Vol. XII 



19 



jj^'l Shufeldt, Osteology of Cereopsis novoi-hollandicB. 225 



relatively speaking, the preacetabular portion of it is longer and 

 somewhat narrower than the same portion in representatives of 

 the genera Chen, Branta, and Anser (fig. 15, Plate XXX., fig. 25, 

 Plate XXXI.), while it more nearly resembles that part of the pelvis 

 in Olor, either buccinator or coliimhianus . In these Swans, however, 

 the last lumbar vertebra fuses with the first dorso -lumbar, and 

 consequently protrudes beyond the ilia, which it does not do in 

 the case of the Geese. 



Such species as Hymenolcemus malacorhynchits and Chenonetta 

 juhata possess the Duck-like form of pelvis, so comparisons would 

 avail us nothing in this respect with them. (Fig. 5, Plate XXIX., 

 and fig. 26, Plate XXXIV.) 



In Cereopsis, from the somewhat stunted prepubis to the rounded 

 antero -lateral angle of the ilium, on either side, the rather roughish 

 border is concave inward, while the side of the bone above it is 

 quite uniformly concaved, and faces for the most part outward, 

 having nearly the same depth throughout. For rather more than 

 its middle third, the inner, which at the same time is the superior, 

 border fuses completely with the superior border of the neural 

 spines of the dorso-lumbar vertebrae. Posteriorly, on either side, 

 this border curves outward to bound the post-acetabular area 

 anteriorly. (Fig. 15, Plate XXX.) 



Further, we are to note upon this dorsal aspect of the pelvis 

 of this cereopsine Goose that, mesially, in the post-acetabular 

 area, the sacral vertebra;, as well as a number of the leading 

 uro-sacrals, are most completely fused together and with the ilium 

 on either side. So complete is this co-ossification that but a very 

 few minute vacuities remain between the transverse processes of 

 the vertebrse, and these afford but the barest hint as to the 

 limits of the bones involved in the amalgamation. In the case 

 of the last three uro-sacrals, their transverse processes become 

 more and more distinct or individualized as we proceed backward. 

 The transverse processes of the leading one of these three 

 articulate, in the usual manner, with the ilia ; the next following 

 one is much smaller, and its transverse processes do not reach 

 the iliac borders ; while the last one, which closely resembles a 

 true caudal vertebra, is grasped by the ilium upon either side — 

 a conspicuous process being thrown out to meet the ends of the 

 transverse processes of the vertebra. This is well shown in 

 fig. 15 of Plate XXX. There is no approach to anything like 

 this in any pelvis of Goose or Swan thus far examined by me. 



On lateral view there is to be observed the large, circular 

 acetabulum, the inner ring being about one-fourth smaller than 

 the outer. These rings are nearly of equal size in Branta 

 canadensis, but showing only slight differences in their diameters 

 in Olor and Chen. The form of the antitro chanter varies but little 

 among the various genera, and the small, elliptical obtitrator 

 foramen below it is better divided from the large, elongate obtnrator 

 space in Cereopsis than it is in other Geese, or in the Swans. 



From the lower arc of the obturator foramen backward almost 



