metastylid; greater development of hypoconulid ridge, … absence of anteroconid…."

Eozapus is more closely related to Pliozapus than to either Zapus or Napaeozapus (Wilson, 1936:32) but all four genera are in the subfamily Zapodinae. Stehlin and Schaub (op. cit.:158 and 311) relate Eozapus to the subfamily Sicistinae on the basis of similarity in the occlusal pattern of the cheek-teeth of Eozapus and various sicistines. Stehlin and Schaub do not consider other structures such as the elongate hind limbs, the shape of malleus and incus, and the shape of the baculum, in which there is close resemblance to the Zapodinae. It is these structural similarities as well as those, pointed out by Wilson (loc. cit.), in dentition that leads me to place Eozapus in the subfamily Zapodinae. The early Pleistocene Zapus rinkeri Hibbard shows that the Zapus stage of development had already been achieved perhaps as early as the late Pliocene. Hibbard (1951:352) thought that Zapus rinkeri was not structurally intermediate between Pliozapus and any Recent species of Zapus; although the teeth of Z. rinkeri have the broader, shallower, re-entrant folds of Pliozapus, these teeth are higher crowned and have an occlusal pattern resembling that of the Recent species of Zapus. The middle Pleistocene species, Zapus burti Hibbard, progressed essentially to the structural level of the Recent Zapus hudsonius, but the molars were more brachydont, broader crowned, and their enamel folds less crowded. Pleistocene material of pre-Wisconsin age obtained from cave deposits in Pennsylvania and Maryland is most nearly like Zapus hudsonius. One such cave deposit in Maryland contained an example of the Recent genus Napaeozapus, indicating that its history dates from at least middle Pleistocene time.

The Asiatic Recent Genus, Eozapus, has not progressed much beyond the Pliocene stage in zapidine evolution if Pliozapus be taken as a standard; the North American Recent Genus Zapus essentially achieved its present form by early Pleistocene times, and the Recent Genus Napaeozapus achieved its more progressive structure by middle Pleistocene times.

Perhaps Pliozapus and Eozapus represent one phyletic line and Zapus and Napaeozapus a second line, both of which lines evolved from a pre-zapidine stock in the Miocene. As mentioned earlier, Wilson (1936) thinks that Pliozapus is not directly ancestral to Eozapus. Possibly these two genera diverged at an early date; nevertheless, they are closely related primitive forms.

Zapus and Napaeozapus closely resemble each other and both are structurally advanced; Napaeozapus seems to have differentiated at a more rapid rate.

According to Simpson (1947), the occurrence of the same group of mammals on two different land masses is to be taken as prima facie evidence that migration has occurred. Keeping in mind then the present geographic distribution, unspecialized condition of the dentition of Eozapus, and its resemblance to the extinct Pliozapus known from North America but not from Asia, it may be that Eozapus descended from primitive stock of a North American jumping mouse that was forced to the periphery (across the Asiatic North American land bridge) by the more specialized zapidine stock.

Subsequently or perhaps during the migration of the pre-Eozapus stock the zapidine stock may have dispersed transcontinentally, occupying most of northern North America. The unprogressive Macrognathomys and Pliozapus line which remained in North America may have become extinct. Any such period of dispersal and climatic equilibrium ended when glaciers came to cover most of the northern part of the continent and the mammals living there were forced southward by the ice or remained in ice-free refugia within the glaciated area. Later, with melting and retreat of the ice, the jumping mice could have again spread enough to occupy the northern part of the continent. Such glaciation isolated segments of the population and aided their evolution into distinct species.

If it be assumed, as Matthew (1915) did and as Hooper (1952:200) later on the generic level did, that the region of origin and center of dispersal for a given group of animals is characterized by the presence of the most progressive forms, then southeastern Canada and the northeastern United States make up the area of origin and center of dispersal in relatively late time of the subfamily Zapodinae. This area is inhabited by Zapus hudsonius and Napaeozapus, the most progressive members of the subfamily.

As I visualize it, the evolution of the Zapodinae occurred in two stages: the first stage involved the movement of the primitive pre-Eozapus stock to Asia and the second stage involved the dispersal, isolation, and specialization in North America of the more progressive basic zapidine stock into the present genera Zapus and Napaeozapus.

The genus Zapus is one of three living genera in the subfamily Zapodinae. These genera Zapus and Napaeozapus from North America and Eozapus from China have been variously considered as subgenera of the genus Zapus (Preble, 1899) or as three separate genera (Ellerman, 1940).

Figs. 4-15. Three views of the skull and a lateral view of the left lower jaw of each of the Recent genera of the subfamily Zapodinae. × 1.5.

Figs. 4-7. Eozapus s. vicinus, adult, male, No. 240762 USNM, Lanchow, Kansu, China.

Figs. 8-11. Zapus h. pallidus, adult, male, No. 240762 KU, 5¹/₂ mi. N, 1³/₄ mi. E Lawrence, Douglas County, Kansas.

Figs. 12-15. Napaeozapus i. insignis, adult, male, No. 41109 KU, Shutsburg Rd., at Roaring Creek, 600 ft., Franklin County, Massachusetts.

Figs. 16-21. Occlusal views of upper and lower right cheek-teeth, of the three Recent genera of the subfamily Zapodinae. × 12¹/₂.

Figs. 16 and 19. Eozapus s. vicinus, adult (age group 3), male, No. 240762 USNM, Lanchow, Kansu, China.

Figs. 17 and 20. Zapus h. alascensis, adult (age group 2), female, No. 29073 KU, E side Chilkat River, 9 mi. W and 4 mi. N Haines, Alaska.

Figs. 18 and 21. Napaeozapus i. insignis, adult (age group 3), male, No. 41109 KU, Shutsburg Rd., at Roaring Creek, 600 ft., Franklin County, Massachusetts.

Note especially the variation in complexity of occlusal pattern, width of re-entrant folds, and degree of tubercularity.

The remarkable similarity of the body form, post-cranial skeleton, mandibular rami, and general structure of the cranium of Zapus, Napaeozapus, and Eozapus indicate their relationship (see figs. 4-15); however, dissimilarity between the groups in the dentition (tooth number and occlusal