BOSTON PUBLIC LIBRARY 3 9999063176547 BH1 rF Given By ^-X^i 4A* Z) g-a . U. S. DEPARTMENT OF AGRICULTURE BUREAU OF BIOLOGICAL SURVEY NORTH AMERICAN FAUNA 2STo. 49 [Date of publication, December, 1926] A BIOLOGICAL SURVEY OF NORTH DAKOTA I. PHYSIOGEAPHY AND LIFE ZONES II. THE MAMMALS VERNON BAILEY BIOLOGI8T DIVISION OF BIOLOGICAL INVESTIGATIONS BUREAU OF BIOLOGICAL SURVEY WASHINGTON GOVERNMENT PRINTING OFFICE 1826 NORTH AMERICAN FAUNAS Copies of North American Faunas not out of print are for sale, at the prices named, by the Superintendent of Documents, Government Printing Office, Washington, D. 0. No. 1. Revision of the North American Pocket Mice. By C. Hart Merriam. Pp. 36, pis. 4. 1889 [Out of print. \ No. 2. Descriptions of Fourteen New Species and One New Genus of North American Mammals. By C. Hart Merriam. Pp. 52, pis. 8, figs. 7. 1889. [Out of print.] No. 3. Results of a Biological Survey of the San Francisco Mountain Region and Desert of the Little Colorado, Arizona. By C. Hart Merriam and Leonhard Stejneger. Pp. 136, pis. 14, maps 5 (colored), figs. 2. 1890. [Out of print.] No. 4. Descriptions of Twenty-six New Species of North American Mammals. By C. Hart Merriam. Pp. 60, pis. 3, figs. 3. 1890 [Out of print.] No. 5. Results of a Biological Reconnoissance of South-central Idaho. By C. Hart Merriam and Leonhard Stejneger. Descriptions of a New Genus and Two New Species of North American Mammals. By C. Hart Merriam. Pp. 132, pis. 4 (1 colored), figs. 4. 1891 [Out of print.] No. 6. Not issued. No. 7. The Death Valley Expedition: A Biological Survey of Parts of Cali- fornia, Nevada, Arizona, and Utah. Part II. — 1. Birds, by A. K. Fisher. 2. Reptiles and Batrachians, by Leonhard Stejneger. 3. Fishes, by Charles H. Gilbert. 4. Insects, by C. V. Riley. 5. Mollusks, by R. E. C. Stearns. 6. Desert Trees and Shrubs, by C. Hart Merriam. 7. Desert Cactuses and Yuccas, by C. Hart Merriam. 8. List of Localities, by T. S. Palmer. Pp. 402, pis. 15, maps 5, figs. 2. 1893 •_ [Out of print.] No. 8. Monographic Revision of the Pocket Gophers, Family Geomyidae (exclu- sive of the species of Thomomys). By C. Hart Merriam. Pp. 258, pis. 20, figs. 71, maps 4 (colored). 1S95 [Out of print.] No. 9. Not issued. No. 10. Revision of the Shrews of the American Genera Blarina and Notiosorex. By C. Hart Merriam. The Long-tailed Shrews of the Eastern United States. By Gerrit S. Miller, jr. Synopsis of the American Shrews of the Genus Sorex. By C. Hart Merriam. Pp. 124, pis. 12, figs. 3. 1895. [Out of print.] No. 11. Synopsis of the Weasels of North America. By C. Hart Merriam. Pp. 44, pis. 6, figs. 16. 1896 [Out of print:] No. 12. The Genera and Subgenera of Voles and Lemmings. By Gerrit S. Miller, jr. Pp. 84, pis. 3, figs. 40. 1S9G [Out of print.] No. 13. Revision of the North American Bats of the Family Vespertilionidae. By Gerrit S. Miller, jr. Pp. 140, pis. 3, figs. 40. 1897 [Out of print,] No. 14. Natural History of the Tres Marias Islands, Mexico. General Account of the Islands, with Reports on Mammals and Birds, by E. W. Nelson. Rep- tiles, by Leonhard Stejneger. Notes on Crustacea, by Mary J. Rathburn. Plants, by J. N. Rose. Bibliography, by E. W. Nelson. Pp. 97, pi. (map), figs. 2. 1S99 [Out of print,] No. 15. Revision of the Jumping Mice of the Genus Zapus. By Edward A. Preble. Pp. 42, pi., figs. 4. 1890 [Out of print] No. 16. Results of a Biological Survey of Mount Shasta, California. By (1 Hart Merriam. Pp. 179, pis. 5, figs. 46. 1899 [Out of print.] No. 17. Revision of American Voles of the Genus Microtus. By Vernon Bailey. Pp. 88, pis. 5, figs. 17. 1900 [Out of print.] No. 18. Revision of the Pocket Mice of the Genus Perognathus. By Wilfred K, Osgood. Pp. 72, pis. 4 (incl. 2 maps), figs. 15. 1900 [Out of print)] No. 19. Results of a Biological Reconnoissance of the Yukon River Kegion : Gerl eral Account of the Region ; Annotated List of Mammals, by Wilfred B Osgood. Annotated List of Birds, by Louis B. Bishop. Pp. 100, pis. 7 (inci 1 map). 1900 [Out of print.] No. 20. Revision of the Skunks of the Genus Chincha [Mephitis]. By Arthu H. Howell. Pp. 62, pis. 8. 1901 [Out of print.] No. 21. Natural History of the Queen Charlotte Islands, British Columbia; and Natural History of the Cook Inlet Region, Alaska, By Wilfred H. Osgood Pp. 87, pis. 7 (incl. 1 map), fig. (map). 1901— [Out of print.] (Continued on page 3 of cover) lonqiiud* W»t from Greenwich 100* B M 2 ^ ii © rfer I € 1 f>|V I A M S j ilhite Earth { \\/'<»»»'J!te/«?\"| j ^^-y & ~fj \[ 1 ^ ! S>*3s, J4 L E : •TOWN JER .Hanilltorj E M B I N A Draytor\ fe IOUNTRAI L ! W A R —I ; L .-«*!* 1 II GRAND FORKSL ;"aN D FORKS ^ ^^J r~ --^NEW ROCKFORD • 0FESSENDEN j ^N^ \ WELLS j _^_ J^W* M'CLUSKY I " tt-H* .1. J =sL i I F O S^ f E R I G R I G 6 SJ S T E Z lV j ^5j> , ! «CARRlNGTOl!l COOPERSTOWNSj ^• ^fcP^LpW'-'NGS.i ,. j o l i v. e r) Vsr-L BU R L E I G >H j KIDDER B^RVN E S f f I fiedma VALLEY CITY; C L2*ew^'/rtp ; H E^TvJI N G £-fl' j '-""'"ope 6""? ^--. :r\ j Streeter rrG a N j L a/oVre hv^pf ~~- \S / ^A MOURE! ^^w' J Rl CHLAiN D O W M a/ N T A D A^-M s xY^ > 0 '^1^ s i Mc INTOSH | D / C K E yM^ * " WM L^^/. .^ j »Zeeland Canadian Zone Transition Zone Upper Austral Zone A LIFE ZONES OF NORTH DAKOTA Bv VERNON BAILEY BUREAU OF BIOLOGICAL SURVEY US DEPARTMENT OF AGRICULTURE Scut of Miles- kLudcli K T U. S. DEPARTMENT OF AGRICULTURE BUREAU OF BIOLOGICAL SURVEY NORTH AMERICAN FAUNA ISTo. 49 [Date of publication, December, 1926] A BIOLOGICAL SURVEY OF NORTH DAKOTA I. PHYSIOGRAPHY AND LIFE ZONES II. THE MAMMALS VERNON BAILEY BIOLOGIST DIVISION OP BIOLOGICAL INVESTIGATIONS BUREAU OP BIOLOGICAL SURVEY WASHINGTON GOVERNMENT PRINTING OFFICE 1926 _r ' 7 LETTER OF TRANSMITTAL United States Department of Agriculture, Bureau of Biological Survey, Washington, D. C, December 11, 1925. Sir: I have the honor to transmit herewith, recommending that it be published as No. 49 in the series of the North American Fauna, a report on the biological survey of North Dakota, prepared by Vernon Bailey, biologist of this bureau. This report is in two parts, the first treating of the physiography and natural life zones of the State, accompanied, as in similar reports, by a map of the life zones ; and the second, the mammalian life, consisting of notes on the dis- tribution, abundance, and habits of the mammals of the State. Both are based on natural-history explorations conducted by the bureau and cooperating State organizations in North Dakota over many years, the work on the mammals having begun in 1887, and prelimi- nary reports thereon having been published in the annual report of this bureau in 1888, when it was known as the Division of Economic Ornithology and Mammalogy, and in a circular of the North Dakota Agricultural Experiment Station, in 1914, the latter being in the nature of a cooperative report of progress and an appeal for addi- tional local detailed information. The present report comprises a valuable contribution to knowledge and will be useful to farmers, students, and others interested in the distribution, habits, and eco- nomic relations of our wild-animal life. Respectfully, E. W. Nelson, Chief of Bureau. Hon. W. M. Jardine. *'. • Secretary ■ of AgricuUuPe. '..!•-'•••' ■ CONTENTS Page Introduction 1 Paet I. — Physiography and life zones of North Dakota 3 Changing conditions 3 General physiographic features 3 Glacial remains 3 Lowered water levels 4 Drainage systems 4 Elevations and the Badlands 5 Prairie 7 Forest 7 Life zones S Upper Austral Zone 8 Mammals 9 Breeding birds 9 Plants 10 Crop adaptations 10 Transition Zone 11 Mammals 11 Breeding birds 12 Plants 13 Crop adaptations 14 Canadian Zone 14 Mammals 15 Breeding birds 15 Plants - 16 Crop adaptations 16 Part II. — The Mammals of North Dakota _ 17 Introduction '. 17 Present and former abundance 17 Useful and harmful species 18 Indian names of mammals 18 Measurements and weights , T 18 Order Art iodactyla : Hoofed animals — cattle, sheep, goats, antelope, and deer 19 Family Bovidae : Cattle, sheep, and goats 19 American bison, American buffalo _ 19 Audubon mountain sheep 25 Family Antilocapridae : Pronghorned antelope 27 Pronghorned antelope, American antelope, pronghorn 27 Family Cervidae : Moose, elk. caribou, and deer 31 Moose 31 Woodland caribou 32 American elk, wapiti 33 Plains white-tailed deer 36 Mule deer 41 Order Rortentia : Gnawing animals 43 Family Sciuridae : Squirrels, chipmunks, prairie dogs, ground squirrels, and marmots 43 Pale flying squirrel 13 Minnesota gray squirrel, black squirrel 45 Red squirrel, chickaree , 46 Little northern chipmunk- 47 Pale chipmunk. Badlands chipmunk 49 Gray chipmunk 51 in IV CONTENTS Paet II. — Mammals of North Dakota — Continued. Order Rodentia : Gnawing animals — Continued. Family Sciuridae: Squirrels, chipmunks, etc. — Continued. Page Striped ground squirrel, thirteen-lined ground squirrel, leopard squirrel 52 Pale striped ground squirrel, pale thirteen-lined ground squirrel 54 Gray ground squirrel, Franklin ground squirrel 55 Richardson ground squirrel, flickertail- 58 Black-tailed prairie dog 62 Rufescent woodchuck, groundhog 67 Canada woodchuck, groundhog 69 Family Muridae : Old World rats and mice 70 Brown rat, house rat, wharf rat 70 House mouse 72 Family Cricetidae : White-footed mice, harvest mice, grasshopper mice, wood rats, and voles 73 Osgood white-footed mouse 73 Baird white-footed mouse 76 Northern white-footed mouse, deer mouse 77 Badlands white-footed mouse 79 Prairie harvest mouse 80 Maximilian grasshopper mouse 81 Audubon grasshopper mouse 83 Pale bushy-tailed wood rat 86 Red-backed mouse 88 Eastern meadow mouse 90 Drummond meadow mouse 92 Bean mouse, hetunka .. 94 Western upland mouse 98 Little upland mouse 99 Pale mouse 101 Great Plains muskrat 102 Family Castoridae : Beavers 105 Canada beaver 105 Missouri River beaver 108 Family Erethizontidae : Porcupines 114 Yellow-haired porcupine, Rocky Mountain porcupine 114 Black-haired porcupine, Canada porcupine 116 Family Zapodidae : Jumping mice 117 Prairie jumping mouse 117 Family Heteromyidae : Pocket mice, kangaroo rats 119 Maximilian pocket mouse 119 Dusky pocket mouse 121 Kansas pocket mouse 123 Richardson kangaroo rat 124 Family Geomyidae : Pocket gophers 125 Mississippi Valley pocket gopher 125 Dakota pocket gopher 130 Sagebrush pocket gopher 133 Order Lagomorpha : Rabbitlike animals 134 Family Leporidae : Rabbits 134 Nebraska cottontail 134 Black Hills cottontail 137 Wyoming cottontail 137 Varying hare, white rabbit, snowshoe rabbit 138 White-tailed jack rabbit 141 Order Carnivora : Flesh eaters 144 Family Felidae : Cats 144 Mountain lion, cougar, panther 144 Canada lynx . 146 Northern bobcat, mountain bobcat, spotted wild cat 148 Eastern bobcat, wild cat 149 CONTENTS V Part II. — Mammals of North Dakota — Continued. Order Carnivora : Flesh eaters — Continued. I"age Family Canidae: Dogs, wolves, and foxes 150 Gray wolf, buffalo wolf, lobo, loafer 150 Northern coyote, brush wolf 156 Plains coyote, prairie wolf 157 Yellow-red fox 160 Kit fox, prairie fox, swift 163 Family Mustelidae : Weasels, minks, martens, skunks, badgers 166 Long-tailed weasel, ermine 166 Bonaparte weasel, short-tailed weasel 169 Least weasel 170 Black-footed ferret 171 Mink 173 Marten, pine marten, American sable 176 Fisher, pekan, black cat 177 Wolverene, glutton, " Indian devil " 178 Otter 179 Northern skunk 181 Badger 184 Family Procyonidae : Raccoons 1S7 Raccoon, "coon" 187 Family Ursidae: Bears 191 Black bear, cinnamon bear 191 Grizzly bear, big Plains grizzly, silvertip ' 193 Absaroka grizzly 198 Order Insectivora : Insect-eating mammals 199 Family Talpidae : Moles 199 Missouri Valley mole 199 Star-nosed mole 200 Family Soricidae : Shrews 200 Hayden masked shrew 200 Merriam shrew 202 Richardson shrew, black-backed shrew, saddle-backed shrew. 203 Water shrew, marsh shrew 203 Pigmy shrew 204 Short-tailed shrew, mole shrew 205 Order Chiroptera : Winged mammals 207 Family Vespertilionidae : Common bats 207 Hoary bat, great gray bat 207 Red bat, New York bat 209 Large brown bat 210 Silver-haired bat, silvery bat, black bat 212 Little brown bat 213 Yellowstone bat 215 Say bat 216 Little long-eared bat 216 Bibliography 217 Index 221 ILLUSTRATIONS PLATES Page Plate 1. Map of North Dakota, showing life zones Frontispiece 2. Fig. 1. — Short grass prairie of western North Dakota. Fig. 2. — Prairie slough and glacial ridge of central North Dakota 4 3. Fig. 1. — Yellow pines on buttes south of Medora. Fig. 2. — Bad- lands and sagebrush at Medora 5 4. Fig. 1. — Typical lake of the Turtle Mountains. Fig. 2. — Typical aspen forest of the Turtle Mountains 8 5. Fig. 1. — Red River with its forested shores, near Fargo. Fig. 2. — Type of forest along the Red River, near Fargo 8 6. Cottonwood timber along the Missouri River bottoms, near Mandan . 8 7. Buck antelope 2S 8: Heads of moose, northern white-tailed deer, and mule deer 32 9. Fig. 1. — Two bull elk. Fig. 2.— Plains white-tailed deer 33 10. Skins of pale, little northern, and gray chipmunks 4S 11. Fig. 1. — Osgood white-footed mice in captivity. Fig. 2. — Bean mouse. Fig. 3. — Richardson kangaroo rat 76 12. Kansas pocket mouse, prairie jumping mouse, and prairie har- vest mouse 77 13. Fig. 1. — Grasshopper mouse. Fig. 2. — Pale bushy-tailed wood rat 84 14. Pale mouse, little upland mouse, western upland mouse, Drum- mond mouse, bean mouse, and eastern meadow mouse 85 15. Fig. 1. — Dusky pocket mouse. Fig. 2. — Badger 124 16. Mississippi Valley pocket gopher : 125 17. Fig. 1.— Yellow-red fox. Fig. 2.— Black-footed ferret. Fig. 3. — Varying hare, or snowshoe rabbit 1G0 18. Fig. 1.— Plains coyote. Fig. 2.— Otters 161 19. Fig. 1. — Common mole. Fig. 2. — Star-nosed mole 200 20. Short-tailed, Richardson, Hayden, and pigmy shrews 201 21. Fig. 1. — Hoary and silver-haired bats. Fig. 2. — Say bat 208 TEXT FIGURES Fig. 1. Map showing records of three species of chipmunks in North Dakota : Gray chipmunk, little northern chipmunk, and pale chipmunk 48 2. Map showing distribution of the thirteen-lined ground squirrel and its pale western form in North Dakota . 53 3. Map showing records of Franklin ground squirrel in North Dakota- 56 4. Map showing records of Richardson ground squirrel in North Dakota 59 5. Map showing distribution of prairie-dog towns in North Dakota 63 6. Map showing localities where woodchucks are known in North Dakota 67 7. Map showing records of pocket gophers in North Dakota : The Mississippi Valley pocket gopher and the Dakota pocket gopher_ 125 8. Map showing records of three species of cottontail rabbit in North Dakota : The Nebraska cottontail, the Black Hills cottontail, and the Wyoming cottontail 134 VI No. 49 NORTH AMERICAN FAUNA December, 1926 A BIOLOGICAL SURVEY OF NORTH DAKOTA By Vernon Bailey INTRODUCTION In the preliminary survey of the wild life of North America in- formation has been gathered on the birds and mammals of the country at large, and provisional maps of the life zones of the continent and subdivisions of it have been published. Much of the general information gathered on birds and mammals has been given in bulletins, circulars, and annual reports. The present publication is prepared in accordance with the general plan of providing for definite subdivisions more detailed information on the natural life zones and on the distribution and habits of the native species of birds and mammals. Part I discusses the life zones of North Da- kota and Part II the mammals of the State. The publication of the report on the birds will be arranged for separately. The field work on which this report is largely based has been carried on in North Dakota by the Biological Survey at intervals from the year following its first organization as the Division of Economic Ornithology and Mammalogy in 1886. In 1912 a definite plan of cooperation for covering the State comprehensively by field work and for gathering the specimens and notes necessary for a bet- ter understanding of the animal life was entered into between the Biological Survey and organizations in North Dakota, including the State university, the agricultural experiment station, the agri- cultural and geological survey, and other State educational institu- tions. Under this plan field work was carried on each season dur- ing the subsequent four years.1 In addition to the Biological Sur- vey material, the collections of specimens at the agricultural college at Fargo and of those at the biological laboratory at Devils Lake, with the many field notes and reports gathered in connection with these, have been freely drawn upon in the preparation of this report. The Flora of North Dakota, by Herbert F. Bergman (1918), pub- lished in the Sixth Biennial Report of the North Dakota Soil and Geological Survey, has been of great assistance in the preparation of the life-zone report and the map. Also, free use has been had of 1 The field work of the Biological Survey was carried on with the assistance of H. E. Anthony, Alfred Eastgate, Stanley :ikota days old, and that the eyes do not open till the thirtieth day. The number of young in a litter varies widely, but seems to be usually from 7 to 10. A female taken by Sheldon at Fairmount, on May 9, contained 11 embryos, and there are other records of still larger numbers up to 13 (Lee) and 14 (Seton). The full number of mammae in adult females is 12. Apparently but one litter of young is raised in a season, and even for that the time is short for them to mature and lay up sufficient fat and food to carry them through the six months of hibernation. Food habits. — Although a great part of their food consists of seeds, grain, and nuts, they are omnivorous in habits and take besides berries and some green vegetation, numerous insects, and the flesh of mice, birds, or any small animals which they can capture or find dead. Acorns and hazelnuts are eagerly gathered and stored for winter food, but over most of their range only the smaller seeds and nutlets are obtained, unless grainfields are within reach. Seeds and grain are stored for future use, but much soft food that will 54 NORTH AMERICAN FAUNA [No. 49 not keep is eaten as it is taken. The examination of the contents of large numbers of stomachs shows a considerable portion of grass- hoppers, crickets, caterpillars, beetles, ants, cocoons, insect eggs, and even traces of flesh, hair or small mammals, and feathers of birds; also green foliage, the white p'ulp of bulbs and tubers and the fruit of solanum, cactus, and strawberries. The contents of the ground squirrel's capacious cheek pouches give a good index to the selection of seeds and grains. The pouches are often distended with wheat, oats, barley, rye, or any of the cultivated grains that are obtainable, but also are found to contain acorns, hazelnuts, seed of sunflower, cactus, bindweed, goosefoot, puccoon, wild peas and beans, and a great variety of grass seeds. During late summer and fall, all work industriously, laying up their winter stores, quickly filling their cheek pouches and running to the burrows to empty them into the storage cavities near the winter nests. The seeds of native plants are gathered over a con- siderable area. Sometimes a quart or more is found in a storage chamber, and at the edge of a field where an abundance of grain can be rapidly gathered the winter's stores assume much larger proportions. Economic status. — In spring the planted seed is dug up and eaten or stored from the time it is sown until long after it has sprouted. Then the green stalks are eaten during the early summer, and as soon as the grain is headed out great numbers of the heads are cut off for the young kernels, from the very beginning of their formation. Thus, before harvest time the edges of the grainfields have become ragged and thin for a considerable distance into the field. Although depredations of these ground squirrels do not com- pare with those of the more abundant flickertails, their wide distribu- tion over North Dakota and many other States renders them one of the most serious of rodent pests. But for their natural enemies, which are legion, it would be impossible to raise crops within their territory. They are con- stantly preyed upon by many species of hawks, and some owls, and by foxes, weasels, skunks, and badgers, so that in spite of their rapid increase their numbers are usually kept somewhat within bounds. However, it is necessary over much of their range to supplement the work of their natural enemies by the systematic use of poison. Citellus tridecemlineatus pallidus (Allen) Pale Striped Ground Squirrel; Pale Thirteen-lined Ground Squirrel Alhii'wakao of the Cheyennes [Spermophilus tridecemlineatus] var. pallidus Allen, Monogr. North Amer. Rodentia, p. 873, 1ST7. Type locality. — Plains of Yellowstone River, Mont. General characters. — A pale western form of the thirteen-lined ground squirrel, slightly smaller, and with paler tones of buff and lighter brown stripes. Average specimens from the type region measure in total length, approximately 255 millimeters ; tail, 82 ; hind foot, 34. Distribution and habitat. — The striped ground squirrels become gradually paler across the middle part of the State, but not until the semiarid Badlands country is reached west of the Missouri 1926] MAMMALS OF NORTH DAKOTA 55 do the pale forms become clearly recognizable. In the part of the State west and south of the Missouri, they are the only ground squirrels, and here with the prairie dogs they occupy the short-grass plains country in considerable numbers. While sometimes seen in the open, where there is not sufficient grass to conceal them, they are more often found in the better cover of grass and weeds and low bushes. In this region they were originally one of the interest- ing and harmless forms of native life, but since grain farms have spread over it they have become one of the serious problems wil"h which the farmer has to contend. General habits. — In habits these squirrels do not differ from their darker relatives to the eastward, except as a change of environment gives them other kinds of food and local conditions which they seem always ready to meet. In many places some protection is sought for their burrows, such as grassy spots or weedy ground. Sometimes a piece of paper or cloth is drawn over the entrance to the burrow, apparently for concealment or protection. At Parkin, on" June 28, 1916, a burrow was found where fresh earth had been lately thrown out and the entrance was se- curely packed with sand from the inside. As the entrance to this burrow was opened a half-grown young of the species poked its head out of another entrance near by. In the tunnel, about 8 inches below the surface of the ground, was found a large, soft nest in a roomy chamber, with two doors opening out on opposite sides. The nest was made of dry grass, bark fibers, and bits of paper from the railroad track. It was soft and well matted together like a bird's nest, but not covered over. The young had escaped in the branching burrows. Evidently this was their home nest, from which they had not yet begun to make excursions to the world above. The closing of their doors from within was evidently in this case to protect the young from outside enemies. Economic status. — In many places it has been found necessary to poison these squirrels for the protection of grainfields and garden crops; the methods given for the Richardson ground squirrel, or flickertail, will be found to apply equally well to this species. Citellus franklinii (Sabine) Gray Ground Squirrel; Franklin Ground Squirrel Arctomys franldinii Sabine, Trans. Linn. Soc. London, vol. 13, p. 5S7, 1822. Type locality. — Carlton House, Saskatchewan, Canada. General characters. — Largest of the ground squirrels of this region; some- times mistaken for the gray tree squirrel, which it approaches in size and slightly resembles, but from which it differs in slender form, very short ears, and mueh smaller and less bushy tail. Color, dark gray with a brownish wash and a mottled effect in fine, wavy cross lines or scallops over the back. Adults measure in total length 3S8 millimeters ; tail, 136 ; hind foot, 55. Distribution and habitat. — Extending over a wide range in the cen- tral United States and Canada, from Oklahoma and Illinois to the Athabaska River, the large gray ground squirrels cover approxi- mately the eastern half of North Dakota (fig. 3) . Their greatest abun- dance within the State lies within the Red River Valley and west- ward to the- Dakota River Valley, Devils Lake, and the Mouse River. There is an indefinite record for Burleigh County, near Bismarck, 56 NORTH AMERICAN FAUNA [No. 49 and another for Turtle Lake in McLean County, but the most west- ward authentic record is from Kenmare, in the valley of the Riviere des Lacs, where W. B. Bell collected a specimen in 1913. They are particularly animals of open timber and brush land and do not occupy wide stretches of prairie unless there is ample cover for con- cealment. General habits. — Although occasionally - seen up among the branches of low trees, the Franklin squirrels are strictly ground squirrels, living in burrows generally concealed in brush or weed patches, from which weii-Yv*orn trails or runways radiate to other burrows or feeding grounds. They are shy and secretive, keeping much under cover of protecting vegetation, as they are too large and dark colored to be inconspicuous in the open. When frightened they rush for their burrows, usually uttering a trill of alarm and warning to other members of the family. Their voice is much like Fig. 3. — Records of the Franklin ground, squirrel in North Dakota that of the thirteen-lined ground squirrel but is as much heavier as they are larger. It is often heard in a long bubbling trill from a weed patch and is almost birdlike in musical quality. In the timber and brush patches along the Red River Valley, about Stump Lake, Devils Lake, the Sweetwater Lakes, and in the Turtle Mountains, the squirrels are especially numerous and in such situations they are generally the most abundant of the three species of ground squirrel occupying the general region. Throughout the Turtle Mountains they were found along the edges of meadows, fields, and clearing's along roadsides, and in all the open places where woods and small brush served for cover. They gathered around camps or dwellings where there were no dogs or guns and even came into the writer's cabija and helped themselves from the grub box. They persisted in getting into traps set for others long after enough of them had been secured for specimens and most of the trails and runways attributed to other animals proved to belong to them. 1926] MAMMALS OF NORTH DAKOTA 57 Their burrows were generally in groups of three or four, or more, not far apart and evidently connected below ground. They were in all sorts of situations, but a sloping bank, brush heap, old log, or stone pile usually provided the protection sought for their dens. A considerable quantity of earth is usually thrown out in front of one of the burrows but others open out with less conspicuous markings. Many old dens and burrows are located through the brush and woods and one seems always to be convenient when danger approaches. Often the animals will stop at the entrances of their burrows and straighten up in the picket-pin attitude, to make sure whether an enemy is pursuing. If approached too closely, they quickly dive into their burrows with a flirt of the tail and a parting chatter, but if all is quiet they soon reappear cautiously to reconnoiter. Franklin squirrels are easily tamed and make interesting, though rather mischievous, pets. H. V. Williams, at Grafton, had a tame one for which he made a den by burying a box underground. The squirrel carried about a half bushel of grain into this box, and in fall hibernated as usual. When examined in January it was un- conscious, but before its awakening time in spring water ran into the box and it was drowned. While collecting specimens at Fish Lake in the Turtle Mountains, Williams fed one around his tent until it became so tame as to take food from his hand and come to the tent regularly at meal times. It finally became so bold that it would enter the tent and search through the baggage for food. After breaking and carrying off a lot of birds' eggs that had been collected for specimens it had to be killed to prevent further trouble. Hibernation. — With the first freezing weather in fall, usually in September, Franklin squirrels go to their nests deep underground and usually do not reappear until the following April. Before en- tering upon their hibernation they become very fat and depend upon this concentrated form of nutriment to carry them through the win- ter rather than upon the ample stores of food laid up in convenient chambers near their nests. Just when these stores are eaten is not well known, but probably before the squirrels have become entirely inactive in fall and again before the outside food supply is available in spring. Breeding habits. — Their half dozen young are usually born in May or June and by the last of July are half-grown squirrels, out of the burrows, and hunting for their food. Food habits. — Living largely upon nuts, seeds, and grain, these squirrels show an appetite for a wide range of food. The examina- tion of a large number of stomachs and cheek pouches shows their food to consist not only of a great variety of grains and seeds, but also of berries, green vegetation, roots and bulbs, beetles, caterpil- lars, grasshoppers, crickets, ants, and eggs and pupae of insects. They also eat young birds, birds' eggs, and young mice, and are said to kill young chickens. When caught in traps or found dead they are even eaten by their own kind. They feed upon grain from the time the seed is planted until the last bundle is removed from the fields. Unlike the smaller ground squirrels, they do not cut che standing grain, but pull down the heads and in this way destroy the grain even more rapidly. In their capacious cheek pouches seeds of 58 XOETH AMEEICAlSr FAUK'A [No. 49 grain are rapidly carried to their winter storehouses. "Where a large number of the squirrels gather along the edge of a field they will often harvest considerable of the grain after having fed upon it dur- ing every stage of its growth through the summer. Economic status. — To a great extent the Franklin squirrels occupy the limited areas where the other two ground squirrels of the State, the thirteen-lined and the Eichardson, are absent or less numerous. In extensive areas, therefore, they are the dominant species and levy their toll of destruction on the grainfields and gardens that other- wise would be comparatively safe. In some places, however, the three species occupy the same ground and in combined numbers cause enormous losses of crops. Although larger and according to their numbers possibly more destructive to grain than the Eichardson squirrels, the Franklin ground squirrels are apparently less numerous in most of their habitat. They are easily poisoned and their abun- dance may be controlled at comparatively little expense, using the same methods as recommended for the Eichardson, or flickertail. Citellus richardsonii (Sabine) Richardson Ground Squirrel; Flickertail SotiKOta of the Arikaras ; Pinsa of the Dakotas ; Shopka-sop of the Mandans : Tsipd sopa of the Hidat- sas (all, Gilmore). Arctomys richardsonii Sabine, Trans. Linn. Soc. London, vol. 13, p. 589, 1882. Type locality. — Carlton House, Saskatchewan, Canada. General characters. — A plump little ground squirrel much resembling the prairie dog, but about half the size. Color, rich buffy yellow, darkened over the back with obscure mottling and wavy scallops. Ears, minute ; tail, short. Measurements of average adult : Total length, 237 millimeters ; tail, 73 ; hind foot, 45. Ebner gives the usual weight in fall as 16 to 17^ ounces and in spring as 11 to 13 ounces. Distribution and habitat. — From a wide range over southern Saskatchewan, Alberta, and Montana, Eichardson ground squirrels, or flickertails, cover practically all of Xorth Dakota east and north of the Missouri Eiver (fig. 4). They are absent from most of the immediate valley of the Eed Eiver and the wooded bottoms and tim- bered areas generally, being most abundant over the high open prairie of the central part of the State. For some unknown reason they seem to stop at the Missouri Eiver where the prairie dogs begin, although the ranges of the two species overlap slightly in Montana, where no enmity between them is noticeable. The more humid and fer- tile part of the country was occupied by them long before the great wheatfields spread over their range to supply a new and choice food. Of the three species of ground squirrel in the State, these are by far the most numerous and most destructive. General habits. — Originally the flickertails had a continuous dis- tribution over the prairies in great numbers. On some favorite slopes they were so numerous as to suggest a colonial tendency, but apparently this only showed a preference for certain kinds of ground yielding an abundant food supply. In 1887, when much of the prairie was still unbroken, they were living in their primitive manner on such food as the prairie afforded and doing practically no harm except as grainfields and crops en- 1926] MAMMALS OF NORTH DAKOTA 59 croached upon their original range. Their greatest numbers often appeared to be in the areas of the shortest grass and lowest vegeta- tion, possibly because the grasshoppers and other insect life on which they fed to some extent were most easily obtained there. In places the prairie seemed alive with them and they could be seen scamper- ing about together or standing up like picket-pins, while their shrill whistle was heard on all sides. With each call-note their short little tails are flipped up and down, a farewell twinkle being given as they disappear down the burrow, hence the popular name of " flicker- tail."' In 1887 they were often seen also in the main streets of Devils Lake and Bottineau, which were then in their early stages of con- struction, and in 1916, it was most surprising to find them still occupying vacant lots on the edge of the city of Devils Lake. It was a striking illustration of their tenacity in holding to their Fig. 4. — Records of the Richardson ground squirrel in North Dakota original habitat through years of vigorous but sporadic efforts to destroy them. As soon as the grainfields spread over their range they quickly gathered along the edges to feast on this wonderful new and abundant food. They did not long confine themselves to the edges of the field, however, but went into the middle of large cultivated areas and made their burrows in the plowed ground or in the grow- ing grain. No reliable estimate of their numbers can be obtained, but a general idea of their abundance mav be gained from the statements of Elmer T. Judd, of Cando, in a letter of August 1, 1890, in which he says: Au old gentleman here killed 1,500 'gophers,' by actual count, before the first of June. From the first of June to the middle of July, he and a cotton broker from St. Louis, who spends the summer here on his farm, calculated that they killed over 2,500 more. One forenoon they killed 135, as shown by the tails they had captured. 82242°— 26 5 60 NORTH AMERICAN FAtTNA [No. 49 These 4,000 animals were killed on and around the outer edges of one section of land. Breeding habits. — The number of young to a litter is given by Ebner as 6 to 11, with an average of 7 or 8, born in the underground nests mainly in May. By the first of June the young are out of the burrows and find part of their own food while still under the anxious care of their mothers. Small young are occasionally seen much later than the first of June, and apparently the breeding season extends over a considerable period. It has been supposed that flickertails raise two or more litters in a season, but this seems improb- able on account of the brief period between their emerging from hibernation in the latter part of March or early April, and enter- ing hibernation in the latter part of August or early in Septem- ber. This is scant time for even the earliest young to get anywhere near their full growth and lay in sufficient fat to carry them through the winter. By the first of September there are always many in- dividuals that are still small and these are the last to hibernate, presumably because they have not laid up sufficient fat. Even in spring many of those that are seen before the young are born are not nearly full grown and apparently these late young of the previ- ous year are late in breeding. The principal mating season is early in the spring soon after hibernation but sometimes it is as late as the latter part of June. On September 1, 1914, at Bismarck, a few were seen but these were the young of the year, the adults having already gone into their winter dens. At the same time, Silver, who had been studying them at Garrison, for the previous week reported only young of the year caught. At Van Hook on October 16, 1919, the writer saw one out on a warm, sunshiny day after a cold wave, but none had been seen before for some time. Food habits. — During the summer much green vegetation is eaten by the flickertails — largely the leaves and stems of grain, grass, and a great variety of succulent plants — and apparently it would be possible for these rodents, like the prairie dogs, to subsist entirely upon such vegetation were no grain and seeds available. Late in summer and in fall, when the seeds of the prairie plants and grasses begin to ripen, they constitute the principal food of the squirrels. An important part of the summer food consists also of such insects as grasshoppers, crickets, and caterpillars, though these vary greatly with season and locality. At Crosby, in July, 1913, they were found feeding extensively on the little juicy striped-backed army- worm caterpillars, which swarmed over the roads and fields. Some of the squirrels examined had their stomachs half full and others entirely filled with the caterpillars. Where grasshoppers are abundant they are often fed upon extensively, but wherever grain can be obtained it seems to be the favorite food. One flickertail, shot as it ran out from under a shock of grain, had 269 kernels of oats in its cheek pouches. One recorded by Seton had 162 grains of oats in its pouches and another 240 grains of wheat and nearly a thousand grains of wild buckwheat. Their cheek pouches are so capacious that when well filled they often make the head appear more than double its natural size. The stores gathered are rapidly carried home to be deposited in the burrows and large quantities of food are thus provided for future use. No 1926] MAMMALS OF NORTH DAKOTA 61 stores of grain have been found in the hibernating dens, however, and more study is needed to show when it is used. Destruction of crops. — The annual loss in grain crops in North Dakota occasioned by these ground squirrels has been estimated at $6,000,000 to $9,000,000 in addition to the annual expenditure of at least $100,000 of public and private funds to combat their. depreda- tions. Their tendency is to multiply rapidly in a well-settled and cultivated part of the country because many of their natural enemies are destroyed or kept at a distance, and the food supply is most abundant. As soon as they emerge from hibernation early in spring they begin digging up the seed and eating the young grain that has been sown in the fall, and as soon as the spring sowing starts they dig up the new seed and eat or carry it away. When the grain sprouts they dig both sprout and kernel, and after the kernels are entirely exhausted they feast on the young growing grain until it is headed out, when they begin on the young heads, cutting down the stalks and eating the young seed through all its gi owing stages. As soon as the grain is ripe they carry it away as rapidly as possible to their storehouses, and this is continued until the last bundle is removed from the fields. Four thousand of these squirrels on or around the edges of a section of land would remove a considerable portion of the crop, and it is not surprising that they are considered the greatest pest of the region. They seem to have no preference between wheat, rye, barley, oats, or flax, but take whatever is nearest their dens. Natural enemies. — The natural enemies of these ground squirrels are numerous, and but for them the abundance of the animals would be many times greater. Badgers are constantly digging them out and feasting upon them, from early spring until long after they have hibernated or until the ground becomes well frozen and the badgers themselves go into winter quarters. The long-tailed weasels enter their burrows and kill and feed upon them without the least trouble or hindrance and apparently destroy great numbers besides those merely killed for food. At the first appearance of one of these weasels, the squirrels give frantic alarm calls that set the whole prairie community in a panic. They rush to their burrows, but the weasel follows and helps itself to as many as it cares to kill for food or pleasure. This goes on as long as the burrows are open and probably even during the winter, when the weasels can gain access to the dens through the snow, as they are active all winter and sleeping squirrels fill their needs as well as any others. Skunks probably dig out a few, and foxes, coyotes, and bobcats help also to reduce their numbers. Hawks and some owls prey upon them to a greater or less extent. The ferruginous rough-legged hawk apparently feeds upon them almost exclusively where they occur in its neighborhood and brings them in to feed its hungry broods. The Swainson, marsh, red-tailed, and red-shouldered hawks feed on them extensively, and even the bird-catching sharp-shinned and Cooper hawks may occasionally take one. The little sparrow hawks, which feed mainly upon grass- hoppers, probably destroy some of the young ground squirrels, and it is likely that both the short-eared and long-eared owls capture many of them during early evenings or on cloudy days. Gopher 62 NORTH AMERICAN FAUNA [No. 49 snakes feed upon them to considerable extent, but few data are available in regard to some of the most important species of snakes. The protection of such of their natural enemies as are not otherwise harmful in habits is one of the most important measures for the control of these ground squirrels. Methods of destruction. — Most efficient methods of controlling these ground squirrels have been carefully worked out by members of the Biological Survey and the North Dakota Agricultural College and Experiment Station. In campaigns against these squirrels, the most economical preparation of poison that has been found to be effective is grain lightly coated with strychnine and starch in the proportions of 1 ounce of strychnine alkaloid to 1 tablespoonful of gloss starch made into a paste with 1 pint of boiling water and stirred into 20 quarts of oats. A teaspoonful of this coated grain placed near each occupied burrow disposes of a large percentage of the squirrels at the first application and the few that remain can be practically cleaned up at the second application. Well-organized and coordinated work over a large area is necessary for satisfactory results, as no matter how thoroughly the squirrels are cleaned out from one or a half dozen farms they will quickly reinfest the whole area from those remaining. This preparation of poisoned grain is equally successful with the other species of ground squirrels and chipmunks where it is necessary to reduce their numbers or clean them out of a section of country. Ground squirrels as pets. — On a street car from Devils Lake to the Chautauqua Grounds one day the writer saw a boy who had one of these squirrels, which he had caught with a snare earlier in the day. It was about half grown and had become so gentle that he was playing with it and handling it freely, letting it climb up his coat sleeve and carrying it in his pocket or in his cap on his head. It made no attempt either to escape or to bite, but snuggled up to him in a way that suggested the possibility of using these squirrels as pets for children, a vital need that is not well met by any of our domestic animals. Cleaner, neater little pets could not be found. Although quiet in disposition, they have sufficient vivacity to be very attractive. If taken young and well tamed these native squirrels would certainly be far more attractive, interesting, and in- telligent than white mice, rats, or guinea pigs, which seem to be the only small mammals available for this important phase of child development. The supply would be endless and easily obtained, and by using only one sex in one part of the country any danger from recolonization would be avoided. Cynomys ludovicianus ludovicianus (Ord) Black-tailed Prairie Dog Plnspi"sa of the Dakotas ; Achks of the Arikaras ; Shopka of the Man- dans ; Sinhpa or Tsipd of the Hidatsas (all, Gilmore). Arctomys Hdoviciana Ord, Guthrie's Geogr., 2d Amer. ed., vol. 2, pp. 292, 302, 1815. (Reprint by S. N. Rhoads, 1894.) Type locality. — Upper Missouri River, where discovered by Lewis and Clark. General characters. — Prairie dogs might be described as big, husky ground squirrels or little, plump woodchucks, to both of which they are related and 1926] MAMMALS OF NOETH DAKOTA 63 between which they range in size. Although belonging to the squirrel family, they are compactly built for digging and for life on and under the surface of the ground. The ears are minute, the tail short, and the legs short and muscular. The color generally matches well the fresh yellow earth of their burrows, being a yellowish or pinkish cinnamon above and buffy below ; the tip of the tail is blackish, and coarse black hairs are scattered over the upper parts ; the fur is soft and silky in winter, coarse and harsh in sum- mer. Average measurements : Total length, 388 millimeters ; tail, 86 ; hind foot, 62.10 Weight, 2 to 3 pounds. Distribution and habitat. — From a wide range over the Great Plains from western Texas to northern Montana, these prairie dogs extend over that part of North Dakota west of the Missouri River (fig. 5). In this latitude they are all west of the Missouri River, but farther south they occur on both sides. Fortunately they are colonial in habits and have a scattered distribution, so that the Pig. 5. — Distribution of prairie-dog towns in North Dakota country is not fully occupied by them, but the colonies, or " dog towns," have been numerous over the part of the State which they occupy. In 1910, Anthony reported a few prairie dogs on the south side of the river not far from Buford, and many 20 miles south of there. In 1909, prairie-dog towns were reliably reported near Mannhaven, just west of the Missouri River, and on the Little Missouri near Marmarth in the southwest corner of the State. In 1913 there was a considerable dog town east of Sentinel Butte. In 1913, Jewett reported a large colony on the flats about a mile west of Fort Clark, where the prairie dogs were doing considerable damage to crops, another colony on a piece of level prairie about 3 miles east of Oakdale, and many others along the Little Missouri River from Quinion to Medora, with exceptionally large Colonies at the mouth of Ash Creek and near the head of Magpie Creek. Most of the dog towns he found around Sentinel Butte had been destroyed, but a small colony still existed about 10 10 Measurements from North American Fauna No. 40 (Hollister, 1916, pp. 16-17). 64 NORTH AMERICAN FAUNA [No. 49 miles east of there. A considerable dog town was located a couple of miles east of Medora and another along the Northern Pacific Rail- road between Hebron and Glen Ullin. Kellogg, in 1915, found near Goodall an uninhabited dog town that had covered about 400 acres. A small colony on the west side of the river opposite Elbowoods was said to be decreasing in population. About a mile north of Mannhaven a colony was found covering about 100 acres. At Stanton there had formerly been a large colony but it had been de- stroyed by poison. In 1915, Sheldon reported a small prairie-dog town on Deep Creek near the former Dakota National Forest, and other colonies scattered over that general region. At a point about 4 miles northwest of Cannon Ball, he located a town containing about 2,000 prairie dogs and covering an area of approximately 160 acres. Another colony was located near old Fort Rice, covering about 40 acres and containing about 500 animals; still another about 9 miles south of Cannon Ball of approximately 80 acres and about 500 animals. He was told that the Indians had kept them down by shooting them for food. Near Wade, in 1913, Doctor Bell reported them as occurring in scattered colonies. In 1915, U. S. Ebner, in charge of field operations in the rodent- control work of the Biological Survey in cooperation with the North Dakota Agricultural College and Experiment Station, investigated the prairie-dog situation over a part of the range west of the Mis- souri River. He reported small prairie-dog towns covering 25 to 250 acres scattered along the Little Missouri River in Billings County, larger colonies of 60 to 640 acres in the northern part of Dunn County, a number of towns of 20 to 160 acres along Big Beaver Creek in the northern part of Golden Valley County, other towns of 25 to 500 acres in the eastern part of McKenzie County, and some large towns running as high as 600 acres on the Berthold Indian Reservation. In most of these prairie-dog towns he estimated 20 to 40 burrows to the acre. Although these records show only the colonies that have been located, they indicate a very general distribution of prairie dogs over this part of the State, and a careful survey would doubtless show a surprising number of inhabited prairie-dog towns in a region that is rapidly filling up with grainfields. As a general thing the colonies are located on the open level prairie and often on the best of the grain land. In the Badlands they are usually on the fiats and level spaces where the best grass grows, always away from the brushy and barren areas. General habits. — Prairie dogs are highly social in disposition, almost invariably living in colonies. On rare occasions a new lo- cation is chosen where a family or a few prairie dogs have started a colony, but generally there is evidence of their long residence. The old burrows and mounds remain for many years and the sites of ancient prairie-dog towns are marked by little swells of grassy turf scattered over the prairie. A well-populated prairie-dog town on a bright summer morning is as animated as any busy village could well be. At the first ap- pearance of the sun the animals come out of their burrows and begin their breakfasts of grass and roots, most of them busily digging up grass and little plants for food, nibbling off the grass blades and 1926] MAMMALS OF NORTH DAKOTA 65 plant leaves like rabbits, or sitting up holding them in their hands like squirrels. There are always, however, a few on sentinel duty, usually sitting straight up on the highest mounds, or stretching up occasionally to full height from the grass where they are feeding. Some are always scampering from one point to another, and when the young are out there is much playing and scuffling among them. A populous town of prairie dogs, all busy and many of them call- ing back and forth, with a few on sentinel duty, barking in steady little yap-yap-yap-yaps at some real or imaginary enemy, makes an interesting picture. If the enemy really approaches, the barking becomes frantic and is taken up by other members along the line, and there is a general scamper for the nearest burrows. If one walks toward them to within rifle range the panic increases and the nearest animals rapidly disappear clown the burrows with a farewell twinkle of their tails. The barking passes along farther and farther through the town, usually beyond where the enemy can be seen, every prairie dog taking notice and most of them joining in the alarm. Occasionally one of the guards will stretch up to its utmost height and throwing its head back utter a long C hu-r-r-r-r-r-r-r-r-r-r, as if a dozen barks were crowded into one. This seems to be their only note besides the regular yap-yap, and a chuckling, scolding Chu-r-r-r-r-r, after entering their burrows, as if they were grumbling at having been disturbed. The burrows are deep and go down at steep angles, sometimes almost straight down, for 2 or 3 feet and then slope off gradually. A pebble dropped into one can be heard rolling and bounding down, often for 5 or 6 feet, and a prairie dog with a string tied to its hind foot will sometimes take down 12 or 15 feet of string be- fore reaching the end of the tunnel. The burrows are simple and almost never lead out to a second opening. The nest, instead of being at the lowest point, is usually in a chamber well protected from any rain water that may run down the burrow. As a further protection the earth thrown out is carefully placed around the entrance to form a craterlike rim that serves the double purpose of a watch tower and a dike to prevent the entrance of water from heavy rains. Originally the mound is built of the earth brought out of the burrow, but later fresh earth is scraped up from outside and brought back and added to the sides, and when the ground is moist after a rain the mound is carefully formed and patted and pushed with the end of the nose until externally it has the most approved slopes and internally the correct funnel form. A well-kept mound shows numerous dents and dimples where pushed and poked with the pudgy noses of the prairie dogs. Many old burrows with neglected and broken mounds are used, but the main nest burrows are always kept in good condition. Nest material of dried grass and soft plant fibers is carried into the burrows and the old material is occasionally brought out and scattered about the entrance. The cheek pouches of the prairie dogs are small and little used, and apparently no food is stored. Breeding. — The 4 to 6 or 8 young are born early in May, but usu- ally do not appear out of the burrows until the first or middle of June. They are then seen in family groups around the entrance to (36 NORTH AMERICAN FAUNA [No. 49 their homes and always under their mother's watchful eye. At a signal from her they quickly rush to the burrow and disappear. As their experience increases they are left more to their own dis- cretion, but even when half grown if danger appears the mother in- sists on their all getting down the burrow before she will enter. Small young are often seen later in the year, but in the northern part of their range it is doubtful if more than one litter is raised in a season, the late young probably being the first litter of females of last year's brood. Hibernation. — In fall the adults become very fat and the young moderately so. They are always ready to hibernate in case of very cold or stormy weather or deep snow, but do not enter their dens to remain unless cold weather comes. In mild seasons they are some- times active until midwinter and may be seen foraging on warm days when there is no snow. In severe winters, however, they dis- appear for a long period and evidently pass completely into the state of hibernation. They are out with the first warm days of spring and in March, when a few sagebrush tops were the only visible vegetation, the writer has seen them sitting on top of 2 feet of snow through which they had burrowed to the surface. As soon as the snow is off in spring) they find plenty of food in the dry grasses and roots, and their store of fat helps to carry them through the mating season. Food habits. — The food of the prairie dogs consists principally of grass, including seeds, leaves, stems, and roots, but it includes also a variety of other plants, generally everything that grows over the surface of the ground to a considerable distance around their bur- rows. The short blades of grasses are not only eaten off to the ground, but the roots also are dug up and the tender bottoms of many species are eagerly eaten. Other little plants are eaten to the ground and those with edible roots or bulbs are dug up and extermi- nated. Often tall plants, grasses, and weeds that have sprung up in the prairie-dog town are cut down, if not for food, to keep the ground clear and the view unobstructed. An old and well-popu- lated prairie-dog town is often so completely cleared of vegetation that parts of it have to be abandoned, the animals moving on toward the best grass on the margins. In this way parts of the prairie are progressively denuded of vegetation. The stomachs of prairie dogs are relatively large, as in all grazing animals, and at any time of the day except early morning they are found well filled with finely masticated vegetation, usually showing a good combination of green and white pulp from the foliage, stems, and roots of plants, often with streaks of color from various kinds of flowers and seeds. Many ripening seeds are included in their food, and fields of grain tempt them to extend their colonies into this unusual food supply. When the dog towns are plowed up and seeded to grain the occupants cling to the old burrows with great tenacity, opening them up and if left undisturbed living in the midst of wide grainfields. Depredations. — An area occupied by a colony of prairie dogs may usually be considered stocked to its carrying capacity and of little or no value for grazing or agricultural purposes. It may also be considered that the area thus occupied is just so much withheld from 1926] MAMMALS OF NORTH DAKOTA 67 other use, and it is only a matter of determining the area of land given over to these animals to know the extent of the loss in grazing. If a well-populated prairie-dog town is plowed and seeded, prairie dogs will be the ones to harvest the grain unless they are first destroyed. Destruction of prairie dogs. — Fortunately prairie dogs are easily poisoned by the use of oats or other grains coated with strychnine, as described for the Richardson ground squirrel, and a farm suffering severe losses may be reclaimed at comparative^ small expense. Full directions for preparing and using the poisons will be furnished by the Biological Survey on request. Fig. 6. — Localities where woodchucks are known in North Dakota Marmota nionax rufescens Howell Rufescent Woodchuck; Groundhog Marmota monax rufescens Howell, Proc. Biol. Soc. Washington, vol. 27, p. 13, 1914. Type locality — Elk River, Minn. General Characters. — Heavy-bodied animals, with short ears, short legs, and short, bushy tails. Similar in general appearance to the southern and eastern woodchucks, but more reddish brown above and below. Upper parts dark brownish gray, sides and underparts strongly washed with reddish or rusty brown: feet blackish; tail black or dark brown, long-haired and bushy. Average measurements: Total length, 548 millimeters; tail, 143; hind foot, s:'..'1 Weight, about 8 to 12 pounds, but individuals have been recorded as heavy as 13% and is pounds. (Anon., 1900; Fellows, 1881.) Distribution and habitat. — From the Transition Zone of the east- ern United States woodchucks extend across Minnesota and into southeastern North Dakota as far as Devils Lake (fig. 6). In revising the group Howell examined specimens from Fargo, Grafton, and Leonard, in North Dakota; and at the biological laboratory in 1913 there were skins collected near Stump Lake and Devils 11 Measurements from North American Fauna No. 37 (Howell, 1915, p. 20). 68 NORTH AMEBICAN FAUNA [No. 49 Lake. At Wahpeton woodchucks are reported common along the banks of the timbered river bottoms. At Fargo and Grafton they are occasionally found. In 1915, Kellogg collected a half- grown young near Larimore and obtained a specimen at Grafton. While at Manvel, Grand Forks County, he saw their burrows and one young that had been captured. In 1919, Williams reported them becoming more numerous each year at Grafton. Eastgate says they are occasionally found in the forest near the biological lab- oratory at Devils Lake, but that they are by no means common. Ap- parently they fill the forested belts along the rivers, extending west- ward from the Red River Valley and thus reaching the Devils Lake and Stump Lake forested tracts. Although mainly restricted to forested and brushy locations, where no timber is available they will live in the open. Steep banks and sidehills are favorite situa- tions, but in many cases the burrows are found on level ground or under stumps, trees, or stones. Woodchucks are not fastidious as to habitat, the one requisite for their existence seeming to be an ample supply of green food during the summer season. General habits. — These largest and least squirrel-like of the squir- rel family have generally the burrowing habits of the ground squirrels and prairie dogs. They are mainly burrowing and graz- ing animals, occupying the region of rich plant growth rather than the short-grass prairie, and depending to a great extent on cover and concealment for protection. They are not colonial in habits, except as mother and young remain together during the season, but if undisturbed they often multiply so rapidly as to be seriously destructive to crops and forage. Their burrows are extensive, and instead of being one simple tunnel, usually open out in two or more directions from the central den. Near Larimore, Kellogg dug out a den where an old female and her half-grown young were living. There were four entrances and three nests in different chambers. The nests were made of dry grass and leaves and one contained some fresh-cut plants, including nettles. The branches of the den were respectively 6, 7, and 9 feet long, but apparently there were other branches not discovered, as the old woodchuck and her young had been seen to enter the burrow, but could not be found. When first seen, one of the young was up in a basswood tree and when alarmed ran down into the burrow. These woodchucks are good climbers and in places where there are no fences or rocks to serve as watch towers, they are often found up in trees where a good view can be had. They also take refuge in trees to escape from dogs and other enemies. A family of seven young is reported from Manvel, by Kellogg, and this seems to be about the average number for the species. Food habits. — The food of woodchucks consists largely of green vegetation, with which their large stomachs are usually filled. They are particularly fond of clover, alfalfa, or any of the native legumi- nous plants, but will eat grass and growing grain and vegetables with great relish. In fall, flowers, seeds, and grain furnish a richer food from which more rapidly to accumulate their winter fat. Ap- parently they do not lay up stores of food, but depend on finding an ample supply until time to hibernate, and in spring live on their store of fat until green vegetation is available. They usually come 1926] MAMMALS OF NORTH DAKOTA 69 out in spring while the ground is still covered with melting snow. Doctor Bell took a specimen at Fargo on April 29, 1906. Economic status. — The fondness of woodchucks for almost every kind of garden or field crop renders them serious pests wherever they are numerous. Here on the border of their range they may never become troublesome, but they should be watched and their numbers kept down wherever an undue increase is noticeable. It is unfortunate that the habits of so many of our native animals conflict with the interests of man, as their presence would otherwise add much to the interest of life. A few woodchucks in the meadow and along the fences, where their loud, shrill whistle is occasionally heard and where they are seen sitting up in the grass or on the fence watching for danger, or with flopping tails scampering to their burrows, would add a touch of life and interest to any land- scape. Their depredations, however, are too serious to be taken lightly, and it is often necessary to destroy them. Usually they may be shot or trapped, or killed with carbon disulphide placed in their burrows, but they are not easily poisoned, as there is usually an ample supply of their favorite food within reach. Woodchucks as food. — Woodchucks have some value as food ani- mals ; their meat is like that of squirrels, but coarser. Many persons are fond of them, and in the markets they sometimes bring as much as a dollar each. There could be no cleaner or more exemplary animal in food habits and their underground dens are as clean and fresh as the abode of any burrowing animal. When necessary to destroy them their use as food should be encouraged. Marmota monax canadensis (Erxleben) Canada Woodchuck; Groundhog1 [Glis] canadensis Erxleben, Syst. Regni Anim., p. 363, 1777. Type locality. — Quebec, Quebec, Canada. General characters. — In size somewhat smaller than rufescens. Color, strong- ly rufescent. Average measurements : Total length, 513 millimeters ; tail, 10S ; hind foot, 76. Distribution and habitat. — From a range extending across Canada from Nova Scotia on the east to Fort Simpson on the Mackenzie, the small northern form of the monax group of woodchucks barely enters the extreme northeastern corner of North Dakota. One specimen collected at Pembina in 1887 was identified by Howell as belonging to this form. It was taken on the North Dakota side of the river, but after it had been seen to swim across from the Minnesota side. In 1915, Kellogg reported a few woodchucks at Pembina, but did not obtain any specimens. The river at this point has considerable timber along both sides, and woodchucks are likely to pass back and forth freely, if not during the summer, they cer- tainly would early in spring before the river breaks up. At Wal- halla they are fairly common, but by October 1, 1919, they had all denned up for winter. One found in the bank and later mounted was evidently the little dark-colored canadensis. General habits. — The only difference in habits between this form and the more southern rufescens may be attributed to climate and environment. With a longer winter, more snow, and a somewhat different set of plants from which to draw their food supply, these 70 NORTH AMERICAN FAUNA [No. 49 rodents readily adapt themselves to local conditions. In the shorter season they are still able to lay up sufficient fat to carry them through the long, cold winters, and in spring they come out of hibernation even while the snow is still deep. As early as March 18, in north- ern Minnesota, they sometimes come out on 4 feet of snow, making tracks in the soft, melting surface or on the frozen crust and visiting back and forth from one burrow to another, opened out through snow tunnels. This is the mating season, and, accord- ing to Seton (1909, vol. 1, p. 426), the four or five young are born about the end of April. Over this great northern country they are generally harmless, except where locally they come in contact with fields and gardens. Family MURIDAE : Old World Rats and Mice Rattus norvegicus (Erxleben) Brown Rat; House Rat; Wharf Rat [Musi norvegicus Erxleben, Syst. Regni Anim., p. 381, 1777. Type locality. — Norway, where introduced in 1762. General characters. — Size, variable ; ears, small ; nose, long and pointed ; tail, long, nearly naked, and minutely scaly ; color, dull brownish-gray above, light or whitish below, occasionally bluish black. Measurements of average adults : Total length, 415 millimeters ; tail, 192 ; hind foot, 43 ; measurements of a large individual, 46S, 212, and 44, respectively. Weight of large individuals, about 1 pound.12 Distribution and habitat. — The familiar house rats are not native to America, but came over on ships about 1775, and since then have spread over most of this country, except some of the arid interior. They follow railroads and settlements into every part of the country where they can find food and cover, preferring the buildings and habitations of man. It is safe to say that they first entered North Dakota with the early steamboat traffic up the Missouri River. In 1833, Maximilian (Wied, 1839-1841, Bd. 2, pp. 72, 251, 25(3- 257, 1841) found them a great pest among the grain stores of the In- dians at Fort Clark. He says that in the loft of the stores of the fort were 600 to 800 bushels of maize that a great number of Nor- way rats assiduously labored to reduce. They were so numerous and troublesome that no kind of provision was safe from their vo- racity, but their favorite food was the maize, among which they cre- ated much havoc, and it was calculated that they devoured 5 bushels, or 250 pounds, daily. The rats were brought thither by American ships, but as yet had not reached the Minnetaree villages. The fol- lowing winter, in the house which had been built for him among the Mandan Indians, Maximilian says: "We were molested during the night by numerous rats and put my little prairie fox in the loft above us, where some maize was kept and here he did excellent service." In 1887, these large rats were abundant at Fort Buford,. which was then the terminus of the Great Northern Railway. The old build- ings about the fort were filled with them and they were very de- structive. Even in the little adobe hotel they were racing about the room every night until caught in traps. At Grand Forks they were said to have only recently arrived and they were not known 13 Weights of more than 2 pounds have been recorded. 1926] MAMMALS OF NOETH DAKOTA 71 at Pembina, Devils Lake, or Bottineau. In 1909 they were com- mon at Bismarck and Mandan and were said to be at Devils Lake and Rugby Junction, but at Bottineau none had been found. In 1912 they were abundant and troublesome about Fargo, Hankinson, Valley City, Lisbon, Stump Lake, and Grafton. At this time a dead rat was seen at a ranch near Marstonmoor. Stutsman County; Wil- liams reported them at Walhalla, in Pembina County ; and within a few years they had begun to infest the country along the eastern edge of the Turtle Mountains. In 1915 Sheldon found them at Fair- mount, where they were a great pest around barns and granaries. He says that the farmers who tried to raise poultry had considerable trouble with them, as they took the little chickens at every oppor- tunity. During his visit at the Hoffman farm, two of the farm hands, while transferring a quantity of hay from one section of the barn to another, killed about 100 rats in a few hours. Most of these were about half -grown, only 1 adult being killed. At Lidgerwood, in Richland County, Sheldon found them less common than at Fair- mount. On a trip west across the rest of the southern part of the State, however, he did not find any further trace of them. In 1915 Kellogg found them at Wahpeton, at Grafton, and a few at Oakes, in Dickey County ; at Towner, McKenzie County, he reported them as not very common. In 1913 Jewett reported that no trace of them could be found at Sentinel Butte and old settlers living there had never seen them. At Medora, also, none were found. It is prob- able that the rats will not find their way to the scattered farms over considerable portions of the western and more arid parts of the State for some years to come, but eventually they will undoubtedly cover practically the whole State. General habits. — So closely have rats been associated with man and his works and for so long a time, that they have become largely parasitic in habits, seeking the cover and protection of buildings and preying upon the food supplies produced and gathered by man. Their sly, filthy habits, mean appearance, and vicious dispositions have not only won the enmity of mankind, but have done much to instill a dislike for other harmless and more attractive native animals with which their name has become associated. To their destruction of property is added the even more serious menace of conveying disease to man. They are by far the most destructive and dangerous of rodent pests and warfare against them should be relentless. Breeding habits. — A large female rat was sent to the Biological Survey from Fargo by K. F. Bascom, who reported that it had con- tained 12 well-developed fetuses. This is not an unusual number of young at a birth, and the rats breed so rapidly that under favorable conditions of food supply and protection the rate of increase is enormous. Litters of young are said to be produced sometimes at intervals of 25 davs and the breeding season lasts for a large part of the year (Lantz, 1909, p. 16). Food habits. — Probably no rodents are more omnivorous than rats. They accept anything oi an edible nature from fresh or stale meat to young chickens, eggs, fruit and vegetables, grain, nuts, seeds, and even green vegetation. They revel in garbage of all sorts and will often find an abundance of food in city dumps, manure piles, and in the refuse about stables. In a grain-producing region their fondness 72 NOBTH AMERICAN FAUNA [No. 49 for grain leads to enormous losses, as where an abundant supply is available they merely take the germ and ruin far more than they require for food. Control measures. — The depredations of these animals are so serious that it is generally found to be good economy to make build- ings rat-proof, or as nearly so as possible, by means of concrete, brick, stone, and wire mesh. Where grain and other food can be kept away from them their numbers can easily be controlled, but they are so skilful in burrowing under walls and gnawing through wood that special methods are necessary to exclude them. So adept are they in avoiding traps and poison that a combination of rat-proofing, poison, and traps is often necessary to prevent serious losses from them. The most successful methods of combating them are given in Biological Survey bulletins and circulars, which are available for free distribution. Mus musculus musculus Linnaeus House Mouse [Mus] musculus Linnaeus, Syst. Nat., ed. 10, t. 1, p. 62, 1758. Type locality. — Sweden. General characters. — Size small, with slender, tapering tail, pointed nose, and rather small ears. Color, brownish-gray above, buft'y-gray below, usually without any clear white. Measurements of average adults: Total length, 160 millimeters ; tail, 81 ; hind foot, 19. Weight of adult female, 23.5 grams. Distribution and habitat. — Troublesome little Old-World mice have become well established over almost every part of North Dakota, in fact through most inhabited parts of North America. So thoroughly have they become dependent on the habitations of man that little is known of their origin and distribution. They followed quickly on the heels of the first settlement of the country and gen- erally appeared within a few years after the establishment of a ranch or farm, even at a considerable distance from other habita- tions. At almost every place over North Dakota where field work has been done by the Biological Survey, these mice have been re- ported as common or abundant and troublesome about buildings, and in many cases they have been caught in the fields in traps set for native species. At Fargo, in 1912, these little mice were so numerous along the edges of fields and roads that it was difficult to catch other species until enough trapping had been done to reduce the numbers of the house mice. Near Williston, in 1913, they were abundant at the ranches and very destructive of grain in the bins and sacks. In the bunk house at one of the ranches the mice kept up a racket all night, and in the morning there were little piles of oat shells on the floor where grain had been brought in and eaten. Their musky odor was very evident in the room, leaving no doubt as to the identity of the species, and one was shot as it ran across the floor. At Kenmare, in the northwestern part of the State, they were abundant both in town and in the weedy bottomlands, where many were caught in traps. At Mandan, in 1913, Jewett found them abundant in town and also on rough slopes in the surrounding fields to a distance of 2 miles from town. At Glen Ullin he caught them on the sides of buttes a mile from town, and also in the tall grass along the creek 1926] MAMMALS OF NORTH DAKOTA 73 bottoms. At Sentinel Butte they were found at almost every ranch ; also at Fort Clark, and around Oakdale in the Killdeer Mountains. Specimens were also taken at Buford, Bismarck, Cannon Ball, and many other localities over the State. General habits. — House mice are generally imported in boxes and loads of goods where they have made their nests, and are carried long distances on trains or in wagons. They prefer the protection of buildings, but when they have become numerous overflow into the surrounding fields and country wherever food and cover are to be obtained. They breed and increase with great rapidity and but for their natural enemies would soon overrun the fields and render agriculture unprofitable. Cats are generally used to keep down their increase, but serve as a very limited check. The native owls, hawks, and weasels, however, do much to control their abundance. In habits the mice are often filthy, running through the dirt of stables and cellars and then over the food in pantries or kitchens, in this way not only destroying food but distributing disease germs. They are so slender that they can slip through cracks and narrow openings into places supposed to be proof against their entry, and they will also gnaw through a considerable thickness of wood to get at food or grain that is stored. Concrete, plaster, and fine wire mesh are the best protection against their inroads, but in spite of all efforts it is often necessary to resort to poison and trap in order to destroy them. Inverted boxes covering poisoned grain, with small openings through which larger animals can not pass, may be kept in buildings where mice occasionally enter and many may be de- stroyed in this way. Directions for preparing poisoned bait, as well as for trapping these pests, will be furnished by the Biological Survey, United States Department of Agriculture, on request. Family CRICETIDAE : White-footed Mice, Harvest Mice, Grasshopper Mice, Wood Rats, and Voles Peromyscus maniculatus osgoodi Mearns Osgood White-footed Mouse (PL 11, fig. 1) Tepa-uti" of the Ornahas (Gilmore) Hesperomys leucopus nebrasccnsis Mearns, Bui. Amer. Mus. Nat. Hist., vol. 2 (1887-1890), pp. 285, 287, 1890 (not of Coues, 1877). Peromyscus maniculatus osgoodi Mearns, Proc. Biol. Soc. Washington, vol. 24, p. 102, 1911. Type locality. — Calf Creek. Custer County. Mont. General characters. — One of the smaller-sized white-footed mice, of rather pale buffy ochraceous color over the upper parts and pure white below ; tail sharply bicolor. Immature individuals are more bluish-gray above, only the adults being buffy ochraceous. Average measurements of adults : Total length, 158 millimeters ; tail 64 ; hind foot, 20. "Weight of adult male, 20.5 grams. Distribution and habitat. — The little native white-footed mice are abundant over the western, drier part of North Dakota east to the Missouri River Valley, thence grading insensibly into the darker eastern form bairdi. The area of intergradation is mainly east of the river, including such localities as Kenmare, Minot, Napoleon, and Linton. In the Missouri Valley the mice seem to be typical Ia Mice that live in the buffalo skulls, most likely this species. 74 NORTH AMERICAN FAUNA [No. 49 of this western form. They are abundant in almost every locality and situation over the areas they inhabit. At Fort Buford, in 1887, they were found abundant, and again in 1913 about equally abundant over the prairies and Badlands buttes, in marshes, and in wooded bottoms. In 1910 Anthony reported them there as numerous in the brush, among the rocks, and on the hills and prairies. At Fort Clark, Jewett caught them in traps set over a wide range of country, but most commonly among rocky buttes and around wheatfields. Along the Little Missouri Eiver from Medora to Quinion he caught them wherever his traps were set, in the willows along the river, in the sagebrush, and on the rocks and hills. At Oakdale, in the Killdeer Mountains, he also caught them in traps set on rocky slopes, in the brush, along the creeks, and in the swamps around springs. At Sentinel Butte they were found on the open prairie, among rocks on the buttes, and in wet grassy places at the edges of ponds. At Glen Ullin he found them very common, liv- ing in burrows and among rocks all over the country. At Mandan he caught them in the brush, along the river, in rocks on the open prairie, and along fences or borders of wheatfields. At Cannon Ball, Sheldon reported them as inhabiting the grainfields princi- pally, but also the arroyos and sandy bluffs. These reports indicate great abundance, a continuous distribution, and perfect adaptation to a great variety of environment. General habits. — These beautiful little animals, with large eyes, long whiskers, and large, expressive ears, show much intelli- gence by adaptation to a great variety of conditions of life, but are nervous and timid and do not readily accept conditions of domestica- tion. In the woods they climb trees and are fond of living in hollow logs or other cavities, but in the Badlands they find safe re- treats among the rocks, cliffs, and clay banks, and on the prairies they live in natural cavities or abandoned burrows of other animals. They probably dig burrows for themselves when necessary, but usually are able to find plenty of those abandoned by pocket gophers and other burrowing rodents. They often live in the driest situa- tions and seem not to be dependent on a permanent water supply, although they have no objection to wet or marshy ground. Strictly nocturnal in habits, as is indicated by their large, dark eyes, they are rarely seen except when disturbed in their diurnal retreats, and although abundant, they are not generally well known. The plow often turns them out of their underground nests and they are fre- quently disturbed when land is cleared ; after haying and harvesting they are found in haycocks or grain shocks that have been standing for some time in the fields. Wherever they take up their abode they quickly make a soft nest of fine plant fibers and seem perfectly at home if shelter and food are obtainable. They are often so numer- ous as to be very troublesome to the naturalist in search of rare specimens, as they fill his traps night after night until they have been thinned out. Near the mouth of the Cannonball Eiver one evening while it was still light enough to see fairly well, a brown-backed old Peromyseus ran out from under a stone. It darted about nimbly from one stone to another, then stood still and watched for half a minute, its big ears and bright eyes giving it a very animated expression. This 1926] MAMMALS OP NORTH DAKOTA 75 is one of the few times when these little mice have been seen out foraging of their own accord before daylight was entirely gone. They will often run over a person, however, while he is sleeping on the ground, and there is generally plenty of evidence of their pres- ence about camp in the morning. In winter on soft snow their tracks may be found leading from tree to tree, or bush to bush, or from one weed to another where they have run in search of food, but most of their tracks lead to or from holes in the snow which connect with tunnels under the snow or cavities under ground. At Mandan and Cannon Ball late in October, 1919, the writer followed many of their tracks to nest cavities in the ground, and dug down and caught the mice in the hands. All were in old stump holes, where cottonwoods had decayed and left rotten wood or hollow spaces deep in the mellow soil of the forested bottomland. At a depth of 6 inches to a foot below the surface, nests of soft leaves and plant fibers were found, lined with cottonwood cotton and rabbit fur, and in these nests from one to four of the mice were comfortably housed for the winter. When disturbed they came out to see what was the matter and they were tied in a handkerchief or gloves and carried home for further study. Even when the temperature was —15° F. and the snow 11 inches deep the mice were out making long lines of tracks at night, in following which much was learned of their food and other habits. They seemed to know where to go directly to every seed-laden tree, vine, bush or weed, and whether to climb up or dig down to get the seeds or fruit. Breeding habits. — The females usually bring forth four to six young at a litter and they apparently breed several times during the summer. Their increase is rapid, and but for numerous enemies their abundance would be far greater than at present. Food habits — The white-footed mice are dainty feeders. The contents of many stomachs examined show a mass of clean white material so carefully selected and finely masticated that there was no trace of shells or hard parts to show from what kinds of seed it came. Most of their food is of various seeds and grain, although sometimes a bit of green vegetation, some bright-colored flowers or berries, or a few insects are eaten. At Mandan the mice were feed- ing largely on the bullberries, which they gathered nightly from the well-laden bushes, apparently eating both the sweet pulp and seed kernels. There were bits of scarlet skins scattered over the snow and the mouse pellets neatly deposited in a cavity not far from the nest were mostly colored dull scarlet by the berries, while some of the seeds were found in mouse caches. From one cache near the nest a handful of seeds was saved and brought back for identifica- tion. Among them were seeds of chokecherry, woodbine, wild grape, smilax, buffaloberry, hosackia, dogwood, bindweed, two spe- cies of knotweed, two of pigweed, ragweed, Russian thistle, black henbane, sedge, barnyard grass, and dropseed grass. The mice seem fond of any kind of camp food, as flour, meal, oatmeal, grain, meat, butter, bread, or crackers. Rolled oats generally make the most attractive bait with which they can be tempted into traps. They are active throughout the year and do not put on fat to carry them 82242°— 20 6 76 NORTH AMERICAN FAUNA [No. 49 through the winter, but instead store up a limited supply of seeds and grain for winter use or for bad weather when they can not come out and run over the surface of the snow in search of food. Economic status. — The small toll these mice take from grainfields would not in itself cause very serious loss, but added to that of many other species the constant drain on farm products is sometimes serious. They cut some grass in the meadows and eat the seeds of many grasses, thus, to some extent, retarding the forage reproduction and in places taking away so much seed as to form a serious check on the repro- duction of other vegetation. Probably more than any other animals they check reforestation, whether this depends upon naturally or artificially sown seeds. So small, so numerous, and so widely dis- tributed are they that they are not easily controlled, except by their natural enemies, which are numerous. They are favorite prey of all small owls and even of many of the larger owls, and form an important article of diet for weasels, skunks, badgers, foxes, and such of the other small predatory species as occur within their range. Seasonable protection of the species that prey upon them, especially the owls, forms the simplest and most effective means of keeping down their abundance. Peromyscus maniculatus bairdii (Hoy and Kennicott) Baird White-footed Mouse Mus bairdii Hoy and Kennicott, Rpt. Comr. Patents. [U. S.] 1856, p. 92, 1857. Type locality. — Bloomington, McLean County, 111. General characters. — About the size of osgoodi, but colors much darker, often dusky along the back, and less buffy or ochraceous; underparts, white. Measurements of an average specimen : Total length, 150 millimeters ; tail, 60; hind foot, 19. Weight of adult male, from Fargo, 18.5 grams. Distribution and habitat. — From Ohio and Oklahoma the little dark-colored white-footed mice (bairdii) extend over the eastern half of North Dakota and into southern Manitoba. In their typical form they do not reach west of the Missouri River, but at about the one-hundredth meridian they grade insensibly into the paler, more buffy osgoodi. There are specimens from almost every locality where collecting has been done in eastern North Dakota, from Hankinson to Pembina and westward to Linton, Towner, and Kenmare. They are found on the tall-grass prairies in the area of humidity and ample cover, where their dark color is protective in the grassy and weedy shadows. General habits. — These little mice, like their western form, osgoodi, are the most abundant and generally distributed mammals of their region. They live in a great variety of situations, from brushy weedy bottoms in the woods, half -dried tule marshes, and dead-weed rows along the roadsides to the middle of grainfields and out over the wide, open, grassy prairie, making nests and homes in hollow logs or trees, in underground cavities which are found, or if necessary, excavated, or under any cover that will offer a dry bed. From osgoodi of the drier, more open plains farther west, they differ in habits only in adaptation to more abundant plant growth and more nearly continuous grainfields. Where food is plentiful they congregate in great numbers, but where it is scanty they become scarce. At Stump Lake, in a line North American Fauna No. 49, U. S. Dept. Agr., Biological Survey. PLATE I I Fig. I. — Osgood White-footed Mouse (Peromyscus maniculatus osgoodi) in Captivity Slightly reduced Fig. 2.— Bean Mouse (Microtus pennsylvanicus wahema) Captive taken with store of ground beans near Cannon Ball. About two-thirds natural size Fig. 3.— Richardson Kangaroo Rat (Perodipus montanus RICHARDSONl) North American Fauna No. 49, U. S. Dept. Agr., Biological Survey. PLATE 12 Skins of Long-tailed Mice (A) Kansas pocket mouse (Perognathus hispidus paradoxus); (B) prairie jump- ing mouse (Zapus hudsonius campestris); (C) prairie harvest mouse (Bei- thTodoniomys megalotis dychei). Half natural size 1926] MAMMALS OF NORTH DAKOTA 77 of traps set along the sandy beach, 16 of these mice were caught in one night where they were finding a choice food supply in the cockleburs which covered the sandy ground. At Crosby the writer found them feeding largely on little caterpillars. At Valley City many were caught in the rows of tumbleweeds or Russian thistles along the fences, where they were feeding on the seeds of these weeds under the cover of which they found ample protection. Eastgate reported 19 caught in his line of 41 traps the first morning after he camped near this place. They are often found in the haycocks and wheat shocks in the fields, and if these are left for a consider- able time the mice are sure to make their nests in them and do more or less mischief. In breeding and food habits these mice are essentially the same as the western osgoodi. Their injury to crops is somewhat greater because of the more general cultivation of land over their range. Their enemies are practically the same and if only given a fair opportunity will keep down the too rapid increase of these interest- ing but destructive little rodents. At Fargo in the well-cleared city parks not a mouse could be caught, while in the uncleared woods near by, where weeds and bushes protected them from the little owls, these mice filled the traps the first night. Peromyscus leucopus noveboracensis (Fischer) Northern White-footed Mouse; Deer Mouse [Mus sylvaticus] noveboracensis Fischer, Synop. Mamin., p. 318 (p. 14, 5.), 1829. Type locality. — New York. General characters. — Largest of the white-footed mice of North Dakota, with relatively long tail and dark colors; upper parts of adults, dark buffy or tawny-gray, with more blackish or dusky about face and ankles than in aridulus; underparts and lower half of tail, pure white. Young and immature, bluish gray or plumbeous. Measurements of adult male from Fargo : Total length, 185 millimeters ; tail, 78 ; hind foot, 22. Weight, 27.5 grams. Distribution and habitat. — The northern form of the white-footed mouse is the common deer mouse or woods mouse of the northeastern United States and southeastern Canada from Nova Scotia to eastern North Dakota. It is largely a forest species, rarely found far from forest or brushland and seems not to extend over the open prairie country. Specimens taken at Fargo, Moorhead, and at Manvel, just north of Grand Forks, are fairly typical of this eastern form and certainly referable to it rather than to the paler aridulus of the Mis- souri Valley. They probably have a continuous range along the timber of the Red River Valley, but seem not to be abundant. General habits. — Deer mice are largely forest dwellers, but have a wide range of adaptation and will go anywhere that safe cover and an attractive food supply lead them. From their original homes in hollow trees and logs or in the ground under old trees and stumps they readily follow the rail fences and brushy fence rows around the fields, taking up their quarters under grain shocks, haycocks, or haystacks, or entering new buildings erected in the clearings. They are great climbers and will run up the trunks and branches of tall trees or up vines and through bushes in search of 78 NORTH AMERICAN FAUNA [No. 49 seeds or berries, or over walls and timbers of buildings, with ease and skill. They are strictly nocturnal and very timid, nervous little sprites, with long, sensitive whiskers and large, thin, delicate ears that are constantly changing form and expression, apparently catching the faintest sounds; their large, prominent black eyes are owl-like in their adaptation to the darkness of night. Apparently they can see fairly well in the daytime, but are rarely natural or at ease in the light and lose much of their vivacity and beauty as usually seen when driven out of their diurnal beds or captured and held as unwilling prisoners. They are among the most beautiful and ex- pressive of our small native rodents, to which the unfortunate name of " mouse " is generally applied, and but for their occasional mischief and nocturnal habits might be as interesting and popular as many of our song birds. In fact, they are not without voices, and certain individuals have a fine squeaking trill that might well be called a song. They have many little squeaks and low notes that doubtless mean much to them if little to us. A more common means of communication, however, consists of a rapid tapping with their finger tips on any hard surface or thin material, which produces a sound suggestive of the drumming of minute woodpeckers. These vibrations vary in length and tone and doubtless mean much to them in the way of communication. Breeding habits. — Nests containing young are frequently found under grain shocks or haycocks, or are plowed out of hollows below the surface of the ground. They are usually as soft, well built, and well lined as those of any bird ; and the delicate, naked young are found resting on silk or cotton wool from various plants or on feathers or fur or other equally soft materials provided by the parents. Usually 4 to 6 young are born at a time and apparently several litters are raised each year. The mammae of adult females are six in number, arranged in two posterior or inguinal pairs and a single pair of anterior or pectoral. Often when suddenly dis- turbed the mother rims from the nest with 5 or 6 young, each cling- ing securely to a nipple, as she drags them rapidly to some safe cover. Food habits. — Although the greater part of their food consists of seeds, grain, and nutlets, deer mice also are fond of berries, fruit, and a great variety of such foods as the human species regards as its own and exclusive perquisite. This often leads to trouble, for the little moonlight people get into fields, gardens, granaries, and even cellars and pantries and help themselves, always to the best there is to be had. The fact that they consume large quantities of seeds of noxious weeds is generally overlooked and some easy method of lessening their abundance is sought. Unfortunately, this often takes the form of keeping cats, which may scare some of the mice away, while the cats live largely on song birds. If little owls could be kept instead, there would be no more trouble from the mice. In fact, there are usually enough little owls to keep down the abundance of the mice, where brush, weeds, and rubbish are removed so the mice will have no protecting cover. 1926] MAMMALS OF NORTH DAKOTA 79 Peromyscus leucopus aridulus Osgood Badlands White-footed Mouse Wiyashpena M [moon nibblers] of the Dakota Indians (Gilmore). Peromyscus leucopus aridulus Osgood, North Arner. Fauna No. 28, p. 122, 1909. Type locality. — Fort Custer, Yellowstone County, Mont. General characters. — A pale buffy western form of tbe northern white-footed mouse quite distinct from Peromyscus osgoodi with which often associated. Differs in larger size, relatively larger ears and longer, less sharply bicolor tail, and in lacking tbe tiny wbite tuft of hair at upper anterior base of ear ; otherwise the color and markings of the two species are practically identical. The young and immature are slaty gray. Measurements of type specimen: Total length, 177 millimeters ; tail, 73 ; hind foot, 22. Weight of adult female, 27 grams. Distribution and habitat. — The Badlands white-footed mice probably have a wide distribution over North Dakota, but are much less numerous than the smaller species and have not been so thor- oughly collected. There are specimens from along the Missouri River Valley at Cannon Ball, Mandan, Sather, Fort Clark, Oakdale, Williston, and Buford, but to the eastward there are no more speci- mens of this group until we find noveboracensis in the Red River Valley. Typical specimens may be expected only from the Missouri Valley and westward. Apparently these are not prairie dwellers, as specimens have been taken only in timbered flats along the streams. General habits. — At the mouth of the Cannonball River they are comparatively common and in August, 1915, Sheldon collected a series of 17 specimens in the forest of the river bottoms. In June of the following year the writer found them common there in the forest and caught them in traps set in thickets and at the bases of hollow cottonwood trees on the river bottoms. In one hollow tree, about 4 feet from the ground, one of the mice was found in a well- made nest lined with the silky down from the cottonwood seeds, and in another hollow cottonwood an old female was caught well up in the cavity of the trunk. At Mandan late in October, 1919, when the ground in the bottomland woods was covered with 11 inches of soft snow, some of these mice were tracked to their nest cavities in hollows where old stumps had decayed. In one of these honey- combed, rotten-wood cavities a nest was found about a foot below the surface of the ground and four of the five occupants were caught as they came out. There were an adult male, two adult females, and two immature of the year in the blue coats. They came out of a nest in one of the side cavities where a root had de- cayed, but they had free access to all parts of the porous wood from deep in the ground to the leaf-covered surface. Curiosity seemed to bring them up to see what was disturbing their home and they were caught and put in handkerchief and gloves and kept M There is some confusion as to which species of mouse this name should apply. Doc- tor Beede gave it as one of the names of the bean mouse (Miorotus p. wahetna, p. 94), but it is not the name in common use by the Dakotas and does not suggest a diurnal, ground-dwelling species, but rather a wholly nocturnal and partly arboreal one. The mice that nibble the edge of the full moon until it is all eaten up must be good climbers, and Doctor Gilmore thinks the name probably applies to one of the white-footed mice. As it is doubtful whether the Dakotas distinguished the very similar forms of Peromyscus, this most arboreal of their species is chosen for the beautiful name. 80 NORTH AMERICAN FAUNA [No. 49 for many months as interesting pets. The nest was a large, soft, warm ball of dry leaves and plant fibers lined with cottonwood cotton and was evidently the home of a family. In captivity they were very friendly and sociable, making a happy family in one nest with four of Peromyscus maniculatus osgoodi. Like other members of the group, Badlands mice are strictly noc- turnal in habits and are rarely seen except as caught in traps for specimens or driven out of their diurnal retreats. When seen by daylight, they are beautiful little animals with beady black eyes, large expressive ears, and long trembling mustaches, which give them a keen and animated expression. They are quick and agile in habits, running with long leaps, and climbing rapidly and skilfully over the trunks and branches of trees. In fall they do not become very fat, but lay up supplies of winter food and continue active throughout the coldest weather. Their delicate lines of tracks may often be seen from tree to tree, or from some old log to a stump or brush heap, or centering around a hole in the snow through which they have access to the surface of the ground, where their winter nests and stores are hidden. Over much of their range they are found only in limited numbers, but in certain localities are exceed- ingly numerous and at times become very mischievous around out- buildings and granaries. As they avoid the open country, they are less mischievous in grainfields than are their smaller and more gen- erally distributed relatives. They are not usually considered a serious pest, but locally they add their little to the constant tax of such rodents upon farm crops. Reithrodontomys megalotis dychei Allen Prairie Harvest Mouse (PI. 12) Reithrodontomys dychei Allen, Bui. Amer. Mus. Nat. Hist., vol. 7, p. 120, 1895. Type locality. — Lawrence, Kans. General characters. — A slender little mouse with large ears ; buffy brown upper parts, and white underparts, distinguished from the house mouse, which it somewhat resembles in size and color, by slenderer and not notice- ably tapering tail, and by pure white feet and underparts. From both the house mouse and white-footed mouse it is still better distinguished by a longitudinal groove down the front surface of each upper incisor. Average measurements of adults : Total length, 133 millimeters ; tail, 52 ; hind foot, 16. Distribution and habitat. — From the Upper Sonoran plains and prairie regions of the Central States the little prairie harvest mice come into North Dakota along the Missouri and Dakota River Valleys. There are specimens from Cannon Ball and Fort Clark in the Mis- souri Valley, Ellendale, Ludden, and Oakes in the Dakota Valley, and from farther east at Lidgerwood and Hankinson, and Fargo in the Red River Valley. At Cannon Ball, Sheldon found the mice quite common over the prairie and on sandy flats, but more abundant along the brushy borders of grainfields and even out in the fields. At Fort .Clark, Jewett caught four in traps set near small burrows at the edge of a wheatfield and on the high prairie. At Ellendale, Sheldon found them fairly common in the grass along the fences, in brushy places, and occasionally in wheatfields, and at Lidgerwood he took 1926] MAMMALS OF NORTH DAKOTA 81 two specimens in tules at the edge of the lake. At Hankinson, W. B. Bell found them common in the tumbleweeds along the fences on sandy soil, and at Fargo Murie took four specimens on the grassy river bank at the edge of a field. General habits. — These little harvest mice live mainly on the sur- face of the ground under cover of grass and low vegetation. Their tiny runways may be distinguished from those of meadow mice by being narrower. The harvest mice like the open ground, but must have sufficient cover to protect them from a host of enemies over- head. In places they apparently live in small burrows, but generally their trails seem to terminate at neat little nest balls on the surface or in low bushes and weeds. The nests are rarely found more than 8 or 10 inches from the ground, and more often they are lightly placed on the surface under some ample cover. At Hankinson, a harvest mouse was frightened from a pretty little grass nest in a lock of hay ; the nest was a compact ball of fine grass lined with soft fibers, with a tiny opening at one side for a doorway. Breeding habits. — Usually four to six young are brought forth and cared for in these birdlike nests, but at Oakes, on June 4, East- gate took an old female that contained seven embryos. Appar- ently they breed more than once during the season, and in places where there is abundant food and good cover they sometimes become very numerous. Food habits. — The principal part of the food of harvest mice con- sists of seeds, largely of grasses, which are found cut in small sec- tions and drawn down until the seed-laden tops are within reach. The mice are fond of rolled oats and other grains used for trap bait and their presence in the fields indicates a fondness for the growing grains. They do not become fat in fall and evidently do not hibernate. Economic status. — These little mice cover so small a part of North Dakota that they are of slight economic importance, but in areas where they are widely and abundantly distributed their inroads on the grain and forage production materially help to swell the total of rodent depredations. Although they are so small that any arti- ficial means of combating their mischievous tendencies would be futile, effective check is constantly kept on their overabundance by such predatory birds and mammals as small owls, hawks, and probably crows, jays, magpies, and butcherbirds, as well as by weasels, skunks, and badgers. Onychomys leucogaster leucogaster (Wied) Maximilian Grasshopper Mouse (PL 13, fig. 1) Michtika of the Mandans (Maximil- ian)18; Michtik-tak of the Mandans (Gilmore). Hypudaeus leucogaster Wied, Reise in das Innere Nord-America, Bd. 2, p. 99, 1841. Type locality. — Fort Clark, Oliver County, N. Dak. General characters. — Somewhat resembling the white-footed mice, but rec- ognized at once by larger size, heavier build, short, thick, tapering tails, and 16 This name is merely a general term for mice (George F. Will). 82 NORTH AMERICAN FAUNA [No. 49 smaller ears. Legs also shorter and feet heavier, to harmonize with their entirely ground-dwelling habits. Upper parts dark drab-brown, darkest along the back ; underparts and lower half and tip of tail white ; immature speci- mens, dark slaty gray ; occasional individuals nearly black. Average measure- ments of adults : Total length, 164 millimeters ; tail, 42 ; hind foot, 22. Distribution and habitat. — In his revision of the genus Onychomys, Hollister (1915, p. 434), refers all of the specimens from eastern North Dakota to the typical dark-colored subspecies leucogaster as described by Maximilian, Prince of Wied, from specimens taken by him at Fort Clark in 1833. The species as thus restricted covers little more than the eastern half of North Dakota, reaching slightly into western Minnesota and northeastern South Dakota and northward into southern Manitoba. There are specimens from Fort Clark, Linton, Grace, Devils Lake, Minot, Pembina, Sherbrooke, and Han- kinson. None has been taken in the immediate valley of the Red River nor in the Turtle Mountains, but apparently the species covers the rest of the State either in this dark form or in the paler western form. It is strictly a prairie animal occurring neither in the forest nor the dense thickets, but scattered over the open country in bare and exposed situations as well as under the cover of grass, weeds, and low scattered shrubbery. Though widely distributed it is never very abundant locally. General habits. — These anomalous little rodents, like the badger and other predatory animals, are apparently wanderers, to some extent, scattering out singly to cover their hunting grounds to the best advantage. Traps set at different kinds of burrows and holes in the ground over the prairie, under a variety of conditions, catch them apparently at random. Sometimes a whole family will be caught in a little thicket or weed patch or a few may be caught every night along a weedy fence row, where they evidently are hunting for their nocturnal prey, but there seems to be no specific place to look for them and rarely is there any trail, burrow, or sign found that can be unmistakenly attributed to them. Most of the speci- mens taken are caught by accident in traps set for other species. At Fort Clark, Jewett caught, them at small holes or the deserted burrows of ground squirrels and pocket gophers over the prairie, or along the edges of fields. At Hankinson specimens were taken among the sandy dunes, often on bare sand, but also in the rows of tumbleweeds along the fences. Some were caught in burrows of other animals and some in burrows that may have been made by the mice themselves; others in trails made by scraping with the foot in the sand for a distance of 8 or 10 feet ; like many other species of mice, they will follow such a trail and are easily caught in traps set across it. Often they are caught in old badger holes, where ap- parently they are foraging for insects. While mainly nocturnal they are less strictly so than the white- footed mice, and the writer has seen them running through the weeds in the daytime and on one occasion he shot one about 8 a. m. Gen- erally they are unknown to residents of the country, who probably mistake them for the common white-footed mice. Many are doubt- less thrown out of their burrows by the plow, but no one seems to have recorded anything regarding their habitations or home life. At night their fine, prolonged whistle, almost insectlike in pitch and quality, was often heard around the camps, but nearly the whole 1926] MAMMALS OF NORTH DAKOTA 83 summer of 1887 passed before it was discovered to what form the voice belonged. Hibernation. — In fall these mice become moderately fat, but whether they hibernate in this climate is still a question. Farther south closely related species are caught at all seasons, but in a region well covered with snow for a large part of the winter they would have difficulty in procuring a food supply, as apparently they do not lay up stores or make any provision for winter. Breeding habits. — An old female caught at Grace, July 2, 1912, contained four large embryos, which seems to be the usual number of young. The mammae of the females are arranged in three pairs, two pairs of inguinal and one pair of pectoral, and probably like other species of the genus the young %are occasionally five or six in number. Food habits. — Grasshopper mice are omnivorous in their tastes, readily accepting rolled oats, bread, cake, cheese, seeds, or grain as trap bait, but show a decided preference for animal food, as indicated by the examination of a great number of stomachs. At Fort Clark, the type locality, Jewett took a fine series of specimens, both old and young, in traps set around the edges of wheatfields and baited with fresh meat or bacon. He says some were taken in meal-baited traps, but that they prefer meat. Grasshoppers, crickets, beetles, and a great variety of other insects are found in their stomachs, also often the flesh and hair of other mice which have been caught or found dead. At Hankinson many of the other mice in the traps were eaten and the stomachs of grasshopper mice caught near by often proved that they had been the trap robbers. Economic status. — The only complaint of these mice doing any mischief seems to come from their discoverer, Maximilian (Wied, 1839-1841, Bd. 2, p. 101, 1841), in 1833, at the Mandan villages, where he reported them as common over the prairie and in winter coming into the Indians' houses, where all sorts of stores were kept. He says the Mandans call them " mihtick," as they do all kinds of mice. There is no doubt that Maximilian knew the species, as his descrip- tion is full and perfect, but it is suspected that the mice which did the mischief in the Indian stores were mainly the white- footed species, which also were abundant there. At Fort Berthold, in 1872-73, Doctor McChesney (1878, p. 206) reported them abundant and inhabiting the underground caches of the Indians. Before a definite statement can be made as to the destructiveness of these mice, more complete knowledge of their habits will be necessary. The great numbers of injurious insects eaten by them and their destruction of some other and more troublesome species of mice should class them among the highly beneficial mammals. A more detailed study of their habits is likely to prove of practical value. Onychomys leucogaster missourieusis (Audubon and Bachman) Audubon Grasshopper Mouse Mus missourieusis Audubon and Bachman, Quad. North Amer., vol. 2, p. 327, 1851. Type locality. — Fort Union (now Buford), N. Dak. General characters. — Very similar to leucogaster, but slightly smaller and much paler, the upper parts being bufl'y brown, darker in winter pelage, and Onderparts white; immature specimens, slaty gray. Average measurements of adults : Total length, 150 millimeters ; tail, 39 ; hind foot, 21. 