Skip to main content

Full text of "North American fauna"

> 



,\ 5"<\0£ AA 



ilf-Z^i yjlLIi; IL'i-UJJ AjI> <- : ' \ 







Give n B y 



V\ S S^v} o^ ^o ^u\r\e\\Vs> 



U. S. DEPARTMENT OF AGRICULTURE 

BUREAU OF BIOLOGICAL SURVEY 



NORTH AMERICAN FAUNA 



ISTo. 31 

[Actual date of publication, October 19, 1910] 




-y 






REVISION OF THE WOOD RATS OF THE 
GENUS NEOTOMA 



EDWARD A. GOLDMAN 

FIELD NATURALIST, BIOLOGICAL SURVEY 



Prepared under the direction of 

HENRY W. HENSHAW 

CHIEF OF BUREAU OF BIOLOGICAL SURVEY 




WASHINGTON 

GOVERNMENT PRINTING OFFICE 

1910 



U. S. DEPARTMENT OF AGRICULTURE 

BUREAU OF BIOLOGICAL SURVEY 



NORTH AMERICAN FAUNA 

No. 31 

[Actual date of publication. October 19, 1910] 



t 



REVISION OF THE WOOD RATS OF THE 
GENUS NEOTOMA 



EDWARD A. GOLDMAN 

FIELD NATURALIST, BIOLOGICAL SURVEY 




Prepared under the direction of 

HENRY W. HENSHAW 

CHIEF OF BUREAU OF BIOLOGICAL SURVEY 




WASHINGTON 

GOVERNMENT PRINTING OFFICE 

19 10 



LETTER OF TRANSMITTAL. 



U. S. Department of Agriculture, 

Bureau of Biological Survey, 

Washington, D. C, June 22, 1910. 
Sir: I have the honor to transmit herewith for publication as 
North American Fauna No. 31, Revision of the Wood Rats of the 
Genus Neotoma, by Edward A. Goldman, field naturalist of the 
Biological Survey. The wood rats are restricted to North America, 
where they are widely distributed, especially in the United States, in 
the southern part of which they range from the Atlantic to the 
Pacific coasts. They are especially numerous in the arid West, 
where they do some injury to crops and especially to stored grain. 
As the West becomes more densely populated they are likely to inflict 
still greater injuries. In view of this, and especially of the fact that 
in California the Public Health and Marine-Hospital Service has 
recently discovered that these rats may serve as carriers of the plague, 
the definite information concerning the numerous species and their 
distribution contained in this report is important and timely. 
Respectfully, 

H. W. Henshaw, 

Chief, Biological Survey. 
Hon. James Wilson, 

Secretary of Agriculture. 



CONTENTS 



Page. 

Introduction 7 

History and material 7 

Distribution .' 8 

Habits and economic status 10 

General characters 11 

Explanation of measurements 12 

Subfamily Neotominse 13 

Genus Neotoma 13 

Subgenera and minor groups 13 

List of species and subspecies, with type localities 14 

Key to subgenera 15 

Key to species and subspecies 16 

Subgenus Neotoma 20 

Subgenus Homodontomys 86 

Subgenus Teonoma 94 

Index 123 

5 



ILLUSTRATIONS. 



PLATES. 

Page. 
Plate I. Skulls of Neotoma Jloridana, N.f. rubida, N.f. baileyi, N. micropus, 

N. m. planiceps, N. pennsylvanica 108 

II. Skulls of Neotoma albigula, N. a. venusta, N. a. melanura, N. a. 

leucodon, N. latifrons, N. nelsoni 110 

III. Skulls of Neotoma palatina, N. montezumse, N. intermedia, N. i. 

pretiosa, N. i. arenacea, N. bryanti 112 

IV. Skulls of Neotoma anthonyi, N. martinensis, N. mexicana, N. m. 

fallax, N. m. pinetorum, N. m. bullata 114 

V. Skulls of Neotoma navus, N. torquata, N. distincta, N. ferruginea 

chamula, N. f. solitaria, N. f. isthmica 116 

VI. Skulls of Neotoma ferruginea picta, N. parvidens, N. tropicalis, N. 

desertorum, N. I. stephensi, N. goldmani 118 

VII. Skulls of Neotoma (Homodontomys) fuscipes, N. (II.) f. macrotis, N. 

(H.)f.annectens. Upper molars of N. Jloridana, N. (H.) fuscipes. 120 
VIII. Skulls of Neotoma (Teonoma) cinerea, N. (T.) c. saxamans, N. (T.) 

c. orolestes, N. (T.) c. rupicola, N. (T.) c. arizonse 122 

TEXT FIGURES. 

Fig. 1. Distribution of the genus Neotoma 9 

2. Distribution of Neotoma floridana and subspecies 21 

3. Distribution of Neotoma micropus and subspecies 27 

4. Distribution of the Neotoma albigula group 32 

5. Distribution of the Neotoma intermedia group 43 

6. Distribution of Neotoma mexicana and subspecies 55 

7. Distribution of Neotoma navus, N. torquata, N. distincta, N. tropicalis, 

and N. parvidens 62 

8. Distribution of Neotoma ferruginea and subspecies 68 

9. Distribution of Neotoma chrysomelas 75 

10. Distribution of Neotoma desertorum, N. lepida, and subspecies 77 

11. Distribution of Neotoma goldmani 82 

12. Distribution of Neotoma pennsylvanica 84 

13. Distribution of the subgenus Homodontomys (Neotoma fuscipes and 

subspecies) 88 

14. Distribution of the subgenus Teonoma (Neotoma cinerea and sub- 

species) 96 

6 



No. 31. NORTH AMERICAN FAUNA. October, 1910. 



EEYISION OF THE WOOD RATS OF THE GENUS NEOTOMA. 



By Edward A. Goldman. 



INTRODUCTION. 
HISTORY AND MATERIAL. 

The earliest reference in literature to a native American wood rat 
of the genus Neotoma was by Peter Kalm, a who in 1749 cited John 
Bartram, of Philadelphia, as authority for the statement that he saw 
a "great number" of rats which lived among the rocks in the "Blue 
Mountains" of Pennsylvania, came out only at night, and made a 
"terrible noise." The animals seen by Bartram probably were of 
the species now known as Neotoma pennsylvanica, which was des- 
tined to remain without a tenable name until so recently as 1893. 
A bushy-tailed species, based on the description of Lewis and Glark, 6 
was named Mus cinereus by Ord c in 1815, under the assumption that 
the animal was Congeneric with the rats of the Old World. In 1818 
Ord sent a short description and a figure of a wood rat from eastern 
Florida to the Philomatique Society of Paris. The description was 
published in the bulletin of the society for December of the same 
year under the name Mus jioridanus. De Blainville, who prepared 
the account for publication, questioned the applicability of the ge- 
neric name Mus, and in 1825 Say and Ord, d who meanwhile had dis- 
covered the peculiar dental characters of the species, published a 
diagnosis of the genus Neotoma, with Mus jioridanus as type. 

The bushy-tailed wood rats were separated from the round-tailed 
species as the genus Teonoma by Gray* in 1843, but this name has 

«Kalm's Travels (English edition), II, pp. 47-48, 1749. 

b Lewis and Clark (Paul Allen edition), I, pp. 289-290, 1814. 

c Guthrie's Geog., 2d Amer. ed., II, p. 292, 1815. 

dJourn. Acad. Nat. Sci. Phila., IV, pt. 2, pp. 345-346, 1825. 

eList Spec. Mamm., British Museum, p. 117, 1843. 



8 NORTH AMERICAN FAUNA. [no. 31. 

been regarded by most later authors as of not more than subgeneric 
rank. Previous to 1894 seventeen species and subspecies were de- 
scribed by various authors, but of their relationships little was known. 
Preliminary revision of the genus was begun by Merriam in 1894. 
The first of two papers published by him during the year contained 
an abstract of a study of the group, and added 14 new names; the 
second a diagnosis of the subfamily 'Neotominse, and a synopsis of 
the then known members of the genus Neotoma. b As a result of the 
recent exploration of many previously little -known regions, many 
new species and subspecies have been discovered, and since 1894 the 
total number of known forms has more than doubled. The large 
collections now available have rendered it possible to determine the 
validity of names and the status and relationships of the species. 

The present revision is mainly the result of a study of the wood 
rats in the collection of the Biological Survey, numbering over 3,000 
specimens, including topotypes of most of the species. These were 
supplemented by those in the United States National Museum, the 
private collection of Dr. C. Hart Merriam, and a few loaned or submit- 
ted for identification by the American Museum of Natural History, 
the Philadelphia Academy of Sciences, the Field Museum of Natural 
History, and the private collections of E. P. Warren, Joseph Grin- 
nell, and F. Stephens, making a total of over 4,000 specimens. I am 
especially indebted to Dr. C. Hart Merriam for the privilege of using 
his private collection. Acknowledgments are due also for the use of 
the collection of the United States National Museum; to Dr. J. A. 
Allen, American Museum of Natural History, to Witmer Stone and 
James A. G. Rehn, Academy of Natural Sciences, Philadelphia, to 
Dr. D. G. Elliot, Field Museum of Natural History, and to Prof. A. 
L. Herrera, City of Mexico, for the loan of types and topotypes, and 
to Oldfield Thomas and others. 

DISTRIBUTION. 

The genus Neotoma is restricted to North America. It reaches 
from ocean to ocean in the latitude of the lower Mississippi, but the 
species are most numerous along the backbone of the continent from 
Nicaragua and Guatemala northward through Mexico and the south- 
western United States. The bushy-tailed wood rats (subgenus Teo- 
noma) are confined chiefly to the boreal zones in the Sierra Nevada 
and Rocky Mountain regions, and reach their northern limit near lat- 
itude 60°. No representative of the genus is known from the Hud- 
son Bay region or the upper part of the Mississippi Valley. Species 
and subspecies are usually limited to one or two life zones, but groups 

aProc. Biol. Soc. Wash., IX, pp. 117-128, July 2, 1894. 
&Proc. Acad. Nat. Sci. Phila., pp. 225-254, Sept. 24, 1894. 



1010.] 



DISTRIBUTION. 



comprising closely related forms may extend through several zones. 
The vertical range of the members of the mexicana group in southern 
Mexico includes every zone from the tropical coastal plains to the 




Fig. 1.— Distribution of the genus Neotoma. 



lower edge of the Alpine Zone on the high volcanoes. On the south- 
west slope of Mount Orizaba Neotoma torquata ranges to timberline 
at 13,800 feet, an altitude attained by few mammals in North 
America. 



10 NORTH AMERICAN FAUNA. [no. 31. 

HABITS AND ECONOMIC STATUS. 

The habits of all species of wood rats are in general very similar, 
but they differ in many details even in the same species according to 
the varying conditions of local environment. As a rule the members 
of the genus prefer rocky or mountainous areas, where they live com- 
monly in cliffs or caves. Cliffs are especially favored because of the 
protection from enemies afforded by the deep crevices and over- 
hanging shelves. Some species, however, live on level plains or 
brushy hillsides in large conical nests composed chiefly of sticks ; but 
thorny vegetation, bits of cactus, bones, stones, leaves, and almost 
anything else they can carry enter into the construction of their 
homes. The habit of building nests of sticks and of accumulating 
more or less such material about the entrances to their burrows, even 
when in rocky places, is common to most of the species. Many bush- 
els of trash are often piled against a rock or the trunk of a tree or in 
a small cave. These nests, or burrows, have from one to half a dozen 
or more entrances to chambers, both above and below the surface 
of the ground. More or less well-defined runways usually radiate in 
several directions from the entrances into the surrounding vegeta- 
tion or may connect nests many yards apart. Occupied nests may 
be known at a glance by their well-kept appearance. Slight addi- 
tions and repairs are made frequently, and the runways are cleared 
of sticks and leaves. Sure signs of occupation are a few freshly cut 
twigs or leaves laid on or stuck into the upper walls. On desert 
plains a thick clump of cactus or other thorny vegetation is frequently 
chosen as the nest site, and here pieces of cactus are the chief material 
used in construction. Often the entire nest is a bristling mass of 
thorns, and as a further protection some especially spiny sections are 
placed about the openings and along the smoothly worn runways. 
When it is remembered that many of the spines have barbed points 
sharper than needles, which enter the flesh at the slightest touch, it 
is difficult to understand how the builders transport such material 
or are themselves able to reach their dwellings without being pierced. 

Wood rats are expert climbers, and some species, especially mem- 
bers of the fuscipes group, in addition to building surface nests 3 
to 5 feet in height, often place dwellings 20 feet or more from the 
ground among the upper branches of trees. None of the species are 
known to enter water voluntarily, but in Mexico, near the borders of 
lagoons, at least one of the species occasionally builds nests in the 
tops of mangroves, from which a single well-worn route always leads 
through the thick branches out to feeding grounds on the shore, per- 
haps 50 or 75 yards away. Wood rats do not frequent towns, but 
often live in the vicinity of farmhouses, and have been known to 
carry off spoons, knives, forks, pieces of cloth, and many other small 
articles, and add them to the general mass of nest material. 



1910.] GENERAL CHARACTERS. 11 

Wood rats are chiefly nocturnal in habits, but some are partly 
diurnal. Their food is largely determined by varying local conditions, 
but consists mainly of a great variety of green vegetation, including 
grass, leaves, fresh fruit, small bulbs, bark, and cactus stems. Dry 
seeds, nuts, and fungi are also eaten. During successive seasons of 
drought in the Rio Grande Valley, when ordinary food is scarce and 
in consequence most small mammals are greatly reduced in numbers, 
wood rats maintain their usual abundance by recourse to the large, 
soft, juicy cactuses. 

Wood rats have numerous deadly enemies such as owls, hawks, 
snakes, wild cats, civet cats, coyotes, foxes, and probably weasels, 
which serve to keep their numbers in check. Some of the desert 
species are sufficiently numerous to inflict appreciable damage on 
growing crops in fields and gardens and to carry off considerable 
grain stored on farms, but they have not thus far proven as injurious 
as some other rodents. In the arid regions of the Southwest they 
girdle and kill many native shrubs and severely injure cactuses, 
especially during the long dry season when other food is scarce. When 
they attack flat-jointed cactuses they eat the pulp out from one edge 
until only thorny semicircular fragments of the rims are left. Some 
of the smaller, columnar species of Echinocereus have large holes 
gnawed in their bases below or near the surface of the ground, and 
the thorny protection of their exposed upper parts is thus avoided. 
These holes usually result in the death of the plants. Extensive 
injuries to the larger growing plants appear to be connected with local 
fluctuations in numbers of wood rats. Some species of Neotoma, 
especially N. albigula and N. micropus, may sometimes become 
exceedingly abundant in local colonies marked by hundreds of nests 
which may later be found abandoned, and in the vicinity almost all 
shrubs, including such common species as Fouquiera splendent and 
Rhus microphylla, are partly or entirely girdled and killed, the 
injuries all old and evidently dating from the time the nests were 
occupied. 

On the table-land of Mexico wood rats of the albigula group are 
regularly hunted for their flesh, and considerable numbers are sold 
in the markets of San Luis Potosi and other cities to certain classes 
of the native population. 

GENERAL CHARACTERS. 

Externally the wood rats as a group rather closely resemble and 
are often confounded with the Old World rats of the genus Mus, 
which were early introduced into America and now almost universally 
infest the structures of man. In general coloration of upperparts 
the range of variation in the genus Neotoma is from pale shades of 
buff or gray to rich orange buff or ferruginous, always more or less 



12 NORTH AMERICAN FAUNA. [no. 31. 

mixed with black. The underparts are mainly white or pale buff. 
Desert forms are usually paler than those inhabiting more humid or 
heavily forested areas. In the subgenus Teonoma the tail is flattened 
and bushy, like that of a squirrel; in the other subgenera it is round 
and rather scantily haired. 

The molting season is somewhat irregular, especially in the southern 
part of the range of the group. The northern species molt once a 
year, toward the end of summer or fall. The southern forms usually 
molt in early winter, but individuals in worn and in fresh pelage may 
often be seen together. The new coat is generally grayer and brighter 
than the old one, and becomes duller and browner in advanced stages 
of wear. Very old individuals are apt to become rusty brownish in 
worn pelage, and on this account to differ considerably in appearance 
from younger ones of breeding age. 

In tracing the relationships of the species, color and size are of 
much less importance than cranial modifications. Besides the gen- 
eral form of the skull, the chief characters of taxonomic value are 
the following: Comparative size of anterior and posterior molars; 
depth and arrangement of reentrant angles in molar crowns; length 
of molariform toothrows; length and posterior form of palatal bridge; 
form of frontals, nasals, and interpterygoid fossa; and size and form 
of audital bulla?. Every large series of skulls from any given locality 
shows considerable individual variation, but always within definite 
limits. The differences separating species are well marked and 
constant. 

EXPLANATION OF MEASUREMENTS. 

All measurements, unless otherwise stated, are in millimeters, taken 
as follows: 

Total length (measured before skinning). — Nose to tip of tail verte- 
bra?, the body extended. 

Tail vertehrx (measured before skinning). — Base of upper side of 
tail when bent at right angles to body, to end of tail vertebra 3 . 

Hind foot (measured before skinning).- — Heel to end of longest claw. 

Basilar length. — Basion to either incisor at posterior alveolar border. 
- Zygomatic breadth. — Greatest breadth across zygomata. 

Interorbital breadth. — Breadth of interorbital constriction. 

Length of nasals. — Greatest length of nasals. 

Length of incisive foramina. — Greatest length of large palatal 
foramina. 

Length of palatal bridge. — Excavated posterior border of palate to 
posterior end of either incisive foramen. 

Alveolar length of upper molar series. — Greatest length of maxillary 
toothrow at alveolar border. 



1010.] SUBFAMILY NEOTOMINiE— NEOTOMA. 13 

Subfamily NEOTOMINffl Merriam.a 

Characters. — Braincase narrowing gradually anteriorly. Molars 
rooted or semirooted; the crowns prismatic and flat, enamel folds 
not closely crowded; reentrant angles very long, extending below 
alveolar border except in the adult state. Outer walls of antorbital 
vacuities without spines; palate excavated mesially between pos- 
terior molars. 

Remarks. — The subfamily Neotominse as originally restricted 
includes the living genera Neotoma, Xenomys, Hodomys, Nelsonia, 
and Teanopus, and the extinct genera Tretomys and Ptyssophorus. 
The genus Neotomodon, which has been included here by some recent 
authors, is clearly too aberrant and belongs nearer the Cricetinx. 
From the Neotominse it differs chiefly in having the molar crowns half 
tuberculate in early life; in the shortness of the reentrant angles, 
which even in quite young individuals do not reach the alveoli; and 
in the extension of the palatal bridge to the posterior plane of last 
molars. On the other hand, Neotomodon differs in important respects 
from all the other Cricetinse and may be regarded as a somewhat 
annectent genus. 

Genus NEOTOMA Say and Ord. 

Neotoma Say and Ord, Journ. Acad. Nat. Sci. Phila., IV, pt. 2, pp. 345-349, pis. 
XXI-XXII, 1825. Type Mus fioridanus Ord. 

Generic characters. — Molar crowns flat; first and second upper 
molars with middle enamel loops undivided; third lower molar with 
two transverse enamel loops (not S-shaped as in all the other genera 
of the subfamiry); bullae oblique, tapering anteriorly. 

SUBGENERA AND MINOR GROUPS. 

The genus Neotoma is here divided into three subgenera: Neotoma, 
Homodontomys, and Teonoma, under which 70 species and subspecies 
are recognized. The subgenus Homodontomys is characterized for the 
first time. 

The subgenus Neotoma, containing 56 species and subspecies, is 
readily divisible into six rather well-marked, yet closely related* 
groups. The subgenera Homodontomys and Teonoma each include a 
single group of very closely related forms, of which six belong to the 
former and eight to the latter. The arrangement of the list of spe- 
cies and subspecies (see pp. 14-15) with their type localities will in a 
measure show the affinities of the various forms, and to a less extent 
those of the groups to which they belong. 

a Neotominse Merriam, Proc. Acad. Nat. Sci. Phila., p. 228, Sept. 24, 1894. 



14 NORTH AMERICAN FAUNA. [no. 31. 

List of Species and Subspecies, with Type Localities. 

Subgenus NEOTOMA. 

Florida na group: 

Neotomafloridana (Orel) St. Johns River, Fla. 

floridana rubida Bangs Gibson, Terrebonne Parish, La. 

Jloridana illinoensis Howell Wolf Lake, 111. 

floridana baileyi Merriam Valentine, Cherry County, Nebr. 

floridana attivateri Mearns Turtle Creek, Kerr County, Tex. 

micropus Baird Charco Escondido, Tamaulipas, Mexico. 

micropus canescens Allen North Beaver River, Okla. 

micropus littoralis Goldman Alta Mira, Tamaulipas, Mexico. 

micropus planiceps Goldman. . . .Rio Verde, San Luis Potosi, Mexico. 
Albigula group: 

Neotoma albigula Hartley Near Fort Lowell, Pima County, Ariz. 

albigula venusta True Carriso Creek, San Diego County, Cal. 

albigula warreni Merriam Gaume's Ranch, Baca County, Colo. 

albigula melanura Merriam Ortiz, Sonora, Mexico. 

albigula leucodon Merriam San Luis Potosi, San Luis Potosi, Mexico. 

albigula durangx Allen San Gabriel, Durango, Mexico. 

albigula zacatecse Goldman Plateado, Zacatecas, Mexico. 

latifrons Merriam Querendaro, Michoacan, Mexico. 

nelsoni Goldman Perote, Veracruz, Mexico. 

palatina Goldman Bolaiios, Jalisco, Mexico. 

montezumse Goldman Zimapan, Hidalgo, Mexico. 

Intermedia group: 

Neotoma intermedia Rhoads Dulzura, San Diego County, Cal. 

intermedia gilva Rhoads Banning, Riverside County, Cal. 

intermedia pretiosa Goldman Matancita, Lower California, Mexico. 

intermedia arenacea Allen. San Jose del Cabo, Lower California, Mexico. 

intermedia vicina Goldman Espiritu Santo Island, Lower California, 

Mexico. 

intermedia perpallida Goldman. .San Jose Island, Lower California, Mexico. 

abbreviata Goldman San Francisco Island, Lower California, 

Mexico. 

nudicauda Goldman Carmen Island, Lower California, Mexico. 

bryanti Merriam Cedros Island, Lower California, Mexico. 

anthonyi Allen Todos Santos Island, Lower California, 

Mexico. 

martinensis Goldman San Martin Island, Lower California, Mexico. 

Mexicana group: 

Neotoma mexicana Baird Near Chihuahua, Chihuahua, Mexico. 

mexicana fallax Merriam Gold Hill, Boulder County, Colo. 

mexicana pinetorum Merriam. . . .San Francisco Mountain, Ariz. 

mexicana bullata Merriam Santa Catalina Mountains, Ariz. 

mexicana madrensis Goldman Near Guadalupe y Calvo, Chihuahua, 

Mexico. 

mexicana sinalose Allen Tatameles, Sinaloa, Mexico. 

navus Merriam Sierra Guadalupe, Coahuila, Mexico. 

torquata Ward Between Tetela del Volcan and Zacualpam 

Amilpas, Morelos, Mexico. 

distincta Bangs Texolo, Veracruz, Mexico. 

tropicalis Goldman Totontepec, Oaxaca, Mexico. 



1910.] LIST OF SPECIES AND SUBSPECIES KEY. 15 

M< xicana group — Continued . 

Xeotoma parvidens Goldman Juquila, Oaxaca, Mexico. 

ferruginea Tomes Duenas, Guatemala. 

ferruginea chamula Goldman. . . .Near San Cristobal, Chiapas, Mexico. 

ferruginea solitaria Goldman Nenton, Guatemala. 

ferruginea isthmica Goldman Iluilotepec, Oaxaca, Mexico. 

ferruginea picta Goldman Near Chilpancingo, Guerrero, Mexico. 

ferruginea tenuicauda Merriam.. .Sierra Nevada de Colima, Jalisco, Mexico; 

ferruginea ochracea Goldman Atemajac, Oaxaca, Mexico. 

chrysomelas Allen '. .Matagalpa, Nicaragua. 

Desertorum group: 

Neotoma desertorum Merriam Furnace Creek, Death Valley, Cal. 

lepida Thomas Type locality unknown. 

lepida slephensi Goldman ITualpai Mountains, Ariz. 

goldmani Merriam Saltillo, Coahuila, Mexico. 

Pennsylvanica group: 

Neotoma magister Baird Bone caves near Carlisle, Pa. 

pennsylvanica Stone South Mountain, Cumberland County, Pa. 

Subgenus HOMODONTOMYS. 

Fuscipes group: 

Neotoma fuscipes Baird Petaluma, Sonoma County, Cal. 

fuscipes streatori Merriam Carbondale, Amador County, Cal. 

fuscipes annectens Elliot Portola, San Mateo County, Cal. 

fuscipes simplex True Fort Tejon, Cal. 

fuscipes mohavensis Elliot Oro Grande, Mohave Desert, Cal. 

fuscipes macrotis Thomas San Diego, San Diego County, Cal. 

Subgenus TEONOMA. 

Cinerea group : 

Neotoma cinerea (Ord) Great Falls, Cascade County, Mont. 

cinerea drummondi (Richardson).'' Rocky Mountains in latitude 57°," Al- 
berta, Canada. 

cinerea saxamans Osgood Bennett, British Columbia, Canada. 

cinerea occidentalis Baird Shoalwater Bay, Pacific County, Wash. 

cinerea fusca True Fort Umpqua, Douglas County, Oreg. 

cinerea orolestes Merriam Saguache Valley, Saguache County, Colo. 

cinerea arizonse Merriam Keam Canyon, Apache County, Ariz. 

cinerea rupicola Allen Corral Draw, southeastern base of Black 

Hills, S. Dak. 

Key to Subgenera. 
a. Tail terete, not bushy. 

b. Maxillary tooth row much narrower posteriorly than anteriorly; middle lobe of 

last upper molar not divided by inner reentrant angle Neotoma (p. 20) 

b'. Maxillary tooth row slightly narrower posteriorly than anteriorly; middle 
lobe of last upper molar partially or completely divided by inner reentrant 

angle Homodontomys (p. 86) 

of , Tail flattened and bushy Teonoma (p. 94) 



16 NORTH AMERICAN FAUNA. [xo. 31. 

Key to Species and Subspecies. 

[Based on typical adults.] 

I. Subgenus NEOTOMA. 

a. First upper molar with anterointernal reentrant angle shallow or absent, reaching 
not more than half way across anterior lobe. 
b. Smaller (total length less than 320, or tail slightly bushy). 
c. Pelage long, soft, and silky (mainland forms). 
d. Tail normally haired, sharply bicolor. 
e. Bulla? larger. (West and north of Colorado River. )...N. desertorum (p. 76) 

ef . Bulla? smaller. (Northeastern Mexico.) N. goldmani (p. 81) 

d' '. Tail slightly bushy, nearly unicolor. 
e. Color paler, more yellowish. (Northeastern Arizona; northwestern New 

Mexico.) N. lepida (p. 79) 

ef. Color darker, more brownish. (Central Arizona; central-western and 

southwestern New Mexico.) N. I. stephensi (p. 80) 

cf. Pelage short and coarse. (San Francisco Island, Gulf of California.) 

N. abbreviata (p. 49) 
b' . Larger (total length more than 320). 
c. Interpterygoid fossa wider (about 4 or more), broadly excavated near posterior 
plane of molars; bulla? relatively smaller. 
d. Palate concave or emarginate posteriorly. 
e. Tail unicolor or nearly so. 
/. Color duller, more cinnamon. (Eastern Florida; Georgia; South Caro- 
lina.) N.floridana (p. 21) 

/ / . Color brighter, more ochraceous. (Louisiana; Mississippi; Alabama; 

eastern Texas; southern Arkansas.) N.f. rubida (p. 22) 

ef. Tail sharply bicolor. 
/. Tail less than 175. 

g. Maxillary tooth row averaging more than 9. (Nebraska; Colorado; 

Kansas.) N.f. baileyi (p. 24) 

g' . Maxillary tooth row averaging less than 9. (Texas; Oklahoma.) 

N.f. attwateri (p. 26) 

/ / . Tail more than 175. (Illinois.) N.f. illinoensis (p. 23) 

d' '. Palate with posterior median projection, or color slaty grayish. 
e. Color slaty grayish. 
/. Frontals upturned along sides. 

g. Larger (hind foot averaging about 40); color darker. (Lower Rio 
Grande Valley and north through central Texas to southern Kan- 
sas.) N. micropus (p. 26) 

g' . Smaller (hind foot averaging about 37); color paler. (New Mexico; 
western Texas; western Oklahoma; south to Jaral, Coahuila.) 

N. in. canescens (p. 28) 
J'. Frontals not upturned along sides. (San Luis Potosi.) 

N. m. planiceps (p. 30) 

ef. Color brownish. (Southern Tamaulipas.) N. m. littoralis (p. 29) 

cf . Interpterygoid fossa narrower (about 3.2 or less), not broadly excavated near 
posterior plane of molars; bulla? relatively larger. 
d. Rostrum shorter, heavier. (Mohave Desert, Colorado River Valley, and 
eastward to Texas; south over plateau region of Mexico.) 
e. Sphenopalatine vacuities present. 
/. Frontals moderately broad posteriorly, the sides projecting only slightly. 



1910.] KEY TO SPECIES AND SUBSPECIES. 17 

g. Tail sharply bicolor. 
h. Fur on throat and chest pure white to roots. 
i. Smaller (hind foot averaging about 32). (Arizona, except western 
part; New Mexico, except northeastern part; western Texas; 
south through eastern Chihuahua and western Coahuila to 

Jimulco.) N. albigula (p. 31) 

i . Larger (hind foot averaging more than 32). 
j. Paler. (Ranges in United States, except lower Colorado River 
Valley.) 
k. Upper parts pinkish buffy. (Lower Colorado River Valley; 

Mohave Desert.) N. a. venusta (p. 33) 

¥ . Upper parts vinaceous buffy. (Southeastern Colorado; 

northeastern New Mexico.) N. a. warreni (p. 34) 

y . Darker. (Ranges wholly in Mexico.) 

k. Frontals nearly flat above, lateral margins not decidedly up- 
turned nor convex posteriorly. (Western Durango; south- 
ern Chihuahua.) N. a. durangx (p. 37) 

¥ . Frontalswith lateral margins decidedly upturned and convex 
posteriorly. 
I. Paler. (High plateau region from Hidalgo to northwestern 

Nuevo Leon N. a. leucodon (p. 36) 

V . Darker. (Mountains of western Zacatecas.) 

N. a. zacatecse (p. 38) 
h / . Fur on throat and chest more or less plumbeous basally. 

i. Paler. (Sonora.) N. a. melanura (p. 35) 

V. Darker. (Zimapan, Hidalgo.) N. montezumse (p. 41) 

g / . Tail unicolor. (Perote, Veracruz.) N. nelsoni (p. 39) 

/'. Frontals very broad posteriorly, the sides strongly projecting as supra- 
orbital shelves. (Querendaro, Michoacan.) N. latifrons (p. 38) 

e / . Sphenopalatine vacuities absent. (Bolanos, Jalisco.) .N. palatina (p. 40) 
df . Rostrum longer, more slender. (Pacific coast region west of Mohave Des- 
ert and Colorado River Valley, from near Monterey, Cal., to Cape San 
Lucas, including islands along coasts of Lower California.) 
e. Outer sides of hind legs not conspicuously blackish. 
/. Hind foot not more than 37. 

g. Hind foot averaging less than 35. (Mainland forms.) 
h. Upper parts darker, more grayish. (Pacific coast region, except 
most arid portions, from near Monterey, Cal., to Sierra de la 

Laguna, Lower California.) N. intermedia (p. 42) 

¥. Upper parts paler, more buffy. (Desert regions from southwestern 

California to central Lower California.) N. i. gilva (p. 44) 

g' . Hind foot averaging more than 35. (Insular forms.) 
h. Darker. (Espiritu Santo Island, Lower California.) 

N. i. vicina (p. 48) 
h / . Paler. (San Jose Island, Lower California.). V. i. perpallida (p. 48) 
/'. Hind foot more than 37. 

g. Frontals less abruptly broadening between lachrymals. 
h. Tail averaging less than 180; bullae large. 
i. Interpterygoid fossa narrower. (Carmen Island, Lower Califor- 
nia.) N. nudicauda (p. 51) 

i' . Interpterygoid fossa wider. (Margarita Island, Magdalena Island, 
and adjacent coast of Lower California.)... N. i. pretiosa (p. 46) 

52668°— 10 2 



18 NORTH AMERICAN FAUNA. [no. 31. 

h'. Tail averaging more than 180; bullae small. (Cape region of Lower 

California, except high mountains.) N. i. arenacea (p. 47) 

g' '. Frontals more abruptly broadening between lachrymals. (Cedros 

Island, Lower California.) TV. bryanti (p. 51) 

e / . Outer sides of hind legs conspicuously blackish. 
/. Nasals reaching posteriorly beyond plane of lachrymals. (San Martin 

Island, Lower California.) N. martinensis (p. 53) 

/'. Nasals not reaching posteriorly to plane of lachrymals. (Todos Santos 

Island, Lower California.) N. anthonyi (p. 52) 

a'. First upper molar with anterointernal reentrant angle deep, reaching more than 
half way across anterior lobe. 
b. Smaller (hind foot 40 or less). 
c. Ground color above plainer, mainly grayish. (North of latitude 22° 30'.) 
d. Frontals without convex projecting supraorbital shelves. 
e. Bullae larger. 
/. Bullae more rounded. 

g. Averaging smaller. (Mexico; western Texas; southern New Mexico; 
southern Arizona.) 
h. Paler. (Eastern basal slopes of Sierra Madre and desert ranges from 

northern Durango to southern Arizona.) TV. mexicana (p. 54) 

hf . Darker. (Higher parts of Sierra Madre from northern Chihuahua 

and Sonora to Zacatecas.) JV. m. madrensis (p. 60) 

g / . Averaging larger. (Ranges wholly in United States.) 
h. Paler. (Colorado; New Mexico, except western part.) 

JV. m.fallax (p. 56) 

h' '. Darker. (Arizona; western New Mexico). . JV. m. pinetorum (p. 58) 

/ / . Bullae more elongated, narrower anteriorly. (Santa Catalina Mountains, 

Arizona.) JV. m. bullata (p. 59) 

e' '. Bullae smaller. (Western basal slopes of Sierra Madre from southern Sina- 

loa to Sonora.) JV. m. sinalose (p. 60) 

d' . Frontals with convex projecting supraorbital shelves. (Coahuila.) 

JV. navus (p. 61) 
& '. Ground color above richer, mainly bright, ochraceous buff y, or rufous. (South 
of latitude 22° 30'.) 
d. Hind foot less than 40. 

e. Color of sides not encroaching on underparts. 
f. Upper parts orange buffy or rufous. 
g. Upper parts orange buffy. 
h. Nasals truncate or bluntly pointed posteriorly. 
i. Hind foot more than 32. 
j. Nasals moderately broad posteriorly. 
k. Hind foot 36 or more. 

I. Zygomata decidedly narrower anteriorly than posteriorly. 

(Nicaragua.) JV. chrysomelas (p. 74) 

V . Zygomata not decidedly narrower anteriorly than poste- 
riorly. (Arid coastal plains and interior valleys of 
Oaxaca and west-central Chiapas.). JV./. isthmica (p. 71) 
k'. Hind foot less than 36. (Sierra Madre in Guerrero and 

Oaxaca.) N.f. picta (p. 72) 

j'. Nasals attenuate posteriorly. (N en ton, Guatemala.) 

N.f. solitaria (p. 70) 

i' '. Hind foot 32 or less. (Juquila, Oaxaca.) TV. parvidens (p. 66) 

h / . Nasals deeply emarginate posteriorly. (Duenas, Guatemala.) 

N.ferruginea (p. 67) 



1910.] KEY TO SPECIES AND SUBSPECIES. 19 

g / . Upper parts cinnamon rufous. 
h. Larger (hind foot 37 or more). (Mountains of central Chiapas and 

southwestern Guatemala.) N. f. chamula (p. 69) 

h'. Smaller (hind foot less than 37). (Colima; Jalisco; Michoacan.) 

N.f. tenuicauda (p. 73) 
/'. Upper parts ochraceous buffy. 

g. Paler. (Atemajac, Jalisco.) N.f. ochracea (p. 74) 

g' . Darker. (High mountains of Mexico [state]; Morelos; Hidalgo; 

Puebla; Tlascala.) N. torquata (p. 63) 

e / . Color of sides encroaching on underparts. (Toton tepee, Oaxaca.) 

N. tropicalis (p. 65) 

d / . Hind foot about 40. (Texolo, Veracruz.) N. distincta (p. 64) 

b'. Larger (hind foot more than 40). (Appalachian Mountain region.) 

c. Mandibular tooth row less than 9.6 N. pennsylvanica (p. 84) 

</. Mandibular tooth row 9.6 or more (extinct) N. magister (p. 82) 

H. Subgenus HOMODONTOMYS. 

a . Larger (hind foot usually more than 40). 

b. Bullae smaller; palate concave or emarginate posteriorly. (Coast region north of 

San Francisco.) N. fuscipes (p. 87) 

b'. Bullae larger; palate convex or with blunt posterior median projection. (Coast 

region immediately south of San Francisco.) N.f. annectens (p. 90) 

a! ' . Smaller (hind foot usually less than 40). 
b. Upper parts darker (mainly brownish). 
c. Hind foot clouded with dusky. (Coast region from Salinas, California, to north- 
ern Lower California.) N.f. macrotis (p. 93) 

</. Hind foot pure white. (West slope Sierra Nevada, California, from Tehama 

County to Milo, Tulare County.) N.f. streatori (p. 89) 

6'. Upper parts paler (mainly ochraceous buff or grayish), 
c. Upper parts ochraceous buff. (Southern part of Sierra Nevada.) 

N.f. simplex (p. 91) 
(/. Upper parts brownish gray. (Mohave Desert.) N.f. mohavensis (p. 92) 

m. Subgenus TEONOMA. 

a. Sphenopalatine vacuities absent or very small. 
b. Underside of tail not dusky, 
c. Nasals broad, truncate or blunt posteriorly. 
d. Longest hairs in tail usually less than 30. 
e. Upper parts paler, grayish buff to ochraceous buff. (Mountains of Mon- 
tana, Idaho, southeastern Wyoming, and of western Utah ; northern Ari- 
zona; central Nevada and east-central California.). . . . N. cinerea (p. 95) 
e / . Upper parts darker, the buff more obscured by dusky hairs; outer sides of 
ankles more dusky, contrasting more strongly with white of feet. 
(Pacific coast region west of range of cinerea.) . N. c. occidentalis (p. 101) 
d' '. Longest hairs in tail usually more than 30. (Rocky Mountains north of 

United States.) N. c. drummondi (p. 99) 

(f . Nasals attenuate and acutely pointed posteriorly N. c. saxamans (p. 100) 

V. Underside of tail dusky N. c. fusca (p. 103) 

a'. Sphenopalatine vacuities large and widely open. 
b. Upper parts ochracecfus buff, 
c. Larger (hind foot usually more than 40); tail more bushy; color darker. 
(Chiefly mountains in Wyoming, Colorado, and New Mexico.) 

N. c. orolestes (p. 104) 
&. Smaller (hind foot usually less than 40); tail less bushy; color paler. (Ari- 
zona; southern Utah; northwestern New Mexico.) N. c. arizonse (p. 106) 

V. Upper parts cream buff N. c. rupicola (p. 107) 



20 NOKTH AMERICAN FAUNA. [no. 31. 

Subgenus NEOTOMA Say and Ord. 
(Pis. I-VI; PI. VII, fig. 4.) 
Type. — Mus floridanus Ord, from eastern Florida. 

Distribution. — Southern half of North America, from the northern 
part of the United States, except the upper Mississippi Valley, to 
Nicaragua. 

Subgeneric characters. — Maxillary toothrow decidedly broader 
anteriorly than posteriorly; first upper molar with anterointemal 
reentrant angle varying from deep to shallow or obsolete; third 
upper molar with middle loop undivided by deepening of reentrant 
angles; interpterygoid fossa varying from wide to narrow; bullae 
variable in size; frontals constricted near or anterior to middle; tail 
terete, tapering, and short haired; hind foot naked below along outer 
side at least to tarsometatarsal joint. 

Remarks. — The subgenus Neotoma is composed of several minor 
groups, most of which are rather closely related, but not all are of 
equal rank. Neotoma pennsylvanica, whose peculiar characters 
have been pointed out by Merriam, presents a departure from the 
type species. It shows no tendency to subdivide and may be the 
survivor of a group now approaching extinction. Neotoma mexicana 
is a characteristic member of a highly plastic group occupying the 
backbone of the continent from northern Colorado to Nicaragua. 
All the members of this group agree with Hodomys and Xenomys 
in the retention of a long and deep anterointemal reentrant angle 
in the first upper molar. The primitive nature of this character is 
suggested by its presence in Tretomys, the earliest member of the 
subfamily whose upper molars are known. The small group typified 
by Neotoma lepida shows differentiation in the direction of Teanopus — 
perhaps a more modern offshoot, notably in the shortness of the 
anterointemal reentrant angle in the first upper molar (entirely 
absent through wear in adults), the length and slenderness of the 
angle of the mandible, and in the remarkable inflation of the audital 
bullae. Between these three somewhat divergent branches are the 
closely related albigula and intermedia groups, which are nearer the 
type species and tend to bridge the gaps. In the more essential 
characters all the groups agree so closely that further subgeneric 
division seems undesirable. 

oProc. Acad. Nat. Sci. Phila., pp. 240-241, Sept. 24, 1894. 



1910.] 



FLORID AN A GROUP FLORIDANA. 



21 



NEOTOMA FLORIDANA (Ord). 
Florida Wood Rat. 
(PI. I, figs. 1, la; PI. VII, fig. 4.) 

Mus floridanus Ord, Bull. Soc. Philom. Paris, pp. 181-182, December, 1818. 

N[eotoma] floridana Say and Ord, Journ. Acad. Nat. Sci. Phila., IV, pt. 2, p. 346, 
1825. Type from St. Johns River, Fla, probably near Jacksonville, Duval 
County. (See Bangs, Proc. Boston Soc. Nat. Hist., XXVIII, p. 184, Mar., 1898.) 
No type specimen designated. 

Distribution. — Atlantic coast region from South Carolina to Sebas- 
tian, Fla. Austroriparian Zone. 




A/SOrOMA FLOW DANA. 
QIIIIl " f- POB/DA. 

- F. BA/IM. 

- F. ATTWATFPt. 
F. IU/NOENS/S. 



Fig. 2. — Distribution of Neoto ma floridana and subspecies. 

General characters. — Size large; ears medium; tail long, nearly 
concolor, scantily haired; color dark. Related to N. micropus, but 
differing in color and cranial characters. 

Color. — Winter pelage : General color above pale cinnamon, bright- 
est along sides, becoming broccoli brown or brownish drab on head 
and outer sides of legs, much darkened over top of head and back by 
blackish hairs; feet white; under parts creamy white, the fur basally 
plumbeous along sides of belly; tail dusky above, slightly paler below. 



22 NORTH AMERICAN FAUNA. [no. 31. 

Skull. — Skull large and rather elongated; ascending branches of 
premaxillae very long, reaching posteriorly far beyond nasals ; f rontals 
broad, interorbital constriction anterior to middle; palatal bridge 
shorter than incisive foramina; sphenopalatine vacuities rather 
small; interpterygoid fossa very large, broadly excavated, rounded 
anteriorly; presphenoid deeply constricted; bullae small, short, and 
rounded; first upper molar with anterointernal reentrant angle 
moderately developed. Compared with N. micropus, the skull is 
more elongated; rostrum longer; nasals narrower posteriorly; palate 
without posterior median projection (present in micropus); bullae 
more rounded. 

Measurements. — Average of 3 adults from Enterprise, Fla. : Total 
length, 409; tail vertebrae, 189; hind foot, 38.5. Skull (average of 
same): Basilar length, 40.4; zygomatic breadth, 25.7; interorbital 
breadth, 6.4; length of nasals, 19.5; length of incisive foramina, 11; 
length of palatal bridge, 6.8 ; alveolar length of upper molar series, 8.9. 

Remarks. — Neotoma Jloridana is the type species of the genus. 
Aside from its subspecies it requires comparison with only N. micro- 
pus, from which it differs decidedly in color. No specimens from 
western Florida have been examined, but the essential differences 
separating typical N. Jloridana from A T . /. rubida are so slight that 
intergradation may be expected to occur in that general region. 

Specimens examined. — Total number 52, from localities as follows: 

Florida: Enterprise, 5; Fort Gardner, Kissimmee River, 4; Kissimmee, 1; 
Lake Harney, 15; Lake Kissimmee, 1; Miceo, 1; San Mateo, 1; Sebas- 
tian, 1. 

Georgia: Riceboro(Leconte Plantation), 17; St. Simon Island, 1; Savannah, 2. 

South. Carolina: Frogmore, 3. 

NEOTOMA FLORIDANA RUBIDA Bangs. 

Ruddy Wood Rat. 

(PI. I, figs. 2, 2a.) 

Neotoma Jloridana rubida Bangs, Proc. Boston Soc. Nat. Hist., XVIII, No. 7, pp. 185- 
186, March, 1898. Type from Gibson, Terrebonne Parish, La.; No. 2872, $ ad., 
collection of E. A. and 0. Bangs; collected by F. L. Small, April 4, 1895. 

