The baculum of _Zapus_ differs from that of _Eozapus_ as follows: base less expanded horizontally; shaft slenderer; distal end less spade-shaped except in _Z. trinotatus_. The tip is less expanded in _Z.

princeps_ and is still less so in _Z. hudsonius_. In _Napaeozapus_ the tip is lanceolate, the base is narrow, and in lateral view the shaft is slender and curved (see figs. 22-31).

-----------------------+-----------------------+----------------------- _Eozapus_

_Zapus_

_Napaeozapus_ -----------------------+-----------------------+----------------------- P4--Small

Smaller

Absent

M1--Four wide l.a.b.i.al

Four moderately

Three narrow l.a.b.i.al re-entrant folds

narrow l.a.b.i.al

re-entrant folds of equal length;

re-entrant folds

of unequal length, paracone and

of unequal length;

1st long, 2d and 3d metacone largest

1st and 3d longer

shorter; paracone cusps; anterior

than 2d, 4th

and metacone larger cingulum large.

shortest; paracone

than in _Zapus_ and

smaller than in

_Eozapus_; anterior

_Eozapus_; metacone

cingulum absent.

largest cusp;

anterior cingulum

small.

M2--Four wide l.a.b.i.al

Four moderately

Narrow l.a.b.i.al re-entrant folds;

narrow l.a.b.i.al

re-entrant folds, 2d short, others

re-entrant folds

variable in number, of equal length

of unequal length,

often as many as 6; but longer than

1st and 3d long,

anterior and 2d; anterior and

2d and 4th short;

posterior cingula posterior cingula

anterior and

small; occlusal large; occlusal

posterior cingula

pattern complex.

pattern simple.

moderately large;

occlusal pattern

moderately complex.

M3--Three wide l.a.b.i.al

Two moderately

Three narrow l.a.b.i.al re-entrant folds

narrow l.a.b.i.al

re-entrant folds of unequal length,

re-entrant folds

of unequal length, 1st short, 2d and

of equal length;

1st long, 2d and 3d long; anterior

anterior and

3d short; anterior and posterior

posterior cingula

and posterior cingula low,

moderately large;

cingula large; small; occlusal

occlusal pattern

occlusal pattern pattern simple.

moderately complex.

complex.

m1--Anterior oblique

No anterior

No anterior re-entrant fold

re-entrant fold; 4

re-entrant fold; separating equal

moderately narrow

narrow lingual sized protoconid

lingual re-entrant

re-entrant folds and metaconid

folds of equal

variable in cusps; 3 wide

length, 1st joining

number, often as lingual re-entrant

1st l.a.b.i.al

many as 4; folds of equal

re-entrant fold,

anteroconid well length;

4th joining 2d

developed, encloses anteroconid absent;

l.a.b.i.al re-entrant

1 or 2 small lakes; occlusal pattern

fold; anteroconid

occlusal pattern simple; mesoconid

well developed,

complex; mesoconid present.

encloses

absent.

small lake; occlusal

pattern moderately

complex;

mesoconid absent.

m2--Four wide lingual

Four moderately

Narrow lingual re-entrant folds

narrow lingual

re-entrant folds, of unequal length,

re-entrant folds,

variable in number, 1st short, other

1st and 2d long, 3d

may be as many as 5; 3 equal and long;

and 4th short, 1st

anteroconid large, anteroconid

joins 1st l.a.b.i.al

encloses moderately large;

re-entrant fold

complex folds from occlusal pattern

and 4th joins 2d

1st l.a.b.i.al re-entrant simple.

l.a.b.i.al re-entrant

fold; occlusal

fold; anteroconid

pattern complex.

large; occlusal

pattern moderately

complex.

m3--Three wide lingual

Three moderately

Narrow lingual re-entrant folds

narrow lingual

re-entrant of near equal

re-entrant folds

folds variable in length; antero-

of unequal length,

number, as many as 3; conid absent;

1st and 2d long, 3d

anteroconid present; occlusal pattern

short; anteroconid

occlusal pattern simple; 1 l.a.b.i.al

absent; occlusal

complex; 2 l.a.b.i.al re-entrant fold.

pattern moderately

re-entrant folds.

complex; 1 l.a.b.i.al

re-entrant fold.

