PURBECK FORMATIONS. 69 



bered alveolus of the last premolar indicates the usual progressive increase of size. Part 

 of a similar socket, with the hind root and portion of the crown of the first molar {m i), 

 is definable ; it is followed by the first and second cones of m 2. The basal cingulum is 

 shown in both teeth. 



In the fore part of the left ramus, showing the outer surface (PI. IV, fig. 1 c), the 

 characters of that side of the crown of the canine are seen ; it is more convex than the inner 

 side, and has not the longitudinal basal depression. The anterior outlet of the dental canal 

 is beneath the alveolus of the first premolar. Part of the crown of the second {p 2) is pre- 

 served. The third {p ) shows the height of the main cone, which wants the apex and 

 the right ramus. The last premolar and the first true molar (?« 1) are represented by the 

 basal parts of the crown. The interruption or partial development of the cingulum is 

 here noticeable. The mandibular rami in the present specimen show the convex lower 

 outhne and the gradual curve upward to the incisive alveoli, which appears to be charac- 

 teristic of Triconodonferox, or at least to diflferentiate it from Triconodon mordax. 



Species 3. — Triconodon occisor. Plate IV, figs. 2, 2 a, 2 b. 



I next enter upon the description of the partially dislocated rami of the same man- 

 dible (PL IV, fig. 2, nat. size ; a and b, teeth magn. 2 diam.), showing the formal characters 

 of the mandible in Triconodon mordax, but belonging to an animal too superior in size 

 to be referable to that species ; from which also the present more decidedly and instruc- 

 tively differs in a minor relative size of the first molar (in 1) as compared with the other 

 two molars. 



At first sight the left ramus (fig. 2, /) seems identical, answering as it does, in both 

 size and shape, with the type of the genus (fig. 7, PI. III). But if the first two molars 

 be compared in the two specimens, those of the present show a marked increase of antero- 

 posterior extent, notwithstanding the smaller proportion of m 1 to m 2. This differential 

 character comes out well in the figures of each mandible which are magnified in the same 

 degree in PI. IV, fig. 2 b, and PI. Ill, fig. 7 a respectively. Furthermore, although some 

 signs of immaturity are recognisable in the type of Triconodon mordax, the mandible, here 

 compared, of Triconodon occisor is of a still younger individual ; only half of the last molar 

 (/« 3) has advanced into view beyond the base of the coronoid, and not more than two 

 thirds of the crown of the canine (c) and of that of the last premolar [p 4) have risen 

 above the socket. 



The fully grown animal of the present species would equal in size that of Triconodon 

 ferox, but the shape of the mandible and the proportions of the molars forbid a reference 

 of the present specimen to that species, 



I cannot, with a recognition of the above defined characters, withdraw from the 