Distribution. — Lower Mississippi Valley and Gulf coast, from 
southwestern Alabama to eastern Texas, north to eastern Arkansas. 
Austroriparian Zone. 

General characters. — Size averaging slightly larger than N . jloridana; 
color redder; differing also in slight cranial characters. 

Color. — Fresh pelage: Upper parts vary from cinnamon to dark 
ochraceous buff, purest along cheeks and sides, becoming brownish 
drab on middle of face and outer sides of legs, moderately darkened 
over top of head and back by dusky hairs ; feet white ; under parts 



1910.] FLORIDANA GROUP ILLINOENSIS. 23 

creamy white; tail nearly unicolor, blackish above, in some speci- 
mens slightly paler beloAv. 

Skull. — Closely resembling that of N.jloridana, but slightly larger; 
nasals broader posteriorly; incisive foramina usually shorter; palatal 
bridge averaging longer; presphenoid less constricted. 

Measurements. — Average of 6 adults from Houma, La.: Total 
length, 404; tail vertebrae, 195; hind foot, 41. Skull (average of 
same): Basilar length, 42.7; zygomatic breadth, 26.9; interorbital 
breadth, 6.8; length of nasals, 20.1; length of incisive foramina, 
10.1; length of palatal bridge, 8.9; alveolar length of upper molar 
series, 9.4. 

Remarks. — N. j. rubida is a slightly differentiated form, probably 
intergrading with jloridana in western Florida; with attwateri in east- 
ern Texas; and with baileyi in northeastern Texas and southern Ar- 
kansas. Its rich coloration readily separates it from the two latter 
subspecies. 

Specimens from Tallulah, in northeastern Louisiana, are less red- 
dish in color than the typical form from farther down the Mississippi, 
and their more distinctly bicolor tails suggest gradation toward 
illinoensis. 

Specimens examined. — Total number 87, from the following locali- 
ties: 

Louisiana: Franklin, 2; Houma (near type locality), 15; Tallulah, 3. 

Alabama: Castleberry, 1; Mobile Bay, 8. 

Arkansas: McGeb.ee, 3. 

Mississippi: Bay St. Louis, 20; Hancock County, 6; Washington, 6. 

Texas: Sour Lake, 22; Texarkana, 1. 

NEOTOMA FLORIDANA ILLINOENSIS Howell. 

Illinois Wood Rat. 

Neotoma Jloridana illinoensis Howell, Froc. Biol. Soc. Wash., XXII, pp. 28-29, March 
23, 1910. Type from Wolf Lake, Illinois; No. 167752, ? ad., U. S. National 
Museum (Biological Survey Collection) ; collected by John Johnson, January 12, 
1910. 

Distribution. — Swamp region of southern Illinois, and southward 
to northeastern Arkansas. Austroriparian division of Lower Austral 
Zone. 

General characters. — Most like N. f. rubida; size and proportions 
about the same; color grayer; tail bicolor instead of dark all round. 
In color rather closely resembling N.f. baileyi in worn summer pelage, 
decidedly darker than baileyi in winter coat; tail darker above and 
longer. 

Color. — Upper parts dull buffy mixed with black, the black hairs 
more numerous over top of head and back; face grayish; outer sides 
of hind legs brownish; under parts white, the fur pure white to roots 



24 NORTH AMERICAN FAUNA. [no. 31. 

except that it is plumbeous basally along flanks and in some speci- 
mens across belly; feet white; tail blackish above, dull white below. 

Cranial characters.— Skull most like that of N. f rubida in general 
form, but zygomata more squarely spreading anteriorly, the sides 
more nearly parallel; posterior border of palate emarginate, instead 
of evenly concave as in the other forms of the florid-ana group ; spheno- 
palatine vacuities reduced to very narrow slits. Compared with that 
of N. f. baileyi the skull has longer rostrum and differs otherwise in 
the same characters as from rubida. 

Measurements. — Average of 8 adult topotypes: Total length, 403 
(390-412) ; tail vertebras, 195 (187-203) ; hind foot, 38 (36-40). SJcull: 
Average of 5 adults: Basilar length, 40.8; zygomatic breadth, 25.3; 
interorbital breadth, 6.8; length of nasals, 19.9; length of incisive 
foramina, 10; length of palatal bridge, 8.7; alveolar length of upper 
molar series, 9.3. 

Remarks. — Although a well-marked subspecies, N. f. illinoensis 
seems more nearly related to N.f. rubida than to any other form, and 
the ranges of the two may meet on the swampy bottom lands border- 
ing the Mississippi River. Gradation of N.f. rubida toward illinoensis 
is suggested by less reddish color of upper parts and more distinctly 
bicolor tails of specimens from Tallulah, in northeastern Louisiana, 
and from McGehee, in southeastern Arkansas, when compared with 
typical rubida from farther down the Mississippi. In general color 
illinoensis approaches N.floridana but is distinguished by grayer face 
and more sharply bicolored tail. 

Specimens examined. — Total number 22, from the following locali- 
ties: 

Illinois: Wolf Lake (type locality), 15. 
Arkansas: Big Creek (Turrell P. O.), 7. 

NEOTOMA FLORIDANA BAILEYI Merriam. 

Bailey Wood Rat. 

(PI. I, figs. 3, 3a.) 

Neotoma baileyi Merriam, Proc. Biol. Soc. Wash., IX, p. 123, July 2, 1894. Type from 
Valentine, Cherry County, Nebr. ; No. £$}£, 9 ad., Merriam Collection; collected 
by Vernon Bailey, June 16, 1888. 

Neotoma campestris Allen, Bull. Amer. Mus. Nat. Hist., VI, pp. 322-323 [author's sepa- 
rates issued November 7], 1894. Type from Pendennis, Lane County, Kans.; 
No. |Jf|, $ ad., American Museum Natural History, New York; collected by 
W. W. Granger, May 8, 1894. 

Neotoma floridana baileyi Bailey, North American Fauna, No. 25, pp. 109-110, October 
24, 1905. 

Distribution. — Upper Sonoran and Carolinian divisions of Upper 
Austral Zone from southwestern South Dakota to southern Kansas, 
west to Pueblo, Colo. 



1910.] FLORIDANA GROUP BAILEYI. 25 

General characters. — Similar to A 7 , jloridana, but pelage longer and 
fuller; color grayer; tail shorter, longer haired, bicolored; cranial 
characters also different. In summer pelage closely resembling N. f. 
attwateri. 

Color. — Fresh winter pelage: Upper parts creamy buff, varying to 
buffy gray, clearest along sides, thinly overlaid dorsally with dusky 
hairs; feet and under parts white, the fur basally plumbeous along 
sides of belly; tail sharply bicolor, brownish gray above, white below. 
Worn pelage (spring, summer, and fall) : Above, varying shades of 
dark rusty brown. 

SJcull. — In general form similar to that of N. Jloridana, but nasals 
shorter, broader posteriorly; ascending branches of premaxillse 
shorter; zygomata more widely spreading posteriorly; incisive fo- 
ramina shorter; palatal bridge longer. Compared with N. f. rubida 
the rostrum is shorter; nasals shorter, narrowing more abruptly pos- 
teriorly. From N. f. attwateri the skull differs chiefly in heavier 
dentition. 

Measurements. — Average of 5 adult topotypes: Total length, 369; 
tail vertebrae, 160; hind foot, 39.7. Skull (average of same) : Basilar 
length, 40.4; zygomatic breadth, 26; interorbital breadth, 6.7; length 
of nasals, 18.7; length of incisive foramina, 9.2; length of palatal 
bridge, 8.6; alveolar length of upper molar series, 9.2. 

Remarks. — Color variations due to seasonal differences in pelage 
are perhaps greater in baileyi than in any other round-tailed member 
of the genus. The pale winter pelage contrasts strongly with the 
dark rusty brownish summer coat. Although differing remarkably 
in proportions, color, and cranial characters horn. Jloridana, complete 
intergradation through attwateri and rubida seems certain. In essen- 
tial characters baileyi and attwateri are much alike and in summer 
pelage often difficult to separate. Specimens from Oklahoma and 
Arkansas are intermediate. 

Topotypes of N. campestris and specimens from other localities in 
western Kansas and eastern Colorado are paler than typical baileyi 
and represent a slightly differing geographic race, which seems 
scarcely worthy of recognition by name. Specimens from eastern 
Kansas are somewhat darker than typical baileyi. 

Specimens examined. — Total number 76, from the following locali- 
ties: 

Nebraska: Cody (10 miles south), 5; Haigler, 1; Valentine (type locality), 12. 
Colorado: Olney, 12; Pueblo, 1; Tuttle, 2; Wray, 1. 

Kansas: Cedarvale,9; Fort Riley, 1; Hays, 7; Pendennis, 23; Trego County, 1. 
South Dakota: Spring Creek (18 miles southeast of Rapid City), 1. 



26 NORTH AMERICAN FAUNA. [no. 31. 

NEOTOMA FLORIDANA ATTWATERI Mearns. 
Attwater Wood Rat. 

Neotoma attwateri Mearns, Proc. U. S. Nat. Mus., XIX, pp. 721-723, July 30, 1897. 

Type from Lacey's Ranch, Turtle Creek, Kerr County, Tex.; No. HIM. ? ad -> 

American Museum Natural History, New York; collected by H. P. Attwater, 

December 10. 1895. 
Neotoma floridana attwateri Elliot, Syn. Mamm. North Amer., Zool. Ser., II, p. 157, 

1901. 

Distribution.— Mainly Lower Sonoran and Austroriparian divisions 
of Lower Austral Zone in central Texas, passing into N. f. baileyi in 
Oklahoma and northern Arkansas. 

General characters. — Same as N. f. baileyi, but winter pelage darker 
and teeth smaller. 

Color. — Winter pelage: General color of upper parts pale vinaceous 
buff, purest along sides, moderately overlaid with blackish; face and 
outer sides of legs grayish; feet and under parts white, the fur along 
sides of belly basally plumbeous; tail brownish black above, white 
below. Summer pelage: Varying above from ochraceous buff to dark 
rusty brown. 

Skull. — As in N. f. baileyi, but teeth smaller. 

Measurements. — Average of 2 adults from the type locality: Total 
length, 366; tail vertebrae, 167; hind foot, 39. Slcull (average of 
same): Basilar length, 39.7; zygomatic breadth, 25.4; interorbital 
breadth, 6.8; length of nasals, 19.3; length of incisive foramina, 9.5; 
length of palatal bridge, 8.1 ; alveolar length of upper molar series, 8.9. 

Remarks. — This form differs from baileyi mainly in color of winter 
pelage and size of teeth. Specimens from Oklahoma, northern 
Arkansas, and Missouri have the dark winter colors of attwateri, but 
approach baileyi in dentition. Intergradation with rubida probably 
occurs in eastern Texas and southern Arkansas. 

Specimens examined. — Total number 71, from the following locali- 
ties: 

Texas: Cooke County, 1; Gainesville, 1; Ingram, 8; Kountze,2; Lacey's Ranch, 
Kerr County (type locality), 11; Navasota, 3; Rock Springs, 2; Victoria, 1. 

Arkansas: Batesville, 2; Cotter, 2; Pettigrew, 2; Rich Mountain, 8; Womble 1. 

Missouri: Marble Cave, Stone County, 5. 

Oklahoma: Chattanooga, 1; Mount Scott, Wichita Mountains, 6; Ponca 
Agency, 1; Red Fork, 4; Red Oak, 1; Savanna, 1; Stillwell, 8. 

NEOTOMA MICROPUS Baird. 

Baird Wood Rat. 

(PI. I, figs. 4, 4a.) 

Neotoma micropus Baird, Proc. Acad. Nat. Sci. Phila., VII, p. 333, April, 1855. Type 
from Charco Escondido, Tamaulipas, Mexico; No. Wr (skin now lost), $ , U. S. 
National Museum; collected by Lieutenant Couch. 

o See Merriam, Proc. Acad. Nat. Sci. Phila., p. 244, Sept. 24, 1894. 



1910.] 



FLORIDANA GROUP MTCROPUS. 



27 



Neotoma macropus surberi Elliot, Pub. Field Columb. Mus., Zool. Ser., I, No. 14, pp. 
279-280, May 9, 1899. Type from canyon 3 miles west of Alva, Okla.; No. 6755, 
$ ad., Field Columb. Mus., Chicago; collected by Thaddeus Surber, February 20, 
1899. 

Distribution. — Southeastern Colorado and southern Kansas, south 
through Oklahoma and central Texas to southern Tamaulipas, 
mainly in Lower 
Sonoran Zone. 

General charac- 
ters. — Size large; 
fur short and rather 
harsh; color pale; 
tail rather short, 
thinly haired, bicol- 
ored. Related to 
N. floridana, but 
color paler; cranial 
characters dis- 
tinctive. 

Color . — Winter 
pelage: Ground col- 
or above pale ecru 
drab, purest along 
cheeks and sides, 
moderately ob- 
scured by overlying 
dusky hairs ; feet 
and under parts 
white, the fur white 
to roots on pectoral 
and inguinal re- 
gions; tail black- 
ish above, grayish 
below. 

Skull. — Simil ar 
in general form to 
that of N.jloridana, 
but more angular; 
rostrum heavier; 
nasals narrower 
posteriorly; outer 

sides of frontals *" IG ' 3-— Distribution of Neotoma micro-pus and subspecies. 

more upturned, forming prominent, slightly projecting supra-orbital 
ridges; zygomata more widely spreading posteriorly; interptery- 
goid fossa very broad, as in N.jloridana, but usually encroached upon 
anteriorly by median projection from palate (absent in floridana) . 




28 NORTH AMERICAN FAUNA. [no. 31. 

From N. in. littoralis and N. m. canescens the skull differs chiefly 
in larger size. Compared with that of N. m. planiceps the skull is 
larger, with frontal region arched and upturned along sides (nearly 
flat in planiceps) and nasals broader posteriorly. 

Measurements. — Average of 4 adults from Nueces Bay, Tex.: 
Total length, 351; tail vertebra 1 , 163; hind foot, 41. Skull (average 
of same) : Basilar length, 40.3; zygomatic breadth, 26.1; interorbital 
breadth, 5.9; length of nasals, 19; length of incisive foramina, 10.4; 
length of palatal bridge, 8.1 ; alveolar length of upper molar series, 9.4. 

Remarks. — The specimen from Santa Rosalia, Chihuahua, men- 
tioned by Baird in his original description of N. micropus, now a 
much faded skin without skull, is probably N. albigula, the species 
common in that locality. In northern and western Texas and eastern 
Coahuila this form passes into canescens and in extreme southern 
Tamaulipas into littoralis. Its larger size, slightly darker color, and 
shorter fur at all seasons serve to separate it from canescens. From 
littoralis it may be known externally by its grayer, less brownish 
color. Topotypes of N. m. surberi are inseparable from typical 
micropus. 

Specimens examined. — Total number 203, from the following 
localities : 

Tamaulipas: Bagdad, 2; Camargo, 11; Charco Escondido, 1 (type); Forlon, 1; 
Matamoros, 2; Mier, 5; Nuevo Laredo, 12; Soto la Marina, 7. 

Coahuila: Las Vacas, 1; Sabinas, 2; Saltillo, 1. 

Nuevo Leon: China (16 miles south), 1; Doctor Cos, 1; Linares, 1; Rodri- 
guez, 5. 

Texas: Alice, 4; Beeville, 4; Blocker Ranch, Dimmit County, 2; Brazos, 6; 
Brownsville, 11; Comstock, 2; Concho County, 1; Corpus Christi, 6; Co- 
tulla, 2; Del Rio, 11; Devils River (mouth), 11; Dos Hermanos, Webb 
County, 1; Eagle Pass, 18; Fort Clark, 5; Henrietta, 1; Laredo, 10; Las 
Moras Creek, Maverick County, 1; Newlin, 3; Nueces Bay, 6; Pinto Creek, 
Maverick County, 1; Rio Grande City, 2; Rockport, 4; Roma, 4; San 
Angelo, 4; San Antonio, 7; San Diego, 2; Santo Tomas, Webb County, 3; 
Sauz Ranch, Cameron County, 2; Sycamore Creek (mouth), 3; Tebo, Tay- 
lor County, 1; Vernon, 1. 

Oklahoma: Alva, G; Woodward, 3. 

Kansas: Sun, 1. 

Colorado: Monon, Baca Count}-, 1. 

NEOTOMA MICROPUS CANESCENS Allen. 

Hoary Wood Rat. 

Neotoma micropus canescens Allen, Bull. Amer. Mus. Nat. Hist., III., pp. 285-287, 
June 30, 1891. Type from "North Beaver River, Indian Territory [now Okla- 
homa], near the boundary line between the Indian Territory and New Mexico;" 
No. f§|}, 9 , American Museum of Natural History, New York; collected by 
Richardson and Rowley, October, 1889. 

Distribution. — From southeastern Colorado, northwestern Okla- 
homa, and northern and western Texas, west in New Mexico to the 



1910.] FLORIDANA GROUP LITTORALIS. 29 

Rio Grande Valley and south to southern Coahuila, mainly in Lower 
Sonoran Zone. 

General characters. — Closely related to N. micropus, but averaging 
smaller; fur longer and softer; color still paler. 

Color. — Fresh winter pelage: Ground color above pale ashy gray, 
clearest along cheeks and sides, thinly overlaid with blackish; feet 
and under parts white, the fur pure white to roots on throat and on 
pectoral and inguinal regions; tail varying from grayish brown to 
blackish above, white below. 

Skull. — Averaging smaller than that of N. micropus, but not 
otherwise different. 

Measurements. — Average of 4 adults from Capitan Mountains, New 
Mexico: Total length, 330; tail vertebras, 137; hind foot, 36. Skull 
(average of same): Basilar length, 37.4; zygomatic breadth, 24.3; 
interorbital breadth, 5.8; length of nasals, 17.3; length of incisive 
foramina, 9.9; length of palatal bridge, 7.2; alveolar length of upper 
molar series, 8.3. 

Remarks. — N. m. canescens is closely related to N. micropus and by 
recent authors has not been considered separable; but the characters 
given, although rather slight, are so constant over the higher area 
west of the range of typical micropus that it seems to merit recogni- 
tion. Intergradation with micropus occurs in western Texas and 
eastern Coahuila. 

Specimens examined. — Total number 165, from the following 
localities: 

Oklahoma: North Beaver River, 3 (including type). 

Colorado: La Junta (18 miles south), 1. 

New Mexico: Ancho, 5; Capitan Mountains, 15; Carlsbad, 7; Corona, 3; 
Cuervo, 2; Deming (8 miles east), 1; Fort Sumner, 11; Jiearilla Moun- 
tains, 2; Magdalena (10 miles southeast), 1; Manzano Mountains, 1; Rin- 
conada, 2; Roswell, 3; Sandia Mountains, 1; Santa Rosa, 8; Socorro (10 
miles northeast), 2; Tularosa, 7; Tucumcari, 1. 

Texas: Adams, 6; Alpine, 4; Altuda, 12; Big Spring, 3; Chisos Mountains, 5; 
Colorado, 4; Dryden, 1; El Paso, 1; Fort Hancock, 1; Kent, 4; Lips- 
comb, 2; Lozier, 1; Marathon, 1; Miami, 1; Mobeetie, 1; Monahans, 22; 
Presidio County, 1; Samuels, 1; Sierra Blanca, 1; Stanton, 6; Toyah, 1; 
Toyahvale, 3; Valentine, 1. 

Coahuila: Jaral, 1; Monclova, 7. 

NEOTOMA MICROPUS LITTORALIS Goldman. 

Coast Wood Rat. 

Neotoma micropus littoralis Goldman, Proc. Biol. Soc. Wash., XVIII, pp. 31-32, Feb- 
ruary 2, 1905. Type from Alta Mira, Tamaulipas, Mexico; No. 92952, $ ad., 
U. S. National Museum (Biological Survey Collection); collected by E. A. Gold- 
man, April 10, 1898. 

Distribution. — Arid Tropical Zone in southern Tamaulipas, Mexico. 
General characters. — Similar to N. micropus, but differing in some- 



30 WORTH AMERICAN FAUNA. [no. 31. 

what smaller size, decidedly brownish color, and slight cranial 
characters. 

Color. — Slightly worn pelage: Upper parts nearly uniform grayish 
brown, moderately darkened on middle of face, top of head, and 
along back by blackish hairs; under parts white, the fur plumbeous 
basally except on throat and on pectoral and inguinal regions, where 
the hairs are pure white to roots; nose dusky; feet white; tail sharply 
bicolor, blackish above, whitish below. Young (half-grown) : Upper 
parts much browner than in N. micropus of same age. 

Skull. — In general form the skull agrees with that of N. micropus, 
but averages smaller; dentition usually less heavy; interpterygoid 
fossa narrower, encroached upon anteriorly by a short but more or 
less spinous projection from palate, as in N. micropus. 

Measurements. — Average of 3 adults from the type locality: Total 
length, 358; tail vertebrae, 167; hind foot, 38.2. Skull (type): 
Basilar length, 37.5; zygomatic breadth, 24.2; interorbital breadth, 
5.8; length of nasals, 17.4; length of incisive foramina, 10.2; length 
of palatal bridge, 7.6; alveolar length of upper molar series, 8.8. 

Remarks. — This subspecies is an Arid Tropical form of micropus, 
at all ages easily recognizable externally by its browner, less gray 
coloration. Along the coast north of Alta Mira its range probably 
merges into that of micropus. 

Specimens examined. — Five, all from the type locality. 

NEOTOMA MICROPUS PLANICEPS Goldman. 

Rio Verde Wood Rat. 

(PI. I, figs. 5, 5a.) 

Neotoma micropus planiceps Goldman, Proc. Biol. Soc. Wash., XVIII, p. 32, February 
2, 1905. Type from Rio Verde, San Luis Potosi, Mexico; No. 82105, $ ad., U. S. 
National Museum (Biological Survey Collection); collected by E. W. Nelson 
and E. A. Goldman, January 16, 1897. 

Distribution. — Plains of southern San Luis Potosi. Lower Sonoran 
Zone. 

General characters. — Size slightly smaller than N. micropus; color 
more buffy ; skull flatter and less angular. 

Color. — Fresh pelage: Upper parts pale buffy gray, somewhat 
obscured by dusky hairs, which are most abundant along median 
line of back; under parts white, the fur pale plumbeous basally 
along sides of belly and inner surface of hind legs; nose, eyelids, and 
ankles dusky; feet white; tail brownish black above, grayish below. 

Skull. — Similar in general to that of N. micropus, but smaller and 
less arched; frontals flatter above, the sides not upturned nor pro- 
jecting as supraorbital shelves; braincase more smoothly rounded, 
bulging posteriorly below lambdoid crest; nasals attenuate pos- 



1910.] ALBIGULA GROUP ALBIGULA. 31 

teriorly, the ends pointed and deeply emarginate ; ascending branches 
of premaxillse very long, reaching posteriorly beyond nasals nearly 
to interorbital constriction; interpterygoid fossa very broad, as in 
N. micropus. 

Measurements.— Type: Total length, 351; tail vertebrae, 167; 
hind foot, 38. Skull (type): Basilar length, 38; zygomatic breadth, 
23.5; interorbital breadth, 5.6; length of nasals, 17.1; length of 
incisive foramina, 9.5; length of palatal bridge, 9.1; alveolar length 
of upper molar series, 9. 

Remarks. — This apparently well-marked form inhabits the eastern 
terraces of the Mexican plateau region in southern San Luis Potosi. 
Intergradation with micropus is not satisfactorily shown by present 
material but no doubt occurs in southwestern Tamaulipas. 

Specimens examined. — One, the type. 

NEOTOMA ALBIGULA Hartley. 

White-throated Wood Rat. 

(PI. II, figs. 1, la.) 

Neotoma albigula Hartley, Proc. Calif. Acad. Sci., 2 Ser., IV, pp. 157-159, pi. XII 

May 9, 1894. Type from vicinity of Fort Lowell, Pima County, Ariz.; No. 

1336, 9 ad., Museum of Stanford University; collected by W. W. Price and 

R. L. Wilbur, June 14, 1893. 
Neotoma intermedia angusticeps Merriam, Proc. Biol. Soc. Wash., IX, 127, July 2 

1894. Type from southwestern corner of Grant County, N. Mex., 4 miles north 

of Mexican boundary; No. ffff, $ ad., Merriam Collection; collected by A. W. 

Anthony, April 12, 1886. 
Neotoma intermedia albigula Merriam, Proc. Acad. Nat. Sci. Phila., p. 248, September 

24, 1894. 

Distribution. — Northern New Mexico to southern Coahuila, Mexico, 
and from central Texas to western Arizona. Upper and Lower 
Sonoran zones. 

General characters. — Size medium ; throat and pectoral region pure 
white to roots of hairs; tail bicolor. Similar in general to N. inter- 
media, but differing in important characters. 

Color. — Fresh pelage: Upper parts dull pinkish buff, brightest along 
sides, varying in some specimens to clear pinkish buff, thinly overlaid 
with blackish ; outer sides of legs tinged with vinaceous ; under parts 
and feet white ; tail grayish brown above, white below. Worn pelage: 
Upper parts more or less ochraceous buffy or rusty. 

Skull. — Similar to that of N. intermedia, but larger; rostrum much 
heavier; palate concave posteriorly (convex in intermedia); bullae 
larger; first upper molar with anterointernal reentrant angle shallow, 
as in N. intermedia. 

Measurements. — Average of four adults from the region of the type 
locality: Total length, 328; tail vertebras, 152; hind foot, 33.5. 



32 



NOKTH AMERICAN FAUNA. 



[no. 31. 



Skull (average of same): Basilar length, 36.3; zygomatic breadth, 
22.0; interorbital breadth, 5.8; length of nasals, 16.5; length of 
incisive foramina, 8.8; length of palatal bridge, 7.8; alveolar length 
of upper molar series, 8. 

Remarks. — In extreme western Arizona N. albigula passes into a 
larger form, N. a. venusta. It intergrades in central Sonora, Mexico, 
with N. a. melanura, a subspecies ranging to the southward along the 




Fig. 4.— Distribution of the Neotoma albigula group. 

western basal slopes of the Sierra Madre. Specimens from Rio 
Puerco and San Rafael, X. Mex., are dichromatic. In addition to 
normally colored individuals are darker ones in winch the pinkish 
buff of sides spreads entirely across the belly. In specimens showing 
the dark phase the throat and chest are pure white, as in the normal 
individuals, and contrast strongly with the intensified color of the 
areas over which the fur is always basally dark. Externally albigula 
closely resembles N. m. fallax, and the two often range together. 



1910.] ALBIGULA GROUP VENUSTA. 33 

N. albigula may be known by the pure white, instead of plumbeous, 
basal color of fur on throat and chest. 

Neotoma intermedia angusticeps was based on a rather abnormal 
specimen of N. albigula. 

Specimens examined. — Total number 491, from localities as follows: 

Arizona: Big Sandy Creek, Mohave County, 1; Bisbee, 1; Camp Verde, 1; 
Carrizalillo Spring, 2; Dos Cabezas, 1; Eagle Mountain (latitude 31° 47', 
longitude 30° 150, 10; Fort Bowie, 1; Fort Grant, 1; Fort Huachuca, 9; 
Fort Whipple, 1; Granite Mountains, 1; Holbrook, 1; Huachuca Moun- 
tains, 19; Hualpai, 8; La Osa, Pima County, 4; Little Meadows, Mohave 
County, 1; New River (30 miles northwest of Phoenix), 1; Oracle, 2; 
Phoenix, 2; Pinal Mountains, 1; San Bernardino Springs, 1; Santa Rita 
Mountains, 2, Tinajas Altas, Yuma County, 1; Tombstone, 1; Tucson, 3; 
Tucson (75 miles southwest), 1; Tule Well, 2. 

New Mexico: Abiquiu, 2; Alma, 2; Ancho, 2; Animas Mountains, 4; Ani- 
mas Valley (Lang Ranch), 1; Bear Spring Mountains, Socorro County, 5; 
Big Hatchet Mountains, 5; Burley, 4; Cabezon, 1; Cabra Spring, San 
Miguel County, 2; Cactus Flat (20 miles north of Cliff), 7; Canyon de 
Chelly, 2; Capitan Mountains, 18; Carlsbad, 3; Chama Canyon, 1; Ciene- 
guilla (near Rinconada), 3; Corona, Lincoln County, 2; Cuchillo, 3; Cuervo, 
Guadalupe County, 1; Datil Mountains, 1; Deining, 5; Dog Spring, Grant 
County, 3; Espanola, 1; Fairview, 1; Faywood, 1; Florida Mountains, 6; 
Fort Wingate, 2; Gallina Mountains, 1; Gallup, 3; Garfield, 1; Glenwood, 
San Francisco River, 3; Grant, 16; Grant County (southwest corner), 
9; Gila, 1; Hachita, 5; international boundary, 100 miles west of El 
Paso, 15; Isleta, 1; Jarilla, 4; Jicarilla Mountains, 16; Kingston, 2; La- 
guna, 12; Lake Valley, 1; Lamy, 2; Las Cruces, 1; Las Palomas, 1; 
Magdalena Mountains, 2; Malpais Spring, Otero County, 1; Mangos Valley 
(near Tyrone), 4; Manzano Mountains, 14; Mesa Jumanes, 1; Organ Moun- 
tains, 4; Playas Valley, 4; Pleasanton, 1; Redrock, 5; Ribera, 1; Riley, 
1; Rinconada, 5; Rio Alamosa, 2; Rio Puerco, 8; Roswell (40 miles west) 
1; San Andres Mountains, 15; Sandia Mountains, 1; San Mateo Mountains 
(Indian Butte), 1; San Pedro, 2; San Rafael, 2; Santa Clara Canyon, 
Santa Fe County, 4; Santa Rosa, 10; Silver City, 7; Socorro, 6; Stinking 
Spring Lakes, 6; Tucumcari (25 miles southwest), 1; Tularosa, 5; Weed, 
2; Wingate, 1. 

Texas: El Paso, 6; Franklin Mountains, 20; Guadalupe Mountains, 2; Kent, 
1; Llano, 1; Marfa, 1; Paisano, 1; Sierra Blanca, 1; Stanton, 1; Toyah, 1. 

Chihuahua: Chihuahua, 21; Ciudad Juarez, 2; Colonia Diaz, 8; Escalon, 3; 
Guzman, 4; San Luis Mountains, 1; Santa Eulalia, 1; Santa Rosalia, 11. 

Coahuila: Jaral, 2; Jimulco, 1; Monclova, 1. 

Durango: Mapimi, 2. 

Sonora: Hermosillo, 2; Magdalena, 2; Nogales, 1; Patagonia Mountains, 1; 
Pozo de Luis, 4; San Bernardino River (near international boundary), 
1; San Jose Mountains, 4; Santa Cruz, 2; Sonoyta, 4. 

NEOTOMA ALBIGULA VENUSTA True. 

Colorado Valley Wood Rat. 

(PI. II, figs. 2, 2a.) 

Neotoma venusta True, Proc. U. S. Nat. Mus., XVII, p. 354, November 15, 1S94 
[author's separates issued June 27, 1894]. Type from Carrizo Creek, San Diego 
County, Cal.; No. fjfff, $ yg., U. S. National Museum ; collected by F. Stephens. 

52668°— 10 3 



34 NORTH AMERICAN FAUNA. [ho. 31. 

Neotoma cumulator Mearns, Proc. U. S. Nat. Mus., XX, p. 503, January 19, 1898 
[author's separates issued March' 5, 1897]. Type from Fort Yuma, Cal. ; No. 60348, 
$ ad., U. S. National Museum; collected by Dr. E. A. Mearns, April 2, 1894. 

Neotoma desertorum grandis Elliot, Pub. Field Columb. Mus., Chicago, Zool. Ser., Ill, 
No. 14, p. 247, January, 1904. Type from Cameron Lake, Sierra Nevada, Kern 
County, Cal.; in Field Museum of Natural History; collected by Edmund Heller. 

Distribution. — Colorado River Valley from northwestern Arizona 
to Gulf of California and west through southern California to eastern 
basal slopes of southern Sierra Nevada, San Bernardino, and San 
Jacinto Mountains. Lower Sonoran Zone. 

General characters. — Closely related to N. albigula, but larger; 
differing also in slight cranial characters. 

Color. — Same as N. albigula. - 

Skull. — Skull similar to that of N. albigula, but usually larger, 
more angular, and in adults more arched across anterior roots of 
zygomata; nasals narrower posteriorly. 

Measurements. — Average of 5 adults from Fort Yuma, Cal. : Total 
length, 396; tail vertebra, 182; hind foot, 38. Skull (average of 
same): Basilar length, 38.5; zygomatic breadth, 24.1; interorbital 
breadth, 6; length of nasals, 17.3; length of incisive foramina, 9.9; 
length of palatal bridge, 7.6; alveolar length of upper molar series, 
8.4. 

Remarks. — This subspecies is a large western offshoot of the 
albigula group, grading into typical albigula east of the Colorado 
River in western Arizona and northern Sonora. 

The type of venusta is a very young individual whose characters 
are not very apparent unless comparison is made with specimens of 
similar age. This fact may have led to the publication of N. cumu- 
lator and N. d. grandis, which were based on fully adult specimens of 
the same species. 

Specimens examined. — Total number 103, from localities as follows: 

California: Borrego Spring, San Diego County, 5; Cameron Lake, 2; Carriso 

Creek (type locality), 3; Carriso Creek (20 miles east), 1; Colorado River, 

at monument 204, 6; Fort Yuma, 26; New River, 1; Pilot Knob, 7. 
Arizona: Adonde, 1; Beale Spring, 50 miles west of Prescott, 2; Brawley, 4; 

Colorado River, opposite Needles, 2; Dolans Spring, Mohave County, 3; 

Ehrenberg (15 miles southwest), 4; Fort Mohave, 2; Mineral Park, 2; 

Mud Spring, Mohave County, 2; Nortons, Yuma County, 1; Yuma, 14. 
Lower California: Cocopah Mountains (east base), 3; Colonia Lerdo, 1; 

Gardners Lagoon, 2; head of Hardy River, 2; Seven Wells, Salton River, 

2; Volcano Lake, 5. 

NEOTOMA ALBIGULA WARRENI Merriam. 

Warren Wood Rat. 

Neotoma albigula warreni Merriam, Proc. Biol. Soc. Wash., XXI, pp. 143-144, June 9, 
1908. Type from Gaume's Ranch, Baca County (northwest corner), Colo, (alti- 
tude 4,600 feet); No. 151051, $ ad., U. S. National Museum (Biological Survey 
Collection); collected by Merritt Cary, November 28, 1907. 



1910.] ALBIGULA GROUP — MELANUEA. 35 

Distribution. — Plains region of southeastern Colorado and north- 
eastern New Mexico. Upper Sonoran Zone. 

General characters. — Similar in general to N. albigula, but color 
decidedly grayer; hind foot larger; audital bullae smaller. 

Color. — Fresh pelage (November) : Upper parts pale buffy gray, 
lightest along cheeks and sides, slightly darkened on top of head and 
back by dusky-tipped hairs, the buffy element becoming purer in a 
pale, but rather distinct, lateral line along sides of belly; feet and 
under parts white; the fur basally pure white on throat and on pec- 
toral and inguinal regions, as in albigula; outer sides of hind legs 
slate-grayish, faintly tinged with buffy; tail brownish above, white 
below. Worn pelage (May) : Upper parts strongly suffused with 
vinaceous buffy quite different from albigula in corresponding stages 
of wear. 

Skull. — Closely resembling that of typical albigula, but audital 
bullae decidedly smaller, less inflated. 

Measurements. — Type: Total length, 324; tail vertebrae, 137; 
hind foot, 37. Average of 6 adult topotypes: 313; 133; 36. SJcull 
(average of 7 adults): Basilar length, 36.3; zygomatic breadth, 23.5; 
interorbital breadth, 5.8; length of nasals, 16.5; length of incisive 
foramina, 8.5; length of palatal bridge, 8; alveolar length of upper 
molar series, 8.5. 

Remarks. — This subspecies is based in part on a fine series of 8 
specimens, all topotypes, kindly loaned by E. R. Warren. In 
grayer, vinaceous instead of pinkish buffy coloration, it differs con- 
spicuously from N. albigula, the only form with which close compari- 
son is necessary. Externally it resembles a geographical neighbor, 
N. m. canescens, but well-marked cranial characters place it in the 
albigula group. Specimens from Clayton, N. Mex., are not quite 
typical and appear to be grading toward albigula. 

Specimens examined. — Total number 13, as follows: 

Colorado: Gaume's Ranch, Baca County (type locality), 10. 
New Mexico: Clayton, 3. 

NEOTOMA ALBIGULA MELANURA Merriam. 

Sonora Wood Rat. 

(PI. II, tigs. 3, 3a.) 

Neotoma intermedia melanura Merriam, Proc. Biol. Soc. Wash., IX, pp. 126-127, July 
2, 1894. Type from Ortiz, Sonora, Mexico; No. J|fH» <? yg- ad., U. S. National 
Museum (Biological Survey Collection); collected by Vernon Bailey, November 
13, 1889. 

Distribution. — Western basal slopes of Sierra Madre in southern 
Sonora and southwestern Chihuahua. Lower Sonoran and upper part 
of Arid Tropical zones. 



36 NORTH AMERICAN FAUNA. [no. 31. 

General characters. — Color darker than N. albigula; tail black 
above; skull less massive, with decidedly smaller audital bullae. 

Color. — Fresh pelage: Upper parts between dark buff and dull pink- 
ish buff, lightest along sides, darkened over back by dusky hairs; 
middle of face and outer sides of legs pale drab; underparts dull 
white; feet pure white; tail black above, whitish below. 

Skull. — Similar to that of N. albigula, but lighter; rostrum more 
slender; teeth slightly smaller; bullae much smaller. 

Measurements. — Average of 2 adults from Alamos, Sonora. Total 
length, 347; tail vertebrae, 160; hind foot, 37.5. Skull (average of 
same): Basilar length, 37.5; zygomatic breadth, 23.2; interorbital 
breadth, 6.1; length of nasals, 17.9; length of incisive foramina, 9.6; 
length of palatal bridge, 7.9; alveolar length of upper molar series, 
8.3. 

Remarks.— Specimens from the type region, Ortiz (type locality), 
and Batamotal, are not t} T pical, but show intergradation in both color 
and cranial characters between albigula and a more differentiated form 
ranging farther southward along the west slope of the Sierra Madre. 

Specimens examined. — Total number 39, from localities as follows: 

Sonora: Alamos, 7; Batamotal, 2; Camoa, 3; Guaymas, 16; Ortiz, 1 (type); 

Presidio, 8. 
Chihuahua: Batopilas, 2. 

NEOTOMA ALBIGULA LEUCODON Merriam. 

White-toothed Wood Rat. 

(PI. II. figs. 4, 4a.) 

Neotoma leucodon Merriam, Proc. Biol. Soc. Wash., IX, pp. 120-121, July 2, 1894. 
Type from San Luis Potosi, San Luis Potosi, Mexico; No. 50137, $ ad., U. S. 
National Museum (Biological Survey Collection); collected by E. W. Nelson, 
August 4, 1892. 

Distribution. — Mexican plateau region from western Nuevo Leon 
south to the northern part of the State of Mexico. Upper and Lower 
Sonoran zones. 

General characters. — Larger than X. albigula; color darker; cranial 
characters distinctive. 

Color. — Upperparts creamy buff, clearest along sides, becoming 
dark pinkish buff in worn pelage, moderately darkened dorsally by 
blackish hairs; nose and middle of face grayish; outer sides of 
forearms and hind legs varying from gray to drab gray; feet and 
underparts white, the fur on pectoral and inguinal regions pure white 
to roots; tail blackish above, white below. 

Skull. — Larger than that of N. albigula; rostrum and dentition 
heavier; supraorbital ridges more developed, forming slightly pro- 
jecting frontal shelves. Compared with N. a. venusia the skull is 



1910.] ALB1GULA GROUP DURANG.E. 37 

less arched across anterior roots of zygomata; supraorbital ridges 
more developed and projecting. 

Measurements. — Average of 6 adult topotypes: Total length, 358; 
tail vertebra 3 , 157; hind foot, 38.5. Skull (average of same) : Basilar 
length, 40.5; zygomatic breadth, 26.2; interorbitai breadth, 6.1; 
length of nasals, 17.3; length of incisive foramina, 9.9; length of 
palatal bridge, 8.6; alveolar length of tipper molar series, 9. 

Remarks. — In southern Coahuila and western Nuevo Leon leucodon 
intergrades with albigula, the form ranging thence to the northward. 
Although widely separated geographically, leucodon is rather closely 
related to venusta, but differs in color and in the cranial characters 
already noted. 

Specimens examined. — Total number 75, from the following localities : 

San Luis Potosi: San Luis Potosi (type locality), 9; Hacienda La Parada, 9. 

Aguas Calientes: Chicalote, 13. 

Coahuila: Carneros, 1; La Ventura, 2; Saltillo, 2; Sierra Encarnacion, 10. 

Guanajuato: La Quemada, 3. 

Mexico: Marques, 2. 

Jalisco: Lagos, 2. 

Nuevo Leon: Santa Catarina, 3. 

Queretaro: Tequisquiapan, 1. 

Tamaulipas: Jaumave, 3; Miquihuana, 5. 

Zacatecas: Berriozabel, 9; Zacatecas, 1. 

NEOTOMA ALBIGULA DURANGO Allen. 

Durango Wood Rat. 

Neotoma intermedia durangse. Allen, Bull. Amer. Mus. Nat. Hist., XIX, pp. 602-603, 
November 12, 1903. Type from San Gabriel, northwestern Durango, Mexico; 
No. 21185, $ ad., American Museum of Natural History, New York; collected 
by J. H. Batty, February 20, 1903. 

Distribution. — Eastern basal slopes of the Sierra Madre west of 
and above the range of N. albigula, from central Durango to south- 
western Chihuahua, Mexico. Upper Sonoran Zone. 

General characters. — Similar in external appearance to N. albigula, 
but slightly larger; skull larger; dentition much heavier. 

Color. — General color above mixed buff and black, varying in 
some specimens to pale cream buff and black, palest on face, becom- 
ing more or less rusty brown in worn pelage; underparts and feet 
white, the hairs white to roots on breast and throat; tail brownish 
black above, white below. 

Skull. — Very similar to that of N. albigula, but larger; frontal 
region somewhat narrower posteriorly; dentition much heavier. 

Measurements. — Average of 5 adults from the type locality (from 
original description). Total length, 350; tail vertebrae, 157.7; hind 
foot (without claw), 32. Skull (an adult topotype) : Basilar length, 



38 NORTH AMERICAN FAUNA. [no. 31. 

37.6; zygomatic breadth, 23.3; interorbital breadth, 5.8; length of 
nasals, 16.7; length of incisive foramina, 9; length of palatal bridge, 
7.7; alveolar length of upper molar series, 9.1. 

Remarks. — This form is very closely related to N. albigula, from 
which it differs chiefly in slightly larger size and heavier dentition. 
Specimens examined. — Total number 9, from localities as follows: 
Durango: Durango, 1; Inde, 1 ; Rancho Bailon and Rancho Santuario (north- 
western Durango), 4; San Gabriel (type locality), 2. 
Chihuahua: Parral, 1. 

NEOTOMA ALBIGULA ZACATEC2E Goldman. 

Zacatecas Wood Rat. 

Neotoma leucodon zacatecse Goldman, Proc. Biol. Soc. Wash., XVIII, p. 30, February 
2,1905. Type from Plateado, Zacatecas, Mexico; No. 90957, £ ad., U.S. National 
Museum (Biological Survey Collection); collected by E. W. Nelson and E. A. 
Goldman, September 4, 1897. 

Distribution. — Sierra Madre in western Zacatecas, Mexico. Transi- 
tion Zone. 

General characters. — Size much larger and color decidedly darker 
than N. albigula. Nearly related to N. a. leucodon, but darker; 
upper lip dusky instead of grayish or whitish ; skull more arched and 
otherwise peculiar. , 

Color. — Partly worn pelage: Upper parts pale cinnamon, lightest 
along cheeks and sides, becoming redder on rump, well darkened along 
median dorsal area by black-tipped hairs; chest and inguinal region 
pure white; belly dull creamy white; upper lip dusky; feet white; 
tail well haired, sharply bicolor, brownish black above, white below. 

SJcull. — Similar to that of N. a. leucodon, but more arched across 
anterior roots of zygomata; rostrum more decurved; frontals 
longer ; maxillary arm of zygoma heavier ; upper incisors smaller. 

Measurements. — Type: Total length, 358; tail vertebrae, 162; 
hind foot, 37. Skull (type): Basilar length, 39.1; interorbital 
breadth, 6.3; length of nasals, 16.9; length of incisive foramina, 9.8; 
length of palatal bridge, 8.8 ; alveolar length of upper molar series, 8.6. 

Remarks. — This subspecies is a dark, mountain representative of 
the albigula group. Specimens from Valparaiso, Zacatecas, are 
grading toward leucodon. 

Specimens examined. — Five, all from Zacatecas, as follows; 

Plateado (type locality), 1; Valparaiso, 4. 

NEOTOMA LATIFRONS Merriam. 

QUERENDARO WOOD RAT. 

(PI. II, figs. 5, 5a.) 

