340 NEW YORK STATE MUSEUM 



containing the Anseres, the Pygopodes, the flamingoes, the Stork- 

 Heron assemblage, the Accipitres, and the Steganopodes, were more 

 or less nearly allied to each other. By the extinction of many in- 

 termediate forms, however, in time the gaps standing among these 

 several avian suborders are so wide, and so irregular, that to define 

 their now existing affinities becomes a matter of no little difficulty, 

 and doubtless in some directions is quite impossible. That ancient 

 bird forms with lengthened pelvic limbs occurred in anserine an- 

 cestry there can be no doubt, and the same may be said for the 

 accipitrine stock. Gastornis points to this, and such existing out- 

 lying suborders as the Phoenicopteri (flamingoes), and the Pala- 

 medeae (screamers) strong'ly indicate this for the Anseres, and the 

 Secretary-bird (Serpentarius) bespeaks the same for the accipitrine 

 branch. 



Judging from the osteological characters it presents it would 

 seem that the suborder Palamedeae, containing the family Palame- 

 deidae, belongs to a line of stock that at one time was very close 

 to the ancient anserine one, and consequently this must be indicated 

 in our avian classificatory schemes. The suborder Odontoglossae 

 is separated from the Anseres by a less evident gap perhaps, while 

 the kinship with the Pygopodes "is more remote. Such problems as 

 these can never be settled until accurate and exhaustive and com- 

 parative studies have been made of a great many embryonic stages 

 of the forms that have just been mentioned, from the egg to the 

 adult. We need especially embryological studies of the flamingoes, 

 the screamers, the drivers, the Accipitres, Anseres, and others, 

 and science will be very fortunate if even these studies tend to 

 clear up avian alliances that are now but very imperfectly under- 

 stood. The best that can be suggested at the present time is that 

 the Anseres be included in a suborder, which, in so far as our 

 American avifauna is concerned, contains the family Anatidae ; and 

 this last may be divided into four subfamilies, as above pointed out, 

 viz: the Merginae, the Anatinae, the Anserinae and the Cyg- 

 ninae. The remains of anserine birds, now extinct, have been found 

 from time to time ; and these doubtless were the representatives 

 of other families no longer in existence, though in any scheme of 

 taxonomy that is made to include fossil and subfossil forms, these 

 types must find a place, and in a number of instances the classifiers 

 of birds have done this, as, for example, Fiirbringer, Sharpe, and 

 Stejneger, 



