at the University of Michigan by Dr. E. T. Hooper, The American Museum of Natural History by Mr. G. G. Goodwin, The United States National Museum by Dr. David H. Johnson and the Biological Surveys Collection by Miss Viola S. Schantz. A fellowship from the National Science Foundation made possible the studies at the museums other than at the University of Kansas.

Microtus montanus nanus (Merriam)

Arvicola (Mynomes) nanus Merriam, N. Amer. Fauna, 5:62, July 30, 1891.

Microtus montanus nanus, Hall, Proc. Biol. Soc. Washington, 51:131, August 23, 1938.

Microtus nanus, Bailey, N. Amer. Fauna, 17:30, June 6, 1900 (part).

Microtus montanus caryi Bailey, Proc. Biol. Soc. Washington, 30:29, February 21, 1917.

Microtus nanus nanus, Cary, N. Amer. Fauna, 42:43, October 3, 1917 (part).

Type.—Adult male No. 23853/31253, U. S. National Museum, Department of Agriculture collection, from Pahsimeroi Mountains, Custer County, Idaho; obtained by C. Hart Merriam and Vernon Bailey, September 16, 1890.

Range.—Idaho; southwestern Montana; most of the southwestern half of Wyoming; southward to central Colorado. See figure 1.

Comparisons.—Comparisons with subspecies named as new in this paper will be found in the accounts of those subspecies beyond. From Microtus montanus fusus Hall, the subspecies to the south, M. m. nanus from Idaho differs as follows: averages smaller; slightly darker and less reddish and less yellowish in color; slightly wider braincase (see measurement of prelambdoidal breadth); larger bullae.

Measurements.—Average (= arithmetical mean) measurements of 34 specimens, both male and female, from several localities in eastern Idaho are: total length, 151; length of tail, 39; hind foot, 19.2; condylobasilar length of the skull, 25.0; zygomatic breadth, 15.0; alveolar length of upper molar tooth-row, 6.4; prelambdoidal breadth, 8.9; and lambdoidal breadth, 11.7.

Average and extreme measurements of six adult males from near Pocatello, Bannock County, Idaho, and nine adult males from near Afton, Lincoln County, Wyoming, are, respectively, as follows: total length, 143(135-150), 163(143-179); length of tail, 35.1(33-38), 42.8(36-49); caudal index, 32.0(28.0-33.1), 35.7(30.6-41.9); hind foot, 18.9(18-20), 18.8(17-20); condylobasilar length of skull, 24.4(24.0-26.0), 25.6(24.5-26.2); alveolobasilar length, 14.1(13.7-14.5), 14.6(13.8-15.0); palatilar length, 13.2(12.9-13.6), 13.8(13.2-14.5); alveolar length of upper molar tooth-row, 6.3(6.1-6.5), 6.3(6.0-6.6); depth of braincase, 7.7(7.5-7.9), 8.0(7.7-8.3); lambdoidal breadth, 11.4(11.0-11.7), 12.0(11.3-12.7); prelambdoidal breadth, 9.1(8.6-9.4), 8.7(8.0-9.4); zygomatic breadth, 14.3(13.8-14.7), 15.3(14.4-16.3); interorbital breadth, 3.6(3.5-3.7), 3.5(3.3-3.7). The average length of the nasal bones in the series from Pocatello is 7.1 mm. The averages, which have not been included in Table 1, for three measurements of the series from Carbon County, Wyoming, are as follows (Encampment, males; Encampment, females; Savery, males; and Savery, females, respectively): alveolobasilar length, 14.4, 14.3, 14.5, 14.3; interorbital breadth, 3.5, 3.4, 3.5, 3.4; depth of braincase, 7.8, 7.6, 7.9, 7.6. Additional measurements are included in Table 1 for other series.

Discussion.—The name Microtus montanus caryi Bailey is here placed in synonymy under M. m. nanus (Merriam). Vernon Bailey (1917) in his description of caryi made four assumptions that have been found to be entirely or partly invalid. First, he assumed that this is an "extreme variant which gradually changes in characters across Nevada and Utah, and reaches its maximum variation in Wyoming." The differences pointed out in subsequent descriptions of subspecies found in the above area do not show a gradual change in any character, or in the number of characters, nor is caryi an extreme when compared with the other subspecies. Second, Microtus nanus was not, as Bailey assumed, a different species than Microtus montanus. Third, he assumed that the characteristics of adults of nanus were adequately ascertainable from the thirteen topotypes available to him. Subsequent sampling from Idaho shows that the series of specimens available to Bailey was made up mostly of young and subadult animals. Finally, caryi does not occupy as Bailey stated "the meadows along streams in the arid sagebrush country of the Bear River, Green River, and Wind River valleys" exclusively, or characteristically. When the localities from which the species actually is known are plotted, it seems that the arid basin serves as a barrier and that the species is more commonly and abundantly found in montane meadows in the Transition and Canadian life-zones.

Certain samples, here assigned to M. m. nanus, that vary from the average of the subspecies deserve comment. For example, mice from the area in Wyoming southwest of the Green River (in the Uinta Mountains) have relatively smaller feet, but are larger in both total length and size of skull. Specimens from near Afton, Lincoln County, Wyoming, are relatively large in both total length and size of skull. This series and specimens from Teton County, Wyoming, are intermediate between nanus from Idaho and the newly named subspecies from near Cody, Park County, Wyoming, described below, in terms of both darkness and the amount of reddish color. Mice from Laramie County are more richly reddish-brown. The specimens from near Savery, in Carbon County, Wyoming, are darker. The alveolobasilar length relative to the condylobasilar length is smaller in the series from along Deer Creek, 16 mi. S, 11 mi. W Waltman, Natrona County, Wyoming. The series from the southern tier of counties in Wyoming and some of the specimens from Colorado have relatively wider zygomatic arches. The specimens from southern Sweetwater County, Wyoming, are relatively paler, have a relatively longer tail and longer hindfeet, lesser condylobasilar length, and wider braincase. Most of these variations are of questionable significance; they may be chance variations owing to errors in sampling.

Much of the south-central part of the state is relatively low and relatively arid. This area includes the arid basin of the Green River and its major tributaries and the arid Red Desert along the continental divide in Sweetwater County. This area might have acted as a barrier to the mice; gene flow might have been prevented between the populations of the western part of the state and those farther east in the Medicine Bow Mountains and Laramie Mountains. Nevertheless geographic variations of subspecific worth have not taken place. The barrier has either not been of as long duration, or has not been so complete and effective, as the other barriers in the state, namely the Absaroka Range, the Big Horn Basin, the Shoshone Basin, and the valley of the North Platte River. These four barriers presumably have led to the differentiation of the two subspecies that are newly named beyond. Each of the two areas which is set apart by these barriers and in which one of the newly named subspecies has evolved is small; therefore there is a lesser amount of suitable habitat available for each of the newly named mice than there is for M. m. nanus. It is conceivable, therefore, that in periods of adverse conditions in each of the small areas the size of the