EEPOET ON THE EXPECTATION OF BEIXITAM CAVE. 
541 
almost impossible that fig. 3 should also belong to the same ; whilst if we regard the 
peculiar form of fig. 4, it seems difficult to believe that it also does not belong to a third. 
I am, however, inclined to the opinion that it may belong to the upper jaw of the same 
species as fig. 3, and also that the canine teeth in the present collection indicate the 
existence it may be of three, and certainly of two, distinct species of TJrsus. In this 
view we might perhaps, though with doubt, refer fig. 1 to If. spelceus , and fig. 2 to If. 
prisons, whilst I have little doubt the two small teeth belong to If. circtos. This con- 
clusion is supported by the numerical data given in the Table. The mean dimen- 
sions of the canine teeth found at Brixham are 108x67. Nine teeth are decidedly 
above this mean, whose average dimensions are 124 X 72 ; whilst six are as much below it, 
Avhose size is about 90 X 60. The dimensions of the intermediate eight are very uni- 
formly 100x65. Among the teeth belonging to the first category there are four whose 
mean size is nearly 140x80, and four of 120x67. Now the former may fairly be 
referred to U. spelceus , whilst the smaller may represent either the female of that species 
or the large male of If. prisons. The teeth in the second category might in like manner 
be referred to If. prisons , whilst the smaller teeth in the third category may be assigned 
to If. arctos, in which species, from the measurements I have taken, the mean size of 
the canines may be given as 92 x 60. In If.ferox it is about 103 x 70, and in If. mari- 
timus the same. 
But it is in the molar teeth, and more particularly, perhaps, in those of the lower jaw, 
that we have to seek the most marked and distinctive dental characters. 
These depend upon (1) the absolute and relative dimensions, and (2) on the form of 
the various teeth. As regards the former point, the materials that I have employed for 
comparison are contained in the subjoined Table, in which all the measurements I have 
been able to make or obtain of the molar teeth in most of the known and admitted 
species of recent and fossil Bears, including those from the Brixham Cave, are given. 
Upon these figures I will first offer a few explanatory remarks, and then proceed briefly 
to discuss the principal morphological characters. 
In doing this, however, although most of the existing species are included in the 
Table, it will be needless for the present purpose to pay any particular attention to 
more than three or four, viz. : — 1. If. ferox, Richardson, 1825 [If. Jiorribilis, Ord, 1815) ; 
2. If. arctos, Linn A ; 3. If. maritimus , Desrn. ; 4. If. horriceus, Baird. 
The numerical data upon which I have gone in this inquiry are contained in the 
following Table f : — 
* Under which I include Myrmarctos Eversmannii, Gray, U. isabellinus, Horsf., U. syriacus, Hempr. & 
Ehr., and all the other varieties of the European and North Asiatic Brown Bear, which have sometimes 
received distinct specific names, believing, as I do, that, regarded as subjects of palceontology, it would be 
impossible to distinguish among them. 
t Eor the purpose of facilitating comparison the dimensions recorded in this Table arc graphically represented 
in the “ Odontograms ” on Plate XLVII. These have been constructed on the plan described in Proc. Eoy. Soc. 
1870, vol. xviii. p. 544. 
