THE OSTEOLOG Y OF SINOPA—MA TTHEW. 221 



Hojylophoneus has the transverse processes of the cervicals, including 

 the atlas, greatly extended posteriorh\ A distinct inferior lamina 

 develops on C 5-6. The dorsals are not unlike Slnojxu but the posterior 

 ones develop powerful anapophyses, which are continued into the lum- 

 bars with decreasing strength. The lumbars are not so long, their spines 

 much like those in Shiopa^ transverse processes considerably shorter. 

 The sacrum is not so wide anteriorly, the spines higher. The trans- 

 verse processes of the two iirst caudals are expanded into plates, in the 

 next three they slant backward, and are not so stout as in Sinopa; the 

 transverse lamellie of the middle caudals are much less developed. 

 The neural arches are continuous as far as the thirteenth caudal, but 

 the middle caudals are smaller and the tail shorter. 



RIBS AND STERNUM. 



There are thirteen ribs. They are distinguished, especially the 

 anterior ones, by exceptional shortness and rather broad, flattened 

 shafts. The flattening of the shaft is more uniform from end to end 

 than in dogs or viverrines, in which it is restricted to the distal middle 

 section of a few ribs, and the others are much rounder and somewhat 

 more slender and elongate. I do not lind any Carnivora presenting 

 the rib features of Suwpa. The}' do not appear to be marsupial char- 

 acters, but are seen in earl}- ungulates and (except the length) highly 

 developed in the more recent ones. 



Five sternal segments are preserved, A\\ of them are of the narrow 

 elongate type usually seen in Carnivora, and show none of the flatten- 

 ing observed by Wortman in Mesonyx and mentioned by him as a 

 characteristic marsupial feature. It is also characteristic of most 

 ungulates and of many other groups. 



APPENDICULAR SKELETON, DESCRIPTION AND COMPARISONS. 



Forelimh. — The limbs are remarkably small in comparison with the 

 proportions of skull and backbone. They do not exceed those of a 

 large domestic cat in length, although somewhat stouter, while the 

 backbone (exclusive of the tail) is one-fourth longer than in that 

 animal and the skull nearly twice as long. 



The scapula is incomplete, the anterior border not being preserved 

 on either side. Its general form, so far as comparison can be made, 

 agrees best with Canis^ being rather long and narrow for a carnivore, 

 and the upper border at right angles to the posterior. The spine is 

 about as high as in Canis, considerabl}' lower and less overhanging 

 than in Felis or Viverra. The acromion is much better developed 

 than in Canis; apparently considerably more than in Felts or Viverra. 

 Its tip is broken off,, but it projects considerably bej^ond the glenoid 

 cavity. The coi'acoid process is short but very distinct, as in Felis 



