those between the last dorsal and first free caudal) 

 being as constantly fourteen, while the free caudals, excluding the 

 pygostyle, are five or six in number. 



The following table shows the relations in this respect of 

 Storks, Flamingoes, and those Geese which I have examined. 



[In this table the dorsal vertebrae are separated by the + sign : 

 those in front of it being free, those behind it being covered by the 

 ilium. Thus 5 + 3 means five free dorsal vertebrae and three 

 covered by the ilium.] 



Further, in all the ahove-named Storks and in Phanicopterus 

 the radius is more than twice as long as the first metacarpal, while in 

 all the Geese that I have examined it is less than twice as long. 



I think, therefore, that while the skull and larynx of Phoenico- 

 pterus, together with its webbed feet and the characters of its bill, 

 undoubtedly connect it with the Lamellirostres, yet the rest of its 

 organs — its air-cells, its muscles, its alimentary canal, its vertebral 

 column, and the characters of its wing-bones — show close relation- 

 ship with the Storks. 



In fact, if we assume that the typical Chenomorphce are descended 

 from a group similar to that now represented by the Screamers, with 

 their simple desmognathism and complete muscle-formula, we may 