Neotoma latifrons Merriam, Proc. Biol. Soc. Wash., IX, p. 121, July 2, 1894. Type 
from Querendaro, Michoacan, Mexico; No. 50135, $ ad., U. S. National Museum 
(Biological Survey Collection); collected by E. W. Nelson, August 8, 1892. 



1910.1 ALBIGULA GROUP NELSONI. 39 

Distribution. — Known only from type locality. Lower part of 
Lower Sonoran Zone. 

General characters. — Similar to N. a. leucodon, but smaller; color 
darker, more ocliraceous; skull with projecting supraorbital shelves. 

Color. — Partly worn pelage: Upper parts ochraceous buffy, purest 
along cheeks and sides, moderately darkened dorsally by dusky 
hairs ; middle of face grayish ; nose and upper lip dusky ; feet white ; 
under parts dull white, the fur on pectoral and inguinal regions only, 
pure white to roots; tail dusky above, dull white below. 

Skull. — Similar to that of N. a. leucodon, but smaller, shorter, and 
relatively broader; frontal region relatively broader and flatter pos- 
teriorly, with lateral margins still more developed, forming strongly 
projecting supraorbital shelves. 

Measurements. — Type: Total length, 350; tail vertebrae, 149; 
hind foot (dry skin), 34.5. Skull (type): Basilar length, 36.9; 
zygomatic breadth, 24.6; interorbital breadth, 5.9; length of nasals, 
16.6; length of incisive foramina, 9.5; length of palatal bridge, 8.3; 
alveolar length of upper molar series, 9.1. 

Remarks. — This form is rather closely related to N. a. leucodon and 
may intergrade with it in southern Guanajuato. 

Specimens examined. — One, the type. 

NEOTOMA NELSONI Goldman. 

Nelson Wood Rat. 

(PI. II, figs. 6, 6a.) 

Neotomanelsoni Goldman, Proc. Biol. Soc. Wash., XVIII, pp. 29-30, February 2, 1905. 
Type from Perote, Veracruz, Mexico; No. 54320, 9 ad., U. S. National Museum 
(Biological Survey Collection); collected by E. W. Nelson, June 3, 1893. 

Distribution. — High plains along the eastern edge of the Mexican 
plateau region in eastern Puebla and extreme west-central Veracruz. 
Upper Sonoran Zone. 

General characters. — Size large; tail stout, well haired, rather 
short; ears large; coloration dark.' Related to N. a. leucodon, but 
differing in darker color and important cranial characters. 

Color. — Partly worn pelage: Upper parts pale cinnamon, heavily 
overlaid with smoky brown or Prout brown, this color predominating 
on top of head and along back, becoming grayish brown on cheeks 
and middle of face; under parts dull white, the deep plumbeous basal 
color showing through everywhere except on a small pectoral area 
where the fur is pure white; nose and upper lip blackish; feet white; 
tail indistinctly bicolor (nearly concolor toward tip), smoky brown 
above, slightly paler and grayer below. Young (about half grown) : 
Above grayish brown tinged with smoky brown. 

Skull. — General outline of skull similar to that of N. a. leucodon but 
nasals more wedge-shaped, pointed posteriorly; with a narrow emar- 



40 NORTH AMERICAN FAUNA. [no. 31. 

gination between the ends; palatal bridge decidedly shorter than 
incisive foramina (in leucodon about equal), and with a short pos- 
terior median projection (absent in leucodon) ; ascending branches of 
premaxillae very long, the ends reaching posteriorly beyond nasals in 
approaching parallel lines to interorbital constriction; dentition as in 
N. a. leucodon. 

Measurements. — Average of 5 adult topotypes: Total length, 349; 
total vertebrae, 154; hind foot, 38. Skull (average of same): Basilar 
length, 39; zygomatic breadth, 25.3; interorbital breadth, 5.7; 
length of nasals, 17 ; length of incisive foramina, 9.7 ; length of palatal 
bridge, 8.4; alveolar length of upper molar series, 8.9. 

Remarks. — This species is a member of the albigula group. Its 
darker color at all ages readily separates it externally from leucodon, its 
nearest known relative. 

Specimens examined. — Total number 11, all from the type locality. 

NEOTOMA PALATINA Goldman. 

Bolanos Wood Rat. 

(PI. Ill, figs. 1, la.) 

Neotoma palatina Goldman, Proc. Biol. Soc. Wash., XVIII, pp. 27-28, February 2, 
1905. Type from Bolanos, Jalisco, Mexico; No. 90959, $ ad., U. S. National 
Museum (Biological Survey Collection); collected by E. W. Nelson and E. A. 
Goldman, September 12, 1897. 

Distribution. — Known only from the type locality in the canyon 
of the Bolanos River. Arid Tropical Zone. 

General characters. — Size large ; ears small ; pelage short and coarse ; 
skull heavy; sphenopalatine vacuities absent; vomer peculiar. 

Color. — Worn pelage (September) : Upper parts pale cinnamon, 
suffused with buffy along cheeks and sides, becoming much darker on 
dorsal region from abundant admixture of black hairs; under parts 
dull white, the basal color plumbeous except on throat, breast, and 
inguinal region; muzzle brownish gray; feet white; tail scantily 
haired, bicolor, blackish above, soiled whitish below. 

Skull. — Skull large and massive, rather smoothly rounded, well 
arched across anterior roots of zygomata; rostrum short and heavy; 
nasals wedge-shaped, reaching posteriorly to plane of orbits; frontals 
long and broad, the sides very slightly upturned ; brain case some- 
what truncate posteriorly; interparietal large and rectangular, with- 
out trace of a posterior angle; mastoids small; interpterygoid fossa 
very broad, evenly rounded anteriorly, slightly constricted poste- 
riorly by curvature inward of pterygoids; hamular process of ptery- 
goids short and heavy; sphenopalatine vacuities completely closed 
by palatines; vomer prolonged posteriorly as a thin vertical plate 



1910.] ALBIGULA GROUP — MONTEZUMiE. 41 

along median line of presphenoid, partially dividing posterior nares, 
and ending in a point at suture between presphenoid and basisphe- 
noid; bullae small and heavy, somewhat pear-shaped and pointed 
anteriorly; dentition of the albigula type. 

Measurements. — Type: Head and body, 187; hind foot, 37. Skull 
(type): Basilar length, 38.5; zygomatic breadth, 24.4; interorbital 
breadth, 5.8; length of nasals, 18; length of incisive foramina, 9; 
length of palatal bridge, 7.6 ; alveolar length of upper molar series, 8.8. 

Remarks. — This remarkable animal appears to be an aberrant mem- 
ber of the albigula group. In general form the skull somewhat 
resembles that of N. a. melanura, but differs too widely in important 
details to require close comparison with any known form. The pos- 
terior prolongation of the vomer to the anterior border of the basi- 
sphenoid is a unique character, and the complete closure of the spheno- 
palatine vacuities has been observed elsewhere in the subfamily only 
in some of the members of the subgenus Teonoma. In developing 
upward and backward the palatines have left small irregular vacuities 
along the palatopterygoid suture, which open outward into the exter- 
nal pterygoid fossse. 

Specimens examined. — One, the type. 

NEOTOMA MONTEZUMA Goldman. 

Montezuma Wood Rat. 

(PI. Ill, figs. 2, 2a.) 

Neotoma montezumx Goldman, Proc. Biol. Soc. Wash., XVIII, p. 29, Feb. 2, 1905. 
Type from Zimapan, Hidalgo, Mexico; No. 81426, $ ad., U.S. National Museum 
(Biological Survey Collection); collected by E. W. Nelson and E. A. Goldman, 
October 17, 1896. 

Distribution. — Known only from the type locality on the high 
plains of western Hidalgo. Upper Sonoran Zone. 

General characters. — Size medium; tail rather short, scantily haired, 
bicolored. Somewhat resembling N. a. melanura, but differing in 
darker color and in marked cranial characters. 

Color. — Partly worn pelage: Upper parts dull ochraceous buff, 
palest on head and along sides, everywhere evenly and abundantly 
darkened by brownish or blackish hairs; under parts dull white, the 
under color of fur plumbeous, except on throat, chest, and inguinal 
region, where it is pure white ; nose, upper sides of forearms, and outer 
sides of hind legs, grayish brown; feet white; tail blackish above, 
whitish below. 

Skull. — In general form similar to N. a. melanura, but larger and 
heavier; maxillary arm of zygoma decidedly heavier; nasals narrower 
posteriorly; dentition much heavier; first upper molar with antero- 



42 NORTH AMERICAN FAUNA. [no. 31. 

internal reentrant angle deeper, but not so deep as in N. mexicana; 
bullae small, as in N. a. melanura, but more pointed anteriorly; 
ascending branches of premaxillse very long, reaching or nearly 
reaching posteriorly to interorbital constriction and somewhat 
broadening toward ends; frontals constricted well forward, the sides 
slightly upturned. 

Measurements. — Type: Head and body, 181; hind foot, 39. Skull 
(type): Basilar length, 38; zygomatic breadth, 24; interorbital 
breadth, 5.9; length of nasals, 17.1; length of incisive foramina, 9.6; 
length of palatal bridge, 7.9; alveolar length of upper molar series, 9. 

Remarks. — Neotoma montezumse, is a member of the albigula group 
with distinctive characters, showing no close affinity with any neigh- 
boring species; its general resemblance to melanura is somewhat 
superficial. The skull shows a slight departure from the albigula 
type in the depth of the anterointernal reentrant angle of the first 
upper molar. Unfortunately the end of the type specimen's tail is 
missing, having been lost before the animal was captured, as shown 
by the healed-over end. 

Specimens examined. — Total number five, all from Hidalgo, as 
follows : 

Zimapan (type locality), 4; Ixmiquilpan, 1. 

NEOTOMA INTERMEDIA Rhoads. 
Rhoads Wood Rat. 
(PL III, figs. 3, 3a.) 

Neotoma intermedia Rhoads, Amer. Nat., XXVIII, pp. 69-70, January, 1894. Type 
from Dulzura, San Diego County, Cal.; No. 1343, <?ad., collection of S. N. Rhoads; 
collected by C. H. Marsh, August 21, 1893. 

Neotoma californica Price, Proc. Cal. Acad. Sci., 2 Ser., IV, pp. 154-156* pi. XI, May 
9, 1894. Type from Bear Valley, San Benito County, Cal.; No. 335, $ ad., Museum 
Stanford University; collected by C. H. Gilbert and W. W. Price, April 2, 1893. 

Distribution. — Lower slopes of southern part of Sierra Nevada and 
coast region of California from Monterey Bay southward and through- 
out the mountains of Lower California to near Cape San Lucas. Upper 
and Lower Sonoran zones. 

General characters. — Size medium; ears large; tail moderately long, 
bicolor. Somewhat similar to N. albigula, but color darker; fur on 
throat and breast basally plumbeous instead of white; cranial 
characters distinctive. 

Color. — Fresh pelage: Upper parts grayish brown, more or less suf- 
fused with pale buff, darkest on dorsal region from more abundant 
admixture of black-tipped hairs; under parts white or creamy white, 
the fur basahV plumbeous; breast usually crossed by a faint buffy 
band; sides of ankles usually dusky; feet white; tail black above, dull 



1910.] 



INTERMEDIA GROUP INTERMEDIA. 



43 



white below. Winter pelage: Upper parts darker, the black-tipped 
hairs more conspicuous than in summer pelage. 

STcuU. — Rather small and light, brain case large and smoothly 
rounded; temporal ridges faint and widely separated. General out- 
line similar to that of N. albigula, but rostrum much more slender; 
nasals broader and more evenly rounded posteriorly; palate convex 
instead of concave posteriorly; bullae smaller, more tapering, and 




Fig. 5. — Distribution of the Neotoma intermedia group. 

turned inward anteriorly; first upper molar with anterointernal 
reentrant angle shallow as in N. albigula. 

Measurements. — Average of 3 adult topotypes: Total length, 325; 
tail vertebrae, 160; hind foot, 31.5. Skull (average of same) : Basilar 
length, 34.6; zygomatic breadth, 21.4; interorbital breadth, 5.5; 
length of nasals, 15.3 ; length of incisive foramina, 8.7 ; length of palatal 
bridge, 7.2; alveolar length of upper molar series, 7.5. 



44 NORTH AMERICAN FAUNA. [no. SI. 

Remarks. — The abundant material examined shows that N. inter- 
media and N. albigula belong to rather nearly related but quite dis- 
tinct groups. Their ranges meet or overlap in southern California 
without trace of intergradation. Typical intermedia inhabits the 
more humid or higher areas, passing into a paler form, gilva, along 
the basal slopes of the mountains and on sandy deserts. Specimens 
from the upper slopes of the Sierra la Laguna in the Cape Region 
of Lower California average more brownish in color than typical 
intermedia; the skulls are very similar, but have rather more broadly 
spreading zygomata and slightly wider interpterygoid fossae than is 
usual in intermedia. But these differences seem too slight for sub- 
specific recognition, although the specimens appear to represent an 
isolated colony, cut off from the main distribution area of inter- 
media by an interposed arm of the range of the larger and paler form, 
arenacea. Specimens of intermedia in fresh winter pelage are usually 
darker than those in the worn summer coat. 

N. californica does not differ appreciably from typical intermedia. 

Specimens examined. — Total number 252, from localities as follows: 

California: Arroyo Seco Canyon (near Pasadena), 1; Bear Valley, San Benito 
County, 4; Bear Basin, Monterey County, 2; Bell Station, 1; Bergman, 6; 
Capistrano, 3; Chihuahua Mountains, 1; Cone Peak, 5; Dulzura (type local- 
ity), 9; El Nido, 2; Fremont Peak, Gabilan Range, 3; Jamul, 1; Pacheco 
Pass, 4; Piute Mountains, 1; Porterville (8 miles east), 2; Posts, 1; Poway, 
1; Priest Valley, Monterey County, 1; Reche Canyon (near San Bernar- 
dino), 13; Redlands, 1; Riverside, 14; San Benito Valley, 3; San Bernar- 
dino, 2; San Bernardino Mountains, 8; San Diego, 1; San Fernando, 4; San 
Francisquito Canyon, Los Angeles County, 1; San Jacinto Valley, 1; San 
Luis Obispo, 3; San Pedro, 1; San Rafael Mountains, 2; Santa Inez, 1; Santa 
Isabel, 4; Santa Lucia Peak, 6; Temescal, 1; Tassajara Creek, Monterey 
County, 3; Tijuana River (mouth), 1; Wildomar, 1; Winchester, 1; Witch 
Creek, 2; Zaca Lake, Santa Barbara County, 2. 

Lower California: Calmalli, 1; Comondu, 15; El Potrero (near Mulege), 2; 
Ensenada, 9; Ensenada (20 miles east), 2; La Grulla, San Pedro Martir 
Mountains, 1; Palomar, San Pedro Martir Mountains, 1; Paso Hondo (16 
miles north of La Purisima), 1; Rancho San Antonio (west base San Pedro 
Martir Mountains), 9; Rancho Viejo (15 miles east of Alamo), 4; Rosario, 5; 
San Andres, 6; San Bruno, 5; San Fernando, 3; San Ignacio, 7; San Isidro, 8; 
San Quintin, 24; Santana, 1; San Telmo, 1; Santa Rosalia Bay, 1; Santo 
Domingo, 2; Sierra la Giganta, 1; Sierra la Laguna, 10; Tecate River, 1; 
Tijuana, 2; Trinidad Valley (northwest base San Pedro Martir Mountains), 
3; Vallecitos, San Pedro Martir Mountains, 3. 

NEOTOMA INTERMEDIA GILVA Rhoads. 

Yellow Wood Rat. 

Neotoma intermedia gilva Rhoads, Amer. Nat., XXVIII, p. 70, January, 1894. Type 
from Banning, San Bernardino County, Cal.; No. 1665, $, Academy of Natural 
Sciences, Philadelphia; collected October 26, 1893. 

Neotoma descrtorum sola Merriam, Proc. Biol. Soc. Wash., IX, p. 126, July 2, 1894. 
Type from San Emigdio, Kern County, Cal.; No. f^fH- <? ad -> n - S. National 



1910.] INTEEMEDIA GEOUP GILVA. 45 

Museum (Biological Survey Collection); collected by E. W. Nelson, October 24, 
1891. 
Neotoma bellafelipensis Elliot, Pub. Field Columb. Mus., Zool. Ser., Ill, pp. 217-218, 
June, 1903. Type from San Felipe, Lower California, Mexico; type in Field 
Museum of Natural History; collected by Edmund Heller in March or April, 1902. 

Distribution. — Arid plains and basal slopes of mountains mainly 
along the eastern border of the range of N. intermedia, from Stanley 
in Fresno County, Cal., south through northeastern Lower California 
to the Santa Clara Mountains on the west side of the Peninsula. 
Lower Sonoran Zone. 

General characters. — Same as N. intermedia, but color paler, upper- 
parts more buffy or yellowish, less obscured by blackish hairs. 

Color. — Fresh winter pelage: Upperparts varying from creamy buff 
to pale ochraceous buff, purest along sides, thinly overlaid with black- 
ish hairs ; underparts and feet white; ankles less dusky than in N. inter- 
media, the inner sides often pure white; tail brownish gray above, 
white below. 

Skull. — As in N. intermedia. 

Measurements. — An adult from Cabezon, Cal. : Total length, 330 ; 
tail vertebra?, 160; hind foot, 34. Skull (of same): Basilar length, 
33.2; zygomatic breadth, 20.9; interorbital breadth, 5.5; length of 
nasals, 15.9; length of incisive foramina, 8.6; length of palatal bridge, 
6.8; alveolar length of upper molar series, 8.1. 

Remarks.— N. i. gilva is a desert form differing from typical N. inter- 
media only in paler, more yellowish coloration of upperparts. It 
intergrades with intermedia along the slopes of the mountains or 
wherever the climate becomes more humid. In color gilva sometimes 
closely resembles lepida, but the pelage is shorter and coarser. 

Large series of specimens show considerable range of variation in 
color and from some localities average somewhat yellower than from 
others, but there are no characters by which felipensis can be satis- 
factorily separated from typical gilva. Through transposition, prob- 
ably while in the hands of the skull cleaner, the skull of a southern 
member of the fuscipes group was numbered to correspond with a skin 
of gilva, and the composite specimen later became the type of sola. 
This mismatched skull belongs either to macrotis or simplex, both of 
which had already been described, but it can not be definitely assigned 
to either, owing to their close agreement in skull characters. The 
skin, however, of the type of sola is certainly gilva, as is a topotype 
taken at the same time. Hence the name seems to belong in synonymy 
under the latter form. 

Specimens examined. — Total number 118, from localities as follows: 

California: Banning, 1 (type); Big Pine Mountain, 1; Cabezon, 2; Carriso 
Plains, 2; Cuyama Valley, 1; east base of Coast Range Mountains, San 
Diego County, 1; Fort Tejon, 17; Frazier Mountain, 1; Jacumba, 7; Kern 
River (south fork), 4; Kern River (15 miles northeast of Bakersfield), 1; 



46 NORTH AMERICAN FAUNA. [no. 31. 

Milpitas Ranch, Monterey County, 1; Mount Pinos, 2; Mountain Spring, 
San Diego County, 4; Mount Waterman, San Gabriel Mountains, 1; Palm 
Springs, 1; Paraiso Springs, 1; San Emigdio, Kern County, 2; San Gorgonio 
Pass, 8; San Jacinto Lake, 1; San Jacinto Mountains (San Andres Canyon), 
2; Seven Oaks, San Bernardino Mountains, 1; Stanley, 6; Tehachapi, 2; 
Tejon Pass, 2; Vallecitos, 2; White Water, 3. 
Lower California: Aguaje de la Natividad, 1; Calamahue, 3; Canyon Vicente, 
1; Esperanza Canyon, east base San Pedro Martir Mountains, 1; Nachoguero 
Valley, 1; Pozo San Augustin (20 miles east of San Fernando), 2; San Angel 
(20 miles west of San Ignacio), 2; San Felipe Bay, 20; San Francisquito, 6; 
Tinaja Santa Clara, Santa Clara Mountains, 2; Yubay, 2. 

NEOTOMA INTERMEDIA PRETIOSA Goldman. 

Matancita Wood Rat. 

(PI. Ill, figs. 4, 4a.) 

Neotoma intermedia pretiosa Goldman, Proc. Biol. Soc. Wash., XXII, p. 139, June 25, 
1909. Type from Matancita (also called Soledad), 50 miles north of Magdalena 
Bay, Lower California, Mexico (altitude 100 feet); No. 146123, $ ad., U. S. 
National Museum (Biological Survey Collection); collected by E. W. Nelson and 
E. A. Goldman, November 17, 1905. 

Distribution. — West coast and islands of Lower California, from San 
Jorge (southwest of Comondu) south to Margarita Island. Lower 
Sonoran Zone. 

General characters. — Size much larger and color paler than N. inter- 
media. Similar in size to N. i. arenacea, but color paler; tail shorter; 
audital bullae larger. 

Color. — Worn pelage: Upperparts very pale drab gray, purest 
along cheeks and sides, overlaid on top of head and back by dusky 
or rusty hairs; feet and underparts white; tail blackish above, grayish 
below. 

SJcull. — Similar in general to that of N. intermedia, but very much 
larger and more angular; nasals narrower posteriorly; supraorbital 
ridges more prominent. In size the skull is similar to that of N. i. 
arenacea, but dentition heavier; bullae much larger. 

Measurements. — Type: Total length, 370; tail vertebrae, 165; hind 
foot, 41. Average of 6 adult topotypes : Total length, 374; 166; 39.3. 
Skull (average of same): Basilar length, 39.4; zygomatic breadth, 
24.6; interorbital breadth, 5.8; length of nasals, 18.2; length of 
incisive foramina, 10.5; length of palatal bridge, 7.6; alveolar length 
of upper molar series, 9.3. 

Remarks. — This large form occupies a position between intermedia 
and arenacea and probably intergrades with both. In cranial char- 
acters it approaches N. bryanti, but differs markedly in color. Adult 
specimens taken in November are undergoing the change from worn 
to fresh pelage. 



1910.] INTERMEDIA GROUP ARENACEA. 47 

Specimens examined. — Total number 41, from localities in Lower 
California, as follows: 

Magdalena Island, 12; Margarita Island, 10; Matancita (type locality), 12; 
San Jorge, 7. 

NEOTOMA INTERMEDIA ARENACEA Allen. 

Cape Wood Rat. 

(PI. Ill, figs. 5, 5a.) 

Neotoma arenacea Allen, Bull. Amer. Mus. Nat. Hist., X, pp. 150-151, April 12, 1898. 
Type from San Jose del Cabo, Lower California, Mexico; No. 98, 3,112, $ , British 
Museum; collected by Dane Coolidge, August 6, 1896. 

Distribution. — Coastal plains and basal mountain slopes in the 
Cape region of Lower California, north at least to La Paz. Mainly 
overlapping portions of Lower Sonoran and Arid Tropical zones. 

General characters. — Size much larger than N. intermedia. Related 
to N. i. cursoria, but color darker; tail longer; audit al bullae much 
smaller. 

Color. — Fresh pelage: Upper parts between buff and ochraceous 
buff, clearest along cheeks and sides, moderately darkened over top 
of head and back by overlying dusky hairs; feet and under parts 
white; ankles brownish or dusky; tail indistinctly bicolor, brownish 
above, grayish brown below. 

Measurements. — Average of 5 adult topotypes: Total length, 383; 
tail vertebrae, 188; hind foot, 38.5. Skull (average of same): Basilar 
length, 38.8; zygomatic breadth, 23.7; interorbital breadth, 5.9; 
length of nasals, 17.4; length of . incisive foramina, 10.2; length of 
palatal bridge, 7.2; alveolar length of upper molar series, 8.2. 

Skull. — Similar in general form to that of N. intermedia, but larger 
and more angular; supraorbital ridges more prominent; bullae rela- 
tively much smaller. In size the skull is similar to that of N. i. 
cursoria, but dentition much lighter and bulla? much smaller. 

Remarks. — This large form inhabits the subtropical coastal plains 
and valleys and the Lower Sonoran basal slopes of the Sierra la 
Laguna in the Cape region of Lower California from La Paz to Cape 
San Lucas. Complete intergradation with intermedia probably occurs 
to the northward of La Paz and on the slopes of the Sierra la Laguna. 
Some of the specimens taken early in January are in fresh pelage, 
while others are much worn. 

Specimens examined. — Total number 36, all from the Cape region 
of Lower California, as follows : 

Cape San Lucas, 8; El Sauz, 1; La Paz, 11; San Jose del Cabo (type locality), 
14; Santa Anita, 2. 



48 NORTH AMERICAN FAUNA. [no. 31. 

NEOTOMA INTERMEDIA VICINA Goldman. 

Espiritu Santo Wood Rat. 

Neotoma intermedia vicina Goldman, Proc. Biol. Soc. Wash., XXII, p. 140, June 25, 
1909. Type from Espiritu Santo Island, off east coast of Lower California, Mex- 
ico; No. 146803, $ ad., U. S. National Museum (Biological Survey Collection); 
collected by E. W. Nelson and E. A. Goldman, February 9, 1906. 

Distribution. — Espiritu Santo Island, Gulf of California, Mexico. 
Lower Sonoran Zone and upper border of Arid Tropical Zone. 

General characters. — Most nearly related to N. i. perpallida, but 
color darker and tail averaging slightly shorter. Size slightly larger 
than N. intermedia; color above more grayish; under parts purer 
white; tail more scantily haired; skull differing in various details. 
Somewhat similar to N. i. arenacea, but smaller; color paler; audital 
bullae relatively larger. 

Color. — Fresh pelage: Upper parts dull grayish or brownish drab, 
lighter along cheeks and sides, moderately darkened over top of 
head and back by overlying dusky hairs; feet and under parts white; 
tail blackish above, grayish below. 

SJcull. — About as in N. i. perpallida; averaging larger and more 
angular than in N. intermedia; nasals narrower posteriorly; inter- 
pterygoid fossa wider; dentition heavier; bullae slightly smaller. The 
skull differs from that of N. i. arenacea in smaller size, relatively 
heavier dentition, and relatively larger bullae. 

Measurements. — Type: Total length, 335; tail vertebrae, 156; hind 
foot, 37. Average of 5 adult topotypes: 344; 162; 34.5. 

Skull (average of 4 adults): Basilar length, 35.6; zygomatic 
breadth, 22.8; interorbital breadth, 5.6; length of nasals, 16.4; length 
of incisive foramina, 9.3 ; length of palatal bridge, 7.2 ; alveolar length 
of upper molar series, 8.1. 

Remarks. — This island form differs from N. i. perpallida mainly in 
darker color. The close agreement of vicina and perpallida in gen- 
eral characters seems rather remarkable in view of the fact that the 
island of San Francisco, lying between their insular ranges, is inhab- 
ited by the very distinct species N. abbreviata. 

Specimens examined. — Thirteen, all from Espiritu Santo Island, 
Lower California. 

NEOTOMA INTERMEDIA PERPALLIDA Goldman. 

San Jose Island Wood Rat. 

Neotoma intermedia perpallida Goldman, Proc. Biol. Soc. Wash., XXII, pp. 139-140, 
June 25, 1909. Type from San Jose Island, off east coast of Lower California, Mex- 
ico; No. 79061, $ yg. ad., U. S. National Museum (Biological Survey Collection); 
collected by J. E. McLellan, August 4, 1895. 



1910.] INTERMEDIA GEOUP ABBREVIATA. 49 

Distribution. — San Jose Island, Gulf of California, Mexico. Lower 
Sonoran Zone. 

General characters. — Most nearly related to N. i. vicina, but color 
still paler and tail averaging slightly longer; skull about the same. 
Similar in color to N. i. gilva, but under parts purer white and tail 
not basally buffy; tail longer, more scantily haired; skull differing in 
details. 

Color. — Fresh pelage: Upper parts dull grayish or brownish drab, 
lighter along cheeks and sides, moderately darkened over top of head 
and back by overlying dusky hairs; feet and under parts white; tail 
blackish above, grayish below. Worn pelage: Upper parts becoming 
browner or more rusty, mainly through fading of dark hairs. 

Skull. — About like that of N. i. vicina. Much as in N. intermedia 
and N. i. gilva, but somewhat larger; bulla? slightly smaller. Dif- 
fering from that of N. nudicauda in much smaller bullae. 

Measurements. — Average of six adults: Total length, 359 (345-375). 
STcull ( average of four adults) : Basilar length, 31 ; zygomatic breadth, 
22; interorbital breadth, 5.7; length of nasals, 16.8; length of incisive 
foramina, 9.4; length of palatal bridge, 7.2; alveolar length of upper 
molar series, 8.6. 

Remarks. — The San Jose Island wood rat differs from N. i. vicina of 
Espiritu Santo Island mainly in paler color. This difference in color 
is within the range of variation in some mainland forms, and but for 
their insularity perpallida and vicina might be included under one 
name. The small island of San Francisco, lying only a short distance 
off the southern end of San Jose Island, is inhabited by N. abbreviata, a 
distinct species with which perpallida requires no close comparison. 

The Biological Survey collection contains two specimens of this 
form. Outram Bangs has kindly furnished for examination a fine 
series of additional specimens — 11 skins and skulls collected on San 
Jose Island by W. W. Brown, jr., for John E. Thayer. 

Specimens examined. — Thirteen, all from San Jose Island, Lower 
California. 

NEOTOMA ABBREVIATA Goldman. 

Short-tailed Wood Rat. 

Neotoma abbreviata Goldman, Proc. Biol. Soc. Wash., XXII, pp. 140-141, June 25, 1909. 
Type from San Francisco Island (near southern end of San Jose Island), Gulf of 
California; No. 12260, $ ad., Museum of Comparative Zoology; collected by 
W. W. Brown, jr., February 22, 1909. (John E. Thayer expedition.) 

Distribution. — Known only from San Francisco Island, in the Gulf 
of California. Lower Sonoran Zone. 

General characters. — Size small; color very pale; tail decidedly 
shorter than head and body— shorter than in any other known mem- 
ber of the intermedia group. Somewhat like N. intermedia, but size 
52668°— 10 4 



50 NORTH AMERICAN FAUNA. [no. 31. 

smaller, color very much paler, tail relatively shorter and skull very 
different. Similar in color to N. i. perpallida of the neighboring island 
of San Jose, but not very nearly related; size smaller; tail relatively 
shorter; skull differing in numerous details. 

Color. — Fresh pelage: Upper parts pale drab gray, palest along 
cheeks and sides, the dark hairs over top of head and back scarcely 
numerous enough to alter the general color, but producing a grizzled 
effect; posterior part of back suffused with cinnamon or pale rusty; 
under parts dull white, the fur basally plumbeous; feet white; tail 
pale grayish above, white below. 

Slcull. — Somewhat like that of intermedia, but smaller, more angu- 
lar; nasals broader posteriorly, the ends more rounded; frontals nar- 
rower posteriorly, the constriction nearer middle, upper surface more 
deeply channeled along median line; bullae shorter, more rounded; 
posterior border of palate slightly emarginate (evenly convex in inter- 
media) ; mastoid process of squamosal less spatulate at tip, or less 
developed upward, leaving mastoid more exposed below lateral exten- 
sion of supraoccipital ; supraoccipital more angular, sharply ridged 
posterior to union with interparietal; outer surface of squamosal with 
a distinct eminence or ridge descending obliquely toward orbit. Com- 
pared with that of perpallida the skull of abbreviata is decidedly 
smaller; nasals broader and more rounded posteriorly; frontals rela- 
tively narrower posteriorly, the constriction nearer middle, broad 
anteriorly between lachrymals, and tapering more abruptly between 
maxillae and premaxillae owing to crowding by nasals; incisive 
foramina more decidedly longer than palatal bridge; bullae shorter, 
more rounded. 

Measurements. — Type: Total length, 305; tail vertebrae, 130; hind 
foot, 33. Average of six adult topotypes, 304 (295-310); 130 
(125-135); 33 (30-35). Skull (average of five adults): Basilar 
length, 34.1; zygomatic breadth, 21.6; interorbital breadth, 5.3; 
length of nasals, 15.4; length of incisive foramina, 8.4; length of pal- 
atal bridge, 7.5; alveolar length of upper molar series, 8.4. 

Remarks. — Although clearly a member of the intermedia group, this 
small insular wood rat is a distinct species, separable from its nearer 
relatives by small size and abbreviated tail. It is known only from 
the small island of San Francisco, situated a short distance from the 
southern extremity of the larger island of San Jose. When the slight 
geographic separation of these two islands is considered, the wide 
difference between the wood rats inhabiting them seems remarkable. 

I was indebted to Outram Bangs for the opportunity to describe the 
present species, which is based on a fine series of specimens collected 
for John E. Thayer. 

Specimens examined. — Ten, all from San Franeisco Island. 



1910.] INTERMEDIA GROUP BRYAN TI. 51 

NEOTOMA NUDICAUDA Goldman. 

Carmen Island Wood Rat. 

Neotoma nudicauda Goldman, Proc. Biol. Soc. Wash., XVIII, pp. 28-29, February 2, 
1905. Type from Carmen Island, Lower California, Mexico; No. 79073, 9 yg. 
ad., U. S. National Museum (Biological Survey Collection); collected by J. E. 
McLellan, October 14, 1895. 

Distribution. — Known only from Carmen Island, in the Gulf of Cali- 
fornia, Mexico. Lower Sonoran Zone. 

General characters. — Size medium; ears large; tail moderately long 
and nearly naked. Similar in general to N. i. arenacea, but color 
paler; audital bullae decidedly larger; interpterygoid fossa much 
narrower. 

Color. — Type (worn pelage) : Upper parts pale grayish buff, tinged 
with brownish along cheeks and sides, the back slightly darkened by 
black-tipped hairs; under parts white, the fur basally along sides of 
belly pale plumbeous; ears grayish brown; feet white; tail bicolor, 
brownish above, dull whitish below. 

Skull. — Similar to that of N. arenacea, but bullae much larger; inter- 
pterygoid fossa much narrower; maxillary arm of zygoma heavier; 
antorbital foramina larger; nasals truncate posteriorly, reaching 
plane of orbits; jugal rather long; dentition of the intermedia type, 
but rather light; supraorbital ridges well developed and sharp, as in 
N. i. arenacea. 

Measurements. — Type: Total length, 330; tail vertebra?, 155; hind 
foot, 40. Skull (type): Basilar length, 36; zygomatic breadth, 22.6; 
length of nasals, 16.3; length of incisive foramina, 9.3; length of 
palatal bridge, 7.1; alveolar length of upper molar series, 8.9. 

Remarks. — In external appearance nudicauda closely resembles 
perpallida, which inhabits the neighboring island of San Jose. In 
combination of cranial characters, however, it differs from all the other 
members of the intermedia group, and compares about as well with 
cursoria or vicina as with arenacea. 

Specimens examined. — One, the type. 

NEOTOMA BRYANTI Merriam. 

Bryant Wood Rat. 

(PI. Ill, figs. 6, 6a.) 

Neotoma bryanti Merriam, Amer. Nat., XXI, No. 2, pp. 191-193, February, 1887. 
Type from Cedros Island (also called Cerros Island), Lower California, Mexico; 
No. ffll) <? » Merriam Collection; collected by Walter E. Bryant. 

Distribution. — Cedros Island, off west coast of Lower California, 
Mexico. Upper Sonoran Zone. 

General characters. — Size very large; ears large; tail moderately 
long, bicolored. Related to N. i. cursoria, but color much darker 
and cranial characters different. 



52 NORTH AMERICAN FAUNA. [no. 31. 

Color. — Slightly worn pelage: Ground color of upper parts rich 
creamy buff, clearest along cheeks and sides, becoming much darkened 
over top of head and back by admixture of black-tipped hairs ; under 
parts creamy white, suffused in some specimens with pinkish buff; 
ankles more or less conspicuously dusky; feet white; tail brownish 
above, grayish below. 

Skull. — Similar to that of N. i. cursoria, but averaging larger and 
more angular; frontals broadening more abruptly anteriorly between 
lachrymals, upper surface deeply channeled, lateral margins strongly 
upturned in nearly straight posteriorly diverging lines ; interpterygoid 
fossa narrower ; dentition heavier. 

Measurements. — Average of 8 adult topotypes : Total length, 377 ; 
tail vertebrae, 168; hind foot, 38.1. Skull (average of 5 adults): 
Basilar length, 40; zygomatic breadth, 24.5; interorbital breadth, 
5.4; length of nasals, 18; length of incisive foramina, 9.9; length of 
palatal bridge, 8; alveolar length of upper molar series, 9.8. 

Remarks. — This large insular species is more nearly related to N. i. 
cursoria, which inhabits the coast and islands farther south, than to 
the form found on the adjacent mainland. The skulls of bryanti and 
cursoria indicate rather close relationship, but the darker color of the 
former readily distinguishes it. The fine series of topotypes in the 
Biological Survey collection shows that the species has whitish under- 
parts. The concolor back, sides, and belly in the type specimen are 
due to the burning off of the tips of the hairs, thus exposing the 
plumbeous basal color above and below. 

Specimens examined. — Total number 27, all from Cedros Island, 
Lower California. 

NEOTOMA ANTHONYI Allen. 

Anthony Wood Rat. 

(PI. IV, figs. 1, la.) 

Neotoma anlhonyi Allen, Bull. Amer. Mus. Nat. Hist., X, pp. 151-152, April 12, 1898. 
Type from Todos Santos Island, Lower California, Mexico; No. xMffi $ ac l-> 
American Museum of Natural History, N. Y.; collected by A. W. Anthony, May 
11, 1897. 

Distribution. — Todos Santos Island, off west coast of Lower Cali- 
fornia, Mexico. Upper Sonoran Zone. 

General characters. — Size medium; tail rather short, well haired, 
sharply bicolor; outer surface of hind legs conspicuously blackish; 
nasals not reaching plane of lachrymals. Related to N. intermedia, 
but differing both in color and cranial characters. 

Color. — Worn pelage: Above grayish brown, palest along sides, 
becoming much darker over top of head and back by admixture of 
dusky hairs; under parts dull whitish, faintly suffused across throat 



1910.] INTERMEDIA GEOUP — MARTINENSIS. 53 

and belly with pale pinkish buff; upper surface of forearms dusky 
brown ; outer surface of hind legs and inner surface of ankles blackish ; 
feet white; tail bicolor, brownish black above, soiled whitish below. 

Skull. — Similar in general to that of N. intermedia, but larger and 
heavier; nasals broader; maxillary arm of zygoma heavier, less 
squarely spreading; frontals more elevated near lachrymals, deeply 
channeled along median line; incisive foramina longer, slightly con- 
stricted near maxillopremaxillary suture; interpterygoid fossa wider; 
basisphenoid broader anteriorly; presphenoid broader posteriorly at 
point of union with basisphenoid; dentition of the intermedia type, 
but heavier. 

Measurements. — Average of 10 adult topotypes : Total length, 338; 
tail vertebrae, 149; hind foot, 36.3. Skull (average of 8 adults): 
Basilar length, 37.3; zygomatic breadth, 23.6; interorbital breadth, 
5.6; length of nasals, 17.1; length of incisive foramina, 10.1; length 
of palatal bridge, 7.5; alveolar length of upper molar series, 8.2. 

Remarks. — In color this species closely resembles N. martinensis, 
but the skull shows nearer relationship to N. intermedia, the animal 
inhabiting the adjacent mainland. The conspicuous black markings 
on the outer sides of the hind legs, as in anthonyi and martinensis, 
have not been observed elsewhere in the genus. 

Specimens examined. — Total number 50, all from Todos Santos 
Island, Lower California. 

NEOTOMA MARTINENSIS Goldman. 

San Martin Island Wood Rat. 

(PI. IV, figs. 2, 2a.) 

Neotoma martinensis Goldman, Proc. Biol. Soc. Wash., XVIII, p. 28, February 2, 1905. 
Type from San Martin Island, Lower California, Mexico; No. 81074, 9 ad. U. S. 
National Museum (Biological Survey Collection); collected by A. W. Anthony, 
July 17, 1896. 

Distribution. — San Martin Island, off west coast of Lower Cali- 
fornia, Mexico. Upper Sonoran Zone. 

General characters. — Size medium; tail moderately long, thinly 
covered with short hairs ; ears large ; outer sides of hind legs conspicu- 
ously blackish. In color closely resembling N. anthonyi, but 
underparts purer white; ears larger; tail more scantily haired; cranial 
characters very different. 

Color. — Rather worn pelage: Ground color of upper parts creamy 
buff, purest along lower parts of sides, heavily darkened on head and 
over back by overlying dusky hairs; under parts creamy white, the 
fur everywhere deep plumbeous basally; upper sides of forearms 
dusky brown; outer sides of hind legs and inner sides of ankles con- 
spicuously blackish; feet, including the sides of the soles to point well 



54 NORTH AMERICAN FAUNA. [no. 81. 

above heels on hind feet pure white; tail brownish black above, 
grayish below. 

Skull. — Size medium, rather smoothly rounded, high and well 
arched across anterior roots of zygomata; temporal ridges faintly 
developed and widely separated; nasals very long, abruptly narrow- 
ing posteriorly, and reaching well beyond plane of lachrymals; 
ascending branches of premaxillse reaching posteriorly beyond nasals 
nearly to interorbital constriction, the ends beveled internally and 
somewhat divaricating; frontals narrowly constricted between orbits, 
broadening posteriorly, the upper surface nearly flat and without 
well-developed lateral ridges; zygomata narrower anteriorly than 
posteriorly; palate truncate or slightly convex posteriorly ; interptery- 
goid fossa narrow and further constricted posteriorly by curvature 
inward of pterygoids; bullae rather small, somewhat pear-shaped, 
the meatus very large and widely open, with upper posterior border 
projecting laterally beyond upper anterior border; dentition of the 
intermedia type. 

Measurements. — Average of 4 adult topo types: Total length, 352; 
tail vertebrae, 165; hind foot, 37.6. Skull (average of same): Basilar 
length, 37.4; zygomatic breadth, 23.2; interorbital breadth, 5.5; 
length of nasals, 17.9; length of incisive foramina, 9.7; length of 
palatal bridge, 7.4 ; alveolar length of upper molar series, 8.3. 

Remarks. — This species, a member of the intermedia group, closely 
approaches anthnnyi in external appearance, but the skull indicates 
no very near relationship to any known form. The remarkable pro- 
longation of the nasals posteriorly beyond the plane of the lachry- 
mals, while in anthonyi they end decidedly in front of this plane, is 
one of numerous cranial characters which separate the two species. 
The conspicuously black hind legs distinguish both species externally 
from all the other known members of the genus. 

Specimens examined. — Total number 24, all from San Martin 

Island. 

NEOTOMA MEXICANA Baird. 

Mexican Wood Rat. 

(PI. IV, figs. 3, 3a.) 

Neotoma mexicana Baird, Proc. Acad. Nat. Sci. Phila., VII, p. 333, April, 1855. Type 
from near Chihuahua, Chihuahua, Mexico. Skull No. 1674 (skin No. 289 lost), 
U. S. National Museum; collected by John Potts. 

Distribution. — Desert ranges along the eastern side of the Sierra 
Madre in Chihuahua and northwestern Durango, and thence north- 
ward in the mountains to western Texas, southwestern New Mexico, 
and southeastern Arizona. Upper Sonoran and Transition zones. 

General characters. — Size medium, tail moderately long, bicolor; 
upper parts grayish ; first upper molar with anterointernal reentrant 
angle deeper than in N. floridana; frontals constricted near middle. 



1910.] 



MEXICANA GROUP — MEXICANA. 



55 



Color. — Fresh pelage: Upper parts grayish burl' or buff gray, palest 
on head, moderately darkened over back by overlying blackish hairs, 
becoming in worn pelage of old adults somewhat rusty brown; 
underparts dull white, the fur everywhere deep plumbeous basally; 
feet white; tail brownish above, white below. Young (nearly full 
grown): Above varying shades of buffy gray thinly mixed with 
black. 

Skull. — Rather small and light; rostrum relatively longer and 
more slender than in N. Jloridana or N. albigula; zygomata more 




Fig. 6.— Distribution of Neotoma mexicana and subspecies. 

squarely spreading; the sides nearly parallel; nasals ending poste- 
riorly near anterior plane of orbits, only slightly exceeded by pre- 
maxillae (nasals and premaxillse reaching farther posteriorly in fallax) ; 
frontals narrower posteriorly than in fallax, constricted near middle; 
interparietal rectangular; palate truncate or slightly convex pos- 
teriorly; interpterygoid fossa narrow; bullae of medium size, short 
and well rounded ; dentition much as in N. jloridana and N. albigula, 



56 NORTH AMERICAN FAUNA. [no. 31. 

but first upper molar with anterointernal reentrant angle deeper, 
cutting more than half way across anterior lobe. 

Measurements. — An adult from Santa Eulalia (near type locality), 
Chihuahua: Total length, 327; tail vertebrae, 149; hind foot, 34. 
Skull (of same): Basilar length, 35.6; zygomatic breadth, 23.3; 
interorbital breadth, 5.3; length of nasals, 16.6; length of incisive 
foramina, 9.1; length of palatal bridge, 8.2; alveolar length of upper 
molar series, 8.3. 

