154 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. 



E THE POSSIBLE ANCESTRY OF PHENACOMYS 

 LONGICAUDUS TRUE 



The known characters and distribution of the members of 

 the genus would seem to support the theory that the subgenus 

 Arborimus is derived from the subgenus Phenacomys, or that 

 both subgenera are derived from a common ancestral stock 

 not very different from either; and that the isolation of por- 

 tions of the parent stock was an important factor in their 

 differentiation. 



It is furthermore possible if not probable that both Phena- 

 comys longicaiidus and Phenacomys alhipes are descended 

 from the same species, doubtless a member of the subgenus 

 Phenacomys. Compared with albipes, longicaiidus would 

 seem to be a little more specialized. The two species may 

 exemplify successive migrations of similar individuals from 

 a common center, longicaudus being of the first wave, alhipes 

 of the second. Longicaiidus would in this way have had 

 time to become more specialized than alhipes. 



It is almost impossible to avoid the inference that the long 

 tail in Phenacomys longicaudus and the arboreal habit are in 

 some way associated. 



Allen has recently shown (Bull. Amer. Mus. Nat. Hist., 

 SIf, 1915, p. 166) that the tail in different groups of tree 

 squirrels is developed in proportion to their exclusiveness as 

 tree dwellers, the ratio of tail length to total length varying in 

 the different groups from about 40 to 52 per cent. Ground- 

 living species of Phenacomys (except alhipes) have ratios of 

 25 per cent, alhipes (ground living) has an average ratio of 

 38 per cent, and longicaudus (tree dwelling) of approximately 

 40 per cent. 



As implied above, it seems certain, from the close general 

 similarity between the tree mouse and its ground living rela- 

 tives, that the long-tailed arboreal species is derived from 

 some short-tailed terrestrial form. 



The possible connection between the long tail of P. longi- 

 caudus and its arboreal habitat suggests a train of puzzling 

 questions. 



Did P. longicaudus acquire its long tail and then take to 

 the trees? Or did it take to the trees and then gradually 