-----------------------+-----------------------+------------------------

[Ill.u.s.tration: FIGS. 22-31. Dorsal and lateral views of the bacula of the Recent genera (and species of the genus _Zapus_) of the subfamily Zapodinae. 10.

FIGS. 22 and 27. _Eozapus setchua.n.u.s_ (after Vinogradov, 1925:585).

FIGS. 23 and 28. _Zapus t. trinotatus_, adult, No. 94596 MVZ, 1-1/4 mi. ENE Amboy, 350 ft., Clark County, Washington.

FIGS. 24 and 29. _Zapus p. princeps_, adult, No. 20870 KU, 3 mi. S Ward, Boulder County, Colorado.

FIGS. 25 and 30. _Zapus h. pallidus_, adult, No. 22954 KU, 4 mi. N, 1-3/4 mi. E Lawrence, Douglas County, Kansas.

FIGS. 26 and 31. _Napaeozapus i. insignis_, adult, No. 41110 KU, Shutsburg Rd., at Roaring Creek, 600 ft., Franklin County, Ma.s.sachusetts.]

EAR OSSICLES.--The auditory ossicles are of three types which differ only slightly. These ossicles possibly are more conservative than some other structures because the ossicles are not so much affected by the molding influence of the environment.

Instances of variation in the auditory region in mammals in general are small, even at the family level; therefore, these differences in the subfamily Zapodinae are offered as additional support for recognizing _Eozapus_, _Zapus_, and _Napaeozapus_ as distinct genera. The distinctive features are chiefly in the malleus and incus; the stapes, however, differs slightly and, therefore, it too is described (see figs.

32-34).

In _Eozapus_ the head of the malleus is narrow, oblong, and rounded dorsally and attaches to the body by a long, slender, abruptly recurved neck. The body is weakly pointed ventrally and rounded dorsally. A beaklike manubrium malleus composed of anterior projecting external and internal spines extends from the body to the tympanum. The incus has a dorsally rounded body with an anterior downward snoutlike projection with which the malleus articulates. The short limb of the incus is broad basally and narrows somewhat distally. The long limb is narrow and its articulating lenticular process is a flat circular structure. The limbs of the stapes are wide-spread and heavy. The neck is short and wide with a large circular articulating surface.

In _Zapus_ the head of the malleus is angular with an anterior projecting point and is flattened in dorsal aspect. The neck is slender, elongate, and gently curved away from the long limb of the incus. The body is pointed dorsally and rounded ventrally, the reverse of the condition in _Eozapus_. There is a beaklike manubrium malleus composed of internal and external anteriorly projecting spines extending from the body to the tympanum as in _Eozapus_. The incus has a rounded body with a long angular limb articulating via a small lenticular process with the stapes. The short limb is narrow but does not taper distally as in _Eozapus_. The limbs of the stapes are relatively narrow, weak, and gently curved. The neck is longer and more slender than that of _Eozapus_.

In _Napaeozapus_ the head and neck of the malleus resemble those of _Zapus_ but are less robust. The body is more rounded dorsally, having the curved dorsal surface directed anteriorly rather than posteriorly (as in _Zapus_) and the lateral surface is nearly flat instead of curved as in the other genera. The manubrium resembles that of _Eozapus_ and _Zapus_. The body of the incus is flattened dorsally but otherwise rounded. The long limb of the incus is angular and longer than that of _Zapus_. The short limb of the incus is broad at the base and tapers distally. The limbs of the stapes are narrow, weak, and abruptly curved.

The neck is more slender and elongate than in _Zapus_.