Remarks. — N. mexicana is a typical member of a large group, the 
differentiation of which into a number of closely related species and 
subspecies has been determined by the varying conditions of envi- 
ronment along the principal mountain ranges of the continent from 
Colorado to Nicaragua. From N. ferruginea of Guatemala it dif- 
fers widely in color (ferruginea is bright orange buff), but rather 
near relationship is shown by agreement in many important cranial 
characters, and the two species almost meet in Zacatecas and Tepic 
through the intergradation of intervening forms. N. mexicana grades 
into fallax in southern New Mexico and into madrensis along the 
eastern slopes of the Sierra Madre in western Chihuahua and Du- 
rango. Externally mexicana and albigula are much alike, but the 
former may be known by the plumbeous instead of pure white basal 
color of fur on throat and breast. In the last lower molar a small 
accessory anteroexternal reentrant angle is sometimes present. 
This character occurs sporadically in nearly all of the members of 
the mexicana group. 

Specimens examined. — Total number 24, from the following local- 
ities : 

Chihuahua: Chihuahua, 1; Santa Eulalia (near Chihuahua), 3. 

Durango: La Cienega de las Vacas (northwestern Durango), 1. 

Texas: Fort Davis, 2; Guadalupe Mountains, 3; Paisano, 1; Valentine, 1. 

Arizona: Chiricahua Mountains, 3; Rincon Mountains, 1. 

New Mexico: Animas Peak, 8. 

NEOTOMA MEXICANA FALLAX Merriam. 

Colorado Wood Rat. 

(PI. IV, figs. 4, 4a.) 

Neotoma fallax Merriam, Proc. Biol. Soc. Wash., IX, pp. 123-124, July 2, 1894. Type 
from Gold Hill, Boulder County, Colo.; No. ffjf, $ ad., Merriam Collection; 
collected by Denis Gale, November 1, 1889. 

Distribution. — Mountains of Colorado and northern and central 
New Mexico. Upper Sonoran and Transition zones. 

General characters. — Size averaging larger than N. mexicana; color 
similar; closely related to N. m. pinetorum, but grayer and differing 
slightly in cranial characters. 



1910.] MEXICAN A GROUP FALLAX. 57 

Color. — Fresh pelage (October): Upperparts creamy buff or pale 
grayish (becoming more or less rusty brown in worn pelage), clearest 
along sides, darkened over top of head and back by black-tipped 
hairs; feet white; underparts white, the fur everywhere basally 
plumbeous; tail grayish or brownish above, white below. Young 
(one-third grown) : Above grayer than in pinetorum of similar age. 

SJcull. — Larger than that of N. mexicana; nasals and premaxillae 
reaching farther posteriorly beyond anterior plane of orbits; frontals 
broader posteriorly. Closely resembling that of N. m. pinetorum, 
but slightly smaller; nasals averaging broader posteriorly and reach- 
ing farther beyond anterior plane of orbits; frontals broader poste- 
riorly ; third lower molar in many (but not all) skulls with two outer 
reentrant angles, the anterior very small and disappearing with 
advancing age. 

Measurements. — Average of 3 adult topotypes: Total length, 331; 
tail vertebrae, 150; hind foot, 33. Average of 9 adults from Love- 
land, Colo.: 334, 144, 33.5. SJcull (an adult topotype): Basilar 
length, 35.8; zygomatic breadth, 22.2; interorbital breadth, 5.2; 
length of nasals, 17.3; length of incisive foramina, 9.5; length of 
palatal bridge, 7.7; alveolar length of upper molar series, 8.9. 

Remarks.— This is a variable form approaching mexicana in south- 
ern New Mexico and grading into pinetorum in western New Mexico. 
Specimens from Copperton, Mount Taylor, and Grant, N. Mex., are 
dichromatic, some being of the usual color, while others are much 
darker, with a pinkish buffy band across the chest, and the belly 
more or less distinctly washed with this color. A young adult indi- 
vidual from Grant is dark slaty above with a pinkish buffy wash 
overspreading entire underparts. The additional outer reentrant 
angle often present in the third lower molar occurs irregularly in 
nearly all of the members of the mexicana group. Externally fallax 
may be known from albigula, which often occurs at the same local- 
ities, by the plumbeous basal color of the fur on the throat and breast. 

Specimens examined. — Total number 278, from localities as follows : 

Colorado: Arboles, 2; Arkins, 2; Ashbaugh Ranch, Montezuma County, 1; 
Boulder, 4; Canyon City, 5; Colorado Springs, 1; Coventry, 2; Fisher 
Peak, 1; Fort Collins (5 miles southwest), 1; Gold Hill (type locality), 21; 
Grand Junction, 2; Loveland, 16; Martinsen, 1; Trinidad, 5. 

New Mexico: Arroyo Seco, 1; Beal (Upper Vermejo River), 1; Boulder Lake, 
1; Burley, 1; Capitan Mountains, 3; Catskill (7 miles southwest), 7; 
Chama River (25 miles southwest of Tierra Amarilla), 1; Chusca Moun- 
tains, 7; Clayton, 9; Copperton, 15; Coyote Creek (8 miles north of Gua- 
dalupita), 1; Datil Mountains, 5; Folsom, 13; Fruitland, 2; Gallina, Rio 
Arriba County, 1; Gallina Mountains, Socorro County, 3; Gallinas Moun- 
tains, Rio Arriba County, 2; Gallo Canyon (35 miles southwest of Corona), 
1; Gallup, 2; Glorieta, 3; Gran Quivera, 1; Grant, 2; Halls Peak, 7; Horse 
Lake, 5; Hoskin's Ranch, Colfax County, 1; Jicarilla Mountains, 3; Lamy, 
1; Manzano Mountains, 52; Monica Canyon, San Mateo Mountains, 5; 



58 NORTH AMERICAN FAUNA. [no. 31. 

Mora, 1; Mount Sedgwick, 1; Mount Taylor, 10; Pecos (10 miles north), 
1; Pecos River (8 miles north of Pecos), 2; Rio Puerco, 1; Sandia Moun- 
tains, 5; San Mateo Peak, 1; Santa Rosa, 3; Sierra Grande, 9; Stinking 
Spring Lakes, 9; Tres Piedras, 1; Twining, 1; Vermojo Park (near Beal), 
1; White Mountains, Lincoln County, 1; Willis, 9; Zuni Mountains (Bear 
Ridge), 2. 

NEOTOMA MEXICAN A PINETORUM Merriam. 

San Francisco Mountain Wood Rat. 

(PI. IV, figs. 5, 5a.) 

Neotoma pinetorum Merriam, Proc. Biol. Soc. Wash., VIII, pp. 111-112, July 31, 
1893. Type from San Francisco Mountain, Arizona (altitude 8,000 feet); No. 
Illlf, 9 ad., U. S. National Museum (Biological Survey Collection); collected by 
Vernon Bailey, August 16, 1889. 

Distribution. — Plateau region from San Francisco Mountain, 
Arizona, north to the Grand Canyon and southeastward along the 
Mogollon Mesa to the Mogollon and Mimbres Mountains in western 
New Mexico. Transition Zone. 

General characters. — Size larger than N. mexicana; color less gray, 
more ochraceous buff. Closely related to N. m. fallax, but color 
darker, more ochraceous buff, and cranial characters slightly dif- 
ferent. 

Color. — Fresh pelage (October) : Upper parts pale ochraceous buff 
(becoming dull ochraceous buff in worn pelage), clearest along 
cheeks and sides, darkened over top of head and back by black- 
tipped hairs; feet white; under parts white; the fur everywhere 
plumbeous basally; tail brownish or blackish above, white below. 
Young (one- third grown) : Above more buffy than in fallax of cor- 
responding age. 

Skull. — Skull decidedly larger than that of N. mexicana; nasals 
relatively narrower posteriorly. Closely resembling that of N. m. 
fallax, but slightly larger; nasals usually narrower posteriorly, end- 
ing near anterior plane of orbits (usually passing well beyond this 
plane in N. m. fallax) ; frontals relatively narrower posteriorly. 

Measurements. — Average of 4 adult topotypes: Total length, 357; 
tail vertebrae, 164; hind foot, 36.5. Skull (average of same): Basi- 
lar length, 38.8; zygomatic breadth, 24.7; interorbital breadth, 5.4; 
length of nasals, 18.9; length of incisive foramina, 9.3; length of 
palatal bridge, 8.6; alveolar length of upper molar series, 9.4. 

Remarks. — Typical pinetorum seems to be confined to San Francisco 
Mountain, but specimens from the Mogollon plateau region and east- 
ward to the Mogollon and Mimbres Mountains of western New Mexico 
agree closely in color and are referable to it. The cranial characters 
distinguishing pinetorum and fallax are, except near the type locali- 



1910.1 MEXICANA GROUP BULLATA. 59 

ties, variable and somewhat unreliable. This is especially true in 
central western New Mexico, where the two seem to merge and are 
separable only by color, pinetorum being somewhat darker, less pale 
grayish than fallax. Topotypes taken in August are in worn summer 
pelage; those taken in October are in the fresh fall coat. 
Specimens examined. — Total number 72, from: 

Arizona: Escudilla Mountains, 1; Flagstaff, 1; Grand Canyon, 5; San Francisco 
Mountain (type locality), 8; Springerville, 12; Walnut, 1; White Moun- 
tains, 5. 

New Mexico: Beaver Lake, 2; Chloride (10 miles west), 3; Diamond Creek, 
Gila National Forest, 2; Kingston, 2; Luna, Gila National Forest, 3; 
Mimbres Mountains (Big Rocky Creek), 1; Mimbres River (head), 5; 
Mogollon Mountains, 15; Quemado (10 miles southwest), 2; San Francisco 
Mountains, 4. 

NEOTOMA MEXICANA BULLATA Merriam. 

Santa Catalina Mountain Wood Rat. 

(PI. IV, figs. 6, 6a.) 

Neotoma mexicana bullata Merriam, Proc. Biol. Soc. Wash., IX, pp. 122-123, July 
2, 1894. Type from Santa Catalina Mountains, Arizona; No. 16863, $ ad., U. S. 
National Museum (Biological Survey Collection); collected by Vernon Bailey, 
June 1, 1889. 

Distribution. — Known only from the type locality. Transition 
Zone. 

General characters. — Similar to N. mexicana, but color more ochra- 
ceous buffy; bullae peculiar. 

Color. — Slightly worn pelage: Upper parts pale ochraceous buff, 
brightest along cheeks, shoulders, and sides, rather well darkened by 
admixture of dusky hairs, which are most abundant over top of head 
and back; under parts white, the fur basally plumbeous; axillae 
ochraceous buff; pectoral region crossed by faint buffy band; feet 
white, tail grayish brown above, whitish below. 

Skull. — Closely resembling that of N. mexicana, but bullae slightly 
smaller, more tapering anteriorly, and curving inward; dentition as 
in N. mexicana. 

Measurements. — Type: Total length, 335; tail vertebrae, 151; hind 
foot, 34. Skull (type): Basilar length, 35.3; zygomatic breadth, 23; 
interorbital breadth, 5.5; length of nasals, 17.1; length of incisive 
foramina, 9.3; length of palatal bridge, 8; alveolar length of upper 
molar series, 9.1. 

Remarks. — This form is evidently closely related to N. mexicana and 
probably intergrades with it. 

Specimens examined. — One, the type. 



60 NORTH AMERICAN FAUNA. [no. 31. 

NEOTOMA MEXICANA MADRENSIS Goldman 

Sierra Madre Wood Rat. 

Neotoma mexicana madrensis Goldman, Proc. Biol. Soc. Wash., XVIII, p. 31, Febru- 
ary 2, 1905. Type from Sierra Madre, near Guadalupe y Calvo, Chihuahua, 
Mexico; No. 95244, 9 ad., U. S. National Museum (Biological Survey Collection); 
collected August 26, 1898, by E. W. Nelson and E. A. Goldman. 

Distribution. — The Sierra Madre from northwestern Chihuahua 
and northeastern Sonora southward to western Zacatecas. Tran- 
sition Zone. 

General characters. — Size smaller than N. mexicana; color much 
darker, cinnamon instead of grayish; tail long, slender, and thinly 
haired; ears rather large. 

Color. — Worn pelage (August) : Above pale cinnamon, purest 
along sides, strongly darkened dorsally by black- tipped hairs; under- 
parts dull white, the hairs everywhere plumbeous basally; axilla? 
ochraceous buff; feet white; ears brownish; tail distinctly bicolor, 
brownish above, whitish or grayish below. Young (about half 
grown) : Much darker and more fulvous than in N. mexicana. 

Skull. — Similar to that of N. mexicana, but smaller; bulla? slightly 
smaller; first upper molar with anterointernal reentrant angle deep, 
as in N. mexicana. 

Measurements. — Average of 2 adult topotypes: Total length, 314; 
tail vertebra?, 144; hind foot, 33. SJcull (average of same): Basilar 
length, 34.8; zygomatic breadth, 21.7; interorbital breadth, 5.4; 
length of nasals, 17.2; length of incisive foramina, 8.6; length of 
palatal bridge, 8.2; alveolar length of upper molar series, 8.5. 

Remarks. — The range of this subspecies includes all the higher parts 
of the Sierra Madre from Sonora and Chihuahua to Zacatecas. Along 
the east slope of the mountains it probably grades into typical mex- 
icana, and along the west slope into sinalose. From the former it 
may readily be separated by its decidedly darker color at all ages 
and conditions of pelage, and from the latter by the much larger size 
of the audital bulla?. 

Specimens examined. — Total number 31, from localities as follows: 

Chihuahua: Colonia Garcia, 3; Pacheco, 4; near Parral, 2; Sierra Madre, near 

Guadalupe y Calvo (type locality), 8. 
Durango: Cerro Prieto, 1; Coyotes, 7; El Salto, 3; Sierra Madre, near 

Guanacevi, 1. 
Zacatecas: Valparaiso Mountains, 2. 

NEOTOMA MEXICANA SINALOA Allen. 

Sinaloa Wood Rat. 

Neatoma sinalose Allen, Bull. Amer. Mus. Nat. Hist., X, pp. 149-150, April 12, 1898. 
Type from Tatameles, Sinaloa, Mexico; No. 98.3.2.88, 9 a d., British Museum ; 
collected by P. 0. Simons, May 14, 1897. 



1910.] MEXICAN A GROUP NAVUS. 61 

Distribution. — Western slope of the Sierra Madre from southern 
Sinaloa northward to Sonora. Lower Sonoran Zone. 

General characters. — Size smaller than N. mexicana; color darker; 
audital bullae very small. 

Color. — Rather worn pelage: General color above pale cinnamon, 
moderately mixed with black, in worn pelage becoming russet in 
patches; face and sides paler, somewhat buffy; under parts dull whit- 
ish, the plumbeous basal color showing through strongly; feet white; 
ears brownish; tail bicolor, brownish black above, paler below. 

SJcull. — Similar to that of N. mexicana, but smaller; bulla? much 
smaller; tooth row shorter. 

Measurements. — Average of 2 adult topotypes: Total length, 316; 
tail vertebra?, 155; hind foot, 31. Skull (an adult topotype) : 
Basilar length, 33.9; zygomatic breadth, 21.8; interorbital breadth, 
5.7; length of nasals, 16.4; length of incisive foramina, 8.2; length of 
palatal bridge, 8.2; alveolar length of upper molar series, 7.8. 

Remarks. — This form differs markedly from N. mexicana, and is 
most closely related to subspecies madrensis, with which it probably 
intergrades along the higher of the western slopes of the Sierra Madre. 
Its very small bulla? serve to distinguish it from all the other sub- 
species of mexicana. 

Specimens examined. — Total number 8, from localities as follows: 

Sinaloa: Mazatlan, 1; Tatameles (type locality), 2. 
Durango: Chacala, 1. 
Sonora: Alamos, 4. 

NEOTOMA NAVUS Merriam. 

Coahuila Wood Rat. 

(PI. V, figs. 1, la.) 

Neotoma navus Merriam, Proc. Biol. Soc. Wash., XVI, pp. 47-48, March 19, 1903. 
Type from Sierra Guadalupe, Coahuila, Mexico; No. 116895, 9 ad., U.S. National 
Museum (Biological Survey Collection); collected by E. W. Nelson and E. A. 
Goldman, April 26, 1902. 

Distribution. — High mountains of southern Coahuila, Mexico. 
Transition Zone. 

General characters. — Size medium, tail rather long and slender. 
Similar to N. mexicana, but color more ochraceous buff; lateral 
borders of frontal platform projecting as supraorbital shelves; den- 
tition as in N. mexicana. 

Color. — Slightly worn pelage: Upper parts pale ochraceous buff, 
brightest along cheeks, shoulders, and sides; moderately darkened 
by blackish hairs, which are most abundant on middle of face, top of 
head, and along back; under parts white, the plumbeous basal color 
becoming very pale on throat and chest; axilla? ochraceous buff; 
feet white; tail dusky above, white below. 



62 



NORTH AMEEICAN FAUNA. 



[no. 3t. 



Skull. — Similar in general to that of N. mexicana, but frontal plat- 
form narrower anteriorly, the lateral borders bent inward, leaving 
the frontals bulging below them, broadening abruptly posteriorly, 
the borders forming convex, projecting supraorbital shelves; nasals 




FIG. 7. — Distribution of Neotoma navus, N. torquata, N. distincta, N. tropicalis, and N. parvidens. 

rounded or bluntly pointed instead of truncate posteriorly; inter- 
pterygoid fossa narrower, more rounded anteriorly (in mexicana trun- 
cate or encroached upon by blunt projection from palate); bullae 
more tapering anteriorly; dentition as in N. mexicana. 



1910.] MEXICANA GROUP TORQUATA. 63 

Measurements. — Average of 2 adults from the type locality: Total 
length, 340; tail vertebrae, 158; hind foot, 35. Skull (average of 
same): Basilar length, 34.8; zygomatic breadth, 21.3; interorbital 
breadth, 5.4; length of nasals, 16.2; length of incisive foramina, 9; 
length of palatal bridge, 8.1; alveolar length of upper molar series, 
8.6. 

Remarks. — Skulls of N. navus differ markedly from those of all the 
other members of the mexicana group in the peculiar shelf-like devel- 
opment of the frontal region. The species is known only from the 
Sierra Guadalupe, but probably includes in its range other high 
mountains of northeastern Mexico. 

Specimens examined. — Two, from the type locality. 

NEOTOMA TORQUATA Ward. 
Ward Wood Rat. 
(PI. V, figs. 2, 2a.) 

Neotoma torquata Ward, Amer. Nat., XXV, pp. 160-161, February, 1891. Type from 
tunnel of abandoned mine between Tetela del Volcan and Zacualpam Amilpas, 
Morelos, Mexico; No. 380, Museum of Comision Geografica Exploradora, Tacu- 
baya, Mexico; collected by H. L. Ward, October 26, 1890. 

Neotoma fulviventer Merriam, Proc. Biol. Soc. Wash., IX, pp. 121-122, July 2, 1894. 
Type from Toluca Valley, Mexico, Mexico; No. 50165, 9 ad., U. S. National 
Museum (Biological Survey Collection); collected by E. W. Nelson, November 
5, 1892. 

Neotoma orizabx Merriam, Proc. Biol. Soc. Wash., IX, p. 122, July 2, 1894. Type 
from Mount Orizaba, Puebla, Mexico; No. 53653, $ ad., U.S. National Museum 
(Biological Survey Collection); collected by E. W. Nelson, April 20, 1893. 

Distribution. — High mountains of south-central Mexico, from 
northern Hidalgo to eastern Puebla and Morelos. Upper Sonoran, 
Transition, and Boreal zones. 

General characters. — Size medium; ears small; fur of under parts 
basally plumbeous everywhere; tail sharply bicolor. Member of 
ferruginea section of mexicana group, nearest to N. f. tenuicauda, but 
larger; color more buffy ochraceous. 

Color. — Fresh pelage (February) : Ground color of upper parts 
ochraceous buff, brightest along sides, becoming grayish buff on head, 
heavily darkened, especially along back, by admixture of blackish; 
outer sides of legs drab, tinged with vinaceous buff; under parts dull 
white, more or less suffused in some specimens with ochraceous buff; 
axillae ochraceous buff, this color sometimes forming a pectoral band ; 
orbital rings black; forefeet white; hind feet usually white, but 
sometimes clouded with dusky to toes; tail brownish or blackish 
above, whitish below. Worn pelage (November) : Upper parts 
varying shades of pale cinnamon or rusty brown, mixed with black. 
Young (half grown) : Upper parts broccoli brown, tinged with buff, 
and much mixed with black. 



64 NORTH AMERICAN FAUNA. [no. 31. 

Skull. — In general form similar to that of N. f. tenuicauda, but 
larger; nasals longer, reaching plane of orbits; bullae larger; denti- 
tion of the mexicana type; first upper molar with anterointernal 
reentrant angle deep, as in N. mexicana and N. ferruginea. 

Measurements. — Type (from original description): Total length, 
338; tail vertebras, 160; hind foot, 35. Average of 5 adults from 
Salazar, Mexico (state): 349; 162; 34.5. Skull (type): Basilar 
length, 35.7; zygomatic breadth, 22.8; interorbital breadth, 6.1; 
length of nasals, 17.4; length of incisive foramina, 8; length of 
palatal bridge, 8; alveolar length of upper molar series, 9. 

Remarks. — -N. torquata is rather closely related to N. f. tenuicauda 
and may intergrade with it in eastern Michoacan. The skull of the 
type has broader frontals, shorter rostrum, and shorter incisive fora- 
mina than is usual in specimens from the same general region, but 
agrees closely with skulls from Tlalpam and Salazar. Specimens 
taken from December to April are in fresh pelage, and those from 
August to November show varying degrees of wear. 

Individual variation is greater in this species than in most members 
of the genus. The underparts are usually white, but may sometimes 
be ochraceous buffy in specimens taken at the same time and place. 
Among cranial variations the size of the molariform teeth is perhaps 
most notable. N. fulviventer was based on a specimen with ochra- 
ceous buffy underparts and small teeth, and N. orizabx on one with 
white underparts and unusually heavy dentition. Recently accu- 
mulated material shows that these remarkable differences are within 
the range of variation of N. torquata, and neither supposed species can 
be satisfactorily separated from it. 

Specimens examined. — Total number 57, as follows; 

Morelos: Between Tetela del Volcan and Zacualpam Amilpas, 1 (type). 

Federal District: Ajusco, 5; Tlalpam, 6. 

Hidalgo: Encarnacion, 7; Tulancingo, 1. 

Mexico: Amecameca, 1; Mount Popocatapetl, 2; Salazar, 13; Toluca Valley 

(near Lerma), 2; Volcano of Toluca (north slope), 2. 
Puebla: Chalchicomula, 2. 
Tlaxcala: Mount Malinche, 4. 
Veracruz: Cofre de Perote, 2; Mount Orizaba, 3; Xuchil, 6. 

NEOTOMA DISTINCTA Bangs. 

Veracruz Wood Rat. 

(PI. V, figs. 3, 3a.) 

Neotoma distincta Bangs, Proc. Biol. Soc. Wash., XVI, pp. 89-90, June 25, 1903. Type 
from Texolo (near Jalapa), Veracruz, Mexico; No. 9819, $ ad., Bangs Collection 
(Museum of Comparative Zoology); collected by S. N. Rhoads, March 8, 1899. 

Distribution. — Known only from the type locality. Humid Trop- 
ical Zone. 



1910.] MEXICANA GKOUP — TKOPICALIS. 65 

General characters. — Size large; ears small; tail long, unicolor; color 
very dark. Member of mexicana group related to N. torquata, but 
differing markedly in size, color, and cranial characters. 

Color. — Worn pelage: Ground color of upperparts rich russet, some- 
what suffused with Vandyke brown or burnt umber, clearest along 
cheeks and sides, becoming much darkened over top of head and 
back by longer blackish hairs; russet of sides encroaching on under- 
pays posteriorly, leaving a narrow yellowish white area along median 
line of belly; lower pectoral region dull gray; throat white; axillae 
and broad pectoral band ochraceous buff ; outer sides of forelegs hair 
brown, passing into grayish white of forefeet without sharp line of 
separation; nose, lips, and ankles dusky; hind feet to toes grayish 
brown, toes white; tail entirely black. 

Skull. — Similar in general to that of N. torquata, but larger and 
more massive; rostrum much heavier; interorbital region narrower; 
frontals excavated along median line above, the supraorbital margins 
convex posteriorly ; zygomata broadly spreading posteriorly, narrow- 
ing anteriorly (nearly parallel in torquata) ; interpterygoid fossa nar- 
rower; palate with short, median, posterior projection; bullae smaller: 
dentition heavier (incisors remarkably heavy). 

Measurements. — Average of 3 adult topotypes (from original descrip- 
tion): Total length, 394; tail vertebra?, 185; hind foot, 41. Skull 
(an adult topotype): Basilar length, 40.2; zygomatic breadth, 25.5; 
interorbital breadth, 5.5; length of nasals, 18.4; length of incisive 
foramina, 10.7; length of palatal bridge, 9.3; alveolar length of upper 
molar series, 10.2. 

Remarks. — This is a well-marked form most closely related to its 
near geographical neighbor, N. torquata. It lives in the humid trop- 
ical forests on the steep eastern slopes of the plateau region in Vera- 
cruz, while torquata occupies the mountain tops above and only a few 
miles away. The dark color of distincta harmonizes well with its dark 
forest environment. On the higher slopes along the eastern edge of 
the plateau intergradation with torquata may occur, but until this can 
be shown distincta must be accorded full specific rank. 

Specimens examined. — One, from the type locality. 

NEOTOMA TROPICALIS Goldman. 

Totontepec Wood Rat. 

(PI. VI, figs. 3, 3a.) 

Neotoma tropicalis Goldman, Proc. Biol. Soc. Wash., XVII, pp. 81-82, March 21, 1904. 
Type from Totontepec, Oaxaca, Mexico; No. 68593, $ ad., U. S. National Muse- 
um (Biological Survey Collection); collected by E. W. Nelson and E. A. Goldman, 
July 17, 1894. 

52668°— 10 5 



66 NORTH AMERICAN FAUNA. [no. 81. 

Distribution. — Mountains of northeastern Oaxaca, Mexico. Humid 
Tropical Zone. 

General characters. — Size small, tail rather short, slender, thinly 
haired, nearly unicolor; ears small. Member of ferruginea section of 
mexicana group, in color similar to N. f. tenuicauda, but cinnamon 
rufous of upperparts encroaching on underparts, and skull differing 
in important characters. 

Color. — Slightly worn pelage: Upperparts light cinnamon rufous, 
clearest along sides, well darkened by black-tipped hairs, which are 
most abundant on middle of face, top of head, and along back; outer 
sides of legs brownish ; cinnamon rufous of sides encroaching on under- 
parts posteriorly, leaving a narrow, whitish area along median line of 
belly; axillae ochraceous buff, this color forming a more or less well- 
defined pectoral band ; rest of underparts dull whitish, the fur basally 
plumbeous except on throat, where it is pure white to roots; nose 
and ankles dusky; feet irregularly clouded with dusky, the toes of 
hind feet whitish ; tail dusky above, slightly paler below. 

Skull. — Somewhat like that of N.f. tenuicauda, but nasals more 
wedge-shaped, much narrower and longer, reaching or passing plane 
of lachrymals; ascending branches of premaxillae very long, reaching 
beyond plane of lachrymals nearly to interorbital constriction; 
frontals broader and flatter posteriorly; teeth smaller (molars of the 
mexicana type) ; bullae smaller. Compared with that of N. parvidens 
the skull is larger and flatter; brain case larger and more smoothly 
rounded; nasals and ascending branches of premaxillae longer; teeth 
larger. 

Measurements. — Type: Total length, 325; tail vertebrae, 156; hind 
foot, 34. Skull (type): Basilar length, 33.6; zygomatic breadth, 
22.3; interorbital breadth, 5.9; length of nasals, 16.7; length of 
incisive foramina, 8.7; length of palatal bridge, 7.9; alveolar length 
of upper molar series, 8.5. 

Remarks. — N. tropicalis is a small, dark species inhabiting the moist, 
tropical, heavily forested mountain slopes in northeastern Oaxaca. 
It differs too markedly from the other members of the ferruginea 
section of the mexicana group to require close comparison with any of 
them. 

Specimens examined. — Two, from the type locality. 

NEOTOMA PARVIDENS Goldman. 

Small-toothed Wood Rat. 

(PI. VI, figs. 2, 2a.) 

Neotoma parvidens Goldman, Proc. Biol. Soc. Wash., XVII, p. 81, March 21, 1904. 
Type from Juquila, Oaxaca, Mexico; No. 71586, 9 ad., U. S. National Museum 
(Biological Survey Collection); collected by E. W. Nelson and E. A. Goldman, 
February 27, 1895. 



1910.] MEXICANA GEOUP — FERRUGINEA. 67 

Distribution. — Known only from the type locality in the mountains 
of southwestern Oaxaca. Humid Tropical Zone. 

General characters. — Size smallest of the ferruginea section of the 
mexicana group; color cinnamon rufous to orange buff; tail rather 
short and slender. Closely resembling N. f. picta in color, but very 
much smaller; differing also in cranial characters. 

Color. — Fresh pelage: Upper parts varying from cinnamon rufous 
to rich orange buff, brightest along sides, becoming duller on outer 
sides of legs, moderately darkened by black-tipped hairs, which are 
most abundant on face, top of head, and along back; under parts 
including upper lip and part of cheeks white, the fur usually pure 
white to roots on throat, inner sides of forelegs, pectoral, and inguinal 
regions; axillae varying from orange buff to ochraceous buff; forefeet 
and toes of hind feet white; hind feet to toes irregularly clouded with 
dusky (in two out of five specimens pure white) ; tail faintly bicolor, 
dusky above, paler below. 

Skull. — Similar in general form to that of N. f. picta, but smaller, 
lighter, and usually more arched; interorbital width relatively 
greater; rostrum usually more decurved; nasals narrower and more 
wedge-shaped; molars of the mexicana type, but very small. 

Measurements. — Average of 5 adult topotypes: Total length, 306; 
tail vertebrae, 149; hind foot, 32. Skull (average of 2 adults) : Basilar 
length, 316; zygomatic breadth, 20.4; interorbital breadth, 5.2; length 
of nasals, 15.4; length of incisive foramina, 8.6; length of palatal 
bridge, 7.1; alveolar length of upper molar series, 7.8. 

Remarks. — This small richly colored species belongs clearly in the 
ferruginea section of the mexicana group. It is more nearly related 
to N.f. picta than to any other known form, but its small size and 
well-marked cranial differences, together with absence of any hint of 
intergradation, seem to entitle it to full specific rank. 

Specimens examined. — Five, all from the type locality. 

NEOTOMA FERRUGINEA Tomes. 
Guatemala Wood Rat. 

Neotoma ferruginea Tomes, Proc. Zool. Soc. London, pp. 282-284, 1861. Type from 
Duefias, Guatemala; No. 7.1.1.124., 9, British Museum (Tomes Collection); 
collected by Osbert Salvin. 

Distribution. — Known only from the type locality. 

General characters. — Size large; color orange rufous; fur white to 
roots on throat and middle of breast. Related to N. mexicana, 
although differing widely in color. Like N. f picta in color, but 
differing in size and cranial characters. 

Color. — (From original description) : "All the upper parts are of 
bright rufous colour, and all the under parts pure white, the line of 
separation being very clear and distinct. The fur of the back is 



68 



NORTH AMERICAN FAUNA. 



[no. 31. 



mixed with black hairs, giving that part a much darker colour than 
the sides of the body, where these hairs are less abundant. Fur of the 
outer surface of the legs strongly tinged with dusky, inner surface of 
the forelegs whitish, of the hind ones dusky grey. From the chin to 
the space between the forelegs is an elongated patch of fur, which is 
pure white from root to tip * * *. The tail is deep dusky above, 
paler below * * *." 

Skull. — About like that of N.f chamula. Similar in general form 
to that of N. mexicana, but larger; frontals broader posteriorly; nasals 
emarginate (in mexicana truncate) posteriorly; bullae smaller; denti- 
tion as in N. mexicana, but heavier. 

Measurements. — Type (from original description) : Head and body, 
6" 6'" (about 165 mm.); tail vertebra?, 6" (about 152 mm.). SJcull 




Fig. 8.— Distribution of Neotoma ferruginea and subspecies. 

(type a ): Basilar length, 36; zygomatic breadth, 22.6; interorbital 
breadth, 6; length of nasals, 18 ; alveolar length of upper molar series, 8. 
Remarks.— N. mexicana and N. ferruginea differ widely in external 
appearance, but their cranial characters indicate rather close relation- 
ship, so close that complete intergradation through the intervening 
forms may occur, though not shown by the material examined. 
Gerrit S. Miller, jr., has kindly compared for me specimens of the 
nearly related forms in the Biological Survey Collection, with, the 
type of ferruginea in the British Museum, and found the latter to agree 
in color almost exactly with a specimen of picta from Omilteme, 
Guerrero, while its skull is most like one from Volcan Santa Maria, 
Guatemala, here referred to chamula. 

"Measured by G. S. Miller, jr. 



1910.] MEXICANA GBOUP CHAMULA. 69 

NEOTOMA FERRUGINEA CHAMULA Goldman. 

Chiapas Wood Rat. 

(PI. V, figs. 4, 4a.) 

Neotoma ferruginea chamula Goldman, Proc. Biol. Soc. Wash., XXII, p. 141, June 25, 
1909. Type from mountains near San Cristobal, Chiapas, Mexico (altitude 8,400 
feet), No. 76061, $ ad., U. S. National Museum (Biological Survey Collection); 
collected by E. W. Nelson and E. A. Goldman, September 29, 1895. 

Distribution. — High mountains of central Chiapas, Mexico, and 
southwestern Guatemala. Transition and Canadian zones. 

General characters. — Size large; closely related to N.ferruginea, but 
color much darker; fur of under parts basally plumbeous everywhere. 
Similar in size to N.f. isthmica, but color decidedly darker. 

Color. — Slightly worn pelage: Upper parts tawny cinnamon rufous, 
purest along sides, rather heavily darkened by black-tipped hairs, 
which are most abundant over back; outer sides of legs dark hair 
brown, hind legs tinged with brownish buff; under parts dull white, 
the fur everywhere basally plumbeous; nose, lips, and ankles dusky; 
axillse dark orange buff; forefeet white; hind feet clouded with dusky 
to toes, toes white; tail indistinctly bicolor, blackish above, grayish 
brown below. 

SkuU. — Closely resembling that of N.ferruginea. Similar to that 
of N. f. isthmica, but less massive and less arched across anterior 
roots of zygomata ; nasals narrower posteriorly. 

Measurements. — Type: Total length, 363; tail vertebrae, 178; hind 
foot, 39. Skull (type): Basilar length, 37.1; zygomatic breadth, 23; 
interorbital breadth, 5.9; length of nasals, 18.3; length of incisive 
foramina, 10; length of palatal bridge, 8.3; alveolar length of upper 
molar series, 9.3. 

Remarks. — This is a high mountain form differing from ferruginea 
chiefly in color and probably intergrading with it in southern Guate- 
mala. Specimens from Volcan Santa Maria, Guatemala, formerly 
assumed to be nearly typical ferruginea, - prove on comparison with 
the type 6 to differ considerably in color and belong nearer the new 
form here described. On this volcano the wood rats appeared to be 
confined to the pine-covered upper slopes, and the destructive 
eruption of October, 1902, probably extinguished the local colony. 

Specimens examined. — Total number 7, as follows: 

Chiapas: San Cristobal (type locality), 3. 

Guatemala: Hacienda Chancol (about 13 miles north of Huehuetenango), 
1; Volcan Santa Maria, 3. 

a Proc. Biol. Soc. Wash., XVII, p. 79, March 21, 1904. 
b Compared by Gerrit S. Miller, jr. 



70 NORTH AMERICAN FAUNA. [no. 31. 

NEOTOMA FERRUGINEA SOLITARIA Goldman. 

Nenton Wood Rat. 

(PI. V, figs. 5, 5a.) 

Neotoma ferruginea solitaria Goldman, Proc. Biol. Soc. Wash., XVIII, p. 31, February 
2, 1905. Type from Nenton, Guatemala; No. 76908, $ ad., U. S. National Museum 
(Biological Survey Collection); collected by E. W. Nelson and E. A. Goldman, 
December 17, 1895. 

Distribution. — Known only from the type locality, near the upper 
end of the Chiapas Valley. Arid Tropical Zone. 

General characters. — Similar in general to N. ferruginea, but less 
richly colored; skull relatively longer and narrower. Closely related 
to N.f. isthmica, but smaller; skull much less massive and differing in 
important details. 

Color. — Partly worn pelage: Upper parts tawny ochraceous, brightest 
along sides, becoming ochraceous buff on head, outer sides of fore- 
arms and hind legs, well mixed with brownish black over top of head 
and back; under parts dull white, owing to plumbeous basal color of 
fur, except a small area on chin and throat, which is pure white; 
nose dusky; forefeet white; hind feet to toes irregularly clouded with 
dusky; toes white; tail faintly bicolor, brownish black above, dull 
gray below, becoming brownish toward tip. 

SJcull. — Generally similar to that of N. ferruginea, but relatively 
longer and narrower; rostrum more slender; nasals longer, more 
attenuate posteriorly, the ends not emarginate; ascending branches 
of premaxillse longer, reaching posteriorly nearly to interorbital con- 
striction. Compared with N.f. isthmica, the skull is smaller, lighter, 
less arched across anterior roots of zygomata; rostrum more slender; 
nasals longer, more attenuate posteriorly. 

Measurements. — Type: Total length, 338; tail vertebrae, 156; hind 
foot, 35. Skull (type) : Basilar length, 35.5; zygomatic breadth, 22.4; 
interorbital breadth, 5.5; length of nasals, 18.2; length of incisive 
foramina, 9.2; length of palatal bridge, 8.3; alveolar length of upper 
molar series, 8.4. 

Remarks. — In the original description solitaria was compared with 
specimens then assumed to be typical ferruginea, but now known to 
represent a distinct form; hence the disagreement with the present 
description. The only known specimens are from the hot dry canyons 
near the upper end of the Chiapas Valley. In the lower part of the 
valley it probably passes into isthmica. On the slopes of the high 
mountains to the north and east of its range, solitaria probably inter- 
grades with chamula, which inhabits the cool, heavily forested moun- 
tain tops. 

Specimens examined. — Three, from the type locality. 



1910. j MEXICAN A GROUP — ISTHMICA. 71 

NEOTOMA FERRUGINEA ISTHMICA Goldman. 

Isthmian Wood Rat. 

(PI. V, figs. 6, 6a.) 

Neotoma isthmica Goldman, Proc. Biol. Soc. Wash., XVII, pp. 80-81, March 21, 1904. 
Type from Huilotepec, 8 miles south of Tehuantepec, Oaxaca, Mexico; No. 
73187, 9 ad., U. S. National Museum (Biological Survey Collection); collected 
by E. W. Nelson and E. A. Goldman, May 5, 1895. 

Distribution. — From Pacific coast region on south side of Isthmus 
of Tehuantepec, eastward into valley of Chiapas River and northwest- 
ward to Coixtlahuaca, Oaxaca. Arid Tropical and Lower Sonoran 
zones. 

General characters. — Size large, about equaling N. ferruginea; color 
similar; skull much heavier and more arched; tail very long and 
coarsely scaly. Similar in color to N. f. picta; size much larger; 
pelage coarser; skull more massive and higher arched. 

Color. — Worn pelage: Upper parts varying from orange rufous to 
ferruginous, clearest along sides, rather thinly overlaid with dusky- 
tipped hairs, which are most abundant along back; outer sides of 
legs buffy brownish gray; under parts, including upper lip, whitish, 
the fur usually pure white to roots on throat, breast, inner sides of 
forelegs, and inguinal region; forefeet white; hind feet to toes pure 
white or irregularly clouded with dusky; toes white; tail indistinctly 
bicolor, brownish above, paler below. 

Skull. — Skull massive; similar in general form to that of ferruginea 
and chamula but much heavier and more arched across anterior roots 
of zygomata; frontals broader and flatter posteriorly; brain case less 
inflated, more angular. Somewhat like that of N. f. solitaria, but 
larger, heavier, and differing in details. 

Measurements. — Average of 9 adults from the type locality: Total 
length, 371; tail vertebrae, 184; hind foot, 37.3. Skull (average of 
6 adult topotypes): Basilar length, 36.8; zygomatic breadth, 23; 
interorbital breadth, 6.1; length of nasals, 17.4; length of incisive 
foramina, 9.4; length of palatal bridge, 8; alveolar length of upper 
molar series, 8.3. 

Remarks. — In color this subspecies agrees closely with chrysomelas 
ferruginea, picta, solitaria, and parvidens, but differs from all in rather 
well-marked cranial characters. The typical form appears to be con- 
fined to the excessively arid coastal region on the south side of the 
Isthmus of Tehuantepec. Specimens from the interior plateau region 
of western Oaxaca (Oaxaca and Coixtlahuaca) grade toward picta. 
Those from the mountain slopes north of the Chiapas River Valley 
show an approach to chamula, the dark, high mountain form. Along 
the Pacific coast in southern Chiapas and Guatemala, isthmica may 
intergrade directly with ferruginea. 



72 NOKTH AMERICAN FAUNA. [no. 31. 

Specimens examined. — Total number 47, as follows: 

Oaxaca: Coixtlahuaca, 1; Huilotepec (type locality), 17; Juchitan, 3; Oaxaca, 

9; Puerto Angel, 1. 
Chiapas: Canjob, 3; San Bartolome, 2; Teopisca (20 miles southeast), 11. 

NEOTOMA FERRUGINEA PICTA Goldman. 

Painted Wood Rat. 

(PI. VI, figs. 1, la.) 

Neotoma picta Goldman, Proc. Biol. Soc. Wash., XVII, pp. 79-80, 1904. Type from 
mountains near Chilpancingo, Guerrero, Mexico; No. 70050, $ ad., U. S. National 
Museum (Biological Survey Collection); collected by E. W. Nelson and E. A. 
Goldman, December 20, 1894. 

Distribution. — Sierra Madre of Oaxaca and Guerrero, Mexico. 
Transition and Canadian zones. 

General characters. — Size smaller than N. ferruginea; color about 
the same; cranial characters distinctive. Similar in color to N. f 
isthmica, but smaller; skull lighter and less arched. Closely related 
to N.f. tenuicauda, but color brighter. 

Color. — Fresh pelage: Upper parts varying from rich orange buff 
to ferruginous, brightest along cheeks, shoulders, and sides, moderately 
darkened by black-tipped hairs, which are most abundant over top 
of head and back; under parts white (in some specimens suffused 
with pinkish buff), the fur usually pure white to roots on throat, and 
sometimes middle of breast and inner sides of forelegs; forefeet yel- 
lowish white; hind feet to toes irregularly clouded with dusky; toes 
white; tail indistinctly bicolor (occasionally concolor), dusky above, 
paler below. 

Skull. — Much smaller than in N. ferruginea; nasals truncate 
instead of emarginate posteriorly; frontals relatively narrower pos- 
teriorly, the sides more upturned. Similar in general to that of N.f. 
isthmica, but smaller, lighter, and less arched across anterior roots 
of zygomata. From that of N. f tenuicauda the skull differs chiefly 
in slightly larger size and longer nasals. 

Measurements. — Average of 10 adults from the type locality: Total 
length, 347; tail vertebra?, 172; hind foot, 34.5. STcull (average of 
6 adult topotypes): Basilar length, 346; zygomatic breadth, 22.4; 
interorbital breadth, 5.5; length of nasals, 16.8; length of incisive 
foramina, 9.3; length of palatal bridge, 7.7; alveolar length of upper 
molar series, 8.6. 

Remarks. — This mountain form agrees with ferruginea in richness 
of color, but differs markedly in size and cranial characters. Speci- 
mens from mountains near Ozolotepec, Oaxaca, show intergradation 
with isthmica. The range of picta seems to be separated from that of 
tenuicauda by the arid valley of the Balsas Kiver, but the two forms 



1010.] MEXICANA GROUP TENUICAUDA. 73 

are evidently closely related, and intergradation is probable in the 
mountains of western Michoacan. Specimens taken in December 
are in fresh pelage. Those taken in May are slightly worn, but do 
not differ in color. 

Specimens examined. — Total number 33, as follows: 

Guerrero: Mountains near Chilpancingo (type locality), 16; Omilteme, 15. 
Oaxaca: Mountains near Ozolotepec, 2. 

NEOTOMA FERRUGINEA TENUICAUDA Merriam. 

Slender-Tailed Wood Rat. 

Neotoma tenuicauda Merriam, Proc. Biol. Soc. Wash., VII, pp. 169-170, September 
29, 1892. Type from north slope of Sierra Nevada de Colima, Jalisco, Mexico; 
No. fjHjff, £ad., U. S. National Museum (Biological Survey Collection) ; collected 
by E. W. Nelson, April 13, 1892. 

Distribution. — Plateau region of western Mexico from southern 
Zacatecas to Mount Tancitaro, Michoacan. Lower Sonoran to Boreal 
zones. 

General characters. — Size smaller than N. ferruginea; color duller, 
more brownish, under parts dull white, the fur everywhere basally 
plumbeous. Closely related to N. f. picta, but color more brownish. 

Color. — Slightly worn pelage: General color of upper parts buffy 
cinnamon rufous, varying in some specimens to dark ochraceous buff, 
clearest along sides, darkened by black-tipped hairs, which are most 
abundant along back; outer sides of legs brownish, more or less 
tinged with ochraceous; under parts dull white, the fur everywhere 
basally plumbeous; axillae rich ochraceous buff, this color usually 
forming a pectoral band; forefeet white; hind feet to toes more or 
less distinctly clouded with dusky, toes white; lips dusky; tail 
indistinctly bicolor (in most specimens nearly unicolor), blackish 
above, slightly paler below. 

