Individuals from the southernmost geographic races (_Z. h. americanus_ and _Z. h. pallidus_) are the smallest for the species and those from the northernmost subspecies (_Z. h. alascensis_) are the largest. One subspecies, _Z. h. campestris_, from the central part of the range of the species, however, seems to be out of the cline. This form inhabits the eastern foothills of the Rocky Mountains and is a robust animal approaching _Z. princeps_ in size. Seemingly there is no clinal variation in the several qualitative features of the cranium, for instance in the shape of the auditory bullae, shape of the incisive foramina, and shape of the postpalatal notch. On the other hand, the dimensions of the entire skull show that the larger crania are of the northernmost subspecies and the smaller of the southernmost subspecies. NATURAL HISTORY _Habitat._--_Zapus hudsonius_ occurs in low undergrowth usually of grasses or forbs or both, in open coniferous forests, deciduous hardwood groves, or in stands of tall shrubs and low trees, but most frequently in open, moist areas. Quimby (1951:75) notes that jumping mice were more common in the moist lowlands than in the drier uplands. More were in the open type lowlands than in the forested type, and these mice favored habitats normally bordered by small streams affording moist to semi-aquatic living conditions. The reports of Goodwin (1924:255), Christian (1936:416), G. S. Miller (1899:329), Cory (1912:249), Lyon (1936:277), Stoner (1918:123), and others, although concerning widely different parts of North America, indicate that _Z. hudsonius_ selects habitats in vegetation of like form, even though different assemblages of plant species may be involved. An average of 11.91 mice per acre was recorded by Quimby (1951:91) from a study plot at Itasca Park, Clearwater County, Minnesota. He gives the monthly population densities per acre for _Z. hudsonius_ at Centerville, Anoka County, Minnesota, as follows: June 2.78, July 3.57, August 3.10, and September 1.81. Blair's (1940:248) data on bi-monthly population density per acre for _Z. hudsonius_ on the Edwin S. George Reserve, Livingston County, Michigan, are remarkably similar, when adjusted on a monthly basis, to those obtained by Quimby (_loc. cit._). Blair's (_loc. cit._) monthly population densities per acre are as follows: June 3.90, July 3.85, August 3.10, and September 2.00. Townsend (1935:90) estimated population densities per acre for _Z. hudsonius_ in central New York state, at 11 to 72 individuals. As Quimby (1951:92) points out, Townsend's figures are probably too high, as commonly is the case when the moving quadrat technique is used because animals from neighboring areas enter the trapped area to take over the niches made available by their predecessors' removal.