In summary: Only the head and body of the malleus and the short and long limbs and body of the incus are sufficiently consistent within a given group to be of taxonomic importance. The similarity in the morphology of these ossicles indicates a close relationship between all three genera.

_Zapus_ and _Napaeozapus_ resemble one another more than either resembles _Eozapus_. The differences recorded are constant between the described groups and, therefore, are considered to be of taxonomic significance. The differences give basis for dividing the subfamily Zapodinae into the three genera _Eozapus_, _Zapus_, and _Napaeozapus_.

[Ill.u.s.tration: FIGS. 32-34. Lateral views of the left ear ossicles (articulated) of the Recent genera of the subfamily Zapodinae. 20.

FIG. 32. _Eozapus s. vicinus_, adult, male, No. 240762 USNM, Lanchow, Kansu, China.

FIG. 33. _Zapus p. princeps_, adult, male, No. 32858 KU, Medicine Wheel Ranch, 28 mi. E Lovell, Big Horn County, Wyoming.

FIG. 34. _Napaeozapus i. insignis_, adult, male, No. 9544 KU, 3 mi.

W Base Station, Coos County, New Hampshire.]

Distribution of and Speciation in the Genus _Zapus_

Many of the described kinds of the genus _Zapus_ were initially named as distinct species (see Preble, 1899). Subsequently (see Hall, 1931), some of the nominal species were reduced to the rank of subspecies. Only three species in the genus _Zapus_ are recognized in the following account. The concept of species adopted here is, in Mayr's (1942:120) words, this: "Species are groups of actually or potentially interbreeding natural populations, which are reproductively isolated from other such groups." The three species are _Z. trinotatus_, _Z.

princeps_, and _Z. hudsonius_. No hybridization is known where two occur together or where their ranges are adjacent. Each of these species has several geographically contiguous subspecies.

The three species of _Zapus_ are closely related but are not equally progressive. If eastern North America is considered to be the region of origin and center of dispersal of _Zapus_ (see pp. 368-369) the geographically distant species would be expected to be the least progressive, and such seems to be the case. _Zapus trinotatus_ is geographically farthest removed and structurally least progressive.

_Zapus hudsonius_ occurs at the center of dispersal and is the most progressive structurally whereas _Z. princeps_ is geographically and structurally intermediate. Structural progressiveness is postulated for the species that has the simplest (in this instance specialized) baculum and smallest fourth upper premolar. The phyletic branches of the genus _Zapus_ possibly developed from geographic segments of a population radiating from the centrally located progressive group. On continental areas where a species with a wide and continuous range gives rise to several daughter species, geographic isolation is thought to be important in bringing about the formation of species. The unspecialized populations conceivably occupied an area west of the present Rocky Mountains and south of lat.i.tude 50. From later Miocene times on, climatic and geological differentiation occurred in this area, and with the growth of geological barriers and differentiation of habitat these unspecialized populations may have been separated into two ecological groups, one inhabiting the more arid area between the present Rocky Mountains and the present Cascade Range and Sierra Nevada and the other group inhabiting the Pacific coastal region. Isolation of each of these groups probably was not complete. How far differentiation might have proceeded with incomplete isolation can only be guessed, but at least incipient differences probably were present and possibly the animals approached in character those found in these areas today in that the ecology of the region was much the same as now.

In the region between the Rocky Mountains and the present Cascade Range and the Sierra Nevada, the flora (in late Pliocene) became semidesert, which presumably made most of this region uninhabitable for jumping mice. The aridity probably induced local concentration into boreal montane islands, thus possibly displacing the populations of the two species that were in contact.

In Pleistocene times continental glaciation must have interrupted the contacts between the coastal, intermontane (the area between the present Rocky Mountains and the present Cascade Range and the Sierra Nevada), and northern and eastern groups of _Zapus_ or mammals of any genus that occurred over all of this vast region. The advance of the ice southward would have increased opportunity for evolution by interposing barriers that isolated some populations. The populations possibly were re-established in interglacial periods and then were isolated again by another descent of glacial ice.

