30 



PLEISTOCENE MAMMALIA. 



5. The crown of in. 3 is usually more angular 

 behind than in U. horribilis. There is usually 

 no sinus or constriction on the outer border. 

 The grinding surface presents a few coarse 

 folds, but is never tuberculated in the slightest 

 degree (Busk). 



6. The jugal arcade is more circular (Busk aud 



Adams). 



7. The posterior narial openings are wide (Busk 



and Adams). 



The crown of m. 'S is usually less angular behind 

 than in U. arctos. In teeth of the typical 

 triangular form there is no sulcus on the outer 

 border. When the tooth is more elongated it 

 presents a shallow sinus dividing the outer 

 border. The grinding surface is coarsely ridged, 

 rarely tuberculated (Busk). 



The jugal arcade is more elliptical (Busk and 

 Adams). 



The posterior narial openings are of medium 

 width (Busk and Adams). 



The angular crotchet is less thick and incurved 

 than in U. horribilis (Busk). The coronoid 

 process is rather less broad and high (Owen). 



The angular crotchet is thicker aud more incurved 

 than in U. arctos (Busk). The coronoid pro- 

 cess is rather broader and higher (Owen). 



The claws are 

 (Merriam). 



less long and straight The claws are longer and straighter (Merriam). 



The constancy and importance of the above supposed distinctions may now be 

 considered. 



(1) The differences to which Busk refers are very slight, and so far as my 

 own observation goes, quite inconstant and unreliable. 



(2) The skulls of U. horribilis in the Zoological Pepartment of the British 

 Museum do not show any marked projection of the cingulum at the antero-internal 

 corner of pm. 4 , or that the tooth tends to be shorter than in U. arctos. 



(3) Busk considered that the relative length of the interspace between c. and 

 pm. 4, on which Owen laid stress, was not constant. This is also shown by the 

 measurements in the table on p. 11. 



(4) Nearly all palaeontologists have laid stress on the structure of pm. 4, this 

 being specially the case with Busk. Lydekker considered that Busk attached undue 

 importance to the structure of the talon. Brown, too, remarks that two skulls of 

 [J. horribilis in the British Museum do not possess the longitudinal ridges con- 

 sidered by Husk to be characteristic of pm. 4 in this animal, while on the other 

 hand a skull of the Isabelline bear, a variety of I 1 , arctos, possesses them. 



(6) The elliptical character of the jugal arcade is variable. In the case of two 

 grizzly bear skulls in the College of Surgeons' Museum, in No. 850, the jugal 

 arcade is more elliptical than in the brown bear skull No. 836, while in the grizzly 

 bear skull No. 854 it is not more elliptical. The skull from Ballymahon in the 

 British Museum, attributed by Adams to the grizzly bear, has the jugal arcade not 

 more elliptical than the brown beat- skull No. 218e in the British Museum. 



(7) It is generally the fact that the posterior narial opening is wider in the 