Skull. — Much smaller than that of N. ferruginea; nasals shorter. 
Closely resembling that of N. f. picta, but nasals shorter. 

Measurements. — Type: Total length, 340; tail vertebras, 160; hind 
foot, 31. Shall (type): Basilar length, 33.6; interorbital breadth, 
5.3; length of nasals, 15.3; length of incisive foramina, 8.9; length 
of palatal bridge, 7.5; alveolar length of upper molar series, 8.8. 

Remarks. — N.f. tenuicauda is the most northerly form of the ferru- 
ginea section of the mexicana group. Although differing rather 
decidedly in size and color, it probably intergrades with ferruginea 
through picta and isthmica. On the north the range of tenuicauda 
approaches that of N. m. madrensis, but the two forms do not appear 
to intergrade. Specimens from the lower parts of its vertical range 
are not quite typical, but do not differ sufficiently to require 
separation. 



74 NORTH AMERICAN FAUNA. [no. 31. 

Specimens examined. — Total number 29, as follows: 

Jalisco: Bolafios (mountains about 10 miles north), 1; La Laguna, 2; San 
Sebastian, 7; Sierra Nevada de Colima, 1 (type); Talpa, 1; Zapotlan, 5. 
Michoacan: Los Reyes, 3; Mount Patamban, 2; Mount Tancitaro, 2; Zamora, 1. 
Zacatecas: Plateado, 4. 

NEOTOMA FERRUGINEA OCHRACEA Goldman. 

Ochraceous Wood Rat. 

Neotoma ferruginea ochracea Goldman, Proc. Biol. Soc. Wash., XVIII, pp. 30-31, 
February 2, 1905. Type from Atemajac, near Guadalajara, Jalisco, Mexico; No. 
IIHIj <? a d-, U. S. National Museum (Biological Survey Collection); collected by 
E. W. Nelson, May 21, 1892. 

Distribution. — Vicinity of the type locality. Lower Sonoran Zone. 

General characters. — Size smaller than N. ferruginea; coloration 
pale; feet pure white; tail sharply bicolor; skull small, light, and 
elongated. 

Color. — Slightly worn pelage: Entire upper parts light ochraceous 
buff, purest and brightest along cheeks and sides, darkened over dorsal 
region, especially posterior half, by black-tipped hairs; under parts 
dingy white, washed with buffy, becoming clear, strong, ochraceous 
buff on pectoral region; nose and upper lip grayish white; feet pure 
white; tail brownish above, whitish below. 

Skull. — Similar in general to that of N. ferruginea, but smaller and 
more arched across anterior roots of zygomata, nasals more slender, 
reaching posteriorly to plane of lachrymals, the ends pointed instead 
of emarginate; ascending branches of premaxillse relatively longer, 
reaching posteriorly nearly to interorbital constriction; frontals 
longer and much narrower, the sides slightly upturned. Compared 
with that of N. f tenuicauda the skull is longer and narrower, with 
much longer and narrower nasals and longer ascending branches of 
premaxillse. 

Measurements. — Type: Total length, 348; tail vertebras, 161; hind 
foot, 38. Skull (type): Basilar length, 35.5; zygomatic breadth, 
22.9; interorbital breadth, 5.3; length of nasals, 17.9; length of 
incisive foramina, 9.3; length of palatal bridge, 8.4; alveolar length 
of upper molar series, 9.2. 

Remarks. — This subspecies differs considerably in color and cranial 
characters from N. ferruginea, but specimens examined show nearly 
complete intergradation through tenuicauda, picta, and isihmica. 

Specimens examined. — Two, from the type locality. 

NEOTOMA CHRYSOMELAS Allen. 

Nicaragua Wood Rat. 

Neotoma chrysomelas Allen, Bull. Amer. Mus. Nat. Hist., XXIV, pp. 653-654, Octo- 
ber 13, 1908. Type from Matagalpa, Nicaragua; No. 28372, J ad., American 
Museum of Natural History, N. Y.; collected by W. B. Richardson, September 
17, 1907. 



11110.1 



MEXICANA GROUP — CHRYSOMELAS. 



75 



Distribution. — Known only from the type locality, at about 3,000 
feet altitude, in central Nicaragua. 

General characters. — Closely related to N. ferruginea, but appar- 
ently larger; type and topotypes in rather worn pelage are paler, less 
rich orange rufous than the type of ferruginea. 

Color. — Partly worn pelage: Upper parts light orange rufous (paler 
and less intense than in the type of ferruginea) , purest along sides, 
darkened on top of head and back by overlying black-tipped hairs; 
outer sides of legs buffy grayish brown; under parts white, the fur 
plumbeous basally except small areas, mainly on throat and on pec- 
toral and inguinal regions, where it is pure white to roots; forefeet 
white; hind feet white clouded with dusky to toes; toes pure white; 
tail brownish or blackish above, grayish below. 

Skull. — Much like that of N. ferruginea, but larger than the type. 
Compared with chamula, one of the larger forms of the ferruginea 
group, the skull of 
chrysomelas has 
zygomata more 
squarely spreading 
posteriorly, narrow- 
ing anteriorly, the 
sides less nearly par- 
allel than in cham- 
ula; squamosal arm 
of zygoma heavier; 
f rontals broader pos- 
teriorly, the lateral 
border slightly con- 
vex and overhanging 
orbits. 

Measurements 
(from original de- 
scription). — Type: 
Total length 380; 
tail vertebrae, 170; hind foot (dry skin), 36. Average of four adult 
topotypes: Total length, 375; tail vertebrae, 160. Skull (average 
of four adults): Basilar length, 37.3; zygomatic breadth, 24.2; 
interorbital breadth, 6.2; length of nasals, 18.2; length of incisive 
foramina, 9.8; length of palatal bridge, 8.5; alveolar length of upper 
molar series, 9.1. 

Remarks. — The recent discovery of A 7 , chrysomelas in central Nica- 
ragua materially extends the known range of the genus southward in 
Central America. The evident close affinity of chrysomelas with 
other members of the widely distributed ferruginea group points to 
the probable occurrence of wood rats in San Salvador and Honduras 
and the continuous range of the genus from Nicaragua northward. 




Fig. 9.— Distribution of Neotoma chrysomelas. 



76 NORTH AMERICAN FAUNA. [no.81, 

The type and topotypes of chrysomelas in somewhat worn pelage are 
paler, less intense orange rufous than a specimen in the Biological 
Survey Collection, which Mr. Miller compared for me and found 
exactly like the type oiferruginea in color. 

Specimens examined. — Total number 4, the type and three 
topotypes. 

NEOTOMA DESERTORUM Merriam. 

Desert Wood Rat. 

(PI. VI, figs. 4, 4«.) 

Neotoma desertorum Merriam, Proc. Biol. Soc. Wash., IX, pp. 125-126, July 2, 1894. 
Type from Furnace Creek, Death Valley, Inyo County, Cal.; No. |ftft> $ a d> 
U. S. National Museum (Biological Survey Collection); collected by T. S. Palmer, 
January 31, 1891. 

Neotoma bella Bangs, Proc. New Eng. Zool. Club, I, pp. 66-67, July 31, 1899. Type 
from Palm Springs, Riverside County, Cal.; No. 5308, $ old ad., collection of 
E. A. and O. Bangs; collected by E. C. Thurber, April 12, 1896. 

Neotoma nevadensis Taylor, Univ. Cal. Pub. Zool., V, pp. 289-296, pis. 27-29, Feb- 
ruary 12, 1910. Type from Virgin Valley, Humboldt County, Nev. ; No. 8282, ? , 
University of California Museum of Vertebrate Zoology; collected by Annie M. 
Alexander, May 17, 1909. 

Distribution. — Desert regions in southeastern and northeastern 
California, Nevada, eastern Oregon, northern and western Utah, east 
to northwestern Colorado, and south along the west side of the Colo- 
rado River to northeastern Lower California. Upper and Lower 
Sonoran zones. 

General characters. — Size small; tail rather short; ears large; fur 
long, soft, and silky; skull small; bulla? very large. Similar to N. 
lepida, tail shorter haired and sharply bicolored. In general appear- 
ance closely resembling N. i. gilva, but not nearly related. 

Color. — Fresh pelage: Upper parts pale pinkish buffy, purest on 
cheeks and sides, becoming creamy buff on middle of face, moderately 
darkened dorsally by blackish hairs; feet and underparts white, the 
belly in some specimens more or less suffused with pinkish buff; sides 
of neck pinkish buff, this color sometimes spreading across throat; 
tail varying from grayish brown to blackish above, white below. 

SJcull. — Similar to that of N. lepida, but more angular; braincase 
less smoothly rounded; nasals narrower posteriorly; zygomata less 
squarely spreading, broadest posteriorly; frontal region narrower, 
with edges more upturned; bullae very large and rounded; first upper 
molar with anterointernal reentrant angle faint or absent. 

Measurements. — Average of 18 adult topotypes: Total length, 
293 (265-313); tail vertebrae, 131 (120-142); hind foot, 30.1 (28-31). 
Skull (average of 6 adult topotypes): Basilar length, 33.4; zygo- 
matic breadth, 20.6; interorbital breadth, 4.9; length of nasals, 15.3; 
length of incisive foramina, 8.3; length of palatal bridge, 6.6; alveolar 
length of upper molar series, 7.6. 



UtlO.l 



DESERTORUM GROUP DESERTORUM. 



77 



Remarks. — The desertorum group includes the four smallest forms 
in the genus, desertorum, lepida, stephensi, and goldmani, ranging as a 
group from the desert region of northeastern California, eastern 
Oregon, and northwestern Colorado to Lower California and north- 
eastern Mexico. The species are characterized externally by small 
size, large ears, and long, soft, silky fur. Although individuals vary 
considerably, the general characters of desertorum remain rather 
remarkably constant throughout its geographic range. 




/V£V7VM4 OESERTOtWM. 
L£P/M. 
L. Sr£*¥/£A&/. 



Fig. 10.— Distribution of Neotuma desertorum, N. lepida, and subspecies. 

Specimens taken in lava beds are usually darker than those inhab- 
iting lighter-colored rock formations, and at some localities show the 
dichromatic condition noted also in N. albigula and other species. 
Most specimens from St. George, Utah, are dark colored above and 
have the under parts strongly suffused with pinkish buff, but along 
with these is a normally colored one. Specimens from Marysvale, 
Utah, include light and dark colored individuals; one has white under 



78 NORTH AMERICAN FAUNA. [no. 31. 

parts, while others are pinkish buffy across throat and belly. In 
these the tail also varies from blackish to pale grayish above. The 
small number available from Utah and Colorado have tails which 
average somewhat shorter than those from Nevada, California, and 
Lower California, but are within the wide range of variation shown 
in nearly every large series. The tail varies in length from 120 to 
140, reaching beyond these limits in a few extreme individuals. 

The type of N. bella in the richly colored pelage common in old 
adults is yellower than usual in typical desertorum; the skull is more 
massive, with heavier rostrum and anteriorly broader nasals. These 
differences appear to be mainly due to advanced age. Specimens 
from the region of the type locality of N. nevadensis seem inseparable 
from N. desertorum. 

Specimens examined. — Total number 377, from the following 
localities : 

Arizona: Bright Angel Creek (mouth), 2; Fredonia, 1; Jacobs Pool, 1; Lees 
Ferry, 5; Soap Creek (15 miles southwest of Lees Ferry), 1; Trumbull 
Mountains, 1. 

California: Amargosa River, 1; Amedee, 9; Argus Mountains, 3; Barstow, 5; 
Bennett Wells, 22; Cameron, 8; Chucka walla Spring, 1; Coso, 22; Coso 
Mountains, 2; Emigrant Spring, Panamint Mountains, 1; Fairmont, 3; 
Fort Yuma, 2; Funeral Mountains, 6; Furnace Creek, Death Valley, 26; 
Grapevine Spring, 2; Hesperia, 2; Honey Lake, 1; Independence Creek 
(Rex Monte Mill), 1; Inyo Mountains, 7; Lone Pine, 8; Lone Willow 
Spring, San Bernardino County, 2; Long Valley, (22 miles from Bennett 
Wells), 1; Ludlow, 5; Mesquite Valley, 3; Mohave, 3; Mohave River, 3; 
Morongo Pass, 1; Needles, 10; New York Mountain, San Bernardino 
County, 8; Onyx, 19; Oro Grande, 19; Owens Lake, 1; Owens Valley 
(Moran's), 4; Palm Springs, 2; Panamint, 2; Panamint Mountains, 8; 
Panamint Valley, 19; Providence Mountains, 10; Resting Springs, 8; 
Saline Valley, 1; Saratoga Springs, 6; Secret Valley, Lassen County, 1; 
Shepherds Canyon, Panamint Mountains, 1; Twelvemile Spring, near 
Resting Springs, 3; Walker Pass, 2; Warren Well, Mohave Desert, 1. 

Colorado: Rangely, 4. 

Nevada: Ash Meadow (10 miles north), 5; Austin (35 miles southwest), 4; 
Battle Mountain, 7; Bunkerville, 4; Candelaria, 1; Charleston Mountains, 
1; Colorado River, Lincoln County, 2; Deep Hole, 6; Granite Creek, 2; 
Grapevine Mountains, 2; Indian Creek, 1; Monitor Valley, 1; Pahrump 
Valley, 7; Peavine, 2; Pyramid Lake, 3; Rabbit Hole Mountains, 1; 
Reese River, 1; Silver Creek, 1; Smoky Creek Desert, 1; Smoky Creek, 3; 
Vegas Valley, 1. 

Oregon: Vale, 1. 

Utah: Fort Cameron, 5; Henry Mountains, 4; Kanab, 2; Kelton, 1; Little Pine 
Valley (near Hebron), 2; Loa, 1; Moccasin Spring, 1; Pine Valley, 2; 
Promontory, 2; St. George, 3; Santa Clara, 2. 

Lower California: Cerro Prieto (near Volcano Lake), 2; Cocopah Mountains, 2. 



1910.] DESERTORUM GROUP LEPIDA. 79 

NEOTOMA LEPIDA Thomas. 

Thomas Wood Rat. 

Neotoma lepida Thomas, Ann. and Mag. Nat. Hist., 6 Ser., XII, p. 235, September, 
1893. Type locality unknown. Type formerly No. 3898 in Smithsonian Insti- 
tution, now No. 73.6.3.42, British Museum. 

Distribution. — Upper Sonoran Zone in the plateau region of north- 
eastern Arizona, north of the Little Colorado River, and northwestern 
New Mexico south to Gallup, grading to the southward into stephensi. 

General characters. — Most like N. I. stephensi, but smaller; color 
decidedly paler (yellowish instead of dark grayish buff) ; tail slightly 
bushy, as in stephensi. Rather closely resembling desertorum exter- 
nally, but upper parts yellowish instead of pinkish bufTy; tail longer 
haired and not sharply bicolored. 

Color. — Partly worn pelage: Upper parts yellowish or between buff 
and cream buff, clearest along cheeks, shoulders, and sides, becoming 
pale buffy gray on middle of face, darkened over top of head and 
back by dusky-tipped hairs; under parts white or creamy (varied in 
some specimens by a pinkish buffy wash across belly) the fur plumbe- 
ous basally, except small pure white areas on throat, inner sides of 
forelegs, and on pectoral and inguinal regions; feet white; tail pale 
grayish, somewhat darker above than below. Young (about half 
grown) : Duller, less yellowish than adults, but decidedly paler than 
stephensi with less nearly unicolor tail. 

Skull. — Closely resembling that of stephensi, but smaller, with 
relatively smaller interparietal, shorter tooth rows, and smaller 
incisors; first upper molar with anterointernal reentrant angle faint 
or absent (usually absent) , as in the other members of the desertorum 
group. 

Measurements. — Average of six adults from Keam Canyon: Total 
length, 286; tail vertebras, 136; hind foot, 29. Skull (average of 
same): Basilar length, 324; zygomatic breadth, 21.1; interorbital 
breadth, 5.5; length of nasals, 14.8; length of incisive foramina, 8.3; 
length of palatal bridge, 7.3; alveolar length of upper molar series 7.8. 

Remarks. — A wood rat, mammal number 3898 of the Smithsonian 
Institution, was regarded by Coues a in 1877 as an abnormally small 
individual of Neotoma cinerea and listed by him as from Williams 
Spring, Utah, with C. S. McCarthy as collector. Subsequently this 
specimen passed to the British Museum and became the type of N. 
lepida Thomas, from "Utah." Williams Spring, Utah, has been 
generally accepted by later authors as the type locality of the species. 
N. Hollister has directed my attention to the fact that when the 
type was entered in the catalogue of the Smithsonian Institution 
it had no label. An examination of the record shows that the speci- 

« Monographs of N. A. Rodentia, p. 24, 1877. 



80 NORTH AMERICAN FAUNA. [no. 31. 

men was entered in the same handwriting as the preceding numbers, 
at the end of a general collection from various localities, received 
from J. H. Simpson and collected by C. S. McCarthy, who was the 
taxidermist for Simpson's expeditions to the " Great Basin of Utah." a 

The preceding number (3897) was entered as "Neotoma Cinerea" 
and ditto marks follow for number 3898, in the columns for the 
name of the species and those of Simpson and McCarthy, but the 
record is incomplete, and this is explained by the statement in the 
column for remarks that the specimen had " no label." That it came 
from Williams Spring is an assumption without apparent reason 
beyond the mere chance that the preceding entry is from that local- 
ity. The type locality of N. lepida is therefore unknown. 

In March, 1904, specimens of arizonse, desertorum, and an unrec- 
ognized animal in the collection of the Biological Survey from Keam 
Canyon, Arizona, were submitted to Oldfield Thomas for comparison 
with the type of lepida. The Keam Canyon specimen was found 
to agree in many respects, and, in view of the uncertainty regarding 
the type, is now assumed to represent typical lepida. 

N. lepida seems to be a species quite distinct from desertorum, 
although clearly a member of the same group. The ranges of the 
two are completely separated by the effective barrier of the Colorado 
River. Specimens of stephensi from Zuni River, Arizona, and Burley, 
N. Mex., are somewhat intermediate in coloration and in cranial 
characters, also approach lepida. 

Total number of specimens examined 32, from the following 
localities : 

Arizona: Cedar Ridge (30 miles north of Tuba), 3; Keam Canyon, 20; Tuba, 1. 
New Mexico: Blanco, 1; Bluewater, 1; Fort Wingate, 3; Gallup, 1; Wingate, 2. 

NEOTOMA LEPIDA STEPHENSI Goldman. 

Stephens Wood Rat. 

(PI. VI, figs. 5, 5a.) 

Neotoina stephensi Goldman, Proc. Biol. Soc. Wash., XVIII, pp. 32-33, February 2, 
1905. Type from Hualpai Mountains, Arizona (altitude, 6,300 feet); No. 117466, 
9 ad., U. S. National Museum (Biological Survey Collection); collected by F. 
Stephens, July 1, 1902. 

Distribution. — Upper Sonoran Zone along Hualpai, Mogollon, and 
White Mountains across central Arizona and from the Burro Moun- 
tains to the Zuni Mountains in western New Mexico, passing farther 
north into lepida. 

General characters. — Size small; tail slightly bushy; underparts 
pinkish buff. Most like lepida, but larger and decidedly darker — 
dark grayish buff instead of yellowish. Related to N. desertorum, 
but tail longer haired; differing also in color and cranial characters. 

« Report of Exploration Across the Great Basin of Utah in 1859, L876. 



1910.] DESERTORUM GROUP — GOLDMANI. 81 

Color. — Partly worn pelage: Upper parts dark grayish buff, mod- 
erately darkened over top of head and back by dusky hairs, becoming 
pinkish buff along cheeks and sides; under parts usually more or less 
heavily washed with pinkish buff, this color spreading over entire 
belly and irregularly invading other parts; small areas on pectoral 
and inguinal regions, sometimes including throat, pure white; ankles 
dusky; feet white; tail grayish brown above, slightly paler below. 
Young (about half grown) : Decidedly darker than lepida, with more 
nearly unicolor tail. 

Skull. — Closely resembling that of lepida, but larger and heavier. 
Similar to that of desertorum, but less angular; brain case more 
smoothly rounded; nasals broader posteriorly; zygomata more 
squarely spreading, the sides nearly parallel; frontal region broader 
and flatter; dentition and bullae about as in desertorum. 

Measurements. — Average of 3 adults from the type locality: Total 
length, 305; tail vertebrae, 135; hind foot, 31. Skull (an adult 
topotype): Basilar length, 33.8; zygomatic breadth, 22; interorbital 
breadth, 6.7; length of nasals, 16.5; length of incisive foramina, 8.9; 
length of palatal bridge, 7.5 ; alveolar length of upper molar series, 8.5. 

Remarks. — This dark subspecies is much more nearly allied to lepida 
than to desertorum, as shown by closer agreement in important 
cranial characters, and in hairiness of tail. Specimens from Zuni 
River, Arizona, and Burley, N. Mex., are somewhat paler than the 
typical form and in cranial characters also approach lepida. One 
from the lava beds near Grant, N. Mex., is unusually dark even for 
stephensi and may indicate the dichromatic condition seen in albigula 
and fallax and apparently associated with the lava beds of the region. 
In the series from the Burro Mountains the under parts vary from the 
usual pinkish buff across the belly to pure white. 

Total number of specimens examined, 27, from the following 
localities : 

Arizona: Hualpai Mountains (type locality), 5; Springerville (25 miles north), 

2; Zuni River, 3; Walnut, 1. 
New Mexico: Burley, 3; Burro Mountains, 11; Grant, 1; Glenwood, San 

Francisco River, 1. 

NEOTOMA GOLDMANI Merriam. 

Goldman Wood Rat. 

(PI. VI, figs. 6, 6a.) 

Neotoma goldmani Merriam, Proc. Biol. Soc. Wash., XVI, p. 48, March 19, 1903. Type 
from Saltillo, Coahuila, Mexico (altitude 5,000 feet); No. 116894, $ ad., TJ. S. 
National Museum (Biological Survey Collection); collected by E. W. Nelson 
and E. A. Goldman, April 18, 1902. 

52668°— 10 6 



82 



NORTH AMERICAN FAUNA. 



[NO. SI. 



Distribution. — Desert regions in southern Coahuila, Mexico. Lower 
Sonoran Zone. 

General characters. — Smallest known species of the genus. In 
general, similar to N. lepida, but smaller; tail shorter haired; color 
darker; cranial characters distinctive. ' 

Color. — Partly worn pelage: Upper parts creamy buff, purer along 
sides, becoming paler on head, much darkened over back by admix- 
ture of dusky; feet and under parts white; tail blackish above, 
white below. 

Skull. — Similar in general form to that of N. lepida, but smaller; 

ascending branches 
of premaxillse broad- 
er posteriorly; ante- 
rior border of inter- 
parietal more con- 
vex, posterior angle 
more developed; 
bulla? much smaller; 
dentition of the deser- 
torum type, but 
rather heavy. 

Measure ment s.— 
Average of 4 adult 
topotypes: Total 
length, 279; tail ver- 
tebras, 128; hind foot, 
30. Skull (average 
of same): Basilar 
length, 30.8; zygo- 
matic breadth, 18.7; 
interorbital breadth, 
5.4; length of nasals, 14.1; length of incisive foramina, 8.1; length of 
palatal bridge, 6.4; alveolar length of upper molar series, 7.4. 

Remarks. — This form is distinguishable from all other species of 
the genus by its small size. It agrees with the other members of the 
desertorum group in long, soft, silky pelage. 

Specimens examined. — Total number 10, from the following locali- 
ties in Coahuila, Mexico : 

Jaral, 5, Saltillo (type locality), 5. 

NEOTOMA MAGISTER Baird. 




Fig. 11.— Distribution of Neotoma goldmani. 



Cave Wood Rat. 

Neotoma magister Baird, Mamm. North Amer., p. 498, 1857. Cotypea from bone cave? 
near Carlisle, Pa.; Nop. 12206-12214. inclusive, U. S. National Museum; col- 
lected by S. F. Baird. 



1910.] 



PENNSYLVANIA GROUP MAGISTER. 



83 



Distribution.— Pleistocene cave deposits of Pennsylvania. 

General characters. — Similar to N. pennsylvanica, but larger. 

Skull. — Mandible similar to that of N. pennsylvanica, but larger; 
coronoid process longer, more decurved; diastema (mandibular) 
longer; beveled cutting edge of lower incisor longer; upper incisor 
broader and heavier; lower molars as in N. pennsylvanica, but larger, 
toothrow longer. 

Comparative measurements of Neotorna magister and Neotoma pennsylvanica. 



Museum 
number. 

! 


Name. 


Locality. 


<v 

60 

a 

■O 

a 

03 

y, 
a? 


o 


0J." 
03 J-> 

si 

~.° ^ 

«»§ 

o 


a 

03 

So 

si 

5 


o 
E 
o . 

o 

a 
t2 

o a 

o 

(J 


S3 

I® 

^ s 

03 U. r/j 

0J p t. 


i 

12207 ! 




Pennsylvania: Near Carlisle 

do 


Adult... 
Adult... 
Adult. . . 
Young . . 
Young . . 
cf Adult 
cf Adult 

9 Adult 
9 Adult 
cf Adult 


7.1 
7.3 

7.2 
7.2 
7.3 
6.4 
6.8 

6.1 
6.4 
6.7 


32.5 

31.6. 
29.7 

31.1 
29.8 
31.3 


9.7 
9.7 
9.6 
8.4 
8.3 
9.0 
8.2 

8.9 

7.8 
.92 


8.5 

8.4 
7.9 

8.2 
7.6 
7.5 


9.6 


12208 


...do 


10.1 


12206 


..do 


....do 


10.0 


12209 


...do 


...do 


9.8 


12212 


.do 


....do 


9.8 


77513 
82146 


...do 


West Virginia: White Sulphur. 
Virginia: Potomac River near 

Washington. 
do 


9.3 

8.8 


87195 


....do 


9.4 


77512 


...do 


...do 


8.7 


3956 


.do... 


Kentucky: Mammoth Cave 


9.4 









Remarks. — The remains of the type material used by Baird in his 
description of N. magister consist of eight incomplete halves of lower 
jaws and a fragment of a premaxilla containing an upper incisor. 
Of these jaws one of the most valuable for comparative purposes 
lacks only the teeth and angular process. Several of the others 
contain all the teeth but are otherwise imperfect. Three are evi- 
dently fully adult and the others somewhat immature. Collectively 
the jaws show all the parts except the angular process. Six are A T . 
magister and the others, numbers 12210 and 12213, are probably 
N. pennsylvanica. a 

Comparison of this rather scanty material with a considerable 
number of fully adult as well as immature specimens of the living 
species inhabiting the same region shows very close relationship. The 
differences noted, however, appear to be constant, and in the absence 
of material establishing their identity both should receive specific 
recognition. 

Specimens examined. — The type material. 

aFor discussion of status of N. magister see Rhoads, Proc. Acad. Nat. Sci., Phila., 
pp. 213-221, 1894, and Mearns, Bull. Amer. Mus. Nat. Hist., X, pp. 334-335, 1898. 



84 



NORTH AMERICAN FAUNA. 



[no. 31. 



NEOTOMA PENNSYLVANICA Stone. 

Pennsylvania Wood Rat. 

(PI. I, figs. 6, 6a.) 

Rat Bartram, in Kalm's Travels, II (English edition), pp. 47-48, 1749. 

Neotoma pennsylvanica Stone, Proc. Acad. Nat. Sci. Phila., pp. 16-18, February, 1893. 

Type from South Mountain, Cumberland County, Pa. (altitude 2,000 feet); 

No. 156, 9, collection of Witmer Stone; collected by J. G. Dillin, December 2, 

1892. 

Distribution. — Appalachian Mountain region from southern New- 
York to northern Alabama, probably including western North 




Fig. 12.— Distribution of Neotoma pennsylvanica. 

Carolina and northern Georgia, and westward to Mammoth Cave, 
Kentucky, and Lawrenceburg, Tennessee. Alleghenian and Caro- 
linian zones. 

General characters. — Largest of the round-tailed species of the 
genus; tail moderately long, well haired, bicolored; ears large; pelage 
coarse ; cranial characters pronounced ; no closely related living species 
known. 



1910.] PENNSYLVANIA GROUP PENNSYLVANIA". 85 

Color. — Fresh winter pelage: Upper parts grayish buff, becoming buffy 
gray on head, heavily overlaid, especially on back, with blackish; feet 
and under parts white, the fur pure white to roots, except along sides 
of belly, where the basal color is pale plumbeous; axillae creamy buff, 
tail varying from grayish brown to black above, whitish below. 

Skull. — Skull very large, elongate, rather narrow; rostrum long; 
nasals long and narrow, truncate or slightly emarginate posteriorly; 
ascending branches of premaxillae broadening and then narrowing 
abruptly at frontomaxillary suture, reaching posteriorly slightly 
beyond nasals; frontal region broad, constricted near middle, exca- 
vated above, lateral margins upturned and tending to develop a small 
point anteriorly; temporal ridges diverging to near anterior border 
of interparietal, whence they turn abruptly inward, as in N. cinerea; 
interparietal subquadrate or rectangular, much as in N. mexicana; 
palatal bridge about equal to incisive foramina; interpterygoid fossa 
rounded anteriorly, as in N. Jloridana, but narrower; sphenopalatine 
vacuities small; bullae more elongated and tapering anteriorly more 
than in N. Jloridana; mastoid process of squamosal very broad; 
dentition similar to that of N. Jloridana, but first upper molar with 
anterointernal reentrant angle deeper — nearly as deep as in N. mex- 
icana. Compared with that of N. magister, the mandible is smaller, 
with coronoid process shorter, less decurved, and beveled cutting edge 
of lower incisor shorter. 

Measurements. — Average of 2 adults from Renovo, Pa.: Total 
length, 430; tail vertebrae, 198; hind foot, 43. Skull (average of 
same): Basilar length, 45.7; zygomatic breadth, 27.5; interorbital 
breadth, 6.5; length of nasals, 21.7; length of incisive foramina, 10.9; 
length of palatal bridge, 10.1 ; alveolar length of upper molar 
series, 9.4. 

Remarks. — N. pennsylvanica is a well-marked species, requiring no 
close comparison with any known living form. Its exact relationship 
to the fossil N. magister, however, is somewhat problematical. Care- 
ful comparison of the fragmentary type material of the latter with 
N. pennsylvanica shows that the two were related, but differed in 
size and apparently in other important details. 

In 1749 Peter Kalm a cited John Bartram,of Philadelphia, as author- 
ity for the statement that he saw a " great number" of rats which 
lived among the rocks in the "Blue Mountains" of Pennsylvania, 
came out only at night, and made a "terrible" noise. Pennant 6 
described an American rat, apparently from a specimen in the Live- 
rian Museum, and referred to Bartram's observations. In 1792 

a Kalm's Travels (English edition), II, pp. 47-48, 1749. 

b Hist, of Quad., Ed. 3, II, No. 378, pp. 180-181, 1793.— Arctic Zool. I, No. 58, 
p. 130, 1784. 



86 NORTH AMERICAN FAUNA. [no. 31. 

Kerr° used Pennant's descriptions under the name Mus americanus. 
The animal seen by Bartram was probably the species now known as 
Neotoma pennsylvanica, but as Kerr apparently described some other 
species, although including *a reference to Kalm's Travels, the name 
given by him is untenable. 6 

Specimens of pennsylvanica from Tennessee and Alabama have 
somewhat shorter fur, the nasals average narrower posteriorly and 
the premaxillae are more prolonged beyond them than in those from 
Pennsylvania. 

Specimens examined. — Total number 94, as follows : 

Pennsylvania: Drurys Run, Clinton County, 3; Renovo, 5. 

Alabama: Huntsville, 2. 

Kentucky: Hawesville, 4; Lost Creek, 1; Mammoth Cave, 11. 

Tennessee: Lawrenceburg, 1; Walden Ridge (3 miles southwest of Rath- 
burn), 8. 

Virginia: Difficult Run (near mouth), Fairfax County, 2; Franklin, 11; Great 
Falls, 10; Hillsboro, 1; Peaks of Otter, 1; Potomac River (opposite Hog 
Island), 2; Potomac River (10 miles above Washington), 7; Tazewell 
Peak, 4; White Sulphur, 21. 

Subgenus HOMODONTOMYS c nobis. 

(PI. VII, figs. l-3o.) 

Type. — Neotoma fuscipes Baird, from Petaluma, California. 

Distribution. — Pacific coast region from northern Oregon to north- 
ern Lower California. 

Subgeneric characters. — Maxillary toothrow only slightly narrower 
posteriorly than anteriorly ; third upper molar broad and heavy, with 
middle enamel loop partially or completely divided by deepening of 
inner reentrant angle; frontals constricted near middle; bullae large; 
tail terete, tapering, and short-haired; hind foot naked below along 
outer side, at least to tarsometatarsal joint. 

Remarks. — The subgenus Homodontomys differs from Neotoma 
proper chiefly in the more nearly equal width of the molars and in the 
more or less complete division of the middle enamel loop in the crown 
of the third upper molar. The condition of the third upper molar is 
due mainly to the deepening of the inner reentrant angle, resulting in 
four more or less complete loops, instead of three as in Neotoma 
proper. In tins character it approaches Xenomys, but differs widely 
in other respects. In the peculiar combination of cranial and dental 
characters Homodontomys departs as much from Neotoma proper as 
from Teonoma, and has no very near relatives in either subgenus. 

a Animal Kingdom, No. 463, p. 227, 1792. 
& See Allen, Bull. Amer. Mus. Nat. Hist., VII, p. 189, 1895. 

c Homodontomys, from b^bg, the same; ddobf, tooth; /zuf, mouse, in reference to the 
nearly equal width of molars. 



1910.] FUSCIPES GROUP FUSCIPES. 87 

NEOTOMA FUSCIPES (Cooper MS.) Baire. 

Dusky-footed Wood Rat. 

(PI. VII, figs. 1, la, 16.) 

Neotoma fuscipes (Cooper MS.) Baird, Mamm. North Amer., pp. 495-496, 1857. Type 

from Petaluma, Sonoma County, Cal.; No. 2679, $, U. S. National Museum; 

collected by E. Samuels, February, 1856. 
Neotoma monochroura Rhoads, Amer. Nat., XXVIII, pp. 67-68, January, 1894. Type 

from Grants Pass, Josephine County, Oreg.; No. 1739, $ ad., Academy of Natural 

Sciences, Philadelphia, collected by George Kenzer. 
Neotoma splendens True, Proc. U. S. Nat. Mus., XVII, p. 353 [advance sheets issued 

June 27], 1894. Type from Marin County, Cal.; No. 19693, $, U. S. National 

Museum; collected November 25, 1887. 

Distribution. — Pacific coast region from San Francisco Bay north 
to Salem, Oreg. Upper Sonoran and Transition zones. 

General characters. — Size large; tail long, nearly unicolor; ears 
large; molariform toothrow only slightly narrowing posteriorly; 
third upper molar with middle enamel loop partially or completely 
divided by inner reentrant angle. About like N.f. annectens in color, 
but cranial characters distinctive. Nearer to N. f. streatori, but 
larger; color darker; metatarsus usually clouded with dusky instead 
of pure white, as in typical streatori. 

Color. — Fresh pelage: Upper parts light ochraceous buff clearest 
along sides, heavily darkened over top of head and back by black- 
tipped hairs, becoming grayish on facial area; throat, chest, and 
inguinal region pure white ; belly washed with creamy buff or pinkish 
buff, beneath which the fur basally is plumbeous; muzzle and orbital 
rings dusky; ears brownish; forefeet and toes of hind feet white; 
hind feet to toes irregularly clouded with dusky; ankles dusky, this 
color passing downwards on both sides, leaving the heels whitish; 
tail blackish. Worn pelage: Upper parts varying shades of rusty 
brown. 

Skull. — Skull long, rather narrow, and evenly arched; rostrum 
long; ascending branches of premaxillae slender, reaching posteriorly 
well beyond nasals; frontals rather narrow, with sides slightly 
upturned, forming supraorbital ridges; interpterygoid fossa narrow, 
concave or slightly emarginate anteriorly; interparietal subquadrate 
or rectangular; bullae large; molariform toothrow only slightly nar- 
rowing posteriorly; first upper molar with anterointernal reentrant 
angle reaching less than halfway across anterior enamel loop, much 
as in N. jloridana; third upper molar short and broad, the middle 
enamel loop partially divided by deepening of inner reentrant angle. 
Similar in general to that of N.f. streatori, but larger; palatal slits 
about equal to palatal bridge (decidedly longer than palatal bridge 
in streatori), not reaching posteriorly to anterior plane of first molars 
(reaching or passing this plane in streatori); interpterygoid fossa 
narrower; bullae larger. 



88 



NORTH AMERICAN FAUNA. 



[no. 31. 



Fig. 



Measurements. — Average of 3 adults from the type locality : Total 

length, 438; tail vertebrae, 
209; hind foot, 42.3. Slcutt 
(average of 2 adult topo- 
types): Basilar length, 42.5; 
zygomatic breadth, 25.7; in- 
terorbital breadth, 6.5; length 
of nasals, 20.3; length of in- 
cisive foramina, 11.2; length 
of palatal bridge, 9.7; alveolar 
length of upper molar series, 
9.8. 

Remarks. — N. fuscipes and 
its five subspecies form an in- 
teresting series of intergrading 
forms, the habitats of which 
completely encircle the Sacra- 
mento and San Joaquin val- 
leys in California. The only 
break in the series is in the 
region south of Monterey 
Bay, where the range of N.f. 
annectens meets that of N. f. 
macrotis, yet the two forms are 
quite distinct. Apart from 
its subspecies N. fuscipes re- 
quires close comparison with 
no known form. Typical 
specimens do not differ ap- 
preciably in color from those 
of annectens, but well-marked 
cranial characters suffice to 
separate them. Around the 
upper end of the Sacramento 
Valley, fuscipes intergrades 
with streatori, which ranges 
thence to the eastward and 
southward along the west side 
of the Sierra Nevada. 

Specimens from a large 
number of localities show 
many slight geographic varia- 
tions in color. Those from 
within the fog belt imme- 
usually darker than those from the 




13. — Distribution of the subgenus Homodontomys 
(Neotoma fuscipes and subspecies). 



diately along the coast are 



1910.] FUSCIPES GROUP — STREATORI. 89 

dryer valleys often only a few miles inland. Topotypes of mono- 
chroura have the under parts somewhat whiter and the hind feet less 
dusky than typical fuscipes, but these slight differences are within 
the range of variation of the species. N. splendens was based on one 
of the darker specimens of N. fuscipes. 

Specimens examined.— Total number 220, from the following 
localities : 

California: Ager, 1 ; Alton, 2 ; Alton Junction, 4; Bald Mountain, Shasta County, 
1; Beswick, 1; Big Valley Mountains, 4; Bully Choop Mountains, 1; Burney, 
1; Calpella, 1; Camp Meeker, 4; Cassel, 8; Dana, 2; Eel River, near South 
Yolla Bolly Mountain, 10; Freestone, 3; Freshwater Creek, Colusa County, 
1; Glen Ellen, 8; Gualala, 1; Hayden Hill, 1; Hoopa Valley, 4; Hornbrook, 
13; Inverness, 4; Leesville, 2; Lierly's Ranch, 2; Little Shasta, 1; Lower 
Lake, Lake County, 3; Mad River, Trinity County, 4; Marin County, 3; 
Marshall, 6; Mendocino City, 3; Mount George, 8; Mount Saint Helena, 4; 
Mount Tamalpais, 2; Mount Veeder, 6; Nicasio, 24; Novato, 5; Petaluma 
(type locality), 15; Petrolia, 2; Picard, 8; Point Reyes, 7; Requa, 1; Round 
Valley, 1; Salmon Mountains (near Etna Mills), Siskiyou County, 1; Scott 
Valley (near Fort Jones), Siskiyou County, 1; Searsville, San Mateo County, 
2; Snow Mountain, 12; Willits, Mendocino County, 1. 

Oregon: Drain, 2; Elkhead, 1; Gold Beach, 7; Grants Pass, 9; Salem, 2. 

NEOTOMA FUSCIPES STREATORI Merriam. 

Streator Wood Rat. 

Neotoma fuscipes streatori Merriam, Proc. Biol. Soc. Wash., IX, p. 124, July 2, 1894. 
Type from Carbondale, Amador County, Cal.; No. 64439, $ ad., U. S. National 
Museum (Biological Survey Collection); collected by C. P. Streator, April 4, 1894. 

Distribution. — West slope of the Sierra Nevada in California from 
Tehama County south to Porterville, Tulare County. Upper Sono- 
ran Zone. 

General characters. — Smaller than N. fuscipes; general color some- 
what paler; tail sharply bicolor; hind feet from tarsus down, pure 
white. 

Color. — Fresh pelage: Upper parts pale ochraceous buff, brightest 
and clearest along sides, moderately darkened over top of head and 
back by more abundant admixture of blackish hairs; face grayish; 
throat, chest, and inguinal region pure white; belly washed with pale 
buff or pinkish buff, beneath which the fur is basally plumbeous; 
ankles dusky; feet pure white; tail blackish above, whitish below. 

Skull. — In general similar to that of N. fuscipes, but smaller; 
ascending branches of premaxillae slightly heavier; incisive foramina 
relatively longer, reaching anterior plane of first molars; inter- 
pterygoid fossa broader and usually evenly rounded anteriorly; 
palate shorter. In streatori, simplex, and macrotis, the palatal bridge 
is decidedly shorter than incisive foramina (in fuscipes about equal). 



90 NORTH AMERICAN FAUNA. I no. 31. 

Compared with N.f. simplex the frontals are broader and ascending 
branches of premaxillae heavier. 

Measurements. — Average of 10 adult topotypes: Total length, 
375; tail vertebrae, 188; hind foot, 39.5. Skull (average of same): 
Basilar length, 38.6; zygomatic breadth, 24.3; interorbital breadth, 
6; length of nasals, 17.9; length of incisive foramina, 10; length of 
palatal bridge, 7.9; alveolar length of upper molar series, 8.8. 

Remarks. — This form intergrades on the north and west with 
fuscipes and on the south with simplex. Externally it closely resem- 
bles macrotis, but the pure white feet serve to distinguish it. 

Specimens examined. — Total number 46, from following localities : 

California: Blue Canyon, 3; Badger, 1; Carbondale (type locality), 15; Chinese 
Camp, 1; Coulterville, 1; Eshom Valley, Tulare County, 2; Nelson (8 
miles east), 7; Marysville Buttes (near Marysville), 2; Michigan Bluff, 1; 
Milo, 1; Porterville (8 miles east), 3; Tehama, 2; Threerivers, 7. 

NEOTOMA FUSCIPES ANNECTENS Elliot. 

Portola Wood Rat. 

(PI. VII, figs. 3, 3a.) 

Neotoma fuscipes annectens Elliot, Pub. Field Columb. Mus., Zool. Ser., I, p. 201 
March, 1898. Type from Portola, San Mateo County, Cal.; No. 2160, $ ad., 
Field Museum of Natural History, Chicago; collected by Price and Coolidge, 
December 23, 1895. 

Neotoma fuscipes affinis Elliot, Pub. Field Columb. Mus., Zool. Ser., I, pp. 202-203, 
March, 1898. Type from Alum Rock Park, Santa Clara County, Cal.; No. 2183, 
$ ad., Field Museum of Natural History, Chicago; collected by J. Diefenbach, 
May 14, 1895. 

Distribution. — Coast region of California from San Francisco Bay 
to Monterey Bay and thence inland and southward along the small 
valleys and mountain ranges east of the Santa Lucia Mountains to 
Carriso Plains, San Luis Obispo County. Upper Sonoran and Transi- 
tion zones. 

General characters. — Size and external appearance as in N. fuscipes, 
but cranial characters distinctive. 

Color. — Essentially as in N. fuscipes. 

Skull. — In general form like that of N. fuscipes, but palate with a 
blunt posterior median projection (palate in fuscipes concave or 
emarginate) ; palatal bridge less deeply grooved along either side of 
median line at posterior border of incisive foramina; bulhe slightly 
larger, with anterosuperior border of meatus more strongly projecting. 
Compared with those of N. f. macrotis and N. f streatori the skull is 
much larger, with narrow interpterygoid fossa, and decidedly larger 
bullae. 

Measurements. — Average of 14 adults from Mountain View, Cal.: 
Total length, 434; tail vertebras, 216; hind foot, 43.4. Skull (average 



1910.1 FUSCIPES GROUP SIMPLEX. 91 

of same): Basilar length, 41.5; zygomatic breadth, 26.2; interor- 
bital breadth, 5.8; length of nasals, 19.3; length of incisive foramina, 
10.1; length of palatal bridge, 9.5; alveolar length of upper molar 
series, 9.3. 

Remarks.— N. f. annectens does not differ externally from typical 
fuscipes, but is easily recognizable by well-marked and constant cra- 
nial characters. The ranges of annectens and macrotis meet along 
the Salinas Valley, east of the Santa Lucia Mountains, but are not 
known to overlap, and no intergradation of the two forms is shown 
by the specimens examined. N. f affinis was based on specimens 
from near the type locality of annectens and not appreciably different 
from typical examples of the latter form. 

