“Tree-Ducks” of the genus Dendrocygna. 59 
Dendrocygna has the bones of its pectoral arch more anatine 
in character than cygnine. 
Dendrocygna autumnalis has a sternum that might answer for 
that of any ordinary duck, as for example Aix sponsa; while on the 
other hand, it agrees with the sternum of Chen hyperboreus nivalıs, 
all but the difference that the latter bird possesses a small manubrial 
process upon its sternum, which that bone lacks entirely in Dendro- 
cygna. 
With respect to the characters to be found in itsappendicular 
skeleton, Dendrocygna autumnalıs presents an interesting mixture 
indicating anatine, anserine and cygnine affinities, first one and then 
another preponderating as various comparisons are made. It would 
be almost safe to say that these Tree-Ducks, in so far as the 
appendicular partoftheirskeleton goes, and the comparable 
characters it presents, are just as nearly related to the Anatinae 
as they are to the Anserinae or Üygninae, or to either of the 
latter as they are to the former. 
On the whole, however, the characters presented on the part of 
the skeleton in Dendrocygna autumnalis point very clearly to the 
fact, that it has a closer affıinity with the Anatinae, than it has 
with either the swans or the geese, and probably the other members 
of the genus are similarly related. 
On the Systematic Position of the Tree-Ducks. 
Having demonstrated that representatives of the genus Dendro- 
cygna are, morphologically, more nearly related to the Anatinae 
than to any other subfamily of the Anatidae, it now but remains 
to present a taxonomical scheme, which will indicate the position 
they occupy in the system. 
Almost without exception the most eminent ornithologists and 
avian taxonomers of the Old World have, long ago, relegated them 
to a place where they belong among the Anatinae, recognizing, 
as they do, the fact that they are certainly neither swans or 
geese. | 
In the first part of this paper I have already presented the 
views of some of the authorities on this question, as SHARPE, EYToN, 
SALVADORI, and others. 
More than three years ago, I pointed out in my *Osteology of 