84 NORTH AMERICAN" FAUNA [No. 49 Distribution and habitat. — The paler form of the grasshopper mouse comes into western North Dakota from a wide range over the semiarid plains of Montana, Wyoming, and Saskatchewan. There are specimens from Buford, Dickinson, Glen Ullin, and Cannon Ball. Although few specimens have been taken, the mice undoubtedly cover the whole western part of the State, grading insensibly into the darker-colored leucogaster near the type locality of that species, Fort Clark. They belong to the short-grass prairie or plains of the semi- arid region and seem to be generally distributed over the open country. General habits. — One of these mice brought to Audubon at Fort Union on July 14, 1843, was figured and described in his (Audubon, 1851-1854, vol. 2, p. 327, pi; 20-C, 1851) Quadrupeds of North America. Thus the species was made known and named, but nothing whatever learned of its habits. In 1887, with instructions to make a special study of the habits of this mouse, the writer visited Fort Buford. With such crude collecting traps as were available at that time a considerable number of specimens was obtained, most of them being caught alive in little tin box traps. They were common over the hills and prairies, living in burrows of other small animals, as pocket gophers, ground squirrels, other mice, and even in old badger holes. Some were caught at the burrows of the pale field mouse Microtus pallidus, which they were probably hunting for food. Some of the fresh burrows in which they were caught may have been of their own construction, but probably were the burrows of other mice, to which they were only paying visits in order to capturing prey. The bait first used for them was cheese and dough- nuts, but since then a bit of fresh meat has been found much more attractive. One of the mice caught in a box trap was not quite full grown and seemed so gentle and interesting that a cage was made for it and it was kept for some months. From the first it was not in the least alarmed and when handled never offered to bite nor struggled to escape, although in the cage at times it became frantic in its efforts to get back to its natural haunts. Unless very hungry it would sleep all day, but on waking up in the evening, after stretch- ing and gaping and blinking for a while, would become thoroughly roused and eager to get out and hunt for its supper. It did not like a bright light and would show signs of discomfort by blink- ing its eyes, but, with its box faced away from the light, was very bright and animated. At a touch on the box it would come to the front and eagerly take whatever food was put in between the wires. Any insect put inside was quickly caught, and even flies would rarely escape it. From the trap line the writer always brought back plenty of food for the mouse and greatly enjoyed watching it eat the different kinds of insects. In one forenoon it ate 16 crickets, 11 grasshoppers, 1 spider, 1 black bug, and 1 big fly. Its favorite food seemed to be crickets, and it would never touch anything else while there was a cricket in its box. Next to crickets it liked grasshoppers or flies, but did not seem to care much for beetles, although it would eat any kind offered including some ladybugs and a small black species that was common under sticks and stones, and it seemed to relish a potato bug found North American Fauna No. 49, U. S. Dept. Agr., Biological Survey. Plate 13 North American Fauna No. 49, U. S. Dept. Agr., Biological Survey. PLATE 14 Skins of Short-tailed Mice (A) Pale mouse (Microtus pallidus); (B) little upland mouse (Microtus minor); (C) western upland mouse (Microtus ochrogaster haydenii); (D) Drummond meadow mouse (Microtus pennsylvanicus drummondi); (E) bean mouse (Microtus pennsylvanicus wahema); (F) eastern meadow mouse (Microtus pennsylvanicus pennsylvanicus). About half natural size 1926] MAMMALS OP NORTH DAKOTA 85 for it, as it ate all but the wings and legs. One day it ate 12 crickets and 1 spider in 7 minutes; and a little later, 11 grasshoppers, 4 crickets, 1 black bug, and 1 large fly, making 29 large insects and 1 spider eaten in about 4 hours, and it still seemed hungry. On an- other day the mouse ate 28 crickets, 15 flies, 8 grasshoppers, and 2 beetles, in all 53 insects in less than 12 hours. It seemed to relish a common gray moth and enjoyed a black hornet until it came to the tail, when the stinger evidently pricked its nose. Ants were the only insects it ever refused and a few of these in its box would make it violently frantic. To see it eat a large grasshopper was amusing. The mouse would hold the grasshopper upright between its hands and begin on the head, when a few vigorous kicks of the grasshopper would tip it over backward ; but it would never let go until the head was eaten off and and the body devoured. The wings and legs would drop off as it progressed, and if the mouse was still hungry they would be eaten later. If a number of grasshoppers were put into its box at one time, it would first bite off the heads of all, so that none would es- cape, and then finish them at its leisure. One day the mouse killed and ate a small frog, but did not seem to care much for it. Only when very hungry would it eat seeds and green leaves of plants. It pounced like a cat upon a dead white-footed mouse dropped into its box, caught it beside the head near the ear, and began biting with all the ferocity of a carnivore. The bones could be heard to crack, and when taken out a small hole was found broken through the base of the skull. Its teeth had penetrated well into the brain. The dead mouse was returned to its captor, which began to tear and pull off strips of skin and flesh from the neck, shoulders, and head, and ate both of its eyes. Another mouse that was put into its cage was treated in the same way, and a song sparrow that had been acci- dentally killed the mouse bit through the head and then partially ate. It ate part of a mouse of its own kind, thus proving to be cannibal- istic as well as carnivorous. It was fond of bits of fresh meat and especially of brains that were given it while specimens were being prepared. The fierceness shown in attacking mice indicated a habit of cap- turing and killing these animals in their free state. In spite of its savage disposition and carnivorous tastes, it -was the gentlest rodent to handle, as well as one of the most interesting and attractive pets ; others of closely related species generally show similar dispositions. Breeding habits. — On May 30, 1910, H. E. Anthony caught an old female at Buford, containing four embryos. Little is known, however, of the breeding or other home habits of these mice. Hibernation. — One adult caught at Fort Buford in September was so fat as to suggest preparation for hibernation, but the fat was not distributed in a thick layer under the skin as is the cape in most hibernating mammals. The question of their hibernation in the North is still to be determined. Economic status. — Grasshopper mice make extremelv interesting pets and their field of usefulness seems well worth careful investiga- tion. They might in certain cases be used to advantage in keeping down insect pests in greenhouses and other buildings, and it is probable that their carnivorous propensities would render them valuable in combating other more destructive mice. 86 NORTH AMERICAN FAUNA [No. 49 Neotoma cinerea rupicola Allen Pale Bushy-tailed Wood Rat (PL 13, fig. 2) Neotoma rupicola Allen, Bui. Anier. Mus. Nat. Hist., vol. 6, p. 323, 1894. Type locality. — Corral Draw, southeast base of Black Hills, S. Dak. General characters. — About the size of the wharf .rat, but of very different appearance. Ears and eyes, large; mustache, very long; tail, bushy, almost squirrel-like; fur, long and soft; expression, animated. Color of upper parts, pale pinkish buff ; the feet and underparts, white. Young, light buffy gray ; tails, mostly white, not buffy. Measurements of average adults : Total length, 349 millimeters ; tail, 144 ; hind foot, 43. Distribution and habitat. — Wood rats are scattered over the State from the Missouri River Valley westward, or throughout the Bad- lands region. Just where they grade into the slightly darker cinerea is not entirely worked out, but probably somewhat west of the line between North Dakota and Montana. They are nowhere abundant or even common, but seem to be generally distributed wherever there are Badlands cliffs and ledges to afford suitable homes and protection from their enemies. Their range can hardly be con- sidered continuous, but they have managed to scatter from one cliff to another until they have occupied almost every suitable rocky slope. There are specimens from Mikkelson, Oakdale, Wade, and the Little Missouri River near the former Dakota National Forest, and their unmistakable signs have been found at Marmarth, Deep Creek, Medora, the White House Ranch (18 miles southeast of Williston), and near Goodall and Parkin. A few were reported at Cannon Ball, and Doctor Bell caught two at Wade. In 1833, Maxi- milian (Wiecl, 1839-1841, Bd. 1, p. 438, 1839; Bd. 2, p. 89, 1841) re- ported them in the forest at Fort Union, at Cedar Island, and at Fort Clark. In 1869 Cooper (1869, p. 296) reported them in the rock bluffs that border the Missouri River above Great Bend. In 1872, Doctor Allen (1875, p. 42) found them more or less frequent in the timbered portions of the streams west of Fort Rice. Their principal strongholds are cliffs and ledges, but they also occupy the cottonwood forests in the stream valleys, where hollow logs and trees often offer choice homes, and dense masses of bullberry brush and impenetrable cover afford the protection necessary for their exist- ence. They are quick to find and occupy buildings of any sort, and for this reason their presence is soon made known wherever they occur. General habits. — In habits as well as appearance wood rats are entirely unlike the Old World rats, and it is unfortunate that they should have to bear the odious name. Their long mustaches and big ears and eyes give them an even more animated expression than that of the squirrels. While mainly nocturnal, they are occasionally seen by day when disturbed from their cozy nests in cabins or cliffs. They are timid animals, with practically no means of defense against numerous enemies except the barricades of their dwellings. Their favorite home is a cleft in the rocks where narrow cracks ad- mit them to deeper cavities or where the openings can be blocked with sticks, stones, and rubbish to keep out larger animals. This building habit has become so fixed that their first instinct is to gather building material wherever they are.. Even where it is not 1926] MAMMALS OF NORTH DAKOTA 87 needed they often pile up heaps of rubbish, sometimes in front of their doorways, sometimes in large openings that they could never hope to fill. These accumulations remain for years in the caves and caverns, or hollow logs, trees, or cabins, where the wood rats have been; as also do the long black pellets of excrement. In buildings to which they have gained access all small articles of a convenient size are usually gathered into one corner to protect the nest, but sometimes they are piled into a box, cupboard, or stove, or heaped up in some doorway that the rats would like to be able to close. In occupied buildings they are noisy and mischievous, but may soon be discovered and disposed of. Roosevelt (1900c, pp. 66-67), on returning to his ranch on the Little Missouri when it had been unoccupied for some months, found that " within doors the bushy-tailed pack-rats had possession, and at night they held a perfect witches' Sabbath in the garret and kitchen." Half the cotton had been dragged out of a mattress and made into a big, fluffy nest that entirely filled the oven. In 1909 one of their old rubbish houses was found in the rocks not far from Marmarth and two of the animals were said to have been caught in a house at the edge of town not long before. At Medora, Jewett was told that they occasionally came into the houses, but he could not find any while there, although he found signs of their characteristic work in the rocky ledges about 10 miles farther north, and took two immature specimens near Mikkelson. At Goodall, in the Killdeer Mountains, he was told that they rarely came into buildings, but he obtained two specimens in a ledge about a mile south of town. In the little Missouri Valley, about 25 miles south of Medora, the writer found their signs and old building material quite common in the Badlands gulches, in one of which Jewett obtained a specimen. South of the Missouri River, a little below Williston, old signs were found among the rocks in several places, but no specimens were taken. Near Goodall, Kellogg reported one caught by a cat on the Good- all ranch, but apparently they were not common in that vicinity. Near Cannon Ball, Sheldon learned of several that had been caught, one in the cellar of an old sod building, and another near the mouth of the Cannonball River in the forested bottoms. At Wade, farther up the Cannonball River, W. B. Bell collected two specimens in a cave in the cedar-covered buttes, and reported them as frequently entering houses and stables of settlers, where they occasionally did considerable mischief. Breeding habits. — Wood rats, like the other members of the group, probably have two or sometimes four young, which are raised in the soft birdlike nests of their safe retreats. Apparently only one litter of young is raised in a year, so that reproduction is not rapid and there is little danger that these rodents will become very numerous. Food habits. — Examination of stomachs shows that the food of wood rats consists largely of green vegetation. One taken on the Little Missouri River had green foliage and buffaloberries in its stomach. About their occupied dens there are always traces of various plants brought in for food, and apparently these are brought in for the foliage more than for seeds. The rats, however, are eager for rolled oats, grain, fruit, or bacon used as trap bait and are easily 88 NORTH AMERICAN FAUNA [No. 49 caught wherever they occur. They do not become fat in fall nor show any signs of hibernation during the coldest winter weather. Generally they lay up stores of green plants, berries, and seeds, which become dried and well cured for winter food, the dry vegeta- tion forming the great bulk of the winter stores. Economic status. — In North Dakota wood rats are not sufficiently abundant to be of any great economic importance, . although related species in other parts of the country are very injurious to crops, forage, and native vegetation. Here they merely add a feature of interest to the picturesque cliffs of the Badlands, with which they are closely associated. They have a strong and not unpleasant musky odor which apparently comes from the glands of the skin; it in no way affects the flesh, which is sweet and delicate as that of young rabbits. Although too small to be of any value as game animals, wood rats are in every way suitable for food. There is a possibility of their serving as interesting pets for children, if properly tamed and kept within bounds. Evotomys gapperi loringi Bailey Red-backed Mouse Evotomys gapperi loringi Bailey, Proc. Biol. Soc. Washington, vol. 11, p. 125, 1897. Type locality. — Portland, N. Dak. General characters. — In size somewhat smaller than the meadow mouse and of the same general form ; ears small, nearly concealed in the long fur ; tail, short ; legs, short ; eyes, small. Whole back, rich chestnut brown, lighter brown in winter ; sides, gray ; underparts, whitish. Average measurements : Total length, 123 millimeters ; tail, 31 ; hind foot, 18. Weight, 18 to 22 grams. Distribution and habitat. — Red-backed mice, which represent a small, pale form of a wide-ranging Boreal forest species, occupy the woods and thickets of the Transition-Zone plains region of North Dakota and central Minnesota. Specimens have been taken at Pembina, Grafton, Portland, Larimore, Devils Lake, Stump Lake, Valley City, Kathryn, throughout the Turtle Mountains and Pem- bina Hills, at Towner, Williston, Buford, Goodall, Elbowoods, Oakdale, Fort Clark, and Cannon Ball. In most of these localities the mice are abundant in the thickets and timber and it is probable that they occupy every suitable locality within the State. General habits. — In the Turtle Mountains red-backed mice were abundant throughout the woods and brush, and many were caught in traps set under logs, near old stumps, at the bases of trees, in holes in banks, and on smooth ground where the leaves were scraped off to make an easy runway. At Portland, Loring reported them common, many being caught in traps set under logs and roots of trees. At Stump Lake the writer found them numerous in the woods, along the lake shore, and in the thickets out on the prairie. After dark one evening, the writer entered a dense thicket and, scrap- ing away the leaves, set seven traps at random within a radius of 6 feet ; the next morning five contained red-backed mice. At Cannon Ball he found the mice common in the forest on the river flats, and at Fort Clark and Oakdale, Jewett found them in similar situations. At the White House Ranch, 15 miles southeast of Williston, Doctor Bell and his party caught several among the dense growth of reeds 192(5] MAMMALS OF NOETH DAKOTA 89 (Phragmites) at the edge of the woods, and at Fort Buford, An- thony found them common in the brushy river bottoms. Unlike meadow mice, they rarely make trails or roadways, but run at ran- dom over the leaves and open ground in the woods and bushes, or, preferably, under the cover of thickets and brush, so that little sign of them is noticed, even where they are abundant. They are by no means strictly nocturnal; occasionally the little animals are seen rustling about in the leaves and grass and nearly as many specimens are caught during the day as at night. To some extent they burrow in the ground, but any natural cavities in or under logs, stumps, or trees seem to answer their purpose for homes and nests. Breeding habits. — Many of the females taken for specimens are found to contain embryos varying in number from 4 to 6, and oc- casionally 8. These are found at all times through the summer, which would indicate that several litters are raised in a season. As the mice do not hibernate, the breeding season probably covers all but the midwinter period. Food habits. — The examination of stomach contents of red-backed mice shows usually a combination of seeds and green vegetation. Considerable grass and many small plants are found cut and partly eaten on their feeding grounds and some of these are drawn under the logs and brush where the animals live. They are always eager for rolled oats and will take almost any kind of grain or seed used for trap bait. At Walhalla, Williams caught one in his hands in the daytime, as it was eating a piece of bread crust in front of his tent. At Portland, Loring caught a large series of specimens by baiting his traps with meat, and found that occasionally they eat their own kind or other varieties of mice caught in traps where they run. For the greater part of the year, however, their principal food is green vegetation, which they find abundant even under the deep snow of winter, for they plow along the surface of the ground and come in contact with the tender shoots of frozen grass or roots under the leaf mold, or if these are not sufficient they gnaw the bark from bushes and small trees as high up as the snow offers con- cealment from their enemies above. There is some evidence that the red-backed mice may be one of the bean-storing species of the Missouri River region. In attempt- ing to discover what animal was really responsible for these caches of food which the Indians find, specimens of all of the small mice occur- ring at Cannon Ball were trapped, but the only evidence obtained in September, before the mice had begun making their winter stores, was the fondness shown by these mice for the mouse-beans (Falcata comosa) . In one bean patch seven traps set on the bare ground, each baited with a half of one of these large, juicy underground beans, caught four of the mice in one night. The eagerness of the mice for the beans and their abundance in the localities where the beans grow suggests that the species may be in part responsible for the stores which have been an important food of the Indians and a boon to many early explorers. Economic status. — As the red-backed mice occupy mainly wood- land and thickets they are of no great economic importance in the grainfields or meadows, but wherever they come in contact with or- 90 NORTH AMERICAN FAUNA [No. 49 chards and shrubbery, if present in great numbers, they are capable of doing considerable mischief, for trees and bushes girdled by them under the snow are usually killed or seriously injured. Their chief enemies are the small owls and hawks, weasels, skunks, and badgers, and these afford the principal protection we have from their depre- dations. Microtus pennsylvanicus pennsylvanicus (Ord) Eastern Meadow Mouse (PL 14) Mus pennsylvanica Ord, Guthrie's Geogr., 2d Amer. ed., vol. 2, p. 292, 1815. (Reprint by S. N. Rhoads, 1894). Type locality. — Near Philadelphia, Pa. General characters. — Size rather large for a meadow mouse ; ears, tail, and legs, short; body, heavy and compact; fur, long and soft in winter, thin and harsh in summer. Colors, dark brown or blackish above, slightly paler and more grayish below. Average measurements of adults : Total length, 171 millimeters ; tail, 46 ; hind foot, 21.2. An adult female from Grafton measured 162, 34, and 19, respectively, and weighed 43.6 grams. Distribution and habitat. — From a wide range over the northeast- ern United States the abundant meadow mice, 'pennsylvanicus, barely reach into eastern North Dakota along the upper Red River Valley, and even here they are becoming slightly smaller than typi- cal, in a very gradual gradation toward the little drummondi, which continues to the northwest. To this race are referred the Red River Valley specimens from Fairmount, Blackmer, Hankinson, Lidgerwood, Wahpeton, Fargo, Larimore, Grafton, and Drayton. Here, as in most of the range of the species, they inhabit mainly the marshes, but are occasionally also found on the uplands under any tall grass or dense vegetation. General habits. — As their build and color suggests, meadow mice are ground dwellers and spend most of their lives under the shadowy cover of dense vegetation. They burrow in the ground, which is often perforated with their little round tunnels, while over the sur- face they make well-defined roads or runways from one burrow to another or from their nests and burrows to feeding grounds, but always under the protecting cover of grass or other plants. They are partial to moist ground and prefer rough meadows where mois- ture is abundant and where water often stands over part of the sur- face. They are good swimmers and seem as much at home in water as on land, swimming from side to side of little streams or ponds and diving and swimming under water when necessary. Their runways often lead through wet places, but dry banks or hillocks are always sought for their nest cavities. In summer they frequent the meadows and low moist ground, but in winter they leave their low underground burrows and push out under the snow of the up- lands, extending their tunnels along the surface of the ground and building warm grass nests on the surface wherever they find an abundant food supply. In this way they often range over uplands, fields, and orchards in great numbers ; and when the snow goes off in their winter nests exposed, they burrow into the ground again and spring, leaving their network of winter roads, their grass piles, and do not all get back into the meadows until the dry weather of sum- mer compels their return to the moist lowlands. 1926] MAMMALS OF NORTH DAKOTA 91 Apparently they are about as much diurnal as nocturnal, although they appear to be most active in the early evening hours. Some are caught in traps during any part of the day or night, but they seem to fill up the collector's trap line more rapidly during the evening than at any other time. They do not become fat or hibernate and are active throughout the year without regard to weather. At Hankinson, in July, 1912, they were found common in places where there was tall grass on the prairie, but more especially so in the meadows and among the tall tules on wet ground. After the hay was cut on the meadows they gathered under the haycocks in considerable numbers, and as these were loaded and hauled away the mice were forced out into the open stubble, where hawks or owls were constantly hunting them. As many of the extensive tule marshes in this region are never cut, they afford a safe harbor for the mice in which to multiply and from which to spread over the meadows and prairies. Similar conditions were found at Wahpeton, near Fargo and Grafton. Breeding habits. — Meadow mice are very prolific and seem to have no well-defined breeding season. Females taken at any time of year from May to October are found to contain from 4 to 6 embryos and sometimes 8. Occasionally even in winter small young are found in nests or females are found with embryos. At times they increase very rapidly and remain abundant for a period, after which they become scarce, the variation presumably being correlated with pro- tection from enemies and the abundance and quality of food. Their waves of abundance, which at times suggest migrating hordes, are undoubtedly due to favorable conditions for rapid reproduction. Food habits. — During the summer the favorite food of meadow mice consists of tender vegetation, as the young shoots of grass, grow- ing grain, clover, alfalfa, and a great variety of plants. Acceptable food is always abundant. Sometimes the tender bases of grasses are chosen, and again the seed-laden tops are drawn down within reach by cutting the stems into inch-long sections. Seeds and grain are al- ways favorite foods, when they are available late in summer and in fall, but in winter the mice thrive equally well on the frozen grass and roots that they get under the snow, or on the bark from bushes or trees which they gnaw off with their sharp cutting teeth, under cover of deep snow. They are particularly fond of the bark of many fruit trees, espe- cially apple, but will eat the bark of almost any tree or shrub that is not too thick and hard for them to gnaw into. Aspens and willows are found peeled and killed by them, but most of the larger woodland trees are protected by their harsh outer bark. In trapping them rolled oats are a favorite bait, but they will take any kind of grain or seeds and are specially fond of fresh meat, fat pork, or bacon. Their carnivorous propensities are seen in trapping, as they almost invar- iably eat their own kind or other mice found dead in traps. Economic status. — Over a wide range of rich farming country these are the most abundant mice in meadows, fields, timber, and orchards, and their total destruction of forage, hay, crops, trees, and shrubs annually causes an enormous loss in agricultural products. Among these losses nothing is more exasperating than the destruction of a few choice fruit trees in an orchard, or some choice shrubbery in 82242°— 2G 7 92 NORTH AMERICAN FAUNA [No. 49 the yard, during the period of deep snow in winter. Orchards are infrequent in North Dakota, so that the few choice, hardy fruit trees that can be raised are of special importance, and if the bark is gnawed from the base, even on one side, the trees are often weak- ened so as to be unable to resist the severe climate to which they are exposed. In wild-grass meadows the hay is not of sufficient value for the mice to cause much loss, but in meadows of clover and timothy and fields of alfalfa their mischief is much more serious. They enter grainfields as soon as the growing crop is sufficiently high to afford protection, and cut the grain shoots for food ; then when the grain is headed out, they cut off the base of the stems, drawing down the heads in order to reach the green and ripening grain. After harvest they congre- gate in the grain shocks and if these are left long in the field consid- erable grain is eaten or shelled out and destroyed. As either these meadow mice or closely related species with similar habits cover all of North Dakota, the total loss from them is by no means insignifi- cant. The importance of placing every possible check on their increase is obvious. That they can be successfully poisoned when necessary has been demonstrated, but the expense suggests this method as a last re- source. The most practical method of controlling the abundance of such small rodents is by protecting their natural enemies, among which the owls and certain species of hawks are foremost. The little owls, during the dusk of evening and all night long, are watching for them and miss no opportunity to pounce upon an imwary mouse that exposes itself. The marsh hawk, or mouse hawk, as often called, sailing low over the meadow and prairie, with eyes intently fixed on the ground, drops suddenly into the grass and secures a mouse more often than it does any other prey. Many other hawks feed upon them extensively, as do also foxes, badgers, skunks, and weasels. But for these enemies the mice would overrun the farms with disastrous results. Microtus pennsylvanicus drummondi (Audubon and Bachman) Druinmond Meadow Mouse (PI. 14) Arvicola drunvmon&ii Audubon and Bachman, Quadr. North Amer., vol. 3, p. 166 [1854]. Type locality. — Rocky Mountains, vicinity of Jasper House, Alberta, Canada. General characters.— Similar to pennsylvanicus, but much smaller and slenderer, and slightly lighter, more yellowish brown in coloration. Average measurements of adults : Total length, 145 millimeters ; tail, 39 ; hind foot, 17.8. Distribution and, habitat. — None of the North Dakota specimens of the little prairie or meadow mice are typical, but they are too small and slender-skulled to be called pennsylvanicus and can best be referred to drummondi, toward which they are slowly grading to the northwestward. The large series of specimens now available from many localities over the State show conclusively continuous range and complete intergradation between the two forms, and, drummondi is here placed as a subspecies of pennsylvanicus. To 1026] MAMMALS OF NORTH DAKOTA 93 it are referred specimens from Crosby, Lostwood, Kenmare, Towner. Turtle Mountains, Walhalla, Pembina, Sweetwater Lakes, Devils Lake, Stump Lake, Portland, Valley City, Lisbon, La Moure, Oakes, Ludden, Napoleon, and Dawson. This carries the range over the high glacial-prairie region between the Red and Missouri River Valleys to the southern boundary of the State and marks its south- ern limit from a wide range over western Canada to Alaska. Over their part of the State Drummond meadow mice are very abundant and occupy the high open prairie, rich bottomlands, and grassy meadows. To some extent they are found in woods and thickets, but primarily they are dwellers in grasslands, wherever the low vegetation affords food and cover. In the Turtle Moun- tains, the writer found them abundant in marshes, meadows, banks, and grassy fields, and even in damp woods, but they were most numerous in the meadows, where the ground was perforated with their runways. In spots nearly half the grass had been cut down for food, leaving the earth strewn with the fragments. At Wal- halla, in the Pembina Hills, Williams reported them inhabiting woods, grainfields, and meadows. On Bird Island, in the aim of Devils Lake, where cormorants nest, the grass was full of run- ways. At Stump Lake the runways were found in the prairie grass apparently without regard to whether the ground was wet or dry. At Valley City East gate reported them common along the river valley, around the marshes, and in prairie meadows, and the writer caught them around some seepage springs high up on the side of the bluff. At Lisbon, Doctor Fisher caught one in a swampy thicket ; and so on over the State they have been reported from a great variety of localities. General habits. — If their habits differ at all from those of yenn- sylvanicus, it is only in a more ready adaptation to high open ground, such as the grassy prairies of their range afford. Their more open habitat may well account for this slightly paler color- ation. Their little roadways are often conspicuous through the prairie grass, especially where the old grass has fallen down and made a protecting cover over the surface of the ground. Old win- ter nests are found scattered over the surface, but rarely are they occupied during the summer; the principal nests are then in un- derground cavities to which the burrows lead. Occasionally an occupied nest is found in some old haycock or grain shock that has been left out over winter, and at Crosby a nest was found occupied by young in a heap of last-year's weeds by the roadside. The nest, as usual, was made of soft grass blades, built into a neat hollow ball, clean and fresh, and wijth a soft lining inside. It was placed in a slight depression in the ground, where it was well protected from rain and snow by the mass of matted vegetation overhead. The four small young inside, with their eyes just opened, were of a beautiful golden-brown color, quite different from the sooty, or slaty gray, young of pennsylvanicus of the same age. Breeding habits. — In the latter part of June, 1912, in the Tur- tle Mountains, great numbers of these mice were caught in traps so that many were thrown away after series were selected for speci- mens. There were all sizes and ages, from little fellows just out of 94 NORTH AMERICAN FAUNA [No. 49 the nest to nearly full-grown young of the year, indicating at least two litters of young of that season, while many of the females con- tained small or large embryos, usually 6, 7, or 8. The mammae are arranged in two posterior and two anterior pairs, so that 8 is proba- bly the normal maximum number of young. Apparently breeding continues throughout the summer, if not throughout the year, and reproduction is so rapid that only through a host of enemies are their numbers kept down to a safe limit. Food habits. — Grass and weed stems are found cut in little sec- tions near the runways and on the feeding grounds of these mice and over considerable areas where much of the grass has been cut by them. The mice are fond not only of seeds but also of grain, and enter fields readily and help themselves to growing crops. Although never fat, they are always well fed and their stomach contents show various mixtures of green plant tissue, white pulpy root and bulb tissue, and the meal or dough of finely masticated seeds and grain. Economic status. — Over an immense area of rich grain-producing land these mice swarm in greater or less abundance, varying with the seasons and with the abundance of their enemies. It would be almost safe to predict that at times, through disturbance of normal conditions in the agricultural development of the country, these mice will increase so as to do serious injury to crops. In such case it may become necessary to use artificial means of destroying them, but as with other small rodents, a wise protection of their enemies will generally produce sufficient check on their abundance. The destruc- tion of weasels for fur and too great a reduction of skunks and badg- ers are likely to have a marked effect on the abundance of these mice, while any wanton destruction of owls and mouse-feeding hawks would certainly be followed by an inordinate increase in the num- bers of the rodents. Microtus pennsylvanicus wahema 16 Bailey Bean Mouse; Hetunka (PI. 11, fig. 2; pi. 14) Hintunka of the Dakotas; Gipdpuli of the Hidatsas ; Sakch of the Ari- karas ; Biddbaho itdhu of the Hi- datsas (all, Gilmore). Microtus pennsylvanicus wahema Bailey, Journ. Mamm., vol. 1, p. 72, 1920. Type locality. — Glendive, Mont. General characters. — A pale form of pennsylvanicus, slightly smaller and very much paler. and grayer than the eastern meadow mouse, which it repre- sents in the arid Badlands region. Upper parts huffy gray ; sides clear gray, underparts and feet and lower surface of tail pale gray or buffy white. Measurements of type specimen : Total length, 178 millimeters ; tail, 43 ; hind foot, 20. Weight of adult female, from Cannon Ball, 30.8 grams. Distribution and habitat. — Bean mice occupy the Badlands section of the Missouri River Valley and range westward over southwestern North Dakota and eastern Montana. There are specimens from near the mouth of the Cannonball River, Bismarck, Mandan, Fort Clark, 10 miles south of Williston on the west side of the river, Oak- 18 Omaha name contributed by Doctor Gilmore, Itrtshunga wahema, the burying mouse, from its habit of storing food in the ground. 1926] MAMMALS OF NORTH DAKOTA 95 dale, Glen Ullin, and Sentinel Butte. This indicates a continuous range over the Badlands and sagebrush semiarid section of the State. In places the mice are found in marshy bottoms, but more often in the long grass of draws and on grassy benches of the ridges and buttes. Near the White House Ranch, about 12 miles south of Williston, a fine typical specimen was caught in 1913, on a grassy bench near the top of the Badlands border of the valley. At Fort Clark Jewett found others on high grassy slopes back of the valley bottom, and at Mandan on the high ridges wherever the grass was sufficiently dense to hide the animals and their runways. At Oakdale he found them about a small marshy place near a spring, where they were occupying the tall grass of a limited area. At Glen Ullin he found them common in the tall grass on moist ground along Curlew Creek, where their fresh runways were abundant. At Sentinel Butte he took specimens high up on the grassy slopes of the large butte south of town and others in the grass of a small slough near the station. Near Cannon Ball, Sheldon took specimens on the flats near the mouth of the Cannonball River and also on the high buttes to the north. In June, 1916, the writer collected them on the river flats near the mouth of the river and saw abundant signs in grassy places over the prairie and fields ; and on October 30, 1919, took two of the mice at their nest and bean cache on the flats near the mouth of the river. General habits. — In habits these mice differ from pennsylvanicus and drummondi only as their more arid habitat places them more in the open, sparsely covered, and grayer soil of this semiarid re- gion, where they are evidently more exposed to light and to the numerous enemies overhead. The light-colored soils and minerals and general gray tone of sagebrush and prairie plants is evidenced in the color and, to some extent, in the habits of these little animals. They are less uniformly distributed over the area than are species in more fertile regions and in places they seem almost colonial, so locally are they gathered in the most favorable spots. In summer they were not easily caught in traps, as they seemed not to care for any bait that was offered them, and the few specimens taken merely ran through the traps set in their little runways under weeds and grass. There has long been a question of whether these could be the bean-storing mice of the Indians of the Upper Missouri River, men- tioned by Lewis and Clark and other early explorers as laying up such ample stores of wild beans, bulbs, and tubers, for a winter's supply of food that they formed one of the important sources of food supply for both Indians and whites. To decide this question many specimens of these, as well as of the other species of mice living along the Missouri Valley, were collected in localities where the Indians said the mouse stores were especially abundant. Some of the Indians and white men, who were familiar with the mouse stores, picked out the present species as the one which they had seen running away from the beans, but others were just as positive that the storers were the red-backed mice, white-footed mice, harvest mice, grasshopper mice, and pocket mice, while some thought the deposits were made by pocket gophers, ground squirrels, or chip- munks. Even the weasel was accused of storing these winter sup- 96 NORTH AMERICAN FAUNA [No. 49 plies. The stores were frequently described and all seemed to agree as to their contents. With pockets filled with underground beans of Falcata comosa (maka ta omnicha) and long tubers of wild artichoke, Helianthus tuber osa (pangi), and the little white tender roots of wild morn- ing-glory, the writer was able to question the Indians intelligently about the stores and the way they were found and gathered and cooked. Although the mouse bean seemed to be the principal part of the stores that were sought by the Indians, the artichokes and morning-glory were said to be usually found with them, and one Indian insisted that the tipsin, Psoralea esculenta, was also sometimes "found in the caches. One man insists that when driven away from their stores the mice often climb and take refuge in trees. In describing the cache the Indians say that the mouse burrows enter the ground from several sides and the cavity where the food is stored often holds a peck or a pailful of beans and tubers. One Indian, who makes a special business of gathering these beans in the autumn, positively asserted that he could find enough in a day to fill a 2-bushel sack. The method of finding the stores is by noting either the burrows and runways centering at a certain point, or the tracks of the mice in the fresh snow leading to and from them. With a sharp stick the ground is probed, the cavities are soon discovered, and the beans removed. The fresh and wholesome vegetables were at one time an important adjunct to the meat diet of these hunting Indians, but at the present time their fields of vegetables and grain furnish an ample variety of food and the mouse stores are sought only by a few. The Indians claimed that it would be impossible to find the stores until late in October or early in November. In October, 1919, six years later, the writer returned to the mouth of the Cannonball Kiver in the hope of being able to settle the ques- tion of identity, and on next to the last day of the month succeeded in finding his first cache of beans and capturing the mouse with them. The night before he had trampled down the soft snow and in the morning found several fresh mouse holes made during the night, en- tering different sides of a mass of snow and leaves. Digging in one of these holes with the left hand the writer saw a mouse soon pop out on the other side, only to be caught in the right hand, and placed in a glove, and carried home alive. The cavity was then carefully dug out and examined. A warm nest of grass and soft plant fibers was found about 6 inches below the surface in a cavity where an old stump had decayed. In another cavity near the nest was a small collection of the mouse beans or ground peanuts, with artichokes and a few roots of the wild morning-glory. As the season had been very dry and both mice and beans were scarce, the cache was meager, but the cavity, which would have held several quarts or a peck, showed the old skins and remains of the previous year's collections. The store would doubtless have been added to until the ground froze so hard that no more beans could be dug. Though there may be other mice which store these beans, this meadow mouse is the first one actually caught at its cache and identified. The Indians describe the cache as easily recognized by the little roads leading up to it from all sides, and tell how the mice drag 1926] MAMMALS OF NORTH DAKOTA 97 home loads of the beans on leaves. They have many legends and stories relating to these mice and their stores, which have been well translated by Doctors Beede and Gilmore, stories telling of the re- spect and reverence of the Indians for their little helpers, the mice people, of the payment in corn or other food for the beans taken, of the punishment of the hard-hearted woman who took all of the beans and left no exchange in food, and of the threat to fight any white man who attempted to capture or injure the mice or take their stores. History. — The use of these ground beans evidently dates far back. Mr. "Will (1917, p. 66) in speaking of the mythical origin of the Hidatsas from a hole in the ground in the vicinity of Devils Lake, says: "At that time the people cultivated ground beans and wild potatoes, two crops that were not really cultivated at all but merely gathered." Apparently the first white men to mention the beans were Lewis and Clark (1893, p. 161, 263), in 1804, on their visit to the " Kicaras " (Hidatsas), when among other presents of food they were given " a large rich bean which they take from the mice of the prairie, which discover and collect it." The next spring their " Bird Woman," Sacagawea, gathered these food stores of the mice, which must have lasted over winter. In 1833 Maximilian (Wied, 1843, p. 276) includes this bean among the plants used by the Mandan Indians as food under the name " feverolles (Fabia minor equina), a fruit resembling the bean which is said to grow in the ground but which I did not see." Again, Father De Smet (1905, p. 655), an early missionary to the Indians of the Upper Missouri, as he left Fort Union in 1851 wrote in his journal : The earth pea and bean are also delicious and nourishing roots, found com- monly in low and alluvial lands. The above-named roots form a considerable portion of the sustenance of these Indians during winter. They seek them in the places where the mice and other little animals, in particular the ground- squirrel, have piled them in heaps. In 1855 Lieut, G. K. Warren (1856, p. 78) wrote: The groundnut, or Apios tuberosa, is very useful to the Indian. It grows very abundantly along the river bottoms, and is gathered in large quantities by a kind of wood-mouse for his winter store. The squaws make a business, during the months of October and November, of robbing these little animals, and I have often seen several bushels of the tubers in a single lodge. They are boiled with dried buffalo meat, and make a rich and palatable dish. Thus a long and useful career has been shown for these little ani- mals and we can well appreciate the feeling of regard for them still held by the older Indians. Now, however, that most of their range has become valuable grain land, their services are no longer needed and their inroads on grain, grass, and other crops are likely to prove as serious as those of other related species in agricultural areas. It is safe to say, however, that they will not be exterminated nor their numbers greatly reduced by the presence of the white man's civili- zation. The only danger is that under cover and stimulus of cul- tivated crops they may increase to such abundance as to become a menace, but if their natural enemies, owls, hawks, and weasels, are given a fair chance' any overabundance will be effectively checked. 98 NORTH AMERICAN FAUNA [No. 49 Microtus ochrogaster haydenii (Baird) Western Upland Mouse (PI. 14) Arvicola {Pedomys) haydenii Baird, Mamm. North Amer., p. 543, 1857. Type locality. — Fort Pierre, S. Dak. General characters. — A medium-sized field mouse of the subgenus Pedomys, with short ears, legs, and tail, the tail about twice as long as the hind foot. Color dull gray with a cinnamon tone, only slightly paler below. Fur long and lax, giving a pepper-and-salt effect of light-tipped hairs over dark underfur. Measurements of adult female from type locality: Total length, 180 mil- limeters ; tail, 47 ; hind foot, 22. Distribution and habitat. — The pale western form of Microtus ochrogaster of the central prairie States occupies the semiarid Plains region from Kansas to Montana, and comes into North Dakota west of the Missouri River. There are specimens from Cannon Ball and Wade, and the writer saw runways and burrows near Stanton that undoubtedly belong to this subspecies. Unlike the meadow mice, they avoid low or wet ground and usually are found on the high, dry prairie in rather open situations. In many places they occupy little thickets of rose and wolfberry bushes, but their characteristic runways and burrows are often found on the open ground, fully ex- posed to view. General habits. — At Cannon Ball, the upland mice were found to be common over the prairie and on the dry valley bottoms. In places they were living under a good cover of prairie grass, where their little roadways over the surface of the ground led to the bur- rows and some old surface nests that had evidently been used during the winter. In other situations they lived in the thin prairie grass, where their runways were easily followed. In some locations they were living near the edges of thickets, where it would have been an easy matter for them to gather the mouse beans had they been inclined to store them. Specimens were easily caught by setting traps across the runways, baited with rolled oats, or even set un- baited, as in running along their roads the mice would trip over the trigger and spring the traps. To a certain extent they seemed colonial in habits, but probably this is merely because in a good location the family increases until the place is well stocked before the members of the colony begin to scatter out. At times they be- come very numerous locally, but, generally, the open nature of their habitat exposes them to so many enemies that they do not last long. Food habits. — The food of this mouse is largely green vegetation, including the stems and leaves of grass and a great variety of little plants that are found cut in sections in their runways. It also eats the flowers and seeds of many plants, is usually eager for rolled oats or other kinds of grain used as trap bait, and will often eat its own kind found dead in traps. Preference for high and dry ground brings it much in contact with cultivated fields, where it finds choice food in the green or ripening crops. Breeding habits. — Females taken for specimens often contain four to six embryos and the mammae are arranged in two posterior and one anterior pairs. Apparently they breed many times during the season and are only a little less prolific than meadow mice. 1026] MAMMALS OF NORTH DAKOTA 99 Economic status. — In North Dakota these mice have been so little observed that any injury to crops has escaped attention, but in other parts of their range, where farms and orchards have been of longer standing, they have been known to occasion serious losses by killing fruit trees and by destroying grain and grass in fields and meadows. Potentially they are dangerous occupants of any agricultural region and with unchecked abundance might become a serious pest. Microtus minor (Merriain) Little Upland Mouse (PI. 14) Arvicola austerus minor Merriam, Amer. Nat., vol. 22, p. 600, 1888. Type locality. — Bottineau, N. Dak. General characters. — Smaller even than (Irummondi, with short ears, short tail, and coarse, lax fur. Color, coarse pepper-and-salt gray, produced by pale- buff tips of long hairs over black underfur ; underparts but little paler. Adults measure in total length approximately 140 millimeters ; tail, 33 ; hind foot, 17. Distribution and habitat. — In a range extending from southern Minnesota to Edmonton, Alberta, the little upland mice cover ap- proximately the eastern half of North Dakota. There are specimens from Bottineau, Kenmare, Starkweather, Goodall, Devils Lake, Stump Lake, Valley City, Sherbrooke, Oakes, Lidgerwood, Fair- mount, Hankinson, and Blackmer. Over the prairie they are usually found on dry ridges or sandy soil, in which they delight to burrow. They seem to avoid low, damp ground and their habits as well as their fur mark them characteristic upland mice, a group quite apart from typical meadow mice. General habits. — Apparently colonial in habits, the upland mice are usually found abundant in favorite spots and in no others for long distances. Often their burrows enter the ground in groups of half a dozen or more and are more or less connected below the surface. Some of these groups suggest a family colony, and others are more extensive and scattered along for a considerable distance in irregular formation. At Bottineau, in the summer of 1887, these mice* were abundant over the dry prairie in small colonies, usually on mellow, somewhat sandy soil. At Kenmare, near the top of a high ridge or point of the prairie running out on the edge of Des Lacs Valley, their little runways and burrows were found numerous over the dry slope. The ground was cov- ered with a network of fresh trails through the short prairie grass and there were three sets of burrows, in each of which 10 or 12 holes entered the ground within a radius of 2 or 3 feet. These seemed to be family or colony dens and several of the mice were caught around each group. Fresh earth was being thrown out on all sides and from each opening a trail led off to the feeding grounds or to other dens and burrows. A number of traps were set and in one night about 20 of the mice were caught. Many were 3'oung of the year and of various sizes, but enough adults were obtained for a good series of specimens. Mouse traps were sunk in the ground across their runways and baited with rolled oats and ripe and green wheat, all of which were eagerly 100 NORTH AMERICAN FAUNA [No. 49 accepted as bait. Near Blackmer, Sheldon and the writer found four distinct colonies in an alfalfa field and one on the prairie sod on the Clarey farm, not far from the station. Those in the alfalfa fields were the most extensive, covering from 2 to 3 square rods of ground each and consisting of 20 to 50 burrows and in- numerable trails. The ground was thickly perforated by the bur- rows and generally half the alfalfa had been killed over the range of the colony. Much was cut and eaten on the surface, but considerably more was killed from below, evidently by having the roots eaten off in winter. As pasturing kept the crop low, there was no trouble in finding the mice, observing their habits, and obtaining a good series of specimens. A pair of short-eared owls were nest- ing in the adjoining field, and served to keep the mice within bounds, but if the alfalfa had been allowed to grow to full height the mice coidd have increased without interference. At Valley City, the writer caught one on the high prairie under tumbleweeds, where a few of their old trails were found, though the mice seemed to be scarce. At Sherbrooke, Loring took six speci- mens in traps baited with meat and rolled oats, set along their beaten runways through the weeds. On the Peterson farm, 10 miles west of Portland, he took two in the daytime in runway traps, and at Portland caught others in similar manner. At Towner, Kellogg secured a specimen in an upland meadow, and at Goodall he found a colony on the sandy flats close to the river bank. On the short-grass prairies these mice are exposed to view from overhead, but on the dark prairie soil in their little roadways they are protectively colored, and their habit of keeping close to their burrows and darting quickly from one burrow to another seems to be their main protection against numerous enemies. Breeding habits. — As in other members of this subgenus (Pedomys) the mammae of the females are arranged in two pairs inguinal and one pair pectoral. Females have been taken contain- ing four and eight embryos, but the normal maximum number of young is probably not more than six. Evidently the young are born at irregular times throughout the season, but the length of the breeding season and the number of litters have not been definitely determined. Food habits. — Grass stems and many prairie plants are found cut in sections along the runways of these mice and near the burrows, while in numerous places little prairie bulbs, as those of the wild onion and the blazingstar, have been dug up and eaten. In the alfalfa field at Blackmer both the green leaves and tender stems of alfalfa plants were eaten, and underground the roots had been exten- sively gnawed. The fondness of the mice for rolled oats, grain, and meat, used for baiting traps, indicates a wide range of food. Economic status. — From the nature of their habitat in fields and on the uplands these mice are likely to prove as injurious to crops as any of the other species, and under favorable conditions of food and cover, such as are found in extensive alfalfa fields, they might well become a serious pest. Where exposed to their natural enemies, however, they are not likely to do more than merely swell the total loss chargeable to small rodents. 1926] MAMMALS OF NORTH DAKOTA 101 Microtus pallidus (Merriam) Pale Mouse (PL 14) Arvicola (Chilotus) pallidus Merriam, Amer. Nat., vol. 22, p. 704, 1888. Type locality. — Fort Buford, N. Dak. General characters. — Recognized by its small size, compact form, and very short tail, which is but little longer than its hind foot ; fine soft fur of a light buffy gray color over the upper parts and creamy white below ; ears and nose conspicuously yellow. The type, an adult female, measures in total length, 121 millimeters ; tail, 20 ; hind foot, 18. Distribution and habitat. — The rare little pale mouse (subgenus Lagurus) is known from only a few scattered localities from western North Dakota, Montana, and Alberta. Two localities only are represented by specimens from North Dakota — Fort Buford and Glen Ullin. In September, 1887, the writer first found them on a Badlands butte, 2 miles east of Fort Buford, where they seemed quite common in the half-barren ground just below the top on the north slope. The only reason that could be suggested for their choice of location on the north sides of the hills was that the twilight, their favorite time for activity, was longer on the shady slopes. The vegetation seemed to be about the same all the wa}^ around the summits of the hills and at best was only scantily represented. At Glen Ullin, Osgood collected three specimens in September, 1901. This is on the high dry prairie, but no report was made of the exact location at which they were caught. Apparently this is one of the rare species which occurs only at widely scattered localities, and may be nearing extinction. No mammal has been more sought for by collectors in the region where it occurs, and with so little success. In 1915, and 1919, the writer again visited the butte where the type was collected, but could find no trace of burrows or runways on this or any of the neighboring buttes. General habits. — Apparently all that is known of the habits of the pale mouse is the little gleaned from the few specimens collected at the type locality, where they were living in a colony along the shady slope of the butte. The little round burrows entered the side- hill at frequent intervals along the well-worn runways leading around the slope. In places the runways passed over grassy ground, where they were well packed by the little feet constantly using them. In other places they passed over naked soil and were only detected by the smoothly worn surface. At that time no suitable traps for catching such little animals were available and the mice seemed strangely suspicious of the clumsy box traps. Only four specimens were taken, although the colony was quite extensive and probably contained a dozen or more individuals. Rolled oats and traps now used had not been invented in those days and the mice did not care for any of the baits offered them. Food habits. — A large part of the food of these mice seemed at that time to consist of the flowers of the little silver sage {Artemisia frigida) and the blazing star (Liatris gramini folia) , and the stems and pieces left from these plants were scattered along the runways and about the entrances of the burrows; heads and seeds of winter- fat {Eurotia lanata) also were eaten. Many grasses and other plants 102 2STOBTH AMERICAN FAUNA [No. 49 had been cut, apparently for food. A partly eaten bulb of the blazingstar was found near a runway, where it had been dug up. Corn and oats, and the seeds of cactus and other plants and also bread, cake, and cheese, were placed around the burrows, but it all remained untouched. None of the specimens taken showed any signs of becoming fat and it is improbable that they hibernate, even in this northern latitude. As a young naturalist, for the first time away from his home fauna and among new and strange animals where the thrill of dis- covery was not infrequent, the writer recognized this mouse as some- thing strange and probably new, and it was with the keenest pleasure that a communication was received from Doctor Merriam, stating that he, also, had been unable to identify it as a member of any de- scribed species. Fiber zibethicus cinnamominus Hollister Great Plains Muskrat Zih-zirukka of the Hiclatsas (Maxi- milian) ; Sinkp6 of the Dakotas (Gilruore) : Shantshuke of the Mandans (Will) ; Citakh of the Arikaras (Gilmore). Fiber zibetlucus cinnamominus Hollister, Proe. Biol. Soc. Washington, vol. 23, p. 125, 1910. Type locality. — Wakeeney, Trego County, Kans. General characters. — Size medium for a muskrat, not so large as the more northern nor so small as the southern species. Fur, dense and soft ; ears, short ; tail, long, nearly naked, flattened and rudder-like; hind feet, large and webbed; musk glands, well developed. Measurements of adults : Total length, about 496 millimeters ; tail, 240 ; hind foot, 73 or 74. Weight, about 2 or 3 pounds. Distribution and habitat. — The bright-colored Plains form of the muskrat, as defined by Hollister (1911), covers the central Plains region from Oklahoma to Manitoba, including all of North Dakota. There are specimens in the National Museum collections from Fair- mount, Oakes, Lisbon, Valley City, Grafton, Fish Lake, Wood Lake, Towner, Elbowoods, Grinnell, Buford, and Dawson. It is safe to say there are muskrats in every suitable slough and lake, marsh and stream in North Dakota, in numbers ranging from a few individuals in the smaller ponds to thousands in some of the extensive marsh and lake areas. While it is impossible to obtain a reliable estimate of their numbers, or of the numbers taken for fur each year, they cer- tainly are the most abundant and valuable fur-bearing animals of the State, as they are of the whole United States. General habits. — In the lakes and extensive tule marshes near Hankinson, the writer found muskrats abundant in 1912, and there were many old muskrat houses along the shores and numerous bank burrows leading up from under water along the margins of the lakes. As usual, much trapping kept the animals down to a small part of the number that the lakes could profitably carry. They were com- mon at Wahpeton in the river and sloughs, and at Fargo, where they live in the Red River banks, and at Stump Lake and Devils Lake in the tule-bordered sloughs over the prairie ; they were scarce, however, in the brackish and alkaline water of the lakes. 1926] MAMMALS OF NORTH DAKOTA 103 In the Turtle Mountains they were found in the lakes and sloughs with which this hilly and forested region abounds, and were espe- cially numerous in the beautiful clear water of Gravel Lake, where a novel use was found for them near the fish hatchery, and where trap- ping was not allowed. The lake had been stocked with trout, perch, and bass, and the muskrats were protected and allowed to build their houses along the shores in order to keep breathing holes open to prevent the ice from closing up so completely as to smother the fish. Both fish and muskrats were thriving and multiplying rapidly and the system seemed to be working remarkably well. The muskrats were comparatively tame and it was a pleasure to watch them swim- ming, diving, and feeding out in the water. They would often lie stretched at full length on the surface, eating roots which they held in their hands above the water. Others would sit in round furry balls on the ends of logs or on the edges of their half-submerged houses, munching the green plant stems or tender roots and bulbs, which they had procured from the bottom of the lake or from the grassy banks. Just across the ridge from this lake, at the fish hatchery, other muskrats were doing considerable mischief in the fish-breeding pond by tunneling through the banks and letting out the water. The half dozen animals that were doing this mischief could have been caught with very little trouble and the banks protected, but the feel- ing seemed to be growing that the muskrats were a great nuisance, fostered probably by the lake full of valuable fur just over the ridge. Wherever the lake banks are high enough for burrows the muskrats live mainly in bank dens, but in the wide tule-bordered lakes and sloughs, where the water is so deep that the winter ice will not reach the bottom, they build large winter houses out in the water. Thus the abundance of muskrat houses in one situation is no indication of a greater number of the animals than in adjoining lakes where none are seen. Around the Sweetwater Lakes muskrats are generally abundant in spite of much trapping, as the marshes are very extensive and the conditions ideal for them. About Castleton, Loring reported them wherever an}' water could be found. Sheldon reported them common along the lake shore near Dawson, and Kellogg reported them in Wood Lake, and especially abundant in Muskrat Lake, Sul- ]ys Lake, along Shell Creek, in Turtle Creek, and at many points along the Missouri River and adjoining sloughs and streams, from Grinnell to Bismarck. In 1915 Sheldon found them abundant across the southern part of the State, from Fairmount and Oakes to Napoleon and Cannon Ball and the Badlands farther west. Along the Little Missouri River Valley, in 1013, Jewett found compara- tively few in the creeks and sloughs. Although leading mainly aquatic lives, muskrats are perfectly at home on dry land, and often when their stream or pond dries up will strike out across the prairie to find a new home. Their peculiar tracks, showing the large hind feet and small front feet, with a narrow line where the tail drags, are often seen in dusty roads and in trails between sloughs. They are sturdy fighters, and if cor- nered will combat anything that comes along, but. if taken when young and tamed they make gentle and interesting pets. 104 NORTH AMERICAN EAtTNA [No. 49 They are great builders and work industriously to make the wails of their houses thick and firm before cold weather comes. It is often said that the larger the houses and thicker the walls, the colder the winter is going to be, but even muskrats sometimes make mis- takes in their forecasts. As long as open water is available under- neath, cold weather has no terrors for the animals in their winter homes ; but the thicker and icier the walls of their houses, the safer they are from all enemies except man and his traps. Usually two or more openings lead from the nest chamber in the center of the house down into the water, and as long as these openings are kept clear the animals are free to come and go as far as water extends under the ice. Air holes through the ice are kept open in the vicinity of the houses or bank burrows and apparently the animals obtain plenty of oxy- gen from these and the bubbles lying under the ice, or from the air carried in their dense coats of waterproof fur. Breeding habits. — The young are usually brought forth in bank burrows, apparently sometime in May, and in June they are first seen swimming about as little quarter-grown muskrats. Apparently six to eight to a litter are the usual numbers ; some credit them with two or three litters during a season. Half -grown young occasionally caught in fall are generally supposed to be from second litters, but they may be merely the first litters of late young of the previous year. Apparently the young of the year do not attain full size and weight the first fall, but by the following spring it is difficult to distinguish between most of the yearling and older animals. They are very prolific, have few enemies except man, and will quickly and abundantly restock suitable grounds where they are given pro- tection. Like other rodents, they show no signs of mating for more than a brief temporary period. The whole care of the family de- volves upon the mother, for after the young are born the male has no further place in the family life. Apparently the males fight for supremacy, as occasionally one is caught with its skin cut full of slits, evidently by the incisor teeth of an opponent. Food habits. — In summer the muskrats feed on the tender shoots and stems of numerous grasses, tules, cat-tails, and water plants along the shores, on roots and bulbs, which they take from the bot- toms and banks, and to some extent on mussels and other animal food. In July, 1893, Doctor Fisher reported that in the Sheyenne Eiver, near Lisbon, where they were common, he found piles of mussel shells at various places along the banks where the musk- rats were in the habit of feeding. In Apple Creek, near Bismarck, they were found in the same ponds with the beavers and several were caught in beaver traps. Many little heaps of fresh-water mussel shells were found along the banks where muskrats had been feeding, and Doctor Bell actually saw a muskrat bring up and cut open one of these shells. In many places where they are in the habit of feeding, the accumulation of grass and plant stems builds up little mounds or platforms on which they sit while eating their meals. They are said to be very fond of carrots and parsnips, which are often used for trap bait. Economic status. — Under certain circumstances muskrats do serious damage, as when they get into irrigation ditches, artificial ponds with dams or raised banks, or in roadways through marshes. 1926] MAMMALS OF NORTH DAKOTA 105 Their burrows will quickly destroy ditch banks and dams. In 1916, they had nearly ruined a graded road running west from Devils Lake for about 2 miles through a large marsh. In about 50 places the}'' had burrowed into the sides of the grade and in many cases clear across under the road, causing the surface to break through into the soft mud below. They had also made hollow dens under the road into which passing horses had broken through. The road was graded only about 2 feet above the surface of the marsh, but even if it had been raised much higher the burrows and dens would have been a constant menace. It would have taken at least $100 to repair this road at the time it was examined, and repairs would have been useless as long as the muskrats were left there. This seemed a serious situation, but it could have been controlled with no expense, merely by allowing and encouraging thorough trapping, in this particular marsh, where every muskrat could have been caught at a profit. In very few places in North Dakota, however, is there any complaint of mischief done by muskrats, while the annual in- come from their fur reaches many thousands of dollars, well dis- tributed among the residents of the State. Fur farming with the muskrat in its native marshes has been suc- cessfully carried on in many sections of the country, as fulty de- scribed by Lantz (1910, 1917), in Farmers' Bulletins 396 (issued in 1910) and 869 (issued in 1917) of the United States Department of Agriculture. Family CASTORIDAE : Beavers Castor canadensis canadensis Kuhl Canada Beaver Ah-mik' of the Ojibways; Ah-mistf of the Crees (Seton). Castor canadensis Kuhl, Beitr. Zool., p. 64, 1S20. Type locality. — Hudson Bay. General characters. — Beavers, largest of all our rodents (sometimes weigh- ing 60 pounds or more), are heavy-bodied, strong, powerful animals, with large, webbed, hind feet ; broad, flattened, naked, scaly tails ; dense, fine underfur, and long coarse outer hair of a dark chestnut-brown color ; and short ears and huge chisel-like incisor teeth well adapted for cutting wood. la fresh fall fur they are dark, rich chestnut-brown in color, which fades to a somewhat lighter brown before the spring molt. An adult female from Mouse River, near Towner, collected by Remington Kellogg, July 30, 1915, measured : Total length, 1,150 millimeters ; tail, 400 ; hind foot, 195 ; and weighed 53 pounds ; it is unusually dark brown, but otherwise seems to be typical of the northern beaver. The young of all ages agree closely with the adults in coloration. Distribution and habitat. — Although there is very little material from which to judge, it seems safe to assume that all beavers in the Hudson Bay drainage, including the Mouse River and Red River Valleys, are of the typical form {canadensis), and very different from those of the Missouri River drainage {missouri- ensis). Formerly beavers were abundant in all the streams and many of the lakes of North Dakota, but to-day they are restricted to a few scattered localities where colonies have received sufficient protection to enable them to regain a foothold since the days of overlapping. 106 NORTH AMERICAN FAUNA [No. 49 In 1800 Alexander Henry (1897, pp. 117, 143, 145, 154, 175, 177, 408) said that beaver houses were numerous along Red and Goose Rivers, near Grand Forks, and more numerous than elsewhere on the upper Sheyenne River. Two of his trappers, from a trip up the Red River, brought in 60 beaver skins on November 17, two others, 60 skins from the Hair Hills on Park River, and the next spring two men brought in 30 skins from the vicinity of Grand Forks. Two other trappers on Park River took 25 skins in two days, and so for six years Henry's bands of Indian trappers scoured the branches of Red River and trapped in the Pembina Hills and Turtle Mountains for the furs that were poured out through the waterways eastward, to be shipped to England. As a result of this systematic destruction, Henry, in 1806, further records that where formerly plentiful beavers were becoming very scarce. Following is a partial record of beaver skins taken by his parties from the Red River Valley during the years 1801-1808 : In 1801, Reed River, 832, Park River, 643. For the winter of 1802, Grand Forks, 410; Hair Hills, 200. In 1803, Turtle River, 337; Hair Hills, 30 Pembina River, 550. In 1804, Grand Forks, 356; Hair Hills, 182 Park River, 147; Pembina River, 211. In 1805, Hair Hills, 121 Park River, 160; Pembina River, 829. In 1806, Grand Forks, 342 Pembina River, 776. In 1807, Pembina River, 565. In 1808, Grand Forks, 150; Hair Hills, 53; Pembina River, 339. Although these localities merely indicate the camps from which his men worked out in all directions, the records give a good idea of the fur harvest in its prime, and also of the rapidity with which the beaver was reduced to numbers that no longer paid the trappers for their time and effort. As early as 1848 David Thompson (1916, p. 249) wrote that the beaver had become very scarce in the Red River Valley near the mouth of Park River. The former abundance of beavers in these streams shows condi- tions favorable to their habits and in many instances marks the places where they could now be maintained in considerable numbers as an attractive and profitable form of livestock. In 1887 no trace was found of beavers along the Red River Valley nor were any colonies heard of on the way down the valley to Pem- bina. In 1893 Doctor Fisher noted a few in the Sheyenne River, near Lisbon, but in 1912 Eastgate reported them as extinct there 16 years before, although he found old cuttings and dams. It is possible that there are still a few beavers along the banks of the Red River, but no one has been able to get any trace of them in recent times. In the Turtle Mountains, in 1912, only one colony of beavers was found, and that was carefully protected by the owner of the prop- erty, who was anxious to have them multiply as rapidly as possible. All through these mountains, however, old traces or the former abundance of beavers was found, while dams closing the outlets of ponds, marshes, and lakes showed where they had been responsible for retaining the richness of the land and spreading it out instead of having it washed away by the spring floods. The best of the meadows in this region are all old beaver ponds that have been filled up with silt. There are also numerous lakes where the beavers used to live in the banks, as shown by old burrows, and where to-day 1926] MAMMALS OF NORTH DAKOTA 107 the interesting animals might live in considerable numbers without doing harm. If adequate protection could be afforded they would soon increase and restock this whole region, once a trappers' para- dise. One morning before daylight in 1915, Kellogg counted 15 beavers about 8 miles north of Towner, where they had built a big brush house on the bank of the Mouse River. At this place the water was about 15 feet deep, but a dam had been built part way across the river to increase the depth. In the early days beavers had been very numerous along this stream, and old settlers told Kellogg that its course had often been changed by their dams. At the time of Kellogg's visit there was another colony 4 miles farther up the river. At Kenmare, in 1913, there were complaints of beavers doing great damage to property on Carl Swensen's place on Mouse River, about 20 miles northeast of there. On the bank of the river just be- low the McKinney Bridge, three or four beaver houses and the places where timber had been cut along the borders of the stream were exam- ined. Apparently there were 20 or 30 beavers occupying the half mile of stream examined, and they were said to be equally numerous below there and above to the Canadian line. C. E. Booth, a taxider- mist, reported later that beavers were common in the Mouse River near Minot, and that there were eight dams across the stream just above Burlington. There is considerable small timber scattered along the course of this river and in a great prairie region even small timber is highly prized. At Mr. Swensen's place the beavers had built winter houses along the banks of the stream by piling up the sticks which they had cut, often a wagonload or more, in a heap 5 or 6 feet high, above their rooms and nest chambers in the bank and plastering them over with mud. During the visit the houses were not used to any extent, as the beavers were living mainly in bank burrows, but before winter all of these houses would be re- paired and put in good condition to protect the dens from freezing during the winter. The beavers were not cutting many trees at that time, but seemed to be feeding mainly on the green vegetation along the river banks and on willow stems and roots. Mr. Swensen showed the bank where they had cut trees the previous fall and the writer counted about 40 stumps of small ash, 2 to 6 inches in diameter, about 20 boxelders, and a dozen elm stumps of the same general size. The largest ash which they had cut was about 10 inches in diameter and another about that size had been killed by being girdled. Seven boxelders 8 or 10 inches in diameter, entirely or partly girdled, were either dead or dying. Most of these trees were in a narrow strip about 40 rods long on the bank of the river opposite the ranch house. Mr. Swensen estimated that the beavers had killed 200 or 300 trees for him and more for some of his neighbors. A few of these trees were large enough for fence posts but the greater num- ber were too small to be of any value except for shade and protec- tion from the cold winter winds. The Swensens were much inter- ested in the beavers and their work, but strongly objected to feeding so many of them on their choice trees. It would seem a simple matter for State officials or game wardens to be detailed in such cases to control the abundance of beavers 82242°— 26 8 108 NORTH AMERICAN FAUNA [No. 49 where they were doing mischief, and to capture alive and remove any surplus to other parts of the State where they would be of value in stocking suitable waters. Castor canadensis missouriensis Bailey Missouri River Beaver Capa of the Dakotas (Gilmore) ; Midapa of the Hidatsas (Mat- thews) ; Wahrapa of the Mandans (Will) ; Citukh of the Arikaras (Gilmore) ; Zhaba of the Omahas (Gilmore). Castor canadensis missouriensis Bailey, Journ. Mamm., vol. 1, p. 32, 1919. Type locality. — Apple Creek, 7 miles east of Bismarck, N. Dak. General characters. — Slightly smaller than canadensis; colors, paler and duller brown ; back, bright hazel brown ; sides, duller brown ; and under- parts, smoky gray. Young, same color as adults. Measurements of type (about 18 months old and not full grown) : Total length, 900 millimeters; tail, 270 ; hind foot, 170. Weight estimated at 35 or 40 pounds. Distribution, habitat, and general habits. — Apparently the light- brown subspecies of beaver occupies the Missouri River drainage, at least from Nebraska north and west to Montana. In North Dakota it still occupies the Missouri River and many of its tributary streams. A number of skulls in the National Museum were col- lected by Lieutenant Warren, along the Upper Missouri, probably in North Dakota. There is also a skull from old Fort Stevenson, part of a skull from the Little Missouri, and a broken skull from Medora, besides the type and one immature specimen from Apple Creek, but much more and better material is needed before a satis- factory diagnosis of the form can be given or the details of its dis- tribution fully made known. In 1804-5 Lewis and Clark (1893, p. 194) found beavers abundant along the Missouri River throughout the North Dakota section of their journey, even in close proximity to long-established Indian settlements. At the Mandan villages they speak of two French trappers coming into camp with 20 beavers that they had caught near there. Trappers were then just beginning to find this river a rich field for their fur harvest. In 1833, Maximilian reported 25,000 beaver skins bought during the year at Fort Union (now Buford). Among his many observa- tions along the Missouri River he (Wied, 1839-1841, Bd. 2, pp. 54, 55, 1841) wrote on November 5, from just above the mouth of the Little Missouri : * * * we lay to for the night on the south bank where the forest was com- pletely laid waste by the beavers. They had felled a number of large trees, chips of which were scattered about on the ground. Most of the trees were half gnawed through, broken down, or dead, and in this manner a bare place was formed in the forest. Not far off we saw in the river a beaver den, or as the American sometimes call it a beaver lodge, to which there was a very well trodden and smooth path, which we availed ourselves of to go to and from our boat. Nature appears to have peculiarly adapted these remarkable animals to the large thickets of poplar and willow of the interior of North America, where the whites on their first arrival found them in countiess numbers and soon hastened to sacrifice these harmless creatures to their love of gain. 1926] MAMMALS OF NORTH DAKOTA 109 Ten years later Audubon (1897, p. 76) at Fort Union, wrote in his journal about the beavers " once so plentiful, but now very scarce. It takes about 70 beaver skins to make a pack of 100 pounds; in a good market this pack is worth $500, and in fortunate seasons a trapper sometimes made the large sum of $4,000." Already the quest for rich fur harvests had swept beyond this region, but fortunately, where the beavers had the protection of the deep water and high banks of the larger rivers, it had not quite ex- terminated them. With characteristic tenacity they still cling to their old haunts or merely scatter out to establish new colonies in tributary streams, but the love of gain has not entirely disappeared from the land and these new colonies are rarety able to keep their coats on their backs for any great length of time. At Buford, in 1910, Anthony reported a few in the Missouri and Yellowstone Rivers, apparently about as many as were found there in 1887. In 1913, Doctor Bell and the writer found many signs of their presence along the Missouri River near "VTilliston and about 18 miles to the southeast, on the west side, found a dam where a few were living in a creek. At Fort Clark, in 1913, Jewett reported beavers common along the Missouri River and one colony located on a small creek about a mile south of the town. The willows had been cut for houses and dams, and some were also scattered along the river shores, where they had been used for food. In the Killdeer Mountains, Jewett reported beavers common in all suitable creeks in the region; there was a small colony on Jims Creek, 3 miles south of Oakdale, and another colony on Charlie Bob Creek on the east slope of the mountains. Their dams and houses were well protected by the owners of the land. At Medora he saw several fresh cuttings along the banks of the Little Missouri and beavers were reported to him as common in places above there. From Medora down the river to Quinion, Jewett found beavers in several localities along the Little Missouri and on Magpie Creek. In the river at the mouth of Magpie Creek a few had been caught the previous fall, and on Magpie Creek, near Quinion, a beaver dam of aspen, willow, and chokecherry bushes had been built across the creek. The dam was about 8 feet high and 20 feet long between the creek banks and had formed a pond from 5 to 8 feet deep and half a mile long. The colony had been there for several years and was well protected by the ranchers. In the deep ponds of the Little Missouri River, near what was then the North Dakota National Forest, about 25 miles south of Medora, Doctor Bell and the writer found where beavers had been cutting cottonwood trees and building houses on the banks. Just below the camp they had a large house on the bank of the river made mainly from the branches of several cottonwoods which they had cut down near by. The largest tree cut was about 10 inches in diameter, and others still larger had been cut half way through or the bark eaten from one side. Only cottonwoods and willows had been taken, and as these were abundant and of little value the beavers were not doing serious damage in this section. Along Deep Creek, on the national forest, where there was no timber and only willows and chokecherry bushes, the beavers had made numerous dams and 110 NORTH AMERICAN FAUNA [No. 49 some good-sized ponds. On Bullion Creek, south of Sentinel Butte, a colony had built a dam of willow and chokecherry bushes and maintained a large pond, which kept the creek flowing throughout the year where it had formerly gone dry in summer. In Apple Creek, just east of Bismarck, in 1914, beavers were reported to have destroyed $1,000 worth of timber. To get at the facts, a trip was made to Bismarck and their work all along the stream carefully examined. The beavers were not numerous at that time, but the half dozen old dams that had been cut and broken out showed that the animals had previously been there in much greater numbers. In a distance of about 6 miles, the writer estim- ated 15 to 20 beavers, including two families of young, but there had probably been twice as many the previous year. In all about 75 stumps of small trees that had been cut down were found mainly elm and ash, but 1 oak and 1 boxelder had been cut and 1 cotton- wood had been girdled and killed. Most of these were not 5 inches in diameter, and they would average about 2 inches. Most of the wood, probably 3 or 4 cords, had been hauled to the ranches. The majority of the bushes cut were diamond willows and chokecherry, which are used both for food and for building dams and houses. The actual value of all other timber cut along this creek would not exceed $20. In a prairie country where timber is scarce every little tree has a value for shade and protection as well as for the relief it gives to the monotony of open county, but the beavers also add life and interest to the country, and in addition have a cash value usually greater than that of a few small trees. Other complaints were made of damage done at the same time by beavers along Sweet Briar Creek, just west of Mandan, but when Doctor Bell went to investigate he found a few small trees cut for food and building purposes, but very few beavers were left. Most of them had been caught and the trappers and farmers were clamor- ing for permission to catch the rest. In other places, however, the beavers are given adequate protection by residents who are inter- ested in having them on their farms. In 1915 Kellogg found traces of beavers along the Missouri River and Antelope Creek near Goodall, and reported a fair-sized colony near Expansion, a large colony below Independence, a freshly built dam across Deep Water Creek below Shell Village, and another colony on a lagoon at Armstrong. On the Knife River he found two beaver houses, and near Sather, in Burleigh County, a few houses and some fresh beaver work. Near Sawyer he reported one small colony, and another in a bend of the river near Painted Woods, while from there to Bismarck he found the houses at almost every bend of the river where there were groves of diamond willow and small cottonwoods. At Cannon Ball, in 1916, the residents said that there were still some beavers alon» the Missouri River and also along the Cannon- ball River, its side streams, and old sloughs and channels. At Parkin, about 8 miles above the mouth of the Cannonball, there were a number of beavers in the deep parts of the river, with dens in the high banks. They were cutting willows and cottonwood brush along the shores. One evening as it was getting almost dark a big old fellow came up on the bank of the river and, climbing out 1926] MAMMALS OF NORTH DAKOTA 111 on a stump, reached up and quickly cut off a cotton wood branch about 6 feet long, dragged it to the water, and then swam down the river, towing it after him, eating it under cover of a steep bank below. Farther up the Cannonball, at Wade, in 1913, W. B. Bell reported a considerable number of beavers in both branches of the river and photographed a dam on the south fork just above the juncture of the two streams. They had done some damage here by cutting down cottonwood trees up to 18 inches in diameter. One ranchman, Mr. Twigg, estimated that 300 trees had been cut on his ranch. On October 23, 1910, O. N. Dvergsten wrote to the Biological Survey from still farther up the Cannonball, near Stowers, inquiring what he could do with beavers that were destroy- ing his little trees along the creek. A few of the animals had come there the previous year, built their winter home, and kept on build- ing and cutting his trees in spite of his efforts to discourage them. Their house had been torn out, but they had rebuilt it and insisted on remaining. In 1919, after two years of open season on beavers, many of the colonies had disappeared or had been sadly reduced in numbers. A few traces of their work were found along the Missouri River at Sanish and Bismarck, and there were said to be a few beavers still in Apple Creek and Burnt Creek. Near the mouth of the Cannon- ball River they were very scarce, although they had been fairly com- mon up to 1916. In a deep loop of the Heart River near Mandan late in October there was still a small colony. Here they had cut down a few scrubby cottonwoods and a large number of willows along the bank and had stacked the green branches and sections of trunks in deep water for winter food. The top of this mass of green wood and brush reached to the surface and was securely held together by sev- eral inches of ice. There was one beaver house on the bank and many burrows and dens in the steep banks, which were about 15 feet above the water. Several vent holes opened out from 50 to 80 feet back from the river and warm air was steaming out of them on cold mornings. These beavers were well located for an experimental beaver farm or for a wonderful city-park colony at the edge of Mandan. Beaver houses. — Large beaver houses are .often built out in ponds where the surrounding water is 6 or 8 feet deep, with walls of matted sticks and mud rising 4 or 5 feet above the surface of the water, inclosing safe and comfortable living rooms. The nest chamber, usually just above the water level, has its only doorway leading down through deep water under the house to the pond outside. Bank houses are generally smaller but equally well-built struc- tures of sticks and logs well plastered with mud. They are com- monly built on low banks to protect the dens from outside enemies. In high banks the burrows generally enter water and come up well back in the banks into nest chambers that are unmarked by any external building material. Beaver dams. — The dams are generally built of brush, sticks, limbs, and trunks of trees that have been cut into sections of a convenient size to be carried, dragged, or floated to the desired spot, pushed into place, and covered with mud from above the dam. Well-built dams 112 NORTH AMERICAN FAUNA [No. 49 show a steep lower face of crisscross sticks and a sloping upper face of mud or firmly packed earth. They offer a wonderful resist- ance to floods and the wear of time, and many old beaver dams may be found to-day that have not been used for a century or' more. On small streams beaver dams are usually of a simple type, built across the channel so as to raise the water above them to sufficient depth for good ponds. A depth of 6 or 8 feet is required to protect the houses, dens, and bank burrows, and to insure a winter swimming pool under the ice. Much deeper water' is preferred and the beaver will usually leave and hunt for better quarters if a depth of a least 6 feet can not be maintained. Large and rapid streams are rarely dammed, except by large colonies of beavers left undisturbed for a long term of years. Some of the old dams show great skill and industry, but the best results seem to be due to persistent efforts in the face of many failures, rather than to the high order of mentality usually attributed to the beavers. Food habits. — The food of beavers varies with the season. In summer it is mainly grass and other green vegetation. At Apple Creek, in August and September, the beavers were feeding on coarse water grasses and sedges along the shores of the creek. The grass blades were scattered over the surface of the ponds and lodged against the dams and in many places the banks were well cropped. All of this was waste material that could not be cut for hay or grazed by stock. The stomachs of the beavers collected contained large quantities of green pulp, apparently of this material, with the addi- tion of a little of the bark and twigs and roots of willow, and some other plants that could not be identified. The trees and bushes cut at that time had been used mainly for building material rather than for food. In fall beavers begin to cut down bushes and trees to be stored under water for winter food. Sometimes tons of green brush mixed with limbs and sections of tree trunks are sunk to the bottom in deep, still water, where under the ice it keeps fresh and green and is available all winter. The bark is eaten off the larger stems and the twigs and buds are browsed where they lie or are carried into the houses to be enjoyed at leisure. That willows are the principal winter food, as well as the favorite building material, is evident from the food stores, the remains of meals and structure of houses and dams. Cottonwoods and aspens are preferred for food where available. The hardwoods — elm, ash, boxelder, birch, and even oaks — are sometimes cut for building ma- terial, but rarely for food. On Apple Creek, some elm and ash, one small bur oak about 2 inches in diameter, a small boxelder, a thorn- apple bush, and a few hop vines had been cut, all of them evidently for building material, as they showed no indications of having been eaten. Boxelder and bullberry bushes were abundant along the stream, but were rarely touched by the beavers. One thorn-apple bush full of red fruit had been cut and placed on the dam. The rootlets of willows, which grow in dense masses under water along the banks, are also a choice food for both summer and winter, and in deep water, where beavers are scarce and timid, they get much of their food from these tender roots without exposing themselves on the surface. 1926] MAMMALS OF NORTH DAKOTA 113 Breeding Tidbits. — Usually four to six young are raised at a time and it is doubtful if more than one litter is raised in a year. Increase is therefore not rapid and the young do not get their full growth for several years. Beaver parks. — Near Jamestown, in 1914, W. B. Bell visited a beaver colony that had been protected for a number of years and allowed to build a good dam across the Dakota River. The animals were comparatively tame and could be watched at their work on the dam or on the banks, or swimming about in their pond during the daytime, and were a source of much interest and pride to the community. The beginning of a valuable and educational zoological park was here developing spontaneously without any expense or trouble be- yond the mere protection of the animals. Unfortunately, a grain- field extended down to one edge of the beaver pond and naturally the beavers accepted the grain as a part of their food supply. Even af- ter the grain was cut they pulled the bundles out of the shocks and carried them to the water for food and building material. The loss of grain, though scarcely appreciable, naturally irritated the owner and roused a sympathetic feeling for him and against the beavers, until, as a result, the colony was destroyed. If a woven-wire fence had been placed along the river bank and woven wire wrapped around the bases of a few trees, the beavers might have remained as a harmless and delightful interest for the public. No more interesting or simple and inexpensive zoological park can be maintained by any community than a good beaver colony. Beaver farming. — In many sections of North Dakota conditions are excellent for raising beavers under control and partial or com- plete domestication in small lakes or ponds or in fenced sections of creeks and small rivers on owned or leased land. If beavers were included in the list of fur-bearing animals permitted to be raised under special license (North Dakota, 1923, pp. 317-318), a valuable industry might be added to the State, and much waste and unprofit- able land made to yield returns to the owners. The selection of stock for beaver farming is of great importance, since the dark, richly colored animals, as found in the Hudson Bay drainage or, still darker, from northern Michigan and Wisconsin, have far greater fur value than the light-brown beavers of the Missouri drainage, and as far as possible should be used for breeding stock. _ Beaver meat. — If properly prepared, beaver meat is good and wholesome. In the adults it is dark, tender, rich, and of good flavor. There is usually a layer of fat over the surface next to the skin, and the tail is always of a soft, fatty tissue which if well cooked is especially delicious. Among the trappers beaver tail has always been considered a luxury equal to buffalo tongue. Lewis and Clark (1893, p. 276), in their journal of April 17, 1805, say, "Around us are great quantities of game, such as herds of buffalo, elk, antelopes, some deer and wolves, and the tracks of bears. * * * We obtained three beavers, the flesh of which is more relished by the men than any other food which we have." This is almost the unanimous testimony among trappers. In skinning the beaver care must be taken not to get on the flesh a trace of musk from the large gland located under the skin of the 114 NORTH AMERICAN FAUNA [No. 49 belly. The beaver should be hung up by the head and skinned with- out touching- the meat with the hands. It is impossible to handle the skin without getting the hands scented by this very clinging, al- though not unpleasant odor.17 Family ERETHIZONTIDAE : Porcupines Erethizon epixanthum epixanthuin Brandt Yellow-haired Porcupine; Rocky Mountain Porcupine Pahi of the Mandans (Will) ; Pahi of the Dakotas (Gilmore) ; Apadin of the Hidatsas (Matthews) ; Suunu of the Arikaras (Gilmore). Erethizon epixanthus Brandt, Mem. Acad. Imp. Sci. St. Petersbourg, t. 3 (ser. 6), pt. 2 (Sci. Nat.), p. 390, 1835. Type locality. — Northwestern America. General characters. — Heavy, wide-bodied, short-necked, short-legged ani- mals with short, stout tails, long curved claws, flat, naked soles and an armor of quills ; upper parts densely covered with very keen barbed quills, embedded in black fur and partly concealed by long yellow-tipped outer hairs ; underparts mainly without quills. An adult male from Montana measures in total length 875 millimeters; tail, 314; hind foot, 112. "Weight, approximately 20 to 30 pounds. Distribution and habitat. — From a wide range in the Rocky Mountain region the yellow-haired porcupines reach their eastern limit, so far as known, in North Dakota. They are fairly common in the Missouri Valley and westward in the State, but east of the river valley they are rare and scattered. A specimen collected by U. S. Ebner in the Turtle Mountains in 1914, and now in the collec- tion of the North Dakota Agricultural College, at Fargo, marks the easternmost authentic locality for the species. Near Warwick, just south of Devils Lake, in 1915, Kellogg reported a yellow-haired porcupine killed by two boys the previous year; at Towner on the Mouse River, one killed by Almond Larson in 1905, and another found dead by Clyde Coss in 1911. In 1913 there were reports of porcupines having been killed near Kenmare and Minot, but there was no real clue to the form represented. It was undoubtedly, how- ever, the yellow-haired. At Buford and all the way down the Mis- souri River through the State porcupines have been reported com- mon from 1910 to 1915 by Anthony, Kellogg, and Jewett, and ap- parently their numbers have not changed much since the days of Lewis and Clark, Maximilian, and Audubon. In 1913, Doctor Bell reported them fairly common at Wade, on the Cannonball River, where two had been recently killed near Mr. Wade's ranch and a skull of one obtained for a specimen. In 1919 they were found com- mon about Sanish, in the brushy gulches on both sides of the Mis- souri River. General habits. — Although well safeguarded by their own spiny armament, the porcupines often seek additional safety in the Bad- lands and brushy stream bottoms, in the protection of little caves and hollows in the banks or the dense, thorny cover of buffaloberry thickets. Near Williston, the writer found their characteristic oval 17 For further information on the habits and control of heavers, see U. S. Dept. Agr. Bui. 1078 (Bailey, 1922) and Misc. Circ. 69 (Bailey, 1926). 