SOME PALEARCTIC BEARS. 



89 



measure also the skull of a very old male which was brought 

 home from Thibet by Dr. Sven Hedin. The latter skull has the 

 teeth extremely worn, so that their dimensions are of little value, 

 but the other measurements are u.seful for comparison : — 



Greatest length of skull 



Condylobasal length 



Basicranial length 



Zygomatic width 



Length of nasals 



Length of palate from gnathion 



Width of palate inside middle of w^ 



Distance ivovsx foramen laorymale to gnathion. 



Breadth of brain-case 



Interorbital breadth 



Mastoid breadth 



Hind margin of m^ to front of i' 



Combined length ot f*, ni^, and mr 



P' 



m^ 



m^ 



Combined length of ^4-mg 



i'4 



m\ 



m,<i 



»J3 



Bear skulls are very variable in shape, and it is thus rather 

 difficult to say whether a certain characteristic exhibited by an 

 individual is of any taxonomic value or not. In the pruinosus 

 specimen from Kansu the nasals are rather long, and their pos- 

 terior ends extend much beyond the frontal processes of the 

 maxillai-y. The mesial length of the nasals is also longer than 

 the mesial frontal suture. In Swedish Bears the opposite as a rule 

 prevails, and I believed, therefore, that this might constitute a 

 characteristic of some impoi-tance. This is, however, not the case, 

 because I found later that, in a young Swedish Bear from the 

 province of Jamtland, the nasals were just as in the pruinosics 

 specimen, being considerably longer than the mesial frontal 

 suture. 



For the preseiit it appears very difficult to point out any 

 definite cranial characteristic, which in every case holds good as 

 distinctive between the pruinosits and arctos groups, except those 

 that are dei-ived from the teeth, or stand in connection with their 

 development. 



The teeth of the pruinosus group are, especially the molars, 

 very mvich larger than the corresponding ones of the arcios 

 group {cf. the table above). The same is also the case with p* and 

 p^. On m^ of the former the cingulum. is quite well marked 



