Beyond, however, admitting its affinities with the diprotodonts, Sir Richard Owen does not appear in this later paper to have regarded _Thylacoleo_ as a near relative of any of the existing forms; but in the article on “Paleontology” in the eighth edition of the _Encyclopædia Britannica_, published in 1859, he seems to have considered it allied to _Plagiaulax_ of the Purbeck strata of Dorsetshire, which had been shown by Dr. Hugh Falconer to be probably of herbivorous habits. Sir William Flower, in the aforesaid paper in the Geological Society’s _Quarterly Journal_ for 1868, while agreeing with Owen that _Thylacoleo_ was related to the diprotodont rather than to the polyprotodont carnivorous marsupials, differed from the conclusion that it was a carnivore. While the large cutting premolar teeth were considered by Owen to resemble the carnassial teeth of a lion, Flower was struck by their similarity to the corresponding teeth of the rat-kangaroos and the phalangers. After discussing the other teeth, he concluded that “in the number and arrangement of these teeth ... _Thylacoleo_ corresponds exactly with the modern families _Macropodidæ_ and _Phalangistidæ_, and differs completely from the carnivorous marsupials.” After alluding to the small size of the brain-cavity and the large space for the attachment of the powerful muscles which worked the lower jaw, and suggesting that these features may be only to be expected in a large form as compared with the smaller members of the same group, Flower concluded that the habits of all species with the same general type of dentition must necessarily be similar. And, on these premisses, it was urged that _Thylacoleo_ must in all probability have been a vegetable-feeder. The large premolar may seemingly have been “as well adapted for chopping up succulent roots and vegetables, as for dividing the nutritive fibres of animal prey.” It is further suggested that the nutriment of _Thylacoleo_ “may have been some kind of root or bulb; it may have been fruit; it may have been flesh.” While in conclusion it is argued that the organisation of the animal did not countenance the idea of its preying on the large contemporary marsupials. Omitting reference to Owen’s reply to this reversal of his conclusions, and also to certain comments and additions to the arguments by other writers, we may pass on to a paper by Dr. R. Broom, published in the _Proceedings_ of the Linnean Society of New South Wales for April 1898, and entitled “On the Affinities and Habits of _Thylacoleo_.” In this the author admits that the animal in question, as suggested by Owen in his second paper, and more fully determined by Flower, was undoubtedly a diprotodont, and that it was nearly allied to the modern phalangers. With the latter it is indeed closely connected by the recently discovered extinct _Burramys_, which differs from the existing members of that group by the large size of the secant premolar.