Specimen's examined. — Total number 141, from the following 
localities : 

California: Alum Rock Park, 1; Aptos, 11; Bear Valley, San Benito County, 
5; Berkeley, 3; Black Mountain, Santa Clara County, 1; Boulder Creek, 
Santa Cruz County, 1; Carriso Plains, 1; Fremont Peak, Gabilan Range, 4; 
Laguna Ranch, Monterey County, 2; Mountain View, 25; Mount Hamilton, 
2; Pacheco Pass, 2; Pacheco Peak, 2; Palo Alto, 3; Portola (type locality), 
52; Priest Valley, Monterey County, 12; Redwood City, 4; San Mateo, 1; 
Santa Clara, 1; Santa Cruz Mountains (near Santa Cruz), 1; Santa Cruz 
Mountains (Woodside), 2; Searsville, 2; Soledad, 2; Walnut Creek, 1. 

NEOTOMA FUSCIPES SIMPLEX True. 
Fort Tejon Wood Rat. 

Neotoma macrotis simplex True, Proc. U. S. Nat. Mus., XVII, p. 354 [author's sepa- 
rates issued June 27], 1894. Type from Fort Tejon, Cal.; No. 3651, U. S. National 
Museum ; collected by John Xantus. 

Neotoma fuscipes dispar Merriam, Proc. Biol. Soc. Wash., IX, pp. 124-125, July 2, 1894. 
Type from Lone Pine, Inyo County, Cal.; No. fff£i, $ ad., U. S. National 
Museum (Biological Survey Collection); collected by Vernon Bailey, December 
25, 1890. 

Distribution. — Eastern basal slopes of the Sierra Nevada in Inyo 
and Kern counties, Cal., and through Walker Pass to the foothill 
region at the southern end of the San Joaquin Valley. Upper 
Sonoran Zone. 

General characters.— Size somewhat smaller than N. fuscipes; col- 
oration pale ochraceous, feet white, tail bicolor. 

Color. — Fresh pelage: Upper parts ochraceous buff, palest on head, 
darkened along back by a moderate admixture of black -tipped 
hairs; feet, throat, chest, and inguinal region pure white; belly 
washed with creamy buff, the fur basally plumbeous except along 
median line, where the hairs are white to roots; ears grayish; tail 
grayish brown above, whitish below. 

STcull. — In general similar to that of N. fuscipes, but smaller; inter- 
pterygoid fossa much broader; bullae smaller; incisive foramina 
longer than palatal bridge, as in N. f. streatori and N.f macrotis (in 
N. fuscipes about equal). 



92 NORTH AMERICAN FAUNA. [wo. 81. 

Measurements. — An adult topotype: Total length, 390; tail ver- 
tebrae, 175; hind foot, 39. SJcuU (of same): Basilar length, 40.8; 
zygomatic breadth, 25.3; interorbital breadth, 5.9; length of nasals, 
19.4; length of incisive foramina, 10.8; length of palatal bridge, 7.3; 
alveolar length of upper molar series, 8.7. 

Remarks. — This subspecies is a pale form interposed between the 
ranges of streatori and macrotis, and intergrading with both. In the 
southern part of the Sierra Nevada in Tulare County simplex passes 
into streatori, which ranges thence to the northward along the west 
slope of the mountains. The type of N. f. dispar and other speci- 
mens from Lone Pine east of the Sierra Nevada are richer ochraceous 
and show a greater degree of differentiation than topotypes of simplex, 
but slight geographic and individual variations are numerous in the 
general region, and dispar can not satisfactorily be separated. 

Specimens examined. — Total number 40, from the following 
localities : 

California: Antelope Valley, 1; Buena Vista Lake (north side), 4; Fort Tejon 
(type locality), 5; Gorman, 1; Kern River (15 miles northeast of Bakers- 
field), 1; Kern River (25 miles above Kernville), 1; Lone Pine, 13; Onyx, 
south fork of Kern River, 3; Piute Mountains, 3, San Emigdio, Kern 
County, 3; Santiago Springs, Kern County, 1 ; Tehachapi, 1; Tejon Canyon, 
1; Walker Pass, 2. 

NEOTOMA FUSCIPES MOHAVENSIS Elliot. 

Mohave Desert Wood Rat. 

Neotoma fuscipes rnohavensis Elliot, Field Columb. Mus., Zool. Ser., Ill, pp. 246-247, 
December, 1903. Type from Oro Grande, Mohave Desert, San Bernardino 
County, Cat; type in Field Museum of Natural History, Chicago; collected by 
Edmund Heller. 

Distribution. — The Mohave Desert, in southern California. Lower 
Sonoran Zone. 

General characters. — Smaller than N. fuscipes; color grayish; tail 
long, bicolor. Similar to N.f. macrotis, but more grayish instead of 
ochraceous. 

Color. — Fresh pelage: Above brownish gray, palest on face and 
sides, dorsally strongly mixed with black, giving the animal a some- 
what grizzled appearance; underparts white, in some specimens 
varying to creamy white across belly; feet white, the hind ones 
more or less streaked with dusky to toes; ears grayish; tail brownish 
black above, dull white below. 

Skull. — As in N. f. macrotis. 

Measurements. — Average of 6 adult topotypes: Total length, 386; 
tail vertebrae, 182; hind foot, 37.8. STcull (average of same): Basilar 
length, 38.8; zygomatic breadth, 24.2; interorbital breadth, 5.6; 
length of nasals, 18.3; length of incisive foramina, 10.1; length of 
palatal bridge, 8; alveolar length of upper molar series, 8.8. 



1910.] FUSCIPES GROUP — MACROTIS. 93 

Remarks. — N. f. mohavensis is a pale desert form very closely 
related to macrotis. Concerning its range Mr. Elliot says: "At the 
headwaters of the Mohave, in the San Bernardino Mountains, this 
race meets N. f. macrotis of the Coast Slope." Along the eastern 
slopes of the mountains north of the Mohave Desert it may inter- 
grade with N.f. simplex. 

Specimens examined. — Total number 25, from the following 
localities : 

California: Fish Creek, San Bernardino Mountains, 1; Mohave, 7; Morongo 
Pass, 2; Oro Grande (type locality), 15. 

NEOTOMA FUSCIPES MACROTIS Thomas. 

Large-eared Wood Rat. 

(PI. VII, figs. 2, 2a.) 

Neotoma macrotis Thomas, Ann. and Mag. Nat. Hist., 6 Ser., XII, pp. 234-235, 
September, 1893. Type from San Diego, San Diego County, Cal.; (in alcohol) 
No. 93. 2. 2. 4, $ , British Museum; collected by Prof. Eigenmann. 

Neotoma fuscipes cnemophila Elliot, Pub. Field Columb. Mus., Zool. Ser., Ill, No. 
15, pp. 267-268, March, 1904. Type from Lockwood Valley, Mount Pinos, 
Ventura County, Cal.; No. 12781, $ ad., Field Museum of Natural History, 
Chicago; collected by Edmund Heller, October 6, 1903. 

Distribution. — Pacific coast region from Monterey Bay, California, 
south through the San Pedro Martir Mountains, Lower California. 
Upper Sonoran and Transition zones. 

General characters. — Size smaller than N. fuscipes: color grayer; 
tail distinctly bicolor; skull smaller; interpterygoid fossa much 
broader; incisive foramina decidedly longer than palatal bridge. 
Closely related to N. f. mohavensis, but color darker. 

Color. — Fresh pelage: Upper parts grayish brown, more or less 
suffused with buff or ochraceous buff, palest on head and along 
sides, darkened dorsally by black-tipped hairs; under parts white, 
the fur plumbeous at base across the belly, as in N. fuscipes (in 
some specimens the basally plumbeous area is overlaid with pale 
buffy); ankles dusky; forefeet white; hind feet clouded with dusky 
to toes, toes white; tail brownish black above, whitish below. 

Skull. — Much smaller than that of N. fuscipes: incisive foramina 
relatively longer, reaching anterior plane of first molars; palatal 
bridge shorter; interpterygoid fossa much broader; bullae smaller, 
shorter, and more rounded. Skull closely resembling that of N. f. 
streatori, but interorbital region narrower, and ascending branches 
of premaxillae more slender. 

Measurements. — Average of 7 adult topotypes: Total length, 359; 
tail vertebrae, 170; hind foot, 37. Skull (average of same): Basilar 
length, 37.6; zygomatic breadth, 23.4; interorbital breadth, 5.4; 
length of nasals, 18.3; length of incisive foramina, 9.5; length of 
palatal bridge, 7.6;- alveolar length of upper molar series, 8.7. 



94 NORTH AMERICAN FAUNA. [NO. 31. 

Remarks. — N. f. macrotis is a well-marked subspecies forming one 
of the extremes of the fuscipes series of intergrading forms. Its 
range extends northward along the coast and through the Santa 
Lucia Mountains to Salinas, meeting along the Salinas Valley the 
range of annectens, the other extreme of the series. Specimens from 
Mount Pinos, the type locality of cnemophila, and from the other 
mountains of that vicinity show considerable variation in color, 
depending on the altitude and slope exposure, and grade toward 
simplex. Along the mountain slopes to the south and west of the 
Mohave Desert, macrotis intergrades with mohavensis. 

Specimens examined. — Total number 336, from the following 
localities : 

California: Alhambra, 7; Ballena, 1; Bergman, 1; Buckhorn Canyon, Santa 
Barbara County, 1; Cajon Pass, 1; Capistrano, 2; Carmel River, 21; Car- 
penteria, 1; Coahuila Mountain, 1; Colton, 1; Cone Peak, 2; Cuyama Valley, 
1; Cuyamaca Peak, 2; Del Monte, 6; Dulzura, 25; El Nido, 4; Hesperia, 2 
Indian Canyon, Santa Barbara County, 1; Indian Valley, Monterey County 
1; Jacumba, San Diego County, 2; Jamesburg, 1; Jolon, 6; King City, 3 
Las Virgines Creek, Los Angeles County, 6; Little Pine Canyon, Santa 
Barbara County, 1; Lytle Creek, 25 miles northwest of San Bernardino, 1 
Mansfield, Monterey County, 8; Mesa Grande, 1; Monterey, 4; Morro, 3 
Mount Pinos, 8; Nordhoff, 3; Oceanside, 1; Pacific Grove, 12; Paraiso 
Springs, 7; Pasadena (Arroyo Seco Canyon), 2; Pasadena (Oak Knoll), 1 
Paso Robles, 5; Pine Valley, Monterey County, 7; Pleyto, 1; Port Harford 
4; Posts, 4; Pozo, 3; Salinas, 5; San Bernardino, 7; San Bernardino Moun 
tains, 14; San Diego (type locality), 25; San Fernando, 3; San Gabriel, 1 
San Gabriel Mountains (Heninger Flats), 4; San Gorgonio, 1; San Jacinto 
Mountains, 7; San Luis Obispo, 1; San Miguelito, Monterey County, 2 
mountains near San Simeon, 2; Santa Inez, 1; Santa Isabel, 2; Santa Lucia 
Peak, 4; Santa Monica, 1; Santa Paula, 5; Seaside, 1; Sur, 1; Sur River, 6 
miles from mouth, 4; Sweetwater, San Diego County, 2; Tassajara Creek, 
Monterey County, 7; Tejunga Valley, Los Angeles County, 2; Temescal, 6; 
Twin Oaks, 6; Ventura River, Ventura County, 2; Wildomar, 2; Wilson 
Peak, 1; Witch Creek, San Diego County, 4. 

Lower California: El Rayo, Laguna Hansen Mountains, 1; Ensenada, 1; 
La Grulla, San Pedro Martir Mountains, 7; Laguna Hansen, Laguna Hansen 
Mountains, 4; Nachoguero Valley, 3; Pifion, northwest base San Pedro 
Martir Mountains, 3; Rancho San Antonio, west base San Pedro Martir 
Mountains, 1; Rancho Santo Tomas, San Pedro Martir Mountains, 3; 
Rosarito Divide, 1; San Antonio, 1; San Matias Pass, San Pedro Martir 
Mountains, 4; Santa Eulalia, 3; San Tomas, 1; Tecate Valley, 1; Tijuana 
River (mouth), 1; Vallecitos, San Pedro Martir Mountains, 3. 

Subgenus TEONOMA Gray. 

(PI. VIII.) 

Type. — Neotoma cinerea drummondi (Richardson), from Rocky Mountains, in 

latitude 57°. 
Teonoma Gray, List Spec. Mamm., British Museum, p. 117, 1843. 
Teonoma Merriam, Proc. Acad. Nat. Sci. Phila., p. 242, September 24, 1894 (subgenus). 



1910. J CINEREA GROUP — CINEREA. 95 

Distribution. — Mountains of North America west of longitude 100°, 
from latitude 35° to 60°. 

Subgeneric characters. — Tail large, bushy, and somewhat distichous ; 
sole of hind foot normally densely furred from heel to posterior 
tubercle. Skull large and angular; temporal ridges prominent, 
diverging posteriorly to near anterior border of interparietal, whence 
they turn abruptly inward and again outward in crossing inter- 
parietal to lambdoid crest; frontal region narrow, constricted near 
middle, somewhat depressed, and excavated above along median 
line; maxillary arms of zygomata broad and heavy; bullae large; 
interpterygoid fossa narrow. 

Remarks. — The bushy-tailed wood rats form a rather well-defined 
group regarded by some authors as generic in rank. Some of the 
forms differ widely in size and external appearance, but all are very 
closely related and doubtless intergrade. Merriam has shown that 
N. c. arizonse, although a bushy-tailed species, approaches Neotoma 
proper in cranial characters. A study of the group indicates the 
close relationship of arizonse to N. c. orolestes, which in turn inter- 
grades through N. cinerea with the type form of Teonoma. In the 
Neotominse, differentiation appears in cranial and dental modifica- 
tions, and Teonoma agrees with members of the subgenus Neotoma so 
closely in these more essential respects that the bushy tail must be 
regarded as a comparatively superficial character of not more than 
subgeneric value. 

NEOTOMA CINEREA (Ord.).& 

Gray Bushy-tailed Wood Rat. 

(PI. VIII, figs. 1, la.) 

Mus cinereus Ord, Guthrie's Geog., 2d Amer. ed., II, p. 292, 1815. Type from Great 
Falls, Cascade County, Mont.: "Based on the ash-colored rat of Rocky Mountains 
of Lewis and Clark." 

Neotoma cinerea Baird, Mamm. North Amer., p. 499, 1857. 

Teonoma cinerea acraia Elliot, Pub. Field Columb. Mus., Zool. Ser., Ill, pp. 247-248, 
December, 1903. Type from Hot Springs, Long Canyon, Mount Whitney, Inyo 
County, Cal.; No. 12850, $ , Field Museum of Natural History; collected by Ed- 
mund Heller, July 31, 1903. 

Distribution. — Rocky Mountain region in southern British Colum- 
bia, Montana, Idaho, western Wyoming, Utah, northern Arizona, 

« Proc. Biol. Soc. Wash., VIII, pp. 109-110, July 31, 1893. 

b An extinct species ( Teonoma spelsea Sinclair, Bull. Dept. Geol. Univ. Calif., Pub., 
IV, No. 7, p. 148, PI. 19, July 19, 1905) has been described, based on material from 
Potter Creek Cave, Shasta County, Cal. This material has not been examined by 
me, but on comparing the drawings accompanying the description with numerous 
skulls of N. cinerea I am unable to find any essential differences, except possibly 
the more nearly equal extent of the inner and outer salient angles in the crown of 
the second lower molar of the fossil species. 



96 



NORTH AMERICAN FAUNA. 



[no. 31. 



and thence westward through the mountains of central Nevada to the 
southern part of the Sierra Nevada in California. Canadian Zone 




Fig. 14.— Distribution of the subgenus Teonoma ( Neotoma cinerea and subspecies'). 

and down along cold cliffs and canyons well into the Transition 
Zone. 



1910.] CINEREA GROUP CINEREA. 97 

General characters. — Very large; tail moderately long, bushy, and 
somewhat distichous; hind feet very large; the soles thickly furred 
from heel to posterior tubercle; ears large; fur long, thick, and some- 
what woolly; rostrum elongated; sphenopalatine vacuities absent 
or small. 

Color. — Fresh pelage: Upper parts grayish buff, in some specimens 
varying to ochraceous buff, palest on face, the back much darkened 
by admixture of dusky hairs; lower surface, including upper lip, 
white, the hairs along sides of belly pale plumbeous basally; forefeet 
white, this color insensibly passing over fore limbs into general 
color of upper parts; hind feet white; color of hind legs reaching a 
short distance over tarsus; ears covered with brownish and grayish 
hairs, and faintly edged with whitish; tail above brownish gray, 
below white, usually crossed basally by a pale buffy band. 

Skull. — Skull large, long, and angular; rostrum elongated; frontal 
region narrow, depressed, deeply constricted near middle, chan- 
neled above, the well-developed lateral ridges somewhat approxi- 
mated, the sides of frontals rounded off and bulging below them; 
brain case short, not prolonged anteriorly between orbits; zygoma 
with squamosal root widely spreading, upper surface of maxillary 
root very broad, the outer border passing forward in an evenly 
convex curve to antorbital foramen ; antorbital foramina narrow and 
somewhat constricted above; nasals long, narrowing posteriorly, not 
usually reaching anterior plane of orbits; ascending branches of 
premaxillse passing posteriorly well beyond nasals; temporal ridges 
prominent, diverging posteriorly to near anterior border of inter- 
parietal, whence they turn abruptly inward and again diverge 
slightly in crossing interparietal to lambdoid crest; interparietal 
subquadrate or rectangular; palatal slits much longer than palatal 
bridge; palate usually convex posteriorly; sphenopalatine vacuities 
absent, or present as very narrow slits; bullae large and somewhat 
elongated; teeth similar to those of N. mexicana, but maxillary 
toothrow less narrowed posteriorly; first upper molar with deep 
anterointernal reentrant angle and last upper molar with an anterior 
closed triangle and two confluent posterior loops, as in N. mexicana; 
condyloid process of mandible long and upturned. 

Measurements. — Average of 2 adults from Big Snowy Mountains, 
Montana: Total length, 387; tail vertebras, 162; hind foot, 43. 
Skull (an adult from same locality): Basilar length, 46.1; zygomatic 
breadth, 29.5; interorbital breadth, 5.6; length of nasals, 20.9; length 
of incisive foramina, 12.5; length of palatal bridge, 9.2; alveolar 
length of upper molar series, 10. 

Remarks. — This species and its subspecies form a series of slightly 
differentiated forms which include all the bushy-tailed wood rats. 
52668°— 10 7 



98 NORTH AMERICAN FAUNA. [no. 31. 

The forms in which the sphenopalatine vacuities are large intergrade 
with those in which they are completely closed, so that this charac- 
ter, although remarkably constant, is merely of subspeciiic value. 
Although N. cinerea presents considerable geographic variation in 
color, specimens from the southern part of the Sierra Nevada in Cali- 
fornia are perfectly typical, and there are no characters by which to 
recognize T. c. acraia. Those from Fiddle Creek, Idaho, and Flat- 
head Lake, Montana, are grading toward occidentalis. In the 
eastern part of its range, specimens often closely approach orolestes 
in color, but may be separated by the very small or absent spheno- 
palatine vacuities. Specimens from southern Utah and northern 
Arizona north of the Grand Canyon agree with cinerea in color and 
in the absence of large sphenopalatine vacuities, but those from 
some localities have rather larger, more inflated audital bullae. 
Comparison of specimens of cinerea from Bright Angel Spring, on 
the Kaibab Plateau, with arizonse, from the opposite side of the Colo- 
rado River at Lees Ferry indicates the complete isolation of the two in 
this region by the Colorado River. The Colorado River thus appears 
to limit the dispersal of forms as effectively in the cinerea group 
•as in the desertorum group of wood rats. The fresh autumn pelage 
usually replaces the worn summer coat in late August or September. 
Specimens examined. — Total number 184, from the following 
localities : 

Arizona: Bright Angel Spring, Kaibab Plateau, 3. 

Montana: Bass Creek (northwest of Stevensville), 4; Big Snowy Mountains, 
5; Billings, 2; Birch Creek, 3; Bozeman, 1; Columbia Falls, 1; Darby, 2; 
Flathead Lake, 8; Florence, 4; Fort Assiniboine, 8; Gardiner, 1; Great 
Falls (type locality), 2; Miles City, 1; Milk River, 1; Ravalli, 1; Red 
Lodge, 1; Rock Creek, 1; Stanton Lake, 1; Terry, 1; Upper Stillwater 
Lake, 9. 

Idaho: Bear Lake (east side), 4; Birch Creek, 7; Challis, 1; Fiddle Creek, 2; 
Lost River Mountains (near Arco), 5; Lumhi, 5; Pahsimeroi Mountains, 1; 
Sawtooth, 1. 

Utah: Beaver Mountains (Britt Meadows), 3; Blacksmith Fork, Cache County, 
3; Fish Lake Plateau, 1; Henry Mountains, 3; Ogden, 3; Parawan, 1; 
Provo, 3. 

Wyoming: Fort Bridger, 1. 

Nevada: Arc Dome, Nye County, 4; Carson City, 3; Monitor Mountains, 1; 
Newark Valley (20 miles east of Eureka), 2; head of Reese River, 1; Reno, 
2; White Rock Valley (30 miles southwest of Austin), 1. 

California: Atwells Mill, Tulare County, 1; Bishop Creek, Inyo County, 2; 
Bronco, Nevada County, 3; Dormer, 17; Emerald Bay, 6; Horse Corral 
Meadows, Fresno County, 1; Inyo Mountains, 2; Kern River (Soda Springs, 
north fork), 2; Lake Tenaya, Mariposa County, 1; Little Yosemite Valley, 
1; Mineral King, Tulare County, 1; Mono Lake, 2; Mono Pass (east side), 
2; Mount Unicorn, 1; Mount Whitney, 10; White Mountains, 2; Wood- 
fords, 12. 



1910.] CINEREA GROUP — DRUMMONDI. 99 

NEOTOMA CINEREA DRUMMONDI (Richardson). 

Bushy-tailed Wood Rat. 

Myoxus drummondii Richardson, Zool. Journ., Ill, pp. 517-518, 1828. Type from 
"Rocky Mountains in latitude 57°;" probably near Jasper House, Alberta, 
Canada; No. 42.10.7.6, 9 , British Museum. 

Neotoma drummondii Richardson, Fauna Boreali-Amer., pp. 137-140, 1829. 

Neotoma cinerea drummondi Merri&m, Proc. Acad. Nat. Sci. Phila., September 24, 1894. 

Distribution. — Rocky Mountains of eastern British Columbia and 
western Alberta, north of the range of N. cinerea. Canadian Zone. 

General characters. — Similar to N. cinerea, but fur longer; tail 
more bushy on distal two- thirds; dark color of forelegs ending in a 
sharp line near carpus, in strong contrast with pure white of feet; 
sphenopalatine vacuities absent or very small. 

Color. — Fresh pelage: Above grayish buff moderately overlaid with 
dusky; head slightly paler; underparts, including upper lip, whitish; 
orbital rings dusky; ears clothed with rather long grayish brown hairs 
and edged with whitish; feet white; tail above brownish gray, 
sometimes white at tip, below white, becoming pale buff across base. 

STcull. — Closely resembling that of N. cinerea, but averaging 
slightly larger; dentition slightly heavier; sphenopalatine vacuities 
closed or nearly closed, as in N. cinerea. 

Measurements. — Average of 10 adults from Jasper House, Alberta: 
Total length, 382; tail vertebrae, 167; hind foot, 44. STcull (average 
of 4 adults from same locality): Basilar length, 43.7; zygomatic 
breadth, 26.9; interorbital breadth, 6.3; length of nasals, 19.2; 
length of incisive foramina, 11.7; length of palatal bridge, 9.5; 
alveolar length of upper molar series, 10.6. 

Remarks. — This slightly differentiated form doubtless passes on the 
south directly into N. cinerea, and in southern British Columbia, 
west of the Rocky Mountains, probahly intergrades with N. c. 
occidentalis. Although Richardson says : a " This animal inhabits the 
Rocky Mountains, in latitude 57°," it is evident from his account of 
Drummond's route that the type specimen came from farther south, 
in the vicinity of Jasper House, near latitude 53°. 

Specimens examined. — Total number 26, as follows: 

Alberta: Jasper House (near type locality), 23. 
British Columbia: Glacier House, 2. 
Mackenzie: Fort Liard, 1. 

o Fauna Boreali-Americana, p. 137, 1829- 



100 NORTH AMERICAN FAUNA. [no. 31. 

NEOTOMA CINEREA SAXAMANS Osgood. 

Osoood Bushy-tailed Wood Rat. 

(PI. VIII, figs. 2, 2a.) 

Neotoma saxamans Osgood, North Amer. Fauna, No. 19, pp. 33-34, October 6, 1900. 

Type from Bennett, head of Lake Bennett, British Columbia, Canada; No. 98923, 

$ ad., U. S. National Museum (Biological Survey Collection); collected by W. H. 

Osgood, June 19, 1899. 
Neotoma cinerea saxamans Allen, Bull. Amer. Mus. Nat. Hist., XIX, pp. 544-546, 

October 10, 1903. 

Distribution. — Northern British Columbia west of the Rocky 
Mountains, limits of range unknown. Canadian and Hudsonian 
zones. 

General characters. — Similar in size to N. cinerea, but color decidedly 
darker; tail more bushy; cranial characters pronounced. Closely 
related to N. c. drummondi, but color darker; skull differing in same 
characters as from N. cinerea. 

Color. — Fresh pelage (September) : Upper parts buffy gray, bright- 
est on shoulders and along flanks, darkened everywhere by admixture 
of dusky hairs; under parts white, fur faintly darkened basally along 
sides of belly and hind legs by pale plumbeous; feet white; upper side 
of tail brownish gray on proximal third, becoming slaty gray on distal 
two-thirds, under side white, except a buffy gray band at base. 
Worn pelage (June) : Above more conspicuously buffy than in fresh 
pelage. 

Skull. — Similar in general form to that of N. cinerea, but differing 
in important details; nasals more attenuate, and deeply but narrowly 
emarginate posteriorly, reaching plane of orbits, the posterior ends 
acutely pointed; interorbital region narrow; interparietal broader 
transversely, and sagittal area very wide at this point, the temporal 
impressions turning less abruptly inward; sphenopalatine vacuities 
present, but short and somewhat triangular, widest at palatopterygoid 
suture; ascending branches of premaxillse rather broad, but narrow- 
ing abruptly posteriorly on reaching frontals; basioccipital with a 
sharp, well-developed median ridge; last upper molar variable, the 
crown sometimes trifoliate. 

Measurements. — Average of 4 adult topotypes: Total length, 
409; tail vertebrae, 173; hind foot, 47. Skull (average of same): 
Basilar length, 46; zygomatic breadth, 27.5; interorbital breadth, 
5.5; length of nasals, 21.5; length of incisive foramina, 12.1; length 
of palatal bridge, 9.1; alveolar length of upper molar series, 10. 

Remarks. — In color and cranial characters saxamans departs con- 
siderably from cinerea. Specimens from Inverness, Stuart Lake, and 
River Inlet, British Columbia, however, are nearer occidentalis, and 
complete intergradation seems probable, although not satisfactorily 



1910.] CINEKEA GROUP — OCCIDENTALS. 101 

shown by present material. In both adults and young the third 
lower molar sometimes has a small additional outer reentrant angle, 
anterior in position to the one normally present. In one adult an 
unusual development of the additional reentrant angle had led to an 
abnormal elongation of the tooth. In nearly full-grown young acces- 
sory peglike nonfunctional teeth are sometimes present. They 
appear at the posteroexternal base of the third lower molar. An 
additional reentrant angle is also developed sporadically in N. m. 
fallax and other members of the mexicana group. 
Specimens examined. — Eighteen, from: 

British Columbia: Bennett (type locality), 14; Inverness, 2; River Inlet, 1; 
Stuart Lake, 1. 

NEOTOMA CINEREA OCCIDENTALS (Cooper MS.) Baird. 

Western Bushy-tailed Wood Rat. 

Neotoma occidentalis (Cooper MS.) Baird, Proc. Acad. Nat. Sci., Phila., VII, p. 335, 
1855. Type from Shoalwater Bay, Pacific County, Wash.; No. 572, $ ad., U. S. 
National Museum; collected by J. G. Cooper in 1854. 

Neotoma cinerea occidentalis Merriam, North Amer. Fauna, No. 5, p. 58, July 30, 1891. 

Neotoma c[inerea] Columbiana Elliot, Pub. Field Columb. Mus., Zool. Ser., I, No. 13, 
p. 255, March, 1899. Type from Ducks, British Columbia; No. 4910, Field 
Museum of Natural History, Chicago; collected by C. P. Streator, August 1, 1889. 

Distribution. — From Pacific coast region of southwestern British 
Columbia to northern California (except the narrow coastal strip 
west of the Cascade Mountains in Oregon), and thence eastward over 
the lava beds to south-central Idaho and northeastern Nevada. 
Mainly Transition and Canadian zones. 

General characters. — Size and general appearance of N. cinerea, but 
color darker; ankles more conspicuously dusky; tail bushy and 
sphenopalatine vacuities absent or very small, as in N. cinerea. 

Color. — Fresh pelage (October): Upper parts brownish buff, much 
obscured by dusky hairs which thin out somewhat and leave the 
sides more decidedly buffy; head slightly paler; under parts dull 
white, the hairs pale plumbeous basally, except over pectoral and 
inguinal regions; axillae ochraceous buff; forefeet white, contrasting 
strongly with brownish buff of forearms; hind feet white; ankles 
dusky all around; ears brownish, edged with whitish; tail above 
brownish black, mixed with gray, below white with a buffy band 
across base. Worn pelage (June to August) : Worn summer speci- 
mens average slightly grayer, and old adults are more rusty brown 
than in fresh autumn pelage. Young: Much darker at all ages than 
in N. cinerea. 

Skull. — As in N. cinerea. 

Measurements. — An adult from Olympic Mountains, Washington: 
Total length, 412; tail vertebrae, 180; hind foot, 45. Average of 



102 NORTH AMERICAN FAUNA. [no. 31. 

4 adults from Spokane Bridge, Wash.: 388; 168; 44.5. Skull (aver- 
age of 4 adults from Spokane Bridge, Wash.): Basilar length, 43.7; 
zygomatic breadth, 26.8; interorbital breadth, 5.8; length of nasals, 
20.7; length of incisive foramina, 11.6; length of palatal bridge, 9.1; 
alveolar length of upper molar series, 9.4. 

Remarks. — This subspecies is so much darker than N. cinerea that 
typical specimens are easily separable at all ages. Although ranging 
over widely differing faunal areas, variation in color is not great, and 
specimens from the dry lava beds of northern Nevada and the Snake 
River Valley, in Idaho, agree well with those from the humid forested 
districts nearer the coast. The explanation of this remarkable con- 
stancy of color may be that the dark lava has had the same effect on 
the coloration of the animal as the forested areas, where a dark 
environment is due to climatic conditions. The specimens examined 
si iow that intergradation with N. cinerea occurs along the western base 
of the Rocky Mountains in Montana and Idaho, in north-central 
Nevada, and north of Lake Tahoe in California. Intergradation with 
N. c.fusca occurs at Bissell and Eagle Creek, Oreg. The fresh autumn 
pelage replaces that of summer about the month of September, and 
continues to grow, becoming longer and fuller until winter. N. c. 
columbiana was based on a nearly typical specimen of N. c. occidentalis . 

Specimens examined. — Total number 261, from localities as follows: 

Washington: Almota, 1; Cascade River (head), 1; Campbell's Ranch, Clallam 
County 1; Cheney, 2; Columbia River (opposite John Day River), 
1; Colville, 2; Conconully, 1; Coulee City, 1; Douglas, 2; Fort Spokane, 
2; Goldendale, 5; Hamilton, 1; Marcus, 3; Marshall, 4; Neah Bay, 1; 
Orondo, 2; Quiniault Lake, 1; Rockland, 1; Shoalwater Bay (type locality), 
2 (including type); Soleduc River, Olympic Mountains, 1; Spokane 
Bridge, 7. 

California: Adin, 4; Bald Mountain (8 miles south), Shasta County, 1; Beswick, 
2; Big Valley Mountains, Lassen County, 6; Bronco, 3; Brownell, 1; 
Bunch Grass Spring, Lassen County, 2; Burney, 2; Canyon Creek, 7; 
Dana, 1; Eagle Lake, 2; Goose Lake, 5; Hayden Hill, 1; Honey Lake, 2; 
Lassen Creek, Modoc County, 3; Lassen Peak, 5; Little Shasta, 1; Lower 
Alkali Lake, 1; Madeline Divide, Lassen County, 1; Madeline Plains, 4; 
Milford, 2; Mount Shasta, 4; Old Fort Crook, 1; Orick, 1; Pine Creek, 
Lassen County, 1; Prattville, 1; Preston Peak, 2; Requa, 2; Salmon Sum- 
mit, 1; South Yolla Bolly Mountain, 4; Secret Valley, Lassen County, 1; 
Siskiyou Mountains, 8; Susan ville, 6; Tuledad Canyon, Lassen County, 2; 
Tule Lake, 2; Willow Creek (15 miles east), 2. 

Idaho: Arco, 1; Big Butte, 2; Mullan, 1; Priest Lake, 5; Shoshone Falls, 1. 

Nevada: Alder Creek, Pine Forest Mountains, 2; Bull Run Mountains, 1; 
Cottonwood Range, 2; Granite Creek, Humboldt County, 2; Mountain 
City, 7. 

Oregon: Bissell, 1; Burns, 2; Christmas Lake, 1; Crater Lake, 1; Crooked Lake, 
1; Eagle Creek, 8 miles southeast of Bissell, 1; Empire, 1; Fort Klamath, 
7; Happy Lake, Clallam County, 14; Harney, 12; John Day River 
(Crown Rock), 1; Lake Alvord, 2; Lonerock, 5; Mount Jefferson, 1; Nar- 
rows, 1; Naylox, Klamath County, 2; Pendleton, 1; Plush. 7: Prospect, 



1910.] CINEREA GROUP — FUSCA. 103 

4; Shirk, 3; Siskiyou, 7; The Dalles, 2; Tule Lake, 1; Twelve Mile Creek, 
1; Vale, 3; Wallowa Lake, 1. 
British. Columbia: Ashcroft, 6; Carpenters Mountain, Caribou, 2; Chilliwack, 
1; Ducks, 1; Midway, 1; Nelson, 1, Okanagan, 1; Shuswap, 3; Sicamous, 2. 

NEOTOMA CINEREA FUSCA True. 

Fuscous Bushy-tailed Wood Rat. 

Neotoma occidentalis fusca True, Proc. U. S. Nat. Mus., XVII, p. 354, November 15, 
1894 [author's separates issued June 27, 1894]. Type from Fort Umpqua, Douglas 
County, Oreg.; No. 3370, U. S. National Museum; collected by E. P. Vollum. 

[Neotoma] cinerea fusca Trouessart, Catal. Mamm., p. 544, 1897. 

Neotoma fuscus apicalis Elliot, Pub. Field Columb. Mus., Chicago, Zool. Ser., Ill, 
pp. 160-1G1, April, 1903. Type from Gardiner, Coos County, Oreg.; type in 
Field Museum of Natural History; collected by Edmund Heller in 1901. 

Distribution.— Humid coastal belt west of the Cascade Mountains 
in Oregon. Transition Zone. 

General characters. — Similar to N. cinerea, but color very much 
darker; fur more woolly; ears smaller; upper surface of metatarsus 
dusky on proximal half ; tail bushy, lower surface not white ; spheno- 
palatine vacuities absent. 

Color. — Fresh pelage (October): Upper parts black, mixed with 
grayish tawny, the black predominating dorsally, and black hairs 
everywhere standing out conspicuously; sides brighter; head more 
gray than back; belly and chin whitish, more or less encroached 
upon by general color of upper parts, the line of demarcation indis- 
tinct; hairs on pectoral and inguinal regions pure white to roots; 
inner sides of legs grayish tawny; throat irregularly suffused with 
tawny ochraceous; muzzle and ears dusky, the latter faintly edged 
with whitish; forefeet and under sides of wrists white; hind feet 
white, except proximal half of metatarsus, which is dusky above, 
like ankles; tail above blackish, the thick, woolly, grayish underfur 
showing through, below buffy gray, obscured by blackish hairs. 
Winter pelage (January): Longer and fuller, but not different in 
color from fresh fall pelage. 

Skull. — As in N. cinerea. 

Measurements. — Type: Total length, 395; tail vertebra?, 180; 
hind foot, 44. An adult from Newport, Oreg.: 410; 190; 48. Skull 
(from Newport): Basilar length, 45; zygomatic breadth, 27.9; inter- 
orbital breadth, 5.7; length of nasals, 21.4; length of incisive foram- 
ina, 13.1; length of palatal bridge, 9.4; alveolar length of upper 
molar series, 10.1. 

Remarks. — In external appearance this subspecies differs widely from 
cinerea, but intergradation through occidentalis is probably complete. 
Its range is limited to a narrow strip mainly along the low Coast 
Range west of the Cascade Mountains, where its differentiation is 



104 NORTH AMERICAN FAUNA. [no. 31. 

due to an environment of heavy forest resulting from excessive 
humidity in a relatively warm climate. Specimens of occidentalis 
from Bissell and Eagle Creek, on the west slope of the Cascade Moun- 
tains, are darker than usual in occidentalis and show gradation 
toward fusca. 

Neotoma fuscus apicalis was based on dark-colored specimens with 
tails whitish along under side, and in five out of six more or less 
broadly tipped with pure white. As specimens with white-tipped 
tails occur sporadically in other forms of the cinerea group, this 
can not be regarded as a character of taxonomic importance. The 
name apicalis seems to belong in synonymy under fusca, represent- 
ing specimens approaching occidentalis in lighter color of under parts. 
A specimen in the Biological Survey, also from Gardiner, is typical 
fusca. 

Specimens examined. — Total number 26, all from localities as 

follows : 

Oregon: Beaverton, 2; FL rence, 10; Fort Umpqua, 1 (the type); Gardiner, 7; 
Newport, 1; Seaton, 1; Wells, 3; Yaquina Bay, 1. 

NEOTOMA CINEREA OROLESTES Merriam. 

Colorado Bushy-tailed Wood Rat. 

(PL VIII, figs. 3, 3er.) 

Neotoma orolestes Merriam, Proc. Biol. Soc. Wash., IX, p. 128, July 2, 1894. Type from 
Saguache Valley, 20 miles west of Saguache, Saguache County, Colo. ; No. f f fff , 
$ ad., U. S. National Museum (Biological Survey Collection); collected by J. 
Alden Loring, August 13, 1892. 

Neotoma grangeri Allen, Bull. Amer. Mus. Nat. Hist., VI, pp. 324-325, November 7, 
1894. Type from Custer, S. Dak.; No. Iffy, $ ad., American Museum Natural 
History, New York; collected by W. W. Granger, August 4, 1894. 

Neotoma cinnamomea Allen, Bull. Amer. Mus. Nat. Hist., VII, pp. 331-332, November 
8, 1895. Type from Kinney Ranch, Bitter Creek, Sweetwater County, Wyo. ; No. 
VAV> $ a d-, American Museum Natural History, New York; collected by 
W. W. Granger, July 9, 1895. 

Distribution. — Rocky Mountain region from northern New Mexico 
north through Colorado and Wyoming to southern Montana and 
thence eastward to the Black Hills in South Dakota. Mainly 
Transition and Canadian zones. 

General characters. — Similar in general to N. cinerea, but color 
usually more ochraceous; sphenopalatine vacuities large; tail bushy, 
as in N. cinerea. Differs from both rupicola and arizonse in larger size 
and much paler color. / 

Color. — Fresh pelage (October): Upper parts ochraceous buff, 
purest along sides, darkened over back by blackish hairs; head slightly 
paler; under parts and feet dull white, the hairs over pectoral and 
inguinal regions white to roots; tail above on proximal third grayish 



1910.] CINEREA GROUP — OROLESTES. 105 

buff, becoming brownish buff on distal two-thirds; below white with 
a more or less distinct pale buffy band across base. Young (about 
half grown) : Darker than in N. cinerea. 

Skull. — Similar to that of N. cinerea, but sphenopalatine vacuities 
large (small or absent in cinerea); presphenoid usually more con- 
stricted. Much like that of rupicola, but larger. 

Measurements. — Type: Total length, 413; tail vertebrae, 175; hind 
foot, 41. Average of 3 adults from Costilla Pass, N. Mex.: 394; 169; 
40. Skull (type): Basilar length, 44.5; zygomatic breadth, 28; 
interorbital breadth, 5.6; length of nasals, 19.7; length of incisive 
foramina, 11.3; length of palatal bridge, 9.8; alveolar length of upper 
molar series, 9.3. 

Remarks. — The presence of large sphenopalatine vacuities dis- 
tinguishes orolestes from cinerea, which many specimens closely 
resemble in color. This character, although only of subspecific 
value, is very constant. It is shared with subspecies rupicola and 
arizonx. Specimens of orolestes from northern Colorado and south- 
western Wyoming approach rupicola; those from the river valleys of 
western Colorado grade toward arizonse. In northern and western 
Wyoming orolestes passes into cinerea. Topotypes of N. grangeri 
do not differ from nearly typical specimens of orolestes. Specimens 
from the type locality of N. cinnamomea are paler than typical orolestes 
and grade toward rupicola. 

Specimens examined. — Total number 137, from localities as follows: 

Colorado: Almont, 2; Boulder County, 8; Crested Butte, 2; East Fork of Rifle 
Creek, Garfield County, 2; Elk Mountains, 1; Estes Park, 9; Gold Hill, 6; 
Grand Junction, 1; Lay, 3; Lily, 1; Longs Peak, 1; Meeker, 1; Mount Mar- 
vine, 1; Rangely, 1; Rifle (8 miles west), 1; Saguache Valley (type locality), 
3; Snake River (near Sunny Peak), 4; White River (20 miles east of 
Rangely), 1. 

New Mexico: Agua Frio, 1; Amizett, 1; Chama, 1; Coyote Creek, 4; Costilla 
Pass, 7; Halls Peak, 4; Hermit, 1; Jemez Mountains, 1; Martinez, 1; Moreno 
Valley, 3; Pecos Baldy, 3; Taos, 2; Taos Mountains, 5; Tierra Amarilla, 1; 
Tres Piedras, 1; Twining, 4; Willis, 1. 

Montana: Pryor Mountains, 2. 

South Dakota: Custer, 6; Deadwood, 2; Elk Mountain, 2. 

Wyoming: Bridger Paes, 1; Bull Lake, Wind River Mountains, 5; Casper (18 
miles southwest), 1; Casper Mountains, 1; Deer Creek, Converse County, 1; 
Devils Tower, 2; Ferris Mountains, 1 ; junction Green River and New Fork, 
1; Islay, 5; Kinney Ranch, Sweetwater County, 7; Lake Fork, Wind River 
Mountains, 3; Laramie Peak, 1; Powder River, mouth of Clear Creek, 1; 
Shoshone Mountains, head of Wind River, 2; Sun, 1; Wind River Basin, 2; 
Woods, 1. 



106 NORTH AMERICAN FAUNA. [no. 31. 

NEOTOMA CINEREA ARIZONA Merriam. 

Arizona Wood Rat. 

(PI. VIII, figs. 5, 5a.) 

Neotoma arizonse Merriam, Proc. Biol. Soc. Wash., VIII, pp. 110-111, July 31, 1893. 
Type from Keam Canyon, Apache County, Ariz.; No. fffi ? ad., Merriam 
Collection; collected by J. Sullivan, May 21, 1888. 

Distribution. — Upper Sonoran Zone in northeastern Arizona, 
southeastern Utah, and probably northward along the Green River 
Valley, southwestern Colorado, and northwestern New Mexico. 

General characters. — Similar in general to N. cinerea, but much 
smaller; tail less bushy; color more buffy ochraceous; sphenopalatine 
vacuities large (small or absent in N. cinerea). Nearest to N. c. 
orolestes, but smaller; color paler; audital bullae larger. 

Color. — Fresh pelage: Upper parts varying from pale to bright 
ochraceous buff, purest along cheeks and sides, thinly overlaid with 
dusky; feet and under parts white; tail grayish brown above, white 
below. 

Skull. — Skull much smaller, less angular than that of N. cinerea; 
frontal region less depressed, not deeply channeled above; temporal 
ridges more widely separated; interparietal broader between tem- 
poral ridges; bullae larger, more rounded; sphenopalatine vacuities 
large as in N. c. orolestes (small or absent in N. cinerea); otherwise 
differing from N. c. orolestes as from N. cinerea. 

Measurements. — Average of 2 adults from Holbrook, Ariz.: Total 
length, 347; tail vertebrae, 146; hind foot, 36. Skull (type): Basilar 
length, 38.8; zygomatic breadth, 24.3; interorbital breadth, 5.4; 
length of nasals, 17.5; length of incisive foramina, 10.3; length of 
palatal bridge, 8.2; alveolar length of upper molar series, 9.1. 

Remarks. — Arizonse. differs from cinerea in a number of compara- 
tively unimportant characters, and intergradation through orolestes 
seems certain. Specimens of orolestes from northern New Mexico 
and western Colorado are grading toward arizonse. The bushy tail, 
always present in the cinerea group, diminishes progressively in size 
from north to south, drummondi having the largest and arizonse the 
smallest. The pale coloration of arizonse is probably a result of its 
desert environment. 

