4. In subspecies of any one full species the incisive foramen is larger in animals which inhabit arid areas than in those which inhabit humid areas. Possibly increased area of moist mucosa associated with Jacobson's organ is required in arid areas for maintenance of the necessary keenness of smell. 5. The number of counterfolds in the molariform teeth vary in clinal fashion. Their variation is in response to humidity. Increasing humidity is correlated with increasing number of folds, and decreasing humidity is correlated with decreasing number of folds. 6. Clinal variation correlated with increasing humidity is to be seen also in longer tail and darker color of pelage. 7. The primitive _Proechimys_ probably was large, with a short tail, narrow aristiforms, strongly built skull, and five counterfolds in each molariform tooth. 8. Geographic isolation appears to have been a factor in the establishment of the two subgenera; the arid belt along the São Francisco River and northward to Ceará appears to be uninhabited by _Proechimys_ and constitutes a barrier separating the two subgenera, _Proechimys_ and _Trinomys_. 9. This arid belt probably developed relatively early, since in deposits of late Pleistocene age, remains of the subgenus _Trinomys_ have been found in the area where the subgenus still occurs. 10. The most primitive types occur at the periphery of the range of the genus. 11. Populations from small islands tend to be more primitive than populations on the mainland. Insular populations develop a homozygous condition with resultant disappearance of secondary biotypes. 12. Insular animals ordinarily are larger than their mainland counterpart. TABLE OF MEASUREMENTS TABLE 1.--Measurements (in millimeters) of adults of _Proechimys_ Key: A Length of head and body B Length of tail C Length of hind-foot D Length of ear from notch E Greatest length of skull F Condylo-incisive length G Zygomatic breadth H Length of nasals I Interorbital constriction J Palatilar length K Crown length of cheekteeth