64 ODOB^NUS ROSMARUS ATLANTIC WALRUS. 



nearly of tlie same size, and equally closely packed, and are 

 connected Avith stellate lacunie in some unmbers near the peri- 

 phery of the tooth. This strnctnre, which is not pecnhar to the 

 Walrus, is, nevertheless, a test of aftinity, inasmuch as the form 

 of the lacunai varies in different animals. Tliey are not met 

 with in the tusks of the Proboscidea or the Ilii)popotamus, but 

 occur in the curious incisors of the Dugong. The 'dentinal 



cells ' of the Crag- tusks also resemble those of the Walrus 



In structure the cement exactly resembles that of the Walrus, 

 displaying vascular canals, bone-lacume, and canaliculi, of the 

 same form and disposition ; but the proportion Avhich it bears 

 to the thickness of the other tooth-tissues appears to be larger 

 in the Walrus than in the fossil. 



''From the foregoing remarks it will be apparent tliat we have 

 in these fossil tusks characters which ally them most closely to the 

 large canines of the genus TricUecusy After enumerating the 

 points of form and structure which distinguish these tusks from 

 those of other animals, and those which assimilate them to those 

 of the Walrus, he thus generalizes the results of his investiga- 

 tions: "Lastly, they resemble the large canine tusks of the 

 living Tricliecus in their curvature, varying lateral compression, 

 large surface-furrows, short and wide pulp-cavity, globular 

 ' osseo-dentine', and every detail of minute structure. They 

 differ from them in theii- greater curvature at the point of the 

 tusk, their greater lateral comi)ression, and minor development 

 of cement. 



" I accordingly i^roi^ose to establish the genus Trichecodon to 

 receive the animal thus indicated. The justification of a gen- 

 eric separation must be sought in the fact of the great antiquity 

 of the Eed Crag, and the consequent i^robability of the associ- 

 ation of other and more distinctive attributes with those of the 

 tusks." 



As regards its geological position and associations, Mr. Lau 

 kester adds: " It appears that the Tricliecodon Sui'leyi, lilvc the 



Cetacean remains of the Crag and large Sharks' teeth, 



is a derived fossil in the Eed Crag, belonging properly to the 

 Middle Crag, which is not now observable in this country [Eng- 

 land], but is well developed at and near Antwerp."* 



It thus ajipears that Mr. Lankester was as much, or more, in- 

 fluenced in his generic differentiation of these fossils from their 



* Quarterly Jonrual of the Geological Society of Londoji, xxi, 18C5, pp 

 2r26-231. 



