BY R. BROOM. 65 



Dendrolagus, cannot regain the lost " thumb," and are at most 

 slightly modified Wallabies. All the known Rat-kangaroos — 

 which are, there is little doubt, the more primitive members of 

 the group* — are of small size, and their dentition is invariably- 

 suited to a fibrous vegetable diet, principally of grass. Though 

 in Burramys there are but three molars above, and the fourth 

 below rudimentary, there can be no doubt that in closely related 

 forms the normal number was present, and the changes which 

 would be required to give rise to such a dentition as is met with 

 in the Rat-kangaroos from a Burramys-like ancestor are very 

 slight. The increased grinding work entailed by the tougher 

 vegetable diet would lead to the retention and greater development 

 of the four molars; and while the large cutting premolars would 

 l)e also retained and modified slightly to suit the special require- 

 ments of the various species, the rudimentary premolars being 

 functionless would become lost. In the further development of 

 the Macropodidca which gave rise to the Kangaroos and Wallabies 

 a most interesting change has taken place. Owing to the increase 

 in size of the forms and also to the loose mode of attachment of 

 the jaws to each other, the cutting functions can all be performed 

 by the incisors, and the large premolars which had been functional 

 in the lower forms became much I'educed in size, and in the larger 

 species are of so little importance that they are lost shortly after 

 the animal becomes adult without apparently causing any incon- 

 venience. It will thus be seen that there are fairly good reasons 

 for believing that tlie unusually large development of the last 

 premolar has been brought about in connection with the more 



* The position of Trirlis, De Vis, is uncertain. Lydekker says of it 

 {Palaeontology by Nieholsou & Lydekker, Vol. ii. p. 1286), "there is a 

 minute tooth behind the lower incisor corresponding to the tooth iu the 

 Phalanyeridce, conimonly reckoned as the representative of the canine 

 [2Qd incisor — Thomas]." If this observation be correct it is certainly an 

 1 iiteresting Phalangeroid character, but De Vis informs nie that the dentition 

 in the adult jaw is " I^; C" ; P*; M^' 2- 3. 4." it will thus for the present 

 be safer to omit consideration of this form. ^"^ ' iP ,i^ 



