“Tree-Ducks” of the genus Dendrocygna. 45 
present paper, — not only of Dendrocygna but of other ducks and 
some of the geese. 
As for the Cygninae, the morphology of the sternum in them 
has long been known through many figures and descriptions of it, 
which have, from time to time, been published. It is very different 
from the sternum as we find it in Dendrocygna autumnalis or D. 
bicolor, and a comparison of the bone, in the case of these Tree- 
ducks, with the sternum of Olor columbianus, for example, would be 
a more or less profitless task. Beyond each possessing a „pair of 
notches” in the sternal body, posteriorly, no other characters in any 
way agree. 
Briefliy it may be said, however, that in Olor the sternum has 
a large manubrium; eight facets on either costal border; the 
entire keel is hollow to admit a loop of the trachea, which latter 
is colled in an osseous box on top of the sternal body, — to say 
not a word with respect to other differential characters. 
In Dendrocygna autumnalis there may be six or seven artieular 
facets on either costal border for the costal ribs, as I have already 
shown when describing the latter above. The first two are close 
together, while the rest are each separated by an oblong, shallow 
concavity, in each of which small, pneumatic foramina appear. Quadri- 
lateral in outline, the thin, superior border of either costal process 
— the processus lateralis anterior of GAnow — has, at its 
middle, a low but distinct process, which is more conspicuously 
developed in Branta. This process is rather broad, too, in Aix 
sponsa, but not in Harelda, Aix galericulata and many other ducks. 
It is of unusual form in Polysticta stelleri, as not only its angles are 
produced as processes, but the aforesaid apophysis is spine-like, and 
there is still another one, also spine-like, on its anterior sharp border. 
In Olor they are low, with their inner surfaces looking almost 
directly upward. Mergus has them much as they exist in Dendro- 
cygna, though the little process on the upper border is absent.) 
Seen upon its dorsal aspect, the body of the sternum in Dendro- 
cygna autumnalis (454) is very much concaved, this concavity being 
deepest anteriorly, and very gradually diminishing, as we proceed 
backward, to an indefinite, transverse line between the anterior 
points in the contours of the “notches”, where it is shallowest, and 
1) SHUFELDT, R. W., Östeology of Birds, figs. 3 and 4, 24, 26, 
27 and tab. 2 of the Anseres. 
