175 



Of all known Marsupialia, recent or fossil, Plagiaulax, so far as its dentition is accu 

 rately determined, is most closely allied to Thylacoleo. In the lower jaw the true 

 molars are similarly reduced to two of small size and tubercular form. One cannot 

 suppose that they were opposed by more tuberculate molars above ; the analogy of Thy- 

 lacoleo (Plate X.) would point to fewer. A character, indeed, of the first of the inferior 

 molars of Plagiaulax, overlooked by Dr. Falconer, would indicate that it worked secto- 

 rial-wise, like the fore part of the anterior lower molar of Thylacoleo, upon the back 

 part of the blade of a large upper carnassial ; I allude to the smooth vertical wall-like 

 surface of the inner side of the outer half of the crown of m 1, in Plagiaulax* . 



Taking the same range of the molary series for comparison as in Thylacoleo and ex- 

 isting Diprotodonts, in reference to the character of size of the last trenchant premolar, 

 the tooth equals in antero-posterior extent one-half of that series in Plagiaulax. But 

 in this more ancient Diprotodont the premolars anterior to the last large one have not 

 undergone the extreme degradation which they show in the tertiary fossil (Thylacoleo) 

 and in some existing Diprotodonts of Australia. They are modified, in Plagiaulax, for 

 sectorial function, and are so combined with the last and largest sectorial as to work with 

 it as one instrument, obliquely ridged and notched at the convex cutting-margin, like a 

 section of a circular saw. I have already f pointed out the advantage of this modifi- 

 cation of carnassial in dividing the integuments and other tissues, tougher and drier 

 than those in Mammals, of the lacertian members of the cold-blooded class which so 

 abounded with the small carnivorous Marsupials in the same Mesozoic period and place. 



If it be admitted that, so far as the lower jaw and its dentition show, Plagiaulax (figs. 10 

 & 15), with its two or three reduced anterior premolars, its suddenly enlarged hind premo- 

 lar, its disproportionately small and few (two) tubercular molars, and its large laniariform 

 upcurved incisor, comes nearest to Thylacoleo (figs. 8 & 14), it is plain, from the antece- 

 dent comparisons with existing Diprotodonts, that there are no grounds for inferring the 

 Macropoda to have been derived from the Paucidentata, or these from Rat-kangaroos. 



What we do learn from consideration of the fossils in question is, the fact of an addi- 

 tional and most interesting modification of the Diprotodont section of the Marsupial 

 order or subclass, unknown before the discovery of these fossils. We further learn that 

 such modification, which, from the extreme reduction of the true molar series, I have 

 been led to take as the character of a " paucidentate" family of Marsupials, was 

 already established at the Purbeck period; yet with modifications interestingly exem- 

 plifying the tendency to the "more generalized condition of structure" as compared 

 with the newer tertiary extinct form. 



But I am here met by another objection. Dr. Falconer, attacking the principle 

 of the tendency to transition in organisms from generalized to specialized structures 

 as they approach in geological position the present time, writes : — " Among other 

 arguments, they insist that the earliest Eocene Mammalia, both carnivorous and 



* ' Prefatory Notice,' &c., pp. 7G, 81 ; Plate IV. figs. 9, 12. 

 t ' Prefatory Notice,' &c, p. 97, Plate IV. figs. 9, 12. 



9* 



