382 PATAGONIAN EXPEDITIONS: PALAEONTOLOGY. 



of two vertebrae more completely fused than in Thylacynus and lacking 

 the dorsal intervertebral fontanelle so conspicuous in the latter genus. 

 The spines are remarkably feeble in contrast with the heavy spine of the 

 last lumbar. The auricular processes are confined almost entirely to the 

 first sacral. The centra are keeled inferiorly, the first having a single low 

 median keel and the second a pair of parallel keels. 



The caudals are remarkably heavy. Those associated with No. 15,170, 

 which have been reproduced in the figures (Pis. LVII, fig. 5 ; LXI, fig. 

 i), are interpreted as the third to the tenth. The proximal ones are 

 short, with stout, posteriorly directed transverse processes. The centra of 

 the more distal caudals increase in length and the transverse processes 

 become wider and shorter. The neural canal is complete as far back as 

 the tenth caudal. It is not possible from the material at hand to deter- 

 mine the length of the tail, but judging from the size of the caudals pre- 

 served, it was probably long and heavy. From analogy with Thylacynus, 

 it has been restored with a vertebral formula of twenty-three. Chevrons 

 are present between the third and fourth caudals and are represented either 

 by the small hatchet-shaped pieces themselves, or by surfaces for their 

 attachment, as far back as the caudal series is preserved (PI. LXI, fig. i). 



A few ribs are associated with both specimens of C. lustratus. The 

 anterior ribs (PI. LVIII, fig. 3) have more cylindrical shafts than in Thy- 

 lacynus. The posterior ribs are slender and sub-round in section. 



Appendicular Skeleton. The scapula (PI. LVII, figs. 2-3) corresponds 

 more closely with that of Prothylacynus than with the corresponding ele- 

 ment in Borhycena or Thylacynus. The neck is short, the glenoid cavity 

 oval in outline and moderately deep and the coracoid prominent, with its 

 anterior margin strongly inflected. In shape the scapular fossae are much 

 as in Prothylacynus. The spine is high, but its free border has been 

 somewhat fractured in both specimens, destroying the acromion. The 

 foramen which perforates the margin of the suprascapular notch in 7Yiy- 

 lacynus appears in Cladosictis some distance posterior to the coracoid 

 border. 



No trace of a clavicle has been found and its introduction in the resto- 

 ration (PI. LXI, fig. i) is conjectural. It may have been rudimentary and 

 not attached to the acromion, as in Thylacynus. 



The humerus (PI. LV, figs. 2, za] is short and heavy, resembling de- 

 cidedly that of Prothylacynus t but differing in the less prominent epicon- 