1926] MAMMALS OF NORTH DAKOTA 115 pellets in the little caves of the Badlands, which seemed to be their favorite dens. Often, however, the animals are met in the open and at night they follow trails and roads for long distances, as shown by their double rows of oval, flat-footed, denticulate tracks in the dust. Although their ordinary gait is not much faster than that of the turtle, they are patient and persistent travelers and sometimes their tracks may be followed for miles. When met with, the porcupine usually attempts to escape, but if crowded, bristles up, erects its quills, and stands at bay awaiting attack. The quills are pointed out at all angles and as the enemy approaches within reach, fierce blows of the heavily armed and muscular tail are struck sideways or upward and the barbed quills thus driven into anything within reach. The common belief that the quills are thrown to a considerable distance has no foundation in fact, although some are occasionally scattered on the ground if the animal is roughly handled. Porcu- pines evidently realize that their lower surface is unprotected, as any effort to turn them over is frantically resisted, and when threat- ened the quickness with which they will wheel and strike is surpris- ing in animals so clumsily built. Their long, very hooked claws enable them to climb trees readily, and the animals are as much at home on the trunks or branches as on the ground. They also climb about in the bushes and seem to enjoy the tops of the very spiny buffaloberry bushes, which prob- ably* give them a feeling of added protection along their own lines of defense. The tops of these bushes are often eaten bare of bark, leaves, and berries and left in a very mutilated condition. The writer has never seen any evidence that porcupines dig burrows, but quite probably they dig out or enlarge some of the cavities in which they dwell. Breeding habits. — The mating season is said to be in October and one or sometimes two young are born early in spring. At birth the young are unusually large and well developed; their eyes are open, and they are provided with a good set of fur, quills, and incisor teeth. They follow the mother until weaned and apparently before they are half grown each one is able to shift for itself and to begin its solitary life. With this slow rate of reproduction the species would soon disappear but for its armored protection. Food habits. — During the summer, porcupines feed on a great variety of green vegetation, accepting apparently almost anything that comes in their way and stuffing their enormous stomachs to the limit of their capacity. At Stanton, Kellogg found one feeding in an alfalfa field with its stomach well filled with alfalfa; he said they were reported to do some damage in the grainfields between Washburn and Bismarck. Jewett reported them as fairly common in the brushy gulches near Sentinel Butte, where they had gnawed the bark from many of the chokecherry bushes. Near Sanish they had eaten the bark and twigs from buffaloberry, black haw, choke- cherry, and rose bushes. In 1913, on the former Dakota National Forest, about 25 miles south of Medora, they were found fairly common in the Badlands gulches and on the forested ridges. Many of the yellow pines had been gnawed more or less extensively by 116 NORTH AMERICAN FAUNA [No. 49 them. On some of the forested ridges about half of the small trees showed peeled spots from which the bark had been eaten and some had been completely girdled and killed. Most of the old trees showed some scars from earlier gnawings. Still farther south, along the Little Missouri, near Marmarth, where yellow pines grow irregularly over the buttes, the writer found fully a fourth of the young trees damaged through having the bark gnawed from them by porcupines.' In some cases the bark had been eaten from the tops and branches; in others the trunks had been girdled, so that many of the trees were either ruined or killed outright. The old pines showed a long struggle with their enemy, the bushy tops and gnarled forms being largely due to the girdling of tops or branches at different times during their lives. Here, as in many other parts of the country, the bark of yellow pines seems to form the favorite food of the porcupines, at least during the winter season. The rough outer coating of bark is rejected and the tender inner growth eaten as it is scraped clean from the wood of the trunk. Apparently the bark from a space the size of a hat is required for a square meal. Any tree that happens to be conveniently near the porcupine's den is sure to suffer and may be stripped of all of its bark from top to bottom. Economic status. — Although most Avild carnivores have become sufficiently accustomed to porcupines either to let them alone or, by taking advantage of their unprotected bellies, to kill and eat them with little harm to themselves, many dogs gain their first knowledge of the species by sad experience. The greatest complaint of the set- tlers against the porcupines comes from this injury to their dogs, for if a dog attacks one recklessly as it would any other animal it may be seriously or fatally injured by the quills. The destruction of crops by porcupines is usually of small consequence, but their destruction of many species of pines and other conifers often causes great loss to the forests within their range. It is not improbable that they are largely responsible for the scarcity of timber in the Badlands region ; were it not for them a fair stand of pines might have spread over this rough country. If reforestation of these areas is attempted, it will be necessary to first eliminate the porcupines, as where they are common no young trees can reach a well-developed maturity. Erethizon dorsatum dorsatum (Linnaeus) Black-haired Porcupine; Canada Porcupine [Eystrix] dorsata Linnaeus, Syst. Nat., ed. 10, t. 1, p. 57, 1758. Type locality. — Eastern Canada. General characters. — Color, black and white instead of black and yellow; upper parts covered with white, black-tipped quills, mixed with black fur and obscured by long black, white-tipped hairs. Usually not so large as the yellow- haired porcupine from farther west. An adult male from Minnesota measures in total length 740 millimeters ; tail, 195 ; hind foot, 115 ; an adult female, 735, 195, and 100 respectively. Weight of female, 16 pounds; of male, probably 20 pounds. Distribution and habitat. — The black-haired porcupines occupy the timbered Canadian Zone area of the northeastern United States and Canada west to the Great Plains, where they probably meet the range of the yellow-haired porcupines. They are common in northern Minnesota, but for North Dakota there seem to be only two or three 1026] MAMMALS OF NORTH DAKOTA 117 probable records and these unsubstantiated by specimens. M. A. Brannon writes that while at the university, at Grand Forks, he had a small black-haired porcupine for a pet, but it met with an untimely death and was not preserved for a specimen. It was given to him and was said to have come from the Red River Valley, near Pembina. H. V. Williams reports a porcupine of the small dark-colored type, almost black, killed at Hamilton, in Pembina County, on July 31, 1916. The description fits this species, which on geographic grounds ought to be found there rather than the large yellow-haired species which has been taken no farther east than the Turtle Mountains; but the young of both species are blackish, so that identification depends in part on age. The boys at the Indian school near Wahpe- ton killed a porcupine on the river bank near town in 1914 and described it but no specific characters could be gathered from the description. Others will probably be found along the Red River Valley, and it is hoped that a specimen may be preserved to de- termine the species positively. Family ZAPODIDAE : Jumping Mice Zapus hudsonius campestris Preble Prairie Jumping Mouse (PI. 13) Zapus hudsonius campestris Preble, North Amer. Fauna No. 15, p. 20, 1899. Type locality. — Bear Lodge Mountains, Wyo. General characters. — A medium-sized mouse with very long, slender hind legs and feet and small front feet ; tail, very slender ttnd longer than head and body; ears, small. Upper parts, bright buffy yellow along sides, darker along the back ; underparts, pure white. Average measurements : Total length, 222 millimeters ; tail, 135 ; hind foot, 30.5. Distribution and habitat. — As its name implies, the prairie jump- ing mouse is a plains species covering practically the whole of North Dakota and the surrounding prairie country. There are specimens in the National Museum from Wahpeton, Fairmount, Blackmer, Han- kinson, Ellendale, Fargo, Harwood, Lisbon, Pembina, Neche, Turtle Mountains, Devils Lake, Fort Totten, Valley City, La Moure, Ludden, Cannon Ball, Fort Clark, Grinnell, and Buford. Speci- mens have also been recorded in the Field Museum from Bottineau, Minot, and Jamestown. Although generally distributed over the State, these jumping mice are found mainly in thickets, weed patches, meadows, or tall grass areas rather than on the high open prairie, where the grass is short and the cover scant. General habits. — Under the protecting cover of bushes, weeds, and tall grass, these timid little jumping mice make their summer homes on the surface of the ground and their winter homes in burrows deep underground. They do not make roads or runways, but go through the grass with long leaps or little hops and occasion- ally with a slow creeping motion on all fours. When startled, they go bounding away with long jumps, suggesting frogs, and usually make two or three leaps before stopping to see if they are pursued. Generally, if the last leap is well noted, one can creep up cautiously and catch the mouse by clapping the hand over it. When caught in this way the mice rarely offer to bite or make much effort to escape, 118 NORTH AMERICAN FAUNA [No. 49 but may be handled and examined freely if held gently in the hollow of the two hands. Evidently they are not entirely nocturnal, as they are often startled from their feeding grounds in the daytime, but more often they are disturbed in their nests, from which they bound away when one steps close to them in the grass. The summer nests are placed on the surface of the ground, well concealed under grass or other vegetation; they are neat little balls of fine grass with a tiny opening at one side arid a soft lining in the central chamber. When the grain is cut and the hay mowed the nests are disturbed and the jumping mice go to live in the shocks of grain and cocks of hay, where they are discovered when the hay and grain are being loaded on wagons. As they bound from under cover to the open ground they are somewhat dazed by the light and can usually be watched for some time as they sit blinking in the open or progress by long leaps through the air. Hibernation. — Unlike most of the mice, these little fellows be- come excessively fat in autumn and with the first frosty nights re- tire to their warm underground nests and curl up for a long winter's sleep. The thin oily fat is deposited in a layer of white fatty tissue over the whole inside of the skin as well as over much of the surface of the body and fills the inside cavities until the animal is about twice its natural size and weight. This fat supplies sufficient nutri- ment and fuel for the long winter sleep and probably carries the animal through the early springtime of breeding activities when food is scarce. Breeding habits. — The five or six young are brought forth in the nests usually in May or June, and are barely full grown by the time their winter sleep is to begin. In this latitude it is doubtful whether more than one litter of young is raised in a summer. Food habits. — In the examination of a great many stomachs of these jumping mice, nothing has been found but the fine white pulp of carefully shelled, well-masticated seeds. Generally these are from grasses, although grain and a variety of other plant seeds are eaten. The mice are fond of rolled oats used for trap bait, and are easily caught in a variety of traps set where they are in the habit of run- ning. To obtain the seeds of grass, on which they mainly subsist, they cut off the tall stems as high up as they can reach, draw them down and cut them off again, and repeat this until the seed-laden tops can be taken. Little heaps of grass stems cut in sections about 3 inches long are found through the meadows where the jumping mice live and are unmistakable evidence of their presence, being al- ways much longer than the grass cuttings of meadow mice and other short-legged species. Apparently these rodents do not store up food, but live a very care-free life in the midst of abundance while the summer lasts. Economic status. — Generally the jumping mice are not sufficiently abundant to do any great harm to the yield of grass and grain, but in places over limited areas in the meadows their cuttings might ag- gregate 2 or 3 per cent of the grass. They cut down and eat or de- stroy a small quantity of grain along the edges of some fields, but on the whole are far less numerous and injurious than the meadow mice. Still, they help to swell the total of the tax levied by rodents on farm products and only fail through lack of numbers to form one 1926] MAMMALS OF NORTH DAKOTA 119 of the serious rodent pests. Their natural enemies are the same as those of the other nocturnal mice, chief of which are owls, weasels, badgers, and skunks, through the good offices of which their num- bers are kept within bounds. Family HETEROMYIDAE : Pocket Mice, Kangaroo Rats Perognathus fasciatus fasciatus Wied Maximilian Pocket Mouse Apapsd of the Hidatsas, Zhxzhina of the Dakotas (Gilmore). Perognathus fasciatus Wied, Nova Acta Acad. Caes. Leop.-Carol. Nat. Cur., t. 19, pt. 1, p. 369, 1839. Type locality. — Upper Missouri River near its junction with the Yellow- stone, northwestern North Dakota. General characters. — Considerably smaller than the white-footed mice, with small ears, slender tails, and conspicuous fur-lined pockets on the cheeks, opening externally and not connected with the mouth ; hair, short and glossy ; upper parts, olive gray; underparts. pure white, bordered by a buffy line along each side. Average measurements: Total length, 135 millimeters; tail, 65; hind foot, 17. Distribution and habitat. — Maximilian pocket mice are scattered over a large part of western North Dakota and adjacent areas of the semiarid plains. There are specimens from Buford, Crosby, Minot, Dunseith, Fort Clark, Cannon Ball, Wade, Dawson, Oakes, Bowdon, and the Little Missouri River north of Medora, but the range is probably more extensive and continuous than these scat- tered localities indicate. They are animals of the open prairie, where they live in tiny burrows in the barest situations or on the short-grass plains, for^ unlike most mice, they avoid the cover of vegetation. General habits. — In 1833, Maximilian, Prince of Wied (1839, p. 373), found this anomalous little pocket mouse near Fort Union, at the junction of the Yellowstone and Missouri Rivers, and in 1839 first described it as a new genus and species of rodent. For more than 50 years no more specimens were obtained, and the name was confused under another species and not put in its proper relationship until 1889, when Doctor Merriam (1889, pp. 2, 4, 11) published his revision of the North American pocket mice. In 1887 the writer visited Fort Buford and collected a small series of specimens that served to verify Maximilian's excellent de- scription of the genus and species. At that time he was unac- quainted with animals of their general habits and had only common steel traps, old-fashioned choke traps, and little tin box traps, and knew of no more tempting bait than cheese, bread, cake, or meat, none of which they would touch, so that although instructed to look out for them and get specimens, he then failed to catch any in his traps. Their characteristic little burrows were common and their tiny tracks recognized as undoubtedly belonging to the species, were found every morning about the traps, although no attention was paid to the bait. Other methods were evidently necessary to obtain specimens. During the dusk of evening as the writer walked over the prairie, sometimes one of these little mice would dart over the ground near 120 FORTH AMERICAN FAUNA [No. 49 him, and by dropping his gun and making a quick spring he could catch it in his hands. All but one of those seen were caught in this way, but in that one case his fingers came down on the mouse's tail and the rest of him escaped. This kind of hunting, though exciting at times, nearly ruined his shotgun, which invariably was dropped on the ground at the first move of the mouse, although he resolved each time to lay it down carefully when the next one was seen. The half dozen specimens obtained served to reestablish the identity of the species, but they did not add much to knowledge of its general habits. Only in later years was it learned that with modern traps baited with rolled oats the mice could be caught in abundance wher- ever they occurred; since then naturalists have been able to learn more of their habits. Their little burrows are usually found in groups of two or three on some dry, open spot, often at the edge of a cactus or sunflower patch, or close to sagebrush, and are easily recognized by their very small size. A little fresh earth is occasionally found thrown from some of the burrows, but in most cases the entrances are unmarked and inconspicuous. In 1910, Anthony collected specimens at Fort Buford and reported burrows found sparsely on the prairies and hilltops, usually in the sides of banks or slight elevations. One specimen was taken in an open space in the sagebrush near the river. At Crosby, in 1913, the writer caught one under some old Russian thistle at the edge of a flax field. At Minot, on October 12, 1919, he tracked one over a soft snow from a strawstack to a hole under a furrow, and digging back about 2 feet found it in a cup- shaped nest of soft plant fibers, captured it alive, and kept it for several months for study. At one edge of the nest cavity it had a small collection of seeds, mainly pigeon grass and Russian thistle seeds, which proved its favorite food in captivity. At Fort Clark, Jewett found these mice fairly common about the wheatfields and high dry prairies back from the river, where they were readily taken in traps baited with rolled oats and set near the small bur- rows. At Cannon Ball, Sheldon found them common in the grain- fields on the sandy places and along the flats of the river. At Wade, farther up the Cannonball River, W. B. Bell collected a specimen for the agricultural college museum. In 1892, Theodore Roosevelt caught a specimen on the Little Missouri River, 40 miles north of Medora, which he contributed to the Biological Survey collection. Small, inconspicuous, and mainly nocturnal in habits these little pocket mice, even where most abundant, generally escape the notice of all but naturalists or keen observers. It has remained for a local naturalist, Stuart Griddle (1915), of Treesbank, Manitoba, to study their habits in a careful and thorough manner. In ex- cavating their winter burrows he learned more of them than was ever known before. He found their burrows penetrating as far as 6 feet below the ground, where the winter nests and stores were well protected from frost. Apparently enough seeds were provided to carry them through the winter. Their winter stores con- sist mainly of seeds of noxious weeds, and Criddle's conclusions were that the mice are mainly beneficial in their foods habits. Such careful studies of mammal habits by local naturalists are of ines- timable value for the better understanding of native species. 1926] MAMMALS OF NORTH DAKOTA 121 Hibernation. — These mice are rarely if ever found with suffi- cient accumulation of fat to suggest hibernation, but Criddle says that when exposed to moderately cold atmosphere they become very sluggish and he thinks that they spend much of the winter in sleep. The writer has found them active up to October 6 in Mon- tana, and to October 12 in North Dakota, and they have been taken even later farther south. A captive specimen was active well into the winter, but in a warm house. The question of hibernation is not yet fully settled. Breeding habits . — A female caught on May 13 contained six embryos, and Criddle reports one containing four. The mammae are arranged in two pairs of inguinal and one pair of pectoral on four distinct mammary glands. It seems probable, therefore, that six is the normal maximum number of young. There are no data to indicate more than one litter in a j^ear. Food tidbits. — In 1887 these pocket mice were found feeding mainly on the seeds of pigweed and knot grass, and at Crosby in 1913, they were living under the Russian thistle, which apparently furnished them food as well as cover. At Buford, Anthony reported their pockets filled with small angular seeds, which were probably of knot grass, and at Fort Clark, Jewett reported several caught at the edges of wheatfields with grains of wheat in their pockets. Others have been taken with their pockets filled with grass seeds, lambs- quarters, red root, and tumbleweed, and Criddle found in their homes and pockets seeds of grass, blue-eyed grass, bug seed, wild buck- wheat, and puccoon. He also discovered grasshopper eggs stored in their tunnels and found many places where these had been dug out of the ground. One of the mice that he kept in captivity pre- ferred meal worms to seeds. Economic status. — From the evidence gathered it seems that these mice are very slightly, if at all, harmful, while in many ways they are decidedly beneficial; but there still remains much to be learned of their habits and tastes. Perognathus flavescens perniger Osgood Dusky Pocket Mouse (PI. 15) Perognathus flavescens perniger Osgood, Proc. Biol. Soc. Washington, vol. 17, p. 127, 1904. Type locality. — Vermilion, S. Dak. General characters. — About the size of fasciatus, but more intensely col- ored, with the rich buff on the upper parts much obscured by a wash of bright black, and the underparts chiefly rich, buffy ochraceous. Measurements of type: Total length, 140 millimeters; tail, 68; hind foot, 17. Weight of live adult, 10 grams. Distribution and habitat. — The silky little dusty pocket mice come into southeastern North Dakota from their range over the prairie country of western Minnesota, eastern South Dakota, and the adjoin- ing corners of Nebraska and Iowa. There are specimens from Hankinson, Blackmer, Lidgerwood, Napoleon, and Finley, and the writer picked up a dead one in the town of Parkin, about 10 miles above the mouth of the Cannonball River, too mangled to be saved for a specimen. The range of the form somewhat overlaps that of 122 NORTH AMERICAN FAUNA [No. 49 fasciatus, from which it is entirely distinct. Apparently this is merely a dark-colored prairie form of the paler flavescens of the semiarid Plains region farther south and west. Sandy prairie soil is their favorite habitat and their little burrows are usually found in the mellow and often barren soil among prairie grasses. General habits. — In the old lake-shore sand dunes, a little south of Hankinson, these little animals were found fairly, abundant. On the crests of many of the low ridges or mounds that had once been dunes, from one to a half dozen of their burrows or groups of burrows were found. There was generally a little mound of sand like a small gopher hill, and, whether freshly made or old, the entrances to the burrows were invariably closed. Often two or three other burrows, just large enough for the end of the finger, would be found near the closed one, but these were inconspicuous and rarely showed any trace of dirt having been thrown out. Traps baited with rolled oats and set at any of these holes, or across a long trail made by scraping the foot in the sand, readily caught the mice, for while they do not make trails of their own, they invariably follow any clear road through the grass. Often in the morning their tiny tracks were found over the open, drifting sand. A few speci- mens were taken in traps set near the tracks which led from the burrows to the feeding grounds. Although more easily located on the open sand, the mice were much less numerous there than in the scattered vegetation, which afforded some cover. At Blackmer two were caught in a sandy field where boys said mice were often turned out by the plow. At Lidgerwood, Sheldon found them common in the grainfields and a series of specimens was taken in traps set in the fields. At Parkin the writer found many of their characteristic burrows and tracks in sandy ground near the edge of the town that had just sprung up on the prairie and picked up a dead mouse in the grassy street. Breeding habits. — Three females collected at Elk River, Minn., on July 30 and August 12, 1912, contained four embryos each. The mammae are arranged in two pairs of inguinal and one pair of pectoral, which for the present constitutes our total knowledge of the breeding habits of this species. Food habits. — At Hankinson the traps were baited with a mixture of rolled and whole oats, but as ants carried away most of the rolled oats during the day the whole grain was usually the only attraction for the mice. Most of the specimens caught had in their pockets some of the whole oats, from which they had removed the hulls, and some had also the seeds of needle grass {Stipa spartea), while the pockets of others were entirely filled with these long grass seeds, hulled and neatly packed in little bundles. There were occasionally also a few seeds of bindweeds and small wild beans. Of course, their food varies with the time of year, and at this season, July 19 to 27, the abundant Stipa seeds were just falling to the ground and the mice were busy gathering their harvest. At Lidgerwood, Sheldon found that the pockets of all of those caught in wheatfields con- tained weed seeds, with the exception of one that had gathered up a few particles of cracked corn; some of them also had included a few kernels of oats from his trap bait. The one picked up at Parkin had its cheek pouches full of little bean seeds, probably of Astragalus, 1926] MAMMALS OF NOllTH DAKOTA 123 which was common there. In Minnesota the writer found where the mice had been feeding extensively on the seeds of sand bur, one of the most troublesome of weed grasses. In the underground winter storerooms of these mice there were seeds of two species of pigeon grass, a few other grasses, and wild buckwheat. In captivity their favorite food has proved to be first of all the pigeon-grass seeds from their own winter stores, then Rus- sian thistle seed, millet, wild sunflower, hemp, and rolled oats. They nibble a little cabbage, turnip, cooked potato, lettuce, celery, or green grass, but apparently more for the moisture than for food, as in a dry, furnace-heated house, they become very thirsty and eagerly suck water from saturated cotton or drink from a small dish. None of the animals caught showed any indications of becoming fat as in hibernating species, but it is evident that they store up much food in the form of small seeds. Economic status. — Too scattered in their distribution to be of any serious consequence one way or another, the habits of these little mice appear to be mainly harmless. Their consumption of weed seeds probably counterbalances any possible mischief in grainfields. Perognathus hispidus paradoxus Merriam Kansas Pocket Mouse (PI. 12) Perognathus paradoxus Merriam, North Amer. Fauna No. 1, p. 24, 1889. Type locality. — Banner, Trego County, Kans. General characters. — Size, large; tail, long; ears, small; pelage, glossy but coarse and hispid ; external cheek pouches, conspicuous ; upper parts, yellowish-brown with scattered black hairs over the back ; sides, clear yellow- ish ; underparts, white. Average measurements of adults : Total length, 222 millimeters; tail, 10S ; hind foot, 20. Distribution and habitat. — These large pocket mice have an exten- sive range from Mexico over the Lower and Upper Sonoran semi- arid plains region to western South Dakota, and one specimen has been taken in North Dakota. This was collected by Doctor Bell, in August, 1913, at "Wade, on the Cannonball River. The specimen is now in the agricultural college collection, at Fargo, and is of special interest as marking the northern limit of the known range of this species. It is a large female, measuring in total length 220 milli- meters, tail 114, and hind foot 27, and was caught in a trap set on the prairie at the edge of a sandy area on the Wade ranch. At this locality the species represents an element of the Upper Sonoran Zone, which is sparingly shown also by the native vegetation. General habits. — Over their wide range these mice are generally scattered and not abundant, but occasionally get into the collector's traps set in open country. They live in burrows of their own con- struction, which are often recognizable by their size and form, as they are larger than ordinary mice burrows and not so large as those of kangaroo rats. Moreover, they often go straight down into the ground like a smooth auger hole, around the entrance of which no trace of earth is found. Always at some place not far away, however, is a burrow at which considerable earth has been thrown 82242°— 26 9 124 NOKTH AMERICAN FAUNA [No. 49 out, showing that the unmarked openings are those that have been opened from below. Sometimes the burrow at which the earth is thrown out is closed at the entrance; at other times it is left open. The underground habits of the pocket mice are little known, ex- cept that specimens taken often have their cheek pouches well filled with seeds, grain, or trap bait, which they are carrying home, evi- dently to be stored for food. They are very fond of rolled oats and are readily caught in traps baited with them. A great variety of seeds is eaten, but the mice do not usually show any signs of accu- mulating fat for winter, and it is doubtful whether they regularly hibernate. Over most of their range farther south they may be caught at any time during the winter. Perodipus niontanus richardsoni (Allen) Richardson Kangaroo Rat (PL 11, fig. 3) Dipodops richardsoni Allen, Bui. Amer. Mus. Nat. Hist., vol. 3 (1890-91), p. 277, 1891. Type locality. — Beaver River, Beaver County, Okla. General characters. — Big head and short body, long brush-tipped tail, long hind legs and feet, small hands, and ample fur-lined cheek pouches combine to produce a most unique and striking appearance. Upper parts, bright buff y-y ell ow with a white band crossing each flank and white spot over each eye; underparts and stripe along each side of tail, white. Measurements of Montana specimen : Total length, 264 millimeters ; tail, 145 ; hind foot, 40. Distribution and habitat. — Richardson kangaroo rats are common in eastern Montana and western South Dakota, and undoubtedly occur in North Dakota, although no specimens have been taken and the only actual evidences of their presence are some groups of burrows described by Doctor Bell, at Wade, on the Cannonball River. He describes groups of large burrows on a strip of sandy ground on the Wade ranch, with considerable earth thrown out around the entrances, exactly as had been found around their dens at Glendive, Mont., and in other parts of their range. The species can only tentatively be included in the North Dakota list, but should be watched for and will undoubtedly be found in a few localities over the western part of the State. The animals can not fail to be recognized, and usually their burrows and the long-paired tracks of their hind feet are unmistakable. General habits. — As indicated by their large, dark eyes, the kangaroo rats are strictly nocturnal, and for this reason are rarely seen except as caught in traps or accidentally driven out of their burrows. They are gentle, timid little animals, depending entirely on speed and their deep dens for protection. In running they hop along on their hind feet, and when hard pressed take flying leaps through the air, balanced by their long, tufted tails. The little front feet are used as hands and rarely allowed to touch the ground. Food habits. — The food of this species consists of a great variety of seeds and grain, which are gathered and carried in the cheek pouches to the dens, to be eaten at leisure. Most of the rats collected for specimens are found with more or less food and sometimes with the pouches distended with various seeds or grains. North American Fauna No. 49, U. S. Dept. Agr., Biological Survey. PLATE 15 Fig. I. — Dusky Pocket Mouse (Perognathus flavescens niger) Photograph of captives kept for study. Slightly reduced Fig. 2. — Badger (Taxidea taxus taxus) "Topsy," a pet at the Agricultural College (photographed by W. C. Palmer) North American Fauna No. 49, U. S. Dept Agr.. Biological Survey. Plate 16 Mississippi Valley Pocket Gopher (Geomys bursarius bursarius) Photographed in the act of digging a burrow in the prairie. About one-fourth natural size 1926] MAMMALS OF NORTH DAKOTA 125 Economic status. — It is perhaps fortunate that these interesting rodents do not reach farther into the State, as in grain-producing country they often levy a considerable tribute on the crops. Where they are abundant, the quantity of grain carried away, eaten, and stored in their dens for future use is sometimes a serious loss. Family GEOMYIDAE : Pocket Gophers Geomys bursarius (Shaw) Mississippi Valley Pocket Gopher (PI. 16) Mus bursarius Shaw, Trans. Linn. Soc. London, vol. 5, p. 227, 1S00. Type localily. — Unknown ; somewhere in the Upper Mississippi Valley. General characters. — Characterized by heavy build, large front feet, and long, heavy, digging claws, conspicuously grooved upper incisors, and deep fur- lined pockets on the cheeks extending back under the skin to the shoulders ; eyes and ears, small ; tail, small and nearly naked at tip ; fur, short, smooth, and glossy. Color, light chestnut-brown above, slightly paler on the belly. Average measurements : Total length, 270 millimeters ; tail, 80 ; hind foot, 35. A large female at Grand Forks measured 290, 75, and 35 millimeters, and weighed 14 ounces. Fig. 7. — Records of two species of pocket gophers in North. Dakota : Triangles, the Mississippi Valley pocket gopher; circles, the Dakota pocket gopher; dot in circle, type locality. The paler sagebrush pocket gopher from the extreme western part of the State is not shown on this map Distribution and habitat. — The Mississippi Valley pocket gophers enter eastern North Dakota and range as far west as Ludden, Oakes, Larimore, Valley City, 10 miles west of Portland, Manvel, Grand Forks, and to the vicinity of Pembina (fig. 7) . At Hankinson, in 1912, the writer found them abundant over the prairies, especially in mel- low, sandy soil. On the high clay hills south and west of Lake Elsie they were scarce or entirely absent from extensive areas. A fondness for mellow soil seems to be a potent factor in outlining the range of the species. At Fairmount, Sheldon found them occupying mellow 126 NOETH AMERICAN FAUNA [No. 49 soil along the river and a few scattered out on the wide, low prairies. At Oakes he found a few and at Lidgerwood they were abundant and very destructive to crops. At Wahpeton, Kellogg and the writer found them common along the river valley, and noted a few over the prairies, where they were doing considerable mischief in grain and alfalfa fields. At Lisbon, in 1912, Eastgate took a few specimens, but reported the animals scarce. In 1892, Loring took specimens at Valley City, Castleton, Wheatland, Buffalo, Erie, Portland, and vicinity. At Fargo, in 1912, their hills were abundant over the val- ley, except on some of the farms where the pocket gophers had been trapped. The hills were large and a long row of them across a green field of young wheat showed up strikingly. At Larimore, in 1915, Kellogg was told by the residents that these large pocket gophers had been the original species there, but the little gray form, Thomo- mys talpoides rufescens, had come in recently; at Manvel, Grand Forks County, he collected a specimen of Geomys and reported it as the common gopher of that region and especially numerous along railroad tracks. At Grafton only Thomomys was caught, but the large hills of Geomys were seen at Minto, 10 miles farther south. In 1916, the writer took Geomys just across the Red River from Pem- bina, where gopher hills were common on a strip of mellow soil, and specimens have been taken at Emerson, just above the Manitoba line. Thus the range has been rather fully worked out and found to extend on the west little beyond the old beach lines of post-glacial Lake Agassiz. General habits. — For a distance of more than 1,000 miles, roughly from Pembina to El Paso, the ranges of Geomys and Thomomys meet without any extensive overlapping, Geomys occupying gen- erally the mellow soil of the fertile valley country and Thomomys the higher, drier, and often more sterile soils to the west. The rea- sons for this division of territory have caused much speculation, and to obtain some evidence on the question, the writer made a special effort to get living specimens of both to test their dispositions when placed together. A live Thomomys was placed in the cage with the larger Geomys. Without a moment's hesitation the old 14-ounce Geomys pounced upon the 5-ounce Thomomys and began to chew it up, catching it by the ribs and crushing its bones, ribs, neck, skull, shoulders, and legs. When convinced that it was entirely dead the Geomys left it and showed no further interest in the victim. Its bones were broken to bits, but the skin was not cut through, prob- ably because the teeth of the Geomys had been dulled on the wires of its cage. This fierce animosity seems to afford a reasonable ex- planation of the division of range"between the two genera, the larger and more ferocious occupying the choice, fertile portion of the country and leaving the rest to its weaker relative. To test further the disposition of Geomys, two that had been caught alive were placed near together, the old female that had chewed up Thomomys, and a half-grown young male caught near her home and quite probably one of her last spring's young that had been long ago sent out to dig its own way in the world. As they met face to face both hissed, struck out with their hands, and clashed their incisors together, the larger forcing the smaller one backward, but they did not clinch, and neither gave the other a chance to get 1926] MAMMALS OP NORTH DAKOTA 127 a hold. Again and again they jumped at each other, hissing and blowing, striking or pushing with their hands, and striking their incisors together with loud clicks, but doing no damage. The smaller animal was constantly forced backward and evidently would have retreated into its burrow had it been within reach. They were separated before any damage was done, but not until they had fully demonstrated the fact that they are not sociably inclined. Later, while in a cold room in a little hotel, the writer placed a meadow mouse, Microtus drummondi, in the glass bowl with the smaller Geomys. Both were chilly and it was hoped they would keep each other warm. At first " Mike " jumped at " Geo " and bit and squeaked at him, but did not stir up any trouble, so he went over to one side of the bowl and made aWst for himself in the grass. It was thought they were going to be friendly and would be company for each other, but later in the evening Mike was heard to squeal ; when the writer reached him Geo was making his bones crack. It was too late to intervene, and when Geo let go Mike was limp and dead. He was left to see what would happen, and in the morning the victim's bones were found broken to bits, although his skin was intact and no attempt had been made to eat him. While the Micro- tus may have started the trouble, for its disposition is not amiable, this further demonstration of the unsociable nature of Geomys is worth recording. Later, after Geomys had become perfectly tame and was no longer interested in eating the writer nor in running away from him, its real nature and disposition were more apparent. It had no objection to being picked up and petted, but if startled would throw up its head as if ready to bite, so that it seemed safer to avoid its nose. If not startled, it would take food and climb into the hand to be taken up and carried about. It would make a large, warm nest in its nest box by carrying in grass, paper, cotton, or any soft material until the box was well filled, then going in would stuff the doorway full and remain buried in its nest, sometimes for 12 or 24 hours at a time. It would sleep longer and eat less in cold weather than in warm. When awake it would insist on strenuous exercise, eating, chewing up nest material, digging, scratching, and gnawing, or, if out of its box, running around the room for an hour' or more at a time at a steady, rapid trot. At first it would butt into every object encoun- tered as it followed the walls around and around, but later seemed to recognize and avoid every obstacle. Finally it became so familiar with the room that it would run in a large circle, missing all the fur- niture, unless something was moved into its path, when it would promptly bump into it. Its eyes were generally kept open while running, but in a lighted room they seemed to be of little help. In the dark it seemed to see well at close range, and when a nut was held in the fingers well inside its nest door it would take the nut gently without touching the fingers with its teeth. This, however, may have been due to the sensitiveness of its abundant short mus- tache more than to sight. Its hearing seemed very dull, except for certain sounds. A touch, scratch, or jar on his house or its nest box would rouse the animal instantly from sleep and put it on the alert, while loud talking, 128 NORTH AMERICAN FAUNA [No. 49 music, or open-air sounds seemed to make no impression on it. At first a puff of air or a door being opened across the room would attract its attention, and it could be stopped in its headlong race across the room by a quick puff of the breath. Mentally it seemed dull and apathetic, although physically power- ful and energetic. It has never shown any play instinct, but was probably too old when captured. The animal was unable to swim. When put in a bathtub half full of water it floated with its head and back well out, but kicked or tried to run with its usual one-foot-at-a-time gait, and made no progress whatever. Apparently pocket gophers are unable to swim, and this may account for some peculiarities in their distribution in other parts of their range. Pocket gophers have been supposed to have no voice. When caught in a trap or held in the hands against their will they make a hiss- ing or blowing sound by forcing the breath rapidly out and in. This was supposed to be their only sound, but the tame pet on several occasions when hurt or troubled made a low, throaty chur, chur, chur in a complaining tone that seemed to be a real voice. These powerful little burrowing animals live solitary lives almost entirely below the surface of the ground, and most of the time in total darkness behind closed and well-packed doorways. Their eyes and ears are of little use to them and have become almost rudi- mentary, but their tails, with sensitive tips, serve an important function in guiding their retreats in their shuttlelike motions back and forth through their extensive tunnels. With their powerful claws thejr dig up the earth and push it before them to some point where a temporary opening is made through which it is thrust to the surface of the ground. The little mounds, or gopher hills, that dot the fields and prairies where they live are rapidly made. Load after load of the loose earth is pushed in front of the hands and breast to the entrance and thrown out with a little toss until the gallery is cleaned, and the last few loads are firmly packed in the entrance to close the burrow. Sometimes a few quarts and sometimes a bushel of earth are thrown out in one heap, but there is always the little circular dent, where the last load was pushed up and left in the mouth of the burrow, and often the direc- tion of slope to the burrow below may be known from the greater quantity of earth on one side of the doorway. Later another door- way is opened up to the surface 10 or 20 feet away and another hill thrown up, and so on, day after day, until a long line of Mils is formed, or a group if the burrows wind about and among each other. In 1887, the writer counted the fresh hills thrown up by three pocket gophers 12 days after a rain, and the number of mounds that had not been rained upon were 28, 35, and 40. These hills averaged about 6 quarts of earth each, or approximately 17 quarts a day thrown out by one pocket gopher. In summer the tunnels are about 10 inches or a foot below the surface, but in winter they run deeper and probabty keep below the frozen earth, except at the entrance, where many are kept open to the surface. From these openings the animals push their way through the snow along the surface of the ground, leaving tunnels that later are filled with the loose earth from their burrows. 1926] MAMMALS OF NORTH DAKOTA 129 Pocket gophers do not become fat or actually hibernate, but they store up food to some extent, probably for winter use. Breeding habits. — Long and widely known as these animals have been, it seems strange that there is so little information available regarding their breeding habits. Once on a Minnesota farm, two naked young were found in a nest chamber in the burrow. Their eyes were closed, their skin was delicate, pink, and hairless, and their little round heads and fat chubby hands were almost baby- like. The number of young, as shown by embryos in females col- lected for specimens, is 2 to 6, with apparently 4 the most common. The mammae of the females are arranged in two pairs of inguinal and one pair of pectoral. Only the small young are found in the burrow with the mother. As soon as they are old enough to dig for themselves, and before half grown, they branch off into new gal- leries, which finally become closed behind them when their solitarj' careers begin. Most of their lives are solitary, but in the mating season in spring a male and female are occasionally caught in the same burrow. The male soon leaves, however, and takes no further interest in the family affairs. Their reproduction is not rapid, but they are so well protected from enemies above that they increase steadily unless their abundance is controlled by artificial means. Food habits. — The food of pocket gophers consists entirely of vegetable matter, largely roots encountered in their underground tun- nels but also a great variety of green plants from above ground. "When the opening is first made to the surface, the pocket gopher examines the plants close by and usually cuts them and fills its pockets before throwing out the earth, sometimes making several trips back to empty its pockets and fill them again before it throws out the earth and closes the doorway. Thistles, dandelions, clover, alfalfa, and leguminous plants generally are favorite foods, but grass, grain, and a variety of other plants are taken as encountered, and the pockets are often stuffed with leaves and stems intended for food or nest material. The many little bulbs, as wild onions, lilies, and the tuberous roots of native plants, are sought for food, but the soft and tender roots of many other plants are eaten, as well as the bark from even the woody roots of shrubs and trees. The contents of stomachs of pocket gophers usually show a combination of green plant tissues and the finely chewed white or light-colored pulp of roots and bulbs. At times ripe grain is eaten, but generally green food seems to be preferred, or is more easily obtained. Economic status. — In many localities pocket gophers are among the most destructive of rodent pests, as they prefer many of the cultivated crops to wild food and steadily gather into fields where potatoes, turnips, or other root crops are raised, and also into fields of clover and alfalfa and the best ot tame-grass meadows. In grain- fields they do extensive damage, but are partly kept out by the plow- ing of the land and by the long period of scant food in the stubble. In orchards and dooryards they also do much damage, eating the roots from fruit trees and ornamental shrubs, and often killing many of the choicest varieties. Their destruction is imperative in any well- kept agricultural land, and in limited areas this is not difficult. They are easily trapped or poisoned, and detailed methods for their 130 NORTH AMERICAN FAUNA [No. 49 most economical destruction have been worked out by the Biological Survey. Circulars or leaflets giving the best methods can be had on application. Thomomys talpoides mfescens Wiecl Dakota Pocket Gopher Machiohpka of the Mandans (Max- imilian) ; Mdnica of the Dakotas (Gilmore) ; Cipans of the Arikaras (Gilmore) ; Eipapude of the Hi- datsas (Gilmore). Thomomys rufesccns Wied, Nova Acta, Acad. Caes. Leop.-Carol. Nat. Cur., t. 19, pt. 1, p. 378, 1839. Type locality. — Fort Clark, N. Dak. General characters. — Smaller and slenderer than the Mississippi Valley pocket gopher, which comes into eastern North Dakota. Upper incisors, not noticeably grooved except in a fine line near inner edge of each tooth ; large fur-lined cheek pockets on each side of face reaching back under skin to shoulders ; front feet and claws, large ; hind feet, comparatively small ; tail, nearly naked at tip; fur, short, smooth, and glossy. Color of upper parts, dull brownish-gray ; underparts, buffy-gray, often with white markings on chin, throat, and breast. Measurements of adults : Total length, about 240 millimeters ; tail, 70 ; hind foot, 31. Weight of adults, 5 or 6 ounces. Distribution and habitat. — The Dakota pocket gophers cover the greater part of North Dakota and extend into eastern South Dakota and southwestern Manitoba. Their range covers practically the whole State except the low part of the Red River Valley, south of Grafton, and the western edge of the State, where a slightly different form occurs at the junction of the Yellowstone and Mis- souri, and probably along the Little Missouri River Valley. The eastern border of their range in the State is marked by Pembina, Drayton, Grafton, Larimore, Portland, Valley City, and a point 4 miles southeast of Ellendale, in an irregular line following closely the old shore line of Lake Agassiz, and also marking the western edge of the range of the larger Mississippi Valley pocket gopher, Geomys bursarius. The cause for this limitation of range may be due to antipathy of the two species, or to combination of factors; nowhere do the two overlap to any great extent. The fact that Thomomys avoids low or wet ground and is partial to high, dry prairies may be one of the determining factors of this border line. Over all the high open prairie country and often in the timbered areas of the Turtle Mountains and Pembina Hills, these pocket gophers are found in dry meadows, fields, clearings, openings in brushy land, and sometimes even in scattered timber. At Pembina, in 1887, the writer found them common everywhere, except in the thickest growths of trees along the river, and took specimens on both sides of the river as well as on both sides of the border line. A few were found in fields, but they were most abundant over the unbroken prairie, where their favorite food plants were growing. In 1892, Loring succeeded in trapping a specimen at Portland, but in six days' subsequent trapping found no others, so that evidently this was somewhat beyond their regular eastern limit. At Larimore he found them abundant 4 miles west of town, but none farther east, and at Sherbrooke he found them common, as also at Valley City and Jamestown. In the northwestern corner of the State, 1026] MAMMALS OP NORTH DAKOTA 131 about Kenmare and Crosby, pocket gophers were comparatively scarce in 1915, as their characteristic little mounds were noticed only in scattered localities. At the type locality of the species, which was visited in 1909 to obtain specimens for determining the validity of Maximilian's name, the pocket gophers were common over prairie and river flats on both sides of the river, occupying both the dry, sandy bottomlands and the high heavy-soiled prairie. Later in 1913, Jewett also found them common over that part of the valley, in the Killdeer Mountains, and farther south in the vicinity of Glen Ullin and Mandan. In 1893, Fisher reported them very common at Bismarck. In 1915, Sheldon traced them across the southern part of the State, from a point 4 miles southeast of Ellendale, westward continuously to Napoleon, Dawson, and Cannon Ball. The same year Kellogg traced them across the northern part of the State from Grafton to Devils Lake, Towner, the Missouri River at Oak- dale, and thence down the river to Bismarck. Thus the reports cover practically the whole State and indicate fairly definitely the range and abundance of the species. General habits. — In many places throughout their range these pocket gophers will average one or more to the acre and their total numbers over the State are enormous. Their presence can always be recognized by the little mounds of earth heaped up in the prairie grass and containing usually from 2 quarts to a peck of earth. Many of these mounds, however, are enlarged by repeated excavations until they contain a bushel or more, and some measured at Pembina, in 1887, were 3 by 3 feet, and 7 inches high ; 4 by 4 feet, and 10 inches high ; 4 by 5 feet, and 6 inches high ; and 4 by 5 feet, and 7 inches high. These, however, were all composite mounds where the earth had been thrown out several times on successive days. Practically the whole life of the animals is spent underground, where they burrow continuously from point to point, usually 6 inches to a foot below the surface of the ground, bringing out the loose earth by pushing it to the surface in the familiar little mounds, then securely closing the doorways, so that no enemy can enter their homes. Sometimes the row of mounds stretches away for 50 to 100 yards in almost a straight line; they are usually G to 8 feet apart, but sometimes 10 to 20 feet, while between some of the larger hills a space of 27 feet has been measured. More often the tunnels wind about and they sometimes form groups, where one of the animals has worked all summer on a few square rods of ground, so that the lines of old and new mounds crisscross and overlap. The burrows pass through ground that is full of choice food in the form of roots, bulbs, and tubers. Some green food is gathered and tucked into the pockets at the entrance of their burrows, but aside from this the animals rarely come out on the surface of the ground unless for a few seconds at a time when they are throwing out the earth. The earth is pushed out in front of them in little loads about half the size of their bodies, and so quickly that it has the appearance of being thrown from them. Most of the people living in the country where they are abundant never see them, and often their rightful name of pocket gopher is misapplied to the ground squirrels. In winter they go deeper so as to escape the 132 NOKTH AMEEICAN" FAUNA [No. 49 frost, but keep their burrows open to the surface and often come out under the snow and tunnel long distances to obtain green vegetation, afterwards filling these surface tunnels with earth from below. They do not become very fat and evidently are active throughout the winter. Breeding habits. — Apparently but one litter of young is raised in a season and judging from the immature specimens caught in July and August these are born some time in June. A record of five embryos, about one-third developed, taken at Carberry, Manitoba, June 29, 1892, by Ernest Thompson Seton (1909, vol. 1, p. 567), seems to furnish the only positive data available for this subspecies, although records for other forms of the same group, with the same arrangement of mammae, two pairs of inguinal, two pairs of ab- dominal, and one pair of pectoral, indicate a normal litter of sis young. Practically nothing is known of the nest and underground habits of these animals, and the small young seem not to have been recorded. Few animals are more solitary in habits, and only during the mating season in spring are a male and female occasionally trapped from the same burrow. The male soon leaves, however, and probably never sees the young. The mother cares for her family until they are about half grown, when they start burrows of their own and are soon shut off from parental care, each beginning a life that is to be mainly solitary. Although breeding but once a year, their increase is comparatively rapid, as they are unusually well pro- tected from enemies. Food habits. — These pocket gophers live almost entirely on roots and green vegetation, and although they are very partial to certain species of plants, they will eat almost anything that comes in their way if better food is not available. The prairie clover (Psoralea argophylla,), prairie turnip (Psoralea esculent a) , and wild licorice bush (Glycyrrhizale'pidota) are apparently their favorite wild foods over much of the prairies, and their mounds often become very numerous where these plants are abundant. In their pockets are found the leaves and stems of a great variety of other plants, in- cluding grass, lupines, and other legumes, and occasionally roots and tubers, but apparently these are not often brought to the surface. Sometimes the pockets are found stuffed so full of green vegetation that they more than double the apparent size of the animal's head. They are used only for carrying food and not, as is sometimes re- ported, for carrying earth out of the burrows. To what extent roots, tubers, and bulbs are stored for winter food is not well known, but occasionally well-filled storage cavities are found along the lines of the tunnels. Economic status. — Next to the ground squirrels, these gophers are generally the most destructive rodent pests of the region where they live. Although for ages they have been industriously plowing and mellowing the prairie soil, burying the surface vegetation and enriching and improving the land, they at once become the farmer's enemy when occupying the ground with his crops. Even on the prairies they destroy or consume much of the choice grass that would otherwise be available for stock, and cover up and prevent the economical cutting of much of the wild hay on the prairie and the best parts of the dry meadows. 1926] MAMMALS OF NORTH DAKOTA 133 In fields, gardens, and orchards, however, they do the most harm. Entering through their safe tunnels, they find choice food in the clover and alfalfa fields, and if nothing better can be found will live all summer on the green stems, leaves, and heads of grains. In vegetable gardens they are even more destructive, cutting the peas and beans above the ground and drawing them into their burrows to be eaten, or, without the risk of appearing at the surface, taking the onions and turnips or following a row of potatoes and cleaning the tubers from each hill in succession. Nowhere is their mischief more exasperating than in a clean and well-kept orchard, where, lacking other food, they often eat the bark from the roots of the trees and leave them to die, or even cut off so many of the roots that the trees dry up and tip over with the first wind. Fortunately, pocket gophers are easily controlled, and it is only necessary to know how to poison or trap them in order to protect crops and trees. "Where only a few are doing mischief, the simplest method is to trap them by merely opening their doorways and set- ting traps that will catch them as. they come out to close the openings. Armed with a few modern traps and an old table knife, anyone can, with a little practice, catch all the pocket gophers in an ordinary garden or orchard without much loss of time. Where the mischief is on a larger scale, poison is a more rapid and eco- nomical control measure. Simple directions can be obtained from the Biological Survey for the most effective methods of administer- ing poison. Although excellent food and in every way perfectly suitable as a food animal, pocket gophers are not large enough to be of importance as gamp. In places where it is necessary to catch con- siderable numbers of them, however, they can be used to advantage as food, and if properly dressed and cooked are as good as rabbit or squirrel. Th.omom.ys talpoides bullatus Bailey Sagebrush. Pocket Gopher Thomow.ys talpoides bullatus Bailey, Proc. Biol. Soc. Washington, vol. 27, p. 115, 1914. Type locality. — Povrderville. Mont. General characters. — Very similar to rufesccns but noticeably lighter and brighter colored, with conspicuously larger audital bullae. Measurements of type specimen; total length, 238 millimeters; tail, 72; hind foot, 30. Weight of female, from Buford, 5 ounces. Distribution and habitat. — The arid sagebrush-valley form of pocket gopher occupies mainly the Yellowstone and Missouri Valleys of Montana, but comes into North Dakota at Buford and is probably the form occupying the Badlands part of the Little Missouri Valley, as specimens have been referred to it from the valley just below the southwest corner of the State. In an arid, open habitat, often with sandy or light-colored soils, these pocket gophers have become adapted to their environment in coloration, but in general habits show only such differences from rufescens as are occasioned by the conditions under which they live. Over the open range country they are of little economic importance, but as many of the valleys are brought under irrigation with intensive cultivation, they become of serious consequence and their destruction is necessary to satisfactory returns from the cultivated areas. 134 NORTH AMERICAN FAUNA [No. 49 Order LAGOMORPHA : Rabbitlike Animals Family LEPORIDAE : Rabbits Sylvilagus fLoridanus similis Nelson Nebraska Cottontail Wahhoos of tbe Chippewas (Wil- ■ son) ; Manshtin-sapana of the Dakotas (Gilmore) ; Monstinga of the Omahas (Gilmore), gen- eric term. Sylvilagus fLoridanus similis Nelson, Proc. Biol. Soc. Washington, vol. 20, p. 82, 1907. Type locality. — Valentine, Nebr. General characters. — Rather compact, with relatively short ears and short legs. Colors essentially similar in summer and winter. Upper parts, rusty gray, darkened by numerous long, black-tipped hairs ; rump, clear dark gray ; throat, belly, and under surface of tail, pure white. Adult measurements : Total length, approximately 408 millimeters ; tail, 52; hind foot, 99; ear from notch, 50. An adult male taken at Fargo by O. J. Murie on November 7, 1919, measured 405, 52, 98 millimeters, and weighed 2 pounds 1% ounces. Fig. 8. — Three species of cottontail rabbit in. North Dakota : Circles, the Nebraska cotton- tail ; triangles, the Black Hills cottontail ; squares, the Wyoming cottontail Distribution and habitat. — Over the central Great Plains from Kansas to near the Canadian line Nebraska cottontail rabbits occupy the stream valleys and thickets of a mainly prairie region (fig. 8). In North Dakota there are specimens from Fairmount, Oakes, Fargo, Kathryn, Portland, Valley City, Larimore, Grafton, Hawks Nest, Dawson, Stump Lake, Sweetwater Lakes, Towner, Oakdale, Stanton, Fort Clark, Deapolis,. Bismarck, Cannon Ball, and Winona. _ These are mainly brush rabbits, not commonly found at any great distance from the wooded or brushy bottoms. Apparently they have not yet reached the Turtle Mountains, although since the settlement of the country their range seems to be slowly extending northward in the 1926] MAMMALS OF NORTH DAKOTA 135 State. In 1887 there was found no trace of them in North Dakota, nor nearer than Fort Sisseton, S. Dak., and Browns Valley, Minn. Eastgate says they first reached Larimore in 1900. In 1913 Mr. Booth, a taxidermist, reported that cottontails were abundant at Minot, and had first arrived about 1890. No trace of them was found farther north, at Kenmare, Crosby, or Bottineau. In 1912, on a wagon trip from Linton, in the vsouthern part of the State, to Stump Lake, no trace of cottontails were found until Hawks Nest Butte was reached, where a specimen was obtained and tracks were seen in the timber, and on the south side of Stump Lake the rabbits were common in the timber. In 1919, at Walhalla, Eugene D'Heiley told the writer that cottontails came there in 1912, but soon disappeared, though they were still found at Neche, 20 miles farther east. In 1913 Jewett collected a specimen at Fort Clark and one at Oakdale in the Killdeer Mountains, and reported them as fairly common in the thickets and brushy gulches. In 1915 Kellogg took specimens at Larimore and reported them common at Manvel, Grand Forks County, and at Grafton, Walsh County. He was told that they were common at Drayton, Pembina County, although he did not find any. At Towner he took one immature specimen in the meadow and saw several others, and he reported the species com- mon along the Missouri Kiver from Stanton to Bismarck. General Tidbits. — These little short-legged rabbits are such an easy prey to dogs, wolves, and foxes that it is necessary for them to keep within the protecting cover of thickets or dense vegetation. Usually where they occur their roadways or trails may be found in every thicket or leading from one thicket to another. At Hankin- son, they were found abundant in the woods and brush patches around the lake shores and in the thickets among the sand dunes. On the Hankinson ranch cottontails were frequently seen in the dooryards and about the buildings, in spite of several dogs and cats which were constantly hunting them. A family of half -grown cottontails living in some burrows under the roots of a tree gave the dogs a great deal of exercise in chasing them to cover and digging and barking at their burrows. The rabbits did not seem to care and were getting the best kind of training for life on the ranch. At Fairmount, Sheldon reported them as frequenting farms and deserted buildings. At Lisbon and Valley City, Eastgate reported them as very common in the thickets and in both the natural timber and planted groves. As the country fills up with farm buildings, orchards, and garden shrubbery, these rabbits seem to increase in abundance and extend their range on the open prairie, where formerly it was impossible for them to exist because of numerous native enemies. In the older, more settled parts of the State they are conspicuously most abundant. Along the Missouri River bottoms, Avhere the thickets are dense and often thorny, they find the most perfect pro- tection and satisfactory conditions of environment. At Washburn, in 1909, the writer found them abundant all over the brushy river bottoms, where in summer they had the added protection' of hosts of mosquitoes, which rendered hunting almost impossible. At Fort Clark, in 1913, Jewett reported them common in the wooded and brushy bottomlands, to which they were closely restricted; on a 136 NORTH AMERICAN FAUNA [No. 49 short walk along the river-bottom roads in the evening, he would usually see five or six. At Cannon Ball, in 1916, they were found very common in the brushy bottoms along the Missouri and Cannon- ball Eivers, and at Parkin, a few miles up the Cannonball, they were common in the wooded and brushy bottoms. One was^ living in a lumber pile in the middle of the new town just starting up on the prairie, and a bulldog spent much of his. time chasing _ it from one lumber pile to another, but the rabbit seemed to realize its advantages and not to worry over the noisy demonstrations of the dog. Breeding habits. — Cottontails are prolific breeders and usually raise several litters of young in a season. At Fairmount, on May 28, Sheldon collected a female which was nursing young. On the Sheyenne River, north of Valley City, Eastgat© took one on May 17, 1912, which contained seven embryos; and at Grafton, on May 10, 1912, Williams took one containing six small embryos; one col- lected by Jewett at Fort Clark, on July 22, 1913, contained five small embryos. Although born in a naked, blind, and helpless con- dition, the young develop rapidly and are soon able to shift for themselves, leaving the mother to resume her parental duties with a new family. Food habits. — Babbits are mainly grazing animals, and their list of food plants includes in large proportion both native and culti- vated vegetation. They take the leaves and tender blades from grasses, clovers, and most of the wild leguminous plants with which they come in contact, and are especially fond of the cultivated clovers, alfalfa, and most garden vegetables. They also eat the bark and buds of many shrubs and small trees in summer and in winter depend largely upon browse and bark for their food. At Kathryn, in Barnes County, Eastgate reported them feeding in the evenings along the edges of the grainfields, where it was common to see six or eight at a time. They always find an abundance of food and as one kind of vegetation dies or dries up, other plants are ac- cepted in its place. In times oi deep snow the rabbits forage out from their well- protected burrows and pick buds and green tips and branches from the shrubs and such plants as are exposed above the snow, eveiy increase in the depth of snow lifts them to a fresh supply. Their runways in the snow are always packed and frozen, so that a rapid retreat to safe cover is assured, and as more food is needed the run- ways are extended farther out through the brush or from one thicket to another. Economic status. — Numerous inquiries among farm residents made it evident that these rabbits are not generally considered a pest, although where abundant they occasionally do considerable mis- chief. The small quantity of grain that they cut along the edges of a field, the forage crops eaten, and the fruit trees and shrubbery occasionally killed or damaged, is readily forgiven them because of their value as food and game. To many of the country boys they furnish the only available hunting, and usually before the winter is far advanced they have become so scarce as to leave barely enough to restock the country the following spring. In this northern clime they accumulate considerable fat during the fall, and are 1926] MAMMALS OF NORTH DAKOTA 137 among the choicest rabbits for food, being especially healthy, plump, tender, and well flavored. They have also a market value, and if the}* ever become overabundant, ample protection against damage to crops and trees may be had by extending the hunting season. In rare cases it may be necessary to poison those around orchards and gardens. Full directions for destroying them in this way will be furnished by the Biological Survey on request.18 Generally, how- ever, the few individuals that are doing mischief can be shot and utilized for food. The young of the year are especially delicious, broiled or fried, while the old individuals, well stewed with a little bacon or fat pork, afford an acceptable variety for any table. SyTvilagus nuttailii grangeri (Allen) Black Hills Cottontail Nis of the Arikaras, and Itak- shipisha of the Hidatsas (Gil- more ) . Lepus sylvaiicus grangeri Allen, Bui. xliner. Mus. Nat. Hist., vol. 7, p. 264, 1895. Type locality.— Hill City, S. Dak. General characters.— About the size of similis, but lighter gray, with slightly longer ears and distinctive skull characters. Slightly smaller than bailey i, with shorter ears and feet. Very similar in color, but brighter rusty on nape and legs. Average measurements : Total length, 3S5 millimeters ; tail, 46 ; hind foot, 95; ear (measured dry), 56. Distribution and habitat. — The little pale-gray Black Hills rabbits barely reach into extreme western North Dakota from their wide dis- tribution over the arid interior of Nevada, Utah, Wyoming, and Mon- tana (fig. 8). There are 5 specimens from Buford, 2 from Goodall, 2 from Medora, and 1 from Mikkelson on Roosevelt Creek, 23 miles north of Medora. They occup3'' the same Badlands country with baileyi but appear to confine themselves mainly to the dense thickets along the stream courses. At Medora, Jcwett reported them as not common, but a few were found in the banks of the Little Missouri River, where a female was shot on January 15 as she sat in front of her burrow. Only three were seen in this locality. Farther down the river a few were seen usually in thick growths of buffaloberry bushes. In other localities the writer has found them taking shelter among rocks and in IioIIoav banks, but more often under dense growths, as sagebrush or thorny thickets of bullberry bushes. The three distinct species of cottontails of North Dakota have amicably or otherwise divided the ground among themselves, in a way that seems best to fit the needs of each, grangeri taking the place of similis in the arid brushy bottoms, while baileyi occupies the rougher and more open uplands. Sylvilagus audubonii baileyi (Merriam) Wyoming Cottontail Lepus baileyi Merriam, Proc. Biol. Soc. Washington, vol. 11, p. 14S, 1897. Type locality. — Spring Creek, Bighorn Basin, Wyo. General characters. — Size about the same as .si in His, but with oars and legs conspicuously longer and colors lighter. Upper parts, light gray with a buffy tinge, neck clear buffy: uuderparts, white; tail, large and puffy and three- 18 See Farmers' Bui. 702 (Lantz, 1916). 138 NORTH AMERICAN FAUNA [No. 49 quarters white, with relatively narrow stripe of gray above. Measurements of adult male, from Little Missouri River : Total length, 399 millimeters ; tail, 48; hind foot, 102; ear (measured dry), 67. Measurements of type specimen, 418, 50, 100, and 94. Distribution and habitat. — The long-eared Wyoming cottontails come into extreme western North Dakota along the Little Missouri Valley (fig 8). There are specimens in the Biological Survey col- lection from Marmarth, the former from North Dakota National Forest, Sentinel Butte, and the Little Missouri Eiver, 25 miles north of Medora, and one in the agricultural college collection, at Fargo, collected by Doctor Bell, at Wade, on the Cannonball River. _ At Parkin, near the mouth of the Cannonball, the writer recognized these long-eared rabbits as common in 1916 around the Badlands buttes, and Sheldon reported one seen at the Palace Buttes a little north of the mouth of the river. General habits. — To a great extent these are Badlands cottontails, and instead of keeping to the brushy bottoms they are more often found along the broken slopes and among the rock piles and Bad- lands gulches of the roughest parts of the country. At Parkin the writer found them along the steep slopes of the high butte near town, running from one rock pile to another and taking refuge under the rocks and in washed-out cavities of the Badlands slopes. Their long ears, big white tails, and yellow-gray color mark them at once as different from the short-eared and more compact little Nebraska or Black Hills cottontails of the brushy bottomlands. Along the north- ern edge of the North Dakota National Forest early in August of 1913 they were found abundant in the banks of the river valley and in the rough gulches of the rocky slopes of the Badlands, where they would quickly gain cover in some rock pile or washed-out hol- low in the banks or else take refuge in an impenetrable jungle of buffaloberry bushes or tangle of brush that offered equally good pro- tection. About 8 miles south of Sentinel Butte Jewett obtained a specimen at the entrance of a deep crevasse in a rocky gulch on the side of the big butte. In an open country, where life frequently depends on getting quickly to safe cover, these rabbits have devel- oped long ears and long legs for quick hearing and rapid flight. In other ways they have the general habits of most cottontails. As food they are equally as good as the brush-inhabiting species and as game generally more difficult to shoot. lepus americanus americanus Erxleben Varying Hare; White Rabbit; Snowshoe Rabbit (PI. 17, fig. 3) [Lepus] americanus Erxleben. Syst. Regni. Anim., p. 330, 1777. Lepus bishopi, Allen, Bui. Amer. Mus. Nat. Hist., vol. 12 (1899), p. 11, 1900; type from Mill Lake, Turtle Mountains, North Dakota.19 Type locality. — Fort Severn, Keewatin, Canada. General characters. — About midway in size between jack rabbits and cotton- tails. Ears and legs, moderately long ; tail, small ; feet, large and hairy, especially in winter. In summer upper parts dark buffy gray, with blackish on tips of ears and top of tail ; feet, buffy brown ; chin and middle of belly, 19 This form, based on an abnormal skull in the type and only specimen available at the time of its description, appears on examination of a good series of specimens from the type region to be typical americanus. 1926] MAMMALS OF NORTH DAKOTA 139 whitish ; lower surface of tail, gray. In winter pure white, except black narrow border of ear, and dark eyes. Fur, very long and soft ; on soles of feet, long, dense, and coarse. During change from white winter coat to gray summer coat, after loss of the long white cover-hairs and before gray summer coat comes to the surface, there is a short time when the yellow underfur is exposed ; also during the fall change from gray to white the color is much mixed and often patched. Average measurements of adult specimens from North Dakota: Total length, 451 millimeters; tail, 34; hind foot, 125; ear from notch (measured dry), 60. Weight, 3 to 3% pounds (Seton). Distribution and habitat. — Snowshoe rabbits, which turn from gray- in summer to white in winter, are more or less common in the forested areas of the Turtle Mountains, Pembina Hills, around Devils Lake, and along the wooded parts of the valleys of the Red, Mouse, and Missouri Rivers. There are specimens in the Biological Survey col- lection from Grafton,20 the Turtle Mountains, Devils Lake, Stump Lake, Elbo woods, and Buford. Throughout the timbered and brushy areas of the Turtle Mountains they are especially abundant and specimens have been collected near Metagoshe Lake, Fish Lake, Diansley, Birchwood, and Mill Lake. On January 21, 1913, W. B. Bell reported one collected near Fargo and mounted for the agri- cultural college collection, and in 1919, Murie reported them as occasionally found there. In 1887, the writer was told that they were found in the woods near Grand Forks, and at Pembina he found them common. At Kenmare, in 1913, he found them common in the thickets and woods of the side gulches along the Des Lacs Valley, where their trails and signs were abundant and several of the rabbits were seen. C. E. Peck said that in the fall and winter the boys killed them there by dozens in the thickets of aspens and other northern trees and shrubs. Mr. Booth, a local taxidermist, was certain that they were common in the woods along the Mouse River near Minot. At Buford, in 1910, Anthony reported them common in the brushy river flats, where their well-beaten runways and patches of peeled willow brush were conspicuous and where several were seen and one specimen obtained. At Elbowoods, in 1915, Kellogg collected one specimen and reported them as quite common in the forest along the river bottoms. At Stanton and Sather he reported them scarce. At Cannon Ball, in 1916, the writer found their unmistakable signs and trails in the thickets along the river bottoms and was told by -the residents that they were not very common. At a spot wdiere one had been killed and eaten and its fur scattered about, the writer collected the tail as positive proof of the species. At Devils Lake, in 1916, Kellogg found one of these rabbits dead on Sullys Hill and saved the skull for a specimen. The following year a young one wras taken on the north shore of the lake about a mile from the town of Devils Lake and signs of them were found throughout the woods along the north and south shores. Williams reports them abundant at Grafton at times. 20 Two specimens collected at Grafton on March 30, by IT. V. Williams, are in the yellow spring coat after the disappearance of most of the white outer fur. In one of these a spot of the new summer coat is shown and this agrees with the buffy-gray color of americanus rather than with the warm brown of Lrpus amrricanus phaconotus Allen of Minnesota. Although the type locality of phaeonotus is just across the Red ttivcr Valley at Hallock, Minn., the specimens from Grafton are evidently nearer to the typical subspecies than to the Minnesota varying hare. 82242°— 20 10 140 NORTH AMERICAN FAUNA [No. 49 General habits. — The varying hares are strictly woods rabbits, de- pending on dense forest and thickets for cover, protection, and food. They rarely come into the open, except along the edges of brush patches, where they can quickly dash back out of sight into their well-beaten trails and runways, which carry them under the brush in perfect safety from most of their enemies. In summer their dusky-gray colors render them invisible in the brushy shadows, and as they sit with ears low on their backs they seem fully aware of the advantage of their protective coloration and often allow passersby almost to step on them before bounding away into the thickets. Though mainly nocturnal in habits, they are usually seen in the eve- ning or early morning sitting in the roads or trails that wind through the forest, and in a good rabbit year, when their numbers are at the maximum, a late or early drive along the wood roads usually sends them hopping out of the way at frequent intervals. At times they become very scarce, and often for a period of several years are seldom seen. Many theories have been advanced to ac- count for the waves of abundance and scarcity, which seem to be more or less periodic, but much remains to be learned by close and continuous observation of the real causes. A very full account of their fluctuations through the north country is given by Preble (1908, p. 199), in North American Fauna No. 27; Seton (1909, vol. 1, pp. 621-652) also gives an interesting account of their habits in his Life-histories of Northern Animals. Breeding habits. — On June 18, 1916, some one found a very young rabbit that had been killed by a dog in a patch of silver-leaf bushes on the shore of Devils Lake, about a mile from town. It was not so large as one's fist and had evidently been dug out of the nest or hollow in the leaves of the little brush patch, and as it had just been killed it made an excellent specimen and showed the beautiful long crinkly, coarse gray fur of the juvenal coat. Although appar- ently not a week old, its fur was very long, soft, and full, and the color even more highly protective than in the adults. Apparently the dog had eaten or carried away the other members of the family, so the number in the litter could not be determined. Usually with this species there are 3 or 4 young at a birth, and farther north in Canada Preble records 2 to 6 embryos. A female examined at Fort Clark by Maximilian in 1833 contained 4 embryos. The species is generally supposed to raise 2 or 3 litters of young during a sum- mer, but data on this point are meager. Food habits. — In summer these rabbits feed on a great variety of green vegetation, including grasses, grains, many of the wild and cultivated clovers and leguminous plants, and some buds and leaves of shrubbery. In winter they depend mainly upon the bark and buds of a great variety of shrubs and eat higher up as the snow becomes deeper. In spring the bushes neatly clipped at various levels show the depth of snow from which the rabbits fed at differ- ent times during the winter; often these clippings reach 4 or 5 feet above the surface of the ground. The large chisel-like incisors of the rabbits will cut bushes up to the size of lead pencils as smoothly as if done by a knife, and they also serve to remove the bark from fallen branches and even the trunks of small trees when other food is not abundant. Sometimes whole thickets of willow and aspen 3926] MAMMALS OP NORTH DAKOTA 141 are denuded of bark as high as the rabbits can reach, and even some of the young forest trees are thus injured or killed. Except in years of unusual abundance the rabbits find an ample food supply in the buds and tender tips of the winter browse without doing much harm. They are usually plump and sometimes show considerable fat even during the coldest of winter weather. Economic status. — In newly settled sections of wooded country where the snowshoe rabbits are abundant they sometimes do consid- erable harm in cutting the young trees and shrubbery in winter, and may take a small portion of the growing crops in summer. Their value as food and game animals, however, is sufficient to outweigh by far the little damage they occasionally do. In many parts of the country where once common, they have been practically ex- terminated" from extensive areas by persistent hunting. There is great danger that, without reasonable protection in restricted areas of their range, such as that about Devils Lake and in the scattered timber patches along the Mouse River, and even in the brushy bot- toms along the Missouri, they may be killed off to the point of ex- termination. Among all the rabbits of the State they are the most desirable as food and game and from their habit of keeping entirely within the brush they are less likely to do serious harm to crops. Except in years of extreme abundance, their seasonal protection with that of other game would seem a wise precaution. At any time when their numbers become too great the protection could be re- moved and they would soon be reduced by local and market hunters. Lepus townsendii canipanius Hollister White-tailed Jack Rabbit Warchu of the Arikaras, and M side of Devils Lake, poisoned 700 wolves and obtained many beavers, otters, foxes, and minks. In some notes from Valley City, furnished by Morris J. Kernall, is one of John Hailand, who settled there in 1878. At that time he says timber wolves were seen occasionally, though they were not numerous. He saw one killed there that he thought must have weighed more than 100 pounds. In the early eighties, Roosevelt (1900c, p. 66), on returning to the house at his ranch on the Little Missouri River when it had 154 NORTH AMERICAN FAUNA [No. 49 been closed for many months, found in the dusty trails in the ravines, many tracks of the timber wolves. "Once or twice in the late evening we listened to their savage and melancholy howling." Even then the great numbers of wolves had gone with the buffalo either to the skin market or farther west. In 1894 Roosevelt sent the skulls of two old and four young wolves, killed 20 miles south of Medora, to the Biological Survey. In 1887 the wolves were practically gone from most of the country across the State, and even at Fort Buford, which was then the ter- minus of the Great Northern Railway, they were very scarce. In 1910 Anthony was told that they were still found in the country south of the river, but no definite records were obtained. In 1913, at Minot, the writer was told that a few wolves were still to be found in that part of the country, and the same year Jewett reported them rare in the Killdeer Mountains, but all too plentiful in the Badlands section along the Little Missouri from Quinion to Medora, and he was told of several colts that had been killed by them during the summer. In the vicinity of the former North Dakota National Forest it was stated that the large wolves were then getting scarce but that a few years previously they had killed many calves. The same year at Wade, on the Cannonball River, Bell reported a fewT although they had been pretty thoroughly trapped out by profes- sional wolf trappers employed for the purpose by the stockmen's association. In 1915, Kellog reported one wolf that had been fol- lowed for three days by a trapper near Warwick, south of Devils Lake. At Elbowoods he saw the skin of one that had been killed on the Indian Reservation, where a drove of sis were said to be still at large. Farther down, at Painted Woods, he saw two cross the river and a few tracks on the east side. At Cannon Ball, in 1916, the writer was told that a few large wolves were still in the country a little farther west, but that the great numbers of the animals had disappeared with the buffalo. Doctor Beede told the writer that the old Indians, in talking of hunting trips when the buffalo were still abundant, claim that three wolves would pull down and kill any buf- falo, even an old bull. On January 1, 1922, a large wolf wTas shot by Mr. Bennett near Harwood, about 10 miles north of Fargo, and the skin, which was said to be 7% feet long, was tanned for a rug and kept by the hunter. The wolf had been tracked for two days by dogs and hunters from the vicinity of Breckenridge, Minn., but was shot on the North Dakota side of the river. At the present time there are probably a few wolves left in the least-settled parts of the rough Badlands region west of the Mis- souri River, but it is to be hoped that a very few years will see the last of these destructive animals in this State. General habits. — Few animals show greater intelligence and re- sourcefulness than wolves in adapting themselves to such conditions of climate and environment as will afford them a sufficient supply of food. From the Arctic barrens to the steaming swamps of Florida they have been at home wherever game was abundant, but nowhere more numerous than over the Plains in the days of the great buffalo herds. In habits they are hunters and rovers and often to a con- 1926] MAMMALS OF NORTH DAKOTA 155 siderable extent migratory, although in their home life they are domestic and as closely restricted to their home grounds as any car- nivore could well be. The breeding dens, which in this prairie country usually consist of burrows in banks and sidehills, are the home centers from which the faithful parents make regular excur- sions for food until the 37oung are old enough to leave the den and accompany them on hunting trips. Then they are freebooters until the next breeding season, when the adults generally endeavor again to occupy the old den or dig another in its vicinity. The fact that the old wolves pair for the breeding season is thoroughly proved, and there is much evidence to indicate that the pairing is for life or for as long as the two are able to keep together. While the young- are small and as long as they remain in the den, the male is alwa}7s on guard or foraging for food to bring home for its mate and young, and as soon as the young leave the den it leads the pack and appar- ently does much of the killing. The wolf pack usually consists of a family, the two adults, and 6 to 10 young of the year, but there are apparently authentic accounts of larger wolf packs where presum- ably two or more families have temporarily joined. Breeding habits. — Wolves do not breed until 2 years old, but the family pack keeps together until about midwinter or later. The young are generally born in March, although there are records of pups late in February, and a few late litters are born in April. The young vary normally from 6 to 10, but there are records of 11, 12, and 13 in a den. At first they are dull black in color, but by the time they are a month old and begin to appear at the entrance of the den they have faded to a dull clay color or yellow gray. Usually they do not leave the den until July or August, when nearly half grown and able to accompany their parents on hunting trips and take care of themselves in case of emergency. A pack of growing, hungry young wolves in fall and early winter requires a large supply of meat which is obtainable only from large game or domestic stock. Economic status. — Apparently a considerable time intervened between the destruction of the buffalo herds and the introduction of domestic cattle in sufficient abundance to provide an easy food supply for the wolves. This scarcity of food, together with the activity of those trapping and poisoning wolves for their skins, reduced the number of wolves and made the cattle industry possible over most of the open plains country. The last of the wolves, how- ever, took up their residence in the roughest and least occupied sections, where they are extremely difficult to dislodge, and with their natural intelligence and long years' experience with man and his traps, guns, and poisons, they have become one of the most difficult animals to capture or destroy. In some sections of the country they had shown their ability actually to increase in the face of all human efforts and inventions for their destruction until the recent concerted efforts of Federal and State wolf hunters proved too much for them. One wolf was known to kill 125 head of cattle in 10 months, valued at the time at $5,000. In Custer County, of the adjoining State of South Dakota, one wolf killed $25,000 worth of cattle in seven years. Although it is probable that 82242°— 26 11 156 NOBTH AMERICAN FAUNA [No. 49 the wolves can never be exterminated over much of the northern forest area of the continent, it has been clearly demonstrated that they can be practically eliminated from the open stock range of the Western States. Canis latrans latrans Say Northern Coyote; Brush Wolf Mes-ta-chd-gan-es of the Gjibways (Seton) Canis latrans Say, Long's Exped. Rocky Mountains, vol. 1, p. 16S, 1823. Type locality. — Engineer Cantonment, near Blair, Washington County, Nebr. General characters. — Largest of the coyotes ; ears, erect and pointed ; conspicuous capelike mane over shoulders ; fur, long and soft in winter, short and harsh in summer ; color, light brownish-gray, darker and more fulvous in summer ; underparts, whitish, tail tipped with black. Measurements of female from Elk River, Minn.: Total length, 1,219 millimeters; tail, 394; hind foot. 179 ; weight of adult male from Beemer, Nebr., 36 pounds ; of one from Fort Dodge, Iowa, 40 pounds. Distribution and habitat. — In the absence of specimens from the type region of Canis latrans it has been customary to refer to this original form of the group, the large brush wolves of Iowa, Wis- consin, and Minnesota. In 1897 Merriam made a study of the coyotes with the material then available, which was not sufficient to establish the matter of intergradation between latrans and nebracensis of the plains region farther west. The great quantity of material since col- lected seems to establish this connection, but the definite outlines of areas occupied by each form remains to be worked out in a compre- hensive study of the group as a whole. Specimens from near Grand Forks and Grafton indicate that these brush wolves come into east- ern North Dakota, but how far west they extend is not at present known. The wolf skins collected by Alexander Henry and his trap- pers along the Red River Valley from 1800 to 1808 probably included many of these large coyotes or brush wolves. At Hankinson in 1912 there were said to be a few coyotes, but they were seen only at rare intervals. In the region about Fargo the same year they were said to be very scarce. At Wahpeton in 1915 Kellogg reported them as common, and was told of a litter of five dug out of a den 2>y2 miles south of town on May 30. At Larimore, in Grand Forks County, he reported them as fairly plentiful, and near Grand Forks he reported one killed during the previous winter and a bunch of six seen at one time; at Grafton a few killed each winter, and at Drayton, Pembina County, as not very common, but a few killed each winter. Near Grand Forks during the winter of 1918-19 they were reported as unusually common and destructive to stock. At Grafton H. V. Williams, in a letter of March 21, 1919, says that several were killed during the winter. In the Turtle Moun- tain region they have been reported as common by Williams and Eastgate, and while there is some doubt as to which form occurs there, in the absence of specimens the writer is inclined to consider them as probably the large northern form. Some of the residents describe them as a large coyote or small wolf and others says that they are small and pale, but the relative characters can not be re- liably determined without actual specimens for comparison. A skull of a small female collected at Valley City by Morris J. Kernall in 1913 is apparently intermediate between latrans and nebracensist 1926] MAMMALS OF NORTH DAKOTA 157 as it does not show decided characters of either. Coyotes are still found over practically all the State, but the specimens from the west- ern part seem to be all referable to the smaller, paler nebracensis. General habits. — These large coyotes generally inhabit a partly timbered, partly open country, but readily adapt themselves to either type where game or livestock furnish a satisfactory food supply. In habits they differ little from other species of coyotes except in adapting themselves more readily to forest conditions and in de- pending more on game and livestock for their food. They are most persistent in the destruction of sheep and calves, and have long rendered impracticable the keeping of small herds of sheep on farms over much of their range. Just how much they have had to do with the destruction of deer and other large game will never be known. Canis latrans nebracensis Merriaru Plains Coyote; Prairie "Wolf (PI. IS, fig. 1) Mica or Mieaksica of the Dakotas (Riggs and Williamson) ; Mikasi of the Omahas (Gilniore) ; Sche'kd of the Mandans (Will) ; ll6tsa of the Hidatsas (Matthews) ; Stshirits pitkatsh of the Arikaras (Gilmore). Canis pallidus Merriam, Proc. Biol. Soc. Washington, vol. 11, p. 24, 1S97. Canis nebracensis Merriam, Science, vol. 8 (n. s.), p. 782, 1S98. (Substituted for pallidus, which was preoccupied.) Tiipe locality. — Johnstown, Brown County, Nehr. General characters. — Slightly smaller than latrans, with lighter denti- tion and paler colors ; upper parts, light buffy gray, back of ears buffy ; tail with black tip ; underparts whitish. Unfortunately there are no measure- ments or weights available from the type region, nor of any specimens that may be considered typical of this form. One measured by Kellogg at Fort Totten was as follows: Total length, 1.193 millimeters; tail, 380; hind foot, 205, but the specimen was not obtained. In distinguishing a coyote from a wolf, the nose, foot, and tooth measurements are always sufficient. In the coyote the nose pad measures approximately seven-eighths of an inch wide, the heel pad of front foot, 1 inch wide, and the greatest diameter of canine tooth at base, 0.3 inch. Distribution and habitat. — At the present time plains coyotes are distributed over practically all of North Dakota and are es- pecially common over the western half of the State. Apparently they have held their own and even increased since the destruction of game herds, for filling the country with domestic livestock and poultry gives them a food supply often more easily obtained than the original wild game. In the early days of trapping and explora- tion, little mention is made of coyotes, and apparently they were less common or less conspicuous than the large wolves. Alexander Henry, in the beginning of the eighteenth century, does not mention them, but possibly he did not discriminate between them and the larger wolves. Lewis and Clark, in 1805, rarety mentioned them on their trip up the Missouri River, while they frequently spoke of the wolves seen and killed. On April 24, 1805, at a point about 13 miles above the mouth of the Muddy River (1893, p. 280), they spoke of the hunters returning " with four deer, two elk, and some young wolves of the small kind." In 1833 Maximilian (Wied, 1839- 158 NORTH AMERICAN FAUNA [No. 49 1841, Bd. 2, pp. 97, 98, 278, 307, 1841) frequently refers to them and gives the Indian names used by the Mandans, Minnetarees, Arikarees, Dakotas, and Blackfeet. He says, " The prairie wolf is numerous over the prairies and in winter comes occasionally into the Indian villages to pick up whatever he can in the way of refuse." Again, in December, he writes, " During the night we heard the barking of the prairie wolves {C arils latrans Say), which prowled about looking for any remains of provisions." On February 26, 1834, he notes, " The prairie wolves now prowl about in couples." In 1843 Audubon (1897, p. 160), on his trip up the Missouri River to Fort Union, made little mention of coyotes, but reported one seen at Fort Union and one shot by Harris, on September 3, below the mouth of the Cannonball River. A series of skulls from the upper Missouri River in North Dakota, collected by F. V. Hayden in 1850, are still in the National Museum collection. Roosevelt (1900c, p. 63), at one of his hunting camps in the Little Missouri River country, in the eighties, enjoyed the " wild, mournful wailing of the coyotes. They were very plentiful round this camp ; before sunrise and after sundown they called unceasingly." At Valley City, Major White told Morris J. Kernall that the coyotes were more abundant there in 1882 than in 1913, and John Hailand, who settled at Valley City in 1878, told him that they were then more numerous than in 1913. In 1887 coyotes were reported as common at many localities over the State, and in 1893 Doctor Fisher, in stopping off at various localities from Bismarck to Fargo, reported them more or less common at all places visited. In 1912 Eastgate reported them at Tolna, on the south side of Stump Lake, along the Sheyenne River, 30 miles north of Valley City, in the sand hills near Kathryn in Barnes County, and that a few were shot every winter near Lisbon, in Ransom County, where they did considerable damage to sheep and poultry. In 1909 the writer found them common about Marmarth, in the southwestern corner of the State, and in 1912 was told that at Tolna one man had brought in 50 scalps during the winter for the bounty, and that a den of young had been found a year before near Stump Lake. At Kenmare, in 1913, they were said to be fairly common over that general region. At Minot, Mr. Booth said that they had been scarce in that vicinity until the sheep industry developed in 1890, when they became unusually numerous. Since the country has settled up, however, few sheep are kept and the coyotes are becoming comparatively scarce. At Crosby the writer was told that they were scarce, but that a man living there, hunting with a couple of wolf hounds, occasionally brought in one. At WiUiston and Buford, in the same year, they were said to be fairly common and a consider- able number of tracks were seen. In the Killdeer Mountains, in 1913, Jewett reported them in numbers over the entire region, as also along the Little Missouri River south to Medora. He also found them common about Fort Clark, where they were heard bark- ing nearly every evening in July, and their tracks were found in the mud along the river flats. At Mandan they were reported com- mon, and a few skins were seen. At Glen Ullin, in July, Jewett. heard them barking near the town. At Sentinel Butte he found them fairly numerous all over the region, but doing little damage on the farms, which were mainly devoted to raising grain. On June 1926] MAMMALS OF NORTH DAKOTA 159 5 he located a den containing young in a rocky slope on the side of Sentinel Butte, but was unable to get the animals, which were the*n well grown. On the former North Dakota National Forest, about 25 miles south of Medora, in 1913, coyotes were found abundant, as also along the Little Missouri River near there, where their tracks were seen and the animals heard barking and howling every night. At "Wade, on the Cannonball River, in August, 1913, W. B. Bell reported them abundant. In 1915, Sheldon reported them as fairly common at Cannon Ball, Dawson, and Ellendale, and Kellogg reported them at Tokio, on the Sullys Hill National Park, at Towner, Grinnell, and all the way down the Missouri River to Bismarck. At a point several miles above Shell Village he saw six at one time running up a steep slope of Badlands. Near Elbowoods he saw five one evening, and while camping heard them every evening at dusk on both sides of the river. At Elbowoods he saw one looking at him through the wil- lows, and at Stanton found them hunting rabbits, which they dug from the burrows. Near Sather he reported them common and doing considerable damage by killing turkeys at the farms. At Cannon Ball, in 1916, Mr. Underhill said that they were common and had troubled him a good deal by killing his chickens at his farm on the river flats. At Devils Lake in June, 1916, the writer found coyotes common in the timber of Sullys Hill Park and all around the lake. Mrs. Falger told him that they howled every night around the dump heaps near the Chautauqua grounds, and their tracks were found everywhere on the sandy patches. On the south side of the lake they were common in and around the park, which was then being fenced. General habits. — Unlike the large wolves, coyotes adapt them- selves readily to conditions of civilization, and if a food supply is available they seem to thrive as well in a thickly settled country as on the open range. They are always ready to match their wits against dogs, traps, and guns, and usually have no trouble in holding their own and increasing if enough poultry, sheep, pigs, young stock, and dead animals can be found for food. They are not entirely dependent on such food, however, as they will get along comfortably on ground squirrels, pocket gophers, mice, rabbits, game birds, eggs, grasshoppers, and fruit. They are sly and to some extent foxlike in their habits, will come close to buildings at night, and usually are not permanently deterred by being chased away by dogs. Breeding habits. — Coyotes are prolific breeders, usually producing five to nine young in a litter. They often live in close proximity to farms and ranches, raising their young successfully, unless hunted with unusual persistence by one familiar with their habits. A few miles south of Ellendale, in the spring of 1915, Sheldon reported nine young captured by a farmer. On May 12, 1913, an old coyote and nine pups with eyes not yet open, were taken from a den about 20 miles west of Valley City. The old one was killed and two of the pups were kept alive in the Valley City Normal School grounds, where they were seen when they were about half grown. Two of the young and the skull of the mother were saved for specimens by Morris J. Kernall. 160 NORTH AMERICAN" FAUNA [No. 49 The young are born in burrows or cavities among rocks or in the sides of Badlands buttes, where they find abundant safe retreats until old enough to venture out in pursuit of game under the leader- ship of their parents. At Parkin, a little way up the Cannonball River, in June, the writer found a family of nearly half -grown pups living in the brushy gulches on the side of one of the big buttes just east of town. While exploring the sides of the butte for chip- munks, the writer suspected the presence of young when one of the old coyotes began barking and howling in plain view at midday in the open valley. He soon found the tracks of the half -grown pups in one of the rocky gulches and saw where they had wallowed clown the grass under the brush near the den. The anxious parents fol- lowed and would appear on first one side and then the other, at every turn doing their best to attract attention and lead away from the family. During the latter part of summer and the fall, coyotes usually hunt in family parties, but by the beginning of winter they have mainly scattered out singly or in pairs. Unlike the wolves, they seem to begin breeding when 1 year old, and, late in January, when the mating season begins, they are usually found in pairs. Economic status. — During the past coyotes have not reduced the number of injurious rodents sufficiently to protect the crops, and it has become necessary to resort to artificial means instead of depend- ing upon coyotes as a natural aid. Hence their value in this respect may be overestimated. On the other hand, their destruction of live- stock and game is in many localities so great as to make the raising of small herds of sheep impracticable, except where protected by wolf -proof fences, while the loss of other stock, poultry, and game which they destroy over the State, is very serious. In States like Montana, Wyoming, Colorado, and New Mexico, where stock rais- ing is one of the most important industries, the annual loss from coyotes is estimated at hundreds of thousands of dollars each year. The bounty system has long ago proved worse than ineffective, but the present system of cooperation between the State and the Federal authorities in employing expert trappers promises satisfactory re- sults. The full cooperation of residents throughout the coyote-in- fested regions is of the utmost importance in keeping down the numbers of the pests. The dens should be located and the young captured whenever possible. It is not probable that coyotes will ever be exterminated over the whole country, but their control over extensive areas can certainly be predicted. Vulpes fulva regalis Merriam Yellow-red Fox: (PL 17, fig. 1) EhchokuscM of the Hidatsas (Maxi- milian) ; Hir&tt-sa of the Mandans (Maximilian) ; Hirutse (Will). Vulpes regalis Merriam, Proc. Washington Acad. Sci., vol. 2, p. 672, 1900. Type locality. — Elk River, Sherburne County, Minn. General characters. — Slightly larger than the eastern red fox; slender and light, with erect, sharp ears, slender muzzle, and long tail. Winter fur, very long, full, and soft; tail, very large and fluffy; summer fur, short and North American Fauna No. 49, U. S. Dept. Agr., Biological Survey. PLATE 17 Fig. I. —Yellow-red Fox (Vulpes fulva regalis) Mounted in Agricultural College collection. Much reduced Fig. 2. — Black-footed Ferret (Mustela nigripes) Mounted specimen from Fort Rice. Much reduced -«»-•- 71>W. Fig. 3. — Varying Hare, or Snowshoe Rabbit (Lepus americanus americanus) Mounted specimen in winter coat. Much reduced North American Fauna No. 49, U. S. Dept. Agr., Biological Survey. PLATE I! Fig. I. — Plains Coyote (Cams latrans nebracensis) A captive at Mandan Fig. 2. — Otters (Lutra canadensis canadensis) Captives in National Zoological Park, Washington, D. C. 1926] MAMMALS OF NORTH DAKOTA 161 thin; tail, slender. Color in winter, rich orange yellow, paler and more straw yellow over face, back, sides, and legs; underparts whitish, tip of tail always white, back of ears, feet, and ankles black. In summer coat darker, richer orange. Young blackish, soon fading to yellowish brown or yellowish gray, with tip of tail white. Measurements of tyne of species, adult male from Elk River, Minn. : Total length, 1,117 millimeters ; tail. 420 ; hind foot, 170. Of adult female collected at Oakes, N. Dak. : Total length, 990 millimeters ; tail, 381 ; hind foot, 165. Seton (1909, vol 2, p. 707) gives the weight of one taken at Carberry, Manitoba, as 10 pounds. The darker forms of this fox, called cross-fox, silver fox, and black fox, seem to be of rare occurrence in the prairie country, but they are merely different degrees of melanism occasionally found among red foxes, and are more common farther north and west. Distribution and habitat. — The group of red foxes has not been sufficiently worked up to show the limits of range of the various forms, but all of the specimens available from North Dakota seem to be referable to the large yellow-red fox described from southern Minnesota, and apparently covering a great part of the northern Plains country. While probably never very numerous over the State, foxes were evidently much more so in the earlier trapping days than at present. In 1801 Alexander Henrv (1897, pp. 184, 198, 221, 245, 259, 281, 422, 440) reported, among" other furs taken in the Red River Vallev, 82 red-fox skins from Reed River and 102 from Park River; in 1802, 20 from Grand Forks and 29 from the Hair Hills ; 1803, 23 from Pembina River, 61 from Turtle River, and 78 from the Hair Hills; 1804, 8 from Grand Forks, 38 from Hair Hills, 4 from Park River, and 12 from Pembina River. In 1805 he reported 5G red, cross, and silver foxes from the Hair Hills, 91 from Salt River, and 31 from Pembina River; in 1806, 171 from Grand Forks, 256 fromJPembina River; in 1807, 34 from Pembina River; in 1808, 2 from the Hair Hills, 6 from Grand Forks, and 28 from Pembina River. In 1833 Maximilian (Wied, 1839-1841, Bd. 1, pp. 431^32, 1839) stated that about 2,000 red-fox skins, 200 to 300 cross foxes, and 20 to 30 silver foxes were brought annually to the fur traders at Fort Union. On February 15, 1805, Lewis and Clark (1893, p. 235) mentioned a large red fox killed at their winter quarters among the Mandans at what is now Fort Clark. Maximilian (Wied, 1839-1841, Bd. 2, pp. 86-87, 98, 1841), in 1833, while at Fort Clark, wrote in his journal: "The red fox (Canis fulvus) is very handsome and at the same time common, though by no means so numerous as the wrolves." He compared many specimens and found them in general very similar, although "the fur dealers make a different species out of every slight variation." Generally they are lighter and brighter colored than the European fox and from their beautiful fur might well be called " goldfuchs." He further says: "The black or silver fox {Canis argentatus) is met with 60 or 70 miles farther north, but it is occasionally seen here, and its skin is highly prized, being sold for $60." In 1856 F. V. Hayden (1875, p. 91) collected three skulls of these foxes from Fort Union, and reported the different varieties as cross, silver, and black, which he said were well known among the traders "and are much valued. A skin of the Silver variety . . sellinc for $100." At Valley City John Hailand told Morris J. Kernall that red foxes were numerous when he came there in 1878. In 1887 the writer was told that they were common about Devils Lake and on 162 NORTH AMERICAN FAUNA [No. 49 the prairie near Bottineau. In 1912 Eastgate collected an old female near Oakes, and nine young, about a month old, were found in the den. The same year the writer was told that there were a good many red foxes about Hankinson, where they were considered one of the standard fur animals. At Fargo there were said to be still a few in that part of the valley, and at Valley City they were re- ported as common, many being caught in winter. . At Stump Lake there were said to be a few, and in the Turtle Mountains they were reported common, but mainly on the surrounding prairie. At Ken- mare in 1913 they were said to be rather scarce, although a few were caught in winter. At Minot Mr. Booth, the taxidermist, said that they had been numerous in the early nineties, but that the great increase in coyotes since sheep were brought into the coun- try had apparently resulted in a corresponding decrease among the foxes. At Crosby in 1913 the writer was told that a few were to be found in the glacial hills to the south, but that they were rather scarce. South of Medora, on the Little Missouri River, they were said to be scarce, and there seemed to be some doubt of their being there at that time. From Grafton on April IT, 1914, Williams sent the skin of a fox pup a day old from a litter of five born in cap- tivity; again in 1917 he sent seven young, only 4 days old; in 1919 he reported red foxes fairly common and several killed every winter in the vicinity. In 1915 Kellogg found them fairly common at Wahpeton and reported a few skins taken each year by hunters. A litter of three was dug out on the farm of J. Brandt, 7 miles west of town, on May 23. At Larimore, Kellogg reported a few red foxes, and stated that a number of skins were taken each year at Manvel. At Drayton, in Pembina County, he reported a number trapped each year and two tame young ones kept in captivity. At Fort Totten he was told that red foxes had been common in that region 30 years before, but were then scarce. At Towner no records were obtained, nor at Lostwood, on the Missouri River. At Fairmount in 1915 Sheldon reported them as abundant at one time, but of late years rarely seen, although occasionally one was trapped or shot by some of the farmers. In 1916 there were said to be still a few in the country about Cannon Ball. General habits. — The yellow foxes typify all that is sly, cunning, and crafty, with a peculiar combination of timidity and boldness. They are skilful hunters, but by farmers are often looked upon as cunning thieves, because of their excessive fondness for poultry. To the fur trapper they are the acme of all that is difficult and inspiring in his craft. In keen senses, alertness, and intelligence they are excelled by few wild animals. For most of the year they are hunters, depending for subsistence on mice, ground squirrels, pocket gophers, rabbits, and birds, which they are usually able to- capture in abundance. At times they find grasshoppers and other insects acceptable, and in the blueberry season they almost live upon these delicious berries. They also eat a great variety of other berries and fruit. Although the animals are mainly nocturnal, their catlike eyes are well adapted to both night and day, and on rare occasions a fox may be seen prowling over the prairies or meadows watching for mice or small game in broad daylight. Generally, however, dur- ing the day they remain in their burrows or curl up on a knoll where 1926] MAMMALS OF NORTH DAKOTA 163 they can see on all sides, and at night range far and wide in search of food. Breeding habits. — The young are generally from five to nine in number, and are brought forth in the breeding burrows prepared by the parents, or sometimes in hollow logs or trees. Near Oakes, East- gate saw nine young about a month old that had been dug out of their den on June 5, 1912. Williams reported five and seven young in the litters of his tame foxes at Grafton. Alexander Henry reports five young caught by one of his men at Park Kiver, October 18, 1800, but these may not have been the full litter, as at that season they would be large enough to be separated. The young when first born are dull black, with a conspicuous white tip of the tail extending up from a half to three-quarters of an inch. Their eyes are said to open in eight or nine days, and by that time the black has begun to fade and the yellow-gray is appearing over the head. About the time they begin to run about, when three or four weeks old, they are usually of a dirty gray color, unless destined to represent the darker cross, silver, or black forms, in which case they retain the dark black and shoulder stripes of the cross fox or the entirely dusky fur of the future silver or black. In all cases, however, the white tip of the tail and the black posterior surface of the ears distinguish them from young coyotes, dogs, or kit foxes. Both parents are very attentive to the young, the male taking its full share in hunting and guarding the den. Apparently the young do not follow their parents in the hunt, even when well grown, as more than one fox track is rarely seen at a place in any season. Their methods of still hunting are more likely to be successful singly than in groups. Economic status. — The destruction by these foxes of game, poultry, and lambs is often discounted by the value of their skins and the interest they offer in hunting and trapping. They are known to be very destructive to lambs, however, in Michigan doing great damage among them. Apparently their fur value is sufficient to keep their numbers down to a minimum, or they might otherwise prove a pest and a serious check on the abundance of many species of small game. Their value in fur farming has not yet been fully determined, but the best grades of silver and black are among the most remunerative of fur animals to be raised under domestication. Many reports, and even books, have been published on the subject of fox farming, and experiments along this line indicate that in the future much of our choicest fur will be that produced under domestication. Vulpes velox hebes Merriam Kit Fox; Prairie Fox; Swift Ihoichka of the Hklatsas (Gilmore) ; Ohcha of the Mandans (Maxi- milian) ; 8ongina of the Dakotas (Gilmore) ; Ciuaku of the Ari- karas (Gilmore). Vulpes velox hehes Merriam, Proc. Biol. Soc. Washington, vol. 15, p. 73, 1902. Type locality. — Calgary, Alberta. Canada. General characters. — A very small fox with a rather short black-tipped tail, long, dense winter fur, and short, harsh summer hair. In winter the color is mainly dark buffy gray with orange sides, legs, and lower surface of tail, and light buffy belly. Tip of tail black and patches on sides of nose blackish. 