224 Shufeldt, Osteology of Cereopsis novcc-hollandiac. [,^f'A"rii 



appendage, the distal extremity of any one of which is truncate 

 almost transversely to its long axis. Any one of these dorsal ribs 

 possesses a somewhat enlarged head, a laterally compressed 

 tubercle, and a slightly constricted neck. On the posterior aspect 

 of the latter, between the capitulum and tuberculum, there is, 

 in the first three pairs of these dorso-vertebral ribs, an unusually 

 large and open pneumatic foramen. There are also smaller 

 openings of this kind on the anterior aspect of the necks and else- 

 where. 



The hcemapophyses or costal ribs are also highly pneumatic, and 

 these bones are likewise laterally compressed, with increasing 

 lengths as we follow them backward. 



While considering the ribs, it will be as well to note that there 

 are two pairs of perfect sacral ribs, agreeing in all respects with 

 those of the dorsal series, barring their gradually increased lengths 

 and the lengths of their long costal ribs, each pair of which reach 

 the costal borders of the sternum in a perfect articulation. The 

 latter pair of these two pairs of " dorso-lumbar " ribs do not 

 support uncinate processes, as do the first pair. 



The next following pair of ribs are rudimentary and largely 

 fused, in the matter of their articulation, with the ilia and their 

 vertebrae. Here, both the pleurapophyses and haemapophyses 

 are long and slender ; the latter, on the right side in this skeleton, 

 reaches the sternum to articulate with the ultimate facet on its 

 costal border ; while, upon the left side, the hamapophysis does 

 not reach the costal border of the sternum by 2 or 3 millimetres — 

 that is, it is a " floating costal rib." 



There is, posterior to these latter ribs, another rudimentary pair 

 of " sacral ribs " with "floating haemapophyses," and, posterior to 

 these again, there may be a final rudimentary pair, with or 

 without floating ribs. As to the exact osteology of these latter, 

 I cannot be quite sure, as they are broken off in the skeleton at 

 hand, and a part of another skeleton before me does not include 

 the ribs. 



Fig. 25 of PI. XXXIIl. of the present paper exhibits very well the 

 broken and imperfect ultimate ribs in this skeleton (No. 18,571, 

 Coll. U.S. Nat. Mus.), while the " skeleton " of Cereopsis novce- 

 hollandicc (No. 19,734, Coll. U.S. Nat. Mus., marked "Body") 

 consists of the cervical and dorsal vertebrae complete ; the four 

 free caudal vertebrae and pygostyle ; the femora ; the imperfect 

 hyoid arches ; a perfect larynx and the first three or four rings 

 and semi-rings of the trachea. 



Branta canadensis and Anser albijrons have four pairs of ribs 

 belonging to the dorso-lumbar vertebrae, articulating with them 

 as in Cereopsis, beneath the eaves of the ilium upon either side, 

 while Chen hyperborcits nivalis possesses but three pairs of them, 

 the last pair being very rudimentary and slender. 



The Pelvis and Coccygeal Vertebra. — Agreeing with a number of 

 other species of existing AnserincB, Cereopsis possesses an elongate 

 and comparatively narrow pelvis — indeed, actually as well 



