E. LÖNNBERG, MAMMALOGY OF ECUADOR. 95 



America, or some land territory in connection with this con- 

 tinent. The known distribution of these animals appears at 

 least to be most easily explained by such a theory and there 

 are also other analogous facts speaking for the same as for 

 instance most strikingly is the case with the distribution of 

 the tree frogs, Hylidae.^ This is practieally identical, as far 

 as is known, with that of the Marsupials thus with a great 

 number of forms in the Australian and Neotropical regions 

 and extending from the latter to North America. From this 

 continent they may by way of the former northern land 

 connection have wandered through Asia to Europé (ef. extinct 

 Didelphyidae), But the Oriental and Ethiopian regions have 

 not been accessible for either group. 



When the primitive Marsupials arrived to Austrah'a the 

 natural conditions of that continent were such, that the 

 majority of the invaders were induced or obliged to assume 

 arboreal habits. This was the case, as far as can be con- 

 cluded now, with all the Caenolestes-like ones. These had 

 then not yet been specialised and their hallux was not re- 

 duced. It therefore easily could be adapted for climbing 

 and become opposable. At the same time the syndactylism 

 originated in correspondence to this. These now arboreal 

 descendants of the primitive Caenolestes-like Marsupials 

 adapted themselves gradually to different kinds of diet etc. 

 and were subjected to changes with regard to their alimen- 

 tary canal etc. in consequence of this. Their dentition altered 

 also in some degree more or less, the canines and posterior 

 incisors became reduced, while in some cases the median 

 incisors as well were subjected to specialisation for gnawing 

 or other purposes. In Kangaroos and some Phalangerids 

 the changes in shape of the incisors especially the lower ones 

 have not gone far from the original one of the Caenolestes- 

 like ancestors, as already has been pointed out, and it does 

 not appear difficult to derive the various types of dentition 

 found in the recent Diprotodonts from such a one as is ex- 

 hibited by Caenolestes and its close allies. It appears also to 

 be ashorter way of natural development between the denti- 

 tion of Caenolestidae and the Diprotodonts than between the 

 former and the Polyprotodonts. 



1 Ortmann (1902) gives also proofs from the distribution of fishes, 

 moUuscs and decapod crustacea. Rep. Princeton Univ. Exp. Fatagonia. 