164 NORTH AMERICAN FAUNA [No. 40 In summer the upper parts are more reddish gray. The type, adult male, measures in the flesh, in total length, 844 millimeters ; tail, 312 ; hind foot, 130. The weight, as given by Audubon (1851-1854, vol. 2, p. 14, 1851), is 8V2 pounds. Seton (1909, vol. 2, p. 700) gives 4% pounds as the weight of an adult from Saskatchewan. Distribution and habitat. — Apparently the kit foxes of the north- ern Plains at one time covered the whole of the prairies of North Dakota, but at present they are restricted to the western part of the State, and even there they have become very scarce. In 1800 they were one of the common fur animals of the Red Elver Valley. Alexander Henry (1897, pp. 184, 221, 245, 259), in his journal, recorded, among other skins taken by his trappers in the spring of 1801, 9 kit foxes from Reed River and 7 from the Park River; in 1803, 1 from the Turtle River and 23 from the Hair Hills; in 1804, 17 from the Hair Hills; and in 1805, 26 from Hair Hills and 31 from the Salt River. In " A Story of '53," which describes the fur-trading station at Walhalla, Charles Cavileer says 400 to 600 skins of kit foxes were obtained in a season, but none have been seen since the buffalo disappeared. In 1805, Lewis and Clark (1893, p. 271), on their journey up the Missouri River, spoke of the Assiniboine Indian camp, 25 miles above the mouth of the Little Missouri River, where the Indians wyere trading dried meat, grease, and skins of wolves and small foxes to British traders for liquor. In 1833, Maximilian (Wied, 1839-1841, Bd. 2, p. 51, 1841) , while at Fort Clark, reported prairie foxes frequently seen. Just below the mouth of the White Earth River he speaks of finding traces of large bears and seeing the prairie fox come out of its burrow ; and later he obtained speci- mens that were kept alive and furnished material for interesting study. Audubon (1851-1854, vol. 2, pp. 15-16, 1851) recorded them a little north of Fort Clark in 1843, and again at Fort Union, where specimens were obtained. He brought back a live one from Fort Clark to his home in New York. In 1862, F. V. Hayden (1862, p. 142) reported 50 to 100 caught every winter near each of the trading posts along the Missouri River. There are a few specimens in the National Museum collection, taken by Culbertson in 1850 at Fort Union, and others taken by Coues in 1873 on the Souris River, where he reported them common. At Grafton, in 1915, Remington Kellogg was told that the last kit fox caught in that region had been taken by Olaf Dahal in 1876. At Minot, in 1909, the writer saw several skins at a taxidermist's shop, brought in to be mounted during the previous winter, and was told that there were still a few swifts in that region, but that they were so scarce that those caught were usually preserved as curiosities. A mounted specimen in the Williams collection at Graf- ton was taken near Williston, December 16, 1911. In 1913, a few mounted specimens and several skulls were examined in the col- lection of Mr. Allen, the taxidermist, at Mandan, who said that none had been brought in in recent years. At Sentinel Butte, Mr. Crawford said that kit foxes used to be common, but had become very scarce since the country had settled up. In 1915 Kellogg reported them as very scarce in the vicinity of Goodall in McKenzie County, where they had formerly been common. He said that they were very easily trapped, poisoned, or caught by dogs, so that they did not last long after the country became settled. 1926] MAMMALS OF NORTH DAKOTA 165 General habits. — At Fort Union (Buford), on October 16, 1833, Maximilian (Wied, 1839-1841, Bd. 2, p. 37, 1841) wrote in his journal : The little prairie fox is so hungry and therefore so tame that it often visits the environs of the fort, and we found these pretty little animals among the circles of turf which were left on the removal of the Indian tents. Here they remain in the daytime and at nightfall collect and look for the remains of provisions in the neighborhood of buildings. Our dogs frequently pursue them, but their extreme swiftness enables them to escape and to retreat to their burrows, where easily caught by setting snares. On his return trip down the Missouri River to Fort Clark, in 1833. he (Wied, 1839-1841, Bd. 2, pp. 256-258, 1841) brought one of these little foxes with him as a pet and gave an interesting account of its habits. At one time, he says : During the night we were disturbed by great numbers of rats, and I placed my little tame prairie fox in the loft where the corn was kept and there he did excellent service. * * * This pretty and very tame little fox afforded us much amusement during the long winter evenings. He was nearly a year old but always glad to play with anyone. Would scratch or pat one on the clothing with his paw as he came quickly by and then make great bounds into the air as if he were pouncing upon a mouse or rat. He was very cunning and noticed everything and was delighted to be petted and stroked. He would often take some object in his mouth and shake and carry it about, dash away, hide it, look roguishly with head on one side, then come bounding back with all sorts of antics. We taught him to shake hands like a little dog, and he always offered his paw when he wished to be rubbed and petted. To rest he would roll up in a heap and cover his nose and face with his bushy tail. In cold weather he would get so close to the fire that he burned off much of his fur. He ate little, but drank often, though only a little at a time. He was very fond of rats and mice, and as with all such animals, caught them by the head. He usually chewed like a cat on one side of his mouth, using the sharp-edged molars, then licked his lips and usually one little paw. When no longer hungry he would bury the rest of his prey in the ground or in a corner, push it down with his nose, and cover it exactly as do others of the dog kind. His voice was a very loud bark, repeated three or four times in suc- cession. It is very similar to that of the European fox, but louder and rougher. It has a wonderful ring to it, and one is astonished to hear such a loud voice from such a tiny animal. Late in the fall (October 31) Maximilian (Wied, 1839-1841, Bd. 2, pp. 47-49, 1841) reported an abundance of buffaloberries, which after the frosts were very palatable. "With this fruit we refreshed our bears and mj little fox, to which they offered an agreeable variety in their food." The wild grapes, however, he says were very poor and did not suit the taste of even the little fox. In 1845, at Fort Union, Audubon (1897, pp. 116, 130), while rid- ing over the prairie, saw a swift dart from a hole under the feet of Harris's horse. Harris gave chase and gained upon the beautiful animal with remarkable quickness, overtaking it and firing at it several times, but to no purpose as it doubled and cut about in such a manner that it escaped into a ravine. A few days later Harris succeeded in shooting one, which was saved for a specimen. At Fort Clark a captive kit fox was given to Audubon, (1851-1854, vol, 2, p. 16, 1851), who carried it back to his home near New7 York. It had been kept for some months in a loft without food other than the rats and mice which it caught there. In its new home it was fed on birds, squirrels, the flesh of other animals, and any kind of fresh meat, and grew fatter every day. This probably accounts for the weight of 8y2 pounds, which seems much for this little 166 NORTH AMERICAN FAUNA [No. 49 animal in its wild state. Seton's (1909, vol. 2, p. 700) weight of 4% pounds for an adult specimen from Saskatchewan seems nearer the probable average weight of wild individuals. These dainty little foxes are among the most graceful and sprightly of native carnivores. They glide over the prairie as lightly and smoothly as passing shadows and are so quick in their motions as to have inspired fabulous stories, of their speed. They are said to be tamed easily and to make interesting pets, but they are of relatively small value in the fur market. Seton records several dens, in each of which a pair of kit foxes were found guarding five young. There is much to be learned in regard to their hunting, feeding, and breeding habits. As they are easily hunted, trapped, and poisoned, they are rapidly disappear- ing over a large part of their range and if doomed to extinction it is important that a closer study be made of their home life before it is too late.23 Family MUSTELIDAE: Weasels, Minks, Martens, Skunks, Badgers Mustela longicauda longicauda Bonaparte Long-tailed Weasel; Ermine Ohsisa of the Hidatsas, and Malich- pach-piraka of the M a n d a n s (Maximilian.) Mustela longicauda Bonaparte, Charlesworth's Mag. Nat. Hist., vol. 2 (n. s.), p. 38, 1838. Type locality. — Carlton House, Saskatchewan, Canada. General characters. — One of the largest of our weasels, with slender body and long tail. In summer, upper parts, yellowish brown, darker on face, with tip of tail black ; underparts and usually feet and toes, yellow, varying from rich buff to deep orange. In winter pure white, except for the black tip of tail and usually a light-yellowish wash or stain on the belly, hind legs, and tail. Measurements of an average-sized male from Crosby, N. Dak. : Total length, 445 millimeters ; tail, 150 ; hind foot, 50 ; of an average female from Lostwood : 369, 121, and 40, respectively. A male from Treesbank, Manitoba, measured 457, 163, 49, and weighed 13 ounces. The change from summer brown to winter white comes usually in November or with the first permanent snows. A male collected at Jamestown, November 1, and another at Castleton, November 3, 1892, are nearly white, with only a mixture of brown hairs over the back sufficient to produce a brownish gray. In the agricultural college collection at Fargo an adult male taken October 24, 1912, has the back, top of head, neck, and tail brown, and the sides and under- parts, including the lower surface of tail, white, while an adult female taken November 2, 1912, is pure white, except for the black tip of the tail. A female taken at Valley City on November 13 is pure white with only a trace of sulphur yellow on the tail. Three specimens collected at Valley City, by Morris J. Kernall, on October 27, 1912, show three stages in the fall change ; one has the back mostly brown, with sides, belly, and tail mostly white ; another is mainly white, with a little brown on the back; and the third is in the full white winter coat. The spring change from white to brown comes approxi- mately with the normal disappearance of the winter snow, but is not repre- sented in the North Dakota series of specimens. 23 Gray fox (Urocyon cinereoarjenteus) (Schreber). There is only one record for the State and this seems doubtful. In 18SS, at Fort Clark, Maximilian (Wied, 1839-1841, Bd. 2, p. 86, 1841) wrote: "The gray fox (Cams cinereo argent&u-s) and the cross fox (Canis decussatas) are likewise found here." This record seems very questionable, as no other report of their occurrence is to be found so far north in the prairie country. It seems probable that Maximilian confused the silver-gray fox of the genus Vulpes with this species, or merely applied the wrong name to the silver-gray. Perhaps the note refers to furs traded by the natives at Fort Clark, but really brought from points farther south. The northern edge of the Black Hills in South Dakota is. the nearest point at which gray foxes are known to occur. 1926] MAMMALS OF NORTH DAKOTA 167 Distribution and habitat. — The long-tailed weasels cover the northern Plains country and are the most common of the weasels over practically all of North Dakota. They are common all over the prairie part of the State, and the forest of the Turtle Mountains region is sufficiently open to attract them. They are prairie dwell- ers, ranging over the wide open expanse of country and making their homes in the burrows of the numerous rodents on which they prey. General habits. — These large weasels may often be seen over the prairie, running rapidly from one to another of the ground squir- rels' burrows, and when alarmed taking refuge in the burrow nearest at hand. In 1833 Maximilian found their skins among those much prized by the chiefs of the Indian tribes along the upper Missouri River, but in those days they seemed not to have been included among the marketable furs of the white trappers. In 1887 the writer took specimens of this weasel at Devils Lake and Bottineau in traps set for pocket gophers and ground squirrels, and again while at Devils Lake in 1914 was surprised to see one thrust its head out from under a board sidewalk where crowds of people gather to take the electric car for the lake. In 1892 Loring collected speci- mens at Castieton, Valley City, Larimore, and Jamestown, taking most of them in traps set in the burrows of pocket gophers. In 1909, while riding over the prairie near Lemon, in the southwestern part of the State, the writer saw one running from one burrow to another of the 13-lined ground squirrels. The squirrels were greatly excited and were calling shrilly back and forth over the surround- ing prairie, evidently passing along the word of great danger. At Buford, in 1910, Anthony reported the weasels quite common, some of them making themselves at home among the ranch buildings for several days. At Lisbon, in 1912, Eastgate reported them fairly common, but rarely seen. During the summer, living where the pocket gophers and ground squirrels are thickest, they destroy large numbers of these rodents, rarely entering poultry yards or killing chickens. Many are trapped during the winter. At Valley City, in 1912, one of these weasels came close to the writer in its pursuit of ground squirrels, running quickly from one burrow to another. On seeing him it stood erect, tall, and straight, as a snake will often raise its head to look over the top of the grass. It then ran into a badger hole, but, full of curiosity, soon reappeared and raised its head and neck in full view to watch. It was needed for a specimen, and the writer hoped to find out what it was eating, but its stomach was empty. At Crosby, in the northwestern corner of the State, while driving over the prairie, a large weasel was seen running from burrow to burrow, while the ground squirrels from far and near were uttering shrill whistles in a panic of fright. To obtain it for a specimen, it was only necessary to frighten it down a burrow and then wait a moment until it reappeared and raised its head and neck from the grass for inspection. As usual, its stomach and intestines were entirely empty, in spite of the fact that it was evidently engaged in killing squirrels. In 1915, Sheldon collected a very large, dark-colored male near Oakes, in Dickey County, and reported the weasels fairly common through- out that part of the State. The same year Kellogg reported them 168 NORTH AMERICAN FAUNA [No. 49 common all across the central part of the State, and down the Missouri Eiver Valley from Goodall to Bismarck. At Hankinson and Wahpeton it was said that a good many weasels were caught each winter for fur, and at Grand Forks Kellogg reported many brought in to the fur market, where they sold for $1.25 each. He also reported considerable numbers of them taken for fur at Drayton, Towner, and other points along his route across the State. In the Turtle Mountains the writer was told that weasels were considered by the trappers as one of the important fur animals. At Wade Bell reported them fairly common and was told that one trapper had caught 46 during the previous winter. Breeding habits. — Of the breeding habits of these weasels there seems to be little definitely known. At Manvel, Grand Forks County, Kellogg was told of a litter of 11 young found by William Brown, but this seems a large number for any weasel to have at one time. Apparently they do not multiply very rapidly, as their abundance seems never to increase beyond a few scattered individ- uals found over the country. Food habits. — The actual determination of the food of weasels is difficult, as examination of stomach contents rarely shows a trace of any food and generally the whole intestinal tract seems to be empty. The weasels when seen are usually chasing ground squirrels or putting their heads out of the burrows of squirrels or pocket gophers, which they enter freely, and where they find the occupants helpless against their attacks. From their well-known habit of killing many more animals than they can eat and the ease with which they can capture the ground squirrels and pocket gophers, it is evident that they are killing for the sheer lust of it as well as for a little blood, which they take from each individual and which is quickly digested. In places where a weasel remains for some time, the ground squir- rels and pocket gophers usually disappear, but generally the weasels are great wanderers, covering new hunting grounds every day. In winter, when the burrows are filled with snow, the weasel tracks show that mice are the principal game sought. Open spaces under logs and brush or fallen grass are entered through the snow and often the tracks reappear on the surface a considerable distance away. In soft snow the weasels often force their way down to the surface of the ground and plow tunnels through the snow, evidently in pursuit of mice and small game. If game is not to be found in sufficient abundance, they will feed on any frozen meat or old carcass that is available, and on rare occasions they find their way into henhouses and sometimes do serious mischief before they are discovered and checked. It is not improbable that they kill some wild birds and possibly eat the eggs, but there is little evidence of their doing so in a country where ground squirrels, pocket gophers, and mice are abundant. Economic status. — Ordinarily many weasels are caught in traps set for other fur-bearing animals such as minks, martens, and foxes, but where these large weasels occur in considerable numbers and bring a good price, the trappers seem to devote their attention especially to catching them. This is easily done, as they are en- tirely unsuspicious and are easily attracted by bait of fur or feathers 1926] MAMMALS OF NORTH DAKOTA 169 scattered around or above the trap. It seems a great mistake, how- ever, in a region of numerous rodent pests to destroy the greatest enemy of such animals. Even if it is possible to destroy by arti- ficial means all of the ground squirrels and pocket gophers over a considerable extent of country, the mice and smaller rodents still remain in abundance, and if their increase goes unchecked serious losses of crops are sure to follow. The occasional mischief done by weasels in the poultry yard can usually be prevented by a little care in making the buildings tight and secure by wire mesh. The value of weasel fur, winch is sold as " ermine," is in most cases far less than the economic value of the animals as rodent destroyers. Mustela cicognanii cicognanii Bonaparte Bonaparte Weasel; Short-tailed Weasel Mach-schipka of the Mandans (Maximilian). M[ustela] cigognanii [sic] Bonaparte, Charlesworth's Mag. Nat. Hist., vol. 2 (n. s.), p. 37, Jan., 1838. Type locality. — Northeastern North America. General characters. — A medium-sized weasel with moderately short tail. In summer, upper parts light brown ; underparts white, usually tinged on belly with sulphur yellow ; winter coat, pure white or slightly tinged on belly, hind legs, and tail with sulphur yellow ; tip of tail, always black for about an inch at end. Measurements of large adult male, from Walhalla : Total length, 338 millimeters ; tail, 9S, hind foot, 44. Weight, 6 ounces. Female much smaller, one from New York State measures 260, 72, and 31 millimeters, respectively. Distribution and habitat. — A specimen of the Bonaparte, or short- tailed, weasel from North Dakota was collected by H. V. Williams at Stump Lake on May 6, 1913. Another specimen examined in the collection of the biological laboratory at Devils Lake in 1914 was taken near there by Eastgate. A specimen listed under richardsoni in the catalogue of the Field Museum from Minot is undoubtedly also this species (Elliot, 1907, p. 449). In 1833, Maximilian (AVied, 1839-1841, Folio, Tab. 13), at the Mandan villages, had a drawing made by his artist, Carl Bodmer, of one of the Mandan chiefs, Mato-Tope, dressed in full regalia and wearing many skins of both large and small species of white weasels with black-tipped tails. In December, 1912, Eastgate, on a trip from Bottineau to St. Johns, along the edge of the Turtle Mountains, reported both the large and small weasels very common and says : " I was never out of sight of their tracks in the soft snow and saw many skins of the larger kinds with the trappers' furs. The trappers did not bother to skin the small weasels." General habits. — The writer's experience with these weasels has been mainly in Minnesota, where in the early eighties they were common and often came about the buildings m winter during the time of deep snow, and got into mink traps set along the streams and lake shores. Their tracks were found everywhere, but mostly in the woods or along fence rows and through thickets. On the meadows they would run from one haystack to another, or along the creek banks, where they would find or make openings to the surface of the ground under fallen grass or reeds. Here they were 170 NORTH AMEBIC AN FAUNA [No. 49 always hunting mice and the small animals that remained active under the snow during the winter. In places where the weasels were most abundant the mice always became noticeably scarce before spring, and when the snow disappeared the mouse crop seemed always to be at its lowest ebb. In the early pioneer days of log barns and rough buildings these weasels would occasionally make their homes, in the barns and out- buildings for a time during the deep snows, and remain as long as there were rats and mice for them to feed upon. Occasionally they would get into the poultry houses and clean out the mice without doing any damage to the poultry, and when the mice were gone they would leave the building and go to the woods or find other hunting fields. There are, however, many authentic reports of their destroy- ing large numbers of chickens and apparently killing them for sport as well as for food. Generally, however, the larger weasels are much more destructive to poultry where it is unprotected. The small size of these weasels, especially of the females, seems to limit their prey largely to mice and small rodents, and the number killed by one of these tireless, bloodthirsty little animals during the course of a year must be enormous. Economic status. — The snowy white skins of these weasels in winter make some of the choicest ermine, but their small size for- tunately limits their value, and many of those caught in traps set for minks, martens, and other animals are not even saved by the trappers. Their value as mice and rodent destroyers seems far to outweigh their fur value and greatly to overbalance the relatively small amount of damage done to poultry and game. In most parts of the country it would seem advisable to protect the weasels, al- though they are generally hardy animals, well able to protect them- selves unless the price of their skins runs high enough to induce trappers to make special efforts to get them. Mustela rixosa rixosa (Bangs) Least Weasel Hitunka-san (white mouse) of the Dakotas (Gilmore) Putorius rixosus Bangs, Proc. Biol. Soc. Washington, vol. 10, p. 21, 1896.. Type locality. — Osier, Saskatchewan, Canada. General characters. — Smallest of all native weasels, full-grown individuals measuring about 6 or 8 inches in length. The tail is very short, without black tip at any time. In summer the upper parts are dark brown, underparts, white ; in winter the whole animal is pure white. An adult male from Grafton, N. Dak., measured in the flesh : Total length, 202 millimeters ; tail, 39 ; hind foot, 25. A smaller male from Bowdon measured 155, 34, and 21 millimeters ; and an adult female from Alaska, 165. 18, 21 millimeters, respectively. Distribution and habitat. — The tiny least weasel occupies the Boreal Zones of much of the northern part of the continent. There is one specimen in the Biological Survey collection from North Dakota, taken by H. V. Williams at Grafton, October 24, 1913. An- other specimen in the Williams collection was examined by Reming- ton Kellogg in June, 1915. Kellogg also reported a specimen in the collection of the State university, taken at Fort Totten, July 21, 1913, and he was told that the species is trapped occasionally in the timber around Devils Lake. At Manvel, Grand Forks County, 3 926] MAMMALS OF NORTH DAKOTA 171 he was told by trappers that a very small weasel was occasionally caught, but was not saved, as its fur had no value. At Fort Buf ord Anthony reported a small weasel that might be of this species. At Tolna, near Stump Lake, Eastgate reported the species as "said to occur." At Bowdon M. C. H. Murie took an adult male in brown summer fur July 27, 1918. In 1833 Maximilian (Wied, 1839-1841, Bd. 2, p. 98, 1841) collected a specimen which was evidently of this species at Fort Clark, but unfortunately it was lost on the return journey. His measurements of Qy2 inches for total length, 1*4 inches for tail, and 7y2 lines for hind "foot, and his statement that in winter it becomes " ganzlich weiss " seem to identify it beyond question. Apparently these little weasels are very scarce even in the midst of their range, and it is not surprising that so few have been taken in North Dakota on its extreme border. General habits. — Apparently least weasels are strictly mouse hunt- ers, and their small size enables them to follow the runways and underground burrows of almost any mouse. The specimens taken by collectors are usually caught in mouse traps. They are such in- conspicuous animals, either in the dark-brown summer coat or pure- white winter coat, that it is not surprising that they pass unnoticed; but the fact that with all the trapping for the different small rodents few of these weasels have been found seems unquestionable evidence of their rarity. Occasionally fine tracks are seen in the snow that may have been made by this species, but these probably in many cases may be attributed to the very small females of the short-tailed weasel. In habits they do not differ from other weasels, except as limited by their diminutive size. Economic status. — Although too small to do any serious harm to poultry or to be of any value for fur, these little animals certainly serve as a valuable check on the increase of mice. Studies in south- ern Manitoba by Stuart Criddle (1926) have demonstrated their effectiveness in controlling the colonies of Microtus minor, and they are undoubtedly equally beneficial in other parts of their range, and with other species of mice. If they could be domesticated, it seems probable that they might be of value in destroying mice around buildings, and that poultry and other animals would be safe from them. If a family of young could be obtained for breeding purposes, it might be well worth while to test their usefulness. Mustela nigripes (Audubon and Bachman) Black-footed ferret (PL 17, fig. 2) Etopta sapa of the Yankton Sioux; Xazi of the Mandans ; Tahu akii- kahak napish of the Hidatsas (all, Gilmore). Putorius mqripcs Audubon and Bachman, Quadr. North Amer., vol. 2, p, 207, 1851. Type locality. — Fort Laramie, Wyo. General characters. — A large, heavy-bodied weasel with rather large ears, short tail, and short fur. Color, creamy yellow with a wash of brown over 82242'— 26 12 172 NORTH AMERICAN FAUNA [No. 49 middle of back and top of head; feet, legs, tip of tail, and mask across face and around eyes, blackish. An adult male measured by Osgood, total length, 529 millimeters ; tail, 130 ; hind foot, 65 ; and adult female from Quinion, measured by Jewett, 510, 128, and 61, respectively. Distribution and habitat. — The black-footed ferret, like the black- tailed prairie dog, has a wide range over the Plains country from Texas to Alberta. A few have been taken in western North Dakota, west and south of the Missouri River. In 1910, the late Howard Eaton told Gary of a ferret skin which he had bought at a Crow Indian fair, and said to have come from the Little Missouri River near Medora, where he has since reported them near his old ranch. In 1912 one was snared by some Indians near Fort Rice and given to H. C. Fish, curator of the Historical Society Museum, at Bis- marck, and later was given to Bell for the agricultural college col- lection at Fargo. On June 20, 1913, Jewett collected a fine' adult female near Quinion between the Killdeer Moimtains and Medora. Describing the incident, he writes: While driving along the road through a large prairie-dog town about 2 o'clock in the afternoon, I saw a ferret's head disappear into a prairie dog's burrow only a few yards distant from the horses' feet. I jumped out of the wagon without stopping the team and almost immediately the head of the ferret reappeared and I shot it. It proved to be an adult female, evidently with young, as the mammae contained milk. I had been told by old settlers that there were no ferrets in this region, and when I showed the specimen to several no one knew what it was, so they are evidently quite rare in this part of the State. In 1915, at Stanton, Kellogg saw a mounted specimen in a taxider- mist's shop, which was said to have been killed near there. General habits. — At his ranch on the Little Missouri River in the eighties, Roosevelt (1900c, pp. 85-86) writes of the ferret: It makes its home in burrows, and by preference goes abroad at dawn and dusk, but sometimes even at midday. It is as blood-thirsty as the mink itself, and its life is one long ramble for prey, gophers, prairie dogs, sage rabbits, jack rabbits, snakes, and every kind of ground bird furnishing its food. I have known one to fairly depopulate a prairie-dog town, it being the arch foe of these little rodents, because of its insatiable blood lust and its capacity to follow them into their burrows. Once I found the bloody body and broken legs of a poor prairie hen which a ferret had evidently surprised on her nest. Another time one of my men was eye-witness to a more remarkable instance of the little animal's blood-thirsty ferocity. He was riding the range, and being attracted by a slight commotion in a clump of grass, he turned his horse thither to look, and to his astonishment found an antelope fawn at. the last gasp, but still feebly struggling, in the grasp of a ferret, which had throttled it and was sucking its blood with hideous greediness. He avenged the murdered innocent by a dexterous blow with the knotted end of his lariat. Most of the records of the black-footed ferret throughout its range are from prairie-dog towns, where ferrets are almost invari- ably found running from burrow to burrow or taking refuge in the underground retreats. Evidently their principal prey consists of prairie dogs, although so far as is known they have never been seen actually catching and killing one. While apparently very useful in destroying prairie dogs, they are so rare that little impres- sion is made upon the population of extensive prairie-dog towns. With the abundance of easily procured food it seems strange that they should remain so scarce. It is possible, however, that this very abundance has in some way pauperized the species until reproduc- tion is restricted. 1926] MAMMALS OF NORTH DAKOTA 173 As in other weasels, the mammae are arranged in 3 pairs well back, 2 of these pairs close together in the inguinal region, and 1 pair a little farther out on the posterior part of the abdomen. Ap- parently nothing is known of the breeding habits or of the number of young at a birth. Lutreola vison letifera (Holiister) Mink DokslWca of the Dakotas (Riggs and Williamson) ; Daktsua of the Hi- datsas (Matthews). Naksua (Gil- more) ; Monika sunt ike of the Man- dans (Will), Mini-gasundek (Gil- more) ; Eruch of the Arikaras (Gilmore). Mustela vison letifera Holiister, Proc. U. S. Nat. Mus., vol. 44, p. 475, 1913. Type locality. — Elk River, Minn. - General characters. — Size, medium, not so dark as the average of minks from farther north. In prime early-winter fur the color is a rich dark brown, darkening to blackish on the tail, with usually a white patch on chin, throat, or breast. Later in the winter and in spring the color fades out to a paler brown and in summer the short, harsh fur is yellowish brown. Measure- ments of large adult male from Lake Irwin, North Dakota : Total length, 697 millimeters ; tail, 230 ; hind foot, 81. Of adult female from same place. 561, 17S, and 67. Weights of the two, 3 pounds 12 ounces, and 2 pounds 5 ounces, respectively. Distribution and habitat. — Until the minks have been thoroughly revised the limits of range of the different forms will necessarily remain somewhat in doubt. There are so few specimens from North Dakota that it is not possible to say whether more than one form is represented in the State, nor to determine the extent of the range of letifera,. It is quite probable that specimens from the northern part of the State could be referred to the larger, darker lacustris described by Preble from Manitoba, Canada. In the Biological Survey collection there is a female collected at Stanton, on October 6, 1915, and an immature male from Fargo, taken December 27, 1918. In the National Museum collection are four skulls taken by Coues on the Souris River in 1873, and there is a specimen in the agricultural college collection at Fargo, taken on Apple Creek, near Bismarck, in 1914. Although never numerous, minks seem to have been fairly common along most of the streams in the State. In 1801, Alexander Henry (1897, pp. 184, 198, 221, 245, 259, 281, 422, 440) reported 68 skins irom Reed River and 29 from Park River; in 1802, 6 skins taken at Grand Forks; in 1803, 39 taken on the Pembina River, 3 on the Turtle River and 8 in the Hair Hills; 1801, 13 at Grand Forks, 2 in the Hair Hills, and 2 on Pembina River; 1805, 14 in the Hair Hills, 5 on Salt River, and 41 on Pembina River; 1800, 35 at Grand Forks, and 141 on Pembina River; 1807, 21 on Pembina River; and in 1808, 7 in the Hair Hills, 18 at Grand Forks, and 63 on Pembina River. At Fort Union, in 1883, Maximilian (Wied, 1839-41, Bd. 1, pp. 431, 432, 1839) reported a few thousand mink skins brought in by the trappers each year. In 1873 Coues (1877, p. 175) reported many minks taken on the Mouse River. In 1887 they were reported |74 NORTH AMERICAN FAUNA tNo. 49 common at Harwood, Grand Forks, Pembina, Devils Lake, and in the Turtle Mountains. In 1912 the writer was told that a good many were caught around the lakes in the vicinity of Hankinson each winter and that in the Turtle Mountains they were considered the most important fur ani- mal of the region. At Stump Lake and along the Sheyenne River, Eastgate reported them as occurring in limited numbers, and at Kathryn, in Barnes County, he reported six caught by one trapper during the winter of 1912. At Lisbon, Eansom County, he reported them as rather rare, but found on every river and creek, and on many of the deeper sloughs and lakes. At Fairmount, in 1915, Sheldon reported them as becoming rare, although a few were trapped each winter along the Bois de Sioux Kiver. At Wahpeton a few were said to be caught along the river each year. At Larimore and Manvel, in Grand Forks County, Kellogg reported quite a num- ber trapped each year. At Grafton he reported them fairly common along Park Kiver, where many were trapped in winter, and at Dray- ton, "in Pembina. County, a good many trapped by the half-breeds in winter. At Devils Lake he says they were not very common, but a few were taken each year, and at Towner he saw the tracks of one on the banks of Mouse River and learned that a few were trapped in winter. At Goodall he reported them quite common along the creeks and river, and near Elbowoods an Indian had caught six on Shell Creek during the preceding winter. At Stanton he took one specimen on Knife River and along the river near Sather and Wogansport he saw a few tracks, but considered the animals rather scarce. At Bismarck, in 1914, the writer caught one in a beaver trap set on Apple Creek, and at Wade, on the Cannonball River, in 1913, Bell reported them as fairly common. In 1919, Murie reported them in fair numbers along the Red River near Fargo and a few on the James River and near Bowdon, and the writer found tracks along the Heart River near Mandan, in the Pembina Hills, and along the Red River near Grand Forks. Wherever there are streams or exten- sive lakes, minks seem to be holding their own fairly well over the State and will probably never be entirely exterminated even by per- sistent trapping and a rather high value on their skins. The days of the professional trapper seem nearly at an end, and if the minks have been able to withstand his skill for more than a hundred years they will doubtless persist for a long time with only local trappers to contend with. General habits. — Minks are semiaquatic animals, usually found near streams, where they do much of their hunting for small game, both in the water and on the banks. They are great hunters, with some of the bloodthirsty ferocity of the weasel, always eager to kill whatever they claim as game. Eastgate reports digging out a mink den at Sweetwater Lake, where he found 9 full-grown muskrats, 4 ducks, 5 coots, several smaller birds, some mice, and one small jack rabbit, that had been killed and brought in for food. This was undoubtedly a breeding den, as it is only during the breeding season that the mink remains in one locality long enough to bring in such stores of food. For most of the year minks are wanderers over somewhat extended hunting grounds. In winter, when their habits can best be observed by watching their tracks, the same mink usually 1926] MAMMALS OF NORTH DAKOTA 175 makes its round every few days with the varying regularity of a free lance. An abundance of safe retreats are found in the hollow banks of streams and lakes, often in muskrat burrows or houses, the owners of which have been killed or driven out in terror of their lives. A hollow tree or log is often used as a refuge or resting place. Minks climb trees readily when hard pressed by dogs and on sev- eral occasions while hunting raccoons the writer has shaken a mink out of the topmost branches of a tree for the waiting dogs below. Even then the dogs are not sure of their game, as the mink is wea- sel-like in its quickness at dodging and avoiding enemies. But if cornered, minks never refuse a fight with anything that comes their way and often terrify a dog by their savage screams as with light- ninglike motions they fasten their keen teeth into his nose or lips. Their pungent odor, from an amber-colored liquid carried like that of the skunk in two glandular sacs surrounding the anus, is used as a method of defense, and though quite different from that of the skunk it is equally offensive to man or beast. Breeding habits. — The five or six young are usually born in May and zealously cared for in the den by the mother mink until old enough to follow her on her hunting trips. Before the trapping sea- son begins in early winter they are practically full grown. The male has no part in the family affairs after the brief mating season, and as soon as the young are large enough to capture their prey the family disperses, and each is thereafter a solitary hunter. Food habits. — The natural food of minks consists mainly of ro- dents, birds, fish, and crustaceans. Among the rodents the muskrat is one of the favorites, and empty muskrat houses with a small round hole in one side usually indicate a family that has been destroyed by a mink. Sometimes a small pond will be entirely depopulated of muskrats before the mink leaves the vicinity, but in larger bodies of water the muskrats appear to escape to other houses or burrows and do not return until the mink has departed. Meadow mice ap- parently furnish considerable food for minks. Rats and rabbits are also captured for food, and wild ducks and other waterfowl, small birds, game, and poultry are equally acceptable. In places small fish furnish a large part of the food of minks, which often capture fish as large or larger than themselves. Crawfish and other crusta- ceans are greatly sought wherever they can be found and in many places form the principal food, as shown by the scattered droppings along the trails or about the dens. Frogs are eaten, but are evi- dently not a favorite food. At times dead animals, and especially frozen carcasses, are eaten when other food is not available, but live game that they can kill for themselves seems always to be pre- ferred to all else. In captivity they will eat bread or cereals soaked in milk and many table scraps, but only when fresh meat and blood are not to be had. Economic status. — Locally the minks have been known to do con- siderable damage to poultry. At Willows, N. Dak., in December, 1886, David H. Herman wrote to the Biological Survey that a mink killed all of his hens one night and the next night spent its time trying to climb up the sides of the house to get at those hung up from the previous night's kill; the third night it came back and killed the rooster, the only remaining bird of the flock, and was 176 NOKTH AMERICAN" FAUNA [No. 49 found breakfasting on it in the morning. One of Mr. Herman's neighbors also lost 51 fowls in one night, the mink being killed with a stick in the henhouse the following morning. In 1912 Eastgate reported minks at Lisbon as doing some damage to poultry dur- ing fall and winter. So serious is this occasional damage that near streams or lakes it is generally necessary to protect poultry houses with some kind of mink-proof structure. The destruction of game, and especially waterfowl, is probably far more serious than is generally supposed, but the guilt is not easily divided between minks, weasels, skunks, and foxes. On the other hand, considerable credit is due the minks for destruction of rats, mice, and other troublesome rodents. Their fur value usually as- sures them protection during the season when fur is not prime. In the absence of statistics of annual fur values it seems safe to assume that minks alone contribute many thousands of dollars a year to the local trappers over the State. Although minks are easier to trap than foxes, the boy who can catch his half dozen in a season without devoting undue time to his trap line can take considerable satisfaction in his skill. Sufficient information has not been obtained to determine the practicability of domesticating minks, but with proper handling they have in some instances proved successful on a small scale. Consider- able has been written on their management in captivity, but further tests are necessary to show that they can be produced economically. Martes aniericana americana (Turton) Marten; Pine Marten; American Sable [Mustelal americanus Turton, Linnaeus, Syst. Nat, vol. 1, p. 60, 1806. Type locality. — Eastern North America. General characters. — About the same size as the mink, with longer legs, larger ears, longer and softer fur, and more bushy tail. Color usually lighter, more yellowish brown than the mink, varying from dull orange to dark chest- nut ; throat usually light yellow to deep orange. Measurements of adult male from Montana : Total length, 615 millimeters ; tail, 200, hind foot, 93 ; of adult female, 565, ISO, and 83, respectively. Distribution and habitat. — At present there are probably no martens in North Dakota, but in 1801 Alexander Henry (189T, pp. 184, 198, 245, 259, 281, 422, 440) recorded among others taken, 26 marten skins from Reed River and 36 from Park River ; in 1802 he reported 13 from the Hair Hills; 1803, 9 from Pembina River, 26 from the Turtle River, and 47 from the Hair Hills; 1804, 21 from Grand Forks, 3 from the Hair Hills, 1 from Park River, and 5 from the Pembina River; 1805, 6 from the Hair Hills and 3 from the Pembina River; 1806, 4 from Grand Forks and 271 from Pembina River ; 1807, 75 from Pembina River ; 1808, 2 from the Hair Hills, 6 from Grand Forks, and 69 from Pembina River. In his " Story of '53 " regarding the fur trade at Walhalla, Charles Cavileer says 700 martens were taken one winter. In 1833 Maximilian (Wied, 1839-1841, Bd. 1, pp. 431-432, 1839) gave a list of the approximate number of furs bought at Fort Union during the year, and among these, marten skins numbered 500 or 600. These, however, were undoubtedly brought down the river from wooded country farther north and west. Apparently martens 1926] MAMMALS OF NORTH DAKOTA 177 were originally fairly common in the timbered sections of north- eastern North Dakota, but the beauty and value of their fur caused the early destruction of the species in that part of the State. General habits. — Martens are timber-loving animals and are rarely found away from forests or the vicinity of trees. They are not only Boreal in range, but largely arboreal in habits, seeking much of their prey under cover of brush and trees and pursuing squirrels and chipmunks up tree trunks and among the branches. They are rarely found along streams, but range at large through the woods, where their winter tracks may be distinguished from those of the mink by larger feet and longer reach. There is nothing recorded of their food habits in this region, but in other parts of the country their natural food consists largely of mice, squirrels, rabbits, and birds ; they are also known to eat berries, insects, and a variety of animal and vegetable foods. Economic status. — It is not probable that martens, even with careful protection, would ever return to restock the limited forest areas of North Dakota, but in captivity they give some promise of becoming of practical value for fur farming. In the northern part of the State, especially the forest area of the Turtle Mountains and Pembina Hills, their fur should become dense and fine, as the Boreal climate represents their original habitat. If the experiments being carried on in raising martens prove successful, North Dakota should be found well adapted to the industry. Martes pennanti pennanti (Erxleben) Fisher; Pekan; Black Cat [Mustela] pennanti Erxleben, Syst. Regni Aniin., p. 470, 1777. Type locality. — Eastern Canada. General characters. — About twice the size of tbe marten, -with relatively long legs, long tail, and coarse fur. Colors, blackish with a grizzled cape over top of head, neck, and shoulders. A large male measured in total length 1,020 millimeters ; tail, 400 ; hind foot, 143 ; a female, 835, 343, and 115, respectively. Distribution and habitat. — Fishers belong to the Boreal Zone forests of the northern part of the continent, and in the early trap- ping days reached into northeastern North Dakota. On September 26, 1800, Alexander Henry (1897, pp. 103,- 122, 184, 198, 221, 245, 259, 281, 422, 440) reported one seen at the mouth of the Park River, and on October 19 wrote in his journal at the same locality that some fishers were brought in daily by the trappers. In the spring of 1801 he recorded 108 fisher skins from the Reed River and 70 from the Park River; in 1802, 23 from Grand Forks and 57 from the Hair Hills; 1803, 69 from the Pembina River, 98 from the Turtle River, and 111 from the Hair Hills; 1804? 36 from Grand Forks, 30 from the Hair Hills, 16 from the Park River, and 21 from the Pembina River; 1805, 74 from the Hair Hills, 14 from the Suit River, and 25 from the Pembina River ; 1806, 59 from Grand Forks and 140 from the Pembina River; 1807, 78 from the Pembina River; 1808, 46 from the Hair Hills, 14 from Grand Forks, and 29 from the Pembina River. Apparently the animals were not uncommon then, as the number of skins usually ran higher than that of mink and marten and many of the other fur bearers that were being 178 NORTH AMERICAN FAUNA , [No. 49 taken. In 1853 Charles Cavileer, at Walhalla, reported 400 fisher skins a year as not an unusual number obtained by the fur company of which he was agent, but many of them doubtless came from beyond the borders of the State. In 1833 Maximilian (Wied, 1839-1841, Bd. 1, pp. 431-432, 1839) , in listing the approximate number of skins annually brought in to Fort Union (now Buford), gives the fisher as 500 to 600. Some of these may have come from the Turtle Mountains, Souris, and the Mouse River country, but probably more of them were brought down the Missouri and Yellowstone from farther west. At the pres- ent time there are certainly no fishers within the State and there seem to be no authentic records of their occurrence since the early trapping days. General habits. — Fishers, like martens, are mainly forest-dwelling animals, seeking their prey of small mammals, rabbits, squirrels, and birds among the trees and brush and wandering at large over the woodland areas. They are expert climbers and pursue and capture squirrels in the treetops. The common name applied to them is an evident misnomer, as they are not known to catch fish or to frequent streams or bodies of water. The names " black cat " and " pekan " are also used for them, but less commonly than that of " fisher." Economic status. — The fur of the fisher, although hidden by long, coarse hairs, is full, soft, and durable, and the general effect of prime skins made into wearing apparel is very pleasing. They are counted among the more valuable furs, and have always brought a high price in the fur market. For this reason the animals have dis- appeared or become scarce over much of their original range, but are still taken in some numbers in northwestern United States, Alaska, and Canada. Gulo luscus (Linnaeus) "Wolverene; Glutton; "Indian Devil" Eh-tupah of the Hidatsas, Mato-ka of the Mandans (Maximilian). [Ursus] luscus Linnaeus, Syst. Nat., ed. 12, t. 1, p. 71, 1766. Type locality. — Hudson Bay. General characters. — A heavily built little animal with short ears, short legs, and short, bushy tail. Fur soft and light, covered with long coarse overhairs. Color dark brown, or blackish, with a yellow gray band along sides and across rump, and more or less gray over top of head and shoulders ; throat, breast, and sometimes belly usually with irregular white spots. A very large male from Alaska, collected by Charles Sheldon, measured in total length, 1,070 millimeters ; tail, 218 ; hind foot, 190 ; and weighed 36 pounds ; an adult female from northern Mackenzie, measured by Preble, 920, 200, and 175 milli- meters, respectively. Distribution and habitat. — The wolverenes are Boreal animals ex- tending across the northern part of the country and southward into the high mountain region. In North Dakota they apparently occu- pied at least the northeastern part of the State in the early trapping days and possibly the north weastern part. In 1801 Alexander Henry (1897, pp. 184, 198, 221, 245, 259, 281) reported, among other skins taken by his trappers, 2 wolverenes from the Reed River and 3 from the Park River; in 1802, he reported 3 from the Hair Hills; 1803, 1926] MAMMALS OF NORTH DAKOTA 179 4 from the Pembina Kiver ; 1804, 3 from Grand Forks, 1 from the Park River, and 2 from the Pembina River ; in 1805, 1 from the Hair Hills and 5 from the Pembina River ; in 1806, 1 from Grand Forks and 10 from the Pembina River. They were not mentioned by Maxi- milian among the skins brought in at Fort Buford in 1833, but in 1842 Harris included them in his list of mammals of the upper Mis- souri territory from Fort Leavenworth to Fort Union. A specimen brought from Fort Union by Mr. Culbertson in 1850, for the National Museum collection, probably, as Baird (1857, p. 182) says, was brought to Fort Union from some of the posts toward the Rocky Mountains. The fact that both the Minnetaree and Mandan Indians have names for this animal is suggestive of its occasional occurrence in the upper Missouri region of North Dakota. Howard Eaton wrote, under date of June 19, 1919, that while he never saw one during his residence in the Little Missouri country in the seventies, a hunter named Henry Bennett told him of poisoning one at the mouth of Cherry Creek, near the Killdeer Mountains. Apparently there are no recent records of occurrence in the State. General habits. — Wolverenes are found mainly within timbered sections of the country, but are great wanderers and at times may strike out over open country in search of new hunting grounds. They are omnivorous hunters and scavengers and have the reputa- tion of being gorging gluttons, a fact which has given them one of their common names. Although valuable as fur animals, they are in bad repute with the trappers from their habit of robbing traps and breaking into caches of food and supplies. They are said to have from two to four young, and like most of the family they have three pairs of mammae arranged close together on the posterior part of the abdomen. Their underfill- is soft and lax, of a gray-brown color, mainly obscured by the long, glossy outer hairs, which in prime skins have a well-spaced and pleasing effect aside from the beautiful and striking pattern of coloration. Prime skins usually bring a high price in the fur market, partly no doubt from their rarity, but mainly from their intrinsic beauty and durable quality. Lutra canadensis canadensis (Schreber) Otter (PL 18, fig. 2) Ptan of the Dakotas (Riggs and Wil- liamson) ; Pehtakd of the Mandans (Will) ; Midapoka of the Hidatsas (Matthews); Citapat of the Ari- karas (Gilmore). Mustela lutra canadensis Schreber, Siiugthiere, pi. 126b [1778]. Type locality. — Eastern Canada. General characters. — Body, long and slender; tail, tapering and muscular; legs, short ; feet, webbed ; ears, small ; fur, dense and glossy. General color, rich dark brown slightly lighter below and with grayish brown on throat and cheeks. Measurements of adult male from Canada: Total length, 1.220 milli- meters; tail, 482; hind foot, not given (Audubon, 1851-1854, vol. 2, p. 4, 1851) ; in inches, 48, 19, respectively; of female, 1.150, 408, and 137 millimeters, respectively. Judging from a medium-sized female from Idaho which weighed 19 pounds, the weight of a large male may be estimated at 25 pounds. 180 NORTH AMERICAN FAUNA [No. 49 Distribution and habitat. — A few otters are still found along all the principal streams in North Dakota and around some of the larger lakes. Although never very abundant, they were evidently much more so in the early trapping days than at present. Owing to their peculiar habits and disposition they hold their own better than many of the more common fur bearers and will undoubtedly remain for generations a part of the North Dakota fauna. In 1801-1808 Alexander Henry (1897, pp. 184, 198, 221, 245, 259, 281, 422, 440) reported 60 otter skins from Keed River, 49 from Park River, 117 from Grand Forks, 24 from the Hair Hills, 322 from Pembina River, 34 from Turtle River, and 12 from Salt River. In 1833 Maximilian (Wied, 1839-1841, Bd. l,pp. 431-432,1839) reported 200 to 300 skins brought in annually at Fort Union, and he fre- quently speaks of the use of otter skins or otter tails for decorations among the Indians. Henry (1897, p. 85) speaks of shooting four otters from the canoes in one day near the mouth of Park River on his way up the Red River, and evidently they were in considerable abundance in the Red River Valley at that time. Lewis and Clark (1893, pp. 175, 272), on October 21, 1804, obtained an otter near the mouth of Heart River, and another was seen and shot at about 30 miles above the mouth of the Little Missouri River on April 14, 1805. A skull from Fort Berthold in the United States National Mu- seum collection was mentioned by Doctors Allen and Coues, but seems to be no longer in the museum. Audubon (1851-1854, vol. 2, p. 11, 1851), on his trip up the Missouri River in 1843, says: "We did not capture any otters during our journey up the Missouri to the Yellow Stone River, but observed traces of them in the small water courses in that direction." In 1913 John Hailand told Morris J. Kernall that there were still a good many otters at Valley City when he settled there in 1878. In 1887 the writer found otter tracks along the northern shore of Devils Lake, and in the Turtle Moun- tains was told that the animals were still fairly common. In 1912 he could get only indefinite reports of their occurrence in the Turtle Mountains, but a more ideal country for them could hardly be imag- ined than this region of numerous lakes and streams well stocked with fish. In 1910 Anthony reported a few otters still caught along the Missouri River near Buford, and in 1912 Eastgate reported them from the Sheyenne River, 3 miles south of Tolna. In 1915 Kellogg reported one seen at the mouth of Antelope Creek near Goodail by Jess VvKlsome two years previously, and at Elbowoods a pair recently seen on a lake at the headwaters of Shell Creek, where they had been common a few years before. General habits. — Otters are largely aquatic in habits, traveling with great ease and speed on or underneath the surface of the water, where much of their food is captured. On land they are slow and awkward except when they " toboggan " over the country on soft snow with considerable speed and evident pleasure. On dry land they are rarely found away from the shores of streams or Jakes, but on deep melting snow they often make long journeys from one stream or lake to another, progressing rapidly in short jumps and long slides on their glossy bellies. They are powerful animals and savage fighters. Few dogs can handle one on land and they will quickly dispose of any dog that they can get into the water. They are 1926] MAMMALS OF NORTH DAKOTA 181 intelligent and, unlike the weasel tribe, have pleasant dispositions and are said to make affectionate and interesting pets. They have few enemies except man, and as more than ordinary trapping skill is required to catch them, they are able to maintain themselves and remain scattered throughout the settled parts of the country in spite of a high price on their beautiful fur. Breeding habits. — The young are usually two to four in number and while small are kept in burrows along the banks. Later they follow the mother on hunting trips until nearly full grown, when they scatter out and each one becomes thereafter a solitary hunter. Food habits. — Apparently the greater part of the food of otters consists of fish, which they pursue and catch in the water. They are rarely found along streams and lakes where fish are not plentiful, but evidently a great part of the fish taken are of the smaller and slower species or the sick or crippled individuals, which fall an easy prey. Crawfish and frogs are also eaten, and it is probable that many waterfowl are captured under favorable conditions. In winter otters travel long distances under the ice, through which they cut holes to the surface when they wish to come out. They are usually in good condition and often covered with a layer of fat like a white blanket under the skin, which serves to protect them from the cold and renders them very difficult to skin for fur. Economic status. — Otter is one of the more valuable and most beautiful of our native furs. It is very durable, especially in the unplucked form, with the glossy overhairs protecting the dense, silky underfill*. Although the price is relatively low for actual value, usually ranging from $10 to $25 for prime skins, it is sufficient to tempt the fur farmer to experiment with raising otters in cap- tivity. As the habits of the animals are becoming better known it is found to be possible to breed them in captivity, and several broods of young have been raised in zoological parks. Further ex- periments and intensive study will be necessary before otters can be recommended for the production of fur in captivity. Mephitis hudsonica Richardson Northern Skunk Ma"ka of the Dakotas (William- son) : Su"kte of the Mandans (Will); Choka of the Hidatsas (Matthews) ; Ho Jiff a (Gilrnore) ; Nichicit of the Arikaras (Gilrnore). Mephitis americana v:ir. hadsonlca Richardson, Fauna Boreali-Americana, pt. 1, p. 55, 1829. Type locality. — Plains of the Saskatchewan. General characters. — Low, heavy-bodied, bushy-tailed animals with planti- grade feet, naked soles, and long digging claws. The most striking pecu- liarity consists of the pair of anal scent glands, which secrete a yellow fluid with a powerful odor. This northernmost and largest form of the genus Mephitis has a very long and busby tail; the color is glossy black with a white stripe between the eyes and a white triangle on the back of the neck connecting across the shoulders with two broad white stripes along the sides of the back and tail ; upper base, lower surface, and tip of tail usually black or washed with black over the surface. The relative amount of black and white varies greatly in different individuals. An adult male from Cannon Ball measures in total length, 710 millimeters ; tail, 300 ; hind foot, 80 ; a female from Towner measures 780, 273, and 91, respectively. Weight of a large adult from the Yellowstone Park, 8% pouiuls. (Seton, 1909, vol. 2. p. 968.) 182 NORTH AMERICAN FAUNA [No. 49 Distribution and hatitat. — The large northern skunks range over most of the northern Plains country and extend south in the mountains to New Mexico. They are found over practically all of North Dakota, ranging alike over the prairie and into the open forest, but are most abundant along the brushy borders of streams and lakes and in the thickets of the gulches. They are much trapped and their abundance varies constantly, but they quickly increase where trapping is relaxed for a short time. So unsuspicious and easily caught are they that by persistent effort any amateur can get most of them in his vicinity. None were reported by the fur trappers of the early pioneer days, evidently because skunk fur was not then considered marketable. Hence no companion can be made of their past or present abundance. In 1878, when John Hailand came to Valley City, skunks were numerous. In 1887, the writer found them common at Pembina, Devils Lake, in the Turtle Mountains, and at Fort Buford. In 1909 D. D. Streeter reported them at Medora, and in 1910 Anthony reported a few at Fort Buford. In 1912 Eastgate reported them at Stump Lake, Valley City, Lisbon, Kathryn, and Bottineau. One trapper near Bottineau had 77 skins and another at Lisbon claimed to have taken 178 during the previous winter. The same year the writer found skunks fairly common in the country about Fargo, Stump Lake, Valley City, and in the Turtle Mountains, where they were said to be one of the principal fur-bearing animals caught in both fall and spring. In 1913 he was told there were a few about Kenmare and along the Mouse River farther west. At Fort Clark, Jewett reported them rather scarce, although a few tracks were seen on the river bottoms near there. In the Badlands, 25 miles south of Medora, a few tracks were found. In 1915 Sheldon reported them as common about Fairmount, Oakes, Dawson, and Cannon Ball across the southern part of the State, and Kellogg reported them common at Wahpeton, Larimore, Manvel, Grafton, Drayton, south of Devils Lake, Towner, Grin- nell, and along the Missouri River at Lostwood, Elbowoods, Good- all, Stanton, and Sather. At many of these localities they were re- garded as the principal fur-bearing animals taken by the trappers, and at Drayton, in Pembina County, Kellogg says that with the mink and muskrat they form the means of support for a large number of persons during the winter months. General habits. — Owing to their confidence in their peculiar de- fensive powers, skunks appear fearless and independent. As a mat- ter of fact they have no other recourse, as their short legs bar escape and their rather weak bodies are unfitted for combat. When met in the path or in the bushes or grass they usually stand their ground, stamp their feet, and with erect and bristling tail make themselves as conspicuous as possible, on the assumption that they will be given plenty of room. Generally their right of way is undisputed, but if closely pressed they about face or throw the body forward and with a quick contraction of the muscular bands surrounding the scent gland force the amber-colored fluid through one or both of the nipplelike ducts to a distance of 10 to 15 feet. The spray is often so fine as to be unnoticed, except by the powerful odor, which at once fills the air and almost stifles one in close proximity. So far as possible, the animals avoid getting the fluid on themselves, and 1926] MAMMALS OF NORTH DAKOTA 183 when undisturbed they usually have little, if any, trace of the odor. Even when caught in traps they rarely discharge their scent unless approached or clumsily handled when killed. If shot so as to break the spinal column or if struck a sharp blow with a club just back of the shoulders, the posterior muscles are paralyzed so that the scent will not be discharged and the animals may be skinned with no un- pleasantness. Skunks generally make their homes in burrows, which they dig in banks or brush patches, or even occasionally in the open. They are mainly nocturnal, but usually leave their dens early in the evening and are often seen abroad before dark and after daylight. They are great hunters, and notwithstanding their short legs often travel long distances in search of food. In fall they become very fat, and usually with the first snows enter the burrows that have been pre- pared for winter use and curl up for the winter's sleep of four or five months. Generally they are out before the last snows are gone in March, and often their tracks are found in the soft, wet snow in spring. Much of the winter's fat is carried over and is needed for the spring mating season or until the supply of insects and other summer food becomes available. Breeding habits. — A female taken by Sheldon on May 11, 1915, at Fairmount, contained 4 well-developed embryos showing per- fectly the characteristic white stripes. This was evidently the first litter of a young breeding animal, as usually the number of young is 6 to 10. Nine skunks which Kellogg reported dug out of one burrow, near Wahpeton, in the winter of 1915, probably represented a family that had not been broken up — the mother skunk and 8 young. The mammae in breeding females are usually arranged in 7 pairs, 2 pairs of inguinal, three of abdominal, and 2 of pectoral, or in 2 long rows of 7 each, rather evenly spaced along each side of the ventral region. Food habits. — Apparently the largest part of the food of skunks consists of grasshoppers, beetles, crickets, and other insects and insect larvae, which they catch in the grass or dig up from underground or under decayed vegetation or rotten logs. Their stomachs are large and are usually found well filled with material that is easily identifiable. They eat some mice, especially tlje young of mice and other rodents which they dig out of the nests. To what extent they feed on young and old ground squirrels in North Dakota has not been well determined, but undoubtedly they get some of these among other rodents. They are fond of eggs and the stomach of one taken \>y Sheldon at Fairmount on May 11 contained egg-shells of prairie chickens, as well as remains of five young meadow mice. Eastgate says they are destructive to young chickens during early summer, and to prairie chickens, sharp-tailed grouse, and wild ducks during the breeding season. Along the borders of the Sweetwater Lakes in June, 1916, skunks were found unusually numerous and in the evenings they were often met galloping along the trails at the edge of the lake or climbing about through the reeds and tules where the ducks were nesting, evidently searching for nests containing eggs or young birds. The fact that many of the old ducks had small broods of young and others none may have been attributable to these nest robbers, although minks and other animals may have been in part 184 NORTH AMERICAN FAUNA [No. 49 responsible. The regular breeding grounds of ducks and other water birds should be protected from skunks and such animals by very- thorough trapping during the fur season and, if necessary, during the breeding season of the birds. Economic status. — Skunks are one of the most valuable, because one of the most abundant, of the fur-bearing animals in North Dakota. There are no statistics as to the number taken in the State, but the total number of skunk skins sold in London in 1911 was more than 2,000,000. The average price at that time was $2 each. In 1921, 824,599 skunk skins were dressed by one association, which handles about 90 per cent of the fur dressed in America. The recent average price of raw skins in New York City was $3. This gives only an incomplete record of the skunk fur crop, of which North Dakota furnishes her full share. In many sections, however, the fur of skunks is not their greatest value, as their insect and rodent-destroying habits render them extremely useful adjuncts to agriculture. An overabundance of skunks would not be advantageous in most localities, and in certain areas their numbers should be restricted as much as possible. Their abundance should be Avell controlled by suitable trapping- laws with provision for local modifications. Many of the States have a close season protecting skunks, except when the fur is prime, during late fall, winter, and early spring. There is little danger of their extermination even locally, but their numbers could often be so controlled as greatly to increase their value for fur and other purposes. For fur farming, skunks have been thoroughly tested and in many cases successfully raised in confinement, but the low value of their fur prevents any large returns from the industry. Farmers' Bulletin 587, of the United States Department of Agriculture (Lantz, 1914), gives much practical information on the breeding of skunks for fur. Taxidea taxus taxus (Schreber) Badger (PI. 15, fig. 2) Choka of the Dakotas (Riggs and Williamson) ; Mate- of the Man- dans (Will) ; Amaka of the Hi- datsas (Matthews), Aicaga (Gil- more) ; Sunuh-katuh (flat porcu- pine) of the Arikaras (Gilmore). Ursus taxus Schreber, Saugthiere, Theil. 3, p. 520, 1778. Type locality. — " Labrador and Hudson Bay " (probably really from Mani- toba or Saskatchewan). General characters. — A heavy-bodied, low, wide, powerfully built animal of the weasel family, with short, muscular neck, short ears, short legs, and short tail ; fur, long and light, especially on the sides, which heightens the effect of the wide body. Color of upper parts buffy or brownish gray, top of head and nose blackish with white stripe from nose to back of neck, white markings on cheeks connected with white or creamy throat ; under- pays, plain buff or soiled whitish ; feet and legs, black. Measurements of adult male from Oakes : Total length, 788 millimeters ; tail 133 ; hind foot 120. Of female from Lidgerwood, 730, 150, and 114, respectively. Weight of male from Wisconsin, 23 pounds 6 ounces. (Jackson, 1908, p. 2S.) 1926] MAMMALS OP NORTH DAKOTA 185 Distribution and habitat. — Badgers range over most of the west- ern United States and from southern Mexico to Saskatchewan, and several well-marked forms are recognized. Those ranging over the whole of North Dakota may undoubtedly be referred to the original species. Apparently there is no considerable area in North Dakota where they are not occasionally found. Although most abundant over the prairies, they penetrate into open forested country and even in the Turtle Mountains are found occasionally throughout the more or less scattered timber. Over the prairie country their great- est abundance usually coincides with the abundance of ground squir- rels, which form their principal prey. In the more thickly settled parts of the State they are disappearing, as they are practically de- fenseless and easily destro}red by man unless they can escape into convenient underground burrows. As the time will doubtless come when these useful animals will be very scarce, it seems worth while to give detailed record of their present distribution. In 1887 the writer found them common at Harwood, Grand Forks, Pembina, Devils Lake, Bottineau, Rugby Junction, and Fort Bu- ford. In 1892 Loring reported them common about Sherbrooke and Jamestown. In 1909, they were found fairly common both in and around the Turtle Mountains. In 1910 Anthony reported a few burrows around Fort Buford, but the badgers were more abundant on the other side of the Missouri River. In 1912 they were more or less common at Hankinson, Fargo, Valley City, and Stump Lake, and a few in the Turtle Mountains; and Eastgate reported them about Stump Lake, Kathiyn, and Lisbon. In 1913 the writer found them common at Kenmare, Crosby, and on the Dakota National Forest, south of Medora. Jewett reported a few along the Little Missouri River from Medora to Quinion and many of their bur- rows in the country about Sentinel Butte. In 1915 Sheldon found them fairly common across the southern part of the State at Fair- mount, Lidgerwood, and Oakes. In the same year it was said that there were still a few near Wahpeton, and Kellogg saw the remains of one in the road near Larimore. At Manvel, in Grand Forks County, he reported a number of burrows found in almost any field where the badgers had been digging out ground squirrels. Near Grafton he reported one killed on the Munson farm. At Drayton, in Pembina County, he found where one had been working on the farm where he staid, but it had recently disappeared. Along the south side of Devils Lake he found a number of places where the badgers had been working, but saw none of the animals. At Towner he collected a specimen and reported the animals quite numerous and doing some damage to the roads as well as killing a great many ground squirrels. In one place he counted 18 burrows within a radius of 20 feet. At Grinnell, on the Missouri River, he reported two badgers seen and at Lostwood he considered them fairly numerous, judging by the number of burrows. At Elbowoods, farther down the river, he was told that they were plentiful, and at Goodall quite a few were found. At Stanton he reported them as fairly plentiful over the prairie and one occasionally found on the river bottoms. From Washburn to Bismarck he was told that they were occasionally found. At the Sweetwater Lakes in 1910, Mrs. Bailey saw three alive and one that had been killed. 186 NORTH AMERICAN FAUNA [No. 49 General habits. — Badgers are preeminently burrowing animals, and they depend on their claws not only for unearthing a large part of their game, but also for the construction of both their summer and winter homes. They seem to prefer open country, where they can see to considerable distances and either escape the approaching enemy by retreating to some near-by burrow or, if necessary, by quickly digging a hole in the ground deep enough to protect all but their vicious jaws, which few animals care to approach. Within a few minutes they will sink their burrows until they are out of sight and then pack the earth behind them as they continue to tunnel through the ground to greater depths. A person on foot can easily overtake one as it lopes away on its short legs, but if unarmed or without even a stick or stone, the tables are quickly turned, and he has to run his best to escape having his legs severely bitten. With a camera one can usually obtain good pictures by chasing a badger until it turns and then backing about over the prairie as it comes on in animated pursuit. Occasionally one is seen lying in the sun on the mound in front of the burrow from which it has unearthed a ground squirrel, or loping across the road in its short, floppy gait. Dogs usually pursue, but keep well out of reach of the savage jaws of the badger, and there are very few dogs that do not get the worst of an encounter with one of these strong- jawed, thick-hided animals. In summer the badgers spend most of their time and energy in dig- ging out the various rodents on which they feed, and even after the ground squirrels have denned up for winter continue to unearth and feed upon them for a month or six weeks, until the ground begins to freeze, when they seek their own winter quarters and, well en- sconced in deep burrows, curl up for a long sleep. At this time of year they are always fat and covered with a heavy coat of long fur. From the middle of October to the middle of March they are rarely seen above ground, but with the melting of snow they appear and, still fat and with a still heavier coat of fur, start out on their hunt- ing and mating expeditions. Breeding habits. — Surprisingly little is known of the breeding habits of badgers. The young are apparently brought forth and kept within the burrows until well grown, as few persons have seen a badger outside less than half grown. The mammae are usually in 4 pairs, 2 pairs of inguinal and 2 of abdominal, and the young are usually four in number. Near Stump Lake, on July 23, 1912, a family of four not fully grown young were found in the prairie grass. They were followed to the nearest burrow, where the ]ast of the four was struggling to get inside. It was caught by the tail and then by both hind legs and given a wide swing over the prairie grass to a considerable distance from the burrow. With the camera the writer followed around, teasing and keeping it engaged while taking as many photographs as were wished before it was allowed to return to the burrow where its brothers and sisters had disappeared. Food habits. — In North Dakota the favorite food of the badger seems to be the Richardson ground squirrel, and where the squirrels are most abundant so also are the badgers. Other ground squirrels are dug out and eaten wherever found, as also are pocket gophers, prairie dogs, mice, and other burrowing rodents. Occasionally 19-'6] MAMMALS OF NORTH DAKOTA 187 badgers will feed on some old carcass, and usually they will take any kind of meat with which traps are baited. On rare occasions one will dig under a chicken coop and kill some of the poultry, but this hap- pens so rarely and is so easily prevented as to be of little economic importance. Economic status. — In North Dakota, as in other parts of the country, badgers are generally killed on sight by the residents on the pretext that they catch poultry, kill lambs, and are a danger to horses, which sometimes step in their burrows and, if running, possibly break their legs or injure their riders; or that they make burrows in roads, causing serious bumps to passing automobiles. All of these claims have some foundation in fact, but they are generally over-emphasized to warrant the wanton destruction of a conspicuous and rather ferocious little carnivore that is not swift or skilful enough to protect itself. Or Mie other hand, the badger spends almost all of its time digging out and devouring the most injurious rodent pests of the region, thus saving a large quantity of grain and other crops from destruction. It is unquestionably one of the least harmful and most completely beneficial of the native mammals in the State, and even when the ground squirrels are poisoned and under good control there will still be ample employment for the badgers in digging out pocket gophers, mice, and other small rodents which must be held in check to prevent serious loss to crops. Only recentty have badger skins come into general use as fur. While very durable, warm, and when in prime condition, rather at- tractive, they are certainly worth more to the State when worn by the badger than when made into robes, coats, or muffs. In some States the value of the badger is recognized and the animal is protected by law, but a protection through popular sentiment based on a full knowledge of its useful habits would be much more effec- tive than a legal statute not well enforced. Family PIIOCYOX1DAE : Raccoons Procyon lotor lotor (Linnaeus) Raccoon; " Coon " Wica of the Dakotas (William- son) ; Mikd of the Omahas (Gil- more) ; Isat of the Arikaras (Gil- more) ; Shunte-pusa of the Mandans (Gilmore) ; Sida-buzhe of the Hid- atsas (Gilmore) ; Asebun of the Ojibways ( Wilson ) . [Ur8us] lotor Linnaeus, Syst. Nat., ed. 10, t. 1, p. 48, 1758. Type locaMtjf. — Eastern United States. General character*. — A thick-set, furry little animal with pointed nose, prominent ears, round, furry tail, long naked soles, and strong curved claws. Color, yellowish or silvery gray, with light gray ears, face, and feet ; gray more or loss darkened with black-tipped hairs over the back; a black mask across face, black spots back of ears, and five black rings around tail; long woolly underfur light brown. Measurements of a large male taken near 82242°— 26 13 188 NOETH AMERICAN FAUNA [No. 49 Fargo, by Murie : Total length, 880 millimeters ; tail, 265 ; hind foot, 125. Weight, 24 pounds. A large and very fat male at Elk River, Minn., weighed 30^ pounds. Distribution and habitat. — In the early trapping days raccoons were abundant in the Red River Valley and apparently scarce in the western part of the State. On September 16, 1800, Alexander Henry (1897, pp. 88, 90, 112, 122,' 136, 155, 171) on a canoe trip along Red River, tells us in his journal that his " people saw many raccoons in the course of the day, and shot four." On October 5, at the mouth of the Park River, his party caught 5, and on October 6, 3, and on October 18, 20. After that he records some brought in daily by the trappers ; many of them were very fat, and when stripped of the fat and roasted made excellent eating. On November 7, he wrote: " My men took great numbers of fat raccoons in their traps ;" and on November 18, no more taken, as they had all denned up for the winter. At Park River, on November 30, he reports seven raccoons taken from one hollow tree where they were evidently hibernating. On March 5, 1801, he says : " The snow is gone and raccoons begin to come out in the daytime." During the trapping seasons from 1800 to 1809, he (Henry, 1897, pp. 184, 198, 221, 245, 259, 281, 440) re- ported among others, 37 raccoon skins taken at Red River, 163 at Park River, 144 at Grand Forks, 57 in the Hair Hills, 158 on the Pembina River, 63 on the Turtle River, and 15 on the Salt River. Apparently they were one of the commonest fur-bearing animals of that region. In 1887 the writer found them common near Fargo and at Devils Lake, and in 1895 Loring reported them common at Portland, where he saw skins of some that had been taken in that vicinity. In 1912, in the Turtle Mountains, a resident trapper said that he knew of only three instances of raccoons having been taken in the hills and he considered them decidedly rare. The same year Eastgate re- ported three killed just north of Dion Lake on November 27. East- gate also reported a few on the Sheyenne River, south of Stump Lake, and was told that they were formerly common at Lisbon, but that of late years they had become very rare, only three or four skins being brought in each winter. In 1913, a few coons were reported along the Mouse River, east of Kenmare; and at Minot, Mr. Booth, the taxidermist, said that there were still a few at that time, but that they used to be very common when he first came there in the early eighties. On the Missouri River no mention is made of raccoons by Lewis and Clark, Maximilian, or Audubon, while Hayden (1875, p. 92) in his report on the upper Missouri region for 1855, 1856, and 1857, reports them abundant at Council Bluffs; but the highest point on the Missouri River at which he observed them was about the mouth of the Niobrara River. The fact that Maxi- milian found a name for them among the Minnetaree Indians on the Upper Missouri would indicate that they were not entirely ab- sent from the region at that time. At Fort Clark in 1913 Jewett reported fresh tracks in muddy places along the Missouri River. In 1914 the writer found tracks along Apple Creek, just east of Bismarck, where the animals had been feeding on crawfish and mussels along the creek. 1926] MAMMALS OF NORTH DAKOTA 189 In 1915, raccoons were found common along the river at Wahpeton, where many of the old hollow trees were well scratched up by their claws, and where there were great numbers of frogs in the marshes and an abundance of acorns, all of which offered a feast for them. In the same year Kellogg reported them at Grafton, where he saw three very large dark skins in the Williams collection. At Drayton, in Pembina County, he reported one occasionally captured, and around Devils Lake he found them common in the woods. At Grinnell, on the Missouri River, he reported a few caught, and at Goodall, he saw tracks along the river and was told that two had been taken by trappers two years before. At Elbowoods he saw a few tracks along the river, and near Sather he followed their tracks from the river to a cornfield, where they had been eating the corn and had destroyed entirely two rows. On his way from Washburn down the river to Bismarck he reported a few in the wooded sections of the river bottoms and was told of two trappers who had sold 75 skins taken around Chanta Peta Creek, south of Bismarck. In 1919, Mr. Allen, a taxidermist at Mandan, said that there had always been a few raccoons along the rivers there and one was brought in to be mounted the year before. No records were obtained from the Little Missouri country and the areas west of the immediate Missouri valley. General habits. — In general habits, as well as to a slight extent in appearance, the raccoons resemble the bears. They are very in- telligent and resourceful animals, adapting themselves to almost any environment where food is abundant and the climate not too severe. They are excellent climbers and usually make their homes in hollow trees or logs, but in the absence of such protection they often occupy caves and hollow spaces in banks or cliffs, where they find dark retreats for the daytime and safe dens for their long winter sleep. They are mostly nocturnal in habits, but on rare occasions will come out in the daytime when disturbed or move from one place to another in search of mates in the breeding seasons. Although not very swift runners, they can usually outdistance a man, but are quicklv overtaken by dogs, which are often used in hunting them at night. When pursued, if no hollow tree or rocky retreats are within range, they usually seek protection by climbing up the nearest tree. They are savage fighters and will generally get the best of a dog of approximately their own size. In fall they become very fat, and soon after the first snows fall enter their dens for hibernation and remain until the early spring thaws rouse them to renewed activity. Breeding habits.— Audubon (1851-1854, vol. 2, p. 77, 1851) gives a number of young of the raccoon as four to six, generally brought forth in May. The mammae are usually arranged in two pairs of abdominal and two pairs of pectoral. The young are kept well secreted in hollow trees or caverns and are rarely seen until about half grown, when they begin to follow their mother in search of food. In fall they are still in family parties and if so fortunate as to escape the dogs and trappers until November, they enter hiberna- tion together, the mother evidently selecting a suitable hollow for 190 NORTH AMERICAN FAUNA [No. 49 their winter's sleep. In spring the males are out with the first few warm days, making long journeys from tree to tree in search of mates. Occasionally in early spring a male and female are found together in a hollow tree or hollow log, but for the rest of the year the animals are mainly solitary, except for the mother and young. Food habits. — In tastes raccoons are highly omnivorous, accepting almost any food in the way of fish, flesh, or fowl, fruit, nuts, or corn. In this northern country they feed in summer very largely on craw- fish, mussels, frogs, and fish, and on such birds, eggs, or small mam- mals as they can find or catch. They are especially fond of ripe blueberries, serviceberries, and any kind of sweet fruit. In fall they usually fatten on acorns where these are obtainable, and the northern limits of their range are almost coincident with the north- ern limits of oaks. Often at this season their large stomachs con- tain nothing but the finely masticated pulp of acorn meats and a few shells, and the fattening properties of these rich nuts seem not to be lessened by their bitter and astringent flavor. They are also very fond of unripe corn, and at night will make long trips to cornfields, where they pull down the ears and strip them of their milky kernels. Once started on the green corn, they usually continue to feast on it from the early milky stages until it has become fully ripe. Economic status. — On rare occasions raccoons find their way into poultry yards or houses at night and kill some of the fowls or rob the nests of eggs. It is probable also that they destroy eggs and young of game and other birds occasionally, but there is little mis- chief that can be proved against them. Their raids on cornfields are often of considerable extent, but usualty the animals are discovered and captured by a night hunt with dogs before they have done serious damage. On the other hand, their value as one of the standard fur-bearing animals is usually sufficient far to outweigh the losses from their occasional depredations. Their fur is thick, warm, and light, and the skins, while light, are very strong and durable and specially well adapted for overcoats, robes, caps, and driving gloves. The fur is also used for women's capes and muffs, and when plucked is a fair imitation of plucked beaver fur, although longer and less dense. It is usually one of the rather low-priced furs, but gives good value in warmth and wear. Many people are very fond of the flesh of raccoons, and when fattened on acorns or beechnuts in the north the meat is of good flavor and wholesome. The fat makes a thin oil that is much prized for use on leather. In the pioneer days it was the principal oil for domestic purposes and even for machinery in the frontier settle- ments. Generally the raccoon is considered a valuable fur and game animal and its depredations are easily overlooked. Many of the skins are of little value because taken early in the fall before they have become heavily furred and prime. The young do not get their full growth before entering hibernation in fall, and only a very short season should be allowed for trapping before they den up for the winter. In North Dakota an open season from November 15 to March 15 would probably insure prime fur. 1926] MAMMALS OF NORTH DAKOTA 191 Family URSIDAE: Bears Ursus americanus amcricanus Pallas Black Bear; Cinnamon Bear Wasabe of the Omahas (Gilmore) ; Wachank-shica of the Dakotas (Williamson) ; Konuch-katit of the Arikara (Gilmore), Watu — tame bear; Haschida of the Hidatsas (Maximilian) ; Ischidda of the Mandans ( Maximilian ) . Ursus americanus Pallas, Spicilegia Zool., fasc. 14, p. 5, 1780. Type locality. — Eastern North America. General characters. — A heavily built, powerfully muscled animal, not half so clumsy as it looks when fat and in long fur. Eyes, small and not very keen sighted, but the ears prominent and as sensitive to sound as the nose is keen to scent. Tail, short; feet, rather large and plantigrade with naked soles; front and hind claws, short, curved, and sharp for climbing. Color, mainly black, with usually a yellow-brown muzzle and occasionally a white spot on the breast or throat. Occasionally these bears are entirely brown, normally of a cinnamon color, but varying from yellow-brown to dark brown. Measurements of adult male, from Montana : Total length 1.6S0 millimeters ; tail, 105 ; hind foot, 275 ; in feet and inches, 5.5 feet, 4.1 inches, 10.8 inches, respectively. In Minnesota, where probably the same form occurs, the weight is estimated usually at about 300 pounds for a fully adult male. One killed by A. H. Wilcox ( 1907. p. 100 ) , at Detroit, Minn., weighed 299 pounds. Seton (1909, vol. 2, p. 1052) gives the weight of a large male killed near Winnipeg, Manitoba, as 205 pounds. Distribution and habitat. — In the early days black bears evidently ranged over practically all of Xorth Dakota, but were most abundant along the Red River Valley, in the Turtle Mountains, Pembina Hills, and on the wooded streams of the eastern part of the State. There are a few records of them for the Missouri Valley, but apparently they were never common over the open prairie country or in the Bad- lands farther west. Their greatest abundance seems to have been in the Red River Valley, where from 1800 to 1808 Alexander Henry (1897, pp. 184-440) records them in such numbers as have rarely been known in any part of the country. In September, 1800, near the mouth of Park River, he reported 4 bears killed on the 14th, 6 on the loth, 1 on the lGth, 2 on the 20th, 3 on the 24th, 1 on the 25th, and 1 on the 2Gth of the month, and 40 skins taken by one party of trappers. For the next eight years he reported, among the fur bearers taken in the Red River Valley, 52 black and 20 brown bears on the Reed River, 148 black and 25 brown on the Park River, 64 black and 3 brown at Grand Forks, 131 black and 2G brown from the Hair Hills, 302 black and 75 brown on the Pembina River, 28 black and 12 brown on the Turtle River, and 18 black and 2 brown on the Salt River, making in all 900 bears from this region. At the mouth of the Park River, on September 22, 1800, Henry (1897, p. 101-102) says in his journal : Bears make prodigious ravages in the brush and willows; the plum trees are torn to pieces, and < very tree that bears fruit has shared the same fate; the tops of the oaks are also very roughly handled, broken, and torn down, to get the acorns. The havoc they commit is astonishing; their dung lies about in the woods as plentiful as that of the buffalo in the meadow. Over the rest of the State the records are few and scattered. Along the Missouri River bears were not mentioned by most of the 192 NORTH AMERICAN FAUNA [No. 49 early explorers, although Audubon (1897, p. 133), in 1843, reported the killing of a black bear on the White Earth River, about 60 miles from its mouth, where he says a few are occasionally shot. In 1878, McChesney (1878, p. 202), from Fort Sisseton, S. Dak., just below the southeastern corner of North Dakota, reported them as once very common, but not seen of late years within 50 or 60 miles of the post. At Valley City, J. S. Weiser gave Morris J. Kern all a record of a black bear found near there in 1878. In 1887, there were said to be still a few black bears in the Devils Lake timbered areas, and in the Turtle Mountains they were said to be common. A few were also said to occur in the country about Fort Buford, but not on authority that seems very reliable. At Wade, on the Cannonball River in 1913, W. B. Bell reported black bears seen by Mr. Wade, who had lived there for 41 years. At the mouth of the Cannon- ball, in 1916, the writer was told that a few black bears had been found along the river bottoms up to comparatively recent times. Beede, who had lived among the Sioux Indians there, said that the Indians did not hunt them unless in dire need of food, as the bears were to them semisacred. When one was killed, he saj^s, a bit of its skin was left on a bush or tree as an offering to the spirit world. At the present time there are a few black bears in the Turtle Moimtains and Pembina Hills, where one is occasionally killed, and in the Red River Valley one may sometimes wander in from the heavy woods of northern Minnesota. At Grafton, Kellogg reported two killed by Andrew Monsebroten, five miles Avest of the town, in 1884; two young bears killed by Arthur Blomquist, in 1910, about six miles north of Drayton on the Minnesota side of the river ; and another killed in Pembina County, in 1894, by Jim Spanglo, " the latest record I could get for the county." At Devils Lake, in 1916, Mr. Palmer told of a small brown bear killed near the lake only two years previously. It had been seen in several localities and evi- dently had wandered from the Turtle Mountains. General habits. — The black bears are timber-loving species, de- pending largely on the cover of thickets, swamps, and dense forests for protection, and to some extent also upon the trees for food and winter quarters. They are great wanderers, however, and do not hesitate to strike across wide stretches of open country when in search of a new supply of food, or when driven out of their regular haunts by hunters. Usually, however, their wanderings are along the lines of streams and wooded or brushy patches, where both food and cover are to be found. Food habits. — Few animals are more nearly omnivorous than the black bears, and as they are hearty eaters a great quantity of food is required to satisfy them, especially in fall, when they are preparing for their long 'winter sleep. Acorns, berries, and fruit form a great part of their food in this northern country, but they readily accept any meat or carcass that can be found, such insects as can be procured from ant hills, rotten logs, or overturned stones, and many plants and roots and much succulent vegetation. In summer they gorge themselves on blueberries and serviceberries, the abundant sweet fruit of which seems to appeal strongly to their appetites. In fall, wherever oaks are to be found, bears search for the acorns and gather them, first from the treetops by draw- 1926] MAMMALS OF NORTH DAKOTA 193 ing in and breaking the branches until every acorn can be reached, often making the top of the tree look like an eagle's nest before they have finished with it. Later, as the ripe acorns fall to the ground, the}*- gather them up and eat them to the exclusion of almost every other food. These puckery but rich little nuts rapidly supply the heavy coating of fat necessary for carrying the bears comfort- ably through the long, cold winter hibernation. Hibernation. — In 1800, Alexander Henry (1897, pp. 157, 252-253, 87, 135, 136, 117) wrote in his journals that " bears den in hollow trees along the Red River, but in the Hair Hills on higher ground in holes in the banks. They are hunted by the Indians in the trees." On November 13, 1804, at Pembina, he writes, " My tame bear is making a hole to take up his winter quarters in." On September 6, 1800, he says, " one bear killed up a tree." In another place he records one bear killed November 5 and 10 skins brought in from the Hair Hills, November 6; and on May 1 following, he records 37 bear skins brought in from Grand Forts (p. 177). These dates, however, do not indicate reliable records of the beginning and end of hibernation, as apparently the Indians were in the habit of killing the bears in their winter dens. Usually the bears in northern Minnesota den up with the first heavy fall of snow and cold weather early in November, reappearing with the first warm days late in March or early in April. Their fur is not prime until about the time of hibernation and usually is in the best condition when they come out of their dens in spring. A large number of the skins taken are in almost worthless condition because the bears were killed too early in fall or too late in spring. Economic status. — Over much of the country black bears are now considered one of the valuable game animals and given pro- tection as such in the game laws. In a forested area like the Turtle Mountains it would seem well worth while to protect them until past danger of extinction. With the abundance of wild land, for- est, and lakes, and ample food in the berries and acorns, there is little probability of their doing any serious mischief to crops or livestock in that region. Ursus horribilis horribilis Ord Grizzly Bear; Big- Plains Grizzly; Silvertip Mato or Mato-chota of the Dakotas (Gilmore) ; Mato of the Mandans (Maximilian) ; Mato unknapininde of the Mandans (Will) ; Laeh- pitzi of the Ilidatsas ( Maximilian ) ; Kthiuch of the Arikaras (Maximil- ian), Konuch-tarauHs (Gilmore). UrsuH horrioilus Ord, Guthrie's Geogr., 2d Amer. ed., vol. 2, pp. 291, 300, 1815. (Reprint by S. N. Rhoads, 1894.) Type locality. — Missouri River above the mouth of Poplar River, north- eastern Montana. General cliaracterK. — Size, very large ; skull, long and massive with very heavy molar and canine teeth ; front claws, long and only moderately curved. Fur, long and loose with well-marked mane or crest over "hump" of shoulder. Color, variable from light yellow to dark brown, the lightest individuals even called white by Lewis and Clark and other writers familiar with them. Audubon (1851-1854, vol. 3, p. 149, 1854) says: "We have skins in our pos- 194 NOETH AMERICAN FAUNA [No. 49 session collected on the Upper Missouri, some of which are nearly white, while others are nearly of a rufous tint, and one that was killed by our party, of which we also have the skin, was a dark brown one."24 Maximilian (Wied, 1S39-1841, Bd. 1, pp. 490, 493, 1839) writes: An old bear and two cubs were killed. The mother was " a pale yellowish color ; one of the cubs, which was brought on board alive, was whitish about the head and neck and brownish gray on body; the other was dark brown." Another killed on July 18 was reported as dark brown, with new hair of light gray with yellow tips already appearing; another killed farther up the river on July 21, 1833, was at first supposed to be a black bear, but when shot proved to be dark brown, and as Maximilian suggests, may have been another species of grizzly. Measurements of a small male collected by Maximilian (Wied., 1839-1841, Bd. 1, p. 488, 1839) and supposed to be about 3 years old were: Tip of nose to tip of tail, 6 feet 2 inches 2 lines. A large one measured by Lewis and Clark (1893, p. 298) in northeastern Montana, apparently the type of the species, measured from tip of nose to extremity of hind foot, 8 feet 7% inches, length of front claws, 4% inches. A still larger one killed by the party was said to measure 9 feet from tip of nose to tip of tail, with front claws 6% inches in length. Skull of large male : Basal length, 351 millimeters ; zygomatic breadth, 247; in inches 13.8 and 9.6, respectively. (Merriam, 1918, p. 19.) There seem to be no reliable weights for the adults of this Plains grizzly available, but Lewis and Clark (1893, p. 29S) estimated the weight of a large one as 500 or 600 pounds.25 Distribution, habitat, and general habits. — At the coming of the white man these large grizzlies were apparently common over prac- tically all of North Dakota. In 1800 Alexander Henry (1897, pp. 121, 145, 184, 245, 259, 281, 440), while in the Eed River Valley, wrote in his journal : Grizzly bears are not numerous along Red River, but more abundant in the Hair Hills. At Lac du Diable [Devils Lake], which is about 30 leagues W., they are very common — I am told as common as the black bear [Ursus americanus] is here, and very malicious. Near that lake runs a principal branch of Schian [Sheyenne River], which is partially wooded. On the banks of this river I am informed they are also very numerous, and seldom molested by the hunters, it being the frontier of the Sioux, where none can hunt in safety ; so there they breed and multiply in security. Again, in speaking of the Sheyenne River, he says : " Grizzly bears are to be seen in droves." On his return from a trip to the Shej^enne River to his winter quarters at the mouth of the Park River, near where Grafton now stands, he records : During my absence the hunter had killed a large grizzly bear [Ursus Jior- riMlis] about a mile from the fort. He had seen two males and a female, but the latter escaped. My people having cooked and eaten some of the flesh were taken very ill, and most of them threw it up. This bear had been wounded in the fore leg some time before by an arrow, the iron head of which stuck fast in the bone, and was beginning to rust. During the first trapping season 1800-1801, his men obtained four grizzly-bear skins at the Reed River and two at the Park River. In 1S04 he reports one grizzly bear from the Park River and in 1805, four from the Hair Hills, four from the Salt River, and two from the Pembina River, and in 1806, three from the Pembina River. Meanwhile of the 113 skins of brown bears recorded, it is very prob- 24 Possibly two species. 