Some morphological features of the same kind and degree that differentiate subspecies in one place may not vary geographically in another place. Furthermore the geographic variation in one feature may be only partly correlated with the variation in another feature. The variation in _M. p. pullatus_ is an example: Specimens from near Pocatello, Idaho, are darker than topotypes of _modestus_ but specimens from Fremont County, Idaho, are indistinguishable from topotypes of _modestus_ (Davis, 1939:315). I have examined a number of mice from the Bitterroot Valley in western Montana and the color value for 12 adults is 2.7. They are slightly but not significantly paler than topotypes of _modestus_. This is a result of the cline mentioned above and does not indicate relationship with _modestus_. Some average measurements of 10 skulls from this series are as follows: condylobasilar length, 24.9; zygomatic breadth, 14.4; interorbital breadth, 3.4; lambdoidal breadth, 11.5; prelambdoidal breadth, 9.1; alveolar length of upper molar teeth, 6.5; and depth of braincase, 7.6. Average external measurements of 9 specimens are as follows: total length, 157; length of tail, 36; length of hind feet, 19.4. Mice from the Bitterroot Valley were compared with topotypes of _modestus_ by the "method of pairs," and _modestus_ had a larger foramen magnum in 6 of 6 pairs (Confidence Limit .97) and larger first upper molar teeth in 5 of 6 pairs (C. L. .75). A comparison of topotypes of _pullatus_ with _modestus_ shows a similar difference in the teeth, _modestus_ being larger, but in the size of the foramen magnum there is no difference. A comparison of the measurements of _pullatus_ (Bitterroot Valley), _pullatus_ (near topotypes), and _modestus_ (topotypes) shows that the two series of topotypes differ significantly in condylobasilar length (of borderline significance), zygomatic breadth and lambdoidal breadth (both of which vary greatly with age), and length of molar series; the specimens from the Bitterroot Valley agree with _pullatus_ rather than _modestus_ in all of these characters. The specimens from the Bitterroot Valley are smaller than _pullatus_ (topotypes) in total length; they more closely resemble _modestus_ in length of tail, and the hind foot is shorter than in either _modestus_ or _pullatus_ (which do not differ significantly). Specimens from western Montana resemble _modestus_ in certain respects but in most respects resemble topotypes of _pullatus_ and are referred to _pullatus_. Specimens from Coeur d'Alene, Idaho, and from Blackfoot, Montana (marginal records of _modestus_, Hall and Cockrum, 1953:410) may be referred to _drummondi_. Marginal records of _pullatus_ in Montana to my knowledge are: Florence, Ravalli Co.; Highwood Mtns., Chouteau Co.; 7 mi. NE Hilger, Fergus Co.; 10 mi. NW Park City, Stillwater Co. (all represented by specimens in the USBS). The line separating _pullatus_ and _drummondi_ is tentatively drawn as shown in Figure 1.