1868.] FLOWER TIIVLACOLEO. 311 



sor close to the middle line, directed obliquely upwards and for- 

 wards. In the number and arrangement of these teeth, therefore, 

 Thylaccleo corresponds exactly with the modern families Macropo- 

 didce and Phalangistidce, and differs completely from the carnivorous 

 marsupials. The same may be said of the canines, which are entirely 

 absent in the lower jaw, and represented above by a very small, 

 simple, conical tooth (fig. 1, c), exactly as in the Phalangistidce and 

 the greater number of Macroprodidce, especially Hypsiprymnus (fig. 

 2, c). Between the canine in the upper jaw, and the great trenchant 

 premolar, there are moreover two small simple teeth, with rounded 

 crowns, which must be regarded as anterior premolars. In the pre- 

 sence of these teeth Thylacoleo diff'ers from ffypsiprymnus and all 

 the Macropodidw; but in Phalangista one or more of such small teeth 

 in front of the large posterior premolar (undoubtedly homologous 

 with the great premolar of Hypsiprymnus and therefore probably 

 with that of Thylacoleo) are the rule. 



One of the most remarkable features in the dentition of this ani- 

 mal is the reduction in number and size of the true molars, of which 

 but one is present in the upper, and two in the lower jaw. This 

 reduction is evidently in relation with the excessive development of 

 the great trenchant premolar. It is interesting to observe a ten- 

 dency to the same occurrence in the recent allied forms. While in 

 the true Kangaroos, where the premolar is relatively small, the true 

 molars increase in size from the first to the fourth, in all the Rat- 

 kangaroos, on the contrary, they decrease, and the last is always 

 notably smaller than the others. In many species this is quite 

 rudimentary ; and it is occasionally altogether absent on one or both 

 sides, reducing the number of true molars to three, as in specimens 

 of H. Ogilhyi in the British Museum. The form of the molars dif- 

 fers greatly in different members of the Macropodidce ; but these 

 teeth in Thylacoleo, in their rudimentary and stunted condition, offer 

 no characters by which they can be compared very closely with any 

 of the known forms. 



There is nothing, in the structure of such portions of the cranium 

 and mandible of Thylacoleo as are known, which belies the conclu- 

 sions arrived at as to its affinities from the dentition. It differs 

 much from any of the now existing Macropodidce in the small rela- 

 tive size of the brain-cavity, and in the great development of the tem- 

 poral ridges, which meet in the middle line at the top of the skull, 

 forming a sagittal crest. This is probably only a difference of the 

 kind always observable in comparing large with small species of a 

 natural group, and not an indication of its affinity with the smaller- 

 brained carnivorous Dasyures and Opossums. In the phytophagous 

 Phalangers, moreover, the sagittal crest is as strongly developed, 

 and the brain-case almost as small as in Thylacoleo, 



In the conspicuous development of the postorbital processes it 

 differs from nearly all the members of the two families to which 

 otherwise it appears most closely related ; but some Hypsiprymni, 

 as H. murinus, have such processes well marked ; and their unim- 

 portance as a family character is shown in the Wombats, where they 



VOL. XXIV. PAKT I. Z 