26 The type specimen of Ursus horribilis collected by Lewis and Clark is lost, but a fine old male skull from near the type locality in eastern Montana, gives reliable char- acters of the species. There is still one skull of a 3-year-old male from Port Buford in the National Museum collection, taken by J. P. Kimball, in 1868. This is the only specimen representing the species from North Dakota and we can only assume that the grizzlies extending across the State were all the same. Old skulls from any part of the State would be of great interest and value as contributions to the National Museum collection. 1926] MAMMALS 01" NO&TH DAKOTA 195 able that some were of the grizzly group. Henry (1897, pp. 422, 221) also reported one grizzly-bear skin in the catch from the Sandhill River, Minn., in 1807, and one from Portage la Prairie, Manitoba. On their way up the Missouri River in 1804, Lewis and Clark (1893, pp. 157, 174,251,274,288-289,298) often referred to the white, yellow, and gra}^ bears. On Fox Island, S. Dak., they saw the tracks of a " large white [grizzly] bear." On October 20, while camped an the river bottom just below where Bismarck now stands, they say : " ^Ve also wounded a white bear, and saw some fresh tracks of those animals, which are twice as large as the track of a man." From Mandan they sent back, among other skins, those of the " yellow bear." The following spring, on the way up the river after wintering at the Mandan villages, they saw one black and two white bears about 30 miles above the Little Missouri, and observed tracks along the river at other places. Near the junction of the Yellowstone with the Missouri River, on April 29, 1805, Captain Lewis, who was on shore with one of the hunters about 8 o'clock, met two white bears. He writes : Of the strength and ferocity of this animal the Indians had given us dread- ful accounts. They never attack him but in parties of sis or eight persons, and even then are often defeated with a loss of one or more of their party. Having no weapons but bows and arrows, and the bad guns with which the traders supply them, they are obliged to approach very near to the bear ; as no wound except through the head or heart is mortal, they frequently fall a sacrifice if they miss their aim. He rather attacks than avoids a man, and such is the terror which he has inspired, that the Indians who go in quest of him paint themselves and perform all the superstitious rites customary when they make war on a neighboring nation. Hitherto those bears we had seen did not appear desirous of encountering us; but although to a skillful rifleman the danger is very much diminished, yet the white bear is still a terrible animal. On approaching these two, both Captain Lewis and the hunter fired, and each wounded a bear. One of them made his escape ; the other turned upon Captain Lewis and pursued him 70 or SO yards, but being badly wounded the bear could not run so fast as to prevent him from reload- ing his piece, which he again aimed at him, and a third shot from the hunter brought him to the ground. He was a male, not quite full grown, and weighed about 300 pounds. The legs are somewhat longer than those of the black bear, and the talons and tusks much larger and longer. ... Its color is a yellowish-brown ; the eyes are small, black, and piercing; the front of the fore legs near the feet is usually black, and the fur is finer, thicker, and deeper than that of the black bear. Added to -which, it is a more furious animal, and very remarkable for the wounds which it will bear without dying." A few days later, May 5, Captain Clark and one of the hunters killed a large grizzly, said to weigh 500 or 600 pounds, and to measure 8 feet 7% inches, from the tip of the nose to the extremity of the hind foot. His front claws measured 4% inches, and his color was of a reddish or bay brown. This specimen, with measurements and description, formed the principal basis of Ord's later description and name of the species; neither skin nor skull can now be found. In 1833, Maximilian (Wied, 1839-1841, Bd. 1, pp. 419-420, 1839) notes in his journal on June 22: Near the great bend of the Missouri (just 3bove the mouth of the Little Missouri River), a large grizzly seen on the prairie on the north bank of the river, and "soon after two others were seen, one whitish, the other of a dark color." From this place on, in their journey up the river, the gray bears become 196 NORTH AMERICAN FAUNA [No. 49 more and more common. Above the mouth of the Knife River at the village of the Minnetarees, many of the Indians wore the large, valuable necklaces made of long bears' claws, and their hand- somely painted buffalo robes were fastened around the waist by a girdle. A few days later Maximilian met a chief of the Assiniboines wearing a necklace of bears' claws, blue glass beads in his ears, and a red flannel shirt. At Fort Union he (Wied, 1839-1841, Bd. 2, pp. 39, 302, 213, 215, 1841) mentions several distinguished men of the Assiniboines who arrived at the post on October 20, among whom was one Manto-Uitkatt (The Mad Bear). At the village of the Minnetaree Indians, about 30 miles above Fort Clark, he found an old chief, Lachpitzi-Sihrisch (The Yellow Bear), of whom he has much to say later. At the Knife River he gives the name of another Minnetaree chief as Lachpitzi-Wah-Kikihrisch (The Bear Hunter). Mato-Tope (Four Bears) was one of the most famous of the Mandan chiefs and a staunch friend of Maximilian. His son was named Mato Berocka (Male Bear). In the folio of plates accompanying Maximilian's Reise in das Innere Nord- America, many of the Indians shown are chiefs or famous hunters wearing grizzly-claw necklaces, among them Sioux, Mandan, Minnetaree, and Crow, who had won the right to wear these trophies of the hunt. In Plate 36 of the folio, Maximilian (Wied, 1839-1841, Folio, pi. 36) shows his hunters attacking two grizzlv bears as described in the text of his journal. On July 18 he (Wied, 1839-1841, Bd. 1, pp. 487-489, 1839) says: The hunters had seen several bears and on the ISth they saw two bears run- ning about on a sand bar in the river. One of these was shot and when mortally wounded rolled over, uttering fearful cries. It was a male about 3 years old and not of the largest size, but was 6 feet 2 inches and 2 lines from tip of nose to tip of tail, and from tail to tip of hairs, 8 inches. His color was dark brown with the points of the hair of a rusty color, but new hair already appearing which was lighter gray with yellow tips. This bear is known to be a very dangerous beast of prey and is willingly avoided by the hunters. . . . It is certain that many white men and Indians have been torn to pieces by these dangerous animals, especially in former times, when they were very numerous and lived to a great age. At Fort Clark, where he spent the winter with the Mandan Indians, he (Wied, 1839-1841, Bd. 2, p. 85, 1841) says: The grizzly bear approaches to within 4 miles of the fort because the Indians, who do not like to hunt them, leave them undisturbed. They are, however, very fond of the flesh of the young bear and the claws are much valued by them for the manufacture of their necklaces. On returning down the river in October, Maximilian (Wied, 1839-1841, Bd. 2, pp. 47-49, 1841) brought with him, among other live animals, some young bears in cages. Near the mouth of the Muddy River, on October 31, he found along the shores an abundance of buffaloberries, which were fed to the caged bears and proved an agreeable variety in their food. Since no game had been killed for several days, the live animals, which would not eat salt pork, were half famished, and the bears especially made an incessant growling, which was in every respect highly disagreeable. The next day an elk was shot and he says: " The "lamentations of my hungry ani- mals were put a stop to." Generally, however, the bears were found feeding on buffalo carcasses, which were often plentifully distributed in the quicksand or along the river banks by floods and breaking ice. 1926] MAMMALS OF NORTH DAKOTA 197 Apparently the Missouri River Valley with its great abundance and variety of large game, wild fruit, and berries, bulbs, tubers, roots, and underground beans, was a paradise for those bears before the days of the rifle. In regard to the breeding habits of the grizzlies, Maximilian (Wied, 1839-1841, Bd. 1, p. 510, 1839) says that only 2 or 3 young are generally raised, but 2 to 4 were sometimes recorded and some of the Indians even claimed that one group of 8 young had been found, but this he considered an exaggeration. In 1843, Audubon (1897, pp. 155-156, 41, 51, 64, 75, 86, 117, 122, 146) on his trip up the river to Fort Union, found the grizzly bears apparently as common as had Maximilian and Lewis and Clark over the same ground 10 and 39 years earlier. Just above Bismarck he found many of their tracks, and near the mouth of the Little Missouri, on August 22, he and his companions killed one and saved it for a specimen. They had seen many tracks the previous day and on the following day saw another bear. In the vicinity of Fort Union he found many tracks around the three conical hills called the Mammalles. On June 19 a grizzly bear was seen just across the river, on June 22 another near there, and others were seen on June 27, July 5, 13, 27, and 30, and August 12, which gives some idea of their abundance at that time. Audubon describes a man at Fort Union who had been attacked by a grizzly in the Black Hills; his face was badly mutilated, one eye had been torn out, and his arm and side were literally torn to pieces, but he lived for years afterward. There are many accounts of bears attacking both Indians and whites, and often without provocation. Audubon (1851-1854, vol. 3, pp. 145-146, 1854) records an attempt to kill an old bear and capture her two young, discovered near the shore from one of the steamers of the American Fur Co. The old bear was wounded and charged the hunters with such fury that they dropped their guns, jumped into the river, and hurriedly made their way back to the steamer. He relates another incident of an attendant at Fort Union picking peas in the garden when he suddenly discovered a large grizzly gathering peas at the other end of the row. He dropped his bucket and fled, and when the hunters arrived they found the bear eating peas out of the bucket. He paid no attention to them as they approached and was shot dead. In 1856, F. V. Hayden collected specimens of the grizzly bear near Fort Clark and a skeleton at Fort Union. At Devils Lake, in 1916, Frank Palmer, who had lived in North Dakota since 1867, told the writer that he had never known of any grizzly bears east of the Missouri River Valley. He said that the Sioux in their own language called some hills near old Fort Ransom, in Ransom County, where the Sheyenne River turns north, "The Bears' Den." Some of the Indians from near Devils Lake used to go clown there and hunt buffalo and Mr. Palmer was with them on one of these trips when he learned the name of these hills. In 1867, he says there were many grizzly bears on the river bottoms about Fort Buford, and also above and below there, and farther west in Montana. While carrying mail from Fort Buford west he often siiw them along the Missouri River bottoms, and they would not always get out of his way. 198 NORTH AMERICAN FAUNA [No. 49 In the Killdeer Mountains, in 1913, Jewett was told by the old settlers that grizzly bears were formerly common over all the country east and north of the Little Missouri River. Frank Donoyer, a veteran buffalo hunter, told of killing several of these bears in the Killdeer Mountains between 1864 and 1870. Dave Warren, assisted by a boy, killed two grizzlies in a gulch near Oakdale in the fall of 1897. From his ranch on the Little Missouri, Roosevelt (1900, pp. 55-56) writes : In the spring and early summer of 1888, the bears killed no cattle near my ranch ; but in the late summer and early fall of that year a big bear, which we well knew by its tracks, suddenly took to cattle-killing. This was a brute which had its headquarters on some very large brush bottoms a dozen miles below my ranch house, and which ranged to and fro across the broken country flanking the river on each side. It began just before berry time, but continued its career of destruction long after the wild plums and even buffalo berries had ripened. Again, he tells of bears attacking his cattle, killing white-tailed deer, attacking one of his cowboys, and killing an Indian near his ranch, and of numerous instances of bears killed under thrilling cir- cumstances during his ranching days on the Little Missouri. In 1887, when the writer visited Fort Buford, there were still a few grizzly bears in the river bottoms in that vicinity, but they were growing scarce. In a letter dated March 30, 1914, Clarence H. Packer, of Minot, states that his father trapped along the river bottoms, 25 miles south of Williston, in 1887, and at that time there were some grizzly or silvertip bears there. In 1889 William B. Mershon (1923, 1925), on a hunting trip along the Little Missouri, reported bear tracks everywhere, the sand bars literally tracked up by them, some of enormous size. He measured one track that was 8 by 14 inches. At the present time there is certainly not a grizzly bear left in the State of North Dakota, and it is doubtful if there is anywhere a living representative of this original species of the grizzly group that was first given a scientific name and status in literature. Its destruction, however, was more inevitable than was that of the buffalo or the other large game animals of the Plains, because, aside from its commercial value and its appeal to the most vigorous sportsmen as a worthy antagonist among large game, its presence in an agricultural and stock-raising region could not be tolerated. Like some of the savage tribes with which it was associated, it has in passing left behind a thrilling record of savage bravery of sur- passing interest to red-blooded Americans. Ursus absarokus Merriam Absaroka Grizzly Ursus absarokus Merriam, Proc. Biol. Soc. Washington, vol. 27, p. 181, 1914. Type locality. — Near head of Little Bighorn River, northern end of Big- horn Mountains, Mont. General characters. — Large, but smaller than Ursus IwrribiUs, with much smaller molar teeth. Color from skin of head and neck only, " Muzzle pale brown, changing to grizzled dark brown on head and face ; a large patch of dark brown free from grizzling on side of face extending from eye to angle of jaw; chin and gular region dark brown (except anterior part of 1926] MAMMALS OF NORTH DAKOTA 199 chin, which has not yet molted the pale old coat) ; top and sides of neck and doubtless body also, strongly grizzled." (Merriam, 191S, p. 93.) Measure- ments of skull of adult male : Basal length, 322 millimeters ; zygomatic breadth, 218; in inches, 12.7 and 8.6, respectively. Distribution and habitat. — Doctor Merriam (1918, p. 93) gives the range of this grizzly as " Laramie and Bighorn Mountains, eastern Wyoming, Black Hills region, South Dakota, and north- ward along Little Missouri to Missouri and Yellowstone Rivers." He says it appears to be a mountain species, while horribilis appar- ently was a Plains species. The only North Dakota specimen consists of a skull with accom- panying skin of head and neck presented to Doctor Merriam for the National Museum collection by Howard Eaton, of Wolf, Wyo. This bear was killed by Mr. Eaton on October 27, 1880, at the mouth of Bear Creek, which empties into the Little Missouri River from the east, opposite Bullion Butte. Apparently its range over- lapped that of the Plains grizzly to some extent, which may account for the supposed discrepancy in color of that species. Evidently this was a darker, browner bear than Ursus horribilis. Order IXSECTIVORA : Insect-eating Mammals Family TALPIDAE : Moles Scalopus aquaticus (machrinoides?) Jackson Missouri Valley Mole (PI. 19, fig. 1) Scalopus aquaticus machrinoides Jackson, Proc. Biol. Soc. Washington, vol. 27, p. 19, 1914. Type locality. — Manhattan, Kans. General characters. — Rather large for the common mole ; a compact little animal with beaklike naked nose, no functional eyes, minute ears, large spade- like front feet with five rigid claws, small hind feet, short, nearly naked tail, and dense plushlike fur of a brassy brown color. Measurements of average adults : Total length, 172 millimeters ; tail, 30 ; hind foot, 22.2. Distribution and habitat. — The northernmost form of the common mole of the eastern United States ranges from Arkansas up through Missouri, Iowa, and Minnesota to Elk River, and has been reported from Ottertail County, near Fergus Falls, and from Crookston. There is no definite record for North Dakota, but at Hankinson, in the southeastern corner of the State, some of the residents re- ported that mole ridges had been seen on the sandy soil in that region. Doctor Bell and the writer were unable to find any trace of moles or ridges, however, and but for their close proximity along the eastern border of the State, little weight should be given to the report. Until specimens are actually obtained from the State this must be considered as a hypothetical species and the identity of the form occurring there doubtful. The characteristic ridges along the surface of the ground, pushed up by these moles in extending their tunnels, are so unmistakable and so well known to those who have lived where they are abundant, that the presence of moles is easily recognized. It is often difficult to obtain specimens, as the moles are not easily trapped, Once discovered in the act of pushing 200 NORTH AMERICAN FAUNA [No. 49 up their ridges, however, they are easily caught by simply pressing down the earth back of them and then quickly scraping them out with hands, feet, or shovel. It is hoped that if the species does occur in North Dakota, specimens may be obtained to add this inter- esting and very useful little animal definitely to the list of the mam- mals of the State. Condylura cristata (Linnaeus) Star-nosed Mole (PL 19, fig. 2) [Sorex] cristatus Linnaeus Syst. Nat., ed. 10, t. 1, p. 53, 1758. Type locality. — Eastern Pennsylvania. General characters. — Starlike disk of sensitive filaments on tip of nose, and no visible eyes or external ears ; front feet wide, flat, and spadelike, but not so large as tbose of the common mole ; bind feet, slender ; tail, large and slightly bairy, usually swollen toward the base. Color, black or dusky, nearly uniform all over. Measurements of average specimens : Total length, 202 millimeters; tail, 78; hind foot, 28. Distribution and habitat. — The star-nosed moles are wide-ranging, Canadian and Transition Zone animals of eastern Canada and the northeastern United States, reaching their previously known western limit of range in central Minnesota and southeastern Manitoba. Seton (1909, vol. 2, p. 1137) records them from the vicinity of Win- nipeg, on the authority of W. R. Hine, who assured him " that specimens have been brought to his taxidermist shop in Winnipeg; unfortunately, they were not kept." This record from Winnipeg on the north and report of occurrence at Fort Ripley, central Min- nesota, mark a close approach to the State line on the east. The animals undoubtedly occur in the Red River Valley and Turtle Mountain country. At Towner, in 1915, Kellogg was told by James Lymburner of an animal answering the description of the star-nosed mole, which had on several occasions been found in his meadow. One was taken to the house and kept in a glass jar for a while as a curiosity, but no specimens were saved. Mr. Lymburner described it as having a long, pointed nose with a ring of soft, fleshy, finger- like projections. Its body was 2 or 3 inches long and its color a bluish black, which would indicate an immature animal. Kellogg says that several other persons described the same animal, but al- though traps were placed all over the meadow where the moles had been seen, none were caught. This seems to be the only available record for the State of this very useful little insectivorous animal, and while it adds the species tentatively to the State list, it only increases the importance of procuring specimens to substantiate the report. Family SORICIDAE : Shrews Sorex cinereus haydeni Baird Hayden Masked Shrew (PI. 20) Sorex haydeni Baird, Mamm. North Amer., p. 29, 1857. Type locality. — Fort Union (now Buford), N. Dak. General characters. — A tiny shrew with slender pointed nose, minute eyes, concealed ears, and slender tail about half as long as its body and three North American Fauna No. 49, U. S. Dept. Agr., Biological Survey. PLATE 19 Fig. I. — Common Mole (Scalopus aquaticus machrinoides) Photograph of fresh specimen. Half natural size «.^Mj _ *r ;* Fig. 2. Star-nosed Mole (Condylura cristatai Photograph of fresh specimen. Half natural size North American Fauna No. 49, U. S. Dept. Agr., Biological Survey. PLATE 20 Skins of Four Shrews (A) Short-tailed shrew (Marina brevicauda brevicauda); (B) Richardson shrew (Sorexarcticus); (C) .Hayden shrew (Sorex cinereus haydeni); (D) Pigmy shrew (Microsorex hoyi). About natural size 1026] MAMMALS OF NORTH DAKOTA 20l times as long as its hind foot. Fur, soft and fine ; color of upper parts, sepia brown, underparts, ashy gray. Measurements of an average adult, from Ken- mare : Total length, 95 millimeters ; tail, 33 ; hind foot, 11. An adult female at Walhalla measured 98, 3S, and 12 millimeters, respectively, and weighed 3.6 grams ; two others each weighed 3.3 grams. Distribution and habitat. — The Hayden masked shrew is a barely recognizable prairie form of the wide-ranging species cinereus (for- merly known as personatus) which with its various subspecies covers most of the northern part of the continent. The subspecies haydeni covers North and South Dakota and the prairie country of most of the surrounding States and Provinces. In North Dakota there are specimens from Buford, from near Williston and Lostwood, and from Kenmare, Bottineau, Birchwood, Walhalla, Grand Forks, Portland, Fairmount. Blaclaner, Oakes, Steele, Cannon Ball, and Selfridge, all in the National Museum collection. There is one from Minot in the Field Museum collection, and specimens from Fargo and Grafton in the agricultural college collection, at Fargo. Al- though rarely seen, they occupy practically every meadow, brush patch, and grove, and are found under rich vegetation almost any- where on the prairie. General habits. — These little masked shrews live on or under the surface of the ground, mainly under the cover of old leaves, grass, and fallen vegetation, where they make endless runwaj^s and tiny tunnels over and through the surface of the mellow soil, keeping almost as completely hidden from view as do moles and pocket gophers. Their long, flexible, sensitive noses apparently take the place of eyes in their dark tunnels and burrows, for their eyes have become mere specks, with apparently very limited range of vision. On rare occasions a shrew is uncovered in turning over a board or log or in loading hay that has stood long on the meadows. More often one is found dead in some trail where killed by a cat, weasel, or little owl, and found too musky for food. On soft snow in winter their tiny double lines -of tracks are fre- quently seen, and often little ridges the size of one's finger show where they have plowed tunnels close under the surface of the snow for passageways from place to place. When caught alive in the hands they struggle and fight with surprising strength and vigor, although their tiny teeth can do no harm and their vain struggles are like those of some vigorous insect. The little that is known of their habits comes mainly from trap- ping them either in their own little runways or burrows or catching them in traps set for meadow mice, whose larger runways they habitually follow. They are usually caught in small traps baited with fresh meat, bacon, or fat pork, and set under logs and in open places scooped out under the leaves and grass~ Traps set for mead- ow mice and baited with rolled oats also catch them, either because they accidentally run against the trigger or in some cases apparently because they stop to feed on the bait. Food habits. — The principal part of their food, as shown by the stomach contents consists of insects, earthworms, and the small ani- mal life found over the surface of the ground. They ravenously devour any kind of fresh meat, whether it be a meadow mouse or woods mouse, or one of their own kind, found in a trap, or a bit of bird meat or beef placed on the trap trigger to attract them, 202 WORTH AMERICAN FAUNA [No. 49 and in winter they will gather in numbers if a piece of frozen meat, lard, or tallow is left under a log for them to gnaw. They un- doubtedly catch many of the small rodents, and especially the young, for food, as they are capable of killing animals larger than them- selves. They apparently never accumulate any fat and are active throughout the year, evidently finding an abundance of frozen in- sects and other food along their tunnels on or below the surface of the ground. Breeding habits. — Very little is known of the breeding habits of shrews, but Stuart Criddle, of Treesbank, Manitoba, sent to the Biological Survey for identification eight half -grown young of this shrew taken on October 14, 1924. They were found dead in a grass nest under a sheaf of brome grass, and near them the head of a partly eaten shrew, probably their mother. So little is known of the habits of shrews that such scraps of information are important. Economic status. — Fortunately these bloodthirsty little animals are not large enough to do any damage to game or domestic poultry. In camps, cabins, and cellars in the wilderness they often gather in winter and become as numerous as some of the wild mice, and in some cases do slight damage by gnawing and soiling meat left within their reach. Any damage along this line, however, is so insignificant as to be negligible, while their constant destruction of insects and probably also their destruction of many small rodents, mark them as beneficial animals. There is still much to be learned of their habits and of the actual species of animals which furnish the bulk of their food. Sorex merriaini Dobson Merriam Shrew Sorex merriami Dobson, Monogr. Insectivora, pt. 3, fase. 1, pi. 23, fig. 6, 1890. Type locality. — Little Bighorn River, about IV-j miles above Fort Custer, Mont. General characters, — About the same size as haydeni, but readily distin- guished by buffy-gray upper parts and white underparts, feet, and lower half of tail. The skull characters show that it belongs to a very distinct group, but external characters are sufficient for easy recognition. Measure- ments of type specimen, preserved in alcohol : Total length, 90 millimeters ; tail, 36 ; hind foot, 11. Distribution and habitat. — The Merriam shrew is known only from the type specimen collected near Fort Custer, Mont., in 1884, by Major Bendire, and an imperfect specimen picked up by Jewett near Medora, on June 30, 1913. The latter was found on top of a dry Badlands butte, where evidently it had been caught by a hawk or a weasel and the head eaten off. Fortunately the skin was saved and the white underparts and the sharply bicolor tail served to identify it as this species. Apparently it is closely associated with the Bad- lands country and additional specimen will undoubtedly be taken in this semiarid region when more naturalists are on the lookout for rare species. Although a great deal of collecting of small mam- mals has been done in that general region, the scarcity of these shrews may be only apparent and due to some peculiarity of habits not yet learned by naturalists. Specimens should not only be saved, but any clue to their habits recorded, so that some light as to whether they are really scarce or merely escape observation may be obtained. 1926] MAMMALS OF NORTH DAKOTA 203 Sorex arcticus Kerr Richardson Shrew; Black-backed Shrew; Saddle-backed Shrew (PL 20) Sorex arcticus Kerr, Animal Kingdom, p. 206, 1792. Type locality. — Severn Settlement, mouth of Severn River, Ontario, Canada. General characters. — Size, rather large ; tail, of medium length ; nose, long and pointed ; eyes, minute ; ears, concealed. In winter, whole back dark brown or black sharply contrasted with buffy brown sides and gray-brown belly. In summer back dull brown but still strongly contrasted with lighter sides and underparts. Measurements of average specimens : Total length, 112 milli- meters ; tail, 40 ; hind foot, 14. Distribution and habitat. — A specimen of the black-backed shrew, now in the National Museum collection, was taken at Pembina by Charles Cavileer in 1855, and another by Robert Kennicott in 1861. One was taken at Fort Sisseton, just below the southeastern corner of the State, in 1877, by C. E. McChesney. Eastgate took one at Stump Lake in 1912, the writer took one at Kenmare in 1913, and Kellogg one at Fort Totten and another at Lostwood in 1915. There is one specimen in the Morris J. Kernall collection, taken at Valley City in 1912, and Kellogg reports one in the H. V. Williams collection taken at Grafton in 1915. This carries the range of the species diag- onally over the northeastern half of the State. It is a wide-ranging species, extending from the Mackenzie, through the Canadian Zone forests, to northern Michigan, Wisconsin, and Minnesota. In North Dakota the records are mainly from forested valleys, marshes, or lake shores. At Grafton, Williams reported catching these shrews on the edge of a marsh near town. Near the east end of Stump Lake Eastgate took one in some cold spring marshes, and at Kenmare the writer caught one in a trap set for meadow mice in a runway under the grass at the lower end of the Upper Riviere des Lacs, not far from cold gulches occupied by aspens and snowshoe rabbits. Usually there is a trace of Canadian Zone conditions where they are found. General habits. — Like most of the shrews, these more conspicuous saddle-backs are known mainly from specimens taken in traps set for small rodents under fallen grass in the meadows or under leaves and dense vegetation or old logs in the woods. Cold, damp places seem to be their favorite haunts in the southern part of their range, where the conditions of Boreal habitat are most nearly approached. These shrews are readily caught in traps baited Avith meat and set across the runways which they follow, but the few specimens taken indicate that they are by no means a common animal in this region. In food, habits, and habitat they seem not to differ from most of the other small shrews. Neosorex palustris (Richardson) Water Shrew; Marsh Shrew Sorex palustris Richardson, Zool. Journ., vol. 3, p. 517, 1828. Type locality. — Marshy places from Hudson Bay to the Rocky Mountains. General characters. — Largest of the long-tailed shrews in the region, with tail about as long as body, hind feet, large and fringed for swimming ; nose, long and pointed ; eyes, minute, and ears hidden in the fur. Upper parts, vel- 82242°— v20 14 204 NORTH AMERICAN FAUNA [No. 49 vety black, with sometimes a trace of brownish or gray ; underparts, silvery white, often clouded with gray or smoky. Measurements : Total length, 160 millimeters ; tail, 70 ; hind foot, 20. Distribution and habitat. — Specimens of water shrews collected at Fort Sisseton, just below the southeastern corner of North Dakota, and from Winnipeg, would imply that this species has a general distribution along the Red River Valley and in eastern North Da- kota, although no specimens are at present available from the State. They belong to a wide-ranging group of species occupying the Canadian Zone practically across the continent from Nova Scotia to Alaska, but generally associated with the marshes of the forest region. The Fort Sisseton record is apparently the only outlying prairie locality for the species. General habits. — Although named for the marshes where they are usually found, these shrews are more than palustrine in habits. As their structure indicates, they are expert swimmers and apparently spend much of their time and obtain much of their food in the water. At Elk River, Minn., where the writer collected them from 1884 to 1887, they were generally taken along the banks of the creek which flowed through the meadow. Traps set at little bur- rows under fallen grass on the creek banks, just above the edge of the water, would occasionally contain one of these shrews, and in winter a few were taken under the ice when the water had fallen and left an air space between two layers of ice. In no locality has the writer ever found them common, or in numbers sufficient to yield more than an occasional specimen among the many other shrews and meadow mice taken in the trap line. In the spring of 1886 a neighbor brought one that he had caught while it was swim- ming about in a small pond of snow water. He said it darted about through the water like a fish and when under the surface seemed coated with silver and even more fishlike. The stomachs and in- testines of those taken are usually found to contain particles of insects and small animals so well masticated that the species are not easily recognized. Of the breeding and other habits little is known. Microsorex hoyi hoyi (Baird) Pigmy Shrew (PI. 20) Borex hoyi Baird, Mamm. North Amer., p. 32, 1857. Type locality. — Racine, Wis. General characters. — Smallest of all North Dakota shrews and until a slightly smaller and closely related species was discovered and described by Preble in 1910, from near Washington, D. C, it was credited with being the smallest mammal in North America. In general proportions it approaches merriam/i and haydeni, but averages a little smaller than either. Upper parts, sepia brown ; underparts, ash gray ; tail, somewhat bicolor, brown above, whitish below. Measurements of specimens from Elk River, Minn. : Total length, 81.7 millimeters; tail, 30.7; hind foot, 10.7; weight, 2.9 grams. Distribution and habitat. — The pigmy shrew, with its several recognized forms, apparently fills the Canadian Zone across the northern part of the continent, but specimens are few and from widely scattered localities. The one record for North Dakota is based on a specimen found dead in a road on the north side of Devils 1926 MAMMALS OF NORTH DAKOTA 205 Lake in 1887. Unfortunately it was in such condition that only the skull could be saved ; but it proves to be Microsorex instead of Sorex personatus, as given in the writer's report for 1887. With all the collecting since done in the State, it seems remarkable that no others have been taken. The nearest localities outside of the State from which specimens are recorded are the Red River Settlement (Winni- peg), Manitoba; and Elk River, Minn. General habits. — Of the habits of these little shrews practically nothing is known except that they are caught in traps with other species in woods, clearings, or meadows. In Ontario, Miller (1897, p. 37) recorded them as invariably found in dry clearings and gardens. Blarina brevicauda brevicauda (Say) Short-tailed Shrew; Mole Shrew (PI. 20) Sorex brevicaudus Say, Long's Exped. Rocky Mountains, vol. 1, p. 164, 1823. Type locality. — West bank of the Missouri River, near Blair, Nebr. General characters. — A large heavy-bodied shrew with the usual small eyes, sharp nose, concealed ears, and short tail. The fur is short, soft, and velvety, varying in color from glossy plumbeous to almost black, with under- pays but slightly paler than the upper parts. The color is unmarked and almost uniform over the body. Measurements of adult male, from Wahpeton : Total length, 137 millimeters ; tail, 30 : hind foot, 17 ; of a female from same place : Total length, 136 ; tail, 28 ; hind foot, 16 ; measurement of a large male from Walhalla: 127, 25, and 19 millimeters, respectively; weight, 23 grams. Distribution and habitat. — There are specimens of the short-tailed shrew from Wahpeton, Fairmount, Oakes, Valley City, Portland, Fargo, Harwood, Grafton, Pembina, Walhalla, Turtle Mountains, Sweetwater Lakes, and Fort Berthold. From a wide range over the Transition and Upper Austral Zones of the northeastern United States and eastern Canada, these large shrews reach their north- western limit of range in North Dakota, extending commonly as far west along the stream and lake valleys as the eastern timber reaches. Over the open prairie and the drier western part of the State they seem not to occur, although an alcoholic specimen in the National Museum is supposed to have been taken in 1856 by F. V. Hayden, at Fort Berthold, on the Missouri River. At Wahpeton they were very common and a number were caught in traps set in meadow-mouse runways along the river and slough banks. At Hankinson the writer caught one in a patch of dense grass under a fence, but it was so badly damaged that it was not saved for a specimen. At Fair- mount, Sheldon found them common along the river banks, where a number of specimens were taken. At Oakes he reported them as fairly abundant along the James River, where they were found in the damp soil along the banks of the stream and also in patches of snowberry bushes. At Larimore, in Grand Forks County, Kellogg reported them, but did not collect specimens; at Manvel he also re- ported them as found about old strawstacks; at Drayton, in Pem- bina County, he was told by the farmers that during harvest they were often seen under shocks of grain. In the Turtle Mountains, near Fish Lake, in Roulette County, the writer caught two in the woods, one in a damp place on low ground and another under an old 206 NORTH AMERICAN FAUNA [No. 49 log near the lake shore. At the Sweetwater Lakes two specimens were taken in a dry marsh in the woods back of the lake shore, where they were living in the meadow-mouse runways under heavy fallen grass. At Portland Loring caught seven under logs and stumps in the woods. General habits. — Most of our specimens were taken in mouse traps baited with small pieces of meat or set in runways where bait was not necessary. The shrews eat many of the mice caught in traps and often leave only a piece of skin and a few bones to show what the trap had caught. In such cases they soon return and are almost cer- tain to get into the trap when it has been reset. They are savage little brutes and very strong and muscular for their size. When caught in small box traps or cans sunk in the ground, they are usually found dead after a few hours, apparently because they are unable to live without an almost constant supply of food. They are caught as readily in the daytime as at night, and at Wahpeton Kel- logg watched one digging a burrow in the ground about noon. Usually, however, they are not seen except as caught in traps or uncovered in moving logs or hay or grain that has been lying long on the ground. They burrow through the mellow soil and make run- ways between the fallen grass and leaves and the surface of the ground. Usually they are in moist, rich places where insect life is abundant, and where they have an ample supply of food while well concealed from enemies whose eyes are keener than theirs. Breeding habits. — Little is known of the breeding habits of these shrews, but an interesting note was obtained by Sheldon at Fair- mount, where on old female, which contained nine embryos, was taken May 28. Immature specimens are often caught in traps, but the very young have rarely been found and little is known of the nest or home conditions. Food habits. — Their principal food consists of insects, earth- worms, mollusks, and the various forms of animal life found on or near the surface of the ground, but these shrews are always eager for any kind of fresh meat. They devour many more desirable species of small mammals caught in traps where their runways are located. The number caught in traps set for meadow mice suggests that they deliberately follow the mouse runways for the purpose of capturing their prey. Even if not able to catch and overpower the full-grown meadow mice, which are nearly twice their size, they will certainly catch, kill, and eat many of the young and immature individuals. Their stomachs and intestines are usually well filled with food, but are never found distended, as are those of rodents, especially mice, after hearty feeding. They are never fat and are active throughout the year, in winter evidently getting their food on or under the surface of the ground, although occasionally coming to the surface of the snow and making long lines of their peculiar little double rows of tracks. They burrow in and out of the snow at will and in midwinter are easily lured to a food supply of frozen meat placed under logs, hay, or fallen vegetation on the surface of the ground. When once in the habit of coming to a food supply they are easily caught in considerable numbers. Economic status. — There are no injurious habits of any conse- quence chargeable to these shrews, although they often come into 1926] MAMMALS OF NORTH DAKOTA 207 cellars and storage places on the farm, and if meat or milk is left within their reach will gnaw and soil the meat and eat the cream around the edges of the milk pans. Conditions where they can do such mischief are, however, rare and unnecessary. Their destruc- tion of insects and great numbers of worms and other small animals inhabiting the richest soils goes on continuously throughout the year and their destruction of small, injurious rodents is undoubtedly of great benefit to agriculture. While many of our small mammals must be considered enemies and destroyed in every possible way, fortunately some, as the shrews, may be classed as wholly beneficial and their presence welcomed on the farm. Of all the small mam- mals, perhaps the shrews with their voracious and often cannibalis- tic natures are least lovable, but they can all be accepted as very useful allies. Order CHIROPTERA: Winged Mammals Family VESPERTILIONIDAE : Common Bats Nycteris cinerea (Beauvois) Hoary Bat; Great Gray Bat (PI. 21, fig. 1) Vespcrtilio cinereus Beauvois, Catal. Peale's Mus. [Philadelphia], p. 15, 1796. Type locality. — Philadelphia, Pa. General characters. — Teeth, 32 ; size, large ; spread of wings, about 16 inches ; ears, short and rounded with black naked rims ; top of feet and tail membranes, furry ; fur, full and soft ; color, yellowish brown, frosted with white above and below ; throat and wing linings, bnffy. Measurements of adult male, from North Dakota specimen : Total length, 130 millimeters ; tail, 60 ; hind foot, 13 ; forearm, 51. Distribution and habitat. — From a wide Boreal range across the northern part of the continent and southward in the mountains, and a winter migratory range to the southern border of the United States, the big hoary, or gray, bats cover at one time or another all of North Dakota. They are undoubtedly far more common than the few scattered records seem to imply, as their nocturnal habits conceal them from common observation. In 1833, at Fort Clark on the Missouri River, Maximilian (Wied, 1839-1841, Bd. 1, pp. 403-104, 1839) collected an adult female, which he described in much detail as to the color and measurements. About 1861, F. V. Hayden (1875, p. 95) collected a specimen at Fort Union (now Buford) and re- ported them as found " all over the United States east of the Rocky Mountains." In 1887 the writer recognized one of these bats on the wing at Pembina, August 3, and saw several about the woods on the north side of Devils Lake, August 6, but obtained no specimens. A specimen from Minot, N. Dak., was recorded in the catalogue of the Field Museum. (Elliot, 1907, p. 514.) A specimen collected June 20, 1913, was sent to the Biological Survey for identification by Daniel Freeman, of the agricultural college at Fargo, and in 1914, Bell and the writer shot four specimens in the little forest area between the lakes on the Hankinson farm near Hankinson. At Wahpeton, in 1915, the boys described a large gray bat which they had found hang- ing in the leaves of a tree, which was undoubtedly this species. At 208 NORTH AMERICAN FAUNA [No. 49 Grafton, Kellogg found one dead in a cow path, but it had been so trampled by cattle as to be worthless for a specimen. At Towner he reported two seen one evening and perfectly identified by their large size, but before he could get his gun they had disappeared. General habits. — These great gray bats are powerful and rapid fliers ; they usually appear rather late in the dusk of evening and are rarely noticed except by bat hunters. During the day they hang head downward in clusters of leaves, usually at the ends of branches of trees. This habit restricts them to the forested areas or to country about ranches, where they can find sufficient foliage for roosting sites. At the Hankinson ranch, where the fine old elms, oaks, ashes, basswoods, and boxelders form heavy foliage and deep shade, they were found to be one of the common species. As they darted swiftly across the narrow spaces between the trees, the collectors had much trouble in shooting four specimens among the considerable number of smaller bats taken during the twilight. Their large size, however, made it possible to obtain all of those killed, while many of the smaller bats were lost in the grass and weeds. Some of these specimens were evidently young of the year, but practically full grown, and probably were born in this particular grove, although strong enough to have flown from a considerable distance. These bats are migratory, and as the cold weather ap- proaches and insect life becomes scarce they move southward at least to the southern border of the United States and probably be- yond for the winter season. Their breeding range has not been well worked out, but apparently they breed mainly in the cooler zones of the Northern States, Canada, and the high mountain areas. Food habits. — Little is known of the food habits of the hoary bats except that they are insectivorous and capture their prey on the wing in swift zigzag flight, most baffling to the collector. At times they seem to be gleaning among the branches of the trees and at other times they circle high over the forest, apparently snapping up the insects that swarm far overhead. Breeding habits. — The specimen collected at Fort Clark by Max- imilian on June 12, 1833, was an adult female containing two large well- developed embryos, which he describes as quite naked, with wings folded over their noses. The specimen from the agricultural college was a female taken June 20, 1913, which contained two large embryos, now also preserved in alcohol. Other specimens have been recorded containing two embryos, and mother bats have been shot while flying about in the evening with two young clinging to their sides. There are two mammae close together on each side of the breast, located in a subaxillary position, and when the mother hangs head downward in the foliage they are just above the two fur-lined cradles formed by the hollows of her folded wings. Apparently the young cling to the body of the mother during her flight and are thus always with her until old enough to use their own wings. The young are surprisingly large at birth and it seems probable that they grow rapidly and do not long burden the parent. If born after June 20 and practically full grown by July 22, their growth and development must be very rapid. Economic status. — Besides being in every way harmless and unob- jectionable, these bats are, through their insectivorous habits, of un- North American Fauna No. 49, U. S. Dept. Agr., Biological Survey. Plate 21 BI33IM Fig. I. — (A) Hoary Bat (Nycteris cineria); (B) Silver-haired Bat (Lasionycteris noctivagans) About two-thirds natural size Fig. 2. -Say Bat (Myotis subulatus subulatus) About two-thirds natural size 1926] MAMMALS OF NORTH DAKOTA 209 questionable value to man. The extent of their destruction of noc- turnal pests and their importance to the welfare of the forests and other products of the country will not be known until a thorough study of their food has been made. Nycteris borealis borealis (Miiller) Red Bat; New York Bat Vespertilio borealis Miiller, Natursyst., Suppl., p. 20, 1776. Type locality. — New York. General characters. — Teeth, 32; size, medium; expanse of wings, about 12 inches ; ears, short and rounded, mainly naked inside and on rims ; top of tail membranes and feet well furred; color, bright rusty or pinkish yellow with slight frosting of white over back and breast. Measurements of large female, from Grinnell, N. Dak. : Expanse, 330 millimeters ; length, 117 ; tail, 52 ; hind foot, 10; forearm, 41; a smaller female from the same place measures 298, 104, 45, and 9, millimeters. Distribution and habitat. — There are specimens of the red bat from Devils Lake, Grinnell, the Yellowstone River (probably Buford), and "Chantee Hills." At Wahpeton some boys reported that they often found a very pretty red-colored bat hanging in the plum trees when they were gathering plums. Across the river from Fargo, Murie often found them hanging in plum trees in July and August, and Williams has one taken near Grafton on August 23, 1919. At Hankinson these yellowish-red bats were accurately described and were said to be often found hanging in the leaves. At Stump Lake, where the wind blew so incessantly that the writer was not able to get any bats, people said that two kinds were found, one red and the other dark brown. At Beaver Creek, 4 miles west of Grinnell, where the Missouri makes its first big bend to the south- ward, Kellogg reported red bats common in the woods and found two hanging in the leaATes of branches. The " Chantee " or " Chartee " Hills the writer has not been able to locate; and another specimen collected by Hayden has no positive locality other than the Yellow- stone River. The red bats are abundant over the eastern United States, mainly in the Austral Zone, and the few specimens from North Dakota may be wanderers after the breeding season or scattered individuals beyond their normal breeding range. General habits. — At least in the northern part of their range these bats are migratory and with the beginning of cold weather move southward until suitable hibernation quarters or a comfortable cli- mate and an ample food supply are found for the winter. In the daytime they hang head downward among the leaves of trees or bushes. In July, 1887, in a little grove on the north side of Devils Lake, the writer picked five specimens from the leaf clusters of box- elder trees, where they were hanging at about the height of the head above the ground. They were well concealed by the leaves, but after the first one was noticed there was no trouble in finding others. Their bright yellow color contrasted sharply with the dark green foliage. In places where common these bats often come out of their diurnal roosts rather early in the dusk of evening and occasionally one is seen flying about on a cloudy day or even in bright sunlight, doubt- less when it has been disturbed or the sunlight has penetrated to its 210 NOETH AMERICAN FAUNA [No. 49 sleeping quarters. They have a quick and erratic flight and when darting back and forth among the trees in the dusk of evening are no easy target for the shotgun. As most of the specimens are ob- tained by evening shooting, it is not so remarkable that they are scarce in collections even where fairly common outside. Food habits. — Little is known of the food habits of the red bat, except that it seems always busy catching insects while on the wing, and the stomachs of those collected for specimens are usually dis- tended with a mass of finely masticated insects of great variety, mainly unidentifiable. Eptesicus fuscus fuscus (Beauvois) Large Brown Bat Aprd/pMaa, of the Mandans ; Ishwa- tdshia of the Hidatsas ; Eupdhu- waki'kadakena of the Dakotas (all, Gilmore). Vespertilcti [sic] fuscus Beauvois, Catal. Peale's Mus. [Philadelphia], p. 14, 1796. Type locality. — Philadelphia, Pa. General characters. — Teeth, 32 ; size, rather large ; expanse of wings, about a foot; ears, prominent and pointed; membranes of ears, wings, and tail, naked ; fur, long and lax ; color, glossy light hair-brown, slightly paler below ; ears and membranes, black or blackish. Measurements : Spread of wings, 324 millimeters ; total length, 117 ; tail, 50 ; foot, 11 ; forearm, 44. Distribution and habitat. — A series of five specimens of these large brown bats taken by Kellogg at Grinnell (where the Missouri River turns southward), August 29 and 30, 1915, and six taken by Sheldon at Cannon Ball, September 2, 1915, give, for the first time, a good representation of this species from the State. The old specimen collected by Hayden at " Fort Union, Nebr." (now Buford, N. Dak.), about 1861, was the only certain previous record on which to admit the species to the State list, although Maximilian in 1833 recorded what was probably this species under his Vespertilio ursinus. In 1912 Kellogg also reported two skins seen in the State university collection, taken at Stump Lake. At Grafton, Williams reports having killed a large brown bat a number of years ago, but no later records have been obtained. At Hankinson, in 1912, the writer saw several bats which he believed to be the species flying among the trees, but did not obtain a specimen. One was also recognized in the evening at the Sweetwater Lakes, but not procured. At Buford, in 1910, Anthony shot a large brown bat that he sup- posed to be of this species, but it fell in the brush on the river banks and could not be found. Having a transcontinental range extending north into the edge of the Boreal Zone, these bats undoubtedly cover the whole of North Dakota in considerable numbers. The specimens from Grinnell and Cannon Ball are decidedly pale and appear to be shading toward the western form, Eptesicus fuscus pallidus Young, but are not sufficiently marked to be referred to it. General habits. — At Cannon Ball Sheldon found these brown bats between the walls of old buildings, and specimens taken 4 miles west of Grinnell by Kellogg were found around a house in the Badlands. Like many other species of bats, they spend the daylight hours hidden away in cracks and dark cavities in walls, roofs, or cornices 1926] MAMMALS OF NOKTH DAKOTA 211 of buildings, in hollow trees, under bark, and less commonly in the clefts of rocks. In the rather late dusk of evening they come out and after quenching their thirst at the nearest pond or stream begin their nightly hunt for winged insects. Usually they are found fly- ing rapidly about buildings and trees or through the open spaces in forests and groves. Their flight is swift and erratic as they snap up one flying insect after another, and it is only an occasional speci- men that can be brought down with fine shot as they zigzag against the twilight sky. Where they are at all common, bat shooting may become very exciting, both from the effort and dexterity required to bring down specimens and also from the possibilit}^ of getting rare species. In some places only one species will be abundant, while others are rare or entirely absent, and in other localities several species and genera will be found equally abundant and hunting over the same ground. Hibernation. — In fall these large brown bats become excessively fat, with a heavy layer of very oily tissue lying under the skin and filling the body cavities. Although probably in part migratory, they seem to hibernate throughout their range, disappearing with the first frosty nights, occasionally reappearing on warm evenings, but again entering their permanent hibernation before the real cold weather begins. They crawl away into the cracks and walls of buildings or any sufficiently sheltered place where they can spend the winter without too much exposure to cold. Occasionally dur- ing the most severe weather, when houses are overheated, some of these bats are roused from their winter sleep in the walls and ap- pear inside the rooms, flying about in good condition, apparently under the misapprehension that summer has arrived. Such speci- mens are usually found to be exceedingly fat, but with empty stomachs. In spring they are one of the first bats to appear with the warm days, the time of beginning plant and insect activity. Migration. — To what extent these bats migrate is not known, but like other species they probably make considerable flights to and from their favorite winter quarters. In the mountains, where they range high in the Canadian Zone late in summer, they undoubtedly return to lower milder levels to find winter quarters. It is doubt- ful whether an extensive north-and-south migration is common to the species. Breeding. — The mating season is late in July or early in August, and the young are born in June of the following year. In June the males and females are usually found in separate localities or, if in the mountains, at different levels. By June 20, females are usual- ly found carrying one or two large embryos, which at birth are very large for the size of the parent. The two mammae are located on the sides back of the wing bases, so that when the mother hangs head downward the suckling young are neatly cradled in the fold of the wing. As with other species of bats, the young are probably carried clinging to the mother's body until able to fly and catch their own food. Their development is evidently rapid, for by July 26 the young are flying, and immature specimens have been collected as early as July 24. Food habits. — Not much is known of the species of insects on which the brown bats feed, but there seems always to be an abundant sup- 212 NOETH AMERICAN FAUKA [No. 49 ply, and soon after the bats have begun flying their stomachs are found distended with a finely ground mass of insect remains. In some localities traces of various beetles are detected in the stomachs. Little is known of the species but it is certain that a vast number of insects are consumed by each bat. Economic status. — The comparative value of different species of bats can not be determined until their food habits have been thoroughly studied and their choice of food and the species of in- sects consumed more fully determined. In general, however, their usefulness can be well compared with that of insectivorous birds, for many of the most destructive insects are active only at night and if by day they escape the birds, they are devoured in millions by these aerial guardians of the night. Lasionycteris noctivagans (LeConte) Silver-haired Bat; Silvery Bat; Black Bat (PI. 21, fig. 1) V[espertilio~\ noctivagans LeConte, Cuv. Anim. Kingdom, McMurtrie ed., vol. 1, p. 431, 1831. Type locality. — Eastern United States. General characters. — Teeth, 36 ; size, medium ; spread of wings, about a foot ; ears, medium, nearly quadrate, about as broad as long, naked ; upper base of tail membrane, hairy; fur, long and soft; color, dark brown, sooty, or black with white-tipped hairs over back and belly ; ears, wings, and feet, sooty or black. Measurements of adult female, from Grafton: Spread of wings, 301 millimeters ; total length, 105 ; tail, 42 ; hind foot, 9 ; forearm, 42. Distribution and habitat. — With a breeding range apparently over the Transition and Canadian Zones across the continent, these bats cover at least all the forested parts of North Dakota during the breeding season and the months of greatest insect activity. There are specimens from Fort Union, Bottineau, Minot, Fargo, Grafton, and Stump Lake. A specimen in the National Museum, collected by F. V. Hayden at Fort Union, has no date but was entered in the catalogue in 1863. On August 23, 1887, the writer found two of these bats under a piece of loose bark on a dead tree in the edge of the Turtle Mountains near Bottineau. On May 12, 1913, Wil- liams took an adult male at Stump Lake; and on July 22, 1915, he collected a less than half grown young at Grafton. It was brought to him alive, and flew about the room before it was killed. An adult female was taken by Kellogg in a pile of fence posts at Graf- ton on June 11, 1915, and one recorded in the Field Museum cata- logue was taken by W. E. Snyder at Minot. A specimen sent to the Biological Survey for identification by Professor Freeman, of the agricultural college, was collected on September 12, 1914. At Wahpeton the writer was told that a black bat was found there. These meager records for the State do not so much indicate the rarity of the species as the difficulty of obtaining specimens and information regarding the habits of a strictly nocturnal species which can not be caught in traps. In reality, these bats are prob- ably common at one season or another in every patch of woods over the State. General Tidbits. — More than most species the black bats are forest dwellers, apparently very largely depending on the cover of loose 1026] MAMMALS OF NORTH DAKOTA 213 bark and hollow trees for their diurnal roosts and keeping mainly among the trees in their nocturnal flight. They usually appear lather late in the evening and, after quenching their thirst at the nearest water, begin their rapid flight between and around the branches of trees. The lateness and swiftness of their flight render them especially difficult to shoot and apparently as many specimens are picked up in their hiding places as are obtained with the shotgun. Food habits. — Like other bats, they are eager in their pursuit of night-flying insects, but the particular species chosen or rejected are not known. Their habit of hunting among the trees would indicate their especial value in forest protection. Breeding habits. — The mammae in these bats are one pair, ar- ranged on the sides just back of the wing base. The young, as indicated by embryos, are usually two, but sometimes only one. The female taken by Kellogg at Grafton on June 11 contained two em- bryos, and the young taken by Williams on July 22 was apparently not more than 2 or 3 weeks old. In his Mammals of the Adirondack Region, New York, Merriam (1884, p. 190) states that females killed during the latter part of June were heavy with young, and that up to July 1 not one had given birth to offspring, but that all killed after July 4 were then suckling their young. He also records the discovery of an old crow's nest which contained embedded in the sticks and litter 13 young bats, with their eyes not yet open. Although not positively identified, they were supposed to be the young of this species. Merriam also says that these bats begin to fly when 3 weeks old, those killed on the first evening weighing only about half as much as their parents. Migration and hibernation. — The black bat is one of the species with a well-established record for migration, appearing in fall and winter far south of its summer range and possibly moving far enough south to avoid the necessity of hibernation. More probably, however, it is prepared to creep away under cover and become dormant for a period of cold weather and scant food supply even in the southern part of its winter range. As with the hoary bat, its migratory habits may well explain the uniformity of characters of the species over a wide range across the continent. My otis lucifugus lucifugus (LeConte) Little Brown Bat V [caper tilio] lucifugus LeConte, Cuv. Aniin. Kingdom, McMurtrie ed., vol. 1, p. 431, 1831. Type locality. — Georgia ; probably Rieeboro. General character. — Teeth, 38; size, very small with small, pointed, naked ears, which laid forward, do not roach to the tip of the nose; wing and tail membranes, naked and dark brown ; fur, soft, glossy hazel brown, bright buffy below. Measurements by Doctor Mearns of specimen from Fort Snelling, Minn. : Expanse, 2G0 millimeters ; total length, 94 ; tail, 41 ; hind foot, 9 ; fore- arm, 38 ; ear from notch at base, 13 millimeters. A very fat male at Elk River, Minn., measured as follows : Expanse, 260 ; total length, 96 ; tail, 37 ; and hind foot, 10 millimeters, and weighed 12.68 grams. Distribution and habitat. — The little brown bats with their sev- eral subspecies cover all but the extreme northern part of North America, but the ranges of the different forms have not been deter- mined and are somewhat indefinite. The type species, lucifwgus, 214 NORTH AMERICAN FAUNA [No. 40 covers the eastern United States and apparently reaches its western limit in eastern North Dakota, grading into carissima toward the western part of the State. The only specimens that seem to be typ- ical of this little dark form are from Hankinson and Fargo, although specimens from Stump Lake, Devils Lake, and Esmond Lake are evidently intermediate in character and could be referred to either lucifugus or carissima. Those from the western part of the State at Towner, Goodall, Cannon Ball, and Bismarck are referred to caris- sima, although not fully typical of that paler western form. Appar- ently one form or the other is common all over the State wherever there are buildings, water, and mosquitoes, without much regard to timber. General habits. — The little brown bats are mainly cave, cliff, or house dwellers, spending the daylight hours hung up in dark crevices, caverns, and rooms, where well protected from the daylight. From their dark retreats they come out at the dusk of evening and after visiting the nearest open water, where their thirst can be quenched as they skim over the surface, they begin hawking after insects, around and around the buildings, under and among the branches of the trees, or over the ponds and marshes wherever insect life is abundant. If the wind is blowing, the bats work in the shelter of buildings or trees or get inside of buildings where insects also take refuge from the wind. They are quick, crooked fliers and by no means an easy mark for the collector, unless they can be silhouetted against the western light for sufficient distance to allow time for a double shot. At Hankinson this seemed to be the most common of the small bats that flew in considerable numbers about the buildings on warm evenings, and the fact that few specimens were taken was due to the limited spaces between the buildings and the treetops. Near Fargo they were numerous over the surface of the river in the evenings, but were usually flying so low among the trees that no clear shots could be obtained against the light portions of the sky. After drinking, they usually left the river to circle about the buildings in town or on the farms, where shooting was not permitted. It is always a great exasperation to a field naturalist to see unknown species of bats circling about buildings and over the streets of towns where he can have no hope of obtaining specimens unless one strays into his room at night and can be captured with a towel. Occasional specimens are thus obtained, but in most cases our meager information of the distribution and habits of bats is very slowly accumulated. Breeding habits. — In these bats the embryos are usually, and per- haps always, one, and the single pair of mammae are on each side back of the wing bases. Of the habits and care of the young so little is known that a most interesting field for close observation remains almost untouched. Food habits. — About the buildings where these bats are usually most numerous the insect life often seems to consist largely of mos- quitoes, flies, and nocturnal beetles and moths. The actual species eaten are not well known, but the bats are certainly industrious gleaners and in a few minutes after they have begun to fly their stomachs are found filled to capacity. 1926] MAMMALS OF NORTH DAKOTA 215 Myotis lucifugus carissima Thomas Yellowstone Bat Myotis (Leuconoc) carissima Thomas, Aim. and Mag. Nat. Hist., vol. 13 (ser. 7), p. 383, 1904. Type locality. — Yellowstone Lake, Yellowstone National Park. General characters. — Teeth, 3S; ahout the size of typical lucifugus; ears, small and pointed ; fur, glossy ; colors, light hazel brown above, buffy below ; ears and membranes, dark brown or blackish ; tail membranes, edged with gray ; young of the year, darker, without gray edges on tail membranes. Meas- urements of typical specimen, from Yellowstone Park: Expanse of wings, 260 millimeters ; total length, 94 ; tail, 40 ; hind foot, 10 ; forearm, 38. Distribution and habitat. — This northern Rocky Mountain form of Myotis ranges at least from western Montana to the Black Hills and over western North Dakota. Specimens from Towner, Goodall, Bis- marck, and Cannon Ball, while not typical, are nearer this species than to the eastern little brown bat, while some of those from the Devils Lake country show intermediate characters. Apparently the intergradation between the two forms is gradual across the middle part of the State. General habits. — On the Mouse River, 8 miles north of Towner, Kellogg collected 28 specimens of these bats on August 2, 1915. They were found behind a barn door and all of the 9 adults were females; of the 17 dark-colored but full-grown young of the year, 8 were males and 9 females. This was evidently a breeding colony from which the adult males were keeping a respectful distance. On September 4 Kellogg collected an adult male at Goodall, at the edge of the forest along Antelope Creek. At Fort Totten on July 15 he collected a male at the edge of the forest and on July 17 a female as it was flying around buildings in the evening. On November 19, 1919, Russell Reid found one dead hanging on the wall of a house in Bismarck. It evidently had been dead for some time, as the tem- perature had been 10° F. as early as October 26. At Cannon Ball, on August 20, Sheldon collected four adult females, part of them in the eaves of an old building, while others were shot in the evening as they flew about, after it had become so dark that it was hard to find them. In the Yellowstone Park, where the subspecies was first discovered, they occupy a large warm cave, called the Devil's Kitchen, in great numbers through the summer, but apparently leave for cooler caves in which to hibernate during the cold season. In the Bitterroot Valley, western Montana, a large breeding colony was found in a bridge over the Bitterroot River; they returned each spring from their winter quarters and many of the old and young were taken for specimens. Breeding habits. — In a large number of specimens examined only one embryo was found in each. The mammae, as in other species of the genus, are two in number, one on each side back of the wing base. Little is known of the breeding habits other than that some females were shot while flying about with the young cling- ing to them, and others collected had evidently left the young at home while hunting for their food. Migration and hibernation. — In fall these bats disappear with the first cold nights and consequent reduction in the abundance of insect life. Some may hibernate in buildings, but apparently most of them resort to caves, the location of which they seem to be 216 NORTH AMERICAN FAUNA [No. 49 familiar with, and here they hang during the winter and remain dormant in the cool, but not freezing, air. In a level prairie country like North Dakota, it may be necessary for them to make long jour- neys in search of winter quarters, but their migration is as imper- fectly known as their other habits. Myotis subulatus subulatus (Say) Say Bat (PL 21, fig. 2) Vlespertiliol subulatus Say, Long's Exped. Rocky Mountains, vol. 2, p. 65, 1823. Type locality. — Arkansas River, near La Junta, Colo. General characters. — Teeth, 38 ; small and very similar in appearance to lucif- iigus, but readily distinguished by its longer ears, which reach, when laid for- ward, well beyond the tip of the nose. From the larger-eared evotis it is distin- guished by darker color of fur and narrower, more pointed ears; ears and membranes, naked and dark brown ; fur, soft and lax ; color, yellowish- brown, slightly paler below. Measurements : Expanse of wings, 247 milli- meters ; total length, 95 ; tail, 41 ; hind foot, 9 ; forearm, 37 ; ear from notch at base, 16. Distribution and habitat. — Miller gives the range of the Say bat as North America, east of the Rocky Mountains, but its distribution is irregular; and although abundant in many places, it is often locally scarce, or absent. Over a wide strip of prairie country from the Gulf of Mexico to Manitoba there are very few records of its occurrence, while to the eastward and westward in rough country where caves are more numerous, the map shows many records. The collected long ago per Missouri." These few, collected by Hayden, Carpenter, and Rothhammer, are in the National Museum collection. Specimens have been taken south, west, and east of the State and may be found at any locality over it, but they are more likely to be found in the western Bad- lands country in close proximity to cliffs and caves. Their presence near Buford is probably accounted for by the many little caves or openings in the rocky cliffs bordering the Missouri Valley. Some of the many small bat's seen along the Little Missouri at Medora, and others along the river south of Bullion Butte, may also have been in part of this species. They are known to roost and hibernate in caves, but of their specific habits our knowledge is very indefinite. Breeding habits. — Examination of females taken for specimens indicates two as the usual number of young of this species, although the mammae are of the same number and position as in other species of the genus Myotis, one on each side back of the wing base. Myotis evotis (H. Allen) Little Long-eared Bat Vespertilio evotis H. Allen, Monogr., Bats North Amer., p. 48, 1864. Type locality. — Monterey, Calif. General characters. — Teeth, 38 ; size rather small ; ears, strikingly large, naked, black ; wing and tail membranes, dusky or black ; fur, soft and lax ; color, glossy buft'y yellowish above, buffy or whitish below. Measurements of an adult male taken near Grinnell by Kellogg: Expanse of wings, 292 milli- meters ; total length, 90 ; tail, 42 ; hind foot, 9 ; forearm, 39 ; ear from notch at base (measured dry), 18. Wlieie UaVCS ttlo IIIUIC UlUIlCluua, inc map on^vvo iiiciiij j only specimens from North Dakota are a few collec1 and labeled "Fort Union." "Fort Buford," or the "Upp 1926] BIBLIOGKAPHY 217 Distribution and habitat. — The little long-eared bat is found in western United States and Mexico, mainly in the Austral and Transition Zones. A single specimen from Beaver Creek, 4 miles west of Grinnell, is the only record of the species for North Dakota and this is a considerable extension of its known range eastward, Loveland, Colo., being its previously known easternmost locality. The specimen, an adult male, found by Kellogg in his room on August 26, 1915, was clinging to the side of a smooth plastered wall. It may represent a stray wanderer in this Badlands region of cliffs and caves, or it may have been within the regular range of the species. If the latter is the case, additional specimens should be obtained from other places. BIBLIOGRAPHY Allen, J. A. 1875. Notes on the natural history of portions of Dakota and Montana Territories. Proc. Boston Soc. Nat. Hist., vol. 17 (1874-75), pp. 33-85. - 1876. The American bisons, living and extinct. Mem. Geol. Survey Kentucky, vol. 1, pt. 2, illus. (Also issued as Mem. Mus. Comp. Zool., Harvard Univ., vol. 4, no. 10.) Amebican Ornithologists' Union. 1910. Check-list of North American birds. Ed. 3, 430 pp., illus. New York. Anonymous. 1900. Woodchucks. Amer. Field, vol. 53, p. 211. Audubon, J. J., and J. Bach man. 1851-[54]. The quadrupeds of North America. 3 vols., illus. New York. Audubon, M. R., and E. Coues. 1897. Audubon and his journals, vol. 2, illus. New York. Bailey, V. 1888. Report on some of the results of a trip through parts of Minnesota and Dakota. Rpt. Commr. Agr. [U. S.] 1887. pp. 426-454, illus. 1922. Beaver habits, beaver control and possibilities in beaver farming. U. S. Dept. Agr. Bui. 1078. 29 pp., illus. 1926. Construction and operation of Biological Survey beaver trap. U. S. Dept. Agr. Misc. Circ. 69, 4 pp., illus. Bailey, V., W. B. Bell, and M. A. Brannon. 1914. Preliminary report on the mammals of North Dakota. North Dakota Agr. Exp. Sta. Circ. 3, 20 pp. Baird, S. F. 1857. General report upon the zoology of the several Pacific railroad routes. Part I. Mammals. 7f>7 pp., illus. Washington, D. C. (U. S. War Dept., Reports of explorations and surveys . . . 1853, vol. 8.) Bergman, H. F. [1918]. Flora of North Dakota. North Dakota Agr. Col. Survey, Bien. Rpt. 6 [1911-12], pp. 151-372, illus. Brackenridge, H. M. 1816. Journal of a voyage up the river Missouri ; performed in eighteen hundred and eleven. Ed. 2. Baltimore. (Reprint in R. G. Thwaites's Early western travels, 1748-1846, vol. 6, pp. 21-106. Cleveland. 1904.) Cooper, J. G. 1869. Notes on the fauna of the Upper Missouri. Amer. Nat., vol. 3, pp. 294-299. Coues, E. 1875. The prairie gopher. Amer. Nat., vol. 9, pp. 147-156. 1877. Fur-bearing animals. 348 pp., illus. Washington, D. C. (U. S. Geol. and Geogr. Survey Ter., Misc. Pub., no. 8.) Criddle, S. 1915. The banded pocket mouse, Perognathus fasciatus Wied. Ottawa Nat., vol. 28, pp. 130-134, illus. 1926. The habits of Microtus minor in Manitoba. Journ. Mammalogy, vol. 7, no. 3, pp. 193-200. 218 NORTH AMERICAN FAUNA [No. 49 Elliot, D. G. 1907. A catalogue of the collection of mammals in the Field Columbian Museum. 694 pp., illus. Chicago. (Field Columbian Mus. Pub., no. 115, Zool. ser., vol. 8.) Fellows, J. O. 1881. Weight of woodchucks. Forest and Stream, vol. 17, p. 29. Hayden, F. V. 1862. On the geology and natural history of the Upper Missouri. Trans. Amer. Phil. Soc, vol. 12, illus. 1875. Catalogue of the collections in geology and natural history. 6. K. Warren's Preliminary report of explorations in Nebraska and Dakota, in the years 1855-56-57, pp. 61-95. (Reprint from Report of the Secretary of War, pp. 673-711, 1858.) Henet, A., and D. Thompson. 1897. New light on the early history of the greater Northwest. The manuscript journals of Alexander Henry and of David Thompson 1799-1814. Edited by E. Coues. vol. 1. New York. HOLLISTER, N. 1911. A systematic synopsis of the muskrats. North Amer. Fauna No. 32, 47 pp., illus. 1915. A systematic account of the grasshopper mice. Proc. U. S. Nat. Mus., vol. 47, pp. 427-489, illus. 1916. A systematic account of the prairie-dogs. North Amer. Fauna No. 40, 36 pp., illus. HORNADAY, W. T. 18S9. The extermination of the American bison. Ann. Rpt. Bd. Regents Smithsn. Inst. 1887, pt. 2, pp. 373-548, illus. Washington, D. C. Howell, A. H. 1915. Revision of the American marmots. North Amer. Fauna No. 37, 80 pp., illus. 1918. Revision of the American flying squirrels. North Amer. Fauna No. 44, 62 pp., illus. Jackson, H. H. T. 1908. A oreliminary list of Wisconsin mammals. Bui. Wis. Nat. Hist. Soc, vol. 6 (n. s.), pp. 13-34, illus. Johnson, G. E. 1917. The habits of the thirteen-lined ground squirrel (Citellus tridecem- lineatus), with especial reference to the burrows. Quart. Journ., Univ. North Dakota, vol. 7, pp. 261-271, illus. Kennicott, R. 1857. The quadrupeds of Illinois injurious and beneficial to the farmer. Rpt. Commr. Pat. [U. S.] 1856, pp. 52-100, illus. Lantz, D. E. 190S. Deer farming in the United States. U. S. Dept. Agr. Farmers' Bui. 330, 20 pp., illus. 1909. The brown rat in the United States. U. S. Dept. Agr., Biol. Sur- vey Bui. 33, 54 pp., illus. 1910. The muskrat. U. S. Dept. Agr. Farmers' Bui. 396, 38 pp., illus. 1914. Economic value of North American skunks. U. S. Dept. Agr. Farmers' Bui. 587, 22 pp., illus. 1916. Cottontail rabbits in relation to trees and farm crops. U. S. Dept. Agr. Farmers' Bui. 702, 12 pp., illus. 1917. The muskrat as a fur bearer, with notes on its use as food. U. S. Dept. Agr. Farmers' Bui. 869, 22 pp., illus. Le Raye, C. 1812. Journal of Mr. Charles Le Raye. J. Cutler's, A topographical de- scription of the State of Ohio, Indiana Territory, and Louisiana, comprehending the Ohio and Mississippi Rivers, and their prin- cipal tributary streams, pp. 158-204. Boston. Lewis, M., and W. Clark. 1893. History of the expedition under the command of Lewis and Clark. 1804-5-6. New ed. by E. Coues. vol. 1. New York. 1926] BIBLIOGRAPHY 219 McChesney, C. E. 187S. Notes on the mammals of Fort Sisseton, Dakota. Eul. U. S. Geol. and Geogr. Survey Ter., vol. 4, pp. 201-218. McLaughlin, J. 1910. My friend the Indian. 416 pp., illus. Boston and New York. Meebiam, C. H. 1884. The mammals of the Adirondack region, northeastern New York. 316 pp. New York. (Eeprint from Trans. Linn. Soc. New York, vol. 1, pp. 9-106, 1882; vol. 2, pp. 9-214, 1884.) 1888. Report of the ornithologist and mammalogist. Rpt. Coninir. Agr. [U. S.] 1887, pp. 399-401. 1889. Revision of the North American pocket mice. North Amer. Fauna No. 1, 34 pp., illus. 1898. Life zones and crop zones of the United States. U. S. Dept. Agr., Div. Biol. Survey Bui. 10. 79 pp., illus. 1918. Review of the grizzly and big brown bears of North America (genus Vrsus) with description of a new genus, Vetularctos. North Amer. Fauna No. 41, 136 pp., illus. Mershon, TV. B. 1923. Recollections of my fifty years hunting and fishing. 259 pp., illus. Boston. [1925]. Notes on hunting conditions in North Dakota in the '80's. Bien. Rpt. State Game and Fish Comm. North Dakota (1923-24) 8, pp. 20-22. Miller, G. S., Jr. 1897. Notes on the mammals of Ontario. Proc. Boston Soc. Nat. Hist., vol. 28, pp. 44. Nelson, E. W. 1925. Status of the pronghorned antelope, 1922-1924. U. S. Dept. Agr. Bui. 1346, 64 pp., illus. North Dakota, State of [1923]. Laws of North Dakota. 611 pp., Fargo, North Dakota. Preble, E. A. 1908. A biological investigation of the Athabaska-Mackenzie region. North Amer. Fauna No. 27, 574 pp., illus. Roosevelt, T. 1900. Hunting the grisly and other sketches. 274 pp., illus. New York and London. 1900a. Hunting trips of a ranchman. 296 pp., illus. New York and London. 1900b. Hunting trips on the prairie and in the mountains. 238 pp., illus. New York and London. 1900c. The wilderness hunter. 279 pp. New York and London. 1919. An autobiography. 647 pp., illus. New York. Ross, A. 1856. The Red River settlement. 416 pp., illus. London. Seton, E. T. 1909. Life-histories of northern animals. An" account of the mammals of Manitoba. 2 vols., illus. New York City. Smkt, P.-.7., DE 1905. Life, letters and travels of Father Pierre-Jean De Smet, S. J. Edited by H. M. Chittenden and A. T. Richardson, vol. 2, illus. New York. Thompson, D. 1916. David Thompson's narrative of his explorations in western America 1784-1812. Edited by J. B. Tyrrell. 582 pp., illus. Toronto. (Champlain. Soc. Pub., no. 12.) Thompson, E. E. (Seton, E. T.) 1886. A list of the mammals of Manitoba. Trans. Manitoba Sci. and Hist. Soc, no. 23. 26 pp.. Illus. United States Department of Agrictjlti re. Weather Bureau. 1919. Climatological data. North Dakota section. Annual summary, 1918. pp. 99-1 OS, illus. Bismarck, North Dakota. 82242°— 26 15 220 NORTH AMERICAN FAUNA [No. 49 Warren, G. K. 1856. Explorations in the Dacota country, in the year 1855. U. S. Con- gress, 34th, 1st Sess., Senate Doc. 76, 79 pp., illus. Wied [-Nexjwied, M. A. P.], Prinz zu. 1839. tJber einige Nager mit ausseren Backentaschen aus dem westlichen Nord-America. Nova Acta Acad. Caes. Leop.-Carol. Nat. Cur., t. 19, pt. 1, pp. 367-384, 1839. 1839-41. Reise in das Innere Nord-America in den Jahren 1832 bis 1834. Bd. 1, illus., 1839; Bd. 2, illus., 1841; folio. Coblenz. 1843. Travels in the interior of North America, 1832-1834. Part II. London ed. (Reprint in R. G. Thwaites's Early Western Travels, 1748-1846, vol. 23. Cleveland. 1906.) Wilcox, A. H. 1907. A pioneer history of Becker County, Minnesota. 757 pp., illus. St. Paul. Will, G. F., and G. E. Hyde. 1917. Corn among the Indians of the Upper Missouri. 323 pp., illus. St. Louis. INDEX Abronia niicrantha, 10. Abronia, winged, 10. Acer rubrum, 13. saccharum, 13. Achillea lanulosa, 13. Alces americanus americanus, 31-32. Alder, 7, 13. Alnus incana, 13. Alum root, 13. Amelanchier canadensis oblongifolia, 16. Ammospiza caudacuta nelsoni, 12. Amorpha canescens, 13. Andropogon furcatus, 13. scoparius, 13. Anemone cylindrica, 13. Antelope, American, 27. pronghorned, 27-31. Anthus spragueii, 13. Antilocapra americana americana, 27- 31. Antilocapridae, 27. Antilope americanus, 27. Apios tuberosa, 97. Aragallus lambertii, 14. Arctomys franklinii, 55. ludoviciana, 62. richardsonii, 58. Arctostaphylos uva-ursi, 13. Aristonetta valisineria, 12. Artemisia cana, 10. caudata, 13. frigida, 13, 14, 101. tridentata, 13. Artiodactyla, 19. Arvieola austerus minor, 99. drummondii, 92. haydenii, 98. pallidus, 101. Ash, 7, 16. green, 13. white, 7, 13. Aspen, 7, 16. Aster, golden, 14. Astragalus, 122. crassicarpus, 14. gracilis, 10. pictus, 10. Atriplex canescens, 10. confertifolia, 10. nuttallii, 10. Avens, white-flowered, 13. Avocet, 13. Badger, 184-187. Bartramia longicauda, 12. Basswood, 7, 13. Bat, black, 212. great gray, 207. hoary, 207-209. large brown. 12, 210-212. little brown, 12, 213-214. little long-eared, 216-217. New York, 209. red, 209-210. Say, 12, 216-217. silver-haired, 15, 212-213. Bat — Continued. silvery, 212. Yellowstone, 12, 215-216. Bear, black, 191-193. cinnamon, 191. grizzly, 193-198. Bearberry, 13. Beardgrass, broom, 13. forked, 13. Beardtongue, large-flowered, 10. pale, 14. slender, 13. Beaver, Canada, 105-10S. Missouri River, 108-114. Bee plant, 10. Betula fontinalis, 13. papyrifera, 16. pumila glandulifera, 16. Birch, shrubby, 16. western, 7, 13, 16. white, 16. Bison, American, 19-25. Bison bison bison, 19-25. Bittersweet, 13. Black-cat, 177. Black haw, 115. Black-eyed susan, 13. Black-footed ferret, 171-173. Blarina brevicauda brevicauda, 12, 205-207. Blazing star, 13, 101. Blue-stem, big, 10. little, 13. Bobcat, eastern, 149-150. mountain, 148. northern, 12, 148-149. Bobolink, 12. Bos bison, 19. Bouteloua gracilis, 10. oiigostachya, 14. Bovidae, 19. Boxelder, 7, 16. Brauneria angustifolia, 14. Brown-eyed susan. 14. Buffalo, American, 19. berry, 7, 13, 16, 115. bur, 10. grass, false, 10. pea, 14. Bulbilis dactyloides, 14. Bunchberry, 16. Bunting, indigo, 9. lazuli, 9. Buteo ferrugineus, 12. platypterus platypterus, 12. Cactus, prickly-pear, 10. slender, 10. Calcarius ornatus, 12. Campanula rotundifolia, 13. Canada lynx, 146-148. Canidae, 150. Canis latrans latrans, 12, 156-157. latrans nebracensis, 12, 157-160. mexicanus nubilus, 150-156. nubilus, 150. pallidus, 157. 221 222 NORTH AMERICAN" FAUNA [No. 49 Carex festucacea, 13. Cariacus virgultus, 41. Caribou, woodland, 32-33. Carnivora, 144. Castilleja sessiliflora, 14. Castor canadensis canadensis, 105- 108. canadensis niissouriensis, 108-114. Castoridae, 105. Celastrus scandens, 13. Celtis oceidentalis, 13. Centrocercus urophasianus, 13. Centronyx bairdii, 12. Cervidae, 31. Cervus canadensis canadensis, 33-36. elaphus canadensis, 33. hemionus, 41. tarandus, 32. Chat, long-tailed, 9. Cherry, choke, 7, 16, 115. pin, 7, 16. red, 16. sand, 10. Chickadee, long-tailed, 9. Chickaree, 46. Chilotus pallidus, 101. Chipmunk, Badlands, 9. gray, 12, 51-52. little northern, 47-49. northern, 15. pale, 49-51. Chiroptera, 207. Chokecherry, 7, 16, 115. Chondestes grammacus grammacus, 9. grammacus strigatus. 9. Chroicocephalus pepixcan, 12. Chrysopsis villosa, 14. Chrysothaninus graveolens, 10. Cinquefoil, shrubby, 7, 13. Citellus franklinii, 12, 55-58. richardsoni, 11, 58-62. tridecemlineatus pallidus, 9, 54-55. tridecemlineatus tridecemlineatus. tus, 12, 52-54. Cleome serrulata, 10. Clover, prairie, 132. silver, 14. Colaptes auratus luteus, 12. cafer collaris, 13. Condylura cristata, 200. Coon, 187. Cornel, 7, 13, 16. Cornus canadensis, 16. femina, 13. Corvus macrourus, 36. Corylus americana, 13. rostrata, 16. Cottontail, Black Hills, 9, 137. Nebraska, 134-137. Wyoming, 137-138. Cottonwood, 7. broad-leaved, 7, 9, 10. Cougar, 144. Coyote, 12. northern, 156-157. Plains, 12, 157-160. Crataegus chrysocarpa, 13. succulenta, 13. Cricetidae, 73. Currant, black, 13. flowering, 7, 10. red, 16. wild, 16. Cyanocitta cristata bromia, 12. Cynomys ludovicianus ludovicianus, 9, 62-67. Dalea, 10. Deer, mule, 41—13. Plains white-tailed, 36-41. Dickcissel, 9. Dipodops richardsoni, 124. Dogwood, shrubby, 7. Dolichonyx oryzivorus, 12 Drymocallis arguta, 13. Duck, canvasback, 12. redhead, 12. ring-necked. 12. Elaeagnus argentea, 13. Elk, American, 33-36. Elm, 7, 16. American, 7. red, 7, 13. white, 13. Eptesicus fuscus fuscus, 12, 210-212. fuscus pallidus, 210. Erethizon dorsatum dorsatum. 116-117. epixanthum epixanthum, 114-116. epixanthus, 114. Erethizontidae, 114. Ermine, 166. Erysimum asperum, 14. Euphorbia marginata, 10. Eurotia lanata. 10, 101. Eutamias borealis, 15. minimus borealis, 47-49. minimus pallidus, 49-51. pallidus pallidus, 9. Evening primrose, 13. low, 10. Evotomys gapperi loringi, 11, 88-92. Fabia minor equina, 97. Falcata comosa, 96. False indigo, gray, 13. lupine, 13. Felidae, 144. Felis hippolestes, 144-146. rufa, 149. Ferret, black-footed, 9, 171-173. Fiber zibethicus cinnamominus, 102- 105. Fisher, 177-178. Flax, yellow, 14. Flicker, red-shafted, 13. yellow-shafted, 12. Flickertail, 58. Flying squirrel, pale, 43-45. Fox, gray, 166. kit, 12, 163-166. prairie, 163. swift, 12. yellow-red, 12, 160-163. Fraxinus lanceolata, 13. pennsylvanica, 13. 1926] INDEX 223 Gaillardia lanceolata, 14. Gaura coccinea,14. scarlet, 14. Gentiana puberula, 13. Gentian, closed, 13. Geomyidae, 125. Geomys bur.sarius, 12, 125-130. Glaucomys sabrinus canescens, 43-45. volans volans, 43. Glis canadensis, 69. Glutton, 178. Glyeyrrhiza, lepidota, 132. Godwit, marbled, 12. Gooseberry, 16. smooth, 13. Grass, buffalo, 14. dropseed, 10. grama, 10. 14. Indian, 13. needle, 122. porcupine, 13. Grasshopper mouse, Audubon, 83-85. Maximilian, 81-83. Greasewood, 10. Grizzly, Absaroka. 198-199. big Plains, 193. Grosbeak, black-headed, 13. rose-breasted, 12. Ground cherry, 13. Groundhog, 67, 69. Groundsel, 14. Ground squirrel, Franklin, 55. gray, 12, 55-58. pale striped, 54-55. pale thirteen-lined. 9, 54. Richardson, 11. 58-62. striped, 52-54. thirteen-lined. 12. Grouse, sage, 13. Gull, Franklin, 12. Gulo luscus, 178-179. Hackberry, 13. Hare, varying, 15, 138-141. Harebell, 13. Harvest mouse, 9. prairie, 9, 80-81. Haw, black, 13. Hawk, broad-winged, 12. ferruginous roughleg, 12. Hawthorn, 13. Hazel, 7, 13. beaked, 16. Hedymeles ludovicianus, 12. melanocephalus papago, 13. Helianthu.s annuus. 10. maximilianus, 13. tuberosa, 96. Hesperomys leucopus nebrascensis, 73. Heteromyidae, 119. Hetunka. 94. Heuchera hispida, 13. High-bush cranberry, 16. Honeysuckle, 13. Hosackia americana, 10. Hypudaeus leueogaster, 81. Hystrix dorsata, 116. Icteria virens longicauda, 9. Icterus bullockii bullockii, 9. galbula, 12. Indian devil, 178. names, 18. Insectivora, 199. Ironweed, 13. Ironwood, 7, 13. Jack rabbit, white-tailed, 11, 141-144. Jay, blue, 12. Jumping mouse, prairie, 11, 117-119. Junco hyemalis, 15. Junco, slate-colored, 15. Juniper, creeping, 13. Rocky Mountain, 7, 13. shrubby, 16. Juniperus horizontalis, 13. scopulorum, 13. Kangaroo rat. Richardson. 9. 124-125. Lacinaria pychnostachya, 13. scariosa, 13. Lagomorpha, 134. Lasionycteris noctivagans, 15. 212-213. Lead-plant, 13. Lepargyrea argentea, 13. canadensis, 16. Leporidae. 134. Lepus americanus, 15. americanus americanus, 138-141. americanus phaeonotus, 139. baileyi, 137. bishopi, 138. sylvaticus grangeri, 137. townsendii campanius, 11, 141-144. Leuconoe carissima, 215. Liatris graminifolia, 101. Licorice, wild, 132. Limosa fedoa, 12. Linum rigidum, 14. Lithospermum linearifolium, 14. Loafer, 150. Lobelia, 13. Lobelia spicata, 13. Lobo, 150. Loco plant, 14. Longspur, chestnut-collared. 12. McCown. 13. Lonicera dioica glaucescens, 13. Lupine, dwarf, 10. Lupinus pusillus, 10. Lutra canadensis canadensis. 179-181. Lutreola vison letifera, 12, 173-176. Lygodesmia juncea, 14. Lynx, Canada, 15. Lynx, canadensis, 15. canadensis canadensis, 146-148. rufus rufus, 149-150. uinta, 12, 148-149. Magpie, 13. Mallow, orange-red false, 14. Malvastrum coccineum, 14. Maple, red, 13. sugar, 13. Marigold, prairie, 14. Marmota monax canadensis, 69-70. monax rufescens, 12, 67-69. 224 NORTH AMERICAN FAUNA [No. 49 Marten, 176-177. pine, 176. Martes americana americana, 176-177. pennanti pennanti, 177-178. Meadow mouse, Drummond, 12, 02-94. eastern, 12, 90-92. Melospiza georgiana, 12. Mentzelia, 10. Mentzelia decapetala, 10. Mephitis americana hndsonica, 181. hudsonica, 181-184. hudsonica hudsonica, 12. Meriolix serrulata, 13. Microsorex hoyi hoyi, 204-205. Microtus minor, 99-100. minor minor, 12. oehrogaster haydenii, 9, 9S-9D. pallidus, 12, 101-102. pennsylvanicus drummondi, 12, 92- 94. pennsylvanicus pennsylvanicus, 12, 90-92. pennsylvanicus wahema, 94-98. Milk-vetch, painted, 10. slender, 10. Mink, 173-176. Minnesota, 12. Mitella nuda, 16. Miterwort, 16. Mole, Missouri Valley, 199-200. star-nosed, 200. Moose, 31-32. Mountain lion, 144-146. Mountain sheep, Audubon, 25-27. Badlands, 9. Mourning dove, western, 9. Mouse, bean, 94-98. deer, 77. house, 72-73. pale, 12, 101-102. red-backed, 88-92. Munroa squarrosa, 10. Muridae. 70. Mus bairdii, 76. bursarius, 125. missouriensis, 83. musculus musculus, 72-73. norvegicus, 70. pennsylvanica, 90. sylvaticus noveboracensis. 77. Muskrat, Great Plains, 102-105. Mustard, treacle, 14. Mustela ainericanus, 176. cicognanii cicognanii, 12, 169-170. cigognanii, 169. longicauda, 12. longicauda longicauda, 166-168. lutra canadensis, 179. nigripes, 9, 171-173. pennanti, 177. rixosa rixosa, 170-171. vison letifera, 173. Mustelidae, 166. Myotis carissima, 215. evotis, 216-217. lucifugus carissima, 12, 215-216. lucifugus lucifugus, 12, 213-214. subulatus subulatus, 12, 216-217. Nanny-berry, 13. Nebraska cottontail, 12. Neosorex palustris, 203-204. Neotoma cinerea rupicola, 9, 86-88. rupicola, 86. Nycteris borealis borealis, 209-210. cinerea, 207-209. Nyroea americana, 12. Oak, 16. bur, 7, 13. Odocoileus hemionus hemionus, 41^13. virginianus macrourus, 36-41. Onychomys leucogaster leucogaster, 81-83. leucogaster missouriensis, 83-S5. Opuntia fragilis, 10. polyacantha, 10. Oriole, Baltimore, 12. Bullock, 9. Orthocarpus luteus, 14. Oryzopsis cuspidata, 10. micrantha, 10. Ostrya virginiana, 13. Otter, 179-181. Ovis canadensis auduboni, 9, 25-27. Owl, burrowing, 9. Pachylophus caespitosus, 10. Paintbrush, 14. Indian, 14. yellow Indian, 14. Panther, 144. Parosela enneandra, 10. Passerina amoena, 9. cyanea, 9. Pedomys haydenii, 98. Pekan, 177. Penthestes atricapillus septentrionalis, 9. Pentstemon albidus, 14. gracilis, 13. grandiflorus, 10. Perissonetta collaris, 12. Perodipus montanus richardsoni, 9, 124-125. Perognathus fasciatus fasciatus, 9, 119-121. flavescens perniger, 9, 121-123. hispidus paradoxus, 9, 123-124. Peromyscus leucopus aridulus, 9. 79- 80. leucopus noveboracensis, 12, 77- 78. maniculatus bairdii, 12, 76-77. maniculatus osgoodi, 9, 73-76. Phalaenoptilus nuttallii nuttallii, 9. Phalarope, Wilson, 12. Phoebe, Say, 9. Physalis lanceolata, 13. Pica pica hudsonia, 13. Pine, Rocky Mountain yellow, 7. yellow, 13. Pinus scopulorum, 13. Pipilo erythrophthalmus erythroph- thalmus, 12. maculatus arcticus, 13. Pipit, Sprague, 13. 1926] INDEX 225 Plantago purshii, 10. Plantain, Indian, 10. Plover, upland, 12. Plum, 16. wild, 13. Pocket gopher, Dakota, 12, 130-133. Mississippi Valley, 12, 125-130. sagebrush, 9, 133-137. Pocket mouse, dusky, 9, 121-123. Kansas, 9, 123-124. Maximilian, 9, 119-121. Pooecetes gramineus confinis, 13. gramineus gramineus, 12. Poor-will, 9. Poplar, aspen, 16. balsam, 7, 16. Populus balsamifera, 16. deltoides, 10. tremuloides, 16. Porcupine, black-haired, 116-117. Canada, 116. Rocky Mountain, 114. yellow-haired, 114-116. Potentilla fruticosa. 13. Prairie dog, black-tailed, 9, 62-67. Prairie-clover, 10. Prickly ash, 13. Procyon lotor lotor, 187-190. Procyonidae, 187. Pronghorn, 27. Prunus americana, 13. pennsylvanica. 16. pumila, 10. Psoralea argophylla, 14, 132. esculenta. 10, 96, 132. lanceolata, 10. tenuiflora, 10. Puccoon, narrow-leaved, 14. Putorius nigripes, 171. rixosus, 170. Pyrola asarifolia, 16. Quercus macrocarpa, 13. Rabbitberry, 16. Rabbitbrush, 10. Rabbit, snowshoe. 138. white, 138. Raccoon, 187-190. Rangifer caribou caribou, 32-33. Raspberrv, 16. red, 13. Rat, brown, 70-72. house, 70. wharf, 70. Rattus norvegicus, 70-72. Recurvirostra americana, 13. Red-backed mouse, Loring, 11. Reithrodontomys dychei, 80. megalotis dychei, 9, 80-81. Rbus glabra, 13. trilobata, 10. Rhynchophanes mccownii, 13. Ribes americanum, 13. aureum, 10. gracile, 13. triste, 16. Rodentia, 43. Rosa blanda, 13. pratincola, 13. Rose, 7, 16, 115. pale, 13. prairie, 13. Rubicola minor, 12. Rubus strigosus. 13. Rudbeckia hirta, 13. Sable, American, 176. Sage, silver, 13, 14, 101. Sagebrush, 13. gray, 10. Salix lutea, 13. missouriensis, 13. serissima, 16. Salpinctes obsoletus obsoletus, 9. Sand cherry, 9. Sarcobatus vermiculatus, 10. Sayornis sayus, 9. Scalopus aquaticus machrinoides, 199- 200. Sciuridae, 43. Sciurus carolinensis hvpophaeus, 12, 45-46. hudsonicus, 15. hudsonicus hudsonicus, 46-47. tridecemlineatus, 52. vulgaris hudsonicus, 46. Sedge, 13. Senecio plattensis, 14. Serviceberry, 7, 16. Canadian, 16. Setochalcis vocifera vocifera. 12. Shadscale, gray, 10. low, 10. Nuttall, 10. Shrew, black-backed, 203. Hay den, 12, 200-202. marsh, 203. Merriam, 9, 202. mole, 205. pigmy, 204-205. Richardson, 15, 203. saddle-backed, 203. short-tailed, 12, 205-207. water. 203-204. Silver-leaf. 13. Silvertip. 193. Skeleton plant. 14. Skunk, northern, 12, 181-184. Skunk bush, 10. Snow-on-the-mountain, 10. Solanum rostratum, 10. Sorex arcticus, 203. brevicaudus, 205. cinereus haydeni, 12, 200-202. cristatus, 200. haydeni, 200. hoyi, 204. merriami, 9, 202. palusti-is, 203. richardsoni, 15. Sorghastrum nutans, 13. Soricidae, 200. Spanish bayonet, 10. 226 NORTH AMERICAN FAUNA [No. 49, 1926] Sparrow, Baird, 12. lark, 9. Nelson, 12. swamp, 12. vesper, 12. western lark, 9. western vesper, 13. white-throated, 15. Speotyto cunicularia hypugaea, 9. Spermophilus tridecemlineatus pal- lidus, 54. Sphyrapicus varius varius, 12. Spiderwort, 13. Spiny solanum, 10. Spiza americana, 9. Squirrel, black, 45. leopard, 52. Minnesota gray, 12, 45-46. red, 15, 46-47. thirteen-lined, 52. Steganopus tricolor, 12. Sterna forsteri, 12. Stipa spartea, 13, 122. Sumac, 13. Sunflower, Maximilian, 13. wild, 10. Swift, 163. Sylvilagus audubonii baileyi, 137-138. floridanus similis, 12, 134-137. nuttalli grangeri, 9, 137. Talpidae, 199. Tamias asiaticus borealis, 47. quadrivitatus pallidus, 49. striatus griseus, 12, 51-52. Taxidea taxus taxus, 184-187. Tern, Forster, 12. Thermopsis rhombifolia, 13. Thomomys talpoides bullatus, 9, 133-137. talpoides rufescens, 12, 130-133. Thorn apple, 16. Tilia americana, 13. Tipsin, 10, 96. Towhee, 12. Arctic, 13. Tradescantia bracteata, 13. Trefoil, bird's-foot, 10. Tripterocalyx micranthus, 10. Turnip, prairie, 132. Ulmus americana, 13. fulva, 13. Upland mouse, little, 12, 99-100. western, 9, 98-99. Urocyon cinereo-argenteus, 166. Ursidae, 191. Ursus absarokus, 198-199. americanus americanus, 191-193. horribilis horribilis, 193-198. lotor, 187. luscus, 178. taxus. 184. Verbena, sand, 10. Vernonia fascicularis, 13. Vespertila fuscus, 210. Vespertilio borealis, 209. cinereus, 207. evotis, 216. lucifugus, 213. noetivagans, 212. subulatus, - 216. . ursinus, 210. Vespertilionidae, 207. Viburnum lentago, 13. opuius americana, 16. Viola nuttallii, 14. Violet, yellow, 14. Vulpes fulva regalis, 12, 160-163. regalis, 160. velox hebes, 12, 163-166. Wapiti, 33. Weasel, Bonaparte, 12, 169-170. least, 170-171. long-tailed, 12, 166-168. short-tailed, 169. Whip-poor-will, 12. White-footed mouse, Badlands, 9, 7980. Baird, 12, 76-77. northern, 12, 77-78. Osgood, 9, 73-76. Wild cat, 149. spotted, 148. Willow, 16. autumn, 16. Missouri, 13. yellow, 13. Wind-flower, cylindric, 13. Winterfat, 10, 101. Winter-lettuce, 16. Wolf, brush, 12, 156. buffalo, 150. gray, 150-156. prairie, 157. Wolverene, 178-179. Woodchuck, Canada, 69-70. rufescent, 12, 67-69. Woodcock, 12. Woodpecker, yellow-bellied, 12. Wood rat, pale bushy-tailed, 9, 86-88. Wormwood, cut-leaved, 13. Wren, rock, 9. Xanthoxylum americanum, 13. Yarrow, 13. Yellow-red fox, 160-163. Yucca glauca, 10. Zapodidae, 117. Zapus hudsonius campestris-, 11, 117-119. Zenaidura macroura marginella, 9. Zone, Canadian, 14-16. Transition, 11-14. Upper Austral, 8-11. Zonotrichia albicollis, 15. o (Continued from page 2 of cover) No. 22. A Biological Investigation of the Hudson Bay Region. By Edward A. Preble. Pp. 140, pis. 14 (incl. 1 map). 1902 [Out of print.] No. 23. Index Generum Mammalium : A List of the Genera and Families of Mammals. By T. S. Palmer. Pp.984. 1904 [Ott* of print.] No. 24. A Biological Reconnaissance of the Base of the Alaska Peninsula. By Wilfred H. Osgood. Pp. 86, pis. 7 (incl. 2 maps). 1904 [Out of print.] No. 25. Biological Survey of Texas : Life Zones, with Characteristic Species of Mammals, Birds, Reptiles, and Plants. By Vernon Bailey. Pp. 222, pis. 16 (incl. 6 maps), figs. 24 (incl. 16 maps). 1905 [Out of print.] No. 26. Revision of the Skunks of the Genus Spilogale. By Arthur H. Howell. Pp. 55, pis. 10 (incl. 1 map). 1906 [Out of print.] No. 27. A Biological Investigation of the Athabaska-Mackeuzie Region. By Edward A. Preble. Pp. 574, pis. 25 (incl. 4 maps) , figs. 16. 190S— Price, $1.25 No. 28. Revision of the Mice of the American Genus Peromyscus. By Wilfred H. Osgood. Pp. 285, pis. 8 (incl. 1 map) , figs. 12 (maps) . 1909- [Out of print.] No. 29. The Rabbits of North America. By E. W. Nelson. Pp. 314, pis. 13, figs. 19 (incl. 16 maps). 1909 [Out of print.] No. 30. Biological Investigations in Alaska and Yukon Territory : 1. East- central Alaska ; 2. Ogilvie Range, Yukon ; 3. Macmillan River, Yukon. By Wilfred H. Osgood. Pp. 96, pis. 5 (1 map), figs. 2 (maps). 1909 [Out of print.] No. 31. Revision of the Wood Rats of the Genus Neotoma. By Edward A. Goldman. Pp. 124. pis. 8, figs. 14 (maps). 1910 [Out of print.] No. 32. A Systematic Synopsis of the Muskrats. By N. Hollister. Pp. 47, pis. 6 (incl. 1 map). 1911 [Out of print.] No. 33. A Biological Survey of Colorado. By Merritt Cary. Pp. 256, pis. 12 (incl. 1 colored map), figs. 39 (incl. 29 maps). 1911 [Out of print.] No. 34. Revision of the Spiny Pocket Mice (Genera Heteromys and Liomijs). By Edward A. Goldman. Pp. 70, pis. 3. figs. 6 (maps). 1911__Pn'ce, 10 cents. No. 35. Life Zones and Crop Zones of New Mexico. By Vernon Bailey. Pp. 100. pis. 16 (incl. 1 colored map), figs. 6. 1913 Price, 45 cents. No. 36. Revision of the American Harvest Mice (Genus Reithrodontomys) . By Arthur H. Howell. Pp. 97, pis. 7, figs. 6 (maps). 1914 Price, 20 cents. No. 37. Revision of the American Marmots. By Arthur H. Howell. Pp. 80, pis. 15, figs. 3 (maps). 1915 Price 20 cents. No. 38. A Review of the American Moles. By Hartley H. T. Jackson. Pp. 100, pis. 6, figs. 27 (incl. 6 maps). 1915 Price, 20 cents. No. 39. Revision of the Pocket Gophers of the Genus Thomomys. By Vernon Bailey. Pp. 136. pis. 8, figs. 10 (Incl. 7 maps). 1915 Price, 25 cents. No. 40. A Systematic Account of the Prairie-dogs. By N. Hollister. Pp. 37, pis. 7, figs. 2 (maps). 1916 [Out of print.] No. 41. Review of the Grizzly and Big Brown Bears of North America (Genus Ursus), with Description of a New Genus, Vetularctos. By C. Hart Mer- riam. Pp. 136, pis. 16. 1918 Price, 35 cents. No. 42. Life Zone Investigations in Wyoming. By Merritt Cary. Pp. 95, pis. 15 (incl. 1 colored map), figs. 17 (incl. 1 map). 1017 Price, J(0 cents. No. 43. The Rice Rats of North America (Genus Oryzomys.) By Edward A. Goldman. Pp. 100, pis. 6, figs. 11 (incl. 10 maps). 1018 Price, 20 cents. No. 44. Revision of the American Flying Squirrels. By Arthur H. Howell. Pp. 84, pis. 7, figs. 4 (maps). 1918 Price, 15 cents. No. 45. A Biological Survey of Alabama : Tart I, Physiography and Life Zones; Part II. The Mammals. By Arthur H. Howell. Pp. 88, pis. 11 (incl. 1 colored map), figs. 10 (maps). 1921 Price, 25 cents. No. 46. A Biological Survey of the Pribllof Islands, Alaska : I, Birds and Mammals, by Edward A. Preble and W. L. McAtee; II, Insects, Arachnids, and Chilopods, by various entomologists, with an introduction by W. L. McAtee. Pp. 255, pis. 15 (incl. 3 maps). 1923 Price, J^O cents. No. 47. Revision of the American Pikas (Genus Ochotona). By Arthur 11. Howell. Pp. 57, pis. 6, figs. 4 (maps). 1924 Price, 15 cents. No. 48. Voles of the Genus Phenacomys: I, Revision of the Genus Phenacomys; II, Life History of the Red Tree Mouse Phenacomys lonrjwaudus. By A. Brazier Howell. Pp. 66. pis. 9. figs. 12 (incl. 5 maps). 1926__Prtce, 20 cents. No. 49. A Biological Survey of North Dakota : I, Physiography and Life Zones; II, The Mammals. By Vernon Bailey. Pp. 232, pis. 21 (incl. 1 colored man), figs. 8 (maps). 1926 Price, 60 cents. U. S. DEPARTMENT OF AGRICULTURE BUREAU OF BIOLOGICAL SURVEY NORTH AMERICAN FAUNA ISTo. oO [Actual date of publication, June 30, 1927] REVISION OF THE AMERICAN LEMMING MICE (Genus SYNAPTOMYS) BY A. BRAZIER HOWELL SCIENTIFIC ASSISTANT, DIVISION OF BIOLOGICAL INVESTIGATIONS BUREAU OF BIOLOGICAL SURVEY J UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON 1927 CONTENTS Page Introduction 1 Geographic distribution of Synaptomys 1 Habits : 1 Material examined and desired 3 Explanation of measurements and color terms 3 Genus Synaptomys Baird 4 History „ 4 Probable relationships 5 Characters :_ 6 Variation • 9 Key to the subgenera of Synaptomys 9 List of species and subspecies with their type localities 9 Subgenus Synaptomys Baird 9 Key to subspecies 12 Subgenus Mictomys True «. 19 Key to subspecies 21 Bibliography of Synaptomys 31 Index 3S ILLUSTRATIONS Pl. 1. Skulls (dorsal views) of subgenus Synaptomys. Synaptomys cooperi cooperi, S. c. stonei, S. c. helaletes, S. c. gossii 30 2. Skulls (dorsal views) of subgenus Mictomys. Synaptomys borealte innuitus, S. b. sphagnicola, S. b. chapmani, S. b. icrangeli 31 T1;XT FIGURES Fig. 1. Map showing approximate distribution, of^. genus Synaptomys — 2 2. Map showing distribution oi sul>genus; Synaptomys 10 3. Palate of subgenus Synaptomys ' 11 4. Enamel pattern ol aiolars of subgenus Synaprornys. 11 5. Incisors of sulv-.enas fcyiiaptoiny'^ — ^1j:_^._^ — 11 6. Map showing approximate distribution of subgenus Mictomys— 20 7. Palate of subgenus Mictomys 20 8. Enamel pattern of molars of subgenus Mictomys 21 9. Incisors of subgenus Mictomys 21 10. Map showing distribution of western forms of subgenus Mic- tomys 23 11. Map showing distribution of eastern forms of subgenus Mic- tomys 28 No. 50 NORTH AMERICAN FAUNA June 30, 1927 REVISION OF THE AMERICAN LEMMING MICE (Genus Synaptomys) By A. Bbaziek Howell INTRODUCTION The genus Synaptomys, comprising those rodents generally known as lemming mice, is confined to North America and is most closely related to the true lemmings, genus Lemmus of the section Lemmi, subfamily Microtinae. The present revision, based chiefly upon the material in the Biological Survey collection, was undertaken early in 1923. Care has been taken to present all available information that should prove of real help to students, without the inclusion of minute descriptions and unnecessary details. GEOGRAPHIC DISTRIBUTION OF SYNAPTOMYS The genus Synaptomys is locally distributed throughout most of the Hudsonian and Canadian Zones of Canada; it enters the United States in the extreme northwestern portion, and eastward is found sparingly throughout the Canadian and in the humid sections of the Transition and certain parts of the Upper Austral Zones, south into Arkansas, Tennessee, and North Carolina. (Fig. 1.) HABITS Synaptomys is not common in collections, but it is by no means certain that it is not more numerous in nature than is generally sup- posed. Except in a very few places, or in years of unusual abun- dance, lemming mice have proved exceedingly difficult to obtain in numbers. Because they are usually confined to bogs and tracts of swampy land, they are rarely if ever of economic importance. The habits of lemming mice are almost unknown. The members of the genus Synaptomys belong at the present day definitely to a boreal fauna, and in the north, although usually found in moist sit- uations, they also occur in dry patches of grass and other low cover, as well as in, bogs. In the districts farther south, however, they have evidently been able to survive only because of the presence or occa- sional cold sphagnum bogs, to which they are almost entirely con- fined in the lower latitudes of the Eastern States. Near Lake Drummond, Va., however, and at Horseshoe Lake, Mo., in Indiana, l NORTH AMERICAN FAUNA [No. 50 and at several other places the genus has been found in grassland, both moist and dry. Observations on food habits have been confined practically to the recording of the presence of cut green grasses in the runways, but judging from the habits of related rodents, these animals may occa- sionally feed upon a variety of bulbs and even insects, as well as succulent herbage. Examination in the Biological Survey of 11 stomachs from Kansas and 1 from Minnesota also showed contents of finely ground grass and sedge leaves and a few insignificant traces of other green vegetation. Further observations on the food habits of Synaptomys are greatly needed. Fig. 1. — Approximate distribution of the genus Synaptomys Well-defined runways are maintained, and burrows are constructed in the ground or through beds of sphagnum. Nests are occasionally placed in tussocks of grass or amid other surface cover, according to published reports (Hahn, 1909a, p. 570). * Collectors, mostly those of the Bureau of Biological Survey, have trapped females containing from four to six embryos, from March 11 to October 7. This indicates that litters are only of moderate size. Probably several litters are borne each year, the period of greatest reproductive activity being largely confined to the warmer months. 