Classification of Birds. 109 



upon the one hand and the Auks upon the other, with the 

 Heliornithidge in the next place as a related branch, and one 

 more nearly so than the Laridse or the Procellariidse. In part, 

 this is believed by Dr. Stejneger to be the relationship, who, 

 however, widely dissociates the Hesperornithidae. While 

 this last relationship is fully appreciated by Professor Fiir- 

 bringer, that eminent authority nevertheless apparently sees 

 no special affinity between an Auk and a Loon or Grebe, and 

 so very widely separates the Colymbo-Podicipites and the 

 Laro-Limicola3 assemblages. 



Passing next to the anserine fowls, one would think that by 

 this time there would be more or less unanimity of opinion 

 among systematic ornithologists as to the affinities and position 

 of such a homogeneous group. As a family tlie existing 

 Anatida^ can but contain the Mergansers, Duclcs, Geese, and 

 Swans, while the outliers, eitlier existuig or extinct, are not 

 as a rule very puzzling forms. The anserine affinities of 

 Palamedea are now pretty generally recognized ; and there 

 can be no question as to the relationships of the extinct 

 Cnevnornis or Cereopsis. Moreover the relation borne by 

 the Flamingoes to the Anseres has been known for a good 

 many years past, and yet, notwithstanding all this, we tind 

 almost as much diversity of opinion among the classifiers of 

 birds as to where this very natural group belongs, as has 

 already been pointed out in regard to the Pygopodes. After 

 a careful examination and comparison of the skeleton of Pala- 

 media, all of our United States Anseres except one or two 

 species, the Flamingoes, including the extinct ones and 

 Falaolodus, and a great many species and genera of Herons, 

 Ibises, Storks, Scopus, and their allies near and remote, and, 

 finally, an equal number of the Steganopodes, I was led to 

 believe a year or two ago that the duck tribe in its widest 

 sense, with the allied suborders containing the Palaraedeidte, 

 the Phoenicopteridse, and their fossil relatives, constituted a 

 group, the nearest related branches to which were the Stegano- 

 podes upon the one hand and the Herodiones upon the other. 

 An opinion practically quite similar to this is entertained by 

 Dr. Sharpe and Dr. Stejneger, while on the other hand 

 Dr. Gadow places the Anseritorraes between the Falconiformes 

 upon the one hand and the Crypturiformes upon the other, 

 which of course is an utterly different view of their relation- 

 ships. To discuss these latter here is obviously out of the 

 question, as it would carry the present paper far beyond its 

 limits. Before turning from the Anseres, however, I desire 

 to say that I have found some interesting osteological points 

 in the skeleton of Dendrocygna autiimnalis^ one of the tree- 



