58 R. W. SHUFELDT, 
the length of the basal joint of the inner toe equals that of the 
basal joint of the middle one — or very nearly equals it; while in 
Olor, as well as in Dendrocygna, the length of the basal phalanx of 
the inner toe is always less than that of the basal phalanx of the 
middle toe. 
This is also the case, however, in Cereopsis novae-hollandiae, and 
Branta canadensis, too, agrees in this matter with the ducks. 
Brief recapitulation of the skeletal characters in Dendrocygna 
as compared with those of other Anseres. 
In the skull, Dendrocygna has a character not found — in so 
far as I know up to the present time — in any other known 
anserine bird, and that is: its orbits are completely surrounded by 
bone, through a meeting and extensive coossification of the extremities 
of the sphenotic process and the lacrymal bone of either side. An 
approach to this condition is found in both ducks and swans in 
certain genera, as in Olangula and in Olor. 
Apart from this character, the skull of Dendrocygna agrees, in 
most particulars, with the skull in any typical species of the genus 
Olor; while, at the same time, it differs in many important points 
with the skull in all the genera of geese examined by me. It 
comes very much nearer the skull of such a duck as Anas platyrhymchus 
than it does to that of any goose or swan. 
Dendrocygna is, in the matter of the morphology of its respira- 
tory and vocal organs, more closely related to certain ducks, 
than it is to any of the geese, and in this respect has no affinity 
with any of the Cygninae with which I have compared it. 
It is an important as well as a very significant fact that, 
in its vertebral column and ribs, Dendrocygna distinctly 
agrees with certain Anatinae, and not with any of the geese 
or swans. 
With respect to its pelvis, it may be said that, in its general 
form, it resembles the pelves of some Cygninae, and to a less 
degree that of certain Anserinae. This refers especially to the 
bone being long and narrow, while in other particulars the 
pelvis of Dendroeygna autumnalis distinetly agrees with that part of 
the .skeleton in certain ducks, and, as a matter of fact, comes very 
much nearer the pelvis of Mergus serrator than it does to the pelvis 
of any of the latter. 