1 Literature citations in parentheses refer to the bibliography beginning on page 31. 1927] REVISION OF THE AMERICAN LEMMING MICE 6 MATERIAL EXAMINED AND DESIRED The present revision is based on a study of 593 specimens, chiefly in the United States National Museum, in which are all the speci- mens listed for the various groups, unless otherwise stated.2 In the case of most of the races here recognized, the material is adequate, but some of the type specimens present problems. Of the 15 forms that have been named, the types of 7 are markedly unsatisfactory, being skulls without skins or skins with fragmentary skulls, and the types of 3 others are immature. In addition, topotypes of several have never been collected. Need for additional material. — Lemming mice are so rarely ob- tained that collectors are likely to seek them whenever possible. Further material from certain critical localities is badly needed, however, and it is hoped that field parties will make special efforts to procure such desiderata. Until more specimens are obtained fur- ther progress in the proper understanding of the relationships of several races can hardly be expected. Specimens of the subgenus Synaptomys are most needed from the area immediately north and west of the Great Lakes, southwestern South Dakota, Arkansas, and all the New England States. In addi- tion, more material illustrating various degrees of intergradation between Synaptomys c. cooperi and S. c. stonei is much to be desired. Specimens of the subgenus Mictomys are greatly needed from the vicinity of Fort Chimo, Ungava, and the coast of southern Labrador, and nothing is known with certainty regarding the subgenus in the interior of eastern Canada as far west as Lake Winnipeg. Speci- mens from immediately south and west of Hudson Bay might prove exceedingly important. Because of excessive cranial variation among the few adult topo- types of Synaptomys b. borealis, more material should be had from the vicinity of Fort Franklin, Mackenzie, as well as from one or two points between that place and the Yukon boundary. Topotypes of S. b. dallij from Nulato, Alaska, and from the lower stretches of the Yukon River are greatly desired. Special search should be made for the genus in southwestern British Columbia and extreme north- western Washington, and intensive trapping in the sphagnum bogs at the base of the Olympic Peninsula of the State of Washington is desirable. EXPLANATIONS OF MEASUREMENTS AND COLOR TERMS All measurements given are in millimeters. With the exception of the type of Synaptomys b. inmiitus. all external measurements listed were taken by collectors in the field from specimens in the 2 For unrestricted access to the collections under his charge, the writer is much indebted to G. S. Miller, jr., of the U. S. National Museum. For the loan of pertinent material thanks are tendered the Museum of Vertebrate Zoology (18 specimens), through Joseph Grinncll ; the American Museum of Natural History (103 specimens), through H. E. Anthony; tlie Field Museum of Natural History (28 specimens), through W. H. Osgood; the Museum of Comparative Zoology (23 specimens), through G. M. Allen; the Royal Ontario Museum of Zoology (2 specimens), through John R. Dymond and L. L. Snyder; the National Museum of Canada (~t specimens), through R. M. Anderson; the Provincial Museum (1 specimen), through Francis Kermode ; the Academy of Natural Sciences of Philadelphia (26 specimens), through Witmer Stone; Washington State College (12 specimens), through W. T. Shaw; G. L. Kirk, Rutland, Vt. (1 specimen); W. E. Saunders, London, Ontario (10 specimens) ; Charles F. Batchelder, Cambridge, Mass. (7 specimens) ; and Donald R. Dickey, Pasadena, Calif. (8 specimens). 4 NORTH AMERICAN FAUNA [No. 50 flesh. For the reason that there is usually slight difference in ex- ternal size (and such as is indicated by the figures is often not substantiated upon direct comparison of the specimens), the impor- tance of field measurement is not stressed. The following measurements of adult crania were taken by the writer with vernier calipers registering in tenths of a millimeter : G ondylobasilar length. — Distance from the posterior border of the condyle to the posterior edge of the alveolus of the incisor on the corresponding side of the skull. Rostral length. — Distance measured from a line passing through the anteriormost portions of the zygomatic process of the maxillae to the anterior tip of the nasals. Rostral breadth. — Greatest breadth of rostrum measured across anterior borders of the anteorbital foramina. Interorbital breadth. — Shortest distance between the orbits. Zygomatic breadth. — Distance across the zygomatic arches at their widest point. Lamibdoidal width. — Greatest width of posterior portion of cra- nium measured across the lambdoidal ridges. Incisive foramina. — Greatest length of incisive foramina. Height. — Height of skull measured from a plane passing from the inferior surface of the bullae to the crown of the third molar. Color terms. — Because of the mixture of shades to be found in any one specimen of lemming mouse, such color names as are applied, including those of Ridgway,3 must be considered only approximately correct. For the same reason minute color descriptions of the dif- ferent races are not included in the present work. Genus SYNAPTOMYS Baird Synaptomys Baird, Mamm. North Amer., p. 558, 1857. Type, Synaptomys cooperi Baird. HISTORY The genus Synaptomys was described and named by Baird in 1857, but material accumulated very slowly, and for many years each instance of the discovery of the genus in a new locality was recorded over and over again. The first species of lemming mouse described was borealis, by Richardson in 1828, but it was referred by him to the genus Arvicola [=Microtus] , and for a long time its proper relationship was a puzzle to systematists. In fact, it was not until 79 years after this specimen was named that the fragments of the skull were removed from the skin and the generic affinity defi- nitely established. When Baird named Synaptomys he also described the species cooperi, which he made the generic type, but unfortunately his single imperfect specimen was from an unknown locality. The race gossii was designated by Coues in 1877, but without a description, and only a list of measurements saved the name from becoming a nomen nudum. It was not until 19 years later that a description of this Kansas race was published, by Merriam, who selected a skull without skin as the type, none having been designated by Coues. 3 Ridgway, R. color standards and color nomenclature. 43 p., illus. Washington, D. C. 1912. 1927] REVISION" OF THE AMERICAN LEMMING MICE 5 Rhoads described stonei from New Jersey in 1893; and in the fol- lowing year True named innuitus from Fort Chimo, Ungava, the type being an alcoholic specimen. Merriam named helaletes, datti, " truei" and wrangeli in 1896, only the first one of which was de- scribed from a first-class adult specimen; in the same year Bangs characterized " fatuus." S< sphagnicola Preble, from New Hamp- shire, appeared in 1899, and medioximiis Bangs, from Labrador, a year later. Preble described " bulletins " from the Mackenzie region in 1902, but the determination of the generic identity of Richard- son's type of borealis, confirmed by the capture of a series of speci- mens at the type locality by Preble in 1903, relegated this name to synonymy. Finally, " andersoni " and chapma?iiy both from British Columbia, were described by Allen in 1903. True in 1894 proposed Mictomys as a full genus, but the degree to which it differs from Synaptomys is certainly not greater than subgeneric, judged by present standards. Two years later Merriam (1896) revised the genus, recognizing eight of the nine forms that had at that time been named, but, with few exceptions, his material was scanty and unsatisfactory. For other history of the subject perhaps mention need here be made only of Genera and Sub- genera of Voles and Lemmings, by Miller (1896). The last new race of this genus was proposed just 20 years before the present study was undertaken, and the material subsequently accumulated has not justified the naming of another. Of the 15 races thus far named — all but one of them as full species — 11 are now recognized as valid under the two species, S. borealis and S. cooperi, the remaining four (" truei" " fatuus" " bullatus" and " cmdersoni " ) being placed in synonymy. The subgenus Synaptomys here comprises the races of S. borealis and the subgenus Mictomys the races of S. cooperi. PROBABLE RELATIONSHIPS No fossil remains of the genus Synaptomys have been found either in this country or abroad, as far as known, so it is of little use to speculate as to where it originated. As it is now confined to the New World, it is at least fairly certain that it is a product of North America, but all the evidence does not necessarily corroborate such a hypothesis. Late in Tertiary and possibly even early in Quar- tenary times the lemmings are believed not to have been so strictly confined to a boreal habitat as at present. In those days, before the appearance of many of the higher and more adaptable species of the Microtinae (at least in western Europe and possibly in America as well), they encountered milder competition, but at present they have mostly retired into habitats to which Microtus is not so partial. That Synaptomys as a genus, but not necessarily as individuals, is adaptable in some respects is indicated by the considerable variety of habitats in which it occurs throughout its range. The evidence at hand, however, indicates that it has a marked intolerance for the immediate presence of other genera of Microtinae, although, of course, it can seldom or never procure for itself complete isolation in this respect. In the warmer parts of its range its presence may be expected only where there is an occasional cold sphagnum bog, 6 NORTH AMERICAN FAUNA [No. 50 now small and dwindling, to which it may be confined throughout a considerable stretch of territory. Externally Synaptomys is certainly much less specialized than either of the other two American genera of lemmings (Lemmus and Dicrostonyx), and in this respect the genus is more primitive. The hypsodont condition of the molars is an advanced modification, but its exact significance is difficult to interpret, for it is impossible now to be sure whether such tooth development in Synaptomys, ,and its lack in Phenacomys, constitutes a case of a single advanced character forging ahead of others in the first instance, or of a primitive character lagging behind in the second. The equal thick- ness of the enamel on the concave and the convex surfaces of the prisms is somewhat primitive, as has often been considered the presence of grooves on the upper incisors. The actual enamel pat- tern is considered most generalized in Lemmus, less so in the sub- genus Synaptomys, and more specialized in Mictomys, because in the lower molars of the last named the buccal triangles, undoubtedly present in some ancestral form, have entirely disappeared. The crania of the less specialized forms of the genus Synaptomys are also more primitive in a number of respects than are those of the other two lemmings, which are of an incipient fossorial type, a development not shared by Synaptomys. In summary, judging as well as possible without the aid of fossil material by which decision might be further influenced, the evidence seems to indicate that on the whole Synaptomys is the most primitive of the North American members of the supergeneric group Lemmi, and although the facts are susceptible of different interpre- tations, according to the relative importance accorded the several indices of evolutional modification, it is evidently more primitive in some ways than many of the lower Microti, such as Phenacomys, in spite of its having a type of molariform dentition of a higher order than occurs in the latter. CHARACTERS Generic diagnosis. — The characters whereby the genus Synaptomys may be most readily distinguished from all other Microtinae are its grooved upper incisors and the extreme shortness of the rostrum (about 25 per cent of the total length of the skull). The molars are rootless, with persistent pulp, as in Microtus, and the mandibular incisors have their roots on the lingual side of the molars. The re- entrant angles are excessively deep on the buccal side of the max- illary molars, and on the lingual side of the mandibular ones, and the third upper molar consists of four simple, transverse loops, peculiarly arranged. The tail is slightly longer than the hind foot. The feet are not especially modified, but the nail on the first digit of the forefoot is flat and strap shaped. The hind foot has six tubercles on the sole. External characters. — Externally there is little or nothing by which the genus may invariably be distinguished, but the shortness of the tail is sufficient to separate it from all but a very few species of American Microtus. In appearance its members are robust, with 1927] REVISION OF THE AMERICAN LEMMING MICE 7 massive heads, in this respect somewhat resembling the other lem- mings, but in the main the superficial appearance is that of voles and meadow mice. Old males may be distinguished from females and from other genera by the color of the hairs overlying the hip glands. These are apparently absent in females, but are well developed in fully mature males. In the latter, some of the hairs growing from the center of the glands are a dingy white, these usually, though not in- variably, extending to the surface of the pelage. These white patches are better developed in the subgenus Mictomys than in Synaptomys. The dorsal coloration does not serve to distinguish the present genus from Microtus, but in several races it averages somewhat richer and more chestnut than the usual meadow mouse, with coarser and somewhat longer coat. Sometimes the anterior portion of the body is slightly grayer than the posterior. The ventral surface is some shade of plumbeous, the exact tint depending upon the degree to which the lighter tips of the hairs have been abraded. In examples in new pelage, the underparts may have a faint suggestion of cream- colored overtone. The feet are dark, the precise color being largety influenced by the quantity of grease present in the dried skin. The very short tail is indistinctly bicolor and for a microtine is moderately well haired. Alcoholics available in such condition that the plantar tubercles can be clearly distinguished number 11 and consist of 5 specimens of stonei, 1 of cooperi, 1 of innuitus, and 4 of gossii. All have six well-developed tubercles, the three anterior ones of which — especially the outer — are unusually large. There are said to be eight mammae in Mictomys and only six in true Synaptomys, but the full number are not always functional in young mothers with small litters. The winter coat differs from the summer one only in that it is longer and denser, and newly acquired pelage is usually, but not always, rich in coloration. Wear is manifested by the abrasion or breaking off of the tips of the hairs, resulting in the proximal colora- tion showing through to a greater extent. The brighter or richer tips to the new hairs of the adults are largely lacking in very young animals, and hence the color of the young averages darker and more plumbeous. Unfortunately, both winter and summer pelages from any single northern locality are not available, and it is difficult to gain an understanding of the precise changes and sequences of pelage. Molt is gradual over the whole animal, there never being anjr well-defined zones or divisions between the old and the new coats, but the old hairs are replaced uniformly over the whole surface. Specimens that seem to be in perfectly fresh coat are available, however, taken during every month of spring, summer, and fall. For this reason it is difficult to be sure whether there is one or two annual molts. The changes of pelage, however, are somewhat irregular, especially in younger animals, and depend fully as much upon the age of the individual as upon the precise march of the seasons. In other words, it seems that a young animal may shed its juvenile coat at some defi- nite stage of development and that temperature has but a partial influence upon such change. In addition, it is not by any means 40440—27 2 8 NORTH AMERICAN FAUNA [No. 50 certain that the changes of coat are precisely the same for all races in different climates. External size is remarkably uniform in Mictomys, but is slightly more variable in Synaptomys. The sexes do not differ in this respect. Skull. — In this genus the skull varies from moderately light to exceedingly massive, but it is less robust than in other American lemmings. The rostrum is considerably depressed and is very short, comprising about 25 per cent of the total length of the skull, and the zygomatic processes of the maxillae are proportionately heavy, as in the other Lemmi, and very deep, but the arches are not widely flaring. The brain case is rather long, the postorbital border quite sharp, and its outline almost right-angled. The interparietal is nar- row in relation to its height. There is practically no ridging upon the brain case, but the interorbital ridges are usually prominent in adults. The pterygoids are short and thickened, and the palate not greatly unlike that of Microtus. Teeth. — The maxillary incisors of Synaptomys are grooved on their anterior surfaces, m young as well as in old individuals, and worn into deep pits on their posterior faces by the narrower lower incisors. The roots of the mandibular incisors terminate at a point slightly anterior to the posterior portion of the third molar and lie wholly on the lingual side of the molars. The molars are rootless and hypsodont, growing throughout life from a persistent pulp. The enamel pattern is characterized by the extreme depth of the reentrant angles on the lingual side of the lower teeth and the buccal side of the upper ones, and are of a style essentially different from all other genera except Lemmus. The enamel is of equal thickness on the concave and convex edges of the prisms, and the reentrant angles contain a moderate quantity of cementum. The upper molars are practically the same in both subgenera. The outer reentrant angles extend to the opposite side of the molars, but the inner ones only halfway toward the buccal side. The first and second upper molars are not otherwise peculiar, the former consisting of five closed enamel spaces, and the latter of four. The third molar has three transverse loops directed outward, and in addition, a fourth situated posteriorly and directed inward, separated from the others by a deep inner reentrant angle. Its shape is distinctive, and its pattern possible of confusion only with Lemmus. The lower molars are described under the subgenera. The question of whether the anterior molariform teeth of the Microtinae are true molars or premolars is not considered in the present study, but the more conservative terminology — first, second, and third molars — is used. The molars, although hypsodont, never project so far beyond the alveoli as do those of most other genera of microtines with hypso- dont dentition, and hence there is less provision made for rapid wear. Therefore, the teeth must be unusually resistant, or else the food is less abrasive than is that of most voles. Of these two theoretical explanations, the former is considered unlikely. The facts as known seem to justify the conclusion that the molars grow at a less rapid rate than in most other genera of the subfamily and hence are of a less pronounced order of hypsodontism. 1927] REVISION OF THE AMERICAN LEMMING MICE \) VARIATION The material now available indicates strongly that the genus Synaptomys contains only two species, one to each subgenus. These, in turn, are further divisible, with tendencies for variation as men- tioned under the subgenera. Key to the Subgenera of Synaptomys Mandibular molars with triangles on outer side Synaptomys. Mandibular molars without triangles on outer side Mictomys. List of Species and Subspecies, with Their Type Localities Subgenus SYNAPTOMYS Synaptomys cooperi cooperi Baird Type locality unknown. cooperi stonei Rhoads Mays Landing, N. J. cooperi helaletes Merriam Lake Drummond Dismal Swamp, Ya. cooperi gossii Coues Neosho Falls, Kans. Subgenus MICTOMYS Synaptomys oorealis oorealis (Richardson) Fort Franklin, Mackenzie, Canada. oorealis dalli Merriam Nulato, Alaska. oorealis chapmani Allen Glacier, British Columbia, Canada. horealis wrangeli Merriam Wrangell, Alaska. oorealis innuitus (True) Fort Chimo, Ungava, Canada. oorealis medioximus Bangs L'Anse au Loup, Labrador. horealis sphagnicola Preble Fabyuns, N. H. Subgenus SYNAPTOMYS Baird Synaptomys Baird, Mamm. North Amer., p. 55S, 1857. Type Synaptomys cooperi Baird. Geographic distribution. — Canadian Zone in eastern Canada west probably through Ontario; thence south through the Upper Austral Zone along the eastern border of the Plains into Arkansas: on the Atlantic slope occurring southward into North Carolina. (Fig. 2.) Essential characters. — Mandibular molars with closed triangles upon the outer sides, and outer reentrant angles well developed. Mandibular incisors relatively robust, not usually excessively pointed. Mammae, 6. Subgeneric diagnosis. — Skull, especially rostrum, in all but one race, very heavy, with maxillary incisors correspondingly so, their outer edges never noticeably unworn nor prolonged into sharp splin- ters of enamel. Mandibular incisors rather short and relatively stout, their tips not being sharply and narrowly pointed. Mandibu- lar molars with closed triangles and well-developed reentrant angles upon the buccal side. Palate (fig. 3) but little different from that of some species of Microtus. Foramen rotundum and foramen ovale practically always separate and distant. Mammae, 6. No hairs at base of ears appreciably brighter than remainder of pelage. White patch upon the hip glands of old males seldom showing conspicuously. 10 NORTH AMERICAN FAUNA [No. 50 Skull. — The skull of the typical race (cooperi) is no larger (smaller than in some forms) nor is the rostrum heavier (although propor- tionately shorter) than in the case of Mictomys, contrary to the usual statement. The skulls of the remaining three subspecies of the present subgenus are larger, however, and the rostra and incisors decidedly heavier. The crown of the first upper molar is situated considerably above a plane passing from the inferior border of the bullae through the crown of the third molar. The interorbital sulcus is seldom obliterated in old age. The inferior surface of the cranium has very few characters of any value in diagnosis. Fig. 2. — Distribution of the subgenus Synaptomys. Open circles indicate type local- ities; solid spots, other specimen records. A, 8. c. cooperi; B, 8. c. stonei; C, 8. o. gossii; D, 8. c. helaletes Teeth. — Maxillary molars exactly as described under the genus. Of the mandibular teeth (fig. 4), the first molar consists of an ante- rior space, corresponding to the usual microtine trefoil, with an incipient triangle (widely open) upon its outer side. In addition, there are three inner loops and a median outer triangle. The second molar consists of two median loops, an irregular, narrower, enamel space anterior to these, and a median outer triangle. The third molar is composed of three triangular spaces — one of them upon the outer side — and a posterior loop. The outer reentrant angles of these molars are well developed, although not to so great an extent as the inner ones. The incisors (fig. 5) are of a deep orange color. The upper ones have the grooves very close to the outer borders, and the 1927] REVISION OF THE AMERICAN LEMMING MICE 11 Fig. 3.— Palate of the subgenus Syn- aptomys— S1. coop- eri helaletes (Dismal Swamp, Va., No. 95876, U. S. Nat. Mas.). Enlarged corners are usually noticeably worn. The mandibular incisors are short and heavy, with tips not excessively slender Remarks. — Taking the average of cranial char- acters, skulls of Synaptomys may be distinguished from those of Mictomys by their proportionately smaller bullae, shorter incisive foramina, more Microtus-like palates, relatively shorter rostra, less attenuate terminations of the ascending branches of the maxillae, the fact that the foramen rotundum and foramen ovale are always separated by a sub- stantial osseous partition, by the greater or lesser tendency of the interorbital ridging to join, by the difference in tilt of the upper molar series, and by the mandibular enamel pattern. The subgenus Synaptomys as now understood contains but a single species, divided into four exceedingly well-marked geographical races. It is considered that further subdivision, based upon the material now avail- able, would be en- tirely unjustified. There has been much uncertainty ex- pressed by various authors regarding the identity of speci- mens from many lo- calities, an uncer- tainty doubtless caused by the ques- tion of the proper status of the type of cooperi and by the fact that the existing material, some- times scanty or in other ways unsatisfactory, was widely scattered. The characters of most use in differentiating the subspecies of the subgenus Synaptomys comprise general size of the skull, shape of the brain case, the width and length of the rostrum, width of incisors, and, to a lessor ex tent, the shape of the incisive foramina and zygo- matic arches. Although there may be some indi- vidual variation in these characters, they are usually reliable. There are no differences in tooth pattern, nor in the relative size of the bullae. Immaturity is indicated chiefly by lack of the normal amount of interorbital ridging, but there are also the other signs of youth usual in the Microtinae. There are no external characters whereby the four races of this subgenus may surely be distinguished; As far as can be told by the dried study skins, the ex- ternal differences are relatively trifling, often inap- preciable, and net at all commensurate with the range of size and development exhibited by the crania. 'Ill" mammae are said to be six in number, inguinal, 1:1, pectoral 2:2, and this seems to be the maximum. Young females may have a smaller number than this actually functional. Fig. 4. — Enamel pattern of molars of the subgf-nns Synap- tomys— £. cooperi stonei (Roan Mountain, N. C, No. 47821, U. S. Nat. Mus.). Enlarged, after Miller, 1896 Fig. 5.— Front view of incisors of Hie subgenus Synap- tomys — 8. cooperi In lull ten (Dismal Swamp, Va., No. 95876, V. s. Nat. Mus.). Kni. i r. : ed 12 NORTH AME^I^O" FAUNA [No. 50 The tooth pattern of this subgenus is less advanced in type than that of Mictomys, but the skull, in most forms, is much more special- ized, though it is difficult to say for just what purposes. The north- ernmost race of the subgenus has the simplest skull, but those to the southward, inhabiting lower zones and living amid surroundings that may logically be assumed to be considerably different from what is normal for the genus, have developed very robust crania and massive dentition, most pronounced in the two southernmost sub- species. Key to the Subspecies of Synaptomys cooperi (Subgenus Synaptomys) (Based on adults) a1 Maxillary incisors very narrow (about 1.1 mm.) ; condylobasi- lar length averaging under 24 ; zygomatic breadth under 16 8. c. cooperi (p. 12) a2 Maxillary incisors wider (over 1.5 mm.) ; condylobasilar length averaging over 24.5 ; zygomatic breadth over 16.5. ft1 Incisive foramina short (averaging 4.7) ; condylobasilar length under 26. c1 Skull not broad for length; rostrum moderate (width under 6) Sf. c. stonei (p. 14) cr Skull very broad for length ; rostrum stubby and massive (width over 6.1) S. c. helaletes (p. 17) o2 Incisive foramina long (averaging 5.7) ; condylobasilar length about 27 8. c. gossii (p. 18) SYNAPTOMYS COOPERI COOPERI Baird Cooper Lemming Mouse (Plate 1, A) Synaptomys cooperi Baird, Mamm. North Amer., p. 556, 1857. Synaptomys fatuus Bangs, Proc. Biol. Soc. Washington 10:47, 1896. Type from Lake Edward, Quebec, Canada. Female adult ; No. 3857, coll. E. A. and O. Bangs : collected by O. Bangs, September 28, 1895 ; original num- ber 3. Type. — Type locality, unknown; immature; No. \ | g fr, U. S. Nat. Mus. ; from Wm. Cooper. Geographic distribution. — Canadian and upper part of Transition Zones in North American east of the Plains from Nova Scotia to Godbout, Quebec, and thence westward to the territory immediately north and west of Lake Superior; south to central Wisconsin and Michigan and in the Catskill Mountains to southern New York. (Fig. 2.) General characters. — A small race with light rostrum, small in- cisors, weak brain case, and little or no interorbital ridging. Color. — As in the subgenus, dorsally somewhat close to the cinna- mon brown of Eidgway. Below, the skins with least worn pelage show a distinct tinge of cream, while the others are of the usual plumbeous. Skull. — The cranium is the smallest of any race of the subgenus and may readily be distinguished by the small size of the incisors (width, 1.1 mm.) and by the narrowness of the rostrum, making" the latter appear longer in comparison with the rest of the skull than in the other races. Interorbital and temporal ridges prac- 1927] REVISION" OF THE AMERICAN LEMMING MICE 13 tically absent in characteristic specimens. Brain case weak and rather long for its width. Zygomatic arches weak and but slightly developed at the supero-inferior expansion of the jugals. Measurements. — Average of collectors' measurements of five largest individuals from Quebec: Total length, 118; tail, 16.5; foot, 18. Average of four adult skulls from Quebec: Condylobasilar length, 23.1; rostral length, 6.1; rostral breadth, 5; interorbital breadth, 3; zvgomatic breadth, 15.1; lambdoidal width, 11.9; incisive foramina, 4^3; height, 8.6. Remarks. — In the original description of the species, Baird stated that the type is "probably from the New England States, or New York; possibly from Iowa or Minnesota." He must have had some reason for making such a statement instead of presuming, as have most of his successors, that the type came from near the Hoboken, N. J., home of Mr. Cooper, from whom it was 6btained. The type is without the head and feet and is torn in many places, and the single paratype is in even worse condition. The posterior por- tion of the skull of the type is missing, only the rostral region, the right zygomatic arch and postorbital border remaining intact, but this is entirely adequate to show that the specimen is subadult, and that it certainly can not be identified with the material from the Central Atlantic States heretofore called cooperi. Among the lemming mice now available it can be exactly matched only by certain examples of the " fatuus " of authors from Ontario, Quebec, and New York State. Hence, fatuus must now be abandoned, as a pure synonym of cooperi. The type locality of the latter can prob- ably never be fixed with exactitude, but it may well be considered as situated somewhere in the northern or western portion of the New England States, or in western New York, as Baird suggested. So far as the present material indicates, skulls from western New York are practically indistinguishable from those taken in Quebec, and the characters of the latter are now considered in greater detail for the reason that the material from this area is the most satisfactory of any available. The chief characters whereby cooperi may be distinguished are the small size of the skull coupled with the narrowness of the ros- trum and almost total lack of ridging, the two latter points being especially diagnostic. The increase in the width of the incisors as the range of stonei is approached is sometimes of great value in an understanding of the degree of intergradation existing in various areas. The measurements given indicate a shorter foot for cooperi, but comparison of skins fails to show an appreciable difference in this respect. Along the southern border of its range this race approaches stonei in varying degree, especially where the Great Lakes do not impose a physical barrier between the two. Such variants are not always uniform as regards minor characters, but these differences are, as a rule, relatively minute. Specimens from northern Wis- consin and Ann Arbor, Mich., show a slight increase in the size of the brain case and a stronger indication of interorbital ridging than is usual in Canadian examples, and the same may be said of the specimens from Minnesota. The latter are unusually bright in coloration, but this may be due to the fact that they were taken 14 NORTH AMERICAN FAUNA [No. 50 early in March, a season during which there are almost no specimens of this race available for comparison. It is considered best to allo- cate such with cooperi, even though some of them, with equal pro- priety, might be considered by others as referable to stonei. Rather unexpectedly the single skull from the Catskill Moun- tains must be called cooperi, although it is not typical. Two specimens from Vermont are intermediate in character and might as well be referred to stonei as to the present race. Those from Essex County, N. Y., vary to some degree, especially in the slightly larger incisors, toward stonei, but only one of these specimens is adult. The skull from Mount Mansfield, badly damaged, is espe- cially questionable as to its identity because of the rostrum. The skull accompanying one of the skins labeled Canoe River, British Columbia, proves to be a true Synaptomys, but this sub- genus has never before been taken west of the prairies, and six other specimens taken at the same time are without skulls. If this example really came from the locality indicated it would almost certainly be perfectly distinct from anything now known, but in many ways it is close to cooperi, although undeniably larger. It was taken by a reliable collector, but because there has very likely been some sort of subsequent confusion in the case of this skull, as occasionally happens in spite of utmost care, it is no more than mentioned here. Specimens examined. — Total number, 88, as follows : Locality unknown: 2 (type and paratype). Maine: Mount Madawaska, 24; east branch Penobscot River, 1; Sebec Lake, 1. Michigan: Ann Arbor, 12. Minnesota: Burntside Lake, 2; Elk River, 4. New Brunswick: Near Bathurst, 45; Tobique Point, l.6 New Hampshire: Mount Moosilauke, l.4 New York: Alfred Center, 1 ; Catskill Mountains, 1 ; Keene Heights, Essex County, 44; Wanakena, 1. Nova Scotia: Digby, 5.7 Ontario: Bryanston. 28; London, 8"; Macdiarinid, Lake Nipigon, l.10 Quebec: Godbout, 3"; Lake Edward, 10 (including type of "fatuus")1*; St. Rose, 5. Vermont: Leicester, l13; Mount Mansfield, 1. Wisconsin: Oonover, 3"; Lac Vieux Desert, l14; Lakewood, 1; Lake St. Germain, 4 ; Long Lake, 2 ; Mercer, 1 ; St. Croix Falls, 1 ; Sayner, 1 " ; Solon Springs, 1." SYNAPTOMYS COOPERI STONEI Rhoads Stone Lemming Mouse (Plate 1, B) Synaptomys stonei Rhoads, Amer. Nat. 27 : 53, 1893. Type. — Collected at Mays Landing (on Egg River), Atlantic County, N. J., by S. N. Rhoads, December 2, 1892. Female subadult; i C. F. Batchelder coll. B Nat. Mus, Canada. 6 Amer. Mus. Nat. Hist. 7 Mus. Comp. Zool. 8 Nat. Mus. Canada, 1 ; W. E. Saunders coll., 1. 9 W. E. Saunders coll. 10 Royal Ontario Mus. Zool. 11 U. S. Nat. Mus., 2 ; Royal Ontario Mus. Zool., 1. 12 Mus. Comp. Zool., 7 ; Field Mus. Nat. Hist., 3. 13 G. L. Kirk coll. "Field Mus. Nat. Hist. 1927] REVISION OF THE AMERICAN LEMMING MICE 15 No. 7567, Acad. Nat. Sci. Philadelphia (formerly No. 567, coll. S. N. Khoads). Geographic distribution. — Lower portion of Transition and north- ern half of the Upper Austral Zones in the United States east of the Plains, from central Wisconsin and Illinois east to the Atlantic coast; occurs as far north as Massachusetts and south in the moun- tains into North Carolina. (Fig. 2.) General characters. — A race with moderate-sized, well-ridged skull with well-developed zygomatic region. Rostrum somewhat inter- mediate in size between the extremes of development as exhibited by cooperi, on the one hand, and helaletes, on the other. Incisors wide (1.7 millimeters). Color. — The coloration of New Jersey specimens is indistinguish- able from that of animals from Quebec. Maryland skins average a shade brighter. Skull. — Adult crania from southern New Jersey are of medium size with well-developed interorbital ridges, which meet in the case of animals of advanced age. Brain case rather long, its length being especially accentuated in thoroughly typical specimens by the nar- rowness of the interparietal. The rostrum is much heavier than in cooperi, as are the incisors, and the zygomatic arches flare strongly, with jugals quite deep supero-inferiorly. The incisive foramina are variable, but are usually relatively long. Measurements. — Average of collectors' measurements of four largest specimens from southern New Jersey: Total length, 127; tail, 21; foot, 20. Average of four adult skulls from the same area : Condy- lobasilar length, 24.9 ; rostral length, 6.4 ; rostral breadth, 5.9 ; inter- orbital breadth, 3.1; zygomatic breadth, 16.7; lambdoidal width, 13; incisive foramina, 4.7; height, 9.9. Remarks. — As the name cooperi is found to belong to the northern- most race of the subgenus, the form occurring in the Middle Atlantic States will have to take Rhoads's name stonei. The type specimen of this is not adult and the cranial points that may be considered most characteristic of the animals in the vicinity of the type locality are not so well developed as could be desired. This race is in many ways intermediate between the light-skulled cooperi, on the one hand, and helaletes and gossii, with their heavy crania, on the other, yet being well defined and rather constant over a wide area, it can not be considered as an intergrade. This inter- mediateness in skull characters is entirely sufficient to distinguish it from the three other races. The larger skull and heavier rostrum, including incisors, in connection with the well-developed ridging, distinguish it from cooperi, and its smaller size and narrower or smaller rostrum, from the remaining two forms. Its foot seems to be a shade longer than that of cooperi, but not sufficiently so for this character to be of practical value in diagnosis. In the northern portion of its range stonei approaches cooperi in varying degree, and there is some little geographic fluctuation of characters in several places, as may only be expected of an animal inhabiting bogs that must often be separated by considerable intervals of country unsuited to the needs of the genus. Such fluctuations, however, are usually slight and on the basis of the material now 40440—27 3 16 NORTH AMERICAN FAUNA [No. 50 available seem not to be sufficiently pronounced or constant to admit of further subspecific division. Specimens from southern Maryland definitely vary toward hela- letes, having the large brain case and stout zygomatic region of that race, but with the rostra more as in stonei, with which they should clearly be placed. The individuals from Roan Mountain, N. C, have slightly weaker skulls, narrower zygomata and wider interparietals than typical, but these differences are not sufficiently pronounced to merit separation. An old female from Wareham, Mass., is certainly closer to stonei, although with a rather weak rostrum, while the three other skulls from this locality are entirely intermediate between this race and cooperi. The individual from Cassopolis, Mich., taken during January, is excessively gray, especially laterally. There is a type of variant in central Wisconsin the skulls of which are no larger than those of cooperi, but fully as stout and ridged, pro- portionately, as are those of stonei. This is most accentuated in in- dividuals from Elkhart Lake and Kilbourn, and these might possibly be considered worthy of subspecific separation if it were not for the fact that the animals from Kelly Lake, Wild Rose, and Rib Hill, all to the north of the two localities first mentioned, have skulls much larger and close to typical stonei. Such being the case, it seems wiser to consider them merely intermediate, to different degrees, between stonei and the cooperi type of lemming mouse occurring in the northern part of the State. The Indiana animals have, at one time or another, been referred to gossii, but in view of the adequate series of the latter form now avail- able, such a decision can not be sustained. Material from the south- ern half of the State is clearly stonei, while two skulls from Hebron and Roselawn, in the northwest section, although undeniably large and with stout rostra, are not sufficiently developed in these respects for inclusion with the Kansas race. Specimens examined. — Total number, 160, as follows : Indiana: Bascom, 8; Bicknell, 1; Brookville, 6; Hebron, 1; Mitchell, 4; Roselawn, 1 ; Salamonia, 5.15 Maryland: Hyattsville, 8; Prince George County, 6. Massachusetts: Wareham, 4.10 Michigan: Cassopolis, l.18 New Jersey: Bear Swamp, Sussex County, 1"; Greenwood Lake, 2"; Mays Landing, 4 18 ; Port Norris, 3 " ; Tuckahoe, 1." North Carolina: Magnetic City, 1; Roan Mountain, 25 10; Weaverville, ll.80 Ohio: Ravenna. 2.21 Pennsylvania: Bushkill Creek, near Cresco, 1"; Krings Station, 5"; Lake Lehigh, 1 * ; Round Island, 2." Tennessee: Rogersville, 1. Virginia: Campbell County, 13 a; Mount Rodgers, 5." West Virginia: Cranberry Glades, 1; Pocahontas County (head of Cran- berry River), 5; Travellers Repose, 4; White Sulphur Springs, S.82 Wisconsin: Camp Douglas, 1; Elkhart Lake, 2; Kelly Lake, 9; Kilbourn, 3; Rib Hill, 2 ; Wild Rose, 7. " Field Mus. Nat. Hist. 16 Mus. Conip. Zool. "Acad. Nat. Sci. Philadelphia. 18 Acad. Nat. Sci. Philadelphia, 3 (including type) ; U. S. Nat. Mus., 1. 19 U. S. Nat. Mus., 19 ; Acad. Nat. Sci. Philadelphia, 6. a" Amer. Mus. Nat. Hist., 10 ; Field Mus. Nat. Hist., 1. aAmer. Mus. Nat. Hist. 23 Field Mus. Nat. Hist., 1 ; Mus. Comp. Zool., 2. 1927] REVISION OF THE AMERICAN LEMMING MICE 17 SYNAPTOMYS COOPERI HELALETES Meebiam Dismal Swamp Lemming Mouse (Plate 1, C) Synaptomys helaletes Merriam, Proc. Biol. Soc. Washington 10 : 59, 1896. Type. — Collected at Lake Drummond, Dismal Swamp, Norfolk County, Va., by A. K. Fisher, October 14, 1895. Female adult; No. 75172, U. S. Nat. Mus. (Biol. Surv. coll.) ; original number 1818. Geographic distribution. — Extreme southeastern Virginia and northeastern North Carolina. (Fig. 2.) General characters. — A race with large, heavy skull and broad, stubby rostrum. Brain case wide and relatively short, and zygo- matic arches much bowed. Incisors very broad (1.9 mm.) (Fig. 5), and incisive foramina rather short. Color. — Evidently averaging a shade brighter and lighter (only winter and early spring specimens available) than the two pre- ceding forms, but many individuals are indistinguishable from them. Skull. — The rostrum of the present form is excessively broad with- out any compensating increase in length, and the zygomatic arches are strongly bowed. The interorbital ridging is moderate, with the sulcus almost, though not quite, obliterated in old age. The brain case is very broad and hence appears rather short, and the interparietal is moderately wide. The incisive foramina are usually short. Measurements. — Average of nine adult topotypes: Total length, 129; tail, 21.4; foot, 20. Average of six skulls of adult topotypes: Condylobasilar length, 25.5; rostral length, 6.6; rostral breadth, 6.3; interorbital breadth, 3.3; zygomatic breadth, 17.2; lambdoidal width, 13.4; incisive foramina, 4.7; height, 10.1. Remarks. — The present form may be most readily distinguished from typical stonei by the much broader, though no longer, rostrum, and correspondingly broader incisors. The zygomatic arches are more strongly bowed and the brain case is considerably wider. The incisive foramina are also, as a rule, shorter. In reality the ridging is less developed in helaletes, the skull gaining its ap- pearance of ruggedness by its great breadth. The feet of this lemming mouse, although no longer, are apparently appreciably stouter than those of its neighbor. As previously mentioned, specimens from southern Maryland, al- though referable to stonei, definitely vary in the direction of helaletes. At the present day, bogs suitable for the presence of this genus are undergoing shrinkage throughout the region in ques- tion, and there may now be a definite hiatus between the ranges of the two forms, but there are not sufficient grounds for consider- ing the Dismal Swamp animals as being more than subspecifically distinct. The two skins from Chapanoke are very light, but the single skull from this locality is entirely typical. Specimens examined. — Total number, 24, as follows : North Carolina: Chapanoke, 2.23 Virginia: Lake Drummond, 22 (including type). 23 U. S. Nat. Mus., 1; Acad. Nat. Sci. .Philadelphia. 1. 18 NORTH AMERICAN FAUNA [No. 50 SYNAPTOMYS COOPERI GOSSII (Cotjes) Goss Lemming Mouse (Plate 1, D) Arvicola (Synaptomys) gossii Coues, Monogr. North Amer. Rodentia, p. 235, 1877. Type. — Collected at Neosho Falls, Woodson County, Kans., by B. F. Goss, 1866. _ Male adult (skull only) ; No. 6915, U. S. Nat. Mus. Geographic distribution. — The west-central Mississippi Valley country, mostly in the Upper Austral Zone, from northeastern Arkansas and southern Illinois into Iowa and extreme southeastern South Dakota (fig. 2). General characters. — A large, rather bright-colored Synaptomys with long, heavy skull having a wide, though long, rostrum. Color. — Dark examples are indistinguishable externally from the other races, but the majority of adults are of a somewhat brighter, richer shade, most noticeable in the case of the series from Rosiclare, 111. Skull. — Long and heavy in all parts, but its width hardly propor- tionate to the length. Rostrum heavy but long, with massive incisors (width, 2 mm.). Interorbital ridging moderately pronounced but never meeting entirely in old age. Temporal ridges weil developed both anterior to the parietals, and in the interparietal-lambdoidal region. The zygomatic arches, although heavy, are relatively un- expanded, thus adding to the appearance of length of the cranium. The incisive foramina are narrow but very long. Measurements. — Average of three adults from Kansas: Total length, 134; tail, 21; foot, 20. Average of four adult skulls from Kansas: Condylobasilar length, 27.1; rostral length, 6.3; rostral breadth, 6.1; interorbital breadth, 3.5; zygomatic breadth, 17.7; lambdoidal width, 13.3 ; incisive foramina, 5.6 ; height, 10.9. Remarks. — In the original citation, Coues gives this name as above, observing that it had been written by Baird upon the labels of the Kansas specimens, and giving a list of these with their measurements. Such a citation may be considered a valid characterization, and hence Coues is sponsor for the name rather than Merriam (1896, p. 60). Coues, however, designated no type, and this was later selected by Merriam, as above. The present race may be distinguished from stonei by its larger size and heavier ridging on the posterior portion of the brain case. The zygomatic arches are proportionately less flaring, and the in- cisive foramina longer and narrower. The feet of this form, as well as of helaletes, seem to be wider and stouter, though no longer, than those of the lighter cooperi and stonei. The trend of development of both gossii and helaletes has been toward an increase in the size of the cranium, but with this excep- tion, large size, the skulls of the two are not particularly similar, and even were their ranges contiguous, there would be no reason for considering them closer to each other than to stonei. The Kansas race has a long type of skull with rostrum not proportionately very broad, and long incisive foramina, while helaletes has a short, broad type of skull with very wide rostrum and usually short incisive foramina. 1927] REVISION OF THE AMERICAN LEMMING MICE 19 Specimens from Rosiclare, in extreme southeastern Illinois, are referred to gossii because of color, but the skulls are intermediate between that and stonei (indicating the true relationship of the Kansas race), and they might almost as well be called one as the other. Although not typical, the series from Horseshoe Lake, Mo., must be called gossii, but the skulls, as long as Kansas ones, have propor- tions a trifle suggestive of stonei. The affinities of the Hillsboro individual are undoubtedly with gossii, but those from Lake City and Odin are too young to be iden- tified with certainty, and they are assigned to this race upon geo- graphic grounds. Hahn states (1909b, p. 522) that he took lemming mice on the Mis- souri River in southeastern South Dakota, which he called gossii because they were " much larger and clumsier, with larger and heav- ier skull " than Indiana examples. The writer has not succeeded in locating this material, but presumes that Hahn was correct in his diagnosis. Specimens examined. — Total number, 47, as follows: Arkansas: Lake City, 1. Illinois: Odin, 1 ; Rosiclare, 11." Iowa: Hillsboro, 1; Knoxville, 1; Marion, l.26 Kansas: Lawrence, 1M; Leavenworth, 7"; Neosho Falls, 7.w Missouri: Horseshoe Lake, 16. Subgenus MICTOMYS True Mietomys True, Proc. U. S. Nat. Mus. 17 : 242, April 26, 1894. (Advance sep- arate.) Type Mietomys innuitus True. Geographic distribution. — Canadian and Hudsonian Zones of Alaska and British America, barely entering the extreme north- western part of the United States, and in the east extending south- ward into New Hampshire. (Fig. 6.) Essential characters. — Mandibular molars without closed triangles upon the outer sides, and practically no outer reentrant angles. Mandibular incisors relatively slender and sharply pointed. Mam- mae, 8. Subgeneric diagnosis. — Neither skull nor rostrum is especially heavy. The maxillary incisors often have the outer corners notice- ably unworn and prolonged into sharp splinters of enamel, after the manner commonly seen in the genus Lemmus. This may be ac- companied by deep pitting upon the oral surface of the upper incisors, the whole formed by the action of the comparatively long and slender, sharply pointed, lower incisors working against the very much wider upper ones. Mandibular molars are without closed triangles or distinct reentrant angles upon the buccal sides. Palate (fig. 7) with a sharply pointed median projection directed backwards — not Microtus-like. Foramen rotundum and foramen ovale usually, though not always, joined, forming a single vacuity. When separated, the osseous partition is exceedingly thin and frag- M Field Mus. Nat. Hist. 28 Mus. Comp. Zool. (no skull, identity inferred). wAmer. Mus. Nat. Hist. 57 U. 8. Nat. Mus., 6 ; Amer. Mus. Nat. Hist., 1. 28 U. S. Nat. Mus., 6 (including type) ; Mus. Comp. Zool., 1. 20 NORTH AMERICAN FAUNA [No. 50 ile. Mammae, 8. A few hairs at the base of the ears distinctly brighter, with a bright rusty tinge, and hairs upon the hip glands of old males usually conspicuously whitish. /Skull. — The skull of Mictomys is smaller and lighter than all but one of the subspecies of the subgenus Synaptomys, and the Fig. 6. — Approximate distribution of the subgenus Mictomys rostrum is lighter and relatively longer, but the interorbital ridges join, thus obliterating the median sulcus, at a rather early age. The bullae are comparatively larger, the incisive foramina longer, the terminations of the ascending branches of the maxillae average more attenuate, the foramen rotundum and the foramen ovale are usually joined into a single vacuity, or else the intervening osseous partition is very fragile, and the median projection of the palate is spinous in character. The crown of the first upper molar is about on a level with a plane passing from the inferior border of the bullae through the crown of the third molar. Teeth. — The maxillary molars are as in the genus. The mandibular molars (fig. 8) are formed entirely of transverse, wedge-shaped loops, four in the first (including anterior " trefoil ") , and three in the sec- ond and third. These differ in pattern from the sub- genus Synaptomys in the absence of an outer tri- angle from each tooth, and in having the buccal borders crenulate, with outer reentrant angles barely indicated, except in the third molar, where one is slightly developed. The incisors (fig. 9) are pale in color, especially upon the outer edges, relatively light, and the grooves of the upper ones are not so close to the outer borders as in the subgenus Synaptomys. The lower incisors are very long, slender, and exceedingly sharply pointed, these terminations fitting into unusually well-defined pits upon the oral surface of the much broader upper incisors. The latter are Fig. 7. — Palate of the subgenus Mic- tomys— &. borea- li8 c h a p m an i (Smoky River, Al- berta, No. 17437, U. S. Nat. Mus.). Enlarged 1927] REVISION OF THE AMERICAN LEMMING MICE 21 Fig. 8. — Enamel pattern of molars of the subgenus Mictomys — 8. borealis innuitus (tvpe, Fort Chimo, Ungava, No. 24729, U. S. Nat. Mus.). Enlarged, after Miller, 1896 usually somewhat unworn upon their outer corners, and there often results a very sharp splinter of enamel that projects at this point. Remarks. — There is really remarkably little geographic variation, both externally and cranially, in the subgenus Mictomys, and although its range can not, as yet, be demon- strated to be uninter- rupted, there is indicated so much continuity in the fluctuation of char- acters that the various forms are likely no more than subspecifically dis- tinct. The chief handi- cap to a proper under- standing of the interre- lationship of the races is the dearth of topotypical material, and it will probabty be many years before this condition can be entirely remedied. External characters are more satisfactory than in the subgenus Synaptomys, but there is slight difference in external size, such, whenever indicated by collectors' measurements, usually proving unreliable when a direct comparison of specimens is made. The characters of most value in diagnosing the various races of this subgenus consist of size of skull; height and proportion of the brain case; length and width of the rostrum (but seldom of the nasals), incisive foramina, and occasionally the bullae; and length of the pterygoid fossae. Some of these characters, however, may be found to fluctuate, especially in regions of diverse faunal tendencies. Immaturity is indicated chiefly by the lack of pronounced interorbital riding. The mammae are said to be 8 in number, inguinal 2 : 2, pectoral 2 : 2, and such seems to be the case, as well as could be told from dried specimens. Young females may develop a fewer number. As previously mentioned, the tooth pattern of Mictomys is con- sidered more specialized than that of the subgenus Synaptomys, and the skull is simpler. Key to Subspecies of Synaptomys borealis (Subgenus Mictomys) o ' Races ranging chiefly east of the mountains of western Canada and Alaska. &1 Coloration dark — rich mahogany; pterygoid fossae short; Athabaska-Mackenzie region 8. b. borealis (p. 22). b * Coloration duller or paler ; pterygoid fossae longer ; eastern North America, c * Skull very small — condylobasilar length 23 : color not known 8. b. innuitus (p. 28). v* Skull larger — condylobasilar length over 24. d1 Color pale and bright (adult skull unknown) 8. b. tnedioximus (p. 29). d2 Color darker; condylobasilar length over 25_ 8. b. sphagnicola (p. 30). Fig. 9. — Front view of incisors of the subgenus Micto- mys— 8. borealis chapmani (Shovel Pass, Alberta, No. 3352, Nat. Mus. Canada). Enlarged 22 NORTH AMERICAN FAUNA [No. 50 a2 Races ranging chiefly west of the Athabaska-Mackenzie region of Canada. &1 Coloration usually bright: incisive foramina short. c" Skull high (9.3) : condylobasilar length about 24 or more. 8. 6. dalli (p. 24). o2 Skull low (8.7) and flat: condylobasilar length averaging less than 24 8. &. wrangeli (p. 26). 6a Coloration dull, with little chestnut: incisive foramina longer 8. 6. chapmani (p. 25). STNAPTOMYS BOREALIS BOREALIS (Richabdson) Richardson Lemming Mouse Arvicola borealis Richardson, Zool. Jour. 3 ; 517, 1828. Synaptomys (Mictomys) oullatus Preble, Proc. Biol. Soc. Washington 15; 181, 1902. Type from Trout Rock, near Fort Rae, Great Slave Lake, Mackenzie, Canada. Male adult; No. 110032, U. S. Nat. Mus. (Biol. Surv. coll. ) ; collected by Edward A. Preble, August 17, 1901 ; original number 4511. Type. — Collected at Fort Franklin, Mackenzie, Canada, by John Richardson. ^ Male; No. 42.10.7.10, British Mus. Geographic distribution. — The Athabaska-Mackenzie region of Canada from Great Bear Lake south to near Edmonton, and east- ward (provisionally) to Lake Winnipeg (fig. 10). General characters. — A dark, richly colored race, with foot some- what short. Skull rather small, and pterygoid fossae short and crowded. Color. — The dorsal coloration of this race is very rich and dark, close to the Argus brown of Bidgway, and plentifully mixed with black-tipped hairs. This richness of color is most pronounced on the rump, the anterior portion of the body often being duller, slightly grizzled, and a trifle paler. The plumbeous of the under surface of the specimens at hand is rather dark and without appreciable tinge of buff. The tail is distinctly bicolor and the feet vary from grayish to almost black. Skull. — The skulls of the few adult topotypes at hand are rather small, with rostrum tapering hardly at all, and high brain cases. The incisive foramina, while not long, are usually quite wide, espe- cially posteriorly. The pterygoid fossae are short and crowded and the pterygoid plates heavy. Measurements. — Average of collectors' measurements of four adult topotypes: Total length, 129; tail, 26; foot, 18. Average of three skulls of adult topotypes: Condylobasilar length, 24; rostral length, 5.9; rostral breadth, 4.9; interorbital breadth, 3.2; zygomatic breadth, 15.2; lambdoidal width, 12.1; incisive foramina, 4.9; height, 9.6. Remarks. — The identity of Richardson's Arvicola borealis was a puzzle for many years, and it was variously ascribed to Arvicola, Microtus, or Phenacomys. Therefore, when E. A. Preble obtained a series of Synaptomys from the Mackenzie region in 1901, apparently differing from anj^thing before characterized, he named it " bullatus" not suspecting the true identity of the species borealis. During 1903 he again made a trip to this region and visited Fort Franklin espe- cially for topotypes of some of Bichardson's species. A series of eight Synaptomys was obtained, and when reporting upon these, Preble (1908, p. 184) stated his conviction that they constituted the 1927] REVISION OF THE AMERICAN LEMMING MICE 23 Arvicola borealis of Richardson. In the interval between Preble's rediscovery of the species and the publication of his report, the frag- mentary skull of the type of borealis in the British Museum had been removed and examined by Osgood (1907, p. 49) who found that it is indeed a Mictomys. Hence, the name bulletins was placed in synonymy by its describer. Synaptomys b. borealis is not sufficiently differentiated from the more eastern races of Mictomys for the interrelationship to be con- sidered more than subspecific in degree, and the future will probably show the range of the subgenus to be practically continuous between Via: 10. — Distribution of the western forms of the subgenus Mictomys. Open circles indicate type localities; solid spots, other specimen records. A, 8. b. borealis ; B, 8. b. dalli ; (', Ef. b. chapinani ; D, .s'. b. urrangeto the two regions. Skins in unworn pelage of borealis as here recog- nized may be distinguished from available ones of medioximus and sphagnicola by the darker, richer color and slightly shorter foot, but there is much variation in the crania, even among the few topo- types. A skull from Echimamish River, just north of Lake Winni- peg, is practically indistinguishable from the type of the small race, ■innuitus, except "for slightly greater height. Until something more than a single typical skull or the latter is available, however, it is unwise to extend its range too greatly. At any rate, this Echi- mamish River specimen is not at all typical borealis and if not 24 NORTH AMERICAN FAUNA [No. 50 innuitus, may be a type of variant, or even another subspecies, in- habiting the country just to the westward of Hudson Bay. Some of the examples of borealis from the vicinity of Athabaska Landing show a slight tendency toward this small type of skull with light rostrum, while others, notably from Smith Landing, indi- cate such a close approach to the dalli- chapmani type of cranium that the state of affairs can hardly be explained otherwise than by assuming that intergr a elation occurs at some point slightly to the westward. Specimens examined. — Total number, 54, as follows : Alberta: Athabaska Landing (Athabaska), 3; 5 miles above Athabaska Landing, 2 ; 30 miles above Athabaska Landing, 18 ; Battle Lake, 1 ; Peace River Landing, 1 w ; 50 miles above Pelican Rapid, Athabaska River, 2 ; Slave River, 25 miles below mouth of Peace River, 1 ; Smith Landing (Fitzgerald), 3; Swift Current. 3. Manitoba : Echimamish River, 1 ; Norway House, 1. Mackenzie: Fort Franklin, 8; Fort Providence, 1; Fort Rae, 6 (including type of "bullatus"); Grandin River, 1; Lake St. Croix. 1; Sarnhk Lake, 1. SYNAPTOMYS BOREALIS DALLI Mebriam Daix Lemming Mouse Synaptomys (Hictomys) dalli Merriam, Proc. Biol. Soc. Washington 10:62, 1896. Synaptomys (Mictomys) andersoni Allen, Bui. Amer. Mus. Nat. Hist. 19:554, 1903. Type from Level Mountain, northern British Columbia, Canada. Female immature ; No. 20467, Amer. Mus. Nat. Hist. ; collected \>j M. P. Anderson, September 11, 1902 ; original number 736. Type.— Collected by W. H. Dall, February, 1867, at Nulato, Yukon River, Alaska. Male immature; No. fgjgf, U. S. Nat. Mus. (Skull and skeleton only.) Geographic distribution. — Hudsonian Zone in Alaska and south to central British Columbia to the eastward of the coast district (fig. 10). General characters. — A rather bright-colored race, with skull of moderate size. Color. — In the case of Yukon animals, the dorsal coloration is close to the Brussels brown of Ridgway, with the usual admixture of blackish hairs. Tail, quite sharply bicolor. Skull. — Of moderate size. Incisive foramina fairly short, and bullae rather small. There is some variation in the length of the pterygoid fossae, those of Yukon examples being shortest. Measurements. — Average of collectors' measurements of three adults from the general vicinity of Fortymile, upper Yukon River Total length, 131; tail, 19; foot, 19. Average of four adult skulls from the same region : Condylobasilar length, 24.1 ; rostral length, 6.1; rostral breadth, 4.9; interorbital breadth, 3.1; zygomatic breadth, 15.5; lambdoidal width, 11.9; incisive foramina, 4.9; height, 9.3. Remarks. — The type of dalli is not only too immature for salient characters to have appeared, but the skin seems not to have been preserved. Hence, it is necessary to assume that the available ma- 20 Nat. Mus. Canada. 1927] REVISION" OF THE AMERICAN LEMMING MICE 25 terial from Charlie Creek, Fortymile, and Eagle, on the upper Yukon River, is typical. In comparison with topotypes of borealis there is little average cranial difference, but skulls of the latter are inclined to have slightly longer rostra, wider brain cases, wider in- cisive foramina, and shorter pterygoid fossae. Skins of dalli, how- ever, are paler and lack the rich mahogany tone of true ~borealis. A large series of specimens from about the base of the Alaska Peninsula exhibit much individual variation, and several of the skulls are very large, with robust rostra. Others from the same or contiguous localities, however, are smaller than what is believed to be typical of dalli, and little can be done toward an exact under- standing of the animals of this general region until both topotypes of dalli, and specimens from the lower Yukon, have been obtained. Contrary to the statement of the describer, " andersoni " proves to be decidedly immature, with skull too young to have developed diagnostic characters of value, and coloration perfectly comparable with that of dalli. At the time of its original description, the type specimen seems to have been compared only with wrangeli and not with dalli, under which name it is now placed. - Intergraduation with chapmani probably occurs somewhere in the mountains of northern British Columbia, but there are too few specimens from the region to allow this statement to be made with- out qualification. The skulls of the three examples from Great Glacier are not very satisfactory, but upon their careful consideration the writer is con- vinced that they may more logically be placed with dalli than with wrangeli. Specimens examined. — Total number, 135, as follows: Alaska: Barabori, Kenai Peninsula, 230; Bettles, 7; Caribou Camp, Kenai Peninsula, 53': Charlie Creek, 1; heart Chulitna River, 1; mountains near Eagle, 3 ; Hope, 1 ; Kenai Peninsula, 1 30 ; Kokwok River, 1 ; Lake Aleknagik, 1 ; Lake Clark, 12 ; Lake Iliamna, 5 ; MeKinley fork of Kuskokwim River, 1; Moose Camp, Kenai Peninsula, 780; Nulato, 1 (the type) ; Nushagak River, 6; Prince William Sound, 931; Seldovia, 38 3a; Sheep Creek, Kenai Peninsula, ll30; Tanana, 1. British Columbia: Atlin, l33; Great Glacier, Stikine River, 3 3l ; Klappan River, 1; Level Mountain. 1 (type of "andersoni")30; McDame Post, 1; Rapirt River. 1 ; Telegraph Creek, 1. Yukon: Chandindu River, 1; mouth Coal Creek, 1; Fortymile, 1; Lake Lebarge, 2; forks Macmillan River, 5; Rink Rapids, 1; Thirtymile River, 1. STNAPTOMYS BOREALIS CHAPMANI Allen Chapman Lemming Mouse (Plate 2, C) Synaptomys (Mictomys) chapmani Allen, Bui. Amer. Mus. Nat. Hist. 19: 555, 1903. Type.— Collected by F. M. Chapman, July 20, 1901, at Glacier, Selkirk Range, British Columbia, Canada. Male adult; No. 16908, Amer. Mus. Nat. Hist. ; original number 7. ' i Ot 0 NORTH AMERICAN FAUNA No. 51 [JULY, 1928] A TAXONOMIC REVIEW OE THE AMERICAN LONG-TAILED SHREWS (Genera SOREX and MICROSOREX) BY HARTLEY H. T. JACKSON BIOLOGIST, DIVISION OF BIOLOGICAL INVESTIGATIONS BUREAU OF BIOLOGICAL SURVEY UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON 3028 NORTH AMERICAN FAUNAS Copies of North American Faunas not out of print are for sale, at the prices named, by the Superintendent of Documents, Government Printing Office, Washington, D. C. Numbers marked with an asterisk [*] are out of print. *No. 1. Revision of the North American Pocket Mice. By C. Hart Merriam. Pp. 36, pis. 4. 1889. *No. 2. Descriptions of Fourteen New Species and- One New Genus of North American Mammals. By C. Hart Merriam. Pp. 52, pis. 8, figs. 7. 18S9. *No. 3. Results of a Biological Survey of the San Francisco Mountain Region and Desert of the Little Colorado, Arizona. By C. Hart Merriam and Leonhard Stejneger. Pp. 136, pis. 14, maps 5 (colored), figs. 2. 1890. *No. 4. Descriptions of Twenty-six New Species of North American Mammals. By C. Hart Merriam. Pp. 60, pis. 3, figs. 3. 1S90. *No. 5. Results of a Biological Reconnoissance of South-central Idaho. By C. Hart Merriam and Leonhard Stejneger. Descriptions of a New Genus and Two New Species of North American Mammals. By C. Hart Merriam. Pp. 132, pis. 4 (1 colored), figs. 4. 1891. No. 6. Not issued. •No. 7. The Death Valley Expedition: A Biological Survey of Parts of Cali- fornia, Nevada, Arizona, and Utah. Part II. — 1. Birds, by A. K. Fisher. 2. Reptiles and Batrachians, by Leonhard Stejneger. 3. Fishes, by Charles H. Gilbert. 4. Insects, by C. V. Riley. 5. Mollusks, by R. E. C. Steams. 6. Desert Trees and Shrubs, by C. Hart Merriam. 7. Desert Cactuses and Yuccas, by C. Hart Merriam. 8. List of Localities, by T. S. Palmer. Pp. 402, pis. 15, maps 5, figs. 2. 1893. *No. 8. Monographic Revision of the Pocket Gophers, Family Geomyidae (exclu- sive of the species of Thomomys). By C. Hart Merriam. Pp. 258, pis. 20, figs. 71, maps 4 (colored). 1895. No. 9. Not issued. *No. 10. Revision of the Shrews of the American Genera Blarina and Notioso- rex. By C. Hart Merriam. The Long-tailed Shrews of the Eastern United States. By Gerrit S. Miller, jr. Synopsis of the American Shrews of the Genus Sorex. By C. Hart Merriam. Pp. 124, pis. 12, figs. 3. 1S95. *No. 11. Synopsis of the Weasels of North America. By C. Hart Merriam. Pp. 44, pis. 6, figs. 16. 1896. *No. 12. The Genera and Subgenera of Voles and Lemmings. By Gerrit S. Miller, jr. Pp. 84, pis. 3, figs. 40. 1896. *No. 13. Revision of the North American Bats of the Family Vespertilionidae. By Gerrit S. Miller, jr. Pp. 140, pis. 3, figs. 40. 1897. *No. 14. Natural History of the Tres Marias Island, Mexico. General Account of the Islands, with Reports on Mammals and Birds, by E. W. Nelson. Reptiles, by Leonhard Stejneger. Notes on Crustacea, by Mary J. Rath- bun. Plants, by J. N. Rose. Bibliography, by E. W. Nelson. Pp. 97, pi. (map), figs. 2. 1899. *No. 15. Revision of the Humping Mice of the Genus Zapus. By Edward A. Preble. Pp. 42, pi., figs. 4. 1890. *No. 16. Results of a Biological Survey of Mount Shasta, California. By C. Hart Merriam. Pp. 179, pis. 5, figs. 46. 1899. *No. 17. Revision of American Voles of the Genus Microtus. By Vernon Bailey. Pp. 88, pis. 5, figs. 17. 1900. *No. 18. Revision of the Pocket Mice of the Genus Perognathus, By Wilfred H. Osgood. Pp. 72, pis. 4 (inch 2 maps), figs. 15. 1900. *No. 19. Results of a Biological Reconnoissance of the Yukon Region; General Account of the Region. Annotated List of Mammals, by Wilfred H. Os- good. Annotated List of Birds, by Louis B. Bishop. Pp. 100, pis. 7 (inch 1 map). 1900. *No. 20. Revision of the Skunks of the Genus Chincha [Mephitis-]. By Arthur H. Howell. Pp. 62, pis. 8. 1901. *No. 21. Natural History of the Queen Charlotte Islands, British Columbia ; and Natural History of the Cook Inlet Region, Alaska. By Wilfred H. Osgood. Pp. 87, pis. 7 (inch 1 map), fig. (map). 1901. r*No. 22. A Biological Investigation of the Hudson Bay Region. By Edward A. Preble. Pp. 140, pis. 14 (inch 1 map). 1902. *No. 23. Index Generum Mammalium : A List of the Genera and Families of Mammals. By T. S. Palmer. Pp. 984. 1904. (Continued on page 3 of cover) North American Fauna No. 51, U. S. Dept. Agr., Biological Survey Plate 1 U. S. DEPARTMENT OF AGRICULTURE BUKEAU OF BIOLOGICAL SUKVEY NORTH AMERICAN FAUNA ISTo. 51 [JULY, 1928] A TAXONOMIC REVIEW OF THE AMERICAN LONG-TAILED SHREWS (Genera SOREX and MICROSOREX) BY HARTLEY H. T. JACKSON BIOLOGIST, DIVISION OF BIOLOGICAL INVESTIGATIONS BUREAU OK BIOLOGICAL SURVEY UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON 1028 U. 3. SUPERINTENDENT OF OOCUMCNTS 28 ADDITIONAL COPIES OF THIS PUBLICATION MAY BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS U.S. GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. AT 50 CENTS PER COPY CONTENTS Page Introduction 1 Distribution and habitat 2 Habits 5 Food and economic status 10 Young 11 Weight 11 Explanations 12 External measurements 12 Cranial measurements 13 Maturity of skulls 13 Teeth 14 Colors 14 Groups 15 Material examined 16 The family Soricidae 16 Subfamilies 17 Pelages and molts 17 Time of molting 17 Manner of molting 18 Variations 18 Geographic variation 18 Individual variation 19 Sexual variation 20 Age variation 21 Seasonal variation 21 History 21 List of generic names used for American long-tailed shrews 26 Key to genera and subgenera of American long-tailed shrews 27 List of American genera, species, and subspecies of long-tailed shrews, with type localities 27 3enus Sorex 30 Subgenus Sorex t 31 Key to subspecies 32 Sorex cinereus group 37 Sorex fmneus group 60 Sorex arcticus group 66 Sorex pribilofensis group 76 Sorex merriami group 78 Sorex sclateri group 82 Sorex longirostris group 83 Sorex dispar group 88 Sorex trowbridgii group 92 Sorex vagrans-obscurus group 101 Sorex stizodon group 147 Sorex veraepacis group 147 Sorex saussnrei group 153 Sorex ornatus group 163 Subgenus Neosorex 175 Subgenus Atophyrax 192 Jenus Microsorex 200 iterature cited _ Z ~~~ I 211 fndex ZJZZJIJIJZZi.JlJl !___ 233 in ILLUSTRATIONS PLATES [Plate 1, frontispiece; plates 2-13, following page 218] Pl. 1. White-lipped water shrew (Sorex palustris albibarbis). 2. Skulls (dorsal view) of Sorex cinereus, 8. fontinalis, 8. lyelli, S. preblei, S. fumeus, S. arcticus, S. tundrensis, S. pribilofensis, 8. merriami, S. leucogenys, 8. sclaterl, S. longirostris, S. dispar, 8. trowbridgii, 8. vagrans, S. durangae, and 8. obscurus. 3. Skulls (dorsal view) of Sorex obscurus, S. yaquinae, 8. pacificus, S. stizodon, 8. veraepacis, S. macrodon, S. saussurei, S. emarginatus, 8. ventralis, 8. oreopolus, S. ornatus, S. sinuosus, S. trigonirostris, 8. juncensis, 8. myops, S. nanus, and 8. palustris. 4. Skulls (dorsal view) of Sorex palustris, S. alaskanus, 8. bendirii, and Microsorex hoyi. Skulls (ventral view) of Microsorex hoyi, Sorex cinereus, S. fontinalis. 8. fumeus, 8. arcticus, 8. tuxdreusis, S. pribilo- fensis, S. menHami, S. leucogenys, 8. sclateri, and £. dispar. 5. Skulls (ventral view) of Sorex longirostris, S. trowbridgii, 8. ragran-s, 8. durangae, S. obscurus, S. yaquinae, S. pacificus, S. stizodon, S. veraepacis, 8. macrodon, 8. saussurei, S. ornatus, S. palustris, S. alasJcanus, and 8. bendirii. Skulls (lateral view) of Sorex cinereus, 8. fontinalis, S. lyelli, S. preblei, S. fumeus, S. arcticus, 8. tUndrensis, and 8. pribilofensis. 6. Skulls (lateral view) of Sorex merriami, 8. leucogenys, 8. solateri, S. longirostris, S. dispar, S. trowbridgii, S. vagrans, 8. durangae, 8. obscurus, S. yaquinae, S. pacificus, S. stizodon, S. veraepacis, 8. saus- surei, 8. ornatus, S. tenellus, S. myops, S. palustris, S. bcr.dirii, and Microsorex Jwyi. 7. Rostra and upper teeth (lateral view) of Sorex cinereus, S. fontinalis, S. lyelli, 8. fumeus, 8. arcticus, 8. tundremls, 8. araneus, 8. pribilofenr sis, S. merriami, and 8. leucogenys. 8. Rostra and upper teeth (lateral view) of Sorex sclateri, S. longirostris, 8. dispar, 8. troicbridgii, 8. vagrans, S. durangae, S. obscurus, 8. yaquinae, S. pacificus, and S. stizodon. 9. Rostra and upper teeth (lateral view) of Sorex veraepacis, S. macrodon, S. saussurei, 8. emarginatus, S. ventralis, 8. oreopolus, S. ornatu,s, S. tenellus, and 8. nanus. 10. Rostra and upper teeth (lateral view) of Sorex palustris, 8. alascensis, 8. bendirii, and Microsorex hoyi. 11. Upper teeth (ventral surface) of Sorex cinereus, S. fumeus, 8. arcliais, 8. pribilofensis, S. merriami, 8. longirostris, fif. dispar, S. troirbridgii, S. obscurus, S. veraepacis, S. saussurei, S. ornatus, 8. palustris, S. bendirii, and Microsorex hoyi, 12. Lower teeth (lateral view) of Sorex cinereus, S. fumeus, S. arcticus, 8. tundrensiS, 8. pribilofensis, 8. merriami. S. longirostris, S. dispar, 8. troicbridgii, 8. vagrans, S. obscurus, 8. yaquinae, S. stizodon, and S. veraepacis. 13. Lower teeth (lateral view) of Sorex saussurei, S. ornatus, 8. palustris, 8. bendirii, and Microsorex hoyi. Lower teeth (dorsal surface) of Sorex obscurus and Microsorex hoyi. Second and third upper nni- cuspids (ventral surface) of Microsorex hoi/i. VI NORTH AMERICAN" FAUNA TEXT FIGURES Paga Fig. 1. Skull of Sorex showing cranial measurements 13 2. Teeth of Sorex showing principal cusps 15 3. Geographic range of the species and subspecies of the Sorex cinereus group 39 4. Geographic range of subspecies of Sorex fumeus 61 5. Geographic range of subspecies of Sorex arcticus and the species S. tundrensis 6T 6. Rostrum of Sorex hydrodrorrms 75 7. Geographic range of species of the Sorex merriami group 79 8. Geographic range of subspecies of Sorex lorvgirostris 84 9. Geographic range of species of the Sorex dispar group 89 10. Geographic range of subspecies of Sorex trowbridgii 93 11. Geographic range of subspecies of Sorex vagrans and the species S. durangae 102 12. Geographic range of the species Sorex obscurus 116 13. Geographic range of five subspecies of Sorex obscurus 118 14. Geographic range of eight subspecies of Sorex obscurus 129 15. Geographic range of Sorex yaquinae and of subspecies of S. pacificus 141 16. Geographic range of subspecies of Sorex veraepacis and the species S. macrodon 148 17. Geographic range of subspecies of Sorex saussurei and the species S. emarginatus 154 18. Geographic range of Sorex sclateri, S. stizodon, S. ventralis, and S. oreopolus 161 19. Geographic range of species and subspecies of the Sorex ornatus group 164 20. Foramina magna of Sorex obscurus and S. ornatus 165 21. Geographic range of species and subspecies of the Sorex palustris group 177 22. Geographic range of subspecies of Sorex bendirii 193 23. Geographic range of subspecies of Microsorex hoyi 201 24. Dorsal view of right third upper incisors of Sorex araneus, S. cinereus, and Microsorex hoyi 201 No. 51 NORTH AMERICAN FAUNA July, 1928 A TAXONOMIC REVIEW OF THE AMERICAN LONG-TAILED SHREWS (Genera SOREX and MICROSOREX) By Hartley H. T. Jackson INTRODUCTION No other group of American mammals having a wide distribution, and in many localities an abundance of individuals, is so little known to the nonprofessional mammalogist as the long-tailed shrews belong- ing to the genera Sorex and Microsorex. Neither are the individuals of any other group of common mammals so seldom seen in life by the professional field mammalogist, nor are the habits of such indi- viduals less known to him. And probably no other group of mam- mals offers so many difficulties and problems in the way of taxonomic study. These arise from numerous features, no one of which may be peculiar to shrews, but the combination occurs in no other large group of mammals. The small size of shrews makes slight errors of measurements, external or cranial, large proportionally, and makes necessary the constant use of the microscope for the study of cranial and dental characters. The absence of color pattern, and a definite color variation between species that in many cases seems scarcely more than individual, makes identification by color alone possible only in a comparatively few instances. The early anastomosis of the separate cranial bones into one compact whole, which occurs while the animal is yet juvenile, makes comparison of the various individual bones of the skull impossible; all outlines of the individual cranial bones are lost in adult shrews. The simple dentition of shrews offers little opportunity for differentiation of form or cusps. Variability of skulls and teeth due to age of the individual is excessive, so much so that the skull of an adult animal may appear entirely unlike the skull of a young animal of the same subspecies; great care must there- fore be exercised in making comparisons to be certain that the indi- viduals are of corresponding age and development. Finally, there is a wide range of individual cranial variation, particularly in the size of the skull ; shrews seem more prone to produce " runty " skulls or abnormal dentition than most other mammals. In popular parlance the American long-tailed shrews might super- ficially be divided into four groups, namely, the long-tailed shrews proper, the saddleback shrews (the arcticus group), the water shrews {palusti'is group), and the marsh shrews (bendirii group). So little are shrews known to the layman that when actually seen they are gen- erally confused with mice, though in reality as closely related to 2 NORTH AMERICAN FAUNA [No. 51 wolves or foxes as to mice. Occasionally, however, they are distin- guished by modifying terms, as in sections of Alaska, where O. J. Murie reports that they are known as " sharp-nosed mice " ; or in the Gallatin country of Montana, where C. W. Richmond says they are called " dormice," or in the Jackson Hole region of Wyoming, where, according to Alexander Wetmore, the water shrews are known locally as " fish-mice." In many localities shrews are known as moles " or " young moles," and the water shrew as " water mole." One of the most peculiar local names is reported by A. K. Fisher, who says that in Dismal Swamp, Va., Sorex longirostris fisheri is called " smell fast." Shrews are known to some of the Indian tribes, who have special names for them; thus, the Chippewas of northern Wisconsin and Minnesota know the little animals as " oke-pa-ku-kue " or " oke- pa-kue-kue " ; the Klamath Indians call S. vagrans " shu-zhi " ; Ver- non Bailey (manuscript notes) reports that the Taos Indians of New Mexico recognize the shrew under the name " pah-ka-che-una "; and the Kwakiool name for the mole or shrew in British Columbia is " kiap-kepu-s " (Dawson, 1888, p. 93). The common shrew of eastern Canada (S. cinereus) was known to the Labrador Eskimos as the ukounavik (Packard, 1866, p. 266). According to Nelson the long- tailed shrew was known to the Alaskan Eskimos as the " u-gu-gi- nuk," and when it was found strayed out on the sea ice by them, it was the subject of a curious superstition. They claim that there is a kind of water shrew living on the ice at sea which is exactly like the common land shrew in appearance, but which is endowed with demoniac quickness and power to work harm. If one of them is disturbed by a person, it darts at the intruder and burrowing under the skin, works about inside at random and finally enters the heart and kills him. As a consequence of this belief the hunters are in mortal terror if they chance to meet a shrew on the ice at sea, and in one case that I knew of a hunter stood immovable on the ice for several hours until a shrew he hap- pened to meet disappeared from sight, whereupon he hurried home, and his friends all agreed that he had had a very narrow escape. (Nelson, 1887, p. 271.) DISTRIBUTION AND HABITAT The genus Sorex is distributed throughout a large portion of the continents of Europe, Asia, and North America. It is absent from the extreme southern and torrid sections and more abundant both in species and individuals in the north-temperate and boreal parts. In North America the range of the genus extends over the entire northern part of the continent from the Arctic Ocean south to northern Florida and Alabama in the eastern United States ; Indiana, Illinois, and Nebraska in the middle United States; and in the higher elevations of the mountains of the West to Guatemala. The range of the water and marsh shrews (subgenera Neosorex and Atophyrax) extends from southern Quebec, Ontario, Manitoba, southern Northwest Territories, northern British Columbia, and southeastern Alaska, south to Pennsylvania, Wisconsin, and Minne- sota in the eastern United States, to Arizona and New Mexico in the Rocky Mountains, to the Mount Whitney region, California, in the Sierra Nevada, and along the Pacific coast to northern California. The range of the genus Microsorex extends from northern Quebec, northern Ontario, and central and western Alaska south to the northern United States. 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 3 The characteristic habitats of long-tailed shrews are moist situa- tions with an abundance of vegetation, such as mossy and grassy banks along streams, meadows, sphagnum bogs, and damp woods, particularly of coniferous trees. There is, however, considerable variation in the dominant habitats among the various species, while certain of the common forms seem less restricted in habitat and may be found at times in associations that could hardly be considered characteristic of the particular species. For example, the species cinereus is of wide distribution and ordinarily may be expected in the normal shrew habitat of damp woods, mossy banks of streams, coniferous swamps, and sphagnum bogs; yet there are numerous records of specimens of this species taken in houses and other build- ings, particularly in the far north, and of other individuals in dry woods and meadows in the eastern United States. The species fumeus is most frequently reported from hemlock woods, but may occasionally occur in meadows near timber. As a general rule, Sorex vagrant prefers damp meadows and S. obscurus the mossy banks of streams; at any rate, in many regions where the two species occur, as, for example, S. v. amoenus and S. o. obscurus in the Sierra Nevada, a majority of each species is trapped in the respective habitats indicated above. In regard to the trapping of a specimen of S. o. obscurus in the Manzano Moun- tains, N. Mex., J. H. Gaut remarks in his field report : A gopher trapped in a tunnel made by the animal was discovered to have a small opening in its side with parts of the intestines gone and immediately upon discovering this fact I removed the specimen and replaced it with a small trap baited with some of the remaining portions of the intestines. An hour later the trap was visited and found to be holding fast one of these little shrews. The hole in which the trap was placed was thoroughly covered with dirt, in such manner that no animal could possibly have entered at that particular place. In the coast region of Washington it is known that S. o. setosus frequently inhabits the runways of moles (Scapanus). The species obscurus is also occasionally found in buildings, but such cases must be considered exceptional. Habitat records of the rare Sorex dispar would indicate that the species usually lives among rocks. It is said that the type specimen was trapped among some large, angular rocks at the head of a wooded talus of loose rock just below low cliffs, which shaded the spot and kept it cool, and that a second specimen was taken in a crevice be- tween some rocks on the bare, open summit of Mount Marcy, N. Y. (Batchelder, 1896, p. 133.) Mearns also records the species as being trapped in hollows under mossy stones, usually in wet balsam or spruce woods, or in weedy swamps, in the Catskill Mountains, N. Y. (Mearns, 1898, p. 356); and at Mount Grejdock, Mass., Copeland caught one specimen under a rock at the edge of a moist grassy clear- ing surrounded by woods, and another near a small brook in swampy woods of spruce, hemlock, and scattered birches thickly carpeted with sphagnum (Copeland, 1912b, p. 162). The first known specimens of S. gaspensis were all caught near small streams. One was trapped among dead tree stumps that were lying partly submerged and almost surrounded by water and shaded by overhanging spruce trees. A second was procured in a trap set in such a position at the foot of a low cliff facing a stream that the animal must have passed through 4 NORTH AMERICAN FAUNA [No. 51 shallow water to reach the trap; a dark, damp forest spread on all sides; deep moss covered the ground and obscured the stream in places, and many trees bore hanging moss. The third specimen was caught along a small stream that came through a narrow canyon on a cool, north slope covered with a forest of spruce and balsam ; the trap was set among driftwood and wet leaves between bowlders. (An- thony and Goodwin, 1924, p. 1 ; Goodwin, 1924, p. 252.) One of the western species, Sorex trowbridgii, may be found in the regulation damp, mossy habitat of long-tailed shrews, but it also dis- plays a marked preference for the drier woods, and in parts of western Washington and Oregon it is frequently most abundant in dry fir timber. The two closely related species Sorex merriami and S. leucogenys have an unusual habitat for shrews, in that they have been found only in an arid sagebrush association. The type and only known specimen of /S. tenellus was also collected among loose rocks on a dry hillside a long distance from water. The water shrews (subgenus Neosorex) are seldom found at any great distance from water, which may be a lake or pond, a brook, or merely a pothole in a swamp, bog, or forest. They seem to prefer a more or less wooded habitat and are rarely found in marshes devoid of bushes or trees. They tend to be more boreal than mem- bers of the subgenus Atophyrax and in the western United States are usually found at the higher altitudes in the Canadian Zone. Nevertheless, in March, 1920, G. G. Cantwell collected three specimens at an altitude of about 300 feet at Rockport, Skagit County, Wash. The streams in which these specimens were caught, however, came down from mountains 4,000 to 6,000 feet high not more than 2 miles distant. The marsh shrews (subgenus Atophyrax) seem to be less truly aquatic than the members of the subgenus Neosorex, and although found in damp woods and other habitats such as are frequented by Neosorex, they also occur in tule or sedge marshes. At Lake Cush- man, Wash., in the midsummer of 1894, C. P. Streator caught three specimens in traps baited with rolled oats and set on an ^ld beaver dam. Shrews of the genus Microsorex do not seem to differ particularly in their habitat preference from certain members of the genus Sorex. as, for example, cinereus. Such reports as are available indicate that the pigmy shrews are most frequently found in damp woods, sometimes of deciduous trees. But they are also found in tamarack swamps and muskegs, and occasionally in marshes and even in dry woods. The type specimen of Microsorex h. washingtoni was found dead in a trail in dry pine woods. E. A. Preble reports one of this genus taken January 4, 1904, in the potato cellar of a dwelling house at Fort Simpson, Northwest Territories, Canada. Of the first two specimens of M. h. wirmemana) collected, one (type specimen) was captured in the decayed interior of a fallen log in mixed woods of maple and other deciduous trees and the other was found in the decayed heart of a dead chestnut tree on a dry hillside some dis- tance from water. The third specimen of wiw/nemama was captured by G. W. J. Blume, who sent it to Wirt Robinson, who in turn pre- 1928] KEVIEW OF AMERICAN LONG-TAILED SHREWS 0 sented it to the United States National Museum. In a letter to Colonel Robinson Mr. Blume states : As I recall the capture of the shrew, I noticed movements in the leaves on one of- the hillsides on my place at Alta Vista, Va., and thinking it to be a mole plowing the surface as they sometimes do in gravelly or very rooty ground, I started'to scratch among the leaves to catch it. The shrew started to run and I caught it in my hands. I think there was a rock pile not far distant which was probably its home. It was in dry. wooded land, probably not over 100 yards from running water but not close to a swamp. There was no dense underbrush, but plenty of natural concealment afforded by the leaves, rocks, old logs, etc., in the vicinity. HABITS Long-tailed shrews are such elusive midgets and such meager defi- nitely planned research has been done on their life history that com- paratively little is known about their habits. That the various species have certain general habits more or less in common is self-assertive, but that the different species also have specialized habits varying to meet their different habitats and environments is also evident. For example, one could hardly assume that the habits of the semi- desert species, Sorex merriami, would be similar in detail to those of S. dispar, which inhabits the comparatively cool, humid coniferous forest region of the eastern United States. Yet almost nothing is known about the specialized habits of either of these species. Shrews are the active, vicious, voracious little imps of the mammal world. They are largely nocturnal, but are not infrequently active during the daytime, particularly under the snow in winter or during cloudy weather at any season of the year. They are apparently active during the entire winter and do not hibernate, although they have small hibernating glands, and it has been erroneously written that they do hibernate. (Arnback-Christie-Linde, 1907, p. 466.) They live for the most part in little burrows and runways underneath logs, rocks, leaves, and grass, where they hunt insects and worms. These run- ways may be made by the shrews themselves or by various species of mice or other shrews. At Tuckerton, N. J., in the summer of 1892, E. A. Preble captured five specimens of S. cinereus under one of several small haystacks scattered over the meadow's. Some of these were kept alive in a deep can for several hours. They con- stantly moved their long snouts in every direction, apparently depending more on the sense of touch and smell than on sight. Resi- dents stated that during especially high tides at this place these shrews would be drowned out of their retreats and would fairly swarm on the driftwood. (Preble manuscript.) Long-tailed shrews are exceedingly quick and active and move with a queer, jerky, trotlike run, starting and stopping abruptly. They may be considered almost strictly terrestrial, although they occasion- ally climb small branches of very low bushes, fallen trees, or herbs. Morris M. Green writes (manuscript) that while watching a deer runway on the north branch of Moose River in the Adirondacks, N. Y., during the summer of 1894, he saw a little shrew no bigger than a thimble, which climbed up a fern stock within 5 feet of him. Another shrew went through his pail of fishworms and ate every one of them. Though in no sense aquatic (except the subgenera Neosorex 6 NORTH AMERICAN FAUNA [No. 51 and Atophyrax), they are good swimmers when occasion demands it of them, Long-tailed shrews evidently have a wide local range, as is witnessed by Nelson's observations in Alaska. After snow falls they travel from place to place by forcing a passage under the snow, and frequently keep so near the surface that a slight ridge is left to mark their passage. On the ice of the Yukon I have traced a ridge of this kind over a mile, and was repeatedly surprised to see what a direct course the shrews could make for long distances under the surface. These minute tunnels were noted again and again crossing the Yukon from bank to bank. (Nelson, 1887, p. 271.) These little animals are exceedingly savage and voracious and will fight and devour one another upon least provocation. Merriam's account of how he confined three of them under a tumbler is familiar to many students : Almost immediately they commenced fighting, and in a few minutes one was slaughtered and eaten by the other two. Before night one of these killed and ate its only surviving companion, and its abdomen was much distended by the meal. Hence in less than eight hours one of these tiny wild beasts had at- tacked, overcome, and ravenously consumed two of its own species, each as large as itself. (Merriam, 1884a, p. 76; 1884c, p. 174.) Over considerable periods of time, these little gluttons, when in captivity, have been known to eat their own weight in meat on an average of once every three hours. Early in the summer of 1900, W. H. Osgood caught two shrews in the same trap on Vancouver Island, British Columbia. One of them, not killed by the trap, proceeded to devour the other and had nearly accomplished it when Osgood visited the trap. H. H. Sheldon reports (manuscript) an instance in August, 1919, at Ogema, Wis., in which the entire tail of a shrew {Sorex c. dnereus) was all that remained in a trap, and about a foot away he found one dead with tail intact but with its head bruised and part of the skin torn from the nose and jaws. This one had evidently fought the one in the trap, the latter losing its tail but winning the fight. George G. Cantwell, in his field report from the Mount Rainier region, Wash., for the summer of 1919, states that on one occasion while he was setting traps beside a small stream a trap on the opposite side containing a freshly caught mouse was visited by a shrew (S. vagrans), which at once started feeding on the mouse. On account of its poor eyesight, the shrew failed to detect the observer only a few feet away, but as a twig snapped the little animal at once dodged into a hole in the bank and did not appear again. In a letter dated July 17, 1889, at Plover Mills, Ontario2 R. Elliott writes that on May 21, 1888, he captured one (tS. c. anereus) under a small log among dry leaves. He placed it in a large bottle with plenty of air and gave it two earthworms, each of which it took by the " tail " and rapidly nipped through and through to the head and left it dead. The shrew afterwards ate part of one of the worms. Mr. Elliott then gave the shrew a May beetle, which it instantly attacked viciously. The sharp feet of the insect seemed to irritate the shrew to an inordinate degree, and at the end of two or three minutes the beetle was torn and entirely eaten. Ten minutes afterwards it was given another May beetle, which, too, was almost entirely eaten, the head and elytra alone remaining. About half an hour later the shrew died. Very little is known in regard to the nests and home life of long- tailed shrews. They build nests of grass and leaves under logs, in 1S2S] REVIEW OF AMERICAN LONG-TAILED SHREWS 7 stumps, and similar situations, but few of these have been found, much less critically studied. On October 14, 1924, Stuart Griddle found eight young of Sorex cinereus haydeni several days old dead in a nest made of brome-grass leaves, with a few ground-cherry leaves on the outside, situated under a brome-grass sheaf near Trees- bank, Manitoba. In the nest there was also the anterior part of the skull of an adult. Judging from the dates of collection of pregnant specimens, the height of the breeding season is June, July, and August, although a few have been taken as late as the last of Septem- ber that had been nursing, and others contained embryos as early as March 29 in Inyo County, Calif., and the middle of May on Prince of Wales Island, Alaska. From the data available as based upon the number of fetuses in specimens trapped, as recorded by collectors, it would appear that the number of young varies in different species. From these records Sorex cinereus appears to have the largest litters, with an average of 12, for 8 specimens, the minimum being 4 and the maximum 10. The average for 5 specimens of S. palustris navigator is 6 fetuses, minimum 5, maximum 7. In 8 litters S. vagram has a minimum of 3 and a maximum of 9, with an average of 5.8. Sixteen records for S. obscurus show an average of 5.4 per litter, with a minimum of 4 and a maximum of 8; of these 16 litters one-half were of 5 each. A single specimen of S. arcticus had 6 embryos, as did also one of S. ornatus. Three embryos were found in a specimen of S. t. trow- bridgii, and 4 in one oi S. t. mariposae. The smallest number of fetuses recorded is 2 from the specimen of S. Jeucogenys from Mount Magruder, Esmeraldo County, Nev. This lone record may give a clue to one of the possible causes for the apparent scarcity of shrews of the merriamd group. Long-tailed shrews are preyed upon by numerous species of ani- mals, although comparatively few of their enemies will eat them except in cases of extreme hunger. It is well known that domestic cats kill numbers of shrews but seldom eat them, and in olden times it was believed that shrews were poisonous to cats. Fragments of shrew skulls and bones are frequently found in owls' pellets; the only record of Sorex longirostris in Alabama is a complete animal, now in alcohol, taken from the stomach of a barred owl. This same species of shrew has been taken from the esophagus of a hooded merganser. (Audubon and Bachman, 1854, p. 250.) Hawks also are known to prey upon shrews occasionally, and one has been re- corded from the stomach of a bear. (Osgood, 1907, p. 63.) Certain fishes, particularly trout, may at times catch them, and A. H. Twitchell in a letter dated September 18, 1917, at Flat, Alaska, re- ports finding during August of that year the remains of six shrews, probably S. tundrensis, in the stomachs of three graylings. Aside from the numbers killed by natural enemies, long-tailed shrews seem to have an unusually high mortality rate, as compared with many mammals, as many of them are accidently trapped in ditches, springs, and wells. It is not at all unusual to find several dead shrews in an uncovered well or spring, and such accidental deaths have been responsible for several specimen records. Although members of the genus Sorex are pugnacious and physically strong in proportion to their size, they seem sensitive to any external 8 NORTH AMERICAN FAUNA [No. 51 shock or stimulus, and individuals are not infrequently found dead on the surface of the ground. Undoubtedly some of these are killed by other animals and abandoned as unfit for food, but many show no signs of injury and appear to have perished merely from nervous shock, extremes of temperature, or the like. O. J. Murie in a field report states that a Mr. Quinn, of McGrath, Alaska, while traveling in the vicinity of Rainey Pass one winter during a con- tinued severe cold spell, found numerous shrews frozen along the trail. The writer recalls a case late in August, 1919, at Ogema, "Wis., when one of the field men in his party, who chanced to come upon a specimen of Sorex arcticus laricorwn, which was running across a road, dropped a felt hat over the animal. He then carefully raised the brim of the hat, expecting to capture the shrew alive, but to his surprise the animal was dead, though apparently not touched by the hat. An interesting account of sensitiveness in a shrew has been described by Gillman (1876) as follows: In the heavily timbered forest in the neighborhood of Cheboygan, Mich., on a cold day in October, 1875, I caught a characteristic full-grown specimen of Thompson's shrew (Sorex Thompsoni Baird).1 The pretty little creature had been busy about an old decayed stump, where it seemed to have its home. It uttered no audible cry, though at first it made several hostile demonstrations, endeavoring to escape, and, seizing my fingers in its mouth, tried to bite them, but the delicacy of its teeth rendered the attempt futile. Having no suitable place in which to deposit it, I carefully wrapped it in paper, allowing its head to protrude, and held it in my hand. Some sportsmen were out shooting on the bay about a mile off, and the reports of their guns came to us from time to time, generally so much muffled by the distance as to be barely distinguish- able, yet the shrew invariably responded to each detonation with a quick, spas- modic movement, evidently of alarm. Holding the animal as I did, the move- ment was immediately perceptible. Though aware that the acuteness of the auditory organs of these animals and their allied genera is most wonderful, I was hardly prepared for so unequivocal a proof of its extreme sensitiveness, which, under the circumstances, I was enabled to test repeatedly in this individual Sorex. It was my intention to preserve the animal alive, and take it with me on my return home for further experiment and study of its habits ; but, to my regret, on unfolding the paper on my way to the house at which I was staying I found the shrew had died. I have little doubt that its death was caused by fright, as I handled it most carefully so as not to hurt it. Long-tailed shrews seldom use a call note and, as a rule, are not noisy animals. The writer has heard Sorex c. cinereus utter a series of sharp squeaks and also a weak purrlike grunt. Charles W. Rich- mond in his field report states that he observed a shrew (probably S. o. obscurus) in Gallatin County, Mont., and says that it frequently stopped to sniff the air suspiciously and occasionally uttered a " little snort." When fighting, either against members of their own species or other enemies, they frequently indulge in much squeaking and make a considerable noise for such mites. In the Mount Hood region, Oreg., in the spring of 1919, George G. Cantwell caught a S. t. trowbridgii in his hand from under a flake of bark. The shrew fought vigorously " with much squealing " and finally squirmed out of the grasp of its captor and disappeared like a ghost, for no con- spicuous hole or apparent cover was in the vicinity to afford concealment. 