If a population occupied the unglaciated coastal region of Oregon and Washington it may have been separated from other populations to the north and east by an ice cap which covered most of the Cascade Range.

The population occupying the intermountain region probably was isolated from the population to the north and west. The formation of glaciers presumably reduced the size of areas available to the populations occupying eastern North America, Alaska, and Canada with the result that they persisted only in areas south of the ice or in ice-free refugia (central and western Alaska) within the glaciated area. According to Axelrod (1948), the flora in the eastern United States during the Pleistocene furnished most of the stock for the revegetation of southern and subarctic Canada east of the Rocky Mountains. Eastern populations of _Z. hudsonius_ (or its progenitors) probably followed the spread of this vegetation and, thus, extended their range into Canada where they crossbred with populations advancing south and east from the refugia in Alaska. Western montane floras, which extended north along the Rocky Mountains and the Cascade and Coast ranges, probably paved the path for a northward migration of populations of the intermountain _Z. princeps_ (or its progenitors). Populations of _Z. princeps_ moved eastward from the present Rocky Mountains, inhabiting the high plains of southern Canada and the north-central United States. In general, _Zapus hudsonius_ occupies the region to the north and to the east of that inhabited by _Zapus princeps_; however, the ranges of the two meet and overlap in central and northern British Columbia and in the high plains area of southern Alberta, Saskatchewan, eastern Manitoba, eastern Montana, North Dakota, and northern South Dakota. In these places of overlap, owing to range expansion following the retreat of the ice, there is no sign of interbreeding, indicating that the populations have attained specific rank.

Populations of both _Z. hudsonius_ and _Z. princeps_ occur together at Indianpoint Lake, British Columbia. Specimens taken there are readily sorted into two groups; none is intermediate. The difference in size between these species there is especially marked; _Z. p. saltator_ there is a large derivative of _Z. princeps_ and _Z. h. tenellus_ is a medium-sized _Z. hudsonius_.

_Z. princeps_ minor and _Z. hudsonius intermedius_ have been taken at several neighboring localities in North Dakota. Although these geographic races are more nearly of the same size (_minor_ is a small subspecies of _princeps_ and _intermedius_ is a moderately large subspecies of _hudsonius_) they do not interbreed. Specimens of _Z. p.

minor_ and _Z. h. intermedius_ have been obtained from an ecologically h.o.m.ogeneous area in the vicinity of Fort Totten and Devils Lake, North Dakota. Values obtained from several measurements of the skull and baculum allow for ready recognition of the two species. The populations from North Dakota are, however, not so widely divergent as are those populations from the area of contact in British Columbia. Perhaps the difference in the degree of distinction between the species at the two areas of contact is indicative of the length and completeness of geographic isolation between neighboring populations.

The ranges of _Z. trinotatus_ and _Z. hudsonius_ are not at present in contact, but the two species differ more strongly than do _hudsonius_ and _princeps_ or _princeps_ and _trinotatus_. Therefore, _trinotatus_ and _hudsonius_ are here considered to be two distinct species.

As pointed out earlier in this discussion, the separation between the progenitors of _Z. trinotatus_ and _Z. princeps_ probably occurred when the present Cascade Range and the Sierra Nevada were being formed. From this time until Pleistocene glaciation an incomplete geographic isolation was in effect between the populations of the Pacific coast and the intermountain populations. Perhaps in the region north of the present Cascade Range there was moderate interbreeding between these populations and the transcontinental form. There may have been a similar zone of interbreeding along the crest of the present Cascades where the intermountain and coastal populations conceivably could have met. At least incipient characters probably were present when in Pleistocene time, continental glaciation further isolated the two populations. Since the retreat of the last ice (Wisconsin) the unprogressive coastal _Z.

