GEESE. 



137 



The greater part of a well-preserved skeleton, from which Dr. Hector demon- 

 strated the affinities of Cnemiornis to the Anatoideae, was found in the same cave 

 which furnished the very interesting specimen of a moa's neck with muscles, skin, and 

 feathers, referred to on a previous page. It differs in several important respects from 

 that described by Owen as C. ccdcitrans, the principal discrepancy being the greater 

 proportional size and somewhat different structure of the humerus, and may be speci- 

 fically distinct. The skull proves the relationship to the Anatoidea3, but is remark- 

 ably robust and short, the palatines are firmly united posteriorly with the vomer, the 

 upper surface of which has a 

 slight groove to receive the 

 prasphenoid ; the basisphenoid 

 has large oval basipterygoid 

 facets; anterior nostrils large, 

 holorhinal; a soft cere has 

 probably covered the basal part 

 of the mandibles, while only 

 the part in front was horny, 

 as in Cereopsis ; the character 

 of the tympanic cavity is quite 

 peculiar, being bridged across 

 by a bony process between the 

 mastoid process and the basi- 

 occipital. The breast-bone is 

 large, deep, nearly square, and 

 without indentations or pro- 

 cesses behind; it has a trace 

 of a low keel in the anterior 

 part, which rises less than one 

 fourth of an inch above the 

 convex surface. The meta- 

 carpal elements are completely 

 fused at both extremities, the 

 length being about two fifths 

 of the length of the humerus. 

 The first six ribs had well- 

 developed uncinate processes. 

 The pelvis is very robust ; the ilium and ischium unite posteriorly, closing the ilio- 

 sciatic foramen behind. In the hind limbs we note the peculiarity of the presence 

 of a high epicnemial process, as in Colymbus, in front of the knee-joint, and hence 

 the name. 



Dr. Hector estimates the height of the bird's back above the ground to have ex- 

 ceeded two feet, and the total length to have been at least three feet. 



That the nearest living relative of Cnemiornis hails from Australia, and is quite 

 peculiar internally and externally, cannot but strike us as natural in view of our expe- 

 rience with other groups. A glance at the admirable full-page cut representing 

 Cereopsis novce-hottandice, the only species of the family CEREOPSID^E, shows us at 

 once a bird in general aspect resembling a goose, but provided with a quite unique 

 bill, a heavy and short frame, very stout feet, and deeply indented palmation of the 



FIG. 63. Skeleton of Cereopsis. 