1 This name is a synonym of Microsorex hoyi thompsoni (Baird), but Gillman's animal was probably Sorex cinereus cinereus Kerr. 1928] EEVIEW OF AMERICAN LONG-TAILED SHREWS 9 The habits of the water shrews (subgenus Neosorex), so far as known, do not differ essentially from those of other shrews, except in adaptations to a more aquatic habitat. Water shrews rank high among the best swimmers of the nonmarine mammals, although op- portunities for observing them in the act of swimming are rarely presented. They can swim, dive, float, run along the bottom of a pool or creek, or actually run on the surface of the water with the greatest ease. In a bog near Rhinelander, Wis., in August, 1906, the author saw one run a distance of about 5 feet across a small pool, the surface of which was glossy smooth. • The body and head of the animal were entirely out of water, the surface tension of the water supporting the shrew, and at each step the animal took there ap- peared to be held by the fibrillae on the foot a little globule of air, which was also discernible in the shadow at the bottom of the pool, exactly as one might notice in the case of the water strider {Gerris remigis). (Miall, 1903, p. 12, 349.) It is probable, however, that this water-walking feat can be accomplished by water shrews only when the water is very still and quiet, and in running or rough water it would seem that the animals would be required to swim. Walter P. Taylor in his field report for Cat Creek, Clallam County, Wash., states that on the evening of September 5, 1921, he saw a Sorex pahistris navigator, which he at first mistook for a frog, in a shallow " running " creek. He noticed that it was walking rather jerkily through the water, at first in water not so deep but that it could touch the rocks beneath, but soon in water that must have been beyond its depth. It did not sink, but remained half exposed, " walking " rapidly along on top of the water. The animal had a dry, fluffy appearance. Edward A. Preble saw one running on the water, July 27, 1910, in a small creek some 25 miles east of Telegraph Creek, British Columbia. He noted that the shrew followed the edge of the stream close to the bank and seemed scarcely to sink at all below the water line but gave the impression rather of running on the water film, progressing at a good rate and making only very slight ripples. At the Three Sisters, Oreg., in July, 1914, Vernon Bailey caught a /S. p. navigator in his hands, tied a string to its leg, and put it in the water. At first it fluffed out its fur and " sat on top of the water like a duck." Lowering itself into the water, it swam rapidly, though using but one foot, to a log, upon which it climbed. Then it would dive and dart about under water like a silver fish, going to the bottom and under logs and sticks, apparently seeing or knowing its way and just where to hide. According to Bailey, it swam with relatively greater speed and skill than the otter, which always seemed to him the most wonderful mammal in the water. George G. Cantwell observed one in the Mount Rainier region, Wash., in the summer of 1919, which ran rapidly through the shal- low water of a swift mountain stream, and swam or dived through the deeper pools with great speed, using all four feet in swimming with the same motions it used in running over the ground. While under water the thick coat of fur of the animal was surrounded by a silvery layer of air, and when the animal came to the surface again it appeared to be dry. 10 NORTH AMERICAN FAUNA [No. 51 A. Brazier Howell (1924, p. 27) states in regard to an individual of S. p. navigator that "it dived and swam under a bank so quickly that I had opportunity to be sure of nothing except that while swim- ming it kicked both hind feet in unison after the manner of a frog." This method of swimming is at variance with the observations of others. An interesting note on this same species made by the late Theodore Roosevelt is extracted from a letter (manuscript) of his dated No- vember 26, 1888 : I was near Kootanei Lake, in British Columbia, and while taking lunch near a small rapid brook I saw a Water Shrew swimming down it. While swim- ming its body looked like a flattened disc studded with silvery bubbles. It ran along the bottom and over the rocks very fast, and swam and dived well. I saw it catch a very minute fish and eat it on a wet, water-washed stone. At last by an under grab I caught it. Its tail was conspicuously longer than its head and body, and it was, without doubt, a Neosorex. I skinned it with my pocket knife and put a little hoop in the reversed skin, but as I was traveling very light, had to put it in my pocket. That afternoon I shot a bear and camped by it, being very hungry. I put the little Shrew skin out on a log and turned away a moment, and to my horror, in the interval the Indian who was with me threw the log into the fire, and of course the skin vanished. I was really very sorry. FOOD AND ECONOMIC STATUS Shrews are chiefly insectivorous in their food selections, but they will eat other flesh and occasionally vegetable matter. Stomach ex- aminations of true Sorex, including representatives of the species cinereus, fumeus, longirostris, vagrans, obscurus, and trowbridgiiy have shown the following among the contents: Hymenoptera, Cole- optera, Diptera (both larval and adult), caterpillars, crickets, spiders, hair and flesh of shrews and mice, and moss, seeds, and other vege- table material. That shrews are not entirely averse to certain vege- table food is attested by the manner in which they will eat rolled oats placed on baited traps. Their food probably does not vary much with the seasons, for shrews, being active in winter and feeding as they do mostly in burrows in the ground and in runways under the leaves, obtain dormant and pupating insects during the colder months of the year. Shrews are known to eat earthworms, although stomach examinations have not shown earthworms in the contents. The food of the water shrew (subgenus Neosorex) does not appear from stom- ach examinations to differ much from that of other shrews. The water shrew is known to eat small fish, however, and is also reported to feed upon fish eggs. It seems probable that a detailed study of its food would show the insects consumed to be more of aquatic species than those taken by other shrews. Also the stomach examinations of marsh shrews (subgenus Atophyrax) show no marked differences from those of other shrews, although 35 per cent of the contents of one stomach was snails. Nothing is known of the food of the mem- bers of the genus Microsorex. Since shrews are such voracious eaters and feed principally upon insects with an occasional dessert of young mice, they are of consid- erable economic value in holding down certain pests of agriculture and forestry. Unfortunately, they are usually mistaken for mice by the layman and killed on sight. In parts of the extreme north, par- 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 11 ticularly in Alaska, long-tailed shrews are sometimes reported, as a nuisance on account of their climbing into caches of fresh meat or fresh or dried fish, some of which they eat and the remainder ruin with their filth. In these same regions they may also become a nuis- ance in houses during winter. Water shrews, also, are occasionally reported to do damage in fish ponds and trout streams by destroying fish and fish eggs, but these shrews are not plentiful enough to do any serious damage except locally, and then but rarely. On the whole, shrews are among the most beneficial mammals. YOUNG Only scant information is available on the young of the genera Sorex and Microsorex. The litter of eight of Sorex cinereus haydeni collected by Stuart Criddle, at Aweme, Manitoba, and already re- ferred to {antea, p. 7), are the youngest long-tailed shrews that the writer has seen. Although of indeterminable age, they are un- doubtedly at least 10 or more days old and more than half the size of adults. They are covered with very short hair, and the tails are proportionately somewhat shorter than in adults. In all probability shrews are born blind, hairless, and, relatively speaking, but slightly developed. Following birth, however, it would seem that develop- ment and growth is comparatively rapid, although they remain in the nest until well along toward maturity. It is this habit of re- maining in the nest until so nearly mature that makes young shrews so scarce in collections. In the Criddle specimens the partly developed molariform teeth and first incisors appear above the alveola, while the unicuspids are still covered by the dermis, and difficult to detect in gross examination. The only other young examined is a litter of five Sorex longirostris fisheri collected in May, 1905, by W. L. Ralph and J. W. Daniel, jr., in Dismal Swamp, Va. The young in this litter appear to be a few days older than those in the litter of S. cinereus haydeni, and offer no juvenile peculiarities not shown in the other litter. WEIGHT Among the species of long-tailed shrews are the smallest of American mammals, and even the larger forms are no bigger than some of the smaller species of mice. The smallest American shrew is Microsorex hoyi winnemana, which may also be the smallest mammal known. There are no weights available for this subspecies, however, nor for any others of the pigmy shrews (Microsorex), ex- cept two male specimens of M. h. hoyi collected by Bernard Bailey at Elk River, Minn. One of these, taken on March 25, 1926, weighed 2.1 grams; the other, collected two days later, weighed 2.9 grams. Neither of these was fat nor was the stomach of the latter so full as that of the first, although the latter weighed more. Two females of Sorex cinereus haydeni weighed by Vernon Bailey at Walhalla, N. Dak., each weighed 3.3 grams, while a third collected 74285—28 2 12 NORTH AMERICAN FAUNA [No. 51 at that time weighed 3.6 grams. Bailey also weighed specimens of jS. c. cinereus at Michigamme, Mich., late in the summer of 1923, the weights ranging from 3.5 to 5.5 grams. In the Stikine River region of British Columbia, in 1919, Joseph Dixon weighed several speci- mens of S. c. cinereus, 15 adults of which averaged 4.5 grams, with a minimum of 2.8 grams and a maximum of 6,1. Part of these speci- mens were approaching S. c. streatori, which is a larger subspecies than typical cinereus. Two males of JS. c. cinereus collected and weighed by O. J. Murie at Fairbanks, Alaska, January 19, 1922, balanced at 2.7 and 2.8 grams, while a female October 14, 1921, from the same locality weighed 2.84 grams. Another female of this same subspecies collected by Murie February 18, 1922, on the South Fork of the Kuskokwim River, Alaska, weighed 2.85 grams. Two males of Sorex v. vagrans collected at Puyallup, Wash., July 4, 1914, were weighed at 7 grams each by T. H. Scheffer. Eighteen specimens of S. o. obscurus from the Stikine region of British Co- lumbia, as weighed by Joseph Dixon in the summer of 1919, aver- aged 6.9 grams, with a minimum of 4.8 grams and a maximum of 8.7 grams. Seventeen western water shrews {Sorex palustris navigator) col- lected by parties from the Museum of Vertebrate Zoology, University of California, in various parts of the Sierra Nevada, but chiefly from the Yosemite region, California, averaged 12.3 grams, the minimum being 9.1 grams and the maximum 19.5 grams. A single individual of the marsh shrew {S. b. bendirii), a female collected October 14, 1914, at Puyallup, Wash., was found by T. H. Scheffer to weigh 12 grams. The above data must not be taken for more than their actual value, for the weighing of a few specimens in the field is unsatis- factory for comparative use. Certain individual specimens may be fat, others lean or emaciated ; some may be heavy with young, others worn and of light weight from the care of young; young and old may offer different weights; some individuals may be gorged with food, others empty ; even the molting process might affect the weight of the animal. For purposes of comparative weights of different species it is therefore essential to weigh individuals that are in every respect in corresponding physical condition, and large series of them. The foregoing weights are therefore merely suggestive of the com- parative weights of a few species. EXPLANATIONS EXTERNAL MEASUREMENTS External measurements of shrews, unless otherwise stated in the text, are in millimeters and are those made by the collector from the animal in the flesh. The following have been used : Total length. — Tip of nose to end of terminal tail vertebra. Tail vertebrae. — Base of tail at superior surface to end of terminal tail vertebra. Hind foot. — Posterior border of heel to apex of longest claw. 192S] REVIEW OF AMERICAN" LONG-TAILED SHREWS 13 CRANIAL MEASUREMENTS Cranial measurements, unless otherwise stated, were made by the author with a vernier caliper. The following (fig. 1) have been employed : Gondylobasal length. — Antero-posterior diameter of skull from anterior median point between bases of first upper incisors to most posterior point of occipital condyle. Palatal length. — Greatest antero-posterior diameter of palate in median line. (This measurement was taken by use of fine-pointed dividers on vernier caliper.) Cranial breadth. — Greatest lateral diameter of skull. Fig. 1. — Skull of Sorex bendirii palmeri. showing cranial measurements employed. Enlarged three diameters Cranial breadth, G-G'. Condylobasal length, A-H. Interorbltal breadth, E-E'. Palatal length, A-F. Maxillary breadth, C-C. Maxillary tooth row, B-D. Interorbital breadth. — Least lateral diameter of skull measured just posterior to maxillary processes. Maxillary breadth. — Greatest lateral diameter of skull through maxillary processes. Maxillary tooth row. — Antero-posterior diameter of upper tooth row between anterior border of second incisor and posterior border of last molar measured at alveolar border. MATURITY OF SKULLS On account of the great differences in the skulls of shrews of differ- ent ages it is essential in making comparisons to have specimens of approximately the same maturity. There are, of course, no sharp 14 NORTH AMERICAN FAUNA [No. 51 age-division points in the life of the animal, but in making compari- sons skulls have been classed as those of animals that were immature, young adult, adult, and old adult. Skulls of immature and old adult animals show more individual variation than those of the young adult and adult and are therefore less satisfactory for taxonomic purposes. In general terms these four classes of skull maturity may be defined as follows : Immature. — Brain case usually moderately high, and unflattened, with sutures not distinctly closed; no sagittal or lambdoidal ridge; teeth usually not fully developed, unworn; first upper incisors pro- truding much beyond premaxillae anteriorly. Young adult. — Brain "case usually high and unflattened, with sutures closed ; sagittal ridge absent or weakly developed, lambdoidal ridge absent ; teeth fully developed, unworn ; first upper incisors pro- truding much beyond premaxillae anteriorly. Adult. — Brain case usually slightly flattened, with sutures closed; sagittal ridge moderately developed, lambdoidal ridge absent or weakly developed; teeth fully developed, usually unworn or slightly worn, sometimes moderately worn; first upper incisors protruding slightly beyond premaxillae anteriorly. Old adult. — Brain case flattened, with sutures closed; sagittal and lambdoidal ridges both usually well developed; teeth usually much worn ; first upper incisors scarcely protruding beyond premaxillae. TEETH Unless otherwise specified, comparisons of relative sizes of uni- cuspidate teeth are as they are viewed from an extero-lateral aspect, while comparisons of relative sizes and shapes of molariform teeth are as the upper molariform teeth 'are viewed from an inferior aspect (that is, looking dorsad). In the detailed examination of teeth a binocular microscope was used, the most satisfactory magnification being obtained with No. 1 oculars and a 40-millimeter objective. The nomenclature of the tooth cusps and other principal elements of the molariform teeth can be determined from the accompanying diagram. (Fig. 2). * COLORS The names of colors used throughout the text are those of Ridg- way (1912). In some cases, where it has been impossible to match the colors of specimens exactly with those of Ridgway, other modi- fying or comparative terms are used. In making comparative studies of the color of mammals, especially those with glossy or iridescent fur, it is essential always to view each specimen from approximately the same angle and to have the light rays from an approximately constant angle. In the author's color studies of moles, the animal was viewed from the anterior end. (Jackson, 1915, p. 20.) In making color observations upon shrews the animal has' been viewed from the posterior end. Diffused day- light from a window was allowed to strike the shrew at an agle of 30° to 45° anterior to a plane perpendicular to the longitudinal axis of the animal. The shrew was then viewed at varying angles, usually slightly laterally, from the light rays but always posteriorly 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 15 to the animal and in the same plane as the reflected light rays; that is, in the plane at an angle of 30° to 45° posterior to a plane perpen- dicular to the longitudinal axis of the animal. GROUPS As a matter of convenience for other workers in the study and identification of specimens, the writer has divided the American long- tailed shrews into assemblages of one or more species, which he calls hyd prd A B FIG. 2. — Teeth of Sore-r bendtrii bendirii. showing principal cusps. En- larged about 10 diameters. A, left upper teeth; B, left lower teeth jne= metacone. ms= mesostyle. mts = metastylc. pa= paracone. ps= parastyle. hy= liypocone. pr= protocone. cnd=entoconid. nied= rnetaconid. pad = paraconid. hyd= hypoconid. prd= protoconid. groups. The author is fully cognizant of the fact that the term " group " as thus used does not, and should not, have any status in the nomenclature of zoological classification and is employed solely for convenience. Nevertheless he has endeavored to bring within each group closely related forms, and, therefore, each group repre- sents more or less a taxonomic unit. Furthermore, an effort has been made to arrange the groups and species in phylogenetic sequence from the more simple morphologically to the more complex, and to arrange subspecies in accordance with intergrading forms. Strict 16 NORTH AMERICAN FAUNA [No. 51 adherence to such a method, however, has not been possible, since linear arrangement can not express what may actually be radial, parallel, or possibly, in the case of subspecies, even partly concen- tric. Although " groups," as previously stated, have been made as a matter of convenience, genera, subgenera, species, and subspecies have been recognized on the strength of structural characters and zoological relationships regardless of convenience in classification. MATERIAL EXAMINED The present revision recognizes 89 forms of 39 species of American long-tailed shrews and is based upon a study of 10,431 specimens, mostly skins accompanied by skulls. Of this number, the genus Sorex comprises 10,293 (subgenus Sorex, 9,369; subgenus Neosorex, 721; subgenus Atophyrax, 203); and Microsorex, 138. Type speci- mens or essentially topotypes of all described forms except Sorex hydrodromus Dobson have been examined. In some groups and species the material has been fairly adequate for a thorough inves- tigation. In others the number of specimens available has been entirely too small for satisfactory conclusions. And always more juvenile specimens Avere needed. The study has been based primarily upon specimens in the collection of the United States National Museum, including therein the Merriam collection and the large and important collection of the Bureau of Biological Survey. Without the cooperation of other institutions and individuals, however, this revision in its present completeness could not have been accom- plished.2 THE FAMILY SORICIDAE The family Soricidae, exclusive of fossil forms, is composed of some 24 currently recognized genera. The family ranges throughout North America, extreme northern South America, and the the trop- ical and temperate regions of Europe, Asia, and Africa. It is a compact, rather homogeneous group, the members of which are small to medium-size mouselike animals, with minute eyes, sharp-pointed snouts, and small ears, the ear conch always being present, though inconspicuous in certain genera. a The author expresses his gratitude and appreciation to each of the following for the loan of specimens or for various other courtesies : Joseph Grinnell, of the Museum of Vertebrate Zoology, University of California ; H. E. Anthony and G. G. Goodwin, of the American Museum of Natural History ; Wilfred H. Osgood, of the Field Museum of Nat- ural History ; Samuel Henshaw and Glover M. Allen, of the Museum of Comparative Zoology of Harvard College ; R. M. Anderson, of the National Museum of Canada ; Witmer Stone, of the Academy of Natural Sciences of Philadelphia ; Manton Copeland, of Bowdoin College ; A. G. Ruthven and Lee R. Dice, of the Museum of Zoology, University of Michi- gan ; George Wagner, of the University of Wisconsin ; W. T. Shaw, formerly of the State College of Washington ; M. H. Spaulding, of the Montana State College ; C. D. Bunker, of the Kansas University Museum of Natural History ; S. A. Barrett, of the Public Museum of the City of Milwaukee ; O. A. Peterson, of the Carnegie Museum ; J. D. Figgins, of the Colorado Museum of Natural History ; L. L. Snyder, of the Royal Ontario Museum of Zoology ; Clinton G. Abbott and Lawrence M. Huey, of the Natural History Museum, San Diego, Calif. ; Francis Kermode, of the Provincial Museum, British Columbia ; Philip Cox, of the Miramichi Natural History Society and the Provincial Museum, New Brunswick ; Frank Smith, of the Illinois State Laboratory of Natural History ; W. L. Burnett, of Colo- rado State College ; Donald R. Dickey, Pasadena, Calif. ; Stuart Criddle, Treesbank, Mani- toba ; C. F. Batchelder, Cambridge, Mass. ; Stanley G. Jewett, Portland, Oreg. ; A. Brazier Howell, Washington, D. C. ; D. E. Brown, Seattle, Wash. ; A. S. Pope, Chicago ; J. Dewey Soper, Ottawa, Canada ; D. E. Kent, Rutland, Vt. ; George I. Kirk, Rutland, Vt. ; E. R. Warren, Colorado Springs, Colo. ; Alex Walker, Tillamook, Oreg. ; Harley B. Sherman, Gainesville, Fla. ; Bernard Bailey, Elk River, Minn. ; and G. G. Cantwell, Palms, Calif. ; and to Oldfield Thomas, of the British Museum (Natural History), who has supplied many notes on specimens in that museum ; and to Gerrit S. Miller, jr., who has allowed absolute freedom in the division of mammals of the United States National Museum. 1928] BE VIEW OF AMEBICAN LONG-TAILED SHBEWS 17 The clavicle is long and slender; humerus relatively long and slender (length more than twice width) ; pelvis relatively broad (width more than one-third length) ; no os falciforme on the fore foot; terminal phalanges of fore foot simple, not bifurcate. The skull is somewhat conoidal, relatively long and narrow, the individual bones anastomosed into one compact whole with but little indication of the sutures; the zygomatic arch is absent, but represented by a rudimentary zygomatic process of the maxilla; audital bullae absent, the tympanic bone annular and not connected with the skull by osseous tissue; exterior pterygoid region angular and not inflated, no exterior pterygoid plate; mandible with double articulation. First upper incisor large, elongated, projecting anteriorly, two- lobed, the anterior lobe the larger; first lower incisor greatly elon- gated, extending anteriorly in line of mandible, the upper edge with two or more slightly developed lobes ; remaining incisors and canines, both upper and lower, and first and second upper premolars if pres- ent, simple unicuspidate ; crowns of upper molars low, W-shaped in inferior outline; crowns of lower molars low, M-shaped in superior outline. SUBFAMILIES The family Soricidae is usually divided into three subfamilies, namely, Soricinae, Crocidurinae, and Scutisoricinae. The last two are not represented in the American fauna. Soricinae is represented in America by five genera, namely, Sorex, Microsorex, Blarina, Cryptotis, and Notiosorex. The present revision includes only the two genera Sorex and Microsorex, which, however, represent a greater part of the American species. PELAGES AND MOLTS The hair of long-tailed shrews is fine, soft, and silky, but not of such velvet-like texture as is found in moles, though it approaches such a degree in the subgenera Neosorex and Atophyrax and some- times in immature individuals of any species. The pelage of shrews seldom shows the high gloss common to that of moles, nor, except in a few localized races, does it display the metallic iridescence char- acteristic of some species of moles in certain pelages. It is usual for mammals to have two types of hairs; one type is short, fine, and numerous, and forms the underfill- ; the other is longer, coarser, stiffer, and comparatively sparse, and forms a pro- tective covering, the overhair. In the long-tailed shrews there is no sharp distinction between underfur and overhair, either in texture or length, although in some individuals a few scattered hairs seem slightly heavier than the majority. Shrews are primitive mammals in many respects, and this lack of hair specialization may be another indication of primitive characters. TIME OF MOLTING So far as known, every species of long-tailed shrew has two molts annually, one in spring and one in fall. Although there is some variation in the time of molting among the different species, particu- 18 NORTH AMERICAN FAUNA [No. 51 larly in the spring, nevertheless individuals of nearly any species may be found in process of molt during May and early in June, and with- out exception of any species the autumnal molt may be looked for late in September or during October, apparently with slight regard to altitude or latitude. The spring molt of Sorex veraepacis seems to be earlier than in other forms, and S. cinereus may also molt as early as early April, while S. bendirii is not apt to begin molting before June. The time of molting is considered in more detail in the discussion under each species in the text following. MANNER OF MOLTING In general, during the spring molt the first appearance of the new fur is on the crown and nape, from where it gradually replaces the old on the head. The molt line then passes caudad over the shoul- ders and back, and ventrad over the sides, in the earlier stages moving more rapidly dorsally than ventrally, but in the later stages appar- ently more rapidly on the ventral parts, since often the entire under- pays are in fresh pelage before the fur on the posterior part of the back has molted. In fact, during spring the rump is nearly always the last part of a shrew to retain the old hair. Often, in the early stages of molting the old fur over the entire body is underlaid with the new, short hairs. Occasionally the mid-dorsal region will molt before the nape and the region over the shoulders do, but such cases are exceptions and seem to occur more frequently in the water and marsh shrews (subgenera Neosorex and Atophyrax) than in true Sorex. The characteristic autumnal molt almost reverses the sequence of that of spring. Usually the first new fur in the fall appears on the rump and posterior half of back; the molt then works cephalad and ventrad, gradually covering the entire animal, the head usually being the last part to change pelage. Sometimes, however, molt may start earlier on the head, leaving the shoulders and anterior portion of the back the last to molt. VARIATIONS GEOGR>PHIC VARIATION Geographic variation in long-tailed shrews manifests itself chiefly in variations of paleness or darkness, in size both external and cranial, in tail length, and in general shape of the skull, partic- ularly in degree of deflation of brain case, in breadth of rostrum and brain case, and in size of teeth and, correlated with it, length of molar tooth row. Geographic variations when constant in character and of commensurate degree may be recognized nomenclatorially as subspecific characters, particularly when such characters occur over a definite geographic area. As a rule there is comparatively little geographic variation in long- tailed shrews, and individual subspecies usually have an extensive geographic range. This is especially noticeable in such forms as Sorex c. cinereus, which ranges nearly across the North American Continent from east to west and has a north and south range from extreme northern Canada to the northern United States; and in S. o. obscurus, which is found with scarcely any variation from north- 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 19 central Alaska to northern New Mexico. As in all groups of mam- mals, there are certain forms' with restricted ranges apparently de- pendent upon peculiar environmental factors or upon more or less complete geographic isolation, but such forms, although superfi- cially similar, are usually specifically distinct from their nearest allies. There are several reasons for 'this lack of pronounced geographic variations in long-tailed shrews. The members of the family Sori- cidae are all comparatively simple colored grays and browns without distinct patterns. As the skull structure and the dentition are also very simple, there is little opportunity for variation in chromatic, cranial, or dental characters. In other words, the mere simplicity of the mammal tends to limit the possibilities for variations. Probably a more important factor in limiting these variations, however, is the plndogenetic age of the group. Shrews are geologically among the oldest of true placental mammals, and as such their characters are deep-seated and fixed. As an example the western water shrew {Sorex palustris navigator) , which inhabits the boreal elevations of many of the mountains of the western United States, might be cited. These shrews from the different ranges are very constant in char- acters and show comparatively little variation, yet their habitats on the. different mountains are often separated by broad expanses of desert or arid plain, which to this species has been an absolute barrier for possibly millions of years. INDIVIDUAL VARIATION The general shape and proportions of skulls of any form of long- tailed shrew, if of corresponding maturity and from the same locality, are seemingly very constant, yet when placed upon percentage basis the variation may amount to as much as 5 per cent from an average. Variation in actual size of skull, based upon any of several measure- ments, such as condylobasal length, greatest length, or breadth of cranium, may be even greater and in rare instances in large series has amounted to 7.5 per cent from the average. There is also a cor- respondingly great variation in the external measurements of total length, tail length, and hind foot, as computed from measurements taken by collectors in the field. There is a tendency for certain skulls of shrews to be " runty," or to have an abnormally shortened rostrum, or abnormal dentition. This does not occur in any great number of specimens, but neverthe- less appears to be more frequent than in most other families of Mam- malia. Thus a skull of Sorex c. cinereus (No. 150083. U. S. Nat. Mus.) from Mount Washington. N. H., has an abnormally shortened rostrum and interorbital region. A specimen of 8. v. vagrans (No. 833087, U. S. Nat, Mus.) from Bear Prairie, Mount Rainier, Wash., has the third upper incisor (second " unicispid ") on the right side bicuspidate. In a specimen of S. />. pcu /feus (No. 9G48. Field Mus. Nat. Hist.) from Eureka, Calif., the left upper first premolar (fifth unicuspid) has two distinct cusps, one directly caudad to the other. A specimen of S. o. obscurus (No. 988, Nat. Mus. Canada) from the mouth of Salmon lliver, British Columbia, has a supernumerary uni- cuspidate tooth interposed between the third and fourth unicuspids; 20 NORTH AMERICAN FAUNA [No. 51 the accessory tooth is smaller than either of the normally third or fourth unicuspids, and considerably smaller than the fourth. Two skulls of JS. o. longicamda (Nos. 74702 and 100570, U. S. Nat. Mus.) from Wrangell, Alaska, have each only four unicuspids on each side in the upper tooth row; the first premolar (fifth unicuspid) is evi- dently the one lacking. One of the most peculiar dental abnormalities occurs in a skull of Microsorex h. hoyi (No. 373, collection of Stuart Criddle, Treesbank, Manitoba) from Aweme, Manitoba. In each of the upper tooth rows one of the unicuspids is lacking, apparently the first one (second incisor), though it may possibly be the second (third incisor). The tooth row is compact, the space that would normally have been occu- pied by the missing tooth being taken up by a slight increase in the postero-anterior diameter of each of the other unicuspids. The in- crease in the size of the unicuspids is particularly noticeable in the case of the third. In normal individuals of Microsorex this tooth is so thin in postero-anterior diameter as to be a mere plate, but in the aberrant specimen this diameter is nearly half the lateral dimension. The two tooth rows are symmetrical with each other and present a striking anomaly. Except for some fading or " rusting," due to wear, the color of shrews of a given species from the same locality and in the same pelage is fairly constant. The fresh pelage is usually a trifle darker and more grayish than the old, and this sometimes gives the appear- ance of actual color variation. Shrews seldom exhibit abnormal color phases, such as melanism and albinism. The author has never seen a melanistic long-tailed shrew. A specimen of Sorex o. obscurus (No. 932, Prov. Mus. British Columbia) collected February 13, 1917, at Okanagan, British Columbia, seems to be a partial albino. This specimen is white on the chin and upper throat, the white extending ventrad and caudad on the left almost to the left fore leg. The color of the eyes of this specimen was not indicated by the collector. Another specimen of JS. o. obscuriis (No. 22060, Mus. Vert. Zool., Univ. Calif.), a male collected July 21, 1915, at an altitude of 10,800 feet at the head of Lyell Canyon, Yosemite National Park, Calif., has a general tone of color over the entire animal of pale ochraceous Duff, the underparts being paler and more whitish ; the base of hairs is pale smoke gray. A skin without skull (No. 241190, U. S. Nat. Mus.) of what is appar- ently JS. tundrensis, received at the National Museum on February 18, 1926, from H. O. Brown, of Shungnak, Alaska, is entirely white, although the color of the eyes is unknown. SEXUAL VARIATION So far as known, there is no sexual variation of color, size, o: proportions in any of the American long-tailed shrews. The adul males of all species have a relatively long and narrow gland on each flank, which develops conspicuously during the breeding season. The relative size of this gland varies with the different species, and in the genus Microsorex it is particularly large in proportion to the size of the animal, being about 9 millimeters long or nearly equal to the length of the hind foot ; it is small in Sorex cinereus, being only about 2 or 3 millimeters long; in JS. obscurus and JS. arcticus it is 192S] REVIEW OF AMERICAN" LONG-TAILED SHREWS 21 about half the length of the hind foot, or about 6.5 millimeters in the former and 7 millimeters in the latter (Preble, 1908, p. 243-249). The use of this gland as a taxonomic character, however, is not satis- factory, since not only is its use as such limited to less than half the specimens available, but also the gland is exceedingly difficult to measure accurately because of its position in the hair and the varying degrees of stretching it receives in different skins as made by different collectors. AGE VARIATION Externally, shrews display little variation with age. As a rule, younger animals appear slenderer than adults and have their tails a trifle more hairy and sometimes slenderer. Cranially, long-tailed shrews display great variation from the juvenile to the senile stage. The brain case flattens and appears to broaden with advancing age; the sutures of the cranium close; the sagittal and lambdoidal ridges develop ; the first incisors gradually grow anteriorly, then inferiorly, producing an entirely different aspect in old age from that of young; and the unicuspids seem to become somewhat swollen and broadened with age. These variations are described in more detail under the heading " Maturity of Skulls," page 13. SEASONAL VARIATION The only pronounced seasonal variation in the long-tailed shrews is in color and length of pelage. In nearly all species the winter pelage is longer and the color at that season is decidedly more grayish than in summer. The color difference between summer and winter fur is very marked in some species, as Sorex cinereus, S. fumeus, S. vagrans, and S. ornatus, in all of which the winter fur not only is more grayish but tends to be actually paler than is summer. In S. arcticus the winter pelage is darker than in summer, producing a more noticeable saddleback effect. The marsh and water shrews (S. bendirii and S. palustris) and S. trowbridgii, species already gray in summer pelage, have paler color in the winter coat. HISTORY The earliest reference in literature to a long-tailed shrew inhabit- ing America is that of Forster, who recorded a specimen sent in by Mr. Graham from the settlement on Severn River, Hudson Bay, under the name /Sorex araneus Linn. (Forster, 1772, p. 370, 380). Forster's specimen was of the species known to-day as S. arcticus Kerr, a shrew superficially like the European S. araneus, but which did not receive a scientific name until 20 years after its discovery, al- though Forster actually noticed differences between the two forms. Forster also had two other specimens of shrews from the same region, which he did not identify (op. cit., p. 381) and which belonged to the species now known as /S. cinereus Kerr. A few years later Pen- nant (1784, p. 139) redescribed the three Forster specimens, virtuallv copying Forster's descriptions, and placed them under "Foetid? shrew, a common name for S. araneus Linn. Forster and Pennant, however, used no Latin binomials. It remained for Kerr (1792, p. 206), basing his descriptions upon Pennant, to give valid names to 22 NORTH AMERICAN" FAUNA [No. 51 both of Forster's species ; the first one he calls " Labradore Shrew — • Sorex arcticus "/ the second species he named " Gray Labradore Shrew — Sorex arcticus cinereus" (vide Jackson, 1925a, p. 55). In November, 1826, Isidor Geoffroy St. Hilaire read an account of this same shrew, which had been named Sorex arcticus cinereus by Kerr, before the Societe d'Histoire JSTaturelle at Paris, and, about two months later, published a description of the animal under the name Sorex personatus (Geoffroy, 1827a, p. 319), a name used for many years for the common long-tailed shrew of the eastern United States and Canada, always erroneously dating, however, from a re- description by Geoffroy published late in the same year (Geoffroy, 1827b, p. 122). In 1828, Richardson was responsible for another name for Sorex arcticus cinereus Kerr, when he described Sorex forsteri (Richardson, 1828, p. 516). In this same paper, moreover, Richardson described the first American water-shrew known, under the name Sorex palustris (Richardson, 1828, p. 517). Thus, at this date, April, 1828, only three species of shrews had been described from the American continent. This is not so suprising, when one recalls the difficulties of collecting small mammals, particularly shrews, in those days when there were no small-mammal traps and when the capture of any small mammal was more or less chance or the result of the hard labor of digging for nests, setting snares, or making deadfalls. Probably the first contribution to the knowledge of the American shrews, which in any way could be dignified hj the title of a revision or review of the group, was that of Bachman in 1837. Bachman listed and described 13 species of shrews, 7 of which were long-tailed ones. Of the 7, 4 were described as new, only 1 of which, his Sorex longirostris (Bachman, 1837, p. 370), stands to-day. His name Sorex richardsonii (op. cit., p. 383), now a synonym of S. arcticus Kerr, was for many years used for the common saddle-backed shrew of America. Bachman also named Sorex cooperi (op. cit., p. 388) and Sorex fimbripes (op. cit., p. 391), both now sjmonyms of S. cinereus Kerr. He also listed Sorex palustris Richardson (op. cit., p. 396), which he had not seen, and Sorex forsteri Richardson (op. cit., p. 386) and Sorex personatus I. Geoffroy (op. cit., p. 398), the last also not seen, both synonyms of S. cinereus Kerr. One year after Bachman's paper appeared, Gray (1838) classified the family into two major divisions, namely, land shrews and water shrews. Under his land shrews were three genera, Corsira, Myosorex, and Sorex. Only Corsira was represented by American species, where he placed Sorex forsteri, S. longirostris, S. cooperi, and S. richardsonii Bachman. Gray also first used the name Blarina, as a subgenus of Corsira, where he placed all the American short-tailed species of shrews then known, and also S. personatus [S. c. cinereus~\ (op. cit., p. 124). Under his water shrews were two genera, Amphiso- rex and Crossopus ; of American species, S. palustris Richardson was included under Amphisorex (op. cit., p. 125) ; and S. fimbripes Bach- man, under Crossopus (op. cit., p. 126). Gray was confused in the actual relationships of many of the species, and his paper added little new, except his attempted arrangement of the species into genera and subgenera. 1928] EEVIEW OF AMEBICAN LONG-TAILED SHEEWS 23 In 1842, the genus Otisorex was named, with Otisorex platyrhinus the type species (De Kay, 1842, p. 22), a name, however, which is a synonym of Sorex cinereus Kerr. De Kay also included Bachman's species Sorex longirostris in the genus Otisorex (op. cit., p. 23.) This same year Duvernoy described Amphisorex lesueurii from Indiana (Duvernoy, 1842a, p. 33). a synonym of S. c. cinereus Kerr, and in another contribution dwelt in considerable detail upon the structure, development, and function of shrews' teeth (Duvernoy, 1842b). This latter paper was supplemented the following year (Duvernoy, 1843), and an essentially modified and revised edition of the whole work with the addition of illustrations was published a few vears later (Duvernoy, 1846). Sundevall, in a synopsis with brief descriptions of the shrews, divides the genus Sorex into three subgenera, the second of which, Sorex proper, he classifies into divisions 1 and 2 (Sundevall, 1843). In the first division he includes the American short -tailed shrews; under the second (" Oorsira Gray, Amphisorex Duvern.") he lists (" omnes mihi ignotae ") five species, namely, S. richardsoni Bach- man, S. forsteri Rich., S. lesueurii Duvernoy, S. personatus Is. Geoffry, and S. longirostris Bachm. (Sundevall, 1843, p. 182-183). The third subgenus recognized by Sundevall is Crossopus, where he allocates S. palustris Rich., but remarks " Non vidi." (Op. cit., p. 187.) Under the heading " Sorices incerti," Sundevall lists among several other species S. fimbripes Bachmah. (Op. cit., p. 188.) Although Sundevall had apparently never seen a specimen of an American shrew, his grouping of the species was probably the best that had been presented up to that time. In 1848 Pomel classified the insectivores into families, tribes, genera, and sections, each with a name and description. He used Hydrogale as a section name under the genus Sorex, but raised it to generic rank in the remark " si ce caractere se confirmait, ce type pour- rait otre erige en un genre distinct: II. fimbripes, l'espece est le sorex fimbripes Bachm." (Pomel, 1848, p. 248.) He also described the genus Galemys (not of Kaup, 1829) and placed therein the American species Sorex palustris under the section Crossopus (op. cit., p. 249.) Pomel placed S. longirostris Baehman in the genus Musaraneus Brisson and in the section Crocidura, thus: [Mu-^ar. (Croc.)] Bachmani (longirostris junior Bachm.) (Op. cit., p. 249.) Pomel's sections were in reality subgenera. A few years later Baird's epoch-marking work on the mammals of North America appeared in which he recognized 13 species of long- tailed shrews. (Baird, 1857, p. 7-56.) Baird described the genus Neosorex with Neosorex navigator the type species. (Op. cit., p. 11.) He also described as new 7 other species, namely: Sorex trow^- bridgii, S. vagrans, S. suckleyi, S. pachyurus, S. haydeni, S. hoyi, and S. thompsoni. In spite of the inadequacy of the material with which Baird worked, of the 8 new specific names that he proposed all except suckleyi and pachyurus are applicable to forms recognized to- day. Baird, however, recognized S. forsteri, S. platyrhinus, S. cooperi, and S. personatus, all of which are synonymous. He listed S. palustris Rich, and S. fimbripes Baehman, species of which he had not examined specimens. (Op. cit,, p. 55.) 24 NORTH AMERICAN" FAUNA [No. 51 In 1867 and 1868 Mivart published a somewhat detailed account of the osteology of the Insectivora (Mivart, 1867, p. 68) and divided the order into seven families, the family Sorices being represented by a single genus Sorex (op. cit., p. 141). Mivart made many com- parisons between Sorex and other genera, but inasmuch as he desig- nated no species in these comparisons the value of his work is nullified. Almost simultaneously with the appearance of the last part of Mivart's osteological work appeared the first part of the account of the shrews of the world by Fitzinger (1868). He recognized 10 species of American long-tailed shrews, 6 of which, however, are synonymous among his other 4 species as known to-day. He de- scribed as new Sorex wagneri (Fitzinger, 1868, p. 512), a synonym of S. longirostris Bachman. Fitzinger evidently did not consult the important contribution of Baird (1857), since he listed none of the species described as new in Baird's work. The results of the important investigations of E. Brandt on the dentition of shrews was published in three sections, which appeared, respectively, in 1869, 1871, and 1874. This study was based upon specimens of nine species of shrews belonging to 5 genera, the denti- tion of which are described in detail. Although Brandt included no American species, his work is valuable in its general application to certain American species and for comparative purposes. Previous to the German issue of this publication (Brandt, 1869-1874) there, had been an edition in Russian (Brandt, 1865). In his " Synopsis of Insectivorous Mammals," Gill (1875) classi- fied the order into families, subfamilies, and genera, giving detailed descriptions of families and subfamilies and a review of the more important works to that date. Gill recognized two genera of Ameri- can long-tailed shrews, namely, Sorex Linnaeus and Hydrogale Pomel, using the latter name to replace that of Neosorex Baird (Gill, 1875, p. 111). Two years later appeared the important studies on American insectivorous mammals by Coues (1877), in which he recognized two genera of American long-tailed shrews, Neosorex and. Sorex, and as a subgenus of Sorex named Microsorex, with the type species Sorex hoyi Baird. Coues did not discuss or list the various, species and subspecies, but described two new species of long-tailed, shrews (op. cit., p. 650), one, Sorex pacificus from Baird MS., a valid species; the other, Sorex sphagnicol-a, now a synonym of S.. arcticus. Later in this same year, Alston (1877) described the first- known long-tailed shrew from Central America under the name Sorex verae-pacis, although the animal had been known to Gray- (1843, p. 79) many years previously. In 1884, Merriam described as a new genus and species Atophyrax bendirii (Merriam, 1884b, p. 217), a large marsh shrew from Klamath- County, Oreg. Shortly afterwards there appeared an important anatomical paper by Parker (1885), in which was described and beautifully illustrated the development of the skull of Sorex vulgaris- es, araneus), not an American species, but one directly comparable,, as far as ontogeny is concerned, particularly with S. arcticus Kerr. The problematical Sorex hydrodromus from Unalaska Island,. Alaska, was described by Dobson in 1889 in a paper in which he also- remarks upon the uselessness of retaining Neosorex as a distinct; 1928] EEVIEW OF AMERICAN LONG-TAILED SHREWS 25 genus. (Dobson, 1889, p. 374.) At the time this paper appeared Dobson was working upon the part on the Soricidae of his mono- graph of the Insectivora, the first number of that part appearing in May of the following year (Dobson, 1890). Unfortunately Dobson's ill health and death prevented him from completing the momentous task, so that all that was published on the Soricidae was the fascicle of six plates, and these bear evidence in misnamed figures of not being proof-read by their careful and able author. In this work, the interesting American species, Sorex merriami, is named and figured. (Dobson, 1890, pi. 23, fig. 6.) During the next three years in three papers Merriam (1890, 1891, and 1892b) described six new forms of Sorex. It was not until 1895, however, when the revisions by Miller and Merriam were published, that a clear presentation of the relationships of the American species as understood at that time was had. Miller's contribution was a review of the members of the genus Sorex (including subgenera Sorex, JVficrosorex, and Neosorex) occurring east of the Great Plains of the United States. He had examined in the British Museum the original specimens of /Sorex palustris, S. forsteri, and S. parvus described by Richardson, which enabled him to describe more ac- curately these specimens and clarify questions of their relationships. He recognized seven species from the eastern United States, one of which, S. fumeus, he described as new (Miller, 1895, p. 50). Mer- riam's synopsis (1895) comprises the most complete account pub- lished of the long-tailed shrews of the entire region of North America and Central America. Merriam recognized 41 species and subspecies, which he included in the single genus Sorex, divided into four sub- genera ; 33 species and subspecies were placed in the subgenus Sorex, 1 in the subgenus Microsorex, 4 in the subgenus Neosorex, and 3 in the subgenus Atophyrax. In this revision, Merriam described 21 new species and subspecies, all of which are recognized in the present revision. And of the 41 forms recognized by Merriam, all except S. sphagnicola (=S. arcticus) and S. vagrans dobsoni (=S. v. monticola) are recognized in the present work, although his S. per^ sonatus here appears under the name S. ci/iereus, and his S. richard- soni as S. arcticus. Merriam's contribution was a big stepping-stone in the climb toward a knowledge of this difficult, group. In 1896 Batchelder described an interesting and distinctive species of shrew from New York under the name Sorex maerurus, which being preoccupied he later renamed Sorex dispar (Batchelder, 1911). During the 30 years following the revisions by Miller and Merriam (1895), there appeared numerous other descriptions of new speciea or subspecies of American long-tailed shrews bv Merriam (1897, 1899, 1900, 1902), Bangs (1899), Elliot (1899, 1903b), Osgood (1901a, 1901b, 1909), Preble (1902, 1910), Nelson and Goldman (1909), Bailey (1913), Jackson (1917, 1918, 1919, 1921b, 1922, 1925a, 1925b, 1925c, 1926), and Anthony and Goodwin (1924); there were also published several papers treating upon the distribution or habits of A-merican forms. During this period, however, only four contribu- tions stand out above the others as needing special mention here. The first of these is an anatomical paper by Arnback-Christie-Linde (1907), which treats in some detail of the muscles of S. pygmaeus and S*. vulgaris, European forms, the latter not far removed from S., 26 NORTH AMERICAN FAUNA [No. 51 arrcticus. In this account, also, the side glands are mentioned as occurring only in males, and the investigator considers that the mem- bers of the family Soricidae hibernate because they are provided with " Winterschlaf drusen oder braunem Fettgewebe," a case where present knowledge of habits apparently does not substantiate a supposed structural adaptation. Hollister's paper (1911) was a brief review of the Sorex of the Eastern United States, in which was described the new species /Sorea fontinalis. Hollister recognized five species from the region. He considered Amphisorex lesueurii Duvernoy to be a synonym of S. longirostris Bachman, and S. fimbripes Bachman to be unidentifiable ; he identified S. acadicus Gilpin with S. personatus Geoffro}^ Grinnell (1913a) discussed the characters, relationships, and dis- tribution of six species and subspecies of Sorex from west-centra] California. Three new forms Avere described, all of which art recognized in the present monograph. Glover M. Allen (1915) described as new Neosorex palustrii acadicus (=S. p. glover alleni Jackson), carefully compared it Witt related forms, and listed all the other known subspecies with ar outline of their respective geographic ranges. LIST OF GENERIC NAMES USED FOR AMERICAN LONG-TAILEI SHREWS Amphisorex Duvernoy, Mem. de la Soc. Mus. d'Hist. Nat. Strasbourg, tome S sig. 5, p. 23, 1835. Type species Sorex hermanni Duvernoy, the animal o which is Sorex araneus tetragonurus Hermann and the skull, Neomy fodiens fodiens Schreber (vide Miller, 1912a, pp. 29, 42, 70). Used generi cally for Amphisorex lesueurii Duvernoy qui Sorex cinereus cinereus Ken Atophyrax Merriam, Trans. Linn. Soc. New York 2 : 217, August, 1884. Typ species Atophyrax oendirii Merriam. A subgenus of Sorex Linnaeus. Corsira Gray, Proc. Zool. Soc. London, part 5, 1837, p. 123, May, 1838. Typ species Sorex vulgaris Linnaeus. A synonym of Sorex Linnaeus to whic Gray referred Sorex forsteri Richardson qui Sorex cinereus cinereus Ken Crocidura Wagler, Isis von Oken 25 : 275, 1832. Earliest available name for th Old World genus of which Sorex leiwodon Hermann is the type specie: Used generically in synonymy for several species of American Sorex b Fitzinger (1S68), who refers to Reichenbach. Croscopus ? Fitzinger, Sitzungber. Kaiserl. Akad. Wissensch., math.-natiirwi; sensch. Classe, Wien, Band 57, Abt. 1, p. 632, 1S6S. Misprint for Crossopu Wagler. Used in synonymy under Crossopus fimbripes qui Sorex cineren cinereus Kerr. Crossopus Wagler, Isis von Oken 25: 275, 1832. Type species Sorex fodieri Beckstein =Sorex fodiens Schreber. A synonym of Neomys Kaup. Firs used for an American species as Crossopus palustris Reichenbach (1847, ] 161), qui Sorex palustris palustris Richardson. Galemys Pomel, Archives Sci. Physiques et Nat., Geneve 9 : 249, Novembe 1848. Included Brachysorex Duvernoy, Crossopus Wagler, and Pachyur de Selys-Longchamps. Not Galemys Kaup (1829), which is a genus ( Talpidae. Included Sorex palustris Richardson. Hydrogale Pomel, Archives Sci. Physiques et Nat., Genewe 9 : 248, Novembe 1848. Type species Sorex fimbripes Bachman. Not Hydrogale Kaup, 182 qui Neomys Kaup. Used to replace Neosorex Baird by Gill (1875, p. Ill Microsorex Coues, Bui. U. S. Geol. and Geogr. Surv. Territories 3 : no. 3, p. 64 May 15, 1877. Earliest available name for the genus of which Sorex Baird is the type species. Described as a subgenus by Coues from B manuscript. Raised to rank of genus by Elliot (1901a, p. 377). Musaraneus Brisson, Regnuni Animale, p. 126, 1762. A synonym of Sorex naeus. Pomel placed Sorex longirostris Bachman in this genus and in " section " Crocidura under the specific name oachmani, thus : " [M (Croc.)] Bachmani (longirostris junior Bachm.)." 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 27 Neosorex Baird, Report Pacific Railroad Survey, vol. 8, part 1. Mammals, p. 11, 1857. Type species Neosorex navigator Baird. A subgenus of Sorex Linnaeus. Otisorex DeKay, Zoology of New York, part 1, Mammalia, p. 22, and plate 5, fig. 1, 1842. Type species Otisorex platyrhinus DeKay, qui Sorex cinereus Kerr. A synonym of Sorex Linnaeus. DeKay also included Sorex longiros- trig Bachman in the genus. Sorax Hollister, Proc. U. S. National Museum 40 : 378, April 17, 1911. Misprint for Sorex. Sorex Linnaeus, Systema Naturae, edition 10, vol. 1, p. 53, 1758. Available name for the genus of which Sorex araneus Linnatus is the type species. KEY TO THE GENERA AND SUBGENERA OF AMERICAN LONG- TAILED SHREWS a1. Unicuspids 5, in superficial lateral view appearing to be only 3, the third and fifth being scarcely, if at all, visible; third uni- cuspid disklike, antero-posteriorly flattened; primary (an- terior) lobe of first upper incisor relatively long and narrow, the length more than twice the width and more than twice ' the length of secondary lobe Genus Microsorex (p. 200) a1. Unicuspids 5, in superficial lateral view appearing to be 5, the fifth sometimes minute and indistinct; third unicuspid not disklike, not antero-posteriorly flattened; primary (an- terior) lobe of first upper incisor relatively broad, the length less than twice the width and usually less than twice the length of secondary lobe Genus Sorex (p. 30) b1. Size smaller; hind foot less than 18; and if hind foot is over 16, color distinctly brown Subgenus Sorex (p. 31) b2. Size larger; hind foot 18 or more; color grayish, never dis- tinctly brown. c1. Rostrum shorter and little down-curved; anterior end of premaxilla scarcely narrower dorso-ventrally than middle portion; dorso-ventral diameter of rostrum measure! at third unicuspid equal about half the diam- eter between anterior border of infraorbital foramen and posterior border of first incisor; posterior end of interior cutting edge of anterior portion of internal basal shelf of first and second upper molars usually without cusplike lobe; hind foot distinctly fimbriate. Subgenus Neosorex (p. 175) c2.. Rostrum relatively longer and distinctly down-curved; an- terior end of premaxilla much narrower dorso-ventrally than middle portion; dorso-ventral diameter of rostrum measured at third unicuspid Less than half the diameter between anterior border of infraorbital foramen and posterior border of first incisor; posterior end of interior cutting edge of anterior portion of internal basal shelf of first and second upper molars usually with distinct cusplike lobe; hind foot slightly fimbriate. Subgenus Atophyrax(p. 192) LIST OF AMERICAN GENERA, SPECIES, AND SUBSPECIES OF LONG- TAILED SHREWS, WITH TYPE LOCALITIES SOREX CINEREUS GROUP Sorex cinereus cinereus Kerr Fort Severn, Ontario (p. 40) . cinereus miscix Bangs Black Bay, Labrador (p. 50). cinereus haydeni Baird Fort Buford, N. Dak. (p. 51). cinereus streatori Merriam Yakutat, Alaska (p. 53). cinereus hollisteri Jackson St. Michael, Alaska (p. 55). fontinalis Hollister Near Beltsville, Md. (p. 56). lyelli Merriam Mount Lyell, Calif, (p. 57). ■preblei Jackson Jordan Valley, Oreg. (p. 58) . 74235—28 3 28 NORTH AMERICAN" FAUNA [No. 51 SOREX FUMETTS GROUP Sorex fumeus fumeus Miller Peterboro, N. Y. (p. 63). fumeus umbrosus Jackson James River, Nova Scotia (p. 65). SOREX ARCTICUS GROUP Sorex ardicus arcticus Kerr Fort Severn, Ontario (p. 68). arcticus laricorum Jackson Elk River, Minn. (p. 71). tundrensis Merriam St. Michael, Alaska (p . 72) . hydrodromus Dobson Unalaska Island, Alaska (p. 74) . SOREX PRIBILOFENSIS GROUP Sorex pribilofensis Merriam St. Paul Island, Pribilof Group, Alaska (p. 76). SOREX MERRIAMI GROUP Sorex merriami Dobson Fort Custer, Mont. (p. 78). leucogenys Osgood 3 miles east of Beaver, Utah (p. 81). SOREX SCLATERI GROUP Sorex sclateri Merriam Tumbala, Chiapas, Mexico (p. 82). SOREX LONGIROSTRIS GROUP Sorex longirosiris longirostris Bachman Cat Island, mouth of Santee River, S. C. (p. 85). longirostris fisheri Merriam Lake Drummond, Dismal Swamp, Va. (p. 87). SOREX DISPAR GROUP Sorex dispar Batchelder Beedes, N. Y. (p. 89). gaspensis Anthony and Goodwin Mount Albert, Quebec (p. 91). SOREX TROWBRIDGII GROUP Sorex trowbridgii trowbridgii Baird Astoria, Oreg. (p. 94). trowbridgii humboldtensis Jackson Mad River, Calif, (p. 96). trowbridgii montereyensis Merriam Monterey, Calif, (p. 97). trowbridgii mariposae Grinnell Yosemite Valley, Calif, (p. 98). SOREX VAGRANS-OBSCURUS GROUP Sorex vagrans vagrans Baird Shoalwater Bay, Wash. (p. 104). vagrans Vancouver ensis Merriam Goldstream, British Columbia (p. 106). vagrans nevadensis Merriam Reese River, Nev. (p. 107). vagrans halicoetes Grinnell Palo Alto, Calif, (p. 108). vagrans amoenus Merriam Near Mammoth. Calif, (p. 109) . vagrans moniicola Merriam San Francisco Mountain, Ariz. (p. 110). vagrans orizabae Merriam Mount Orizaba, Puebla, Mexico (p. 113). durangae Jackson El Salto, Durango, Mexico (p. 114). obscurus obscurus Merriam Lemhi Mountains, Idaho (p. 117). obscurus neornexicanus Bailey Cloudcroft, N. Mex. (p. 123). obscurus parvidens Jackson Bluff Lake, San Bernardino Mountains, Calif, (p. 124). 1923] REVIEW OF AMERICAN LONG-TAILED SHREWS 29 Sorex obscurus shumaginensis Merriam Popof Island, Alaska (p. 125). obscurus alascensis Merriam Yakutat, Alaska (p. 126). obscurus malitiosus Jackson Warren Island, Alaska (p. 128). obscurus elassodon Osgood Moresby Island, Queen Char- lotte Islands, British Colum- bia (p. 130). obscurus longicauda Merriam Wrangell, Alaska (p. 131). obscurus prevostensis Osgood Prevost Island, Queen Char- lotte Islands, British Colum- bia (p. 133). obscurus isolatus Jackson Nanaimo, Vancouver Island, British Columbia (p. 134). obscurus setosus Elliot Happy Lake, Olympic Moun- tains, Wash. (p. 135) obscurus permiliensis Jackson Mount Jefferson, Oreg. (p. 137) . obscurus bairdi Merriam Astoria, Oreg. (p. 139). yaquinae Jackson Yaquina Bay, Oreg. (p. 140). pacificus pacificus Coues Fort Umpqua, Oreg. (p. 142). paciftcus sonomae Jackson Gualala, Calif, (p. 143). SOREX STIZODON GROUP Sorex stizodon Merriam San Cristobal, Chiapas, Mexico (P. 147). SOREX VERAEPACIS GROUP Sorex veraepacis veraepacis Alston Coban, Guatemala (p. 149). veraepacis chiapensis Jackson San Cristobal, Chiapas, Mex- ico (p. 150). veraepacis mutabilis Merriam Reyes, Oaxaca, Mexico (p. 151). macrodon Merriam Orizaba, Vera Cruz, Mexico (p. 152). SOREX SAUSSUREI GROUP Sorex saussurei saussurei Merriam North slope of Sierra Nevada de Colima, Jalisco, Mexico (p. 155). saussurei veraecrucis Jackson Xico, Vera Cruz, Mexico (p. 156). saussurei oaxacae Jackson Mountains near Ozolotepec, Oaxaca, Mexico (p. 157). saussurei cristobalensis Jackson San Cristobal, Chiapas, Mexico (p. 157). saussurei godmani Merriam Volcan Santa Maria, Guatemala (p. 158). saussurei salvini Merriam Calel, Guatemala (p. 159) . emarginatus Jackson Sierra Madre near Bolanos, Ja- lisco, Mexico (p. 159). ventralis Merriam Cerro San Felipe, Oaxaca, Mex- ico (p. 160). or eopolus Merriam North slope Sierra Nevada de Colima, Jalisco, Mexico (p. 162). SOREX ORNATU8 GROUP Sorex ornatus ornatus Merriam Mount Pinos, Calif, (p. 166). ornatus calif or nicus Merriam Walnut Creek, Calif, (p. 168). ornatus lagunae Nelson and Goldman La Laguna, Sierra Laguna, Lower California, Mexico (p. 169). trigonirostr is Jackson Ashland, Oreg. (p. 170). sinuosus Grinnell Grizzlv Island, near Suisun, Calif, (p. 171). juncensis Nelson and Goldman Socorro, Lower California, Mex- ico (p. 172). tenellus Merriam Lone Pine Creek, Alabama Hills, near Lone Pine, Calif. (p. 172). myops Merriam Pipers Creek, White Mountains, Calif, (p. 173). nanus Merriam __. Estes Park, Colo. (p. 174). 30 NOETH AMERICAN FAUNA [No. 51 SOREX PALUSTRIS GROUP Sorex palustris palustris Richardson Between Hudson Bay and Rocky Mountains, Canada (p. 178). palustris hydrobadistes Jackson Withee, Wis. (p. 180). palustris atbibarbis (Cope) Profile Lake, Franconia Moun- tains, N. H. (p. 181). palustris gloveralleni Jackson Digby, Nova Scotia (p. 183) palustris navigator (Baird) Near head of Yakima River. Cascade Mountains, Wash, (p. 184). alaskanus Merriam Point Gustavus, Glacier Bay Alaska (p. 189). SOREX BENDIRII GROUP Sorex bendirii bendirii (Merriam) Eighteen miles southeast of Fori Klamath, Oreg. (p. 194). bendirii palmeri Merriam Astoria, Oreg. (p. 197). bendirii albiventer Merriam Lake Cushman, Olympic Moun- tains, Wash. (p. 198). MICROSOREX HOYI GROUP Microsorex hoyi hoyi (Baird) Racine, Wis. (p. 202) . hoyi thompsoni (Baird) Burlington, Vt. (p. 204). hoyi winnemana Preble Bank of Potomac River, 4 mile; below Great Falls, Fairfa: County, Va. (p. 206) . hoyi intervectus Jackson Lakewood, Wis. (p. 206). hoyi alnorum (Preble) Robinson Portage, Manitobi (p. 208). hoyi eximius (Osgood) Tyonek, Alaska (p. 208) . hoyi washingtoni Jackson Loon Lake, Wash. (p. 209) . Genus SOREX Linnaeus Sorex Linnaeus, Systema Naturae, ed. 10, vol. 1, p. 53, 175S. JWusaraneus Brisson, Regnum Animale, p. 126, 1762. Oxi/rliiii Kaup, Skizzirte Eiitwickelungs-Geschich. und natiirl. Systeir europaischen Thierwelt, p. 120, 1829. Amphisorex Duvernoy, Mem. de la Soc. Mus. d'Hist. Nat. Strasbourg 2, sig. 5 p. 23, 1S35. Corsira Gray, Proc. Zool. Soc. London, part 5, 1837, p. 123, May, 183S. Otisorex, DeKay, Zoology of New York, part 1, Mammalia, p. 22 and pi. 5, fig 1, 1842. Hydrogale Pomel, Archives Sci. Phvsiques et Nat. Geneve 9 : 248, November 1848. Neosorex Baird, Rept. Pacific Railroad Survey 8, part 1, Mammals, p. 11, 1857 Atophyrax Merriam, Trans. Linnaean Soc. New York 2 : 217, August, 1884. Eomalurus Schulze, Schriften des Naturwissenschaft. Yereins des Harzes Wernigerode, 5: 28, 1SD0. Type species. — Sorex armieus Linnaeus. Geographic range of American species. — From the Arctic Ocean south through Alaska, Yukon, Northwest Territories, Quebec, anc Labrador, to central South Carolina, northern Florida, Alabama, anc southern Illinois in the eastern United States, to central Nebraska •in the Great Plains region, to southern Utah and southern Nevada ii the Great Basin region, to southern Lower California on the Pacifi coast, and in the mountains of Utah, Colorado, New Mexico, anc Arizona, and south through the mountains of Mexico to westen Guatemala. Generic characters. — Size small, form murine ; pelage soft and velvetlike tail more or less completely covered with hairs, moderately long, in mos species about three-fourths length of head and body, but varying from one half length of head and body (Sorex tundrensis) to about equal length o 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 31 head and body ( interior edge of cingulum, in unworn teeth sometimes ending in very minute pigmented eUsplet on cingulum. Members of the dnereus group may he distinguished from any of the longi- rostris or vagrans-oiscurus groups by the relatively narrower rostrum of cincrcus, narrower (exlero-interiorly) molariform teeth, and by the relative size of the fourth unicuspid to the third, usually smaller than or rarely equal to the third in cincrcus group; larger than the third in Imujirostris and vagrans- obscurus groups. Smaller than any form of the arcticus, fumcus, or dispar groups, with smaller hind font, and distinctly smaller skull, weaker dentition, 38 NORTH AMERICAN FAUNA [No. 51 the molariform teeth being narrower (extero-interior diameter) ; rostrum decidedly narrower both actually and relatively than in members of arcticus and fumcus groups. Tail relatively longer than in pribilofensis group, and coloration not tricolor ; skull relatively much narrower, particularly inter- orbitally and rostrally ; dentition weaker, the internal ridge on unicuspids less heavily pigmented. Somewhat smaller than either species of the merriami group, darker ventrally and on feet ; skull relatively longer, higher, and nar- rower, not swollen interorbitally (as in merriami group), unicuspidate teeth less crowded and not relatively deeper (supero-inferiorly) than broad (antero- posteriorly) in lateral aspect. Remarks. — The cinereus group, although comprising only eight recognizable forms belonging to four species, has a wide geographic range entirely across the northern half of the North American Con- tinent. In many places in the Canadian and Boreal Zones shrews of this group are among the more common mammals. Although members of the cinereus group are superficially similar both externally and cranially to S. minutus of Europe, the two appear to be actually not closely related. The dentition of S. minutus, par- ticularly as shown in the characters in the first upper incisor, is very different from that of S. cinereus and its related forms. SOREX CINEREUS Kekb [Synonymy under subspecies] Geographic range. — That of the cinereus group, except that part of California inhabited by Sorex lyelli, that part of eastern Oregon inhabited by S. preNei, and that part of Maryland inhabited by /S. fontinalis. (Fig. 3.) Diagnostic characters. — Somewhat larger in all respects than S. lyelU, S. fonti- ntilis, or S. preblei, with correspondingly larger skull, relatively higher brain case (except in certain specimens of 8. c. haydeni), and usually with longer, narrower rostrum ; unicuspidate tooth row relatively and actually longer than in S. lyelli, 8. foutinalis. or 8. preblei, the unicuspids being less crowded and with greater antero-posterior diameter. Subspecies and geographic variation. — The species cinereus is divided into five subspecies: cinereus. miscix, haydeni, streatori, and hollisteri. The species as a whole is rather variable, as is also each of the subspecies. Each subspecies, however, in its extreme form is well defined and has average differences over considerable geographic areas. Beginning from the southern border of the range of the species, there is a general tendency for an increase in size toward the northeast and northwest, which culminates in the subspecies miscix and streatori, each recognizable also by color differences. Still farther northward along the Arctic coast the animal becomes somewhat smaller again, and on the Bering coast of Alaska this tendency reaches a climax in the pale, small, rather short-tailed form hollisteri. The form from the Great Plains region is small, in keeping with other specimens from the southern part of the range of the species, but has additional characters of pale color, short tail, and short, broad rostrum, and is recognized under the name haydeni. Time of molting. — The transition from winter to summer pelage may occur any time between the first of April and the last of June, depending somewhat J upon latitude and altitude. A male of £. e. cinereus in early process of molt I was collected at Hinckley, Minn., March 31, 1890. A female from Fort Totten, Ij N. Y., is at about the same stage, April 4, 1908. Two other females from Jobs « Knob, W. Va. (April 10 and 13, 1897), have the summer fur well advanced ' under the old over the entire backs, the one collected on the earlier date having ' I the underparts in fresh pelage. A male from St. Marys Lake, Mont., has the j summer hair appearing under the worn winter fur over the entire animal (June ' 7, 1895). A male of S. c. haydeni from Fairmount, N. Dak., is in the height of , the molting process (May 22, 1915) ; while another collected on June 7, 1915, I 192S] REVIEW OP AMERICAN LONG-TAILED SHREWS 39 at the near-by locality of Blaekmer is in complete summer pelage. A male from Ekalaka, Mont., collected on May 29, 1916, is about half molted. Three other males have about completed the molt in the Bear Lodge Mountains, Wyo., June 22, 1912, others of the same date and locality being in summer fur. A male 8. c. streatori from Skagway, Alaska, June 1, 1899, and one from Orca, Alaska, June 25, 1S99, are in process of change from winter to summer pelage. The winter pelage is usually acquired during October in the southern part of the range of the species, and in the more northern parts during early Septem- ber or even late August. Two females of 8. c. cinereus, one collected Sep- tember 4. 1893, at Montauk Point, Suffolk County, N. T., and the other taken November 20, 1893, at Wilmington, Mass., are both in early stages of the autumnal molt. A series from Prince Edward Island collected between October Fig. 3. — Geographic range of the species and subspecies of Sorex cinet'i us group 1. 8. cinereus cinereus. '1. 8. c. miscix. 3. 8. c. haydeni. 4. 8. c. streatori. 5. 8. c. holli.strri. 6. 8. fontinalis. 7. 8. lyelli. 8. S. preblei. 27 and November 10, 1897, is for the most part in full winter pelage; two males from this series taken November 3 and 10 are in early stages of transition. Another series from St. Elmo, Colo., is in winter pelage early in October, 1907, and numerous specimens from the mountains of Wyoming arc in process of molt during middle September. A female from the Highwood Mountains, Mont., has the molt well begun, August 22, 1910. Numerous specimens from west-central Alberta were molting the last few days of August and the first week of Sep- tember. Specimens from British Columbia (Bennett, Cariboo Lake, Glenora, and Sicamous) were undergoing change of pelage from the middle to the last of September; and the same is true of a large series from Great Bear Lake, Northwest Territories, many of which are in full winter pelage. Speci- mens from the Cook Inlet region of Alaska were changing from summer to winter pelage between August 24 and September 20, 1900, while a few from the 40 NORTH AMERICAN FAUNA [No. 51 mountains near Eagle, Alaska, show beginning of the molt during the last half of August. The majority of specimens of S. c. miscix are in winter pelage in October. A female from Black Bay, Labrador, is in summer pelage October 3 ; another female is in early process of molt September 24; while a male is at about the same stage October 20. Representatives of 8. c. hay den i from Fort Custer, Mont., are with one exception in full winter pelage the middle of Novem- ber, 1895 ; a female, taken November 13, is still in process of molting. Two females collected October 26, 1893, at Portland, .N. Dak., are in full winter pelage. One taken October 9, 1901, 10 miles south of Cody, Nebr., shows the beginning of the pelage change. Two males of '8. c. streatori from Wrangell, Alaska, had begun to molt September 12, 1895. The majority of specimens in the topotype series of 8. c. hollisteri, collected during the middle of September, 1899, are changing from summer to winter fur; several had already acquired the full winter coat, even as early as September 10. Other specimens of hollis- teri from Nushagak, Alaska, were molting during the middle of September, 1902; and specimens from Kakhtul River, Alaska, show molt as early as August 29, the same year. SOREX CINEREUS CINEREUS Kerr Cinereous Shrew (Pls. 2, a; 4, tt; 5, s; 7, a; 11, a; 12, a) Sorex arcticus cinereits Kerr, Animal Kingdom, p. 206, 1792. Sorex personatus I. Geoff roy- Saint Hilaire, Dictionnaire Cla^sique d'Hist. Nat. 11:319, January, 1827. Sorex forsteri Richardson, Zool. Journ. 3 : no. 12, January-April, 1828, p. 516, April, 1828. Type locality, " Hudson's Bay countries." Sorex cooperi Bachman, Journ. Acad. Nat. Sci. Philadelphia 7 : part 2, p. 3S8, 1837. Type locality, " North Western Territory" Sorex fimbripes Bachman, Journ. Acad. Nat. Sci. Philadelphia 7 : part 2, p. 391, 1837. Type locality, Drury Run, Pa. Corsira forsteri J. E. Gray, Proc. Zool. Soc. London, part 5, 1837, p. 124, May, 1838. Otisorex platyrhinus De Kay, Zool. New York, part 1, Mammalia, p. 22, 1842. Type locality. Tappan, Rockland County, N. Y. S[orex] platyrrhinchus Linsley, Amer. Journ. Sci. and Arts 43: no. 2, p. 346, October 6, 1842. Tyoe locality, Stratford, Conn. (Misspelling or emenda tion of platyrhinus De Kay, with redescription. ) Amphisorex lesueurii Duvernoy, Mag. de Zool., d'Anat. Comp. et de Palaeont., series 2, 4th year, Monog. du Genre Musaraigne, p. 33, November, 1842. Type locality, Wabash River Valley, Ind. Slorex] lesueurii Sundevall, Kongl. [Svenska] Yetenskapsacad. Handl., 1842, p. 1S2, 1843. [Sorex] lesueurii (sic) Reichenbach, Praktische Naturgesch. Menschen unci Saugth., p. 165, 1847. Hlydrogale] fimbripes Pomel. Arch. Sci. Phys. et Nat. 9: 248, 1S4S. S[orex] platyrrhinus (sic) Wagner, Suppl. Schreber Saugth. 5: 547, 1855. Sorex platyrhinus Baird, Report Pacific R. R. Survey 8 : part 1, Mammals, p. 25. 1857. Amphisorex lesuerii (sic) Baird, Report Pacific R. R. Survey 8: part 1, Mam mals, p. 27, 1857. (In questionable synonymy under Sorex cooperi Bach man.) Sorex lesueri (sic) Baird, Report Pacific R. R. Survey 8: part 1, Mammals, p 27, 1857. (In synonymy under &. cooperi Bachman.) Sorex fosteri (sic) Packard, Proc. Boston Soc. Nat. Hist. 10: 266, 1866. Sorex acadicus Gilpin, Proc. and Trans. Nova Scotian Inst. Nat. Sci. 1 : part 2 (erroneously marked vol. 2, part 2), p. 2, 1867. Type locality, Novs Scotia. Amphisorex forsteri Fitzinger, Sitzungsber. Kaiserl. Akad. Wissensch., math natiirwissensch. Classe, Wien, bd. 57, abt. 1, p. 509, 1868. (In synonymy.) Crocidura cooperi Fitzinger, Sitzungsber. Kaiserl. Akad. Wissensch., math natiirwissensch. Classe, Wien, bd. 57, abt. 1, p. 513, 1S68. (In synonymy.) Otisorex platyrrhinus' Fitzinger, Sitzungsber. Kaiserl. Akad. Wissensch., math natiirwissensch. Classe, Wien, bd. 57, abt. 1, p. 584, 1868. (In synonymy.) 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 41 Sorex platyrhynchus Fitzinger, Sitzungsber. Kaiserl. Akad. Wissensch., niath.- natiirwissenscli. Classe, Wien, bd. 57, abt. 1, p. 584, 1868. \,ln synonymy.) Crocidura platyrhyncha Fitzinger, Sitzungber. Kaiserl. Akad. Wissensch., math.- natiirwissensch. Classe, Wien, bd. 57, abt. 1, p. 585, 1868. (In synonymy.) Crossvpus fimbripes Fitzinger, Sitzungber. Kaiserl. Akad. Wissensch., rnath.- natiirwissensch. Classe, Wien, bd. 57, abt. 1, p. 631, 1868. Crocidura fimbripes Fitzinger, Sitzungber. Kaiserl. Akad. Wissensch., math.- natiirwissensch. Classe, WTien, bd. 57, abt. 1, p. 631, 1868. (In synonymy.) Croscopus ? fimbripes Fitzinger, Sitzungsber. Kaiserl. Akad. Wissensch., niath.- natiirwissensch. Classe, Wien, bd. 57, abt. 1, p. 632, 1868. (In synonymy.) Sorex platyrinus Gilpin, Proc. and Trans. Nova Scotian Inst. Nat. Sci. 2: part 2, p. 59, 1869. Sorex acadica ? Gilpin, Proc. and Trans. Nova Scotian Inst. Nat. Sci. 2: part 2, p. 59, 1869. Sorex fimbria t a Holder, Hist, of the Amer. Fauna, part 3, p. 30, 1877. Sorex idalwensis Merriam, North Amer. Fauna No. 5, p. 32, July 30, 1891. Type locality, Timber Creek, altitude 8,200 feet, Salmon River Mountains [now Lemhi Mountains], Idaho. Amphisorex lemeri (sic) Herrick, Mammals of Minnesota, Geol. and Nat. Hist. Surv. Minnesota, Bui. 7, p. 48, 1892. (In synonymy under Sorex cooperi Bachman.) Amphisorex leseurii (sic) Butler, Proc. Indiana Acad. Sci., 1891, p. 163, 1S92. Sorex platyrhinclxus (sic) Miller, North Amer. Fauna No. 10, p. 39, December 31, 1895. Amphisorex leseueri (sic) Miller, North Amer. Fauna No. 10, p. 53, December 31, 1895. (In synonymy under Sorex per.onatus Geoffroy.) Amphisorex lesueuri Merriam, North Amer. Fauna No. 10. p. 00, December 31, 1895. (In synonymy under Sorex per&onaVus Geoffroy.) [Sorex personatus] lesueuri Merriam. Norrh Amer. Fauna No. 10, p. 61, December 31, 1895. [Sorex] plati/rhiirhus (sic) Elliot, Field Columb. Mus. Publ. 45 (zool. series 2) : 366, 1901. Sorex personatus lesneur'.i Miller and Iiehn, Proc. Boston Soc. Nat. Hist. 30 : 235, December 27. 1901. Sorax (sic) personatus lesueurii Hollister, Proc. U. S. Nat. Mus. 40: 378, April 17, 1911. Sorex longirostris lesueurii Hollister. Proc. U. S. Nat. Mus. 40: 380, April 17, 1911. Sorex forest eri (sic) Fleming, Nat. Hist. Toronto Region (Publ. by Canadian Inst.), p. 209, November, 1913. Sorex cinereus cinereus Jackson, Journ. Mamni. 6: 56, February, 1925. Sorex frankatounensis Petersoh, Ann. Carnegie Mus. 16.: 2!»2. March. 1926. Type locality. Frankstown Cave, near Ilollidaysburg, Blair County, Pa. Type specimen. — None now known to exist. Type locality. — Fort Severn, Ontario, Canada. Geographic raiu/e. — Northern Quebec and all of northern Canada, west to central Alaska, northern Kenai Peninsula, western British Columbia (except coastal region) ; south to New Jersey, the moun- tains of North Carolina and Tennessee, central Ohio, southern Indi- ana, northern Illinois, northeastern Iowa, eastern Minnesota, north- ern and eastern Manitoba, northern Saskatchewan, through the mountains of Idaho, western Montana and western Wyoming to northern New Mexico, and northeastern and central Washington (fig. 3). Diagnostic characters. — Size medium ; larger than Sorex fontinalis, S. lyelli, 8. preblei, 8. o. liaydqni or 8. c holHsteri, smaller than 8. o. streatori or S. c. miscix. Darker and more brownish (less grayisb) than miscix, par- ticularly in winter pelage ; skull shorter than that of miscix with relatively shorter and broader rostrum, and lower brain case. Larger and with relatively longer 1 ail than S. c. haydeni; color darker both in summer and winter, par- ticularly on the sides, tending less to development of tricolor pattern; skull averaging slightly larger than that of haydeni, with relatively and actually 42 NORTH AMERICAN FAUNA [No. 51 longer palate, and relatively narrower rostrum. Larger and with slightly longer tail than hollisteri; color darker, particularly in summer, underparts in both winter and summer pelages less whitish ; skull about the size of that of hollisteri or slightly larger, with broader rostrum. Paler than streatori, especially on the underparts ; tail shorter and hind feet smaller ; skull smaller than that of streatori, with weaker rostrum. Color. — Winter pelage: Upper parts grayish fuscous, frequently tending toward chaetura drab or hair brown, extending well down on the sides ; underparts smoke gray or between smoke gray and pale smoke gray ; tail essentially bicolor, chaetura drab or fuscous above, buffy below nearly to tip. Summer pelage: Much more brownish than winter pelage. Upper parts rather variable, fuscous- black, mummy brown, or sometimes Prout's brown, usually extending well onto the sides and gradually blending with colors of underparts ; sometimes with a distinct lateral ribbon of drab or buffy brown ; underparts usually smoke gray, sometimes light grayish olive, or even tinged with deep olive-buff tending toward avellaneous ; tail as in winter. Skull. — Medium in size for the cinereus group (condylobasal length about 15.8 mm.). Smaller than that of S. c.-miseix, with relatively shorter and broader rostrum, and shallower brain case. Averaging larger than that of S. c. haydeni, with relatively and actually longer palate, relatively narrower rostrum, and usually with less densely pigmented dentition. About the size of that of S. c. hollisteri or slightly larger, with broader, less attenuate rostrum. Smaller than that of jS. c. streatori, with rostrum smaller, dentition weaker, and the molariform teeth usually less deeply emarginate posteriorly. Measurements. — Adult male and adult female from Drury Run, Clinton County, Pa. : Total length, 98 ; 95 ; tail vertebrae, 40 ; 39 ; hind foot, 11 ; 12. Average of four adult females from Washington County, R. I. : Total length, 100.5 (99-102) ; tail vertebrae, 41.5 (40^13) ; hind foot, 12.4 (11.8-12.7). Average of three adult females from Mamie Lake. Vilas County, Wis. : Total length, 101.7 (99-103) ; tail vertebrae, 38.7 (38-40) ; hind foot, 12 (12-12). Average of three adult males from Pahaska (mouth of Grinnell Creek), Park County, Wyo. : Total length, 96.3 (95-97) ; tail vertebrae, 41.7 (39-44) ; hind foot, 12 (12-12). Skull: Skull of adult male (teeth moderately worn) and adult female (teeth slightly worn) from Drury Run, Clinton County, Pa.: Condylo- basal length, 15.6 ; 16.0 ; palatal length. 5.9 ; 6.0 ; cranial breadth, 7.6 ; 7.6 ; inter- orbital breadth, 3.1 ; 2.9 ; maxillary breadth, 4.1 4.0 maxillary tooth row, 5.5 ; 5.5. Average of four skulls of adult females (teeth slightly worn) from Wash- ington County, R. I.: Condylobasal length, 15.7 (15.4-16.2); palatal length, 6.0 (6.0-6.0) ; cranial breadth 7.6 (7.6-7.7) ; interorbital breadth, 3.0 (2.9-3.1) ; maxillary breadth, 4.1 (4.0-4.3) ; maxillary tooth row, 5.5 (5.4-5.6). Average of four skulls of adult males ( teeth moderately worn ) from Roan Mountain, N. C. : Condylobasal length, 15.5 (15.1-15.9) ; palatal length. 5.9 (5.8-6.0) ; cranial breadth, 7.5 (7.3-7.6) ; interorbital breadth, 3.1 (2.9-3.2) ; maxillary breadth, 4.1 (4.0-4.2) ; maxillary tooth row, 5.4 (5.3-5.5). Average of five skulls of adult females (teeth slightly worn) from Mamie Lake, Vilas County, Wis.; Condylo- basal length, 15.8 (15.5-16.0) ; palatal length, 6.0 (5.8-6.1) ; cranial breadth, 7.6 (7.5-7.8) ; interorbital breadth, 3.0 (3.0-3.1) ; maxillary breadth, 4.0 (3.9-4.1) ; maxillary tooth row, 5.5 (5.3-5.7). Average of three skulls of adult males (teeth slightly worn) from Pahaska (mouth of Grinnell Creek), Park County, Wyo.: Condylobasal length, 15.5 (15.4-15.5); palatal length, 6.0; (6.0-6.0); cranial breadth. 7.6 (7.4-7.7) ; interorbital breadth, 3.0 (3.0-3.1) ; maxillary breadth, 4.0 (4.0-4.1) ; maxillary tooth row, 5.4 (5.3-5.5). Average of five skulls of adult females (teeth slightly worn) from Fort Resolution, Northwest Territories : Condylobasal length, 16.0 (15.9-16.1); palatal length, 6.0 (6.0- 6.0) ; cranial breadth. 7.8 (7.6-8.0) ; interorbital breadth, 3.0 (2.9-3.1) ; maxil- lary breadth, 4.0 (3.9-4.1) ; maxillary tooth row, 5.5 (5.4-5.6). Remarks. — The common long-tailed shrew of the northeastern United States and Canada was first given recognition with a valid scientific name by Kerr (1792, p. 206), who described the species under the name Sorex arcticus cinereus, basing his description upon the account given by Pennant (1784, p. 139), which in turn was based upon that of Forster (1772, p. 381), neither Pennant nor Forster giving the animal a Latin designation (Jackson, 1925a, p. 55). 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 43 Several years later, November 17, 1826, Isidor Geoffroy Saint Hilaire read an account of the animal before the Societe d'Histoire Naturelle at Paris, and the following year, 1827, he published two descriptions of the species under the name Sorex personatus, which has been generally used for the species since Miller's and Merriam's revisions of 1895. The earlier of these descriptions appeared in Dic- tionnaire Classique d'Histoire Naturelle (Geoffroy, 1827a, p. 319). This volume bears date of January, 1827, on the title page, and was actually distributed before February 10, 1827. The other account, and the one which has heretofore been quoted as the original descrip- tion of personatus, appeared in Memoires du Museum d'Histoire Naturelle (Geoffroy, 1827b, p. 122). The latter volume bears as a date upon the title page only the year, 1827; Sherborn, however (Sherborn, 1914, p. 368), is authority for fixation of its date of pub- lication as December, 1827. Further evidence that the article in the Dictionnaire appeared earlier than the one in the Memoires is fur- nished in a footnote in which Geoffroy states : Ce Memoire a ete compose en octobre 1S26, et lu le 17 novembre suivant S. la Societe d'Histoire naturelle : quelques recueils scientifiques et meme quelques gazettes en ont rendu compte a cette epoque, d'apres ma lecture; et j'eu ai inerxo>t. Soper coll., Edmonton, Alberta. wColo. Mus. Nat. Hist. '-Nat. Mus. Canada, 1. "State Coll. Wash. u Univ. Mich. "Acad. Nat. Sci. Philadelphia. "Amer. Mus. Nat. Hist., 12. -'Mus. Comp. Zool., 2. 48 NOETH AMERICAN" FAUNA [No. 51 Michigan: Alger County, 2 13 ; Ann Arbor, 524; Chippewa County, 313; Fish Hawk Lake, Gogebic County, 513; Honey Creek, Washtenaw County, 1 13 ; Michigamme, 3 ; Palmer, 1 ; Pleasant Lake, Livingston County, 1 M ; Porcupine Mountains, Ontonagon County (T. 51 N., R. 43 W., S. 14), 1 13 ; Rush Lake, Huron County, 1 13 ; Whitefish Point, Chippewa County, l.13 Minnesota: Burntside Lake, 1; Elk River, 60; Fort Ripley, 1; Fort Snel- ling, 31 *; Hinckley, 3; Long Prairie, 1; Minneapolis, 12; Ottertail County, 2 2S ; Princeton, 1 ; Steele County, 1 *■ ; Tower, 1. Montana: Big Belt Mountains (Camas Creek, 4 miles south of Fort Logan), 1; Big Timber (14 miles south), 1; Boulder Creek (8 miles south of Big Timber), 1 ; Bozeman, 1"; Carter (National Bison Range), 1 ; Chief Mountain, 1 ; Crazy Mountains ( near head Big Timber Creek ) , 2 ; Deer Lodge County, 1 ; Dry Creek, 1 ; Fish Creek, Glacier Park, 1 ; Florence, 5^; Highwood Mountains, 2; Hilger (7 miles northeast), 2; Indian Creek, Glacier Park, 1 ; Lake McDonald, 1 ; Little Belt Mountains (Sheep Creek, 16 miles north of White Sulphur Springs), 2; Little Belt Mountains (Dry Wolf Creek, 20 miles southwest of Stanford), 2; Lolo, 1 ; St. Marys Lake, 4 m ; Stevensville, 1 ; Sun River, 1 ; Waterton Lake, 1 ; West Fork of West Gallatin River (Gallatin National Forest, altitude 6,500 feet), 7; Yellowstone (4 miles southwest), l30; Zortman, 1; Zortman (Ruby Creek), 1. New Brunswick: Bathurst (15 miles from, Miramichi Road), 40s; Gulquac Lake, Victoria County, 4 4 ; Hampton, 1 ; Long Lake, Victoria County, 4"; Maugerville, Sunbury County, 231; Point Le Preaux, lc; Resti- gouche River, 1 e ; St. Johns, 1 ; Tobique Point, Victoria County, 3 * ; Tobique River (forks of), Victoria County, 13 4; Tracy Station, Sun- bury County, 1 £ ; Trousers Lake, Victoria County, 40 x ; Yougall, 17.8 New Hampshire: Fabyans, 1; Fitzwilliam, 1; Mount Washington (sum- mit), 4; Ossipee, Carroll County, 1. New Jersey: Beach Haven, Ocean County, l22; Bear Swamp (south side), Burlington County, 322; Bridgeport (near), Gloucester County, l22; Cape May, 16 22 ; Chairville Pond, Burlington County, 3 22 ; Essex County Park, 3 4 ; Haddonfield, 1 ^ ; Mauricetown, Cumberland County, 7 22 ; Mays Landing, 23 32; Millburn, 54; Port Norris, Cumberland County, 8 22 ; South Mountain Reservation, Essex County, 19 4 ; South Orange Reservation, 4 4 ; Tabor, 2 * ; Tuckerton, 5. New Mexico: Pecos Baldy, 1; Pecos Baldy (altitude 11,000 feet), 1; Twin- ing (altitude 10,500-10,700 feet), 6. New York: Adirondacks, 1; Amity ville, Long Island, 1; Berlin (altitude 1,100 feet), 24; Big Moose Lake, 2; Catskill Mountains, 2; Fort Totten, 4 ; Gull Lake, Adirondack Mountains, 1 ; Highland Falls, 1 ; Locust Grove, 7 ; Minerva, 1 4 ; Montauk, 2 4 ; Montauk Point, Suffolk County, 11 ; Mountain View, 3 ; Northwood, 10 * ; Ossining, 1 ; Peterboro, 2 ; Piseco, 1; Point Rock, 1; Tupper Lake, 36; Waterville, 1; West Hampton, 1. North Carolina: Roan Mountain, 4; Roan Mountain (altitude 4,700 feet), 2; Roan Mountain (altitude 6,000 feet), 14; Roan Mountain (altitude 6,300 feet), 1. Northwest Territories : Anderson River region, 5 ; Anderson River region (Fort Anderson), 3; Anderson River region (Lower Anderson River), 2 ; Fort Liard, 1 ; Fort Norman, 2 ^ ; Fort Providence, 10 ; Fort Rae, 23; Fort Resolution, 15; Fort Resolution (Mission Island), 2; Fort Simpson, 11; Fort Smith, 2s9; Fort Wrigley, 2; Franklin Bay, 2; Grandin River, 1; Great Bear Lake (Fort Franklin), 24; Great Slave Lake, 5; Great Slave Lake (Big Island), 1; Harrowby Bay, l4; Horton River (Coal Creek), 24; Kozaryuak River, Coronation Gulf, Is; Lake St. Croix, 3; Mackenzie River (Nahanni River Mountains), 3; Old Fort Good Hope (near), 1; Peels River, 4; Richard Island (east of, 4 Amer. Mus. Nat. Hist. m Mont. State Coll. 0 Acad. Nat. Sci. Philadelphia. 28 Mont. State Coll., 1. 8 Nat. Mus. Canada. ffi Amer. Mus. Nat. Hist., 1. 13 Univ. Mich. 30 D. R. Dickev coll., Pasadena, Calif. 22 Acad. Nat. Sci. Philadelphia. 31 Miramichi Nat. Hist. Soc, Chatham, M Univ. Mich., 1. New Brunswick. 16 Amer. Mus. Nat. Hist, 4. ^Acad. Nat. Sci. Philadelphia, 12. 20 G. G. Cantwell coll., Palms, Calif. 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 49 Mackenzie Delta), l4; Slave River (100 miles below Fort Smith), 3; Toker Point (south of), 3." Nova Scotia: Barren, Victoria County, Is; Barrington Passage, 23 8; Brier Island (Digby Neck, extremity), 1; Camp Point, Cape Breton Island, Is; Cheticamp Lake, Cape Breton Island, 4s; Digby, 12 *; Halifax, 9 7 ; Ingonish Centre, 1 8 ; James River, 5 ; Kedgemakooge Lake, 1 ; Kings County, 66; Little River (Digby Neck, 1/2 mile from shore), 3; Newport, 2.4 Ontario: Algonquin Park, 6; Dows Swamp (near Ottawa), l8; Emsdale. 5; Lac Seul, l8; Humboldt Bay, Lake Nipigon, 3s4; Long Swamp (near Billing Bridge), l8; Lome Park, 2s; Macdiarmid (Lake Nipigon), 5M; Macgregor Bay (District of Manitoba), 3"; Michipicoton Island, 6; North Bay. 2 9 ; Ottawa. 1 ; Rat Portage. 1 ; Sand Lake, 2. Ohio: Cleveland, 1; Ellsworth, 1; Milford Center, 1. Pennsylvania: Drury Run, Clinton County (type locality of finibripes),S; Frankstown Cave, near Hollidaysburg, Blair County (type locality of frankstouiiensis), 1 31a ; Kennett Square, 2"; Lake Ganoga, Sullivan County, 2"; Lake Leigh (North Mountain), 2s2; Summit Mills, Somer- set County, 1." Prince Edward Island: Alberton, l4; Georgetown, S29; Kensington, 249; Lennox Island, 2; Mount Stewart, 7." Quebec: Alymer, 2"; Berry Mountain Camp, Matane County, 1 8 ; Big Island, Biue Sea Lake, l8; Burbridge, 28; Clearwater Lake, l8; Federal Mine, Gaspe County, 28; Fort Chimo, 7; Gaspe Peninsula (Cascapedia River), l4; Gaspe Peninsula (Cascapedia River, Lazy Bogan Mountain), l4; Gaspe Peninsula (Cascapedia River, Loon Lake), l4; Gaspe Peninsula (Cascapedia River, Tracadie), 24; Godbout, 58; Lac Aux Sables, 8"; Lake Edward, 16 e; Mount Albert, 54; Rupert, 2; St. Rose, 4; Seal Lake, l8: Ste. Anne des Monts (Gaspe), 34; Ste. Anne River, Gaspe • County, 4.8 Rhode Island: Washington County, 15. Saskatchewan: Lake Athabaska (Fair Point), 1; Lake Athabaska (Poplar Point), 4'; Lake Athabaska (8 miles northeast of Moose Island), 2'; Lake Athabaska (mouth Beaver River), l.10 Tennessee: Roan Mountain, Carter County (altitude 6,000 feet), 2E; Roan Mountain, Carter County (altitude 6,200 feet), l.22 Vermont: Burlington, 2; Mount Mansfield, 4; Newfane, l4; Pico Peak, 54. Virginia: Mount Rogers. Grayson County (altitude 5,719 feet), 1. Washington: Bauennan Ridge (west end, at Tungsten Mine, altitude 6,800 feet, Okanogan County), 1; Conrad Meadows (3 miles above, south fork Tieton River, altitude 4,200 feet), l21; Curlew (5 miles west. Ferry County, altitude 2.800 feet). 2; Lake Chelan (head of), 1; Lake Keeehelus. I13: Loon Lake. Stevens County (altitude 2,-100 feet), 2; Metaline, Pend Oreille County. 1; Mount Rainier (Paradise Creek, altitude 5.200 feet). 1: Tunk Mountain. Okanogan County (altitude 3.500 feet), 1; Yakima Indian Reservation (Signal Peak, altitude 4,000 feet. Yakima County), 1. West Virginia: Cranberry Glades. Pocahontas County, 5: Jobs Knob, 8s8; Pocahontas County (near head of Cranberry River), 8. Wisconsin: Beaver Dam, 3M; Cataline. Marinette County, 4 s7; ClarksLake, Door County, 2; Connors Lake (18 miles west-northwest of Phillips, Sawyer County), 8; Conover (near), Is8; Crescent Lake, Oneida County, 8; Danbury, 1; Delavan. 7W; Dousman, 1; Eagle River, 2; Elco, 240; Elkhart Lake (Sheboygan Swamp), 6; Ellison Bay, 1; Ellsworth, Pierce County, 1 : Fish Creek. 1 : Florence, 3 ; Herbster. 4 ; Kelley Brook, Oconto County, 2 8T ; Kelly Lake, Oconto County, 11 ; Lac Vieux Desert, 438; Lake St. Germain, Vilas County, 6; Lakewood, 10: Long Lake, Washburn County, 13: Madeline Island (Apostle 4 Amer. Mus. Nat. Hist. M Royal Ontario Mus. Zool. ■Acad. Nat. Scl. Philadelphia. 311(':iniezic Mus. 7 Mus. Comp. Zool., 1. » Mus. Comp. Zool., 6. "Nat. Mne. Canada. so D. R. Dickev coll., 1; Univ. Wis. Zool 10 Mus. Vert. Zool., Univ. Calif. Mus.. 1; Field Mus. Nat. Hist., 1. 13 Univ. Mich. '7 Public Mus. Milwaukee. »Amer. Mus. Nat. Hist, 2. « Field Mus. Nat. Hist. 21 State Coll Wash ""Public Mus. Milwaukee, 2. 22 Acad. Nat. Sd. Philadelphia. "' Dnlv. Wis. Zool. Mus. aAmer. Mus. Nat. Hist., 1. 40 Mus. Comp. Zool. MMus. Comp. Zool., 8. 50 NORTH AMERICAN FAUNA [No. 51 Islands), 3; Mamie Lake, 15; Mather, 1; McAllister, 2; Mercer, 3; Meridean, 2 ; Milton, 1 3S ; Milwaukee, 1 ; Milwaukee County, 1 3T ; Nashotah, 1 ; Newport, Door County, 1 37 ; Ogeina, 6 ; Outer Island (Apostle Islands), 12; Prairie du Sac, Sauk County, 437; Prescott, Pierce County, 3 37 ; Presque Isle (Apostle Islands), 1; Racine, 3; Rhinelander, 84f; Rib Hill, Marathon County, 7; Sand Island (Apostle Islands), 1; Sayner, 22 ss; Solon Springs, ll42; Spread Eagle, l88; Stevens Point, 1 ; Sumner, 2 3S ; Washington Island, Door County, 3 ; Wild Rose, 3 ; Withee, 2. Wyoming: Big Horn Mountains (west slope, head of Trappers Creek, alti- tude 8,500 feet), 19; Big Horn Mountains (west slope, head of Trap- pers Creek, altitude 9,500 feet), 1; Big Horn Mountains (head north fork Powder River ) , 1 ; Big Piney, 1 ; Black Rock Creek, Lincoln County, 1; Cokeville, 1; Evanston, 1; Moran (1 mile north), 5; Moran (Lake Emma Matilda), 6; Pacific Creek, 2; Pahaska (Grinnell Creek), 8; Pahaska (Grinnell Creek, altitude 7,000 feet), 15; Pahaska (mouth of Grinnell Creek, altitude 6,600 feet), 3; Pahaska Tepee (mouth of Grinnell Creek, altitude 6,300 feet), 6; Salt River Mountains (10 miles southeast of Afton, 7,500 feet), 2; Sierra Madre Mountains (south base, Bridger Peak, altitude 8,800 feet), 1; Teton Mountains (Moose Creek, altitude 6,800 feet), 8; Teton Pass (above Fish Creek, altitude 7,200 feet), 12; Valley (altitude 7,500 feet), 2; Yellowstone Park (Mountain Creek), 1. Yukon: Caribou Crossing (Yukon River), 2; Chandindu River (Yukon River), 1; Dawson (near Stewart River), 1; Forks Macmillan River, 1; Fort Selkirk (50 miles below, Yukon River), 1; Lake Lebarge (Yukon River), 1; Yukon River (Carbiou Crossing), 1. SOREX CINEREUS MISCIX Bangs Labrador Cinereous Shrew (Pl. 2, b) Sorex personatus miscix Bangs, Proc. New England Zool. Club 1: 15, February 28, 1899. [Sorex] [merriami] miscix Elliot, Field Columb. Mus. Publ. 45 (zool. series 2) : 376, 1901. Sorex cinereus miscix Jackson, Journ. Mamm. 6: 56, February, 1925. Type specimen. — No. 8651, Museum of Comparative Zoology, Har- vard College, Bangs collection; $ adult (teeth very slightly worn), skin and skull ; collected October 10, 1898, by Ernest Doane. Type locality. — Black Bay, Labrador. Geographic range. — Labrador south of latitude 58° north. (Fig. 3.) Diagnostic characters. — Somewhat larger than Sorex c. cinereus, with paler, more grayish color, particularly in winter pelage. Skull longer than average skulls of S. c. cinereus, with relatively longer and narrower rostrum and higher brain case. Color. — Winter pelage: Upper parts drab more or less tending toward smoke gray, sometimes almost grayish hair brown ; becoming paler on the sides. Underparts between pale olive-gray and pale smoke gray, sometimes tinged with pale pinkish buff. Tail drab above ; avellaneous to light pinkish cinnamon below, nearly to tip. Summer pelage: Darker and more brownish than winter pelage. Upper parts sepia or slightly paler, usually becoming somewhat darker on the rump, and paler on the sides. Undei'parts pale smoke gray sometimes tinged with pinkish buff or avellaneous. Tail about as in winter. Skull. — Similar to that of $. c. streatori. Longer than average skulls of S. c cinereus, with relatively longer, narrower, and more attenuate rostrum, highei brain case, and longer tooth row. Measurements. — Type specimen (adult male) : Total length, 104 ; tail vertebrae 44 ; hind foot, 14. Average of eight adult females from type locality : Tota length, 100.9 (95-110) ; tail vertebrae, 43 (39-46) ; hind foot, 13.1 (12-14) 37 Public Mus. Milwaukee. « Univ. Wis. Zool. Mus., 2. 38 Field Mus. Nat. Hist. ^ Field Mus. Nat. Hist., 10. 19:28] REVIEW OF AMERICAN LONG-TAILED SHREWS 51 Skull: Type specimen (adult male; teeth very slightly worn): Condylobasal length, 16.9 ; palatal length, 6.6 ; cranial breadth, 8.1 ; interorbital breadth, 3.2 ; maxillary breadth, 4.0; maxillary tooth row, 6.0. Average of four skulls of adult females (teeth very slightly worn) from type locality: Conciylobasal length, 16.6 (16.5-16.8) ; palatal length, 6.6 (6.4-6.7) ; cranial breadth, 8.1 (8.0-8.3) ; interorbital breadth, 3.2 (3.1-3.2) ; maxillary breadth, 4.1 (4.1-4.2) ; maxillary tooth row, 5.8 (5.7-5.9). Remarks. — The Labrador form of S. cinereus is at best a poorly defined subspecies, distinguished from typical cinereus chiefly by slight average color differences in winter pelage. The skull of S. c. miscix also averages larger and with narrower rostrum than that of the subspecies cinereus, but there are many skulls of the latter, par- ticularly from the northern part of its range, which are practically inseparable from skulls of miscix. Certain specimens of cinereus from eastern Quebec, Prince Edward Island, Nova Scotia, and New Brunswick show a tendency in color toward miscix but on the whole are nearer S. c. cinereus, to which they have been referred. Specimens examined. — Total number, 75, as follows : Labrador: Black Bay (type locality), 48 i3; Hopedale, 343; L'Anse au Loup, 543; Maddovik, 343; Okak, 443; Paradise, 9; Paradise River (20 miles above mouth), 2; Sandwich Bay, 1." SOREX CINEREUS HAYDENI Baird Plains Cinereous Shrew (Pl. 2, c) Borcx ha yd en i Baird. Report Pacific R. R. Survey 8: part 1, Mammals, p. 29, 1857. Borew personatus hai/dciii Alien. Bui. Amer. Mus. Nat. Hist. 8: 257. November 25, 1896. Borcx cinereus haydeni Jackson, Journ. Mamni. 6: 56, February, 1925. Type specimen. — No. 1685, U. S. Nat. Mus., adult, sex unknown, alcoholic with broken skull unremoved ; collected in 1855 by F. V. Hay den. Type localitij. — Fort Union, Nebr. (later Fort Buford, now Mon- dak. Mont., near Buford, Williams County, N. Dak.). Geographic range. — Extreme east-central Alberta (Islay), southern Saskatchewan, southwestern Manitoba, south through extreme west- ern Minnesota to northwestern Iowa, northern Nebraska, and through eastern Montana to southeastern Wyoming. (Fig. 3.) Diagnostic characters. — Smaller than Surer c. cinereus, with shorter tall; color paler, both in summer and winter, tending more to develop a tricolor pattern, darkest on the back, paler on the sides in a longitudinal ribbon, palest on the underparts; skull slightly smaller than that of 8. c. cinereus, with rela- tively and actually shorter palate and relatively broader rostrum. About the size of 8. c. hollisteri, or slightly smaller, with shorter tail, and tending more to develop tricolor pattern (dorsally usually darker, paler on the sides); ros- trum decidedly shorter and broader than in hollisteri. Color. — Winter pelage: Upper parts averaging paler than corresponding pelage of 8. c. cinereus, usually hair brown or drab, sometimes nearly fuscous posteriorly; color of upper parts usually not extending onto the sides, which are generally rather sharply defined from the back and more nearly the color of the underparts; underparts between pale olive-gray and pale smoke gray, sometimes smoke gray : tail cinnamon brown or Saccardo's umber above, avel- laneous or pinkish buff below, nearly to tip. Summer pelage: Paler than cor- responding pelage of 8. c. cinereus. Upper parts hair brown, drab, or olive- brown, paler on the sides, which frequently are rather sharply contrasted from *a Mus. Comp. Zool. 52 NORTH AMERICAN FAUNA [No. 51 both the back and underparts in a distinct longitudinal ribbon of wood brown or avellaneous ; under parts smoke gray or pale smoke gray, usually faintly tinged with pinkish buff ; tail as in winter. Skull. — Smaller than that of S. c. cinereus (condylobasal length about 15.3), with relatively and actually shorter palate, relatively broader rostrum, and more densely pigmented dentition. About the size of that of S. c. hollisteri, but with decidedly shorter, broader, less attenuate rostrum. Measurements. — Average of five adult females from Lostwood, N. Dak. : Total length, 8S.2 (85-92); tail vertebrae, 32.8 (30-36); hind foot, 11.4 (11-12). Skull: Skulls of two adult females (teeth slightly worn) from Lostwood, N. Dak. : Condylobasal length, 15.3 ; 15.0 ; palatal length, 5.5 ; 5.5 ; cranial breadth, 7.5; 7.5; interorbital breadth, 3.0; 3.0; maxillary breadth, 4.3; 4.4; maxillary tooth row, 5.2; 5.0. Skull of adult female (teeth slightly worn) from Williston, N. Dak. : Condylobasal length, 15.0 ; palatal length, 5.6 ; cranial breadth, 7.3 ; interorbital breadth, 3.0 ; maxillary breadth, 4.2 ; maxillary tooth row, 5.2. Remarks. — The pale, short-tailed form of S. cinereus inhabiting the northern part of the central plains region is confined largely to the Transition Zone. It is a rather variable form particularly in cranial characters, which show7 everywhere in a broad border along its range an approach toward S. c. cinereus. Specimens referable to S. c. haydeni have been examined from as far north as Osier, Saskatche- wan; these are typical of haydeni in color and measurements, but show an inclination toward the subspecies cinereus in having rela- tively longer and narrower rostra than specimens from the type region. The same is true of specimens from Laramie Peak and Fort Steele, Wyo., Portland, N. Dak., and most localities in eastern Mon- tana. Others from southwestern Manitoba are essentially like typical haydeni in external characters but may show cranially a strong tend- ency toward S. c. cinereus. A large series from Aweme, Manitoba, is particularly puzzling and interesting. Nearly all the skins in this series are like typical haydeni in color and measurements, there being only a few that are like typical cinereus in color and only a few that have measurements of that subspecies. The skulls are extremely variable, ranging from a half dozen of those of the extreme style of haydeni (small and flat, with short and relatively broad rostrum) to several of the extreme stjde of 8. c. cinereus (larger and decidedly higher, with long and narrow rostrum), with a few displaying inter- mediate skull characters. The contrast between the extremes in this series, if only the extremes from this locality were considered, is so great as to suggest different species. Certain specimens from the Turtle Mountains, N. Dak., are intermediate and could with about equal propriety be referred to either haydeni or true cinereus were they not surrounded geographically by haydeni. Alcoholic speci- mens, from which the skulls have been removed for study, from Fort Sisseton, S. Dak., and a skin with skull from Browns Valley, Minn., can also about as w^ell be referred to one form as the other. Specimens examined. — Total number 206, as follows: Alberta: Islay, 2.44 Iowa: Sac City (2 miles west), 1; Wall Lake, 1. Manitoba : Aweme, 64 iS ; Carberry, 6 ; Killarney, 1. Minnesota: Browns Valley, 1; Kittson County, 1; Madison, l46; Moor- head, 6. ** J. D. Soper coll., Edmonton, Alberta. 46 Royal Ontario Mus. Zool., 6 ; Stuart Criddle coll., Treesbank, Manitoba, 54. "Acad. Nat. Sci. Philadelphia. 1928] REVIEW OF AMERICAN LON"G-TAILED SHREWS 53 Montana: Crow Agency, 1; Ekalaka (5 miles southeast), 3; Fort Custer, 7 ; Medicine Bow, 1. Nebraska: Bassett, Rock County, 4"; Kennedy, 1; Niobrara River (10 miles south of Cody), 1; Perch, Rock County, 10 47; Two-mile Lake, Cherry County, 1. North Dakota: Blackmer, 2; Bottineau. 1; Cannon Ball, 1: Fairmount (Sioux River), 1; Fargo, 2; Fish Lake (Birchwood), 2; Fort Buford (type locality), 1; Grand Forks, 1; Kenmare, 2; Lostwood (6 miles north), 9; Oakes, 2; Portland, 7; St. John (Fish Lake, 8 miles north of west), 1; Self ridge, 1; Steele, 1; Turtle Mountains (Birchwood), 2; Walhalla, 2; Williston (south of river). 1. Saskatchewan: Indian Head, 12 4S; Osier. 3.49 South Dakota: Beadle County, 1; Custer, 3s0; Custer (16 miles west), 2; Deadwood, 2; Dumont (Black Hills National Forest, altitude 6.100 feet), 5; Elk Mountain (20 miles north), 1; Fort Pierre, 2; Fort Sisse- ton, 8 ; Vermilion, 1. Wyoming: Bear Lodge Mountains (Warren Peak, altitude 6,000 feet), 4; Fort Steele, 1; Laramie Peak (north slope, altitude 8,000 feet), 1; Laramie Peak (north slope, altitude 8,800 feet), 1; Rattlesnake Creek (Black Hills, altitude 6,000 feet), 1; Sherman, 1; Springhill, 1; Sun- dance, 3; Wolf (Eaton's Ranch), 1. SOREX CINEREUS STREATORI Merriam Streatob Cinereous Shrew (Pl. 2, D) Sorcx personatus streatori Merriam, North Amer. Fauna No. 10, p. 62, December 31, 1892. 8orex cinereus streaUrrl Jackson, Journ. Mamin. 6: 56, February, 1925. Type specimen. — No. 73537, U. S. Nat. Mus., Biological Survey collection; $ adult (teeth moderately worn), skin and skull; col- lected July 9, 1895, by C. P. Streator. Type locality. — Yakut at, Alaska. Geographic range. — Pacific coast region of North America from the southeastern part of Kenai Peninsula, Alaska, south to central Washington west of the Cascades. (Fig. 3.) Diagnostic characters. — Darkest of the sixties : largest of the wist American subspecies, with longest tail and largest hind foot. In color most nearly like Sorex c. cinereus but darker, particularly on ventral parts; larger, with longei tail and larger hind foot; skull longer than that of the subspecies cinereus, with longer and usually heavier rostrum, heavier dentition, the molariform teeth usually more deeply emarginate posteriorly. Decidedly larger in :ill respects than 8. c. hollisteri; much darker, the underparts buffy rather than whitish; skull much larger and heavier, with distinctly larger rostrum. About the size of »S'. c. miscix, but very much darker in all pelages; skull much like that of miscix but teeth greater in extero-interior diameter. Color. — Winter pelage: Upper parts scarcely different from corresponding pelage of 8. c. cinereus, underparts decidedly more buffy. Upper parts grayish fuscous or chaetura drab, sometimes tending toward hair brown, extending well down on sides where it becomes slightly more drabblsfa, blending gradually with color of underparts drabblsb avellaneous ; tail fuscous above, avellaneous below nearly to tip. Summer pelage: Much darker and more brownish than in winter. Darker than 8. c. cinereus in corresponding pelage. Upper parts fus- cous-black or fuscous, extending well onto the sides and gradually blending with colors of underparts; underparts heavily tinged with buffy brown or wood brown, sometimes with pinkish lnifT, rarely indicating more or less light grayish olive; tail about as in winter. " Amer. Mus. Nat. Hist. 4» Mus. Coinp. Zool. 48 Acad. Nat. Scl. Philadelphia, 3; Nat. »Amcr. Mus. Nat. Hist., 1. Mus. Canada, 9. 54 NORTH AMERICAN FAUNA [No. 51 Skull. — Large for the cinereus group (condylobasal length 16 millimeters or over), with rather large rostrum. Longer than that of S. c. cinereus, with longer and usually heavier rostrum, longer tooth row, heavier dentition, the molariform teeth usually more deeply and acutely emarginate posteriorly. Larger and heavier than that of S. c. Iwllisteri, with decidedly larger rostrum, and heavier dentition. Much like that of 8. c. miscix but teeth greater in extero-interior diameter. Measurements. — Type specimen (adult male): Total length, 107; tail verte- brae, 50 ; hind foot, 12.5. Average of four other adult males from type locality : Total length, 105.5 (102-111) ; tail vertebrae, 45 (42-48) ; hind foot, 12.8 (12-13). Average of three adult males from Sitka, Alaska: Total length, 104 (102- 107) ; tail vertebrae, 45.8 (45-47) ; hind foot, 13.3 (13-13.5). Skull: Type speci- men (adult male; teeth moderately worn) : Condylobasal length, 16.7; palatal length, 6.3 ; cranial breadth, 7.6 ; interorbital breadth, 3.2 ; maxillary breadth, 4.2; maxillary tooth row, 6.0. Average of four skulls of adult males (teeth moderately worn) from type locality: Condylobasal length, 16.3 (16.0-16.6); palatal length, 6.3 (6.2-6.4) ; cranial breadth, 7.7 (7.5-7.8) ; interorbital breadth, 3.2 (3.2-3.2) ; maxillary breadth, 4.1 (4.0-1.2) ; maxillary tooth row, 5.9 (5.8- 6.0). Average of three skulls of adult males (teeth slightly worn) from Sitka, Alaska: Condylobasal length, 16.6 (16.4-16.8); palatal length, 6.3 (6.2-6.4); cranial breadth, 7.7 (7.5-7.9) ; interorbital breadth, 3.2 (3.1-3.2) ; maxillary breadth, 4.2 (4.1-4.3) ; maxillary tooth row, 6.0 (6.0-6.0). Remarks. — Streator's cinereous shrew is confined to certain islands and a narrow strip of territory along the Pacific coast from the south side of Prince William Sound, Alaska, to the south side of the Strait of Juan de Fuca, Wash. In consideration of the extensive north and south range of the form it retains its characters through- out this area with remarkable uniformity. A short distance inland it intergrades with S. c. cinereus, as it does throughout extreme south- ern British Columbia, where specimens of S. c. cinereus from as far east as Glacier show tendencies toward S. c. streatori in color, size, and cranial characteristics. Specimens examined. — Total number, 365, as follows: Alaska: Anan Creek (mainland), 1"; Bradfield Canal, ll52; Chenega Island (Prince William Sound, 6 52 ; Chichagof Island (Tenakee Inlet), 21; Chickamin River (Behm Canal), 1 52 ; Cordova (Prince William Sound), 16 52; Cordova Bay (head of. Prince William Sound), 7 M ; Ellamar (Prince William Sound), 662; Ellington Island (Prince Wil- liam Sound), 19 52; Etolin Island, 252; Fools Inlet (Wrangell Island), 252; Fort Wrangell, 5°3: Freshwater Bay (Chichagof Island). I5" ; Glacier Bay, I0'2; Glacier Bay (Bartlett Cove), 952; Glacier Bay (Cop- permine Cove), 2s2; Grafton Island ( Prince William Sound), 2s"; Haines, 5; Hasselborg Lake (Admiralty Island), 1 52 ; Hawkins Island (east side of Canoe Passage, Prince William Sound), 3; Hawkins Island (west side of Canoe Passage, Prince William Sound), 352; Helm Bay, l52; Hinchinbrook Island (Northeast Bay, Prince William Sound). 16 m; Hoodoo Island (Prince William Sound), 17 52; Hoonah (Chichagof Island), 762; Horn Cliff (Beach), 1; Idaho Inlet (Chi- chagof Island), 1 52 ; Inian Islands, l51; Juneau, 11; Kelp Bay (Baranof Island), 1 51 : Kupreanof Island, 6; La Touche (La Touche Island, Prince William Sound), 19 52; Loring, 5; Mitkof Island, l52; Mitkof Island (Petersburg), 5; Orca, 1; Peril Strait (Baranof Island), 6 52 ; Port Conclusion (Baranof Island). 351; Port Frederic (Chichagof Island). 352; Port Nell Juan (head of. Prince William Sound), 12 52; Port Snettisham. 252; Quadra Cannery. 251; Quadra Lake. 4M; Red Bluff Bay (Baranof Island), 552: Redoubt Lake (Baranof Island), l51; Revillagigedo, l52; Rodman Bay (Baranof Island), 552; Sitka, 19"; Skagway, 6 : Taku River. 3 52 ; Thomas Bay, 6 52 ; Valdez Narrows (Prince William Sound), 8; Wrangell, 33 56; Takutat (type locality), 6. 