Vol. V] TAYLOR— NEW SUBGENUS OF PHENACOMYS 155 



acquire its long tail? If it got its long tail while still living 

 on the ground, taking to the trees when the tail reached 

 approximately its present length, why has not P. alhipes, 

 which has a tail nearly as long, also adopted an arboreal hab- 

 itat? If Phenacomys longicaudiis took to tree life while still 

 short-tailed, acquiring its long appendage thereafter through 

 some form of environmental or other pressure associated with 

 arboreal life, how is the acquisition of a long tail by the 

 wholly ground-living P. alhipes to be explained? 



But in this connection it ought to be remembered that, in 

 view of the small number of specimens of the latter species 

 which have been taken, and of our ignorance concerning its 

 life history, we are hardly in a position to state positively just 

 where its habitat does lie. 



Perhaps the ancestor of both longicaudiis and alhipes was 

 long-tailed and is extinct and unknown. In this case possibly 

 longicaudus merely selected the arboreal habitat for which its 

 characters already fitted it. There remains the problem of 

 why the similar alhipes, which is to all appearances equally 

 well fitted for tree life, did not also become a tree mouse. 



If longicaudus and alhipes represent successive waves of 

 migration, perhaps longicaudus may be conceived to have 

 attained to the arboreal environment before the development 

 of alhipes. If this were so the prior occupancy of the tree 

 habitat by longicaudus would possibly be sufficient to account 

 for the terrestrial predilections of alhipes. 



It should here be noted that if the hypothesis is true, that 

 Phenacomys longicaudus is derived from some ground-living 

 microtine, we have presented in the phylogeny of the tree 

 mouse an unusual type of migration. The writer has already 

 emphasized (The status of the beavers of western North Amer- 

 ica, Univ. Calif. Publ. Zool., vol. 12, in press) that in general 

 each group of mammals occupies the same ecologic niche in dif- 

 ferent places rather than different ecologic niches in the same 

 place. The Microtina are characteristically terrestrial, with 

 some members adapted to a more or less aquatic, and others 

 to a more or less fossorial, mode of life. Apparently the stock 

 which we now know as Phenacomys longicaudus, in the 

 course of its phylogeny, has broken away from the time-hon- 

 ored group niche in which all other members of its subfamily 



