44 R. W. SHUFELDT, 
the outer angle of the coracoid; but the upper stops short of it by 
about one-fourth the width of the bone at this part of its sternal 
and. Nothing of this kind isto be seen in Branta canadensis, where 
this posterior facet for the sternum is broad and very deep, and in 
no way divided. In Olor columbianus, however, it is much shallower 
than in the Canada Goose, and very thoroughly divided into two 
facets. Of these, the inner one is broad and deep, and occupies 
rather more than half the width of the bone; the other is very 
shallow and narrow, of an elliptical outline, and separated from the 
larger one by an interval of two or three millimeters. Sometimes, 
in the coracoid of Aix sponsa, there is a faint indication of such 
division, while most ducks agree in this particular with Branta, Chen 
and other geese. 
In Polysticta stelleri, however, the division of this facet is even 
better marked than we find it in Dendrocygna autumnalis; so that 
its significance, with respect to affinity, is certainly very obscure, 
even if there be any. Moreover, the manner of the division is quite 
different in Olor, as compared with what it is in a Tree-Duck or in 
Aix. Indeed, the division is practically the same in Dendrocygna and 
Aix, the two differing from what we find in the Swan. 
In all Anseres, the coracoid is a proportionately large, stout 
bone with a big head, broad, expanded sternal end, with a well 
marked “neck” between the two. 
On the whole, then, the bones of the pectoral arch in Dendro- 
cygna — if we may judge from the two North American species 
of the genus — are more anatine than they are either anserine or 
cygnine. 
The Sternum. Unfortunately I have not at hand, at this 
writing, any embryos of Ducks, Geese and Swans from different 
parts of the world; for, were such material available, the study of 
the sterna in them would be likely to throw more light upon 
anserine affinities, with respect to the groups, than with any number 
of comparisons of the breast-bones of adults. Especially do I refer 
to the development of the fore part of the sternum in the chick; 
for, with but few exceptions, too much weight has often been 
attached to the morphology of the posterior moiety of that bone, as 
compared with the far more important characters that are presented 
on the part of its anterior portion. 
So far as the sternum in adult Anseres is concerned, 
there are a number of figures of it given on the Plates of the 