51 D. R. Dickey coll., Pasadena, Calif. 53Am*ei\ Mus. Nat. Hist.. 2. 52 Mus. Vert. Zool. bo C. G. Cantwell, coll., Palms, Calif., 3; s'Ainer. Mus. Nat. Hist. D. R. Dickey coll., 19. w D. R. Dickey coll., 3. 1928] REVIEW OF AMERICAN LOXG-TAILED SHREWS 55 British Columbia: Howe Sound (Gibsons Landing), 1; Metlakatla, 2 5T ; Mount Baker Range (49th parallel, altitude 6,000 feet), 45S; Observa- tory Inlet (Hastings Arm), 1 50 ; Port Simpson, 3; River Inlet (head of Inlet), 4; Sumas, 1.5S Washington: Cedarville, 1; Glacier (altitude 900 feet), 1; Neah Bay. 1; Quiniault River (head North Fork, altitude 4,000 feet, Jefferson County), 1; Skohomish River (Mason County), 1 w ; Whatcom Pass (36 miles east of Glacier, altitude 5,200 feet, Whatcom County), 2.61 SOREX CINEREUS HOLLISTERI Jacksox Hollister Cinereous Shrew (PL. 2, E) Soiree personatus arcticus Merriam. Proe. Washington Acad. Sci. 2: 17. March 14, 1900. (Not Sorex arcticus Kerr, 1792.) Sorex cinereus holHsteri Jackson. Journ. llamm. 6 : 55. February, 1H25. Type specimen. — No. 99305, U. S. Nat. Mus., Biological Survey col- lection; 2 adult (teeth slightly worn), skin and skull- collected Sep- tember 14. 1899, by W. H. Osgood. Type locality. — St. Michael, Alaska. Geographic range. — Western Alaska from Franklin Point south to the head of Cook Inlet (Anchorage). (Fig. 3.) Diagnostic characters. — Similar externally to Sorex c. cinereus but averaging very slightly smaller with shorter tail ; color paler, particularly in summer ; underparts in both winter and summer pelage more nearly whitish; skull about the size of that of 8. c. cinereus or slightly sm;iller. with actually and relatively narrower rostrum and usually with higher brain ease. Much paler than 8. c. streatori both dorsally and ventrally ; smaller, with shorter tail, shorter hind foot, and smaller skull. Color. — Winter pelage: Similar to S. c. cinereus but averaging slightly paler and apparently more flecked with whitish-tipped hairs above, and more dis- tinctly whitish ventrally. Summer pelage: Much paler and more brownish than in winter. Decidedly paler than 8. <'. cinereus in corresponding pelage. Upper parts drab, sometimes almost olive-brown, usually shading into wood brown or avellaneous on the sides : underparts smoke gray or pale smoke gray, frequently tinged with pinkish buff: tail huffy brown or olive brown, sometimes almost clove brown above, avellaneous or light drab below nearly to tip. Skull— Averaging somewhat smaller than that of S. C cinereus, with actually and relatively narrower, more attenuate rostrum and usually with higher brain case. Smaller than that of S. c. Streatori, with decidedly smaller and narrower rostrum, and weaker dentition. Measurements. — Type specimen (adult female): Total length, 108; tail ver- tebrae, 39; hind foot. 12. Average of five adult females from type locality: Total length, 98 (92-104); tail vertebrae, 35.6 (32-38); hind foot, ii.9 (11-12.5). Average of six adult males from Kings Cove. Alaska Peninsula, Alaska: Total Length, 93.2 (90-97); tail vertebrae, 38.2 (36-41); hind foot, 11.7 (11-12). Skull: Typo specimen (adult female; teeth slightly worn): Condylobasal length, 15. 5; palatal length. 8.0; cranial breadth, 7.0: InterorbitaJ breadth, 2.9: maxillary breadth, 3.8; maxillary tooth row. 5.4. Average of five skulls of adult females (teeth slightly worn) from type locality: Con- dylobasal length, 15.7 (15.4-15.9) ; palatal length, 6.0 (5.9-6.1 ) : cranial breadth, 7.0 (7.4-7.8) : interorbital breadth. 3.0 (2.D-3.1) ; maxillary breadth, 4.0 (3.9-4.1) ; maxillary tooth row, 5.5 (5.4-5.G). Average of six skulls of adult males (teeth slightly worn) from Kings Cove, Alaska Peninsula, Alaska: Condylobasal length, 15.3 (15.2-15.5): palatal length. 5.9 (5.8-6.0): cranial breadth 7.5 (7.2-7.8) ; interorbital breadth, 3.0 (2.9-3.1) ; maxillary breadth, 3.8 (3.7-3.9) ; maxillary tooth row, 5.4 (5.3-5.5). 177 Nat. Mus. Canada, 1. "° T'niv. Mich. '•* Mas. ('(imp, ZooL » State Coll. Wash.. 1. =" Nat. Mus. Canada 56 NORTH AMERICAN FAUNA [No. 51 Remarks. — In its extreme form, represented by specimens from the Alaska Peninsula, S. c. hollisteri is a well-defined subspecies. It is confined to the coast region of Alaska from Alaska Peninsula north- ward, specimens from localities but little interior from the coast (vicinity of Lake Clark) showing an approach toward JS. c. cinereus, with which it intergrades in central Alaska. A single skin without skull from St. Lawrence Island, Alaska, is - provisionally referred to hollisteri. Specimens examined . — Total number, 283, as follows : Alaska: Akchookuk Lake, 2; Anchorage (Chester Creek), 1; Becharof Lake (Alaska Peninsula), 2; Bethel, 24 \ ( A* ' at \u\ i l$\ (I i 1 L^-~-*-"- J" \ 1 ' - — ■ 1 1 / "T ""* ; ^^A*£-—- ^— i |; \i4 ■\ <^- \^L V Fig. 4. — Geographic range of the subspecies of Sorex fumcus 1. 8. f. fumcus. 2. S. f. umbrosus. (antero-posterior diameter), the third larger than the fourth; internal ridge extending from apex of unicuspid to about one-half distance toward internal edge of cingulum, moderately pigmented near apex. The fumeus group can be distinguished externally from any forms of the arcticus or pribilofensis grouj>s by its distinctly unicolor (except tail) appearance. In general aspect the skull of fumcus resembles in proportions somewhat that of pribilofensis but is much larger in all respects, with the preorbital region less swollen and the infra- orbital foramen situated relatively farther back, the second upper premolar not lacking the distinct cusplike process on interior edge of basal shelf, and tlie unicuspid teeth without the internal- heavily pigmented ridge from apex to edge of cingulum completely developed. Compared with the skull of any of the 62 NORTH AMERICAN FAUNA [No. 51 arcticus group that of fumeus is flatter throughout, with considerably less deep, relatively narrower, and less angular brain case, less attenuate rostrum, nar- rower interorbitally, infraorbital foramen larger and relatively slightly farther back on rostrum, antero-posterior diameter of unicuspids relatively less, mo- lariform teeth more deeply emarginate posteriorly, and cusps of i1 narrower than in arcticus, the secondary cusp relatively smaller. Externally sometimes superficially like members of the cinereus group, but larger, particularly feet, and tail longer ; skull decidedly larger and heavier than any of the cinereus group, with distinctly heavier rostrum and dentition, and' with unicuspids lack- ing the pigmented ridge extending from apex of tooth to interior edge of cingulum (ridge is incomplete and only partly pigmented, near apex of tooth, in 8. fumeus). Color in winter pelage like S. dispar in summer pelage, but tail shorter and less hairy, and skull much heavier and broader in all proportions ; mesopterygoid space broader ; upper incisors larger, unicuspids heavier and of different relative sizes (in dispar the third and fourth are about subequal)^ Decidedly larger in all proportions than any of the longirostris group and dif-' fering in dentition. Similar in color to certain specimens of the trowbridgii group and not dissimilar in certan cranial features ; differs from any of troio- bridgii group in that unicuspids are relatively much wider (extero-interior diameter), wider than long (antero-posterior diameter), and third unicuspid is larger than fourth. Larger than any of the merriami group and never whitish on the ventral parts ; skull radically different, much larger than that of jS. merriami, more truncate posteriorly, relatively narrower interorbitally, much less swollen orbitally, rostrum relatively longer, more attenuate, and relatively narrower, particularly through infraorbital region ; infraorbital foramina larger and situated farther back ; dentition different. SOREX FUMEUS Mtlleb [Synonymy under subspecies] Geographic range. — That of the fumeus group. (Fig. 4.) Diagnostic characters. — Those of the fumeus group. Subspecies and geographic variation. — The species fumeus includes two sub- species : The typical form, fumeus, and umbrosus. The color of the species is fairly constant throughout its range, a tendency toward a reduction of reddish of the upper parts reaching its climax in the subspecies umbrosus in Nova Scotia. Geographical variation in the skulls is completely swamped by the extensive individual and local variation. Time of molting. — In the southern part of the range of Sorex fumeus the spring molt may begin as early as the middle of April, though most specimens at that time are still in full winter pelage. The transition in the south apparently occurs generally during May, for the majority of June specimens are in summer pelage. A male from Roan Mountain, N. C, shows first indica- tions of the molt April 19, 1893, while another from the same locality is in worn winter pelage May 3. A male from Mount Rogers, Va., has the molt about half completed June 22, 1903, while a second one collected at the same place and time is in complete winter pelage. Middle May specimens from Maryland and most middle and late June specimens from New York and Vermont are in complete summer fur, although a breeding female from Lake George, N. Y., has not fully completed the molt July 10. Two specimens, a male and a female, from Halifax, Nova Scotia, are in process of molt July 18, 1894. A male from River du Loup, Quebec, has acquired practically all the summer fur July 16, 1900. The winter pelage of 8. fumeus is usually obtained in full by the last week in October or first week in November. Occasionally specimens are in complete winter pelage early in October (Roan Mountain, N. C, October 11, 1892; Topsham, Me., October 3, 1915; Digby, Nova Scotia, October 10, 1893), but the transition seems to occur more frequently about that time or the middle of the month. A female from Black Mountain, W. Va., is in summer pelage October 29, 1900, but three others collected between October 28 and November 1, the same year, are in full winter pelage. A female from Digby, Nova Scotia, in worn summer pelage had not begun to molt October 25, 1893; another female from the same place has the winter fur coming in under the summer fur over the entire upper parts and abdomen October 22. 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 63 SOREX FUMBUS FUMEUS Miller Smoky Shrew (Pls. 2, i; 4, w; 5, w; 7, d; 11, b; 12, b) Sorex f um ens Miller, North Amer. Fauna No. 10, p. 50. December 31, 1895. Sorex fumeus fumcus Jackson, Proc. Biol. Soc. Washington 30 : 149, July 27, 1917. Type specimen. — No. 7.7.7.2582, British Museum (No. 2582, collec- tion of Gerrit S. Miller, jr.); 2 adult, skin and skull; collected September 24, 1893, by Gerrit S. Miller, jr. Type locality. — Peterboro, Madison County, N. Y. Geographic range. — New Hampshire, Vermont, Connecticut, Rhode Island, and northern New York, south through northwestern New Jersey, Pennsylvania, and western Maryland, to south-central Ohio and northwestern Georgia; also recorded from Ontario (North Bay) and southeastern Wisconsin (Racine). (Fig. 4.) Diagnostic characters. — Characters given under the species Sorex fumeus will separate S. f. fumeus from all shrews except S. f. umbrosus. It averages slightly smaller than umbrosus, and is distinctly more reddish brown (less grayish) in summer pelage. Color. — Winter pelage: Distinctly grayish. Upper parts in general effect either mouse gray or deep mouse gray, occasionally almost dark mouse gray. Underparts a trifle paler than the upper parts, usually mouse gray, frequently silvery in certain lights. Tail indistinctly bicolor, fuscous above, chamois or honey yellow beneath nearly to tip ; feet chamois, the outer edge dusky. Summer pelage: Decidedly more brownish than winter pelage. Upper parts about olive-brown or slightly darker ; underparts somewhat paler than upper parts, usually drab or wood brown, showing more or less intermixture of deep neutral gray of base of hairs ; tail and feet as in winter. Skull. — Medium in size, relatively short and broad, with short rostrum and relatively short and broad interorbital region ; brain case moderately flattened ; infraorbital foramen large and placed well back ; dentition moderate, molari- form teeth rather deeply emarginate posteriorly ; third unicuspid larger than fourth. Skull decidedly larger and relatively broader than that of S. c. cinereus, with much heavier rostrum and dentition ; about equal in length to that of S. dispar but decidedly broader and heavier throughout, with wider rostrum and heavier dentition ; somewhat smaller than that of S. cinereus, flatter throughout, with less depth of rostrum and brain case, relatively wider and shorter interorbitally, and with smaller unicuspidate teeth. Not essen- tially different from that of S. f. umbrosus. Measurements. — Two adult males from type locality : Total length, 124, 120 ; tail vertebrae, 45, 45 ; hind foot, 13, 13.4. Adult female from Renova, Pa. : Total length, 111 ; tail vertebrae, 45; hind foot, 13. Average of 3 adult females from Cranberry Glades, Pocahontas County, W. Va. : Total length, 117.7 (115- 120); tail vertebrae, 47.7 (47-48); hind foot, 14.3 (14-15). Skull: Average of 3 skulls of adult females (teeth slightly worn) from type locality: CondylO- basal length, 18.1 (18.0-18.3) ; palatal length, 6.8 (6.8-6.9) ; cranial breadth, 8.9 (8.7-9.0) ; interorbital breadth, 3.8 (3.7-3.9) ; maxillary breadth, 5.3 (5.2- 5.4) ; maxillary tooth row, 6.2 (6.2-6.3). Skulls of two adult males (teeth slightly worn) from Travellers Repose, W. Va. : Condylobasal length, 18-6, 18.3 ; palatal length, 7.2, 7.1 ; cranial breadth, 8.8, 8.9 ; interorbital breadth, 4.0, 3.8; maxillary breadth, 5.3, 5.4; maxillary tooth row, 6.6, 6.6. Average of 4 skulls of adult females (teeth slightly worn) from Roan Mountain, N. C. : Condylohasal length, 18.8 (18.0-18.6) ; palatal length, 7.0 (6.8-7.2) ; cranial breadth, 8.9 (8.7-9.0) ; interorbital breadth, 3.9 (3.8-4.0) ; maxillary breadth, 5.2 (5.2-5.3) ; maxillary tooth row, 6.4 (6.2-6.5). Remarks. — Two specimens of S. f. fumeus were in the United States National Museum as long ago as 1855. Both of these have the brain cases broken away and absent, but both have the rostra and teeth in 64 NOKTH AMERICAN" FAUNA [No. 51 fairly good condition. One of them from Carlisle, Pa., was misi- dentified by Baird (1857, p. 22) with S. forsteri Richardson, a syno- nym of S. c. cinereus; the other, from Racine, Wis., he identified as iS. richardsoni (Baird, 1857, p. 24). The Wisconsin specimen, here referred to the subspecies fumeus, is the only one known from that region ; it is in worn summer pelage, very faded, and, with its broken skull, is unsuitable for critical subspecific comparison. Still later, Dobson (1890, pi. 23, fig. 5) misidentified a specimen from Lake George, N. Y., as S. platyrhinus (DeKay), a synonym of S. c. cinereus. The majority of specimens of S. f. fumeus now in collections have come from the Appalachian Mountains, where the form seems to be not uncommon in certain localities. In color, the subspecies is con- stant throughout most of its range, but there is a pronounced varia- tion in the skulls, particularly in the degree of flatness of the brain case. In some cases this appears to be local ; in others all degrees of flatness of the cranium are present at a given locality ; nowhere can a definite geographic range be assigned to these variations. The skulls from Travellers Repose, W. Va., and Renovo, Pa., have as shallow and flat brain cases as any, but specimens from intervening localities in Maryland, from north in New York, and south in North Carolina have skulls with high brain cases. Moreover, some of the skulls from as near Travellers Repose as Black Mountain, and Cranberry Glades, Pocahontas County, W. Va., have high brain cases; and in the large series from Roan Mountain, N. C, are found skulls both with comparatively high brain cases and low ones, and also inter- mediates with varying degrees of depth. In the extreme northeastern part of its range, S. f. fumeus ap- proaches S. f. umbrosus in color. A majority of the specimens from Vermont, New Hampshire, and western Massachusetts shows this tendency, and even one from Tupper Lake in the Adirondacks, N. Y., and another from the Catskills in the same State, are almost as gray as typical umbrosus. Specimens from Peterboro, N. Y., the type locality of the subspecies fumeus, however, are as reddish-brown as those from the southern part of the range of the subspecies. The writer has been unable to examine a specimen, now in the British Museum, recorded from North Bay, Ontario (Miller, 1897, p. 35) ; the locality has been provisionally included on the map (p. 61) in the range of S. f. fumeus. Specimens examined. — Total number, 161, as follows : Connecticut: Monroe, 1. Georgia: Brasstown Bald (altitude 4,700 feet), 1. Maryland: Bittinger, 3; Finzel, 6. Massachusetts: Mount Greylock, 6.71 New Hampshire: Antrim, 1 72 ; Dublin, l72; Intervale, 17S; Mossy Brook (Mount Monadnock), 1; Ossipee, 4; Waterville, 1; Webster, l.72 New Jersey: Culvers Gap, Sussex County, l74; Delaware Water Gap, l74; Greenwood Lake (south end, Passaic County), l.74 New York: Berlin (altitude 1,100 feet), 18; East Greenwich, 1; Hunter Mountain (CatsTiill Mountains), 3; Lake George, 6; Peterboro (type locality), 6; Piseco, 1; Tupper Lake, l.72 North Carolina: Roan Mountain, 5; Roan Mountain (altitude 6,000 feet). 12; Roan Mountain (altitude 8,300 feet), 2; Roan Mountain (Magnetic City), 1. 71 Manton Copeland coll., Brunswick, Me. 73 Univ. Mich. Mus. 72 Mus. Comp. Zool. 7* Acad. Nat. Sci., Philadelphia. 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 65 Ohio: Alma, 1"; Carrollton, 2"; Hopetown, 1"; Overton, 1™ Pennsylvania: Bushkill Creek (7 miles east of Cresco, Monroe County), 1"; Carlisle, 1; Chester County, 1; Eagles Mere, 3 74 ; Fleming, 1; Ganoga Glen, Sullivan County, 374; Ganoga Lake (North Mountain), 274; Krings Station, l74; Lake Leigh (North Mountain), 2 74 ; Mount Pocono, Monroe County, 274; Renovo, 4; Round Island, 374; Summit Mills, 274; Sayre, 2. Rhode Island: Chepachet, 1. Tennessee: High Cliff, 1; Roan Mountain (top), Carter County, 2.74 Vermont: Mount Mansfield, 3; Rutland, I75; Woodstock, 2.71. Virginia: Mount Rogers (altitude 5.719 feet), Grayson County, 2; Paris, 1; Washington (Devils Stairs, altitude 2,000 feet), 1. West Virginia: Black Mountain, 478; Cranberry Glades (head Cranberry River) , Pocahontas County, 10 ; Franklin, 1 ; Rowlesburg, 1 ; Travellers Repose, 5; White Sulphur, 2"; White Sulphur Springs, l71; Winding Gulf, 3. Wisconsin: Racine, 1. SOREX FUMEUS UMBROSUS Jackson Nova Scotian Smoky Shrew 8[orex] fumens (sic) Cox, Canadian Record Sci. 7:118, 1896. (Nomen nudum, misprint for S. fumeus.) Sorex fumeus umbrosus Jackson, Proc. Biol. Soc. Washington 30: 149, July 27, 1917. Type specimen. — No. 150065, U. S. Nat. Mus., Biological Survey collection; $ adult (teeth slightly worn), skin and skull; collected July 29, 1907, by W. H. Osgood. Original number 3140. Type locality. — James River, Antigonish County, Nova Scotia. Geographic range. — Nova Scotia, New Brunswick, southeastern Quebec, and Maine. (Fig. 4.) Diagnostic characters. — Similar to Sorex f. fumeus but averaging slightly larger, and in summer pelage distinctly less reddish brown (more grayish brown) on upper parts. Color. — Winter pelage: Similar to that of S. f. fumeus. Summer pelage: Brown of the upper parts noticeably less reddish than in the subspecies fumeus. Upper parts fuscous-black mixed with grayish ; underparts drab mixed with deep neutral gray of base of hairs; tail bicolor, fuscous-black above, honey yellow, cinnamon-buff, or chamois below nearly to tip ; feet chamois, the outer side dusky. Skull. — Similar to that of flf. f. fumeus, possibly averaging slightly larger. Measurements. — Type specimen (adult male) : Total length, 127; tail verte- brae, 52 ; hind foot, 14. Two adult males from type locality: Total length, 127, 126; tail vertebrae, 49, 45; hind foot. 14. 14.5. Skull: Type specimen (adult male; teeth slightly worn): Condylobasal length, 19.0 "; palatal length, 7.2; cranial breadth, 9.3 ; interorbital breadth, 3.9 ; maxillary breadth. 5.3 ; maxillary tooth row, 6.8. Skulls of two adult females (teeth very slightly worn) from type locality: Condylobasal length, 17.8, 18.5; palatal length. 7.2, 7.2; cranial breadth. 8.5, 9.1; interorbital breadth, 3.9, 3.9; maxillary breadth, 5.0, 5.1; maxillary tooth row, 6.5, 6.6. Remarks. — Although not a strikingly differentiated form, speci- mens of S. f. umbrosus from Nova Scotia are in series readily sep- arable from specimens of S. f. fumeus from central New York and the southern Appalachian Mountains. Intergradation between the 71 Mantim Copeland coll., Brunswick, Me. 78 Univ. Mlcb. Mus. ■"Acad. Nat Sci. Philadelphia. 75 Amer. Mus. \:it. Hist. 79 Field Mus. Nat. Hist. 77 Field Mus. Nat. Hist., 1. 78 In the original description of this subspecies the condylobasal length was misprinted as 24.0 (Jackson, 1917, p. 150). 66 NORTH AMERICAN FAUNA [No. 51 two forms, however, is clearly indicated over the greater part of the New England States and western New Brunswick. Specimens from Maine are referable to umbrosus; while specimens from New Hamp- shire, Vermont, and western Massachusetts (Mount Greylock) can be referred to S. f. fumeus, although displaying an approach toward umbrosus. The small series from Hampton, southeastern New Bruns- wick, is typical of umbrosus in every respect, yet a specimen from Restigouche River, northwestern New Brunswick, shows a strong tendency toward the subspecies fumeus. Specimens examined. — Total number, 62, as follows : Maine: Brunswick, 479; King and Bartlett Lake, l80; Mud Pond, Penobscot County, 2"; North Belgrade, l79; Topsham, 3.** New Brunswick: Bathurst (15 miles from, Miramichi Road), l83; Hamp- ton, 3 ; Maugerville, 2 S4 ; Restigouche River, l.85 Nova Scotia: Barrington Passage, j783; Digby, 13 88; Halifax, 3"; James River (type locality), 6; Little River, Digby Neck, 2; Newport, l.81 Quebec: Gasp6 Peninsula (Cascapedia River, Tracadie), l81; Riviere du Loup, l.85 Table 2. — Cranial measurements of adult specimens of Sorex fumeus group Species and locality 6 "c3 03 O o O be a a "a Is Ph C3 © Ut .a "3 a 03 o 00 is £ « a o £| *g 03 Wear of teeth Remarks S. f. fumeus: New York: Peter- boro. Do 111122 111123 140945 87025 87026 47823 47825 47826 55818 150065 150061 150064 9 9 9 & ; ear, length, 6^ ; elbow to end of middle digit, without claw, 13 ; manus, 6 ; pes, 13 ; distance between tips of first upper incisor and last premolar, ZV-i" (Dobson, 1889, p. 373.) Remarks. — Dobson remarks in regard to his S. hydrodromus: This species is evidently aquatic, like Crossopas foederis, the fringe of the manus and pes being even better developed than in that species ; but in all generic characters it agrees with those of the genus Sorex. While agreeing with Sorex palustris from the adjoining continent of America in external char- acters, it differs from it in the proportions of its teeth, resembling in this respect the section of which S. vulgaris is typical, while S. palustris agrees with those represented by S. vagrans. (Dobson, 1889, p. 374.) Unfortunately the writer has been unable to examine a specimen of this shrew. Merriam, who also never saw the species, placed it with acknowledged uncertainty in the subgenus Neosorex (Merriam, 1895, p. 94-95), probably being led to do so by Dobson's account of the "thick comb-like fringe of stiff hairs along %feSSr?i5 the outer and inner margins of both manus ??v£° <1I?2le'£ers'- o4ftcr and pes." Elliot (1901a, p. 379) and Miller and Rehn (1903, p. 119) recognized Neosorex as a genus and included therein the species hydrodromus, apparently following Merriam's lead. Except for the description of the fringed feet, there is nothing in the original account to associate closely S. hydrodromus with Neosorex, and Dobson clearly intimated the distinct difference between the two forms when he stated: No better proof could be afforded of the uselessness of retaining Neosorex as a distinct genus for the American species characterized by the possession of swimming-fringes in the digits, while the tail is simple, as in Sorex. These species are in fact aquatic forms of the genus Sorex. (Dobson, 1889, p. 374.) The original description, as quoted almost entire in the present account of the species, and the illustration of the upper teeth fit S. tundrensis almost precisely and make it necessary to place S. hydro- dromus in the arcticus group. The known geographic ranges of S. tundrensis and the subgenus Neosorex also support this contention. Until specimens from the type locality are available, the exact status of the form will remain in doubt. The description so closely fits S. tundrensis that the possibility of S. hydrodromus Dobson preoccupy- ing S. tundrensis Merriam" is strongly suggested, and at best jS. 74235—28 G 76 NORTH AMERICAN FAUNA [No. 51 tundrensis may prove only subspecifically distinct from S. hydro- dromus. Table 3. — Cranial measurements of adult specimens of Sorex arcticus group Species and locality d M 03 03 03 °^ •oja a o O ~8> a 09 IS CO "is ■3 03 .a "3 '8 03 o is _, 03 "J 3 J3 +j o o .52 'x 03 Wear of teeth Remarks S. a. arcticus: Alberta— South Ed- monton. Do 69150 69161 69163 69171 69151 69158 69159 69173 73180 73181 186837 186843 186845 186849 186852 186855 186856 38820 99286 99276 99277 99279 131000 131006 131011 131018 131032 131034 & & d" & 9 9 9 9 9 9 0" J1 d1 d1 d1 d1 d1 d1 9 9 9 d1 & & d" d1 d1 18.5 18.5 18.7 18.6 19.1 19.0 18.7 19.1 18.9 18.6 19.0 19.0 18.7 18.6 19.1 19.1 18.8 18.9 18.3 18.0 18.4 18.1 18.0 18.3 18.5 18.0 17.9 17.8 7.2 7.2 7.2 7.2 7.5 7.3 7.4 7.3 7.4 7.2 7.6 7.7 7.5 7.6 7.8 7.8 7.6 7.7 7.1 7.0 7.2 7.1 7.1 7.0 7.2 7.1 6.9 6.7 9.4 9.3 9.3 9.2 9.3 9.4 9.4 9.2 9.2 9.3 9.3 9.4 9.5 9.1 9.2 9.4 9.3 9.3 9.1 9.2 9.1 9.0 9.1 9.1 9.3 9.2 9.0 9.0 3.5 3.4 3.6 3.5 3.7 3.7 3.7 3.7 3.5 3.5 3.8 3.9 3.7 3.8 3.8 3.9 3.7 3.8 3.7 3.5 3.6 3.7 3.9 3.7 3.8 3.7 3.8 3.8 5.0 5.1 5.2 5.1 5.2 5.0 5.0 5.1 5.1 5.0 5.3 5.3 5.1 5.3 5.2 5.3 5.2 5.3 4.9 4.8 4.8 5.0 5.1 5.1 5.1 4.8 5.0 5.0 6.5 6.6 6.6 6.5 6.8 6.5 6.6 6.7 6.6 6.4 6.8 6.8 6.6 6.6 6.9 6.8 6.7 6.7 6.3 6.2 6.4 6.2 6.2 6.3 6.4 6.1 6.3 6.2 Slight ...do ...do. ...do ...do. ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do ...do.. ...do. ...do ...do ...do. ...do ...do ...do ...do ...do Do Do - Do Do .. Do . Do Saskatchewan — Win- gard. Do S. a. laricorum: Minnesota— E Ik River. Do Type specimen. Type locality. Do Do. Do .. Do. Do .. Do. Do Do. Do Do. North Dakota— Pem- bina. S. tundrensis: Alaska— St. Michael. Do Type specimen of S. pachyurus Baird. Type specimen. Type locality. Do Do. Do Do. Alaska— Mountains near Eagle. Do Do Do Do Do SOREX PRIBILOFBNSIS GROUP The pribilofensis group includes a single species: Sorex pribUo- fensis. All necessary group comparisons are made under the species. SOREX PRIBILOFENSIS Merriam Pribilof Shrew (Pls. 2, m ; 4, z ; 5, a' ; 7, h ; 11, d ; 12, e) Sorex pribilofensis Merriam, North Amer. Fauna No. 10, p. 87, December 31, 1895. Type specimen. — No. 30911, U. S. Nat. Mus., Biological Survey collection; $ adult, alcoholic with skull not removed; collected July 29, 1891, by C. Hart Merriam. Type locality. — St. Paul Island, Pribilof Group (in Bering Sea), Alaska. Geographic range. — Known only from type locality. Diagnostic characters. — Size small; tail rather short, hairy; color pattern somewhat like 8. tundrensis, in summer pelage tricolor, in winter distinctly tri- color. Skull short and broad, interorbital region broad, mesopterygoid space 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 77 short, rostrum heavy; dentition moderately heavy, the third unicuspid larger than the fourth. Compared with any of the cinereus group, 8. pribilofensis has a relatively shorter tail, a more tricolor pattern in summer pelage, and in winter pelage has the dark color of the upper parts much more reduced to a longitudinal ribbon ; the skull of £. pribilofensis is relatively much broader and shorter than that of S. cinereus, distinctly broader interorbitally, with a decid- edly heavier rostrum, broader palate, and heavier dentition ; molars relatively broader (extero-interiorly) than in 8. cinereus, the unicuspids larger, with internal ridge from apex to edge of cingulum as in the species cinereus (but more heavily pigmented than in S. c. cinereus). Similar in external propor- tions and general color pattern to members of the arcticus group but smaller, and paler in winter pelage ; skull shorter and broader interorbitally than that of any members of the arcticus group, with flatter, more rotund (less.angular) brain case, shorter, less attenuate rostrum, and with internal pigmented ridge on unicuspid extending from apex to cingulum (in arcticus group this ridge is short and on the internal part of the cingulum only and is scarcely pigmented). Color. — Winter pelage: Distinctly bicolor, the color of the underparts en- croaching upon the back and meeting in a sharp line. Top of nose, crown, ears, nape, back, and rump to base of tail a sharply defined longitudinal ribbon of drabbish hair brown; entire underparts, lips, cheeks, sides of head below eyes and ears, flanks, and sides of body we!! up toward back pale olive gray, usually very faintly tinged with pale olive buff. Tail bicolor ; drab above, pale pinkish buff beneath nearly to tip. Summer pelage: Tricolor, the back sharply defined from sides, which are less clearly defined from the ventral parts. Top of head and back to base of tail a distinct longitudinal strip of brownish (be- tween hair brown and olive brown) ; sides of head and body paler, drab to between wood brown and avellaneous ; underparts smoke may. sometimes slightly tinged or stained with olive buff ; tail as in winter, less hairy. Time of molting. — The evidence at hand seems to show that the transition from winter to summer pelage usually occurs in June or early in July. Out of 24 specimens collected in June, July, and early in August, all except 7 are in complete summer pelage. A female collected June 20, 1896, still retains the entire worn winter pelage. A male, June 22, 1890, has acquired about half of the summer fur, and 2 others, collected July 5, 1914, are in about the same status. A female, August 14, 1895, has the last remnants of the winter fur, while 2 males, collected one day earlier, are approximately one-half molted. The fall change occurs about the middle of October. Eleven of twenty-five specimens collected between October 22 and 26, 1914. are in complete winter pelage ; 11 are in process of molting; 3 are still in summer pelage. Skull. — Short and broad, particularly interorbitally; rostrum moderately .short, heavy; mesopterygoid space short; brain case moderately flattened, not angular; dentition moderately heavy; molariform teeth about as broad (extero- internal diameter) as long (antero-posterior diameter); cusplike processes of basal shelves of molariform teeth reduced; interior edge of basal shelf of sec- ond upper premolar without distinct cusplike process; unicuspid teeth rela- tively large and broad (extero-interior diameter), decreasing gradually in size posteriorly, the third larger than the fourth, the fifth relatively large, but much smaller than the fourth; unicuspid teeth each with internal heavily pigmented ridge from apex to edge of cingulum. Measurements. — Average of 8 adult males from type locality: Total length, 96.2 (92-103); tail vertebrae, 34.7 (32-37); hind foot, 13.8 (13-14.5). Skull: Average of 10 skulls of adult males (teeth slightly worn) from type locality: Condylobasal length. 15.8 (15.4-16.0); palatal length, 5.7 (5.6-6.8); cranial breadth, 7.S (7.7-8.0) ; interorbital breadth. 3.7 (3.6-3.8) ; maxillary breadth, 4.8 (4.7-4.9) ; maxillary tooth row, 5.5 (5.4-5.6). Remarks. — The interesting little Pribilof shrew as far as known is confined to St. Paul Island, in Bering Sea, where it is not uncommon. It is strikingly different from any other American Sorex, showing some affinity toward the cinereus group in the possession of pig- mented ridges on the internal sides of the unicuspid teeth, but, on the whole, it is probably more closely related to the arcticus group. Specimens examined. — One hundred and twenty-one" from the type locality. 8 Mus. Comp. Zool., 1. 78 NORTH AMERICAN" FAUNA [No. 51 Table 4. — Cranial measurements of adult specimens of Sorex pribilofensis group "ca ja a 03 hi o Species and locality d H s> m ■a S3 — be ^S a o O SO a 03 13 Ph 03 "3 '3 03 o 23 a>,Q a *3 3 o 1 Wear of teeth Remarks S. pribilofensis: Alaska— St. Paul 206181 rt1 16.0 5.7 7.9 3.7 4.8 5.5 Slight Type locality. Island. Do 206182 206183 63232 63233 63234 63235 217965 217969 217975 & c? 15.8 15.9 15.8 15.9 16.0 15.5 15.9 16.0 15.4 5.7 5.7 5.7 5. 7 5.7 5.6 5.7 5.8 5.7 7.9 7.9 7.8 7.9 8.0 7.9 7.7 7.7 7.7 3.7 3.6 3.7 3.7 3.8 3.6 3.6 3.7 3.8 4.8 4.8 4.8 4.8 4.9 4.7 4.7 4.8 4.9 5.4 5.4 5.5 5.4 5.6 5.4 5.5 5.6 5.4 ...do ...do ...do.. ...do_. _..do. ...do ...do ...do. ...do Do. Do..*. Do. Do Do. Do Do. Do Do. Do Do. Do Do. Do Do. Do Do. SOREX MERRIAMI GROUP The merriami group includes two species : Sorex merriami and S. leucogenys. Geographic range. — Known only from arid regions of southeastern Montana, southwestern North Dakota, southeastern Washington^ north-central Oregon, Nevada, and southwestern Utah. (Fig. 7.) Diagnostic characters. — In size somewhat larger than S. cinereus, pale (grayish drab above), with distinctly whitish underparts and feet. Skull relatively short and broad, flattened through the brain case, relatively high and swollen interorbitally, with a short, broad rostrum, which compared with that of other members of the genus found within its geographic range, is abruptly truncate anteriorly (nares region). The third upper unicuspidate tooth of most of the west American shrews is smaller than the fourth. Excep- tions to this are found in S. cinereus and S. arcticus, both species that may possibly occur within certain parts of the geographic range of shrews of the merriami group, and, like them, have the third upper unicuspid larger than or, infrequently in 8. cinereus equal to the fourth. Members of the merriami group, however, have the unicuspids relatively narrow and elongate (in lateral aspect), and tending to be more crowded together than in <8. cinereus or S. arcticus. Remarks. — Although the number of specimens available that repre- sent the merriami group is small and insufficient for satisfactory evi- dence as to distribution and variation, nevertheless it shows that the species merriami and leucogenys are very closely related forms dif- fering markedly from any other shrews. The members of this group appear to inhabit arid or semidesert regions, much in contrast with the habitat preference of most species of long-tailed shrews. SOREX MERRIAMI Dobson Merkiam Shrew (Pls. 2, n ; 4, a' ; 6, a; 7, i ; 11, e; 12, f) So-rex merriami Dobson, Monograph of the Insectivora, systematic and anatomi- cal, part 3, fasc. 1, pi. 23, fig. 6, May, 1890. Type spechnen. — No. 186441, U. S. Nat. Mus., Merriam collection, (old No. \%%\ ) ; 2 adult (teeth slightly worn), alcoholic with skull removed ; collected December 26, 1884, by Charles E. Bendire. Type locality. — fort Custer, Bighorn County, Mont. 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 79 Geographic range. — Known only from five localities in south- western North Dakota, southeastern Montana, northern Nevada, north-central Oregon, and southeastern Washington. (Fig. 7.) Diagnostic characters. — Size small, larger than Sorex cinereus, smaller than 8. arcticus; color pale, underparts and feet distinctly whitish ; tail medium in length, decidedly bicolor, whitish below to tip. Skull relatively short and broad, flattened through the brain case, high and swollen orbitally, with short, broad rostrum. Dentition heavy ; unicuspid row relatively short, the fourth Fig. 7. — Geographic range of species of Sorex merriami group 1. 8. merriami. 2. S. Icucogenys. unicuspid smaller than third. Skull somewhat smaller than that of 8. leucogcnys, with flatter brain case and anterior halves of unicuspid tooth rows more approximated and more nearly parallel. Color. — Winter pelage: 7 Upper parts drab, becoming paler on the flanks; underparts, chin, lips, and feet, distinctly whitish, the feet tinged with light buff; tail bicolor, wood brown above, whitish beneath to tip. Summer pelage:* Upper parts grayish drab becoming light drab on the flanks ; underparts nearly white, very faintly tinged with pale olive-buff ; tail and feet as in winter. 7 Based on alcoholic specimen from Elko County, Nev. 8 Based on imperfect skin from Mcdora, N. 1 'ak. 80 NORTH AMERICAN FAUNA [No. 51 Skull. — Short and broad, flattened through brain case, high and swollen orbitally ; rostrum short and broad, particularly broad through region of infra- orbital foramina, which open relatively well forward ; mesopterygoid space short, the sides relatively heavy. Teeth large, the molars as broad (extero- intemally) as long (antero-posteriorly), deeply emarginate posteriorly; teeth densely pigmented ; unicuspid tooth row short, the fourth unicuspid smaller than the third ; unicuspids placed relatively vertical to antero-posterior axis of alveolar borders ; antero-posterior diameter of each unicuspid considerably less than supero-inferior diameter; unicuspids without heavily pigmented internal ridge from apex to edge of cingulum ; first upper incisors small. Measurements. — Type specimen (adult female) (measured from alcoholic specimen after removal of skull) : Total length, 90; tail vertebrae, 35; hind foot, 11.5. Skull. — Type specimen (adult female; teeth slightly worn) : Con- dylobasal length, 15.8 ; palatal length, 6.3 ; cranial breadth, 8.0 ; interorbital breadth, 4.0 ; maxillary breadth, 5.0 ; maxillary tooth row, 5.6. Skull of adult (teeth slightly worn), sex unknown, from Golconda (100 miles northeast of), Nevada : Condylobasal length, 16.2 ; palatal length, 6.4 ; cranial breadth 7.9 ; interorbital breadth, 3.8 ; maxillary breadth, 5.1 ; maxillary tooth row, 5.8. Remarks. — Regarding the type specimen of S. merriami, Merriam has written : * The type and only known specimen of this remarkable shrew was presented to me by Maj. Charles E. Bendire, who collected it at the post garden, on the Little Big Horn River, about a mile and a half above Fort Custer, Mont., December 26, 1884. I sent it, with all my other shrews to Dr. George E. Dobson, who was then engaged on a monographic revision of the Soricidae. Unfortu- nately, owing to Dr. Dobson's continued ill health, all that has ever been pub- lished of this monograph is a fasciculus of plates, showing the jaws and teeth of certain species, with a page of explanation facing each plate (Monog. In- sectivora, Part III, fasc 1, May, 1890). The present species is named and its peculiar dentition shown in PI. XXIII, fig. 6, of this work. But the remarkable shape of the palate and peculiarities of the skull as a whole are not shown. The skull was removed from the alcoholic specimen by Dr. Dobson, and I have sometimes wondered whether by any possible accident it could have been trans- posed with that of some Asiatic species, it is so very unlike all known American shrews. When the specimen was returned the alcoholic bore my original label and number (1001), but the skull was numbered differently (1SS6; its proper number is 4861). Dr. Dobson afterwards wrote me that his number was an error, and that the skull belonged to my alcoholic No. 1001. [Merriam, 1S95, p. 88-89.] Since the time Merriam published the foregoing remarks, four additional specimens of this rare shrew have come to light. Unfor- tunately, none of these is a perfect specimen. On June 23, 1896, Vernon Bailey found a dead shrew in a creek valley, 7 miles south- east of Antelope, Oreg. From this specimen, a mere fragment of skin and body that has been in alcohol, the partly crushed skull has been removed. It agrees well with that of the type of S. mer- riami, except that it seems a trifle higher through the brain case. The remains of a small shrew were found among the rocks on a high butte near Medora, N. Dak., on June 13, 1913, by S. G. Jewett. Some animal had killed the shrew and eaten its head, so that only the skin of the hind half of the body, the hind feet, and tail are available for study. The color of the animal and the habitat where it was found indicate that the specimen is with little doubt iS. merriami. The third specimen was obtained by Edmund Heller, November 26, 1914, at Desert Ranch, Elko County, Nev., where it had been caught by a house cat. The skull of this specimen is slightly larger and with somewhat higher brain case than that of the type of S. merriami, but it is decidedly more nearly like this form than JS. leucogenys. 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 81 The last specimen to make its appearance, a skin accompanied by a broken skull, was collected by George G. Cantwell, November 18, 1919, at the entrance to an old badger digging on top of a " high bunch grass hill," at Starbuck (altitude 645 feet), Columbia County, Wash. It shows no appreciable differences from the type specimen of /S. merriami. Specimens examined. — Total number, 5, as follows: Montana: Fort Custer (type locality), 1. Nevada: Desert Ranch, 100 miles northeast of Golconda, Elko County, 1. North Dakota: Medora, 1. Oregon: Antelope (7 miles southeast), 1. Washington: Starbuck (altitude 654 feet), 1. SOREX LEUCOGENYS Osgood White-cheeked Shrew (Pls. 2, o; 4, b'; 6, b; 7, J) Sorex leucogenys Osgood, Proc. Biol. Soc. Washington 22 : 52, April 17, 1909. Type specimen. — No. 157952, U. S. Nat. Mus., Biological Survey collection; 9 adult, skin and skull; collected August 12, 1908, by W. H. Osgood. Type locality. — Mouth of the canyon of Beaver River, about 3 miles east of Beaver, Beaver County, Utah. Geographic range. — Known onty from type locality and Esmeralda County, Nev. (Fig. 7.) Diagnostic characters. — Essentially like Sorex merriami in color, but slightly larger ; skull larger than that of S. merriami, higher through the brain case, and with anterior halves of unicuspid tooth rows less approximated and less nearly parallel. Color. — Winter pelage: Unknown. Summer pelage: Similar to 8. merriami. Upper parts pale hair brown or grayish drab becoming light drab on the flanks; underparts nearly white, faintly tinged with pale olive-buff; chin, lips, and sides of face below eyes pale olive-buff; feet whitish, tinged with light buff; tail distinctly bicolor, drab above, whitish below, tipped with whitish. Skull. — Larger tban that of S. merriami, with relatively and actually broader and higher brain case, which rises more abruptly in frontal region, the uni- cuspid tooth rows tending to approach each other at the anterior ends with more regularity (in straight line) and with anterior halves less approximated and more diverging posteriorly. Measurement*. — Type specimen (adult female) : Total length, 107; tail verte- brae, 38 : hind foot. 12.5. Adu't female from Mount Magrnder, Nev. : Total length. 105; tail vertebrae, 40; hind foot, 12. Skull: Type specimen (adult female; teeth slightly worn) : Oondylobasal length. 16.5; palatal length, 6.5; cranial breadth, 8.3: Interorbitai breadth, 3.8; maxillary breadth, 5.1; maxillary tooth row, 5.8. Skull of adult female (teeth slightly worn) from Indian Spring, Mount Ma- gruder. Nev.: Condylobasal length, 16.9; palatal length, 6.7; cranial breadth, 8.4; interorbitai breadth, 4.0; maxillary breadth, 5.4; maxillary tooth row, 6.0. Remarks. — The type specimen of S. leucogenys was trapped on a dry Upper Sonoran slope about 200 yards from running water. The Indian Spring specimen is paler than the type specimen, which may be due chiefly to the differences in the make-up of the skins, the former being stuffed considerably fuller than the latter. Its skull has the high brain case and other characters of the type specimen of 8. h'nco- genys and is even slightly accentuated in size. The White Mountains specimen differs from the type only in slightly heavier molariform teeth and in a somewhat less reddish coloration, which is undoubtedly 82 NORTH AMERICAN FAUNA tNo.51 due to seasonal variation. The external measurements as taken by the collector are less than of the type of 0. leucogenys, but the skull does not show a corresponding differentiation ; if anything it is larger. Although the skull of S. merriami from northern Nevada is slightly larger than the type skull of that species and has a trifle higher brain case, the approach toward S. leucogenys does not clearly indicate intergradation between the two forms. Specimens examined. — Three as follows: Nevada: Chiatovich Creek, altitude 8,200 feet, White Mountains, Esmeralda County, 1 ' ; Indian Spring, altitude 7,700 feet, Mount Magruder, Esmeralda County, l.M "Utah: Beaver (3 miles east of), Beaver County (type locality), 1. Table 5. — Cranial measurements of adult specimens of Sorex merriami group Species and locality c M CO "c3 03 a o o 43 ■a a "3 "3 .a ■S 03 g "3 '3 03 u o 03 Si a t— 1 hi o o 03 O "m 03 Wear of teeth Remarks S. merriami: M o n t a n a— F o r t Custer. Nevada — 100 miles northeast of Gol- conda. S. leucogenys: Utah — Beaver River . N e v a d a— M o u n t Magruder. 186441 210121 157952 1K572 9 9 9 15.8 16.2 16.5 16.9 6.3 6.4 6.5 6.7 8.0 7.9 8.3 8.4 4.0 3.8 3.8 4.0 5.0 5.1 5.1 5.4 5.6 5.8 5.8 6.0 Slight ...do ...do ...do.. Type specimen. Type specimen. 1 Collection of Donald R. Dickey, Pasadena, Calif. SOREX SCLATERI GROUP The sclateri group includes a single species — Sorex sclateri. All necessary group comparisons are made under the species. SOREX SCLATERI Merriam Sclater Shrew (Pls. 2, p; 4, c'; 6, c; 8, a) Sorex sclateri Merriam, Proc. Biol. Soc. Washington 11 : 288, July 15, 1897. Type specimen. — No. 75872, U. S. Nat. Mus., Biological Survey col- lection; ? adult (teeth slightly worn), skin and skull; collected Oc- tober 23, 1895, by E. W. Nelson and E. A. Goldman. Type locality. — Tumbala, altitude 5,000 feet, Chiapas, Mexico. Geographic range. — Known only from the type locality. Diagnostic characters. — Size relatively large; tail long, hind foot large; color dark, both dorsally and ventrally; skull large, relatively long and narrow, interorbital region rather elongate, interpterygoid space long, dentition moder- ately heavy, weakly pigmented, the third unicuspid larger than the fourth. Similar in external appearance to S. veraepacis, possibly darker ventrally ; skull decidedly narrower than of 8. veraepacis, the brain case less angular and more flattened, the interorbital region more elongate, and the relative size of the third and fourth unicuspids reversed. 8 Mus. Vert. Zool. 10 D. R. Dickey coll., Pasadena, Calif. 1928] EEVIEW OF AMERICAN LONG-TAILED SHREWS 83 Color. — Winter pelage: Upper parts dark clove brown or dark bister, or almost blackish clove brown ; underparts scarcely paler than upper parts, clove brown or between clove brown and bister ; tail clove brown above, slightly paler below. Summer pelage: Unknown. Time of molting. — A male collected October 22, 1895, has acquired the new fur except on the occiput, nape, and rump, which are in process of molting. The three other specimens, collected October 23, 24, and 25, appear to be in fresh pelage. Skull. — Large, relatively long and narrow; the interorbital and post-palatal regions noticeably relatively elongate; brain case narrow, flattened, gently rotund laterally (not angular) in superior aspect; dentition moderately heavy, weakiy pigmented, tbe third unicuspidate tooth slightly larger than the fourth. Measurements. — Type specimen (adult female) : Total length, 126; tail ver- tebrae, 52; hind foot, 16. Adult male from type locality: Total length, 125; tail vertebrae, 52; hind foot, 16. Skull: Type specimen (adult female; teeth slightly worn) : Condylobasal length, 19.9; palatal length, 8.0; cranial breadth, 9.2; interorbital breadth, 4.4; maxillary breadth, 5.9; maxillary tooth row, 7.4. Skull of adult male (teeth slightly worn) from type locality: Condylobasal length, 19.6; palatal length, 7.7; cranial breadth, 9.2; interorbital breadth, 4.4; maxillary breadth, 5.6+ ; maxillary tooth row, 7.2. Remarks. — Although S. sclateri is very similar to S. veraepacis in superficial external appearance, averaging scarcely a shade darker and more brownish in color, it shows pronounced differences in cranial characters, and the two species are apparently not closely related. Only four specimens of the Sclater shrew are available for study, but in these four the characters diagnostic of the species are very uniform. Specimens examined. — Four, from the type locality. SOREX LONGIROSTRIS GROUP The longirostris group includes a single species. — Sarex longiros- tris. All necessary group comparisons are made under the species. SOREX LONGIROSTRIS Bachman [Synonymy under subspecies] Geographic range. — Atlantic Plain and Piedmont region from northern Virginia and southern Maryland south to northern Florida (Alachua County) and central Alabama (Autauga County) ; eastern and southern Illinois and southwestern Indiana. (Fig. 8.) Diagnostic characters. — Small ; with short rostrum and crowded unicuspid tooth row ; first and second unicuspids about equal in size, the third and fourth decidedly smaller than first and second, the third somewhat smaller than the fourth ; fifth unicuspid very much smaller than fourth, almost minute ; teeth inextensively pigmented. Differs from the cinereus group in its relatively shorter, broader rostrum, shorter and more crowded unicuspid row, third upper unicuspid smaller than the fourth, and antero-posterior diameter of unicuspid teeth less than extero-interior (lateral) diameter; extero-interior diameter of molariform teeth relatively much greater than in those of the cinereus group and first incisors, both upper and lower, relatively smaller. Much smaller than any of the fumeus or arcticus groups, and with different skull proportions and dentition. Sometimes similar to certain forms of the ornatus and vagranx- obscurus groups. Usually more reddish or darker in color than any of the ornatus group, skull higher through the brain case and less depressed inter* orbitaliy, first incisors weaker, unicuspid tooth row and mesopterygoid space shorter. Differs from members of the vagrans-obscurus group in a more flat- tened brain case, shorter unicuspid tooth row with the individual teeth rela- tively less in antero-posterior diameter, more extensively pigmented dentition, and relatively and actually shorter mesopterygoid space. 84 NORTH AMERICAN FAUNA [No. 51 Subspecies and geographic variation. — The species longirostris includes the two subspecies longirostris and fisheri. Insufficient specimens are available to show clearly the variations of the species over its entire range, but the sub- species longirostris appears to be very constant in characters over its compara- tively wide distribution, grading rather abruptly into the localized larger fisheri with its larger and relatively narrower skull. Time of molting. — A male and a female of the subspecies longirostris taken April 16, 1907, at Reevesville, 111., show fresh summer pelage over the entire ventral parts and head, the upper parts being in worn winter pelage ; two males collected on April 16 and 18 of the same year at the same place are in worn winter pelage. The specimen of fisheri from Chapanoke, N. C, collected March Fig. 8. — Geographic range of subspecies of Sorex longirostris 1. £. I. longirostris. 2. 8. I. flslieri. 20, 1897, has the molt beginning on the midback. A male fisheri collected May 23, 1905, at the type locality shows the summer pelage coming in under the worn winter fur on the back, while another male taken at the same time and an unsexed individual collected June 5, 1895, at Dismal Swamp, have obtained the full summer pelage. None of the specimens of S. I. longirostris examined is in process of changing from the summer to the winter pelage. Specimens of this form collected November 27 and 28, 1906, at Olive Branch, 111., have apparently acquired the full winter coat. A specimen of 8. I. fisheri caught October 21, 1S95, has the new fur incoming under the old over most of the back, while the type speci- men, a male collected October 11, has the winter fur on the rump, the fur on the remainder of the back and the abdomen being in the process of renewal. Rematrks. — The species JS. longirostris is composed of only two forms, comprising a compact lot separable from all other groups by 1928] REVIEW OF AMERICAN" LONG-TAILED SHREWS 85 the combined characters above enumerated. The shrews of this spe- cies differ from all other Sorex of eastern America in having the third unicuspid considerably smaller than the fourth, a characteristic common to most species of Sorex of western America, and absent from those of the intervening plains region. For this reason, com- bined with the apparent scarcity of individuals and the uniformity of characters within the species, one might be inclined to believe that it is a relict group of the preglacial fauna occupying the faunally old portion of the southeastern United States. SOREX LONGIROSTRIS LONGIROSTRIS Bachman Bachman Shrew (Pls. 2, q; 5, a; 6, d; 8, b; 11, f; 12, g) Sorex longirostris Bachman, Journ. Acad. Nat. Sci. Philadelphia 7 : part 2, p. 370, 1837. O[tisorex] longirostris De Kay, Zool. New York, part 1, Mammalia, p. 23, 1S42. [Musar[aneus] {Croc[idura])] bachmani Pomel, Arch. Sci. Phys. et Nat. 9: 249, ' 1848. Sorex wagneri Fitzinger, Sitzungsber. Kaiserl. Akad. Wissensch., math.- natiirwissensch. Classe, Wien, bd. 57, abtheil 1, p. 512, 1868. Type specimen. — Not known now to exist. Collected by Alexander Hume. Type locality. — Hume Plantation, swamps of the Santee River [Cat Island, mouth of Santee River], S. C. Geographic range. — Atlantic Plain and Piedmont region (except vicinity of Dismal Swamp, Va., inhabited by S. I. fisheri) from northern Virginia and southern Maryland, south to northern Florida (Alachua County) and central Alabama (Autauga County) ; eastern and southern Illinois and southwestern Indiana. (Fig. 8.) Diagnostic characters. — Size small, with relatively short tail (about equal length of body without head) ; about the size of S. fontinulis, more reddish in color, with distinct cranial characters ; skull relatively short and broad, with flattened brain case and short rostrum; unicuspids short (antero-posterior diameter) and broad (extero-interior diameter), the third smaller than the fourth. Separated from all other American shrews except 8'. I. fisheri by group characters. Smaller than fisheri, with decidedly smaller and relatively broader skull with shorter rostrum. Color. — Winter pelage: Upper parts Prout's brown, or mummy brown, some- times almost fuscous, shading gradually on the sides into cinnamon brown ; underparts tinged strongly with smoke gray mixed with drab, and showing considerable deep neutral gray of base of hairs. Tail indistinctly bicolor, Prout's brown or mummy brown above, cinnamon brown or ochraceous tawny below. Summer pelage: Very slightly paler than in winter iielage. the under- parts, more drab or avellaneous. Tail as in winter. Skull. — Small, with short rostrum, and broad, flattened brain case. Differs from that of S. cinereus and S. fontinalis in group characters. Much smaller than that of S. I. fisheri and relatively broader with shorter rostrum. Measurements. — Adult male from Raleigh, N. C. : Total length, 87 ; tail ver- tebrae, 32; hind foot, 11.5. Adult male from Falls Church, Va.: Total length, 79; tail vertebrae, 33; hind foot, 11. Average of three adult males from Olive Branch, 111.: Total length, 83.7 (79-90); tail vertebrae, 28.3 (27-30); hind foot, 10.3 (10-10.5). Skull: Skull of adult (sex unknown; teeth slightly worn) from Butler, Ga. : Condylobasal length, 14.6; palatal length, 5.1; cranial breadth. 7.3 ; interorbital breadth, 3.2 ; maxillary breadth, 4.2 ; maxillary tooth row, 5.0. Skull of adult male (teeth slightly worn) from Raleigh, N. C. : Condylobasal length, 14.1 ; palatal length, 5.1 ; cranial breadth, 7.2 ; interor- bital breadth, 3.0; maxillary breadth, 4.2: maxillary tooth row, 5.0. Skull of adult male (teeth slightly worn) from Falls Church, Va. : Condylobasal length, 86 NORTH AMERICAN FAUNA [No. 51 14.4 ; palatal length, 5.2 ; cranial breadth, 7.4 ; interorbital breadth, 3.2 ; max- illary breadth, 4.3 ; maxillary tooth row, 5.1. Imperfect skulls of two adults (sex unknown; teeth moderately worn) from Bicknell, Ind. : Palatal length, 5.1, 5.2 ; interorbital breadth, 3.2, 3.0 ; maxillary breadth, 4.6, 4.1 ; maxillary tooth row, 4.8, 4.8. Remarks. — Apparently local in distribution, and either rare or difficult to trap, probably both, this little shrew is represented in collections by comparatively few specimens. The knowledge of its presence in a given locality has frequently been purely accidental. The first one brought to the attention of zoologists was found in a newly dug ditch in the Santee Marshes of South Carolina and upon it was based the original description. Bachman also describes an- other specimen, which was found in the gullet of a hooded merganser (Lophodytes cucullatus) (Bachman, 1837, p. 372), and a specimen in the United States Bureau of Biological Survey collection was taken from the stomach of a barred owl (Strix varia alleni) shot near Autaugaville, Ala. Another had evidently fallen over the cliffs to the shore of Chesapeake Bay, Md., where it was found dead by Marcus Ward Lyon, jr. Although the specimens examined come from widely separated localities, represented in most cases by a single individual, there is little variation in the color or- cranial characters. Indiana and Illi- nois specimens seem to average a trifle less reddish in color than specimens from the Atlantic States, and they may possibly have on the average shorter unicuspid rows and smaller molariform teeth, but the differences are slight and inconstant, and, on the basis of the specimens examined, not marked enough for subspecific separation. The skulls from Raleigh, N. C, seem to average very slightly higher through the brain case than those from Georgia, but the difference is nonessential. The skull from Chesapeake Beach, Md., is relatively somewhat narrower than typical specimens, which probably indicates a tendency toward S. I. fishe?^, though there is no approach in size. The specimen from Falls Church, Va., is almost identical with the Georgia ones, which are assumed to be typical. Unfortunately, efforts to procure specimens of Sorex from the type locality of S. I. longirostHs have proved futile. There seems little doubt, however, of the status of the form. The status of Amphisorex lesueurii Duvernoy, the description of which has been misidentified as that of S. longirostris, is discussed under S. c. cinereus (p. 43-44). /Specimens examined. — Total number, 24, as follows : Alabama: Bear Swamp, 4 miles northeast of Autaugaville, 1. District of Columbia: Washington, 1. Florida: Newnans Lake, near Gainesville, Alachua County, 1." Georgia: Butler, 1; Young Harris, 1. Illinois: Olive Branch, 312; Pistakee Lake, Henry County, l18; Reeves- ville, 4.12 Indiana: Bicknell, 3. Maryland: Chesapeake Beach, Calvert County, 3; Hall, Prince Georges County, 1. North Carolina: Raleigh, 3. Virginia: Falls Church, 1. 11 Harley B. Sherman coll., Gainesville, Fla. 12 Field Mus. Nat. Hist. 13 111. State Lab. Nat. Hist. 1928] REVIEW OF AMERICAN" LONG-TAILED SHREWS 87 SOREX LONGIROSTRIS FISHERI Merriam Fisher Shrew (Pls. 2, r; 5, b) Sorex fisheri Merriain, North Amer. Fauna No. 10, p. 86, December 31, 1895. Type specimen. — No. 75166, U. S. Nat. Mus., Biological Survey collection; $ adult (teeth very slightly worn), skin and skull; collected October 11, 1895, by A. K. Fisher. Type locality. — Lake Drummond, Dismal Swamp, Va. Geographic range. — Known only from Dismal Swamp, Va., and adjacent part (Chapanoke) of North Carolina. (Fig. 8.) Diagnostic characters. — Similar to Sorex I. longirostris but much larger with color usually duller above and more tinged with drab or wood brown on the underparts ; skull distinctly larger in all dimensions than that of the sub- species longirostris, relatively somewhat narrower. Color. — Winter pelage: Upper parts and sides fuscous; underparts moderately tinged with drab or wood brown, showing mixture of deep neutral gray of base of hairs. Tail bicolor, fuscous above, drab below nearly to tip. Summer pelage: Somewhat brighter tban winter pelage. Upper parts Prout's brown, mummy brown, or near fuscous, becoming slightly paler on the sides; under- parts drab or wood brown, rarely showing any trace of the deep neutral gray of base of hairs. Tail as in winter. Skull. — Much larger than that of S. I. longirostris, relatively narrower with longer rostrum. About the size of that of large individuals of S. c. cinereus, but easily distinguished by its flatness of brain case, wide rostrum, and other group characters. Measurements. — Type specimen (adult male): Total length, 108; tail verte- brae, 39 ; hind foot, 12. Two adult females from type locality : Total length, 102, 98; tail vertebrae, 40, 34; hind foot, 13, 11.5. Skull: Skulls of two adult males (teeth slightly worn), type specimen and topotype : Condylobasal length, 15.9, 15.9 ; palatal length, 5.6, 5.4 ; cranial breadth, 7.8, 7.7 ; interorbital breadth, 3.5, 3.5 ; maxillary breadth, 4.4, 4.5 ; maxillary tooth row, 5.4, 5.4. Skulls of two adult females (teeth slightly worn) from type locality: Condylobasal length, 16.4, 15.4; palatal length, 5.6, 5.5; cranial breadth, 8.2, 7.5; interorbital breadth, 3.6, 3.3 ; maxillary breadth, 4.8, 4.5 ; maxillary tooth row, 5.8, 5.3. Skull of adult male (teeth slightly worn) from Chapanoke, N. C. : Condylo- basal length, 15.8 ; palatal length, 5.5 ; cranial breadth, 7.3 ; interorbital breadth, 3.4 ; maxillary breadth, 4.4 ; maxillary tooth row, 5.6. Remarks. — The distribution of S. I. fisheri is evidently restricted to a small area in the Dismal Swamp region of southeastern Vir- ginia and northeastern North Carolina. There is considerable varia- tion in size among the skulls from the type locality, but the smallest skulls of fisheri are distinct^ larger than the largest of S. I. longi- rostris. The single specimen of fisheH available from North Caro- lina barely suggests an approach toward S. I. longirostris. Specimens examined. — Total number, 16, as follows: North Carolina: Chapanoke, Perquimans County, 1." Virginia: Lake Drummond, Dismal Swamp (type locality), 15. "Acad. Nat. Sci., Philadelphia. 88 NORTH AMERICAN FAUNA [No. 51 Table 6. — Cranial measurements of adult specimens of Sorex longirostris group 03 si Si •5 03 >> o Species and locality d a o D it a « "tfl "3 Pi 03 "3 '3 S-i o osi a i— i -6 o 03 Wear of teeth Remarks S. I. longirostris: Georgia- Butler 38425 159415 14. C 13.9 5.1 5.0 7.3 7.2 3.2 3.1 4.2 4.2 5.0 5.0 Slight ...do Young Harris North Carolina — Ral- 81972 d" 14.1 5.1 7.2 3.0 4.2 5.0 ...do eigh. V i r g i n i a— F alls 87190 C? 14.4 5.2 7.4 3.2 4.3 5.1 ...do Church. M ar yland— Chesa- 151738 & 14.3 5.1 7.0 3.2 4.2 5.0 ...do peake Beach. 168737 158834 5.1 5.2 3.2 3.0 4.6 4.1 4.8 4.8 Moderate . ...do Do S. 1. flsheri: V i r g i n i a— L a k e 75166 d" 15.9 5.6 7.8 3.5 4.4 5.4 Very slight Type specimen. Drummond. Do 75168 d" 15.9 5.4 7.7 3.5 4.5 5.4 Slight Tvpe locality. Do 75167 140810 « 10573 9 9 & 16.4 15.4 15.8 5.6 5.5 5.5 8.2 7.5 7.3 3.6 3.3 3.4 4.8 4.5 4.4 5.8 5.3 5.6 ...do ...do ...do Do. Do Do. North Carolina — Chapanoke. 1 Acad. Nat. Sci., Philadelphia. SOREX DISPAR GROUP The dispar group contains two species. — Sorex dispar and S. gaspensis. Geographic range. — Gaspe Peninsula, Quebec ; mountains of eastern New York, western Massachusetts, northeastern Pennsylvania, and southern West Virginia. (Fig. 9.) Diagnostic characters. — Size, medium ; color, dull grayish ; back almost con- color with underparts ; tail, long, moderately hairy ; skull, smooth, nonangular, long and narrow, moderately flattened ; rostrum, relatively long and narrow, depressed ; infraorbital foramen with posterior border lying behind the plane of interspace between m1 and m2; dentition, moderate; unicuspids, relatively narrow (extero-interior diameter), the third about equal the fourth in size. In color, members of the dispar group are not unlike S. fumeus in winter pelage but differ from any member of the fumeus or arcticus groups in rela- tively longer tail ; long, narrow, depressed rostrum ; narrow palate ; in the position of the infraorbital foramen ; weaker dentition, with pm2 with posterior portion of cingulum less expanded internally, and narrower unicuspidate teeth. Compared with any of the cinereus group, either species of the dispar group is larger, with larger, more flattened skull and more depressed rostrum, which is even more accentuated in ratio of length to breadth than that of 8. cinereus;- molariform teeth relatively wider (extero-interior diameter) than in cinereus group, the interior ridge from the apex to cingulum of the unicuspidate teeth less developed, less pigmented. Somewhat like the western S. trowbridgii in color, but with tail inot so clearly bicolor ; and radically different cranially and ventrally. Remarks. — The dispar group has a limited distribution in the eastern United States and Canada, and individuals belonging thereto are apparently rare, since less than 30 are known to exist in collec- tions. The group is clearly differentiated from any other Sorex in. the extreme posterior position of the infraorbital foramen. 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 89 SOREX DISPAR Batchelder Gbay Long-tailed Shrew (Pls. 2, s ; 4, d' ; 6, e ; 8, c ; 11, g ; 12, h) Sorex macrurus Batchelder (nee S. macrourus Lehrnann, 1822, qui Neomys fodiens), Proc. Biol. Soc. Washington 10: 133, December 8, 1896. Sorex (lis par Batchelder (substitute for S. macrurus Batchelder), Proc. Biol. Soc. Washington 24 : 97, May 15, 1911. Fig. 9. — Geographic range of species of Sorex dispar group 1. 8. dispar. 2. 8. gaspensis. Type specimen*. — No. 1384, collection of C. F. Batchelder, Cam- bridge, Mass.; $ adult (teeth unworn), skin and skull, left maxillary process broken away; collected September 9, 1895, by C. F. Batchelder. Type locality. — Beedes (sometimes called Keene Heights), Essex County, N. Y. 90 NORTH AMERICAN FAUNA [No. 5J Geographic range. — Mountains of eastern New York, western Massachusetts, northeastern Pennsylvania, and southern West Vir- ginia. (Fig. 9.) Diagnostic characters. — Similar to Sorex gaspensis but distinctly larger, darker, and tending to be more brownish in color ; hind foot actually larger than in S. gaspensis but in proportion to body length relatively smaller ; skull noticeably larger than that of 8. gaspensis, with corresponding heavier dentition. Color. — Winter pelage: Unknown. Summer pelage: Dull, grayish, scarcely paler ventrally than on back. Upper parts dark mouse gray to deep mouse gray with a perceptible tinge of chaetura black or chaetura drab, in some lights appearing more or less finely flecked with whitish ; underparts about same tone of color as upper parts, scarcely, if any, paler ; tail fuscous-black above, usually paler beneath, particularly basally, chaetura drab or hair brown, sometimes almost as dark as above (fuscous-black), occasionally, when much worn, honey yellow or cinnamon-buff. Upper surfaces of feet more or less clothed with fuscous hairs, particularly on outer edge. Time of molting. — Of the specimens of 8. dispar examined, only one is in process of molting. A male collected September 4, 1896, on Hunter Mountain, in the Catskills, N. T., shows indications of the incoming of the winter pelage under the worn summer pelage on the rump. Another male from the same place, collected a day previous, is in worn summer pelage with no signs of molting. Skull. — Medium in size ; smooth, not angular, moderately flattened ; rela- tively long and narrow, brain case narrow, rather low ; orbital region elongated (antero-posteriorly) ; rostrum relatively long and narrow, depressed; meso- pterygoid space elongate ; molarif orm dentition moderately heavy ; first incisors small; unicuspidate teeth relatively narrow (extero-interior diameter), the first and second about subequal, the third and fourth smaller than the first and second, the third about equal the fourth or possibly slightly smaller, the fifth relatively large but considerably smaller than third ; unicuspids with cingulum weakly developed. Measurements. — Type specimen (adult male) : Total length, 130; tail ver- tebrae, 60; hind foot, 15. Two adult males from Hunter Mountain, Catskill Mountains, N. Y. : Total length, 124, 121; tail vertebrae, 55, 56; hind foot, 15, 14. Two adult females from Hunter Mountain, Catskill Mountains, N. Y. : Total length, 122, 125; tail vertebrae, 56, 58; hind foot, 14.5, 15. Adult male from 4 miles southwest of Pemberton, W. Va. : Total length, 131 ; tail verte- brae, 62; hind foot, 14. Skull: Type specimen (adult male, teeth unworn) : Condylobasal length, 18.2 ; palatal length, 7.0 ; cranial breadth, 8.1 ; inter- orbital breadth, 3.5 ; maxillary tooth row, 6.1. Skulls of two adult males (teeth slightly worn) from Hunter Mountain, Catskill Mountains, N. Y. : Condylo- basal length, 18.0, 17.3; palatal length, 6.9, 6.6; breadth of cranium, 8.0, 7.9; interorbital breadth, 3.3, 3.3 ; maxillary breadth, 4.2, 4.2 ; maxillary tooth row, 6.2, 6.1. Skulls of two adult females (teeth slightly worn) from Hunter Mountain, Catskill Mountains, N. Y. : Condylobasal length, 17.6, 17.9 ; palatal length, 6.6, 6.8 ; cranial breadth, 8.0, 8.1 ; interorbital breadth, 3.3, 3.4 ; maxil- lary breadth, 4.2, 4.3 ; maxillary tooth row, 6.1, 6.1. Skull of adult male (teeth slightly worn) from 4 miles southwest of Pemberton, W. Va. : Condylobasal length, 18.0 ; palatal length, 6.9 ; cranial breadth, 8.0 ; interorbital breadth. 3.4 ; maxillary breadth, 4.5 ; maxillary tooth row, 6.3. Remarks. — So different from any other species of American shrew is S. dispar that, once the animal is known, critical comparisons with other forms in the region inhabited by it are unnecessary. In color it is not unlike 8. fumeus in winter pelage, but it can usually be dis- tinguished by its longer tail (55 or more). Cranially, it differs from all other American shrews in its relatively long, narrow, flat- tened skull with long, narrow, depressed rostrum and peculiar denti- tion. It does not fit into any other group of shrews, showing in minor respects suggestions of the cinereus, fumeus, and arcticus groups, but differing radically from each of them in other characters. Specimens examined. — Total number, 15, as follows: 192S] REVIEW OF AMERICAN LONG-TAILED SHREWS 91 Massachusetts: Mount Graylock, 3.1B New York: Beetles (sometimes known as Keene Heights), Essex County (type locality), 1 ; 18 Hunter Mountain, Catskill Mountains, 8; Mount Marcy (summit), l.18 Pennsylvania: Lake Leigh (North Mountain), Sullivan County, 1." West Virginia: Winding Gulf (4 miles southwest of Pemberton), 1. SOREX GASPENSIS Anthony and Goodwin Gaspe Peninsula Shrew Sorex gaspensis Anthony and Goodwin, Amer. Mus. Novitates, no. 109, p. 1, March 10, 1924. Type specimen. — No. 64190, Amer. Mus. Nat. Hist., $ young adult (teeth unworn), skin and skull, the skull slightly crushed in pterygoid region ; collected September 5, 1923, by G. G. Goodwin. Type locality. — Mount Albert, altitude 2,000 feet, Gaspe Peninsula, Quebec. Geographic range. — Known only from Gaspe Peninsula, Quebec. (Fig. 9.) Diagnostic characters. — Similar to 8. dispar but distinctly smaller, paler, and more grayish (less brownish) in color; hind foot small, actually considerably smaller than in 8. dispar, but in proportion to body length relatively slightly larger; skull distinctly smaller than that of 8. dispar with correspondingly weaker dentition. Color. — Winter pelage: Unknown. Summer pelage: Distinctly paler and more grayish than that of 8. dispar. Upper parts between deep mouse gray and deep neutral gray, or slightly paler ; underparts scarcely paler than upper parts ; tail essentially as in 8. dispar, possibly less fuscous ; feet a shade paler. Time of molting. — A female collected September 7, 1923, shows the intrusion of the winter fur under the summer pelage over the posterior half of the back. Skull. — Essentially similar in proportions to that of 8. dispar but decidedly smaller in all dimensions, the mesopterygoid space apparently relatively shorter, teeth smaller, and dental pigmentation heavier. Measurements. — Type specimen (young adult male) : Total length, 102; tail vertebrae, 47; hind foot, 10.5.18 Adult male and young adult female from type locality: Total length, 100, 95; tail vertebrae, 47, 47; hind foot, 12, 12. Skull: Type specimen (young adult male, teeth unworn) : Condylobasal length, 16.3; palatal length, 6.4 ; cranial breadth, 7.4 ; interorbital breadth, 3.0 ; maxillary tooth row, 5.6. Skulls of adult male (teeth moderately worn) and young adult female (teeth very slightly worn) from type locality: Condylobasal length, 16.1, 15.8 ; palatal length, 6.5. 6.4 ; cranial breadth, 7.3, 7.9 ; interorbital breadth, 3.2, 3.0 ; maxillary breadth, 4.0, 3.7 ; maxillary tooth row, 5.5, 5.6. Remarks. — Although the material at present available shows no specific connection between S. gaspensis and S. dispar, it is possible that additional specimens from the regions now separating the known geographic ranges of the two forms may show intergradation be- tween them. The members of this group of shrews are local in distribution, and apparently scarce where found. Specimens examined .—Three,19 from the type locality.19* M Manton Copeland coll., Brunswick, Me. 18 C. F. Batchelder coll., Cambridge, Mass. "Acad. Nat. Sci., Philadelphia. 18 Evidently an error ; measures 12 millimeters in the dry skin. u Amer. Mus. Nat. Hist. lu* George G. Goodwin, of the American Museum of Natural History, writes under date of Nov. 22, 1927, in a letter to the author : " * * * this summer I took nine speci- mens of Sorex gaspensis in the Gaspe" Peninsula, south of Shickshock Range. I found this shrew to be comparatively common in the Cascapedia Valley. The specimens average slighUy larger than the three that I took at Mount Albert ; hind foot in every case meas- uring 12.5 mm. One specimen had a total length of 115 mm., and 55 mm. for length of tail. They seem to have the same habits as the water shrew [i. e., Sorew palustris glover- alleni]." 74235—28 7 92 NORTH AMERICAN FAUNA [No. 51 Table 7. — Cranial measurements of adult specimens of Sorex dispar group Species and locality d H "3 03 a o O si "6> a a> "os 03 "3 si ■S 03 a) M .Q 'a "a 03 h o 03 is £s si fl .g O o 03 Wear of teeth Remarks S. dispar: New York- U384 83159 83160 83161 83162 2 64189 2 64190 2 64191 c? 9 9 & & d" 9 18.2 17.6 17.9 18.0 17.3 16.1 16.3 15.8 7.0 6.6 6.8 6.9 6.6 6.5 6.4 6.4 8.1 8.0 8.1 8.0 7.9 7.3 7.4 7.9 3.5 3.3 3.4 3.3 3.3 3.2 3.0 3.0 ..... 6.1 4. 2 6. 1 4.3 ; 6.1 4.2 6.2 4. 2 6. 1 4.0 5.5 5.6 3. 7 5- fi Unworn... Slight ...do ...do ...do Type specimen. Hunter Moun- tain. Do Do Do S. gaspensis: Q u e b e c— M o u n t Albert. Do. Type locality. Type specimen. Type locality. Do 1 C. F. Batchelder, coll., Cambridge, Mass. ' Amer. Mus. Nat. Hist. SOREX TROWBRIDGII GROUP The trowbridgii group includes a single species— Sorex trowbridgii. All necessary group comparisons are made under the species. SOREX TROWBRIDGII Baibd [Synonymy under subspecies] Geographic range. — Extreme southwestern British Columbia, west- ern Washington and Oregon, extreme northern California, western California south to San Raphael Mountains, and eastern California south to Kaweah River; chiefly in Transition Zone. (Fig. 10.) Diagnostic characters. — Size medium ; tail moderately long, sharply bicolor, dark above, nearly white (or pale ochraceous-buff) below; underparts of body scarcely, if any, paler than back. Skull medium in size, moderately depressed, the third unicuspid smaller than fourth, ridge extending from apex of unicuspid toward interior edge of cingulum but slightly pigmented and rarely pigmented to cingulum, separated from cingulum by antero-posterior groove, and never ending in distinct cusplet. Compared with 8. obscurus, the color of &. trow- bridgii and subspecies is more sooty, and it differs from any of the obscurus group in the relatively narrower teeth, and in that the internal ridge from the apex of the unicuspid is different. Compared with any of the orn-atus group, 8. trowbridgii is larger, with distinctly larger skull, higher brain case, broader cranium, broader mesopterygoid space, and longer tooth row. Subspecies and geographic variation. — The species trowbridgii is divided into four subspecies : trowbridgii, humboldtensis, montereyensis, and maripdsae. Beginning at the northern edge of the range of the species (subspecies trow- bridgii) and passing toward the south through the region inhabited by hum- boldtensis to that occupied by montereyensis, there is a gradually intensifying of the reddish element in the color and a shortening of the tail, particularly south of the type region of humboldtensis, and a broadening of the palate and rostrum, which is correlated with an increase in size of the molariform teeth. Toward the east and northeast of montereyensis the color of the animal becomes paler, recognized in the subspecies mariposae. Time of molting. — The spring molt usually occurs during June or late in May but may start as early as the last of April. Thus 7 specimens collected at Chehalis, Wash., April 26 to April 30, 1918, are all in more or less worn winter pelage, while 8, collected April 30, show the beginning of the molt on the back. Two males from Neah Bay, Wash., have the molt barely started May 14 and 23, while 1 from Sumas. British Columbia, is considerably more advanced May 26. Three individuals from near Inverness, Calif., have the molt well 1928] EEVIEW OF AMERICAN LONG-TAILED SHREWS 93 started May 27, and June 5 and 8, while 1 from near Cazadero, the same State, is in similar condition June 22. Specimens from Oregon show molt as follows : Swan Lake, June 13 ; Eugene, 2 females, June 18 ; Yaquina Bay, June 19 ; Vida, June 25 ; and Reston, July 7. Specimens show- ing late molt are 1 from Prattviile, Calif., molt nearly completed, July 24; 1 from East Fork Kaweah River, Calif., complete except for pos- terior half of back, July 28 ; and 1 from Canyon Creek, Calif., which still shows barely a trace of the winter fur on the rump, August 11. The fall molt in #. troicbridgii is usually completed by the first or second week of November and frequently by late in October. In rare in- stances it may begin as early as late in July, as in a male and female from Aptos, Calif., col- lected, respectively, on July 23 and 20, 1909, in each of which the fresh winter pelage is begin- ning to appear on the rump. Other evidence of early fall molt is found in 2 males from the south base of Santa Lucia Peak, Calif., Au- gust 26, 1902, and an- other from San Rafael Mountains, Calif., Au- gust 30, 1903. The ma- jority of fall specimens in actual process of molt appear to have been col- lected during October and are represented by specimens from Lake- view, Oreg., October 2; Verdi, Nev., October 15 and 1G ; and from the fol- lowing localities in Cali- fornia : Monterey, Oc- tober 6; Pacheco Peak. October 17 and 18 ; Orick, October 20 and 21 : Mich- igan Bluff, October 28. Five specimens from Point Reyes, Calif., col- lected November 3 to 5, are all in complete win- ter fur except 1 that still shows a trace of the summer fur on the head, and 2 females from Briceland are still in process of molting November 11. Only 3 specimens of the subspecies trowbridgii that show molt are available, all of which have the molt well advanced under the old fur ; they were collected at the following localities in Washington ; Mount Rai- nier, September 18 ; Luke Quinault, September 27 ; and Mount Stewart, October 2. Fig. 10. — Geographic range of subspecies of Sorex troic- b ridg ii t. troirbridpii. t. humboldtensis. 3. B. t. montercyctisis. 4. 8. t. mariposae. 94 NORTH AMERICAN FAUNA [No. 51 Remarks. — The species trowbridgii forms a well-differentiated group separated from others by distinct dental characters in the unicuspidate teeth. Its distribution is limited to a comparatively short range from southwestern British Columbia south over the northern two-thirds of California, where it is confined largely to the Transition Zone. The species appears to be less confined to marshes and damp habitats than some others and is frequently found in dry woods. SOREX TROWBRIDGII TROWBRIDGII Baied Tbowbbidge Shbew (Pls. 2, t; 5, c; 6, f; 8, d; 11, h; 12, i) Sorex trowbridgii Baird, Report Pacific R. R. Survey 8: part 1, Mammals, p. 13, 1857. Sorex trowbridgei True, Proc. U. S. Nat. Mus. 7 : 606, 1885. Sorex trowbridgii trowbridgii Jackson, Journ. Washington Acad. Sci. 12 : 264^ June 4, 1922. Type specimen. — Cotypes No. ^o¥§-> U. S. Nat. Mus.; sex un- known, adult (teeth moderately worn), poorly made skin and skull (cranium broken and right mandible missing) ; received from W. P. Trowbridge, United States Army; skin catalogued July, 1855, skull January, 1857. This (No. xcrrsO is tne onry specimen of which Baird gives skull measurements in the original description. It is hereby selected as the lectotype of Sorex trowbridgii. No. 967, U. S. Nat. Mus., sex unknown, poorly made skin without skull20; collected June 10, 1855, by James Wayne, received from W. P. Trowbridge, United States Army, and entered in the museum cata- logue November, 1855. Type locality. — Astoria, mouth of Columbia River, Clatsop County, Oreg. Geographic range. — Extreme southwestern British Columbia, west- ern Washington and Oregon, and extreme northwestern California (south to mouth of Klamath River). (Fig. 10.) Diagnostic characters. — Color rather dark and grayish ; darker and more grayish than S. t. montereyensis or #. t. mariposae, with relatively longer tail. Skull comparatively narrow, with noticeably narrow rostrum and weak denti- tion. About the color of S. t. humboldtensis, ratio of total length to length of tail vertebrae about the same, averaging a bit smaller. Rostrum and dentition weaker than in humboldtensis. Color. — Winter pelage: Upper parts deep mouse gray to almost dark mouse gray, slightly, if at all, tending toward brownish ; underparts scarcely paler than upper parts, mouse gray to deep mouse gray, sometimes very slightly tinged with drabish; tail sharply bicolor, fuscous, to fuscous-black or chaetura black above, whitish below, sometimes near cartridge buff or light buff, to tip. Summer pelage: A trifle more brownish and possibly paler than in winter. Upper parts between deep mouse gray and chaetura drab to between mouse grap and hair brown; underparts essentially like back, scarcely if any paler; tail about as in winter. Skull. — Comparatively narrow, particularly in rostral region; frontal region arising rather abruptly to moderately elevated brain case ; dentition weak, par- ticularly unicuspids and first upper molariform tooth. Rostrum narrower and dentition weaker than in any other form of S. trowbridgii. Measurements. —Two adult females from type locality: Total length, 115, 130 ; tail vertebrae, 56, 59 ; hind foot, 13, 14. Average of 4 adults females from 20 The skull has been removed from the skin, but is missing. It has not been entered in tne museum catalogue, and Baird (1857) makes no mention of it. 1928] EEVIEW OF AMERICAN LONG-TAILED SHREWS 95 Vida, Oreg. : Total length, 117 (113-120) ; tail vertebrae, 54.3 (52-57) ; hind foot, 13.5 (13-14). Average of 3 adult males from Sumas, British Columbia: Total length. 114.7 (112-117) ; tail vertebrae, 55.7 (55-57) ; hind foot, 13.8 (13.5-14). Skull: Lectotype (adult, sex unknown; teeth moderately worn): Condylobasal length, 17.5± : palatal length, 6.8 ; cranial breadth 8.6± ; inter- orbital breadth, 3.8 ; maxillary breadth, 5.0 ; maxillary tooth row, 6.5. Two skulls of adult females (teeth slightly worn) from type locality: Condylobasal length, 17.6, 17.6 ; palatal length, 6.9, 6.9 ; cranial breadth, 8.6, 8.9 ; interorbital breadth, 3.8, 3.8; maxillary breadth, 4.9, 5.0; maxillary tooth row, 6.4, 6.6. Average of 4 skulls of adult females (teeth slightly worn) from Vida, Oreg.: Condylobasal length. 17.3 (17.1-17.5) ; palatal length, 6.8 (6.7-6.9) ; cranial breadth, 8.6 (8.3-8.9) ; interorbital breadth, 3.8 (3.6-3.9) ; maxillary breadth, 4.9 (4.8-5.0) ; maxillary tooth row, 6.4 (6.3-6.5). Average of 3 skulls of adult males (teeth slightly worn) from Sumas, British Columbia: Condylobasal length. 17.5 (17.4-17.6) ; palatal length. 6.9 (6.8-7.0) ; cranial breadth, 8.6 (8.5-8.8) ; interorbital breadth, 3.9 (3.8-4.0) ; maxillary breadth, 5.0 (4.9-5.0) ; maxillary tooth row, 6.5 (6.4-6.6). Remarks. — Judging from the notes of various collectors of JS. t. trowbridgii, the form is apparently more plentiful in log-strewn forests than in marshes and habitats favorable to certain other shrews, and although confined within a geographic range covered by what is generally known as the humid northwest coast region of the United States, the little mammal is as apt to be found on the drier forested hills as on the moist lowlands. Wherever found the form is easily identified, as shrews go, by the characters pre- viously given, the long, sharply bicolor tail together with the size of the animal and the dark underparts of its body nearly concolor with its back being particularly good recognition marks. Direct intergradation with S. t. humboldtensis is indicated in specimens of JS. t. trovjbndgii from Requa and Crescent City, Calif. Also one of the four skulls from Prospect, Oreg., shows a noticeable approach toward humboldtensis in breadth of rostrum and size of molars, but the three others are like those of typical trowbridgii. Specimens from Siskiyou, Oreg., and Stud Horse Canyon, in the Siskiyou Mountains, and Preston Peak, Calif., shows a tendency toward JS. t. mariposae. A single specimen from Stehekin, at the head of Lake Chelan, Wash., appears a shade paler than average specimens of JS. t. trowbridgii. Specimens examined. — Total number, 219, as follows: British Columbia: Cascade Mountains (altitude 1,000 feet), 1; Douglas, l21; Hope, 4M; Second Summit (altitude 6,000 feet, Skagit River), 1 a ; Sumas, S6.23 California: Crescent City, 524; Klamath River (Happy Camp), 1; Preston Peak (altitude 5,500 feet), 1; Requa, 1 25 ; Siskiyou Mountains (alti- tude 6,500 feet, Stud Horse Canyon), 2. Oregon: Astoria (type locality), 6; Blaine, 3M; Blue River, 1; Drew, 1; Empire, 1; Eugene, 2; Fish Hatchery (2 miles west of Vida), 2; Gold Beach, 4 2ri ; Marshfield. 1 ; Mercer, 1 27 ; Netarts, 6 28 ; Netarts Bay, 227; Parkdale (2 miles west, altitude 1,500 feet), 3; Philomath (5 miles southwest), 4; Portland, 1; Prospect, 428; Prospect (Rogue River), 1 28 ; Reston, 1; Siskiyou, 3; Three Sisters (Alder Springs, altitude 4,300 feet), 1; Three Sisters (north slope, altitude 6,000 feet), 2; Vida, 9; Yaquina Bay, 2. n Nat. Mus. Canada. ^ Field Mus. Nat. Hist. »Mus. Coinp. Zool. 2SD. R. Dickey coll., 1; Univ. Mich., 2. ad. Nat. Sci. Philadelphia, 9; Field ■' B. O. Jewett coll., Portland, Oreg. Has Nat. Hist, 2. ' D. K. Dickey coll. 24 I). R. Dickey coll., Pasadena, Calif., 1. 96 NORTH AMERICAN FAUNA [No. 51 Washington: Aberdeen, 2; Ashford (Nisqually Valley), 1; Blaine, 2; Blew- ett Pass (2 miles south, altitude 3,000 feet, Kittitas County), 1 Blewett Pass (altitude 5,000 feet, Chelan County), l28; Brookfield, 1 Cathlamet, 1; Cedarville, 1; Chehalis, 5; Chehalis (8 miles west), 3 Darrington (altitude 600 feet), 1; Destruction Island, 1; Duckabush, 1; Elwha, l29; Elwha (altitude 425 feet), 2; Everett, 229; Everett (4 miles south), 1; Glacier (altitude 900 feet), 1; Harstine Island, Mason County, 1; Hoodsport, 2; Index, 2; Index (north fork Skykomish River), l29; Kapowsin, 1; Kirkland, 1 25 ; Kirkland (3 miles east), 4; Lake Cushman, 1 30 ; Lake Quinault, 1; Lake Whatcom, 1; Longmire (Mount Rainier, altitude 3,000 feet), 32S; Mount Rainier (Spray Park), l29; Mount Rainier (Obanapecosh Springs, altitude 2,000 feet), 4; Mount Rainier (Mesler's ranch, 1 mile west Rainier Park, altitude 2,000 feet), 1; Mount Rainier (Tahoma Creek, altitude 2,500 feet), 1: Mount Stewart (6 miles south, north fork Teanaway River, altitude 3,500 feet), 2; Mount Vernon, 12; Neah Bay, 8; North Bend (altitude 600 feet) , 1 ; Olympic Mountains (3 miles southeast of Mount Elinor) , 3 ; Point Defiance Park, Tacoma, 228; Potlatch, 42S; Puget Sound, 2; Puyallup, 531; Rockport (altitude 300 feet), 2; Seattle, 232; Stehekin (head of Lake Chelan, altitude 1,079 feet), 1; Steilacoom, 3; Tacoma (5 miles east), l28; Tacoma (6 miles south), 1; Tenino, 2; Tokeland (Shoalwater Bay), 1. SOREX TROWBRIDGII HUMBOLDTENSIS Jackson Humboldt Bay Shrew Sorex trowbridgii humboldtensis Jackson, Journ. Washington Acad. Sci. 12: 264, June 4, 1922. Type specimen. — No. 97271, U. S. Nat. Mus., Biological Survey collection; $ adult (teeth slightly worn), skin and skull; collected June 11, 1899, by Walter K. Fisher. Original number, 914. Type locality. — Carsons Camp, Mad River, Humboldt Bay, Hum- boldt County, Calif. Geographic range. — Coastal region of northern California from mouth of Klamath River south to Point Arena. (Fig. 10.) Diagnostic characters. — In general intermediate between Sorex t. troicbridgii and S. t. montereyensis. About the color of 8. t. troicbridgii, tending to be larger, with tail proportionately to body length about as in the subspecies troicbridgii. Skull larger, and broader in all dimensions than that of &. *. trowbridgii, with heavier dentition. Averaging a shade darker and less brownish than montereyensis, with relatively and actually longer tail ; skull with narrower rostrum and somewhat weaker dentition than in montereyensis. Color. — Winter pelage: Essentially like that of S. t. troicbridgii; possibly averaging a shade darker. Summer pelage: Indistinguishable from that of 8. t. trowbridgii. Skull. — Intermediate in most respects between that of S. t. trowbridgii and that of S. t. montereyensis. Larger and broader than that of 8. t. troicbridgii, with noticeably heavier rostrum and dentition. Rostrum narrower and denti- tion weaker than in S. t. mariposae or montereyensis. Measurements. — Type specimen (adult male) : Total length, 132; tail verte- brae, 62; hind foot, 14. Average of three adult males from Areata, Humboldt County, Calif.: Total length, 131 (129-133) ; tail vertebrae, 60.3 (60-61) ; hind foot, 14.7 (14-15). Skull: Type specimen (adult male; teeth slightly worn) : Condylobasal length, 17.S ; palatal length, 7.2 ; cranial breadth, 8.9 ; interorbital breadth, 4.1 ; maxillary breadth, 5.4 ; maxillary tooth row, 6.7. Average of three skulls of adult males (teeth moderately worn) from Areata, Humboldt County, Calif. : Condylobasal length, 18.2 (18.2-18.3); palatal length, 7.2 25 Field Mus. Nat. Hist. a D. R. Dickey coll., 1 ; G. G. Cantwell 28 D. R. Dickey coll. coll., Palms, Calif., 1. 28 State Coll. of Wash. 32 Field Mus. Nat. Hist., 1. 30 Univ. Mich. 1928] REVIEW OF AMERICAN" LONG-TAILED SHREWS 97 (7.1-7.2) ; cranial breadth, 9.0 (8.9-9.1) ; interorbital breadth, 4.0 (3.9-4.1) ; maxillary breadth, 5.3 (5.3-5.3) ; maxillary tooth row, 6.9 (6.9-6.9). Remarks. — Although in reality intermediate between S. t. trow- bridgii and S. t. montereyensis, the subspecies humboldtensis in its typical form averages larger than either. Externally it is more like S. t. troiobridgii; cranially it is apparently rather nearer monterey- ensis. Some of the specimens from near Mendocino City, Calif., here referred to humboldtensis, could be referred to montereyensis with almost equal propriety in so far as external characters are concerned ; cranially, however, they are much nearer humboldtensis. Specimens from 7 miles north of Hardy, Mendocino County, Calif., have shorter tails than typical representatives of humboldtensis, but in color and cranial characters they are similar to this subspecies. Specimens examined. — Total number, 80, as follows: California: Alton Junction, 2; Areata, 533; Briceland, 2; Carlotta, 8M; Cuddeback, l85; Dyerville, 2M; Dyerville (5 miles south), 3; Eureka, ll36; Fair Oaks, 337; Hoopa Valley, 7; Mad River, Humboldt Bay (type locality), 10; Mendocino, ll38; Orick, 5s8; Sherwood, l39; Trini- dad, 9.M SOREX TROWBRIDGII MONTEREYENSIS Merkiam Monterey Shkew (Pl. 2, u) Borex montereyensis Merriam, North Amer. Fauna No. 10, p. 79, December 31, 1895. Borex montereyensis montereyensis Grinnell, Univ. California Publ. Zool. 10: 188, March 20, 1913. Borex tlrowbridgii] montereyensis Jackson, Journ. Washington Acad. Sci. 12: 264, June 4, 1922. Type specimen. — No. ||ggg, U. S. Nat. Mus., Biological Survey collection; S adult (teeth slightly worn), skin and skull (basi- occipital region slightly broken) ; collected October 1, 1891, by Vernon Bailey. Type locality. — Monterey, Monterey County, Calif. Geographic range. — Coast region of California from Point Arena south to San Raphael Mountains. (Fig. 10.) Diagnostic characters. — Tending to be more brownish than either Borex t. trowbridgii or 8. t. humboldtensis, with relatively and actually shorter tail ; skull broader and heavier than that of either subspecies troxebridgii or hum- boldtensis, the postorbito-frontal region less constricted, dentition heavier. Slightly darker and more brownish than 8. t. mariposae, with apparently aver- age smaller skull with narrower cranium. Color. — More brownish than 8. t. trowbridgii, particularly in winter. Winter pelage: Upper parts chaetura drab or between chaetura drab and fuscous; underparts but slightly paler than upper parts, hair-brown or slightly paler; tail distinctly bicolor, less sharply so than in 8. t. trowbridgii, above fuscous to mummy brown, below near pale orchraceous-buff nearly to tip. Summer pelage: Scarcely paler or more brownish than in winter. Upper parts fuscous, sometimes tending toward chaetura drab or olive-brown ; underparts near hair brown ; tail as in winter. Bkull. — Moderately broad and heavy ; broader throughout and a trifle more flattened cranially than that of 8. t. trowbridgii or 8. t. humboldtensis, with " Mus. Vert. Zool., 3. 87 Mus. Comp. Zool., 1 : Mus. Vert. Zool., 2. M D. R. Dickey coll., Pasadena, Calif. M Field Mus. Nat. Hist., 2 ; Mus. Vert M Mus. Vert. Zool. Zool., 6. •• Field Mus. Nat. Hist., 4 ; Mus. Vert. *> D. R. Dickey coll., 1. Zool., 6. 98 NORTH AMEEICAN FAUNA TNo. 51 rostrum averaging distinctly broader and dentition heavier. Similar to that of S. t. mariposae, possibly averaging smaller with narrower cranium. Measurements. — Type specimen (adult male) : Total length, 120; tail ver- tabrae, 54 ; hind foot, 15. Average of five adult females from Monterey and Pacific Grove, Calif.: Total length, 123 (114-131) ; tail vertebrae, 51.8 (48-56) ; hind foot, 14.3 (13.5-15). Skull: Type specimen (adult male; teeth slightly worn): Condylobasal length, 18.0; palatal length, 7.3; cranial breadth, 9.1; interorbital breadth, 4.2 ; maxillary breadth, 5.6 ; maxillary tooth row, 7.0. Average of five skulls of adult females (teeth slightly worn) from Monterey and Pacific Grove, Calif. : Condylobasal length. 18.2 (18.0-18.4) ; palatal length, 7.3 (7.2-7.4) ; cranial breadth, 9.0 (8.7-9.2) ; interorbital breadth, 4.2 (4.0^.3) ; maxillary breadth, 5.6 (5.4-5.7) ; maxillary tooth row, 6.9 (6.8-7.0). Remarks. — Although for many years treated as specifically dis- tinct from S. trowbridgii, the form S. t. montereyensis is clearly sub- specifically connected with it through humboldtensis. Specimens of montereyensis from Nicasio, Point Reyes, and other points in Marin County, Calif., show in some respects an approach toward hum- boldtensis in that they seem to average somewhat darker than typical montereyensis, and a few of the skulls show narrower rostra; their dentition, however, is heavy, as in true montereyensis, in nearly every individual. Summer specimens from Mount Saint Helena, Calif., are inclined toward S. t mariposae in color. Specimens examined. — Total number, 284, as follows : California: Aptos (5 miles south), 5; Bear Basin (head of Carmel River), 3; Bear Valley (head of Carmel River), 1; Bells Station, Santa Clara County, 1; Berkeley, 340; Berkeley (Strawberry Creek), l41; Berke- ley (first canyon north of Strawberry Creek), 1"; Boulder Creek, 3; Cazadero (7 miles west, altitude 900 feet), S42; Cone Peak, Monterey County, 1; Divide (altitude 800 feet, between Mill Valley and Muir Woods) , Marin County, 1 42 ; Freestone (altitude 300 feet) , 1 ** ; Freemont Peak, Gabilan Range, 2; The Geysers, Sonoma County, 1; Gilroy (near), 343; Gualala, 5**; Gualala (Sonoma County side of Gualala River), 341; Hardy (7 miles north, Mendocino County), 441; Hayward, l42; Inverness, 37"; Lagunitas, 4; La Honda, 215; La Honda, San Mateo County, 13 46; Marin County, 2; Mendocino, 412; Menlo Park, l42; Milpitas Ranch (south base Santa Lucia Peak), 2; Monterey (type locality), 16"; Morro, 2; Mount Hamilton, l45; Mount St. Helena, 2; Mount Tamalpais, 1; Mount Tamalpais (altitude 2,000 feet), 1; Mount Veeder, 16; Nicasio, 68 ** ; Oakland, 649; Olema, 1; Pacheco Peak (summit), 3; Pacific Grove, 2; Palo Alto, l43; Petaluma, 3 4B ; Pine Valley (10 miles northwest Tassajara Springs) , 1 ; Point Pinos (Pacific Grove P. O.), Monterey County, 1; Point Reyes, 24 50; Portolo, San Mateo County, 17 B1; San Luis Obispo, 1; San Rafael Mountains (Peach Tree River), Santa Barbara County, 1; Stevens Creek, Santa Clara County, 242; Tassajara Creek (6 miles below Tassajara Springs), 3; Telegraph Canyon (near Berkeley), l42; Watsonville (10 miles north, altitude 2,000 feet), Santa Cruz County, l52; Woodside (Santa Cruz Mountains, San Mateo County), 1. SOREX TROWBRIDGII MARIPOSAE Gkinnell TOSEMITE SHEEW Sorex montereyensis mariposae Grinnell, Univ. Calif. Publ. Zool. 10 : 189, March 20, 1913. Sorex troworidgii mariposae Grinnell, Univ. Calif. Publ. Zool. 21 : 314, January 27, 1923. 40 Acad. Nat. Sci., Philadelphia. i8Mus. Vert. Zool., 1; Mus. Comp. Zool., 41 D. R. Dickey coll., Pasadena, Calif. 4 ; Field Mus. Nat. Hist., 10 ; Amer. Mus. 43 Mus. Vert. Zool. Nat. Hist., 13. 43 Mus. Comp. Zool. 49 Public Mus. Milwaukee. 44 D. B. Dickey coll., 6 ; Mus. Vert. Zool., 50 Field Mus. Nat. Hist., 5 ; Mus. Comp. 31. Zool., 3 : Mus. Vert. Zool., 3. 45 Field Mus. Nat. Hist. a D. R. Dickey coll., 2; Field Mus. Nat. 46 Amer. Mus. Nat. Hist. Hist., 11; Mus. Comp. Zool., 4. 47 D. R. Dickey coll., 6 ; Mus. Vert. Zool., 6. B2 Univ. of Mich. 1928] KEVIEW OF AMERICAN" LONG-TAILED SHREWS 99 Type specimen. — No. 12979, Mus. Vert. Zool., Univ. California; 9 adult (teeth moderately worn), skin and skull; collected May 27, 1911, by J. and H. W. Grinnell. Type locality. — Yosemite Valley, 4,000 feet altitude, Mariposa County, Calif. Geographic range. — Extreme south-central Oregon, northern Cali- fornia south through the inner coast range, to about 39° 30' N., and southeast of Sacramento Valley to Kaweah River; also extreme west-central Nevada near California line. (Fig. 10.) Diagnostic characters. — Similar to Sorex t. montereyensis but slightly paler and more drabish (less brownish) in color, with apparently average larger skull with broader brain case. Color. — Winter pelage: Upper parts between hair brown and mouse gray; underparts smoke gray tinged with light drab or drab gray, tail bicolor, hair brown to drab above, below pale ochraceous-buff to nearly pale pinkish cinna- mon. Summer pelage: Upper parts hair brown; underparts light drab or slightly paler, sometimes near drab gray ; tail bicolor, hair bown to drab above, near tilleul buff below. Skull. — Similar to that of S. t. montereyensis, possibly averaging larger with broader cranium. Measurements. — Type specimen (adult female): Total length, 121; tail vertebrae, 51 ; hind foot. 14. Average of three adult females from altitudes 4,700 feet to 6,400 feet, Yosemite Park, Mariposa County, Calif. : Total length, 117.7 (116-120) ; tail vertebrae, 50.3 (50-51) ; hind foot, 14.5 (14-15), Skull: Type specimen (adult female; teeth moderately worn): Condylobasal length, 18.6 ; palatal length, 7.7 ; cranial breadth, 9.3 ; interorbital breadth, 4.2 ; max- illary breadth, 5.7; maxillary tooth row, 7.1. Average of three skulls of adult females (teeth slightly worn) from altitudes of 4,700 feet to 6,400 feet, Yose- mite Park, Mariposa County, Calif. : Condylobasal length, 18.6 (18.4-18.8) ; pala- tal length 7.4 (7.3-7.5) ; cranial breadth 9.3 (9.2-9.4) ; interorbital breadth, 4.2 (4.1^1.3) ; maxillary breadth, 5.6 (5.5-5.7) ; maxillary tooth row, 7.0 (6.9-7.1). Remarks. — The Yosemite form of trowbridgii closely resembles S. t. montereyensis, being but slightly paler and more drabish in color, and with slight cranial differences. It apparently intergrades with both S. t. humboldtensis and montereyensis. The four speci- mens from Liarly Ranch, 4 miles south of Mount Sanhedrin, Men- docino County, Calif., are provisionally referred to S. t. mariposae. Two of these specimens, collected in August, 1905, are more brown- ish than typical mariposae and in color show an approach toward montereyensis, to which they could about as well be referred; the other two, collected in August, 1913, are darker and more grayish than typical mariposae and in this respect approach humboldtensis, but the skulls show no tendency toward that form. A single speci- men from South Yolla Bolly Mountain, Calif., is like maHposae in color, but cranially shows a distinct similarity to humboldtensis. Except that the skulls may run a trifle smaller and the dentition appears weaker, specimens from Swan Lake Valley and Lakeview, Oreg., do not differ from those of average typical mariposae. Specimens examined. — Total number, 77, as follows: California: American River (Middle Fork), 1; Beswick, 1; Canyon Creek (altitude 4,600 feet), 3; Carberry's Ranch, 1; Castle Lake (altitude 5,434 feet), 1"; Chinquapin (near, altitude 6,200 to 6,400 feet), 3"; Cisco (altitude 6,000 feet), 4M; Downieville, 1; Dutch Flat (altitude 3,400 feet), 1M; Eldorado County, 2; Emerald Bay, 1; Emerald Bay (Lake Tahoe), 1; Fyffe, 2M; Fyffe (altitude 3,600 feet), l53; Gentrys Big Oak Flat Road (altitude 5,800 feet, Yosemite Park), l63; Giant Forest, Sequoia National Park, 1 ; Glen Alpine Springs, 1 M ; Hayden B" Mus. Vert. Zool. u Mus. Comp. Zool. 100 NORTH AMERICAN FAUNA TNo. 51 Hill, 1; Indian Canyon (east fork, Mariposa County), l53; Jackson Lake, Siskiyou County, l63; Kaweah River (east fork), 1; Liarly's Ranch (8 miles east of Hearst, Mendocino County), 365; Merced Grove Big Trees, Mariposa County (altitude 5,400 feet), 253; Merced Lake (2 miles east), Yosemite Park, l63; Michigan Bluff, 1; Mount Lassen (south base, Mill Creek, altitude 5,000 feet), 2; Mount Shasta (Mud Creek, timber-line), 1; Mount Shasta (Upper Mud Creek), 4; Mount Tallac, Eldorado County, 356; Myers, Eldorado County, 1"; Parker Creek, Warner Mountains, 263; Placerville, l63; Prattville (12 miles northeast), 1; Sequoia National Park (Halsted Meadows), 5; Slipperyford, l68; South Yolla Bolly Mountain, 265; Squaw Creek Valley (Warmcastle Soda Springs), 1; Sweetwater Creek (altitude 3,800 feet, 2 miles east Feliciana Mountain), 253; "The Spring" Yosemite Falls Trail, Yosemite Valley (altitude 4,700 feet), 263; Tower House, Shasta County, 253; Tower House (altitude 1,268 feet), Shasta County, I63; Yosemite, 1; Yosemite Valley (altitude 4,000 feet, type locality), 3.53 Nevada: Verdi, 2. Oregon: Lakeview, 1; Swan Lake Valley, 2. Table 8. — Cranial measurements of adult specimens of Sorex trowbridgii group Species and locality "3 A ■Qja % 03 P, ^tr„ CD x> d M »>3 ■a" a o "3 Is 1 '3 a £ 02 O Ph D 3088 17. 5± 6.8 8.6± 24315 9 17.6 6.9 8.6 89021 9 17.6 6.9 8.9 204440 9 17.5 6.9 8.9 204473 9 17.3 6.9 8.4 204475 9 17.1 6.7 8.8 204476 9 17.1 6.8 8.3 62999 c? 17.5 6.9 8.6 99748 of skull and dentition, and in some respects is rather aberrant ; but it is perfectly connected by gradual intergrades with the rest of the series and can be matched almost perfectly by occasional skulls of amoenus throughout the range of the subspecies. Specimens examined. — Total number, 292, as follows: California: Alvord (Owen Valley), Inyo County, 1; Bear Creek (head, alti- tude, 6,400 feet), Trinity County, l94; Beswick, 1; Bieber (altitude, 4,500 feet), 1; Brownell, Klamath Lake (altitude, 4,300 feet), 1; Buck Ranch, Plumas County, 20 95; Burney (12 miles east, altitude, 4,700 feet, Redding-Bieber Road), 1; Canyon Creek (altitude, 6,000-7,500 feet), Trinity County, 2; Carberrys Ranch, 4; Cassel, 2; Castle Lake (altitude, 5,400 feet), Siskiyou County, 204; Cliff House, 296; Dana, 17; Davis Creek, Goose Lake, 1; Donner (altitude, 7,500-7,900 feet), 3; Dry Creek (altitude, 4,800 feet), Warner Mountains, l94; Fall Lake, Fall River Valley, 3 ; Fort Crook, 11 ; Goose Nest Mountain, Siskiyou County, 2 ; Hornbrook, 3 ; Independence Lake, Nevada County, 7 M ; Lake Audrain, 1 85 ; Lassen Creek, Modoc County, 1 95 ; Lassen Peak, 13 ; Lincoln Creek, Sierra County, 1; Long Valley (Convict Creek, altitude, 6,800 to 6,900 feet), 495; Mammoth, Mono County, 395; Mammoth (head of Owens River, near) (type locality), 2; May ten, 997; Mono Lake, 598; Mount Conness, 1; Mount Dana, 1; Parker Creek (head north fork, altitude, 7,300 feet), "Warner Mountains, 2s4; Phillips, Eldorado County, 295; Plumas County (altitude, 6000 feet), 1; Prattville (12 miles north- east), 2; Rush Creek (head), Siskiyou County, 3s4; Salmon River (south fork), Siskiyou County, 3M; Shasta region, l94; Sierra Valley, l; Sisson, 24"; Spring Garden Ranch (vicinity), Grizzly Mountains, 3; Squaw Creek (head of), Mount Shasta, 2; Squaw Creek (altitude, 7,800 to 8,100 feet), Mount Shasta, 3; Sugar Hill (Goose Lake Mead- ows, altitude, 4,800 feet), Modoc County, l94; Tallac, 3; Upper Ash Creek, Mount Shasta, 1 ; Upper Mud Creek, Mount Shasta, 8 ; Wagon Camp, Mount Shasta (type locality of shastensis) , 5; Warmcastle Soda Springs, Squaw Creek Valley, 2; Warner Creek (Drakes Hot Springs), 1; Williams Butte (altitude, 6,900 feet), 2.94 Nevada: Mountain City, 1; Reno, l94; Ruby Lake, 3; Ruby Mountains, 9. Oregon: Anna Creek, Mount Mazama (altitude 6,000 feet), 1; Crater Lake. 24 *; Diamond Lake, 6; Fort Klamath, 36 2; Klamath Falls, 83; Old Fort Klamath, l95; Plush, Lake County, 1; Warner Creek, Warner Mountains, 1; Warner Mountains, 3. SOREX VAGRANS MONTICOLA Merriam Rocky Mountain Shrew (Pl- 2, T) Sore® monticolus Merriam, North Amer. Fauna No. 3, p. 43, September 11, 1890. Sorex dohsoni Merriam, North Amer. Fauna No. 5, p. 33, July 30, 1891. Type locality, Alturas or Sawtooth Lake, altitude about 7,200 feet, east base of Sawtooth Mountains, Blaine County, Idaho. Si Mus. Vert. Zool. 99 Mus. Vert. Zool., 4 ; D. R. Dickey coll., 13. 93 D. R. Dickey coll., Pasadena, Calif. * D. R. Dickey coll., 6 ; Mus. Vert. Zool., 1. 80 Mus. Comp. Zool. a Amer. Mus. Nat. Hist., 1. H ?JUS- Vert- ZooL' 7- 3 G- G- Cantwell coll., Palms, Calif., 2. 88 Mus. Vert. Zool., 4. 1928] REVIEW OF AMERICAN LONG-TAILED SHREWS 111 Sorex vagrans dobsoni Merriam, North Anier. Fauna No. 10, p. 68, December 31, 1895. Sorex vagrans monticola Merriarn, North Anier. Fauna No. 10, p. 69, December 31, 1895. Type specimen. — No. Will, U. S. Nat. Mus., Biological Survey collection; $ adult (teeth slightly worn), skin and skull; collected August 28, 1889, by C. Hart Merriam and Vernon Bailey. Type locality. — San Francisco Mountain, altitude 11,500 feet, Co- conino County, Ariz. Geographic range. — Extreme southern British Columbia, east- ern Washington, eastern Oregon, Idaho, western Montana, western Wyoming, south through eastern Utah, western Colorado, eastern Arizona, and western New Mexico to southern Chihuahua, Mexico. (Fig. 11.) Diagnostic characters. — Differs from So-rex v. vagrans, S. v. amoenus, S. v. nevadensis, and 8. v. orizabae in its slightly larger size and longer tail, dis- tinctly paler and more grayish coloration (particularly in winter) ; skull with heavier rostrum and larger teeth than in S. v. vagrans, amoenus, or nevadensis. Skull relatively broader, brain case expanding more abruptly anteriorly, shorter maxillary tooth row, and heavier unicuspids than in orizaoae. Similar to S. o. obscurus, but with average shorter tail and smaller foot, weaker and shorter rostrum, and smaller teeth. Color. — Winter pelage: Upper parts hair brown or between hair brown and chaetura drab ; sides and flanks usually a trifle paler than back, drabbish ; underparts pale olive-gray washed with pale pinkish buff; tail bicolor, hair brown or olive-brown above, avellaneous or pinkish buff below. Summer pelage: Upper parts usually between olive-brown and fuscous, sometimes almost hair brown, frequently somewhat darker posteriorly than anteriorly ; sides usually about same color as upper parts, sometimes slightly paler (drabbish) ; under- parts smoke gray washed with avellaneous, vinaceous-buff, or occasionally tilleul buff; tail about as in winter or less clearly bicolor. Skull. — Rostrum and teeth relatively heavy for the species vagrans, heavier than in any other subspecies of S. vagrans except S. v. halicoetes. Skull prac- tically indistinguishable from that of halicoetes. Similar to that of S. o. obscurus, but palate averaging narrower; rostrum smaller, narrower, and more attenuate ; dentition weaker, especially unicuspidate teeth. Measurements. — Type specimen (adult male) : Total length, 107; tail verte- brae, 45 ; hind foot, 12.5. Average of 3 adult males from Mount Thomas, White Mountains, Ariz.: Total length, 107.3 (104-110) ; tail vertebrae, 41.3 (40-43) ; hind foot, 12.8 (12.5-13). Adult male from Alturas (Sawtooth) Lake, Blaine County, Idaho (type locality of