206 Dr. C. "W. Ajidrews on a 



sonic specimens ilio inner cusp is not niucli developed, as 

 in tlie tooth from Bacton (16448) figured by lloynold.s (pi. vi. 

 fijf. G c), and still less in another specimen (i\[ (Jl'.tO) from Side- 

 strand ; on the other hand, in a specimen (M i'A)d5) figured 

 by Keynolds on the plate just quoted (tig. «), tha com- 

 plication is much greater, but in no case does it approach 

 that seen in the normal /i77?4 of the Pleidtocone Ursus 

 spelceus. 



In both Ur.-'Us spelceus and U. savini the maialibular rami 

 of some old individual?, j)robably males, may Ix-come much 

 deepened beneath the posterior niolais. This is especially 

 marked in one very massive mandibular ramus (M G18G) 

 from Bactou ; in this case, however, this peculiarity may 

 have been partly due to a diseased condition. This deepening 

 of the posterior part of the mandibidar ramus is well shown 

 in a specimen from Overstrand (Savin Coll. M G079) figured 

 by E. T. Newton {op. cit. pi. i. fig. 3). In younger indi- 

 viduals, particularly' in the smaller, probably female, jaws, 

 the lower border of the ramus is nearly straiglit. 



Another bear with which Ursus savini must be compared is 

 U. deningeri from the older Pleistocene sands ol Mosbach and 

 Mauer. This species has been described in great detail by 

 V. Reichenau [Abhandl. Geol. Landesanstalt Darmstadt, 

 ]kl. iv. (1901-8) p. 208], and its relationships to other species, 

 especially to the Forest-bed bears and U. sptlccus, have been 

 discussed by Freudenberg [Paljeon. Abiiandl. Bd. xvi, (1913- 

 14) p. 582]. The latter authoi-, though at first inclined (o 

 regard U. deningeri as identical wiih the common Forest-bed 

 species {U. savini), later in his paper states tliat it is really 

 different in several respect.-=. I'he chief diflerences are : — 

 (1) in U. deningeri pni^ is always a nariow cone without 

 the inner tubercle, which is often more or less developed in 

 U. savini ; "(2) in U. savini the third lower molar, though 

 similar in geneial outline, is broader in proportion to its 

 length, a peculiarity still more marked in 7»2~/'"'4« I'l one 

 specimen of Ursus deningeri pni^ is present. 



From his table of average measurements of the teeth in 

 U. savini, U. deningeri, and U. spehvus, Freudenberg believes 

 that U. savini is intermediate between U. spelccns and U. de- 

 ningeri] but the more extensive series of measurements now 

 available shows that, so far as the dimensions of the teeth go, 

 U. deningeri and U. savini are very similar. Nevertheless, 

 the dift'erencos in structure already referred to and the differ- 

 ence of date seem to justify the separation of the two fol'ms 