Specimens examined. — Total number 31, as follows: 

Arizona: Holbrook, 10; Keam Canyon (type locality), 5; Lees Ferry (south 

side of Colorado River), 1; Walpi, 1; Winslow, 5. 
Colorado: Ashbaugh Ranch, Montezuma County, 3; Coventry, 1. 
Utah: Bluff, 1. 
New Mexico: Chaca (Canyon Bonito), 1; Fruitland, 2; Shiprock, 1. 



1910.] CINEREA GROUP — RUPICOLA. 107 

NEOTOMA CINEREA RUPICOLA Allen. 

Pallid Busht-talled Wood Rat. 

(PI. VIII, figs. 4, 4a.) 

Neotoma rupicola Allen, Bull. Amer. Mus. Nat. Hist., VI, pp. 323-324, November 7, 
1894. Type from Corral Draw, southeastern base of Black Hills, S. Dak. ; No. f f f^, 
$ ad., American Museum of Natural History; collected by W. W. Granger, 
August 21, 1894. 

Distribution.— Rig Bad Lands region from southwestern South 
Dakota, through southeastern Wyoming and western Nebraska to 
northeastern Colorado. Upper Sonoran Zone. 

General characters. — Similar in general to N. cinerea, but smaller 
and color much paler; tail bushy, as in N. cinerea, but shorter; 
sphenopalatine vacuities large, as in N. c. orolestes. 

Color. — Fresh pelage (September): Upper parts cream buff, palest 
on head, moderately darkened over back by black- tipped hairs; 
under parts and feet snowy white; ears brownish gray edged with 
white; tail brownish gray above, lighter toward tip, pure white below. 
Worn summer pelage: Above paler with black-tipped hairs less con- 
spicuous than in autumn. Young: Much paler than in N. cinerea. 

Skull. — In general form similar to that of N. cinerea, but smaller; 
dentition heavier; nasals reaching posteriorly to plane of orbits; 
sphenopalatine vacuities large, as in N. c. orolestes. 

Measurements. — Average of 5 adult topotypes: Total length, 349; 
tail vertebrae, 144; hind foot, 43.2. Skull (average of same): Basilar 
length, 40.2; zygomatic breadth, 25.7; interorbital breadth, 5.8; 
length of nasals, 19.1; length of incisive foramina, 10.8; length of 
palatal bridge, 8.9; alveolar length of upper molar series, 9.6. 

Remarks. — This subspecies is closely related to orolestes, its near 
geographical neighbor in the Black Hills, but may readily be known 
by its small size and remarkably pallid coloration. Some of the 
topotypes taken early in September are already in fresh pelage, 
while others are still in the molting stage, with fresh fur appearing 
in patches. 

Specimens examined. — Total number 41, from localities as follows: 

Colorado: Avalo, 5; Pawnee Buttes, 3. 

Nebraska: Glen, 5; Warbonnet Canyon, 2. 

South Dakota: Corral Draw (type locality), 23; Quinns Draw, Cheyenne 

River, 1. 
Wyoming': Bordeaux, 1; Uva, 1. 



PLATE I. 

[About natural size.] 

Fig. 1, la. Neotoma floridana (Ord). Enterprise, Fla. (No. 72571, <? ad., U. S. Nat. 
Mus.). 

2, 2a. Neotoma floridana rubida Bangs. Houma, La. (No. 46144, $ ad., U. S. 

Nat. Mus.). 

3, 3a. Neotoma floridana baileyi Merriam (type). Valentine, Nebr. (No. 5034, 

9 ad., Merriam Collection). 

4, 4a. Neotoma micropus Baird. Nueces Bay, Texas. (No. 43552, 9 & &> U. S. 

Nat. Mus.). 

5, 5a. Neotoma micropus planiceps Goldman (type). Rio Verde, San Luis 

Potosi, Mexico (No. 82105, $ ad., U. S. Nat. Mus.). 

6, 6a. Neotoma pennsylvanica Stone. Renovo, Pa. (No. 57846, $ ad., U. S. 

Nat. Mus.). 
108 



North American Fauna No. 31, U. S. Dept. Agr. Biological Survey. 



Plate I. 




Skulls of Neotoma. 



1, la. N. floridana. 

2, 2a. N. f. rubida. 

3, 3a. N. f. baileyi. 



A, 4a. N. micropus. 

5, 5a. N. m. planiceps. 

6, 6a. N. pennsylvanica. 



PLATE II. 

[About natural size.] 

Fig. 1, la. Neotoma albigula Hartley. Fort Grant, Ariz. (No. 92699, 9 a d-. U. S. 
Nat. Mus.). 

2, 2a. Neotoma albigula venusta True. Fort Yuma, Cal. (No. 99492, 9 ad- 

U. S. Nat. Mus.). 

3, 3a. Neotoma albigula melanura Merriam. Alamos, Sonora, Mexico (No. 

96289, $ ad., U. S. Nat. Mus.). 

4, 4a. Neotoma albigula leucodon Merriam. Chicalote, Aguas Calientes, Mex- 

ico (No. 79010, $ ad., U. S. Nat. Mus.). 

5, 5a. Neotoma latifrons Merriam (type). Querendaro, Michoacan, Mexico 

(No. 50135, $ ad., U. S. Nat. Mus.). 

6, 6a. Neotoma nelsoni Goldman (type). Perote, Veracruz, Mexico (No. 54320, 

9 ad., U. S. Nat. Mus.). 
110 



North American Fauna No. 31, U. S. Dept. Agr. Biological Survey. 



Plate II. 




Skulls of Neotoma. 



1, la. N. albigula. 

2, 2a. N. a. venusta. 

3, 3a. N. a. melanura. 



4, 4a. N. a. leucodon. 

5, •">". X. latifrons. 
ti, (in. N. nelscmi. 



PLATE III. 

[About natural size.] 

Fig. 1, la. Neotoma palatina Goldman (type). Bolafios, Jalisco, Mexico (No. 90959, 
2 ad., U. S: Nat. Mus.). 

2, 2a. Neotoma montezumse Goldman (type). Zimapan, Hidalgo, Mexico (No. 

81426, $ ad., U. S. Nat. Mus.). 

3, 3a. Neotoma intermedia Rhoads. Tijuana, Lower California (No. 81886, $ 

ad., U. S. Nat. Mus.). 

4, 4a. Neotoma intermedia pretiosa Goldman (type). Matancita, Lower California, 

Mexico (No. 146123, $ ad., U. S. Nat. Mus.). 

5, 5a. Neotoma intermedia arenacea Allen (topotype). San Jose del Cabo, 

Lower California, Mexico (No. 146716, $ ad., U. S. Nat. Mus.). 

6, 6a. Neotoma bryanti Merriam (topotype). Cerros Island, Lower California, 

Mexico (No. 81078) $ ad., U. S. Nat. Mus.). 
112 



North American Fauna No. 31, U. S. Dept. Agr. Biological Survey. 



Plate III. 




Skulls of Neotoma. 



1, la. N. palatina. 

2, 2a. N. montezuma'. 

3, 3a. N. intermedia. 



I. in. N. i. pretiosa. 

5, 5a. N. i. arenacea. 

6, 6a. N. bryanti. 






52668°— 10 8 



PLATE IV. 

[About natural size.) 

Fig. 1, la. Neotoma anthonyi Allen (topotype). Todos Santos Island, Lower Cali- 
fornia, Mexico (No. 137176, ? ad., TJ. S. Nat. Mus.). 

2, 2a. Neotoma martinensis Goldman (type). San Martin Island, Lower Cali- 

fornia, Mexico (No. 81074, 9 ad., U. S. Nat. Mus.). 

3, 3a. Neotoma mexicana Baird. Santa Eulalia, Chihuahua, Mexico (No. 57605, 

9 ad., U. S. Nat. Mus.). 

4, 4a. Neotoma mexicana fallax Merriam (type). Gold Hill, Colo. (No. 6345, $ 

ad., Merriam Collection). 

5, 5a. Neotoma mexicana pinetorum Merriam (type). San Francisco Mountain, 

Arizona (No. 246281, 9 ad., U. S. Nat. Mus.). 

6, 6a. Neotoma mexicana bullata Merriam (type). Santa Catalina Mountains, 

Arizona (No. 23774, $ ad., U. S. Nat. Mus.). 
114 



North American Fauna No. 31, U. S. Dept. Agr. Biological Survey. 



Plate IV. 




Skulls of Neotoma. 



1, la. N. anthonyi. 

2, 2a. N. martinensis. 

3, 3a. N. mexicana. 



4, -la. N. m. fallax. 

5, 5a. N. m. pinetcirum. 

6, 6a. N. m. bullata. 



PLATE V. 

[About natural size.] 

Fig. 1, la. Neotoma navus Merriam (topotype). Sierra Guadalupe, Coahuila, Mexico 
(No. 116896, $ ad., U. S. Nat. Mus.). 

2, 2a. Neotoma torquata Ward. Mount Popocatepetl, Mexico (No. 52029, $ ad., 

U. S. Nat. Mus.). 

3, 3a. Neotoma distincta Bangs (topotype). Texolo, Veracruz, Mexico (No. 

126418, $ ad., U. S. Nat. Mus.). 

4, 4a. Neotoma ferruginea chamula Goldman (type). San Cristobal, Chiapas, 

Mexico (No. 76061, ? ad., U. S. Nat. Mus.). 

5, 5a. Neotoma ferruginea solitaria Goldman (type). Nenton, Guatemala (No. 

76908, $ ad., U. S. Nat. Mus.). 

6, 6a. Neotoma ferruginea isthmica Goldman (type). Huilotepec, Oaxaca, Mex- 

ico (No. 73187, 9 ad., U. S. Nat. Mus.). 
116 



North American Fauna No. 31, U. S. Dept. Agr. Biological Survey. 



Plate V. 




Skulls of Neotoma. 



1, la. N. navus. 

2, 'in. X. torquata. 

3, 3a. N. distinotu. 



4, 4a. N. f. chamula. 

5, 5a. N. f. solitaria. 

6, 6a. N. f. isthmica. 



PLATE VI. 

[About natural size.] 

Fig. 1, la. Neotoma ftrruginea picta Goldman (type). Mountains near Chilpancingo, 
Guerrero, Mexico (No. 70050, $ ad., U. S. Nat. Mus.). 

2, 2a. Neotoma parvidens Goldman (topotype). Juquila, Oaxaca, Mexico (No. 

71582, 9 ad., U. S. Nat. Mus.). 

3, 3a. Neotoma tropicalis Goldman (type). Totontepec, Oaxaca, Mexico (No. 

68593, $ ad., U. S. Nat. Mus.). 

4, 4a. Neotoma desertorum Merriam. Furnace Creek, Death Valley, California 

(No. 33139, $ ad., IT. S. Nat: Mus.). 

5, 5a. Neotoma lepida stephensi Goldman (type). Hualpai Mountains, Arizona 

(No. 117466, 9 ad., U. S. Nat. Mus.). 

6, 6a. Neotoma goldmani Merriam (type). Saltillo, Coahuila, Mexico (No. 

116894, $ ad., U. S. Nat. Mus.). 
118 



North American Fauna No. 31, U. S. Dept. Agr. Biological Survey. 



Plate VI. 




Skulls of Neotoma. 



1, la. N. f. picta. 
•J, 2a. X. parvidens. 

3, 3a. N. tropical is. 



I, la. X. desertorum. 
5, 5a. X. 1. stephensi. 
t), 6a. X. goldnmni. 



PLATE VII. 

[About natural size, except figures 4, 16, which are enlarged two diameters.] 

Fig. 1, la. Neotoma(Homodontomys)fuscipes Baird. Marshall, California (No. 140421, 
9 ad., U. S. Nat. Mus.). 
16. Neotoma (Homodontomys) fuscipes Baird. Big Valley Mountains, Cali- 
fornia (No. 48980, $ ad., U. S. Nat. Mus.). 

2, 2a. Neotoma (Homodontomys) fuscipes macrotis Thomas (topotype). San Diego, 

California (No. 46186, ? ad., U. S. Nat. Mus.). 

3, 3a. Neotoma (Homodontomys) fuscipes annectens Elliot. Mountain View, 

California (No. 96164, $ ad., U. S. Nat. Mus.). 
4. Neotoma fioridana (Ord). Riceboro, Georgia (No. 45065, $ ad.,U. S. Nat. 
Mus.). 
120 



North American Fauna No. 31, U. S. Dept Agr. Biological Survey. 



Plate VII. 




Skulls and Upper Molars of Neotoma. 



1, la, lb. N. fuscipes. 

2, 2d. N. f. macrotis. 



3, 3a. N. f. annectens. 

4. N. floridana. 



PLATE VIII. 

[About natural size.) 

Fig. 1, la. Neoloma (Teonoma) cinerea (Ord). Big Snowy Mountains, Montana. 

(No. 67244, 9 ad., U. S. Nat. Mus.). 
2,2a. Neotoma (Teonoma) cinerea saxamans Osgood (type). Bennett, British 

Columbia (No. 98923, $ ad., U. S. Nat. Mus.). 
3,3a. Neotoma (Teonoma) cinerea orolestes Merriam (type). Saguache Valley, 

Colorado (No. 48215, $ ad., U. S. Nat. Mus.). 
4, 4a. Neotoma (Teonoma) cinerea rupicola Allen (topotype). Corral Draw, South 

Dakota (No. 138726, $ ad., U. S. Nat. Mus.). 
5,5a. Neotoma (Teonoma) cinerea arizonse Merriam (type). Keam Canyon, 

Arizona (No. 4980, 9 ad., Merriam Collection). 
122 



North American Fauna No 31, U. S. Dept. Agr. B ological Survey. 



Plate VIII. 




Skulls of Neotoma. 



1, 1'/. N. cinerea. 

2, 2a. N. c. saxamans. 

3, 3a. N. e. orolestes. 



4, la. X. c. rupicola. 

5, 5a. N. c. arizonse. 



INDEX. 



(New names in bold-face type; synonyms in italics.) 



Characters, general, 11-12. 
Distribution, 8-9. 

Economic status, habits and, 10-11. 
Habits and economic status, 10-11. 
History and material, 7-8. 
Hodomys, 13, 20. 
Homodontomys, subgenus, 8G. 
Key to species and subspecies, 16-19. 
Key to subgenera, 15. 
List of species and subspecies, 14-15. 
Material, history and, 7-8. 
Measurements, explanation of, 12. 
Mus americanus, 86. 
cinereus, 95. 
floridanus, 21. 
Myoxus drummondii, 99. 
Nelsonia, 13. 
Neotoma, genus, 13. 

subgenus, 20. 
Neotoma abbreviata, 49-50. 

affinis, 90, 91. 

albigula, 31-33. 

angusticeps, 31. 

annectens, 90-91. 

anthonyi, 52-53. 

apicalis, 103. 

arenacea, 47. 

arizonse, 106. 

attwateri, 26. 

baileyi, 24-25. 

bella, 76, 78. 

bryanti, 51-52. 

bullata, 59. 

californica, 42, 44. 

campestris, 24, 25. 

canescens, 28-29. 

chamula, 69. 

chrysomelas, 74-76. 

cinerea, 95-98. 

cinnamomea, 104, 105. 

cnemophila, 93, 94. 

Columbiana, 101, 102. 

cumulator, 34. 

desertorum, 76-78. 

dispar, 91, 92. 

distincta, 64-65. 

drummondi, 99. 

drummondii, 99. 

durangse, 37-38. 

fallax, 56-58. 

felipensis, 45. 

ferruginea, 67-68. 



Neotoma floridana, 21-22. 
Julvivcntcr, 63, 64. 
fusca, 103, 104. 
fuscipes, 87-89. 
gilva, 44-46. 
goldmani, 81-82. 
grandis, 34. 
grangeri, 104, 105. 
illinoensis, 23-24. 
intermedia, 42-44. 
isthmica, 71-72. 
latifrons, 38-39. 
lepida, 79-80. 
leucodon, 36-37. 
littoralis, 29-30. 
macrotis, 93-94. 
madrensis, GO. 
magister, 82-83. 
martinensis, 53-54. 
melanura, 35-36. 
mexicana, 54-56. 
micropus, 26-28. 
mohavensis, 92-93. 
monochroura, 87, 89. 
montezumae, 41-42. 
navus, 61-63. 
nelsoni, 39-40. 
nevadensis, 76, 78. 
nudioauda, 51. 
occidentalis, 101-103. 
ochracea, 74. 
orizabx, 03, 64. 
orolestes, 104-105. 
palatina, 40-41. 
parvidens, 66-67. 
pennsylvanica, 84-86. 
perpallida, 48-49. 
picta, 72-73. 
pinetorum, 58-59. 
planiceps, 30-31. 
pretiosa, 46-47. 
rubida, 22-23. 
rupicola, 107. 
saxamans, 100-101. 
simplex, 91-92. 
sinaloae, 60-61. 
sola, 44-45. 
solitaria, 70. 
splendens, 87, 89. 
stephensi, 80-81. 
streatori, 89-90. 
surbcrl, 27, 28. 



123 



124 



INDEX. 



Neotoma tenuicauda, 73-74. 

torquata, 63-64. 

tropicalis, 65-66. 

venusta, 33-34. 

vicina, 48. 

warreni, 34-35. 

lacatecae, 38. 
Neotominae, subfamily, 13. 
Neotomodon, 13. 
Ptyssophorus, 13. 
Species and subspecies, key to, 16-19. 



Species and subspecies, list of, 14-15. 
Subgenera and minor groups, 13. 

key to, 15. 
Teanopus, 13, 20. 
Teonoma acraia, 95, 98. 

spelaea, 95. 

subgenus, 94. 
Tretomys, 13, 20. 
Type localities, list of, 14-15. 
Xenomys, 13, 20. 



o 






U. S. DEPARTMENT OF AGRICULTURE 

BUREAU OF BIOLOGICAL SURVEY 

HENRY W. HENSHAW, ChieJ 



NORTH AMERICAN FAUNA 

]STo. 32 

[Actual date of publication, April 29, 1911] 




A SYSTEMATIC SYNOPSIS OF THE 
MUSKRATS 



N. HOLLISTER 

ASSISTANT CURATOR OF MAMMALS, UNITED 

STATES NATIONAL, MUSEUM; FORMERLY 

ASSISTANT, BIOLOGICAL SURVEY 




WASHINGTON 

GOVERNMENT PRINTING OFFICE 
1911 



LETTER OF TRANSMITTAL. 



U. S. Department of Agriculture, 

Bureau of Biological Survey, 
Washington, D. C, January 27, 1911. 
Sir: I have the honor to transmit herewith, for publication as 
North American Fauna No. 32, a Systematic Synopsis of the Musk- 
rats, by N. Hollister, formerly Expert in the Bureau of Biological 
Survey. Our fur bearers have been so reduced in numbers in recent 
years that the muskrat has become of great economic importance 
because of the utility of its fur. In addition its flesh is valuable for 
food. The animals are likely to be still more important in future, 
especially as it has been found practicable to raise them on flooded 
marsh land of no agricultural value. In such areas muskrats may 
be protected and the supply maintained indefinitely. In other 
places they cause serious breaks by burrowing in embankments, and 
the damage done in this way is far in excess of their value as fur 
bearers. Because of its economic relations the muskrat has been 
the subject of numerous legislative enactments. The animal is 
widely distributed over North America, where there are several 
species and numerous subspecies, the interrelations and ranges of 
which have hitherto not been well understood. The present report 
is therefore timely and important, as the ranges of the several forms 
have been worked out and mapped and the whole subject brought 
up to date. 

Respectfully, Henry W. Henshaw, 

Chief, Biological Survey. 
Hon. James Wilson, 

Secretary of Agriculture. 

3 



CONTENTS. 



Page. 

Introduction 7 

History 7 

Distribution and habits 9 

Economic relations 10 

Characters and pelage 11 

Key 12 

Fossil muskrats 13 

Material and acknowledgments 13 

Genus Fiber 14 

List of species and subspecies, with type localities 14 

Existing species 15 

Fossil species 32 

Tables of measurements 34 

Bibliography 35 

Index 47 

5 



ILLUSTRATIONS. 



Page. 

Plate I. Map of the distribution of the muskrats Frontispiece 

II. The muskrat in its haunts 10 

III. Skulls (dorsal views) of Fiber zibcthicus, F. z. macrodon, F. z. spatu- 

latus, F. z. zalophus, F. obsmrus, F. z. albus, F. z. ripensis 40 

IV. Skulls (ventral views) of Fiber zibethicus, F. z. macrodon, F. z. spatu- 

latus, F. z. zalophus, F. obscurus, F. z. albus, F. z. ripensis 42 

V. Skull of F. z. osoyoosensis. Enamel pattern of upper molar teeth of 
Fiber. Enamel pattern of lower molar teeth of Fiber. Hind feet 

of Fiber 44 

VI. Fiber z. macrodon, F. z. cinnamominus , right mandibular ramus. 

Types of F. annectens, F. oregonus, F. nebracensis 46 

6 



No. 32. NORTH AMERICAN FAUNA. April, 1911. 



A SYSTEMATIC SYNOPSIS OF THE MUSKRATS. 



By N. Hollister. 

INTRODUCTION. 
HISTORY. 

The conspicuous North American animals which have no repre- 
sentatives in the Old World, including the skunk, opossum, musk- 
rat, hummingbird, turkey, and others, early attracted the attention 
of explorers and accounts of them were soon taken to Europe. The 
first published description of the muskrat appears in Captain John 
Smith's Description of Virginia, 1612. x Smith writes: " Mussascus 
is a beast of the forme and nature of our water Bats, but many of them 
smell exceeding strong of muske." Under the name of mussascus, 
musquash, or ondatra there are numerous references to the animal in 
the literature of the period succeeding Smith's work. Linnaeus, 
however, at first confounded the American muskrat with his Castor 
moscliatus (== Desmana moschata) of Asia, and it was not until 1766 
that he gave it an independent place in his Systema Naturae, placing 
it in the same genus with the beaver and naming it Castor zibeihicus. 
The specific name zibethicus has been used for the form from eastern 
Canada and the northern United States by most systematists to the 
present day. Tiedemann, in 1808, renamed it Ondatra americana, 2 
and Oken, 1816, used the same specific name. The Linnsean specific 
name has, however, been associated between 1788 and 1840 with no 
less than eight generic names. 

From time to time various mammalogists have described forms of 
muskrats from many parts of North America, giving them specific 
or subspecific rank as the limited material before them seemed to 
justify. Seventeen names have been thus proposed, and with the idea 

1 A Map of Virginia. With a Description of the Countrey, the Commodities, People, 
Government, and Religion. Written by Captaine Smith, sometimes Governour of 
the Countrey, p. 14, 1612. 

2 Zoologie, 1,481, 1808. 

7 



8 NORTH AMERICAN FAUNA. [no. 32. 

of determining their respective validity and importance and of map- 
ping the ranges of the recognizable forms, a study of the large collec- 
tion of skins and skulls of muskrats in the United States National 
Museum was undertaken. 

Sabine, in the Zoological Appendix to Franklin's Narrative, in 
1823, describes a light-colored specimen from Cumberland House, 
Saskatchewan, as Fiber zibethicus albus; and Richardson, in the 
Fauna Boreali- Americana, in 1829, has three varieties, B, C, and D, 
also based upon abnormally colored examples, which he designates as 
nigra, maculosa, and alba. The names nigra and maculosa of Rich- 
ardson are easily disposed of as synonyms of Fiber zibethicus or Fiber 
z. albus, but Sabine's albus itself, with a perfect diagnosis and definite 
type locality, must unfortunately, according to present rules of 
nomenclature, be used for the recognizable geographic race found in 
that region. In 1863, Lord, in the Proceedings of the Zoological 
Society of London, described a species of muskrat from southern 
British Columbia as Fiber osoyoosensis, but this form seems to have 
been lightly considered and his name usually placed in synonyin}^. 

The genus was generally considered monotypic up to 1890, when 
Mearns described a form from Camp Verde, Arizona, as Fiber zibe- 
hicus pallidus. 1 During the next thirteen years eight new muskrats 
were named. Bangs described Fiber obscurus from Newfoundland 
in 1894 2 and Fiber z. rivalicius from Louisiana the following year. 3 
Merriam described Fiber macrodon from the Dismal Swamp of Vir- 
ginia in 1897. 4 Bangs, in 1899, proposed the name aquilonius for the 
Labrador muskrat, 5 and Osgood, in 1900, named the animal from 
northwest America Fiber spatulatus. 6 In 1902 Bailey described a 
new muskrat from the Pecos River, Fiber z. ripensis, 7 and Preble pro- 
posed the name hudsonius for the Keewatin race. 8 The following 
year Elliot described Fiber occipitalis from the coast of Oregon. 9 
No more names were proposed for existing species until 1910, when, 
as a result of the systematic revision in progress by the present writer, 
three recognizable forms found to be still unnamed were described, 
zalophus from the Alaska Peninsula, 10 mergens from Nevada, 10 and 
cinnamominus from Kansas. 11 

•Bull. Amer. Mus. Nat. Hist., II, 280, 1890. 
2 Proc. Biol. Soc. Wash., IX, 133, 1894. 
3 Proc. Boston Soc. Nat. Hist., XXVI, 541, 1895. 
4 Proc. Biol. Soc. Wash., XI, 143, 1897. 
fi Proc. New England Zool. Club, I, 11, 1899. 
6 North Amer. Fauna, No. 19, p. 36, 1900. 
7 Proc. Biol. Soc. Wash., XV, 119, 1902. 

8 North Amer. Fauna, No. 22, p. 53, 1902. 

9 Publ. Field. Col. Mus., Zool. Ser., Ill, 162, 1903. 
I0 Proc. Biol. Soc. Wash., XXIII, 1, 1910. 

11 Proc. Biol. Soc. Wash., XXIII, 125, 1910. 



1911.] HABITS. 9 

DISTRIBUTION AND HABITS. 

Muskrats occur over the greater part of North America from the 
northern limit of trees south to the Mexican border. They are not 
found along the lower Atlantic seaboard, nor, except in southern Lou- 
isiana, over the entire Gulf region. They are unknown also on the 
Pacific slope south of central Oregon. 

As might be expected from the muskrat's extensive distribution, 
its habits vary considerably with local conditions. Over the greater 
part of its range it is noted as a builder of marsh houses, and these 
heaps of aquatic vegetation are a characteristic feature of the marsh 
landscape. The houses, chiefly for winter shelter, are sometimes of 
great size, though commonly the home of a single family. The nest 
chamber is in the center of the heap, above water line, with tunnels 
for entrance and exit running out below the surface. Not all musk- 
rats, however, build houses. Where abrupt banks take the place of 
low, marshy shores, many of the animals seem to prefer holes in 
these banks. In this case the burrows extend from an underwater 
entrance through the bank to a dry nest chamber, near the surface, 
above high-water mark. In many places muskrat houses are un- 
known, all the animals living in these bank homes. 

By far the greatest part of the muskrat's food is vegetable matter 
and many kinds of aquatic and shore-growing plants help make up 
its bill of fare. It often travels a considerable distance from water 
at night to feed on some especially favorite food. There is good 
evidence that the muskrat sometimes eats animal matter, fresh- 
water mussels especially, and occasionally fish, dead birds, and other 
animals. Doctor Mearns observed a muskrat fishing in the Verde 
River, Arizona, and notes it "occasionally coming out upon a log to 
eat the fish it caught." * Mr. E. R. Warren, in his recently pub- 
lished Mammals of Colorado, records the following: 

One [muskrat] was seen in a lake near Crested Butte chasing under water a "water 
dog," Amblystoma tigrinum, which it finally captured by making a sudden dash 
forward and seizing it with its teeth. The rat then came to the surface with its prey 
in its mouth, and not until then was it seen to be a muskrat, for while the chase was 
in progress the observers supposed it to be a mink. 

Breeding habits doubtless vary somewhat with climatic condi- 
tions. Prof. D. E. Lantz, after calling attention to the wide variance 
in the published accounts of the breeding habits of the muskrat, 
gives his information from the best-informed trappers in Maryland. 
The most reliable evidence shows that in this region from 3 to 5 
Utters (normally 3) are produced annually, and that the number 
of young in a litter varies from 3 to 12, or even more, the average 

1 Bull. U. S. Nat. Mus., No. 56, p. 496, 1907. 
7Q5Gno_-n 2 



10 NORTH AMERICAN FAUNA. [no. 32. 

being probably 6 or 8. The young of early spring litters are said to 
breed the fall of the same year. Professor Lantz further summarizes: 

Normally the animals mate in March and the first litter is born in April; a second 
litter is due in June or early July, and a third in August or September. In favorable 
seasons a fourth or even a fifth litter may be produced. The period of gestation is 
possibly no longer than twenty-one days, as with the common rat and probably with 
the field mouse. The young are blind and naked when born, but develop rapidly. 
Outside of low marshes, muskrats are usually born in the underground burrows. 1 

Where the muskrats are not able to find suitable banks for nesting 
burrows, the young are usually born in open nests, made for the 
purpose in the drier parts of the marsh home. 

Actual breeding records noted on labels of female specimens are as 
follows: Summit, Mont., June 18, 1895, 13 large fetuses (V. Bailey); 
Ward, Colo., June 8, 1893, 8 fetuses (J. A. Loring); Newport, R. I., 
April 18, 1900, 6 small fetuses (Dr. E. A. Mearns). E. R. Warren, in 
The Mammals of Colorado, gives breeding records from that State 
as follows: Grand County, May 12, eight good-sized embryos; Lily, 
Routt County, June 1, young 2 or 3 weeks old; Barr, Adams 
County, May 30, very small young, not much larger than adult 
Microtus modestus; Medano Ranch, Costilla County, June 24, seven 
embryos, second litter. 

ECONOMIC RELATIONS. 

The great and ever-increasing demand from the furrier and the con- 
sequent rise in the price of muskrat skins make the animal one of great 
economic importance. During the past few years especially, the 
price of the fur has steadily increased, until during the season of 
1909-10 the choicest dark skins in prime condition netted the trapper 
close to $1 each. Though this exceptionally high price may not be 
maintained, the proper recognition of the beauty of this fur will 
insure its steady favor. With the rapid extermination of the rarer 
fur bearers, especially those species which can not«adapt themselves 
to the changes wrought by the settlement of the country, the muskrat 
will soon become the most important fur-bearing animal of North 
America. With proper protection it should furnish a constant supply 
of choice fur and add to the wealth of the country for generations to 
come. Its well-known adaptability to changed conditions is a strong 
point in its favor. 

While the damage the rodent inflicts on crops is not severe, it 
sometimes destroys grain and vegetables for a limited distance from 
the water's edge. The chief complaint against it, however, is on 
account of the injury it does by burrowing into dams and embank- 
ments of ditches and levees. Instances of serious loss to property 
from this source are numerous, and in certain places unceasing war- 

1 U. S. Dept. Agr., Farmers' Bulletin No. 396, p. 15, 1910. 



North American Fauna No. 32, U. S. Dept. Agr. Biological Survey. 



Plate II. 




mil.] CHARACTERS AND PELAGE. 11 

fare against the burrowers is necessary. In most localities, however, 
the animal's value as a fur bearer justifies its protection throughout 
the breeding season and the months when the pelage is not at its very 
best. 

Of late years the flesh of the muskrat has been extensively sold in 
eastern markets, and during the open season many thousands are 
used as food in the larger cities. In fur farming the chances of suc- 
cess with the muskrat are much greater than with any of the other 
fur bearers, and many reserves are now paying handsome profits to 
owners and trappers. 

CHARACTERS AND PELAGE. 

All existing forms of muskrats are closely related. The majority 
are geographic races of one species, and blend in characters from one 
to another. As a rule the recognized forms are well characterized 
and over a considerable definite area are constant in color, size, and 
salient cranial characters. It is sometimes difficult to identify with 
a particular race specimens from intermediate areas, especially if 
they are immature, or of pelage not comparable with the specimens it 
becomes desirable to study in that connection. In most cases the 
material has been sufficient to settle the boundaries of subspecies with 
considerable accuracy, though many additional specimens will be 
necessary to fill in the numerous gaps in ranges, as shown by the 
colored map. 

The pelage of the muskrat is made up of a thick underfur, the main 
coat, and long overlying darker hairs, which come in slowly as the 
season advances. The great number of specimens examined from all 
parts of the range show that all but one form have only one molt, 
and this occurs during the warm summer months. Fiber rivalicius, 
of the coast region of Louisiana, apparently molts twice a year, 
approximately spring and fall. 

Aside from conspicuous cases of dichromatism, the color of all 
specimens in comparable pelage from any given locality is remarkably 
uniform. The great variation frequently noticed in a large series 
from one region is due mainly to age and season and the degree 
to which the black overlying hairs have appeared. Sexuar variation 
is so absolutely wanting that it was found useless to distinguish 
between the sexes in tables of skin or cranial measurements. Skulls 
in a large series from any particular water are remarkably alike in 
shape and size, but a slight variation between series from near-by 
localities is frequently noted. In the discrimination of forms I have 
endeavored to confine the named subspecies to what appear well- 
characterized geographic races, that combine the essential char- 
acters of all the individuals over a definite area and differ from the 
individuals of all other recognized subspecies in some general and 
common characters. Long and painstaking study of large series of 



12 NORTH AMERICAN FAUNA. [no. 32. 

skulls is necessary to distinguish subspecific characters of real value 
from minor local variations. 

Since no two forms occur in the same locality and the characters 
separating the geographic races are frequently relative, it is obvious 
that an artificial key to the subspecies is of little value in identifying 
specimens. As such a key has important uses, however, as a means of 
ready reference to characters, and in other ways, the following is 
presented : 

KEY TO ADULT MUSKRATS IN FRESH PELAGE. 

Size large; hind foot averaging over 80 mm. 

Coloration darker; blackish or blackish brown. 

Skull with high, sharp interorbital ridge; nasals broadly spatulate anteriorly 

(Pnget Sound region and Rocky Mts.) osoyoosensis (p. 24). 

Skull without specially developed interorbital ridge. 

Largest in the genus; tooth row averaging over 17 mm.; coloration 
blackish (Atlantic coast, Delaware Bay to N. C). Black phase 

of macrodon (p. 18). 

Size less; tooth row averaging under 17 mm. ; coloration dark brown (S. E. 

Canada and N. E. United States) zibethicus (p. 16). 

Coloration lighter; reddish brown or grayish brown. 

Interpterygoid fossa narrow, with borders nearly parallel (western Ore- 
gon) occipitalis (p. 26). 

Interpterygoid fossa much widened posteriorly. 

Upperparts grayish brown, with darker dorsal area (Great Basin region). 

mergens (p. 27). 
Upperparts bright reddish brown; size very large (Atlantic coast, 

Delaware Bay to N. C). Normal phase of macrodon (p. 18). 

Size small; hind foot averaging less than 80 mm. 
Coloration dark; black or blackish brown. 
Skull with high, sharp interorbital ridge. 

Zygomata broadly spreading anteriorly (Alaska and N. W. British 

America) spatulatus (p. 22). 

Zygomata not broadly spreading anteriorly. 

Hind foot averaging 75 mm. ; colors darker with more rusty tinge 

(Keewatin and eastern Saskatchewan) albus (p. 20). 

Hind foot averaging less than 70 mm. ; colors lighter with little rusty 

tinge (Alaska Peninsula) zalophus (p. 23). 

Skull without distinct interorbital ridge. 
Coloration glossy blackish. 

Tail long (averaging over 260 mm.); skull large (Labrador and 

Ungava) aquilonius (p. 19). 

Tail short (averaging less than 230 mm.); skull small and weak 

(Newfoundland) obscurus (p. 15). 

Coloration dull blackish brown ; underparts dark (coast region of Louisi- 
ana) rivalicius (p. 31). 

Coloration pale; reddish or pale brown. 

Larger (tail averaging 240 mm. ; hind foot over 73 mm.) (Great Plains region), 

cinnamominus (p. 30). 

Smaller (tail averaging less than 205 mm.); hind foot less than 70 mm. 

Upperparts cinnamon rufous (Colorado River east to the Rio Grande in 

New Mexico) pallidus (p. 28). 

Upperparts Vandyke brown (Pecos Valley, Texas and New Mexico), 

ripensis (p. 29). 



1911.] FOSSIL MUSKRATS — MATERIAL. 13 

FOSSIL MUSKRATS. 

Fossil remains of muskrats have been found in Pleistocene deposits in 
various parts of the United States. As might be expected, these bones, 
chiefly fragments of skulls and jaws, indicate species identical with 
existing forms or closely related to them. Three species known only 
as fossils are recognizable, two of which are remarkable for their very 
small size, considerably less than that of any species now living; about 
the size of Neofiber alleni. Through the kindness of Mr. J. W. Gidley, 
of the United States National Museum, and Dr. W. D. Matthew, of 
the American Museum of Natural History, I have been able to study 
these fossil muskrats in connection with the living species. One spe- 
cies, Fiber annectens, from the Middle Pleistocene of the Ozark Moun- 
tains, has already been described by Mr. Barnum Brown, 1 and in 
this paper two species, from the Lower Pleistocene of Nebraska and 
Oregon, are named. 

MATERIAL AND ACKNOWLEDGMENTS. 

In the preparation of the following systematic account of the genus 
Fiber, over 1,000 specimens have been studied. These were chiefly 
well-prepared skins, accompanied by skulls, together with many odd 
skulls and skeletons. The type specimens, or virtual topotypes, of 
all described forms have been examined. The range of the animal 
has been well covered and, on the whole, the material has been suffi- 
cient to work out satisfactorily the characters and ranges of the forms. 

While the conclusions herein presented are based largely upon a 
study of the specimens in the United States National Museum (the 
collection of the Biological Survey alone contains over 500 specimens), 
I have been greatly aided by the use of various other collections. 
For the loan of material and for greatly appreciated assistance in other 
ways, it is a pleasure to acknowledge my indebtedness to Dr. C. Hart 
Merriam, to whose private collection I have had free access; to Mr. 
Charles B. Cory and Mr. W. H. Osgood, Field Museum of Natural 
History; to Dr. Joseph Grinnell and Mr. H. S. Swarth, Museum of 
Vertebrate Zoology, University of California; to Mr. Outram Bangs 
and Dr. Glover M. Allen, Museum of Comparative Zoology; to Dr. 
W. D. Matthew, American Museum of Natural History; to Mr. 
Gerrit S. Miller, jr., and Mr. J. W. Gidley, United States National 
Museum; and others, particularly various members of the Biological 
Survey. 

1 Mem. Amer. Mus. Nat. Hist., IX, pt. 4, p. 197, 1908. 



14 NORTH AMERICAN FAUNA. [no. 32. 

Genus FIBER Cuvier. 

Ondatra Lacepede, Tabl. Mamm., p. 9, 1799. Not Ondatra Link, 1795. 

Fiber Cuvier, Lecons d'Anat. Comp., I, tabl. I, 1800. 

Mussascus Oken, Lehrbuch Naturgesch., 3 ter Theil, 2 te Abth., p. 886, 1816. 

Simotes Fischer, Mem. Soc. Imp. Nat. Moscow, V, 444, 1817. 

Moschomys Billberg, Syn. Faun. Scandinavise. ed. 1, I, Mamm., Conspectus A, 1827. 

Type species. — Castor zibethicus Linnaeus. 

Geographic distribution. — Hudsonian, Canadian, Transition, and 
Austral Zones of North America, from the northern limit of trees 
south to the Mexican border of the United States; excepting the 
southern Atlantic seaboard, most of the Gulf States, and the Pacific 
coast south of middle Oregon. 

General characters.— Form robust; legs short, feet large, both 
modified for swimming; feet and toes fringed by short, stiff hairs, 
and toes of hind feet partly webbed; tail long, compressed laterally, 
covered by small scales, and thinly haired. External ear small, 
scarcely extending beyond fur. Fur dense and waterproof; pelage 
supplemented by longer glossy overlying hairs. Strongly developed 
perineal glands secreting a powerful musk; mammae six; plantar 
tubercles five. 

STcull and teeth. — Skull resembling that of Microtus and other 
related genera, but comparatively large and massive; angular, with 
heavy zygomata and long rostrum; posterior border of palate not 
bridged, but with small median spine projecting into the interptery- 
goid fossa. Audital bullae large. Upper incisors without grooves; 
lower incisor passes under m2 and outside of m3 with extremity of 
root at base of condylar process. All molars rooted; ml with an- 
terior loop and four closed triangles; m2 with anterior loop and 
three closed triangles; m3 with anterior and posterior loops and 

two or three closed triangles; ml normally with large anterior loop 
deeply cut by two reentrant angles, five closed triangles, and pos- 
terior loop; m2 with four closed triangles and posterior loop; m3 
with three or four closed triangles and posterior lobp. 

List of Species and Subspecies, with Type Localities. 

EXISTING SPECIES. 

Fiber obscurus Bangs Codroy, Newfoundland. 

zibethicus (Linnaeus) Eastern Canada. 

zibethicus macrodon Merriam Dismal Swamp, Virginia. 

zibethiciLS aquilonius Bangs Rigoulette, Labrador. 

zibethicus albus Sabine Cumberland House, Saskatchewan. 

zibethicus spatulatus Osgood Lake Marsh, Yukon. 

zibethicus zalophus Hollister Becharof Lake, Alaska. 

zibethicus osoyoosensis Lord Lake Osoyoos, British Columbia. 

zibethicus occipitalis Elliot Florence, Oreg. 

zibethicus mergens Hollister Fallon, Nev. 



1911.] FIBER OBSCUKUS. 15 

Fiber zibethicus pallidus Mearne Old Fort Verde, Ariz. 

zibethicus ripensis Bailey Carlsbad, N. Mex. 

zibethicus cinnamominus Hollister Wakeeney , Kans. 

rivalicius Bangs Burbridge, La. 

FOSSIL SPECIES. 

Fiber nebracensis nobis. . .Lower Pleistocene, Niobrara River, Sheridan County, Nebr. 

oregonus nobis Lower Pleistocene, Fossil Lake, Oreg. 

annectens Brown Middle Pleistocene, Newton County, Ark. 

EXISTING SPECIES. 

FIBER OBSCURUS Bangs. 
Newfoundland Muskrat. 
Fiber obscurus Bangs, Proc. Biol. Soc. Wash., IX, 133, September 15, 1894. 

Type locality. — Codroy, Newfoundland. 

Geographic distribution. — Newfoundland. 

General characters. — Size small; hind foot proportionally large; 
color very dark; skull small and weak; parietals large. 

Color. — Fresh pelage: Upperparts dark mummy brown, varying to 
almost black, darkened on back by brownish black overlying hairs; 
sides, chestnut ; underparts chestnut, brighter than in F. zibethicus; 
small spot on chin blackish brown; underfur slate gray; lips pale 
straw yellow or white; nasal pad and tail black; feet brown; nails 
yellow to brown. Worn pelage: Upperparts paler, without overlying 
black hairs ; sides lacking the brightness of the winter pelage. Young: 
Above uniform dusky; below paler. 

Skull and teeth. — Skull very small, weak, and smooth; interorbital 
constriction relatively broad; parietals large; nasals narrow; teeth 
small. 

Measurements. — Average of seven adults from Codroy and Bay 
St. George, Newfoundland: Total length, 500 ;* tail vertebrae, 226; 
hind foot, 76. 

Skull. — Average of 11 specimens from Codroy and Bay St. George, 
Newfoundland: Basal length, 53.2; zygomatic breadth, 34.7; palatal 
length, 34; length of nasals, 19.8; breadth of nasals, 7.4; alveolar 
length of upper molar series, 14.2. 

Type specimen. — No. 1155, Museum of Comparative Zoology 
(Bangs collection). Skin and skull, 9 . Collected by Ernest Doane, 
May 14, 1894. 

Remarks. — Though some winter specimens of Fiber zibethicus are 
as black as some skins of obscurus, the color of the Newfoundland 
animal averages darker at all seasons. It is apparently a well- 
established species. 

Specimens examined. — Total number 17, from localities as follows: 
Newfoundland: Balena, 1; Bay St. George, 12; Codroy, 3; Newfoundland, 1. 

1 All measurements are given in millimeters. 



16 NORTH AMERICAN FAUNA. [no. 32. 

FIBER ZIBETHICUS ZIBETHICUS (Linnaeus). 

Common Muskrat. 

Castor zibethicus Linnaeus, Syst. Nat., ed. 12, I, 79, 1766. 

Castor zibethus Severinus, Tent. Zool. Hung., p. 107, 1779. 

Mus zibethicus Gmelin, Syst. Nat., I, 125, 1788. 

Myocastor zibethicus Kerr, Anim. Kingd. syst. cat., bet. pp. 32 and 33, 1792. 

Mus Myocastor zibethicus Kerr, Anim. Kingd., p. 225, 1792. 

Ondatra zibethicus Link, Beytr. Nat., I, pt. 2, p. 76, 1795. 

Ondatra americana Tiedernann, Zoologie, 1, 481, 1808. 

[Fiber] zibethicus Illiger, Prodr. Syst. Mamm., p. 88, 1811. 

[ Mussascus] americana Oken, Lehrb. Nat., 3 ter Theil., 2 te Abth., p. 886, 1816. 

[Simotes] zibethicus Fischer, Mem. Soc. Imp. Nat. Moscou, V, 444, 1817. 

Mus (Fiber) zibeticus Cuvier, Regne Anim., I, 192, 1817. 

