64 AFFINITIES AND HABITS OF THYLACOLKO, 



Before, however, discussing the relations and hal>its of Thyla- 

 coleo it may be well to make a short digression to consider the 

 origin and probable phylogenetic history of the enlarged premolar 

 as found in Biirramys and carried on into the Macropodidce. 

 Though grooved premolars occur in the Plar/iaulacidce it will be 

 unnecessary at present to discuss that group, as it is certainly not 

 nearly related to the existing Diprotodont Marsupials, and any 

 similar development can only have been due to a parallel develop- 

 ment. 



Let us imagine a small Dromicia-like Phalanger which, from 

 necessity, had to live less exclusively on succulent leaves and 

 other soft substances and had to make up the deficiency with 

 grass. Eucalyptus and other succulent leaves, fruits, and even 

 insects, can be broken and crushed, but grass requires to be cut, 

 and the comparatively feeble and pointed incisors would unaided 

 be unable satisfactorily to finely cut the tougher fibres of the new 

 diet. The sharp-edged premolars would be called in to assist in 

 the dividing process and the increased work given to them would 

 lead to their greater development. It is further not difficult to 

 see the advantage that would result from a serrated edge being 

 acquired, though the exact details by which the serrations would 

 arise could not well, with the meagreness of the data, be more 

 than roughly guessed at. Such a development and specialisation 

 of the posterior premolar would give rise to a form closely 

 resembling Burramys. In the Macropod line of descendants the 

 arboreal life is more or less completely abandoned, and the whole 

 organisation has been modified to suit a ground life and a diet of 

 grass and other fibrous plants and roots. The lower limbs have 

 become lengthened and strengthened to enable the animals to 

 escape their enemies by flight; and the hallux or '• thumb " being 

 a useless encumbrance, no longer requii'ed for grasping the boughs, 

 has been early lost. In only one species of the Macropodidce (Hyp- 

 siprymnodon moscliatus) does the hallux still remain, apparently 

 an ancestral type and one which forms an almost perfect link 

 between the Burramys-like species and the Rat-kangaroo. The 

 few forms which hav^; returned to an arboreal life, such as 