trinotatus_ has expanded its range only slightly, reaching as far as southwestern British Columbia. It seems that ecological difference rather than the barrier formed by the higher elevations is responsible for the limited expansion of range. The population of _princeps_ has extended its range northward to the southern part of the Yukon Territory but does not occur in coastal southern British Columbia because that area already was occupied by _Zapus trinotatus_. The ranges of the two species meet and overlap in southwestern British Columbia. The species occur sympatrically in Manning Park where, according to Carl _et al._ (1952:77), they occupy the same range in the region of Allison Pa.s.s, Pinewoods, and Timberline Valley. These workers remark that no intergradation was apparent between individuals of the two species obtained in the same trap line.

I have examined material of both species from Allison Pa.s.s. There the species differ in color, in the shape of the skull, and in the size and shape of the baculum. Material from Timberline Valley, an area in which Carl _et al._ (_loc. cit._) reported both species, here is a.s.signed to _Z. princeps_. Where bacula have been preserved the ident.i.ty of the species is instantly possible.

In summary: First, a population of jumping mice, possibly a monotypic genus, occurred over most of North America; then this population partly divided into Pacific northwest, intermountain (from the east slopes of the present Rocky Mountains to the east slopes of the present Cascade Range and the Sierra Nevada), and transcontinental (eastern and northern) groups with the least progressive groups peripheral; a further reduction or possibly a complete isolation of these populations followed owing to Pleistocene glaciation (especially in the Wisconsin period); and, finally, the present day contacts were established between these populations which by now have differentiated into species. Conceivably, _Z. burti_ (Blancan age) and _Z. rinkeri_ (mid Pleistocene) may represent stages in the development of _Z. hudsonius_.

ANNOTATED LIST OF SPECIFIC AND SUBSPECIFIC NAMES

(Applied to the genus _Zapus_ since 1899)

Edward A. Preble's (1899) early revisionary account of the genus _Zapus_ provides an annotated list of the names which had been proposed for American jumping mice to that date. The present account supplies in chronological order the names proposed (including the new kinds described by Preble) in the 54 years since Preble's revision. Detailed synonymies are given for each kind under the accounts of the subspecies.

1899 _campestris_ (_Zapus hudsonius_) Preble, N. Amer. Fauna, 15:20, August 8, 1899, applies to the jumping mouse of southeastern Montana, and the Black Hills region of Wyoming and South Dakota.

1899 _minor_ (_Zapus princeps_) Preble, N. Amer. Fauna, 15:23, August 8, 1899, originally applied to the jumping mouse of the prairies of Saskatchewan, but now includes populations of this species from the plains of Canada (southern Manitoba to Canadian Rockies) and northern United States (Montana, North and South Dakota).

1899 _oregonus_ (_Zapus princeps_) Preble, N. Amer. Fauna, 15:24, August 8, 1899, originally applied to the jumping mouse of eastern Oregon, but now applies also to populations from southeastern Idaho, eastern and central Nevada, and extreme northeastern California.

1899 _major_ (_Zapus_) Preble [= _Zapus princeps oregonus_], N.

Amer. Fauna, 15:25, August 8, 1899, arranged as a subspecies of _Zapus princeps_ by Hall, Univ. California Publ. Zool., 37:10, April 10, 1931; here considered a synonym of _Zapus princeps oregonus_.

1899 _nevadensis_ (_Zapus_) Preble [= _Zapus princeps oregonus_], N. Amer. Fauna, 15:25, August 8, 1899, arranged as a subspecies of _Zapus princeps_ by Hall, Univ. California Publ. Zool., 37:10, April 10, 1931; here considered a synonym of _Zapus princeps oregonus_.

1899 _orarius_ (_Zapus_) Preble [= _Zapus trinotatus orarius_], N.

Amer. Fauna, 15:29, August 8, 1899, applies to the animals from southwestern Marin County, California.