PURBECK FORMATIONS. 109 



long and large laniariform canines, longitudinally grooved/ sharp-pointed, with 'a 

 trenchant posterior margin.' 



In the Carnivora these characters distinguish the canines of Felines from those of 

 other families and genera of the order. The correspondence in the number and direction 

 (verticality) of the grooves on the crown of the sharp-pointed posteriorly trenchant 

 canines, as well as in the number of those teeth, in the genera of Qmdrmnana, cited 

 below, and in the genus Felis, is much closer than in the grooved trenchant premolars of 

 Plagiaulax and Hypsiprymnus . 



Now, suppose that our knowledge of the genus Felis was founded upon fossil remains 

 of the mandible and mandibular dentition, from Mesozoic deposits. A Palaeontologist 

 detecting the correspondence of the fossil canine in the vertical grooves and trenchant 

 hinder border, superadded to the common characters of a laniary tooth, with the canines 

 of the existing Monkeys and Baboons, might infer that a Mammal had existed at the 

 Mesozoic period with quadrumanous modification of limbs, scansorial habits and 

 frugivorous diet. 



Dr. Falconer has drawn the parallel inference from the correspondence, such as it is, 

 in the grooved and trenchant premolar teeth of Tlagiaulax with those of Hypsiprymnus, 

 namely, that there existed, at a Mesozoic period, a Marsupial with the saltatorial modifi- 

 cation of limbs, leaping mode of progression, and herbivorous diet. 



Mr. Boyd Dawkins, therefore, draws the same conclusion, though with a much lower 

 degree of likeness, in the assumed ridged and trenchant premolars, in regard to his 

 " Hyjmjyrynmojms Bhceiims of the Watchet Shore." ^ 



The alleged " ample demonstration by Dr. Falconer of the alliance of Plagiaulax to 

 the Kangaroo-rats" is of the same kind and degree as that by which a Fehne might be 

 demonstrated to be a Baboon-ally. 



The proof of the insufficiency of the ground afforded by the similarity of canine teeth 

 is of the same kind as that by which I believe myself to have demonstrated the insuffi- 

 ciency of the ground afforded by the similarity of premolar teeth held to be ' amply demon- 

 strative ' of Plagiaulax being a ' Kangaroo-rat ally.' 



The alleged correspondence of premolars is more than counterbalanced by the con- 

 comitant difference in the series of true molars presented by Plagiaulax and Hypsiprym- 

 nus. To the inference from the fossil jaw of Pelis with canines having not only the same 



' Owen's ' Odontography ;' canines of Hapale (p. 439) : canines of Cehm, — " those above being 

 marked by the deep anterior groove : there is also a second longitudinal groove on the inner side of the 

 crown near its posterior margin" (p. 440) : in Cynocephaliis, besides the grooves, the " trenchant posterior 

 margin " of the long, pointed canine, is specified (p. 442) : in Papio maimon " the anterior longitudinal 

 groove of the canine is very deep, the posterior margin very sharp" (p. 316). The same character of canines 

 is pointed out in Papio porcarms in the ' Catalogue of the Osteology in the Museum of the Royal College 

 of Surgeons,' vol. ii (1853), p. 734, No. 4723. See also 'Anatomy of Vertebrates,' vol. iii, p. 316. 



2 Quarterly Journal of the Geological Society of London,' vol. xx (1864), p. 412. 