Fiber zibethicus, var B, nigra Richardson, Fauna Boreali-Amer., I, 119, 1829 (no 

definite locality). 
Fiber zibethicus, var C, maculosa Richardson, Fauna Boreali-Amer., 1, 119, 1829 (no 

definite locality). 
Lemmus zibethicus Fischer, Synops. Mamm., p. 289, 1829. 
Fiber zibethicus varius Fitzinger, Sitz.-Ber. Kais. Akad. Wiss., LVI, 103, 1867 (new 

name for F. z. maculosa Richardson). 
Fiber zibethicus niger Fitzinger, Sitz.-Ber. Kais. Akad. Wiss., LVI, 103, 1867 (new 

name for F. z. nigra Richardson). 

Type locality. — Eastern Canada; specimens from New Brunswick 
assumed to be typical. 

Geographic distribution. — Southeastern Canada, northeastern and 
east central United States; from New Brunswick and Quebec west 
to Minnesota, and south to northern Georgia and Arkansas, except 
along the Atlantic seaboard south of Delaware Bay. 

General characters. — Size large; tail long; color dark; skull large, 
with zygomata not broadly spreading anteriorly; molars of me- 
dium size. 

Color. — Fresh pelage: Upperparts mummy brown, darkest on head; 
back glossy; sides chestnut to hazel. The darker color on back is 
due to the blackish overlying hairs, the color of the fur being much 
like that of sides. Underparts like sides but paler, approaching tawny, 
shading to whitish on throat and belly; a small spot on chin and hair 
of wrist and heel blackish; lips straw j^ellow; underfur light slate 
gray; nasal pad and tail black; feet dark brown; nails pale straw to 
brown. Worn pelage: Paler and duller throughout; upperparts and 
sides uniform grayish brown, or with a faded reddish mixture; back 
and head with little or no black. Black phase: Upperparts uniformly 
black; cheeks and long hair at base of tail chestnut; underparts 
dark. Young: Back uniform dusky; sides and belly paler; cheeks 
rusty. 

Skull and teeth. — Skull large; zygomata not broadly spreading 
anteriorly; interorbital ridge not especially developed, except in 
extreme old age; parietals large; audital bullae rounded; molars of 
medium size. 



1911.] FIBER ZIBETHICUS ZIBETHICUS. 17 

Measurements. — Average of 7 adults from Lake George and Peter- 
boro, N. Y.: Total length, 563; tail vertebrae, 254; hind foot, 81. 
Skull. — Average of 10 adults from New Brunswick: Basal length, 
60.4; zygomatic breadth, 38.8; palatal length, 38; length of nasals, 
21.2; breadth of nasals, 9.1; alveolar length of upper molar series, 
16.3. 

Remarks. — This form, the common muskrat of the Northern and 
Middle States, is a dark-colored animal; much darker than F. z. 
macrodon in ordinary color phase, and only slightly lighter than 
obscurus from Newfoundland. Specimens from the coast region of 
Massachusetts and Rhode Island average especially black in full 
winter pelage. Specimens from Conanicut Island, Rhode Island, 
have long tails, but occasional specimens from some mainland locali- 
ties match them in this character, and the very slight insular variety 
perhaps developing here is hardly worthy of recognition by name. 
Prince Edward Island specimens show no approach toward obscurus 
and are apparently typical of zibethicus. Specimens from Middle 
and Southern States average less black than New England speci- 
mens, approaching some of the less pronounced examples of nor- 
mally colored macrodon, and have more red than most specimens from 
the Northeast. I have as yet, however, failed to find a single speci- 
men from any inland southern or western locality, east of the Great 
Plains, that can not be matched by some strictly comparable speci- 
men or specimens in the large series of true zibethicus from north- 
eastern United States. Specimens from the lower Hudson Valley 
and Long Island show a decided approach toward macrodon, and 
these two forms probably blend throughout New Jersey and Delaware. 
Specimens from upper Delaware Bay have been referred to macrodon, 
though the discrimination at this point is difficult, and the animals 
could be placed with either form without much violence. 

The black phase appears to be of rare occurrence in typical zibe- 
thicus. I have seen it only from Lake George, New York, and 
Conanicut Island, Rhode Island. Several albinos and partial 
albinos have been examined. 

Remains of Fiber zibethicus are recorded by Leidy x and Holmes 3 
from Pleistocene deposits of New Jersey, Pennsylvania, and South 
Carolina. 

Specimens examined. — Total number 468, from localities as follows: 

Prince Edward Island: Mount Stewart, 2. 
New Brunswick: "New Brunswick," 14. 
Quebec: Lake St. John, 1; Ottawa River, 1. 
Maine: Naskeag, 1. 

1 Synops. Extinct Manim. N. Amer., p. 407, 1869; Ann. Report Geol. Survey Penn- 
sylvania for 1887, pp. 5 and 19, 1889. 

2 Proc. Amer. Ass. Adv. Science, Third meeting, pp. 201-204, 1850. 

79580°— 11 3 



18 NOKTH AMERICAN FAUNA. [no. 32. 

New Hampshire: Charlestown, 2. 

Massachusetts: Belmont, 16; Middleboro, 2; Wilmington, 7; Woburn, 2. 

Connecticut: East Hartford, 2. 

Rhode Island: Conanicut Island, 25; Fort Adams, 1; Newport, 14. 

New York: Adirondacks, 2; Essex County, 68 ;* Fort Totten, 5; Hastings, 

1; Highland Falls, 5; Lake George, 14; Locust Grove, 1; Montauk Point, 

1; Peterboro, 3; Piseco, 2; Saranac Lake, 1; Schroon Lake, 2; Severance, 

22 j 1 Sing Sing, 3; Trousers Lake, 2; Troy, 1. 
Pennsylvania: Allegheny County, 3; Carlisle, 6; Conestoga Creek, 1; Sayre, 1. 
Michigan: Ann Arbor, 1; Isle Royale, 2; New Richmond, 1; Taquahmenaw 

River, 1. 
Wisconsin: Beaver Dam, 2; Conover, 1; Crab Lake, 1; Delavan, 14; Eagle 

River, 4; Green Bay, 2; Lake Koshkonong, 1. 
Minnesota: Elk River, 150 ; 2 Fort Snelling, 5. 
Iowa: Burlington, 1. 

Illinois: Chicago, 1; Fox Lake, 7; Libertyville, 1; Olive Branch, 7. 
Indiana: Effner, 3; Hebron, 2; La Porte, 1; Wheatland, 5. 
Kentucky: Eubanks, 1. 

Tennessee: Highcliff, 5; Roan Mountain Station, 1; Watauga Valley, 1. 
West Virginia: White Sulphur Springs, 4. 
North Carolina: Magnetic City, 1. 
South Carolina: Greenville, 1. 
Georgia: "Georgia," 1; Hogansville, 1. 
Alabama: Reform, 1. 
Arkansas: Mammoth Spring, 4. 

FIBER ZIBETHICUS MACRODON Merriam. 
Virginia Muskrat. 

Fiber macrodon Merriam, Proc. Biol. Soc. Wash., XI, 143, 1897. 
Fiber zibethicus macrodon Stone and Cram, Amer. Anim., 126, 1902. 

Type locality. — Lake Drummond, Dismal Swamp, Virginia. 

Geographic distribution. — Middle Atlantic coast region of the 
United States, from Delaware Bay to Pamlico Sound; inland to 
Washington, Virginia, and Raleigh, N. C. 

General characters. — Size largest in the genus; colors rich and 
bright; normal pelage with much red in fresh and worn state (there 
is a large proportion of specimens in the black phase from some 
localities). Skull large and massive; teeth large. 

Color. — Fresh pelage: Like zibethicus, but lighter and brighter, 
with less black. Upperparts Prout's brown, darker on nose, head, 
and back; sides varying from grayish brown to russet; underparts 
from broccoli brown or drab to bright cinnamon rufous. Specimens 
in perfectly fresh pelage have very little of the bright russet tinge 
which appears in specimens taken later in the winter and spring, the 
hairs of this color coming in as the season advances. Spring speci- 
mens, before the summer molt has commenced, are often especially 
bright and rich colored above and below. Worn pelage: Upperparts 
dull russet; underparts pale cinnamon rufous. Young: Like young 

i Skulls. ■ 122 odd skulls. • 



1911.] # FIBER ZIBETHICUS AQUILONIUS. 19 

of zibethieus. Black phase: Entire upperparts, sides, and middle area 
of imderparts uniform brownish black; flanks with a few hairs of 
rusty brown; cheeks rusty; throat and ventral region whitish, drab, 
or Isabella color; usual small spot on chin black. Between this 
phase and the normal color before described is every degree of vari- 
ance. Some specimens have the back black and the belly gray or 
rusty, some have the belly dark and upperparts nearly normal or 
partly melanistic, while a few are of an intermediate slate color. 

Skull and teeth. — Skull large, with heavy rostrum, posteriorly 
elongated brain case, and elevated frontal; jugal massive, high, and 
rounded above; molar teeth large. 

Measurements. — Average of four adults from the Dismal Swamp, 
Virginia: Total length, 620; tail vertebras, 274; hind foot, 88. 
Skull. — Average of four adults from Dismal Swamp: Basal length, 
65.1; zygomatic breadth, 41,. 7; palatal length, 41; length of nasals, 
22.9; breadth of nasals, 9.9; alveolar length of upper molar series, 
17.7. 

Type specimen. — No. 75940, United States National Museum, 
Biological Survey Collection. 9 adult, skin (black phase) and skull. 
Collected by Dr. A. K. Fisher, October 9, 1895. 

Remarks. — The most remarkable thing about this form is the large 
per cent from certain localities of specimens in the black phase. In 
some marshes on the eastern shore of Maryland over half of the 
muskrats are black. From no locality, however, from which more 
than two specimens are at hand, is the normal phase unknown. 
Specimens from interior localities in Maryland, though retaining the 
color of macrodon, have skulls closely approaching in character those 
of zibethieus from farther north. No specimens from New Jersey 
have been examined, but it seems probable that they will prove 
somewhat intermediate between the two subspecies. 

Specimens examined. — Total number 83, from localities as follows: 

Pennsylvania: Chester County, 2. 

Maryland: Branchville, 1; Broadwater, 1; Cambridge, 1; Forest Glen, 1; 

Jefferson, 1; Kensington, 1; Laurel, 10. 
District of Columbia: Washington, 11. 
Virginia: Arlington, 1; Dunn Loring, 4; Fredericksburg, 1; Lake Drummond, 

Dismal Swamp, 5; Pope Creek, 5; Quantico, 5; Suffolk, 6; Wallaceton, 

Dismal Swamp, 2; Warwick, 8; Washington, 6. 
North Carolina: Currituck, 2; Poplar Branch, 2; Raleigh, 7. 

FIBER ZIBETHICUS AQUILONIUS Bangs. 

Labrador Muskrat. 

Fiber zibethieus aquilonius Bangs, Proc. New England Zool. Club., I, 11, February 
28, 1899. 

Type locality. — Rigoulette, Hamilton Inlet, Labrador. 
Geographic distribution. — Labrador and Ungava. 



20 NORTH AMERICAN FAUNA. [no. 32. 

General characters. — Like zibethicus, but slightly smaller, with hind 
foot actually and relatively smaller, color averaging more blackish 
in summer. 

Color. — Fresh pelage: Almost exactly like Fiber z. zibethicus, but 
slightly brighter and richer colored, especially on sides and under- 
pays. Worn pelage: Averaging darker, with more black, than the 
corresponding pelage of F. z. zibethicus; underparts averaging more 
heavily colored, the overlying hairs of a darker shade. 

Skull and teeth. — Skull almost precisely like that of F. z. zibethicus, 
but averaging a little lighter, with slightly lighter rostrum and very 
slightly smaller teeth. 

Measurements. — Average of four specimens from Hamilton Inlet 
and Lance au Loup, Labrador: Total length, 551 ; tail vertebrae, 262 
hind foot, 74. Skulls. — Basal length, 60.3; zygomatic breadth, 40.1 
palatal length, 37.5; length of nasals, 22.3; breadth of nasals, 9.2 
alveolar length of upper molar series, 15.3. 

Type specimen. — No. 3957, Museum of Comparative Zoology 
(Bangs Collection) . Skin and skull, immature $. Collected by C. H. 
Goldthwaite, August 15, 1895. 

Remarks. — The series of specimens representing this form is far 
from satisfactory. In the available series of 16 skins and skulls there 
is only 1 good adult specimen, 1 and 3 or 4 more which may be called 
subadult. These specimens, together with the remaining series of 
young, indicate a slight form which seems to show enough average 
difference from zibethicus to merit recognition. It approaches the 
Hudson Bay form slightly but is much blacker and the skull is more 
like that of zibethicus. In no important character does it seem to 
resemble the very different F. obscurus from Newfoundland. A large 
series of adult skulls and more skins in full winter pelage are much 
needed, and until these are available the validity of the form can not 
be considered as satisfactorily established. 

Specimens examined. — Total number 16, from localities as follows: 
Labrador: Black Bay, 1; Hamilton Inlet, 1; Lance au Loup, 10. 
TJngava: Forks, 2; Fort Chimo, 2. 

FIBER ZIBETHICUS ALBUS Sabine. 
Hudson Bay Muskrat. 

Fiber zibethicus — albus Sabine, Zool. App. Franklin's Narr., p. 660, 1823. 
Fiber zibethicus, var. D, alba, Richardson, Fauna Boreali-Amer. , I, 119, 1829. 
Fiber zibethicus hudsonius Preble, North Amer. Fauna, No. 22, p. 53, October 31, 1902. 
Fort Churchill, Keewatin. 

Type locality. — Cumberland House, Saskatchewan. 

Geographic distribution. — Waters draining into Hudson Bay from 
the west, in eastern Saskatchewan and Keewatin; north to the 
Barren Grounds. 

1 Museum of Comparative Zoology (Bangs Collection), No. 8947. 



1911.] FIBER ZIBETHICUS ALBUS. 21 

General characters. — Like Fiber z. zibethicus, but smaller, with 
shorter tail and smaller hind foot; faded summer pelage with much 
more rusty; skull with zygomata as in zibethicus, interorbital ridge 
and teeth as in spatulatus. 

Color. — Fresh pelage: Upperparts like zibethicus but paler; back and 
sides uniform Vandyke brown to Prout's brown; muzzle darker; sides 
russet; underparts like sides but paler, shading to whitish on throat 
and belly; small spot on chin and hair of wrist and heel brown; lips 
pale straw yellow; feet grayish brown; tail blackish. Worn pelage: 
Like zibethicus, but with more red; sides lighter. Young: Like cor- 
responding age of zibethicus. 

STcull and teeth. — Skull averaging smaller than in Fiber z. zibethicus; 
interorbital crest higher; molars small. Differs from skull of spatu- 
latus in the shape of the zygomata, which are not so squarely spreading 
anteriorly; bullae more inflated; nasals narrower. 

Measurements. — Average of eight adults from Keewatin: Total 
length, 541; tail vertebras, 239; hind foot, 75. Skull. — Average of 
five adults from Fort Churchill and Echimamish River, Keewatin: 
Basal length, 58.8; zygomatic breadth, 38.1; palatal length, 37.1; 
length of nasals, 21; breadth of nasals, 9.1; alveolar length of upper 
molar series, 15.3. 

Type specimen. — The original specimen of albus, collected at Cum- 
berland House by a Mr. Holmes, was taken to Europe by Franklin's 
First Expedition, and was described by Sabine. Mr. Gerrit S. Miller, 
jr., has searched the records at the British Museum and informs me 
that no such specimen is catalogued in that institution. The "type" 
of tins form is probably not in existence. 

Remarks. — This is a slight form, but one apparently worthy of 
recognition. To a certain degree it combines the characters of true 
zibethicus with those of spatulatus, but it occupies a definite area of 
some extent and is not exactly an intermediate in the ordinary sense 
of the term. To the northwest the intergradation with spatulatus 
is very clear, and specimens from the Athabaska and the Mackenzie 
regions, though unquestionably referable to spatulatus, are certainly 
intermediate. Specimens from near The Pas, Saskatchewan River, 
virtually topotypes, have been recently received at the Biological 
Survey, through the efforts of Mr. E. A. Preble and Mr. R. MacFar- 
lane, and the question as to what form occurs at Sabine's type locality 
is thus definitely settled. Sabine's description, based upon an abnor- 
mal specimen, agrees with this form in everything except color, and 
his name unfortunately must be used. 

Specimens examined. — Total number 14, from localities as follows: 

Keewatin: Echimamish River, 5; Fort Churchill, 1; Hairy Lake, 1; Nelson 
River 1; Robinson Portage, 1; The Pas, 2; York Factory, 3. 



22 „. NORTH AMERICAN FAUNA. [no. 32. 

FIBER ZIBETHICUS SPATULATUS Osgood. 

Northwestern Muskrat. 

Fiber spatulatus Osgood, North Amer. Fauna, No. 19, p. 36, 1900. 

Fiber zibethicus spalulatus Preble, North Amer. Fauna, No. 27, p. 191, 1908. 

Type locality. — Liake Marsh, Yukon. 

Geographic distribution. — Northwestern North America, from the 
Kowak River and Yukon Valley, Alaska, east to the Anderson 
River and south into British Columbia and Alberta. 

General characters. — Size small; hind foot small; color dark, with 
minimum amount of rusty. Skull angular, zygomata broadly spread- 
ing anteriorly; interorbital crest high; parietals small; nasals broad; 
molars small. 

Color. — Fresh pelage: Upperparts glossy mummy brown; nose and 
hips blackish; sides russet ; underparts dull whitish with a cinnamon 
wash; lips, throat, and belly whitish; small spot on chin brown; feet 
grayish brown; tail black. Worn pelage: Varying from dingy yel- 
lowish brown to dark grayish brown; underparts with less coloring 
from overlying hairs than in the fresher coat. Young: Dull grayish 
brown, lighter than in corresponding age of zibethicus or albus; 
underparts grayish white. 

Skull and teeth. — Skull angular; zygomata broadly spreading 
anteriorly; interorbital ridge developed to a high, sharp crest, even 
in comparatively young individuals; bullse small and flattened 
diagonally; jugals small, not rising high above plane of the zygomatic 
arch; parietals small; nasals broad at curve and constricted poste- 
riorly; molars small. 

Measurements. — Average of four adults from northern British 
Columbia and Alberta: Total length, 530; tail vertebras, 232; hind 
foot, 74.5. Skull. — Average of five adults from type region: Basal 
length, 58.2; zygomatic breadth, 38.4; palatal length, 37; length of 
nasals, 20.4; breadth of nasals, 10; alveolar length of upper molar 
series, 14.7. 

Type specimen. — No. 98567, United States National Museum, 
Biological Survey Collection. Skin and skull, 9 . Collected by 
Wilfred H. Osgood, July 3, 1899. 

Remarks. — The northwestern muskrat is a well-defined subspecies 
occupying a large area, over the greater part of which it remains 
remarkably constant in characters. Specimens from the entire 
Yukon Valley and northwestern Alaska, Yukon, northern British 
Columbia, and Alberta are typical, but those from the northeastern 
parts of its range, from the mouth of the Mackenzie River along the 
northern tree limit, are obviously approaching the Hudson Bay form. 

Specimens from the north of Great Bear Lake and from the Lower 
Mackenzie Valley have slightly larger teeth and less broadly spreading 



1911.] FIBER ZIBETHICUS ZALOPHUS. 23 

zygomata. There is also a marked tendency toward the Hudson Bay 
form in the increase of the rusty, throughout the pelage, in all the 
most eastern specimens. Five specimens from Revillagigedo Island, 
Alaska, 1 while typical of spatulatus in size and color, show a slight 
approach toward osoyoosensis in the shape of the audital bullae and in 
the high, rounded jugals. 

Specimens examined. — Total number 67, from localities as follows: 

Alaska: Anklin River, Yakutat Bay, 2; Dangerous River, Yakutat Bay, 1; 
Eagle, 1; Fort Hamlin, 3; Kowak River, 150 miles from mouth, near Jade 
Mountains, 1; Nome, 1; Norton Bay, 1; Nulato, 1; Portage Cove, Revilla- 
gigedo Island, 5; Russian Mission, Yukon River, 2; St. Michaels, 3; 
Yukon, 4. 

Yukon: Lake Marsh, 2. 

Mackenzie: Fort Anderson, 1; Fort Franklin, 7; Fort Liard, 1; Fort McPher- 
son, 7; Fort Resolution, 3; Fort Simpson, 5; Great Slave Lake, 1; Willow 
River, 1. 

British. Columbia: Bennett, 1; Tagish Lake, 1. 

Alberta: Athabaska Lake, 1; Blindman River, 1; Edmonton, 2; Fort Chipe- 
wayan, 4; Henry House, 1; Slave River, 2; Swan Lake, 1. 

FIBER ZIBETHICUS ZALOPHUS Hollister. 

Alaska Peninsula Muskrat. 

Fiber zibethicus zalophus Hollister, Proc. Biol. Soc. Wash., XXIII, 1, February 2, 1910. 

Type locality. — Becharof Lake, Alaska. 

Geographic distribution. — Alaska Peninsula, north to Nushagak and 
east to the head of Cook Inlet. 

General characters. — Size small; tail short; hind foot very small. 
Skull with zygomata not broadly spreading anteriorly; parietals very 
small, even in young animals; interorbital ridge extreme in develop- 
ment into a blade-like crest; molars small. 

Color. — General tone of upperparts bister, darkest on back and 
hips, with little or no rusty coloring; sides like back, but with a 
slight russet tinge; brown spot on chin reduced to a mere streak. 
Underparts creamy white with a cinnamon wash, varying in inten- 
sity and shading to white on throat and hind legs; lips whitish; feet 
grayish brown, fringed with pale buffy hairs; tail black. Worn 
pelage: Upperparts russet to cinnamon, varying greatly in the speci- 
mens at hand, but usually with much more red than in fresh coat, or 
in any pelage of F. z. spatulatus. Young: Like young of spatulatus. 

Skull and teeth.— Skull with zygomata not broadly spreading 
anteriorly as in spatulatus; rostrum and nasals longer; parietals very* 
small, squamosal covering most of area of brain case, even in young 
animals; interorbital ridge extreme in development into a blade-like 
crest; teeth small. 

1 Collection Museum of Vertebrate Zoology, University oi California. 



24 NORTH AMERICAN FAUNA. [no. 32. 

Measurements. — Average of four adults from Lake Clark and Cook 
Inlet, Alaska: Total length, 533; tail vertebrae, 228; hind foot, 69.7. 
Skull. — Average of five adults from Becharof Lake, Tyonek, and 
Lake Clark, Alaska: Basal length, 60.1; zygomatic breadth, 38.8; 
palatal length, 38.4; length of nasals, 21.3; breadth of nasals, 9.7; 
alveolar length of upper molar series, 14.7. 

Type specimen. — No. 131488, United States National Museum, Bio- 
logical Survey Collection. Skin and skull, not sexed. Collected by 
A. G. Maddren, October, 1903. 

Remarks. — This is a very well-marked form, all the specimens agree- 
ing in the uniformity of the characters which separate it from its geo- 
graphical neighbor, F. z. spatulatus. It is completely cut off from 
spatulatus on the east, but on the north it probably intergrades some- 
where between Nushagak and the mouth of the Yukon River. In 
many ways it is the extreme type of the spatulatus style, but it has a 
longer skull with the zygomata shaped more like true zibethicus from 
the East. Compared in color with spatulatus it is grayer in fresh 
pelage and with more red in the worn or washed-out coat. 

Specimens examined. — Total number 21, from localities as follows: 

Alaska: Becharof Lake, 13; Fort Kenai, 2; Lake Clark, 3; Nushagak, 1; 
Tyonek, 1; Ugashik, 1. 

FIBER ZIBETHICUS OSOYOOSENSIS Lord. 

Rocky Mountain Muskrat. 

Fiber osoyooserms Lord, Proc. Zool. Soc. London, 1863, p. 97, 1863. 

Fiber zibethicus osoyoosensis Hollister, Proc. Biol. Soc. Wash., XXIII, 1, 1910. 

Type locality. — Lake Osoyoos, British Columbia. 

Geographic distribution. — Puget Sound region and Rocky Moun- 
tains, from southern British Columbia, Washington, Idaho, and west- 
ern Montana, south in the mountains to northern New Mexico. 

General characters. — Nearest to spatulatus but decidedly larger; col- 
ors darker than in spatulatus, mergens, or cinnamominus. Skull much 
like that of spatulatus but larger, with much larger teeth. 

Color. — Fresh pelage: Upperparts varying from uniform glossy 
mummy brown to black; slightly darker than in spatulatus; sides rus- 
set; underparts usually heavily colored with cinnamon and dark 
russet or brown hairs; throat and ventral region lighter; hips black. 
Worn pelage: Upperparts dull sooty brown, sides and underparts 
paler, usually with few rusty overlying hairs. Young: Seal brown, 
paler beneath. 

Skull and teeth. — Skull large, resembling that of spatulatus in the 
development of the high interorbital ridge and broad nasals, but 
much larger and relatively narrower; molars larger; interorbital con- 
striction great; bullse small, with lateral inflation; rostrum and nasals 
lone:. 



1911.] FIBEE ZIBETHICUS OSOYOOSENSIS. 25 

Measurements. — Average of ten adults from Oroville, Washington: 
Total length, 589; tail vertebrae, 271; hind foot, 83. Skull: Basal 
length, 64.1; zygomatic breadth, 40.5; palatal length, 41.3; length 
of nasals, 23.6; breadth of nasals, 10.1 ; alveolar length of upper molar 
series, 15.9. 

Type specimen. — No. 62.12.30.6, British Museum (Natural His- 
tory). Skin and skull, barely adult, not sexed. Collected by J. K. 
Lord, no date. As this is the only muskrat type not preserved in an 
American museum, the following notes, kindly made for me by Mr. 
Gerrit S. Miller, jr., at the British Museum, August, 1910, are of con- 
siderable importance and should be placed on permanent record: 

Skin in moderately good condition; recently made over; Bomewhat shrunken. 
Head and body, 270; tail, 220 (from skin in present condition); hind foot, 72 (by 
taxidermist); hind foot (now), 69; with claw, 76; ear (by taxidermist), 19. Color 
rather dark but not at all peculiar, belly of the ordinary style, not dusky. Skull 
barely adult; brain case smooth; ridges in temporal region low, not quite joined. 
Condylobasal length, 58 (estimated, condyles cut away); upper length (to front of 
nasals), 57.4; zygomatic breadth, 35; mastoid breadth, 25.4; interorbital breadth, 
6.2; nasals, 19.0; greatest breadth, both nasals together, 9.0; diastema, 20.6; man- 
dible, 40.2; maxillary tooth row (alveoli), 14.6; (crowns), 13.2; mandibular tooth row 
(alveoli), 15; (crowns), 13.6. 

Remarks. — This large, dark-furred muskrat of the spatulatus type 
is readily distinguishable from its neighboring subspecies. From 
spatulatus on the north, with which it clearly intergrades in British 
Columbia, it is distinguished by its larger size and darker colors, as 
well as by the great cranial differences. Specimens from Ashcroft, 
British Columbia, are slightly approaching spatulatus in skull char- 
acters. From its neighbor on the great Nevada desert, mergens, it 
differs especially in its darker color. At Twelvemile Creek, Oregon, 
and at points along the western range in the Rockies it shows a 
decided tendency toward mergens in color and skull. To the east it 
blends into cinnamominus, as shown wherever specimens are available 
from along the border of the Great Plains. I can find no satisfactory 
characters to separate the animals from the southern Rockies, in 
Colorado and New Mexico, from typical osoyoosensis. The speci- 
mens from southern localities average slightly smaller, with smaller 
hind foot, and the skulls have shorter rostra, but some specimens 
can be exactly matched in the series from northern localities, and the 
per cent of specimens exhibiting the slight differences does not seem 
large enough to warrant naming the form. Some skulls from Farm- 
ington, N. Mex., are almost exactly matched in the topotype series 
from Lake Osoyoos (Oroville, Wash.), and the skins are equally black 
above, with the black hips and other characteristics of osoyoosensis. 
Specimens from Rinconada and Costilla River, New Mexico, though 
clearly referable to this form, show a slight approach toward ripensis 
of the Pecos Valley. 



26 NORTH AMERICAN FAUNA. [no. 32. 

Specimens examined. — Total number 131, from localities as follows: 

British Columbia: Ashcroft, 2; Kettle River, 1; Port Moody, 5. 

Washington: Aberdeen, 1; Almota, 1; Chehalis County, 2; Easton, 1; Fort 
Steilacoom, 1; Lake Cushman, 29; Lake Washington, 2; Mabton, 2; Mar- 
shall, 2; Mount Vernon, 3; Oroville, 12; Rockland, 1; Seattle, 6; Touchet, 
7; Walla Walla, 1. 

Oregon: The Dalles, 1; Twelvemile Creek, 1. 

Idaho: Fort Sherman, 1; Lemhi, 1; Packer Meadow, 2; Sawtooth Lake, 5. 

Montana: Corvallis, 8; Florence, 1; High wood Mountains, 1; Summit, 1. 

Wyoming: Opal, 2; Pass Creek, 1; Rock Creek, 1; Valley, 1. 

Utah: Laketown, 2; Ogden, 2; Utah Lake, 1. 

Colorado: Cochotope Pass, 5; Coventry, 2; Crested Butte, 1; Hebron, 2; Hot 
Sulphur Springs, 3. 

New Mexico: Costillo River, 1; Farmington, 4; Rinconada, 2. 

FIBER ZIBETHICUS OCCIPITALIS Elliot. 
Oregon Coast Muskrat. 
Fiber occipitalis Elliot, Publ. Field Col. Mus., Zool. Ser., Ill, 162, April, 1903. 

Type locality. — Florence, Oreg. 

Geographic distribution. — Northern Willamette Valley -and coast of 
Oregon. 

General characters. — Size and general characters of F. z. osoyoosensis, 
but averaging slightly paler or more reddish; skull with extremely 
narrow interpterygoid space. 

Color. — Fresh pelage: Like F. z. osoyoosensis, but with less black; 
color above a uniform chestnut brown; head darker; nose and outer 
sides of legs blackish; sides like back but slightly brighter; under- 
pays heavily washed with bright cinnamon rufous. Worn pelage: 
Duller and darker, showing much more black. 

Skull and. teeth. — Skull like that of osoyoosensis but with less highly 
developed interorbital ridge; interpteiygoid space very narrow, the 
borders nearly parallel. 

Measurements. — Average of four specimens from the type locality: 
Total length, 589; tail vertebrae, 271; hind foot, 83.5. Skull of type: 
Basal length, 64.1; zygomatic breadth, 44; palatal length, 42.5; 
length of nasals, 24.5; breadth of nasals, 10; alveolar length of upper 
molar series, 16. 

Type specimen. — No. 9260, Field Museum of Natural History, 
Chicago. ? , old adult, skin and skull. Collected by Edmund Heller, 
December 16, 1901. 

Remarks. — This form is known from only three localities, and the 
material representing it is rather unsatisfactory. It seems to be a 
well-marked form, but with a limited range. The specimens from 
the upper Willamette Valley are mostly rather immature but clearly 
belong with this form rather than with osoyoosensis. Some of the 
specimens from south of Puget Sound, in Washington, show a slight 



1911.] FIBER ZIBETHICUS MERGENS. 27 

approach toward this form in a narrowing of the interpterygoid 
fossa; but otherwise distinctly belong with osoyoosensis, which 
remains typical south along the Sound to and below Seattle. Addi- 
tional specimens from the coast region of Washington and Oregon 
are greatly needed to work out thoroughly the interrelations of these 
forms. 

Specimens examined. — Total number 10, from the following 
localities: 

Oregon: Beaverton, 1; Florence, 4; Portland, 5. 

FIBER ZIBETHICUS MERGENS Hollister. 

Nevada Mtjskrat. 

Fiber zibethicus mergens Hollister, Proc. Biol. Soc. Wash., XXIII, 1, February 2 y 
1910. 

Type locality. — Fallon, Nev. 

Geographic distribution. — Northern part of the Great Bashi; south- 
eastern Oregon, northeastern California, Nevada, and western Utah. 

General characters. — Size large, colors pale. Differs from F. z. 
osoyoosensis in its much paler color, and from F. z. pallidus in its 
large size and darker colors. 

Color. — Fresh pelage: Above, grayish brown; head and dorsal area 
blackish; cheeks, shoulders, and sides rusty; underparts creamy 
white with central area pale cinnamon or russet; usual spot on chin 
blackish brown. Fall specimens, before the black hairs have come 
in, are sometimes quite rusty above. Worn pelage: Above, uniform 
pale yellowish brown; sides and underparts with little rusty. 

Skull and teeth. — Skull slightly smaller than that of osoyoosensis, 
with shorter rostrum, more broadly spreading zygomata, and heavier 
jugals; much larger than that of pallidus. 

Measurements. — Average of six adults from Fallon and Lovelocks, 
Nev.: Total length, 554; tail vertebra?, 253; hind foot, 80. SJcull. — 
Average of four adults from Fallon and Lovelocks, Nev.: Basal 
length, 62.1; zygomatic breadth, 40.2; palatal length, 39.9; length 
of nasals, 21.4; breadth of nasals, 9.9; alveolar length of upper molar 
series, 15.5. 

Type specimen. — No. 156880, United States National Museum, 
Biological Survey Collection. 9 adult, skin and skull. Collected 
by Stanley E. Piper, April 3, 1908. 

Remarks. — This is a pale desert form of the osoyoosensis type, occu- 
pying the northern part of the Great Basin. It grades directly into 
osoyoosensis on the north and east, but the material from southern 
Nevada and Utah is too scanty to give a good idea of its direct rela- 
tionship with pallidus. At the time of publishing the original descrip- 
tion of mergens, I stated that no intergradation with pallidus was 



28 NOETH AMERICAN FAUNA. [no. 32. 

known. Since that time two specimens have been received from St. 
George, Utah, which show almost intermediate characters. For the 
present these, on account of their large size, have been placed under 
mergens, though a larger series from the same locality and interme- 
diate points may change tins view. 

Specimens examined. — Total number 20, from localities as follows: 

Oregon: Shirk, 1. 

California: Eagle Lake, l; 1 Susanville, l. 2 

Nevada: Fallon, 6; Lovelocks, 4; Paradise, 1; Ruby Lake, 4. 

Utah: St. George, 2. 

FIBER ZIBETHICUS PALLIDUS Mearns. 
Arizona Muskrat. 
Fiber zibethicus pallidus Mearns, Bull. Amer. Mus. Nat. Hist., II, 280, 1890. 

Type locality. — Old Fort Verde (Camp Verde), Yavapai County, 
Ariz. 

Geographic distribution. — Colorado River valley (California, Lower 
California, and Arizona), east to the Rio Grande Valley in New Mexico. 

General characters. — Size small, color uniform rusty red, with no 
overlying black hairs; skull small. 

Color. — Fresh pelage: Back and sides bright rusty red, nearest to 
the cinnamon rufous of Ridgway, with no black in the overlying hairs; 
between eyes and nose Prout's brown; whiskers blackish brown. 
Underparts like back and sides, but paler; a small brown spot on 
chin; edges of lips yellow; nose pad black; tail brownish; feet pinkish 
gray, fringed with buffy hair; nails straw to chrome. Worn pelage: 
Similar, but paler and duller throughout; underparts pale drab with 
little of the rusty tint; throat and ventral region whitish; brown spot 
on chin nearly obsolete. Young: Like adult in summer but still 
paler; very different from the dark young of all the other forms. 

Skull and teeth. — Skull very small; except in size it resembles that 
of F. z. zibethicus in nearly every particular. Skulls of pallidus are 
easily separated from skulls of mergens by their small size, but it is 
difficult to distinguish them from skulls of ripensis, which are also 
small. Skulls of pallidus average slightly larger than those of 
ripensis, with heavier rostrum, wider nasals, and smaller, less inflated 
audital bullae. 

Measurements. — Average of five specimens from the Arizona and 
California banks of the Colorado River: Total length, 431; tail verte- 
bra?, 191 ; hind foot, 66.5. Doctor Mearns's type series of nine speci- 
mens from the Verde River average: Total length, 482; tail, 204. 3 
Skull. — Average of five specimens from the Colorado River: Basal 
length, 54.7; zygomatic breadth, 36.8; palatal length, 34.7; length of 

1 Foot only. 2 Skull only. 3 Bull. Amer. Mus. Nat. Hist., II, 282, 1890. 



1911.] FIBER ZIBETHICUS EIPENSIS. 29 

nasals, 19.2; breadth of nasals, 8.6; alveolar length of upper molar 
series, 14.5. 

Type specimens. — There are two cotypes, both in the American 
Museum of Natural History, New York City — No. 2346, $ , Septem- 
ber 17, 1885, and No. 2348, ? , August 28, 1886; collected by Dr. E. A. 
Mearns. 

Remarks. — This is one of the best of the described forms of the 
muskrat, and it is readily distinguishable at all seasons by its small 
size and peculiar coloration. Its range is restricted and it does not 
seem to be abundant in many localities. New Mexican specimens 
seem to be grading toward the Rocky Mountain form, but the scarcity 
of material from this region and from eastern Arizona is a great 
handicap in working out its relationships. 

Specimens examined.'- — Total number 14, from localities as follows: 

California: Colorado River, 15 miles southwest of Ehrenberg, Ariz., 4. 
Arizona: Camp Grant, 2; Fort Verde, 2; Robert's Ranch, opposite Needles, 

Cal., 1; Springerville, 1. 
New Mexico: Albuquerque, 1; Upper Tularosa River, 3. 

FIBER ZIBETHICUS RIPENSIS Bailey. 
Pecos Muskrat. 
Fiber zibethicus ripensis Bailey, Proc. Biol. Soc. Wash., XV, 119, June 2, 1902. 

Type locality. — Carlsbad (Eddy), N. Mex. 

Geographic distribution. — Pecos Valley, in Texas and New Mexico. 

General characters. — About the size of F. z. pallidus, with shorter 
tail; color much darker. Skull slightly smaller, with larger bullae. 

Color. — Fresh pelage: Upperparts Vandyke brown; a few black 
hairs on back; top of head blackish; muzzle whitish; sides russet, 
brightest on neck and cheeks; underparts much paler; throat and 
ventral regions whitish. Worn pelage: Similar, but much paler and 
with no gloss. Back and sides light brown; underparts dirty white, 
with a faint cinnamon wash over central area. Young: Blackish 
above, drab below; sides and cheeks with a faint rusty tinge. 

Skull and teeth. — Skull very small; differs from that of pallidus in 
the slightly lighter rostrum, narrower nasals, and larger, more 
inflated, bullae. 

Measurements. — Average of six specimens from the type locality: 
Total length, 463; tail vertebra?, 204; hind foot, 68. Skull: Basal 
length, 53.6; zygomatic breadth, 34.6; palatal length, 34.3; length 
of nasals, 18.3; breadth of nasals, 7.7; alveolar length of upper molar 
series, 14.9. 

Type specimen. — No. 109012, United States National Museum, 
Biological Survey Collection. $ adult, skin and skull. Collected 
by Vernon Bailey, July 25, 1901. 



30 NORTH AMERICAN FAUNA. [no. 32. 

Remarks. — This well-marked form, easily distinguished from all 
others, is confined to a small area in New Mexico and Texas, the 
immediate vicinity of the Pecos River, and neighboring streams and 
springs. Specimens from Santa Rosa, N. Mex., in the upper Pecos 
Valley, have more black in the pelage and are slightly larger than 
typical specimens. They apparently show a tendency toward the 
darker form of the Rocky Mountains. 

Specimens examined. — Total number 28, from localities as follows: 

New Mexico: Carlsbad, 6; Santa Rosa, 7. 
Texas: Fort Stockton, 15. 

FIBER ZIBETHICUS CINNAMOMINUS Hollister. 

Great Plains Muskrat. 

Fiber zibethicus cinnamominus Hollister, Proc. Biol. Soc. Wash., XXIII, p. 125, Sep- 
tember 2, 1910. 

Type locality. — Wakeeney, Trego County, Kan. . 

Geographic distribution. — Great central plains region of western 
United States and Canada; from Manitoba south to northern Texas; 
east to central Iowa and west to the Rocky Mountains. 

General characters. — Smaller than F. z. zibethicus or osoyoosensis; 
larger than ripensis. Coloration pale, with much red in both fresh 
and worn pelages. Skull smaller than in zibethicus, with smaller teeth. 

Color. — Fresh pelage: Upperparts and sides cinnamon brown, 
dorsal area only slightly darker, with few black hairs. Nose to fore- 
head and eyes seal brown. Cheeks and underparts creamy clay color, 
lighter on neck, throat, and inner sides of legs. A very small brown 
spot on chin. Feet drab; nails yellowish. Tail dark brown. Worn 
pelage: Varying from wood brown to russet, depending upon the 
state of wear or renewal. Specimens in the short-haired early fall 
pelage are the darkest, and those in extreme faded or washed out 
early summer coat are the lightest. 

Skull and teeth. — Skull smaller than in zibethicus or osoyoosensis 
with smaller teeth; larger than that of ripensis. Compared with 
skulls of zibethicus it has a proportionally shorter and heavier rostrum, 
accompanied by a shortening and widening of the nasals. 

Measurements. — Average of two specimens from the type locality: 
Total length, 496; tail vertebras, 240. Average of hind foot in 
twenty-one specimens from various localities, 73.5. Skull. — Average 
of live specimens from the type locality: Basal length, 56.3; zygo- 
matic breadth, 35.5; palatal length, 35.5; length of nasals, 19.5; 
breadth of nasals, 8.9; alveolar length of upper molar series, 15. 

Type specimen. — No. f$f£, Merriam Collection (in United States 
National Museum). $ adult, skin and skull. Collected by A. B. 
Baker, January 14, 1887. 



1911.] FIBER EIVALICIUS. 31 

Remarks. — This is the muskrafc of the prairie sloughs and streams 
of the interior Great Plains region. In color it most resembles 
Fiber z. pallidus of Arizona, and is thus very different from its nearest 
geographical neighbors. It apparently intergrades with albus on the 
north and with zibethicus and osoyoosensis on the eastern and western 
borders of the Great Plains. Specimens from Carberry, Manitoba, 
though apparently referable to cinnamominus, are somewhat inter- 
mediate between three forms, zibethicus, cinnamominus, and albus. 

Specimens examined. — Total number 84, from the following local- 
ities : 

Manitoba: Carberry, 4. 

North Dakota: Buford, 1; Oaks, 1. 

Montana: Glasgow, 1; Little Dry Creek, 3. 

Wyoming: Bear Creek, 1; Sun, 3. 

South Dakota: Custer, 1; Savoy, 4; Tigerville, 1. 

Nebraska: Beemer, 10; Johnstown, 27. ' 

Colorado: Boulder, 5; Ward, 1; Wray, 2. 

Kansas: Garnett, 1; Manhattan, 1; Wakeeney, 8. 

Iowa: Knoxville, 1. 

Oklahoma: Red Fork, 1. 

Texas: Canadian, 4; Lipscomb, 8. 

FIBER RIVALICIUS Bangs. 

Louisiana Muskrat. 

Fiber zibethicus rivalicius Bangs, Proc. Boston Soc. Nat. Hist., XXVI, 541, July 31, 
1895. 

Type locality. — Burbridge Plantation, near Belair, Plaquemines 
Parish, La. 

Geographic distribution. — Coast region of Louisiana, north to north- 
ern Calcasieu, Pointe Coupee, and Tangipahoa parishes. 

General characters. — Like Fiber zibethicus, but averaging slightly 
smaller; colors duller; underparts darker. 

Color. — Fresh pelage: Upperparts uniform dark brownish black, 
with little of the reddish tints of other forms; sides like back, but 
without the overlying hairs. Underparts pale drab to sepia or 
Prout's brown, the color depending on the number and shade of the 
longer dark-colored hairs; underfur light plumbeous. Worn pelage: 
Paler and duller throughout, usually showing more of the rusty tint 
than the fresh pelage. 

Slcull and teeth. — Skull ahnost precisely like that of Fiber z. zibethi- 
cus, but averaging slightly smaller. 

Measurements. — Average of 10 adults from Belair, La.: Total 
length, 547; tail vertebrae, 233; hind foot, 78. Skull: Basal length, 
60.1; zygomatic breadth, 38.7; palatal length, 38; length of nasals, 
20.3 ; breadth of nasals, 8.4 ; alveolar length of upper molar series, 15.7. 

1 Skulls. 



32 NORTH AMERICAN FAUNA. [no. 32. 

Type specimen. — No. 2719, Museum of Comparative Zoology 
(Bangs Collection). $ adult, skin and skull. Collected by F. L. 
Small, January 31, 1895. 

Remarks. — This form has a very limited range and is known only 
from the southern parishes of Louisiana. In a letter to Dr. C. Hart 
Merriam, Mr. Frank M. Miller, of New Orleans, writes that the musk- 
rat is confined strictly to the area south of a line running along the 
northern boundaries of Calcasieu, Acadia, Lafayette, St. Martin, 
Pointe Coupee, Baton Rouge, Livingston, Tangipahoa, and St. Tam- 
many parishes. The animal is thus completely isolated from other 
forms. The colony doubtless originated from stock from farther up 
the river. The most southern localities from which muskrats are 
known are included within the range of Fiber rivalicius, though they 
are only slightly farther south than the known southern limit of 
Fiber z. ripensis, at Del Rio, on the Rio Grande. This is the only 
form of which the specimens examined show anything like conclusive 
evidence of two annual molts. The Louisiana muskrat apparently 
molts in spring and fall, approximately in May and October. 

Specimens examined. — Total number 36, from localities as follows: 

Louisiana: Abbeville, 5; Belair, 20; Chef Menteur, 1; Gum Cove, 1; Houma, 
4; Iowa, 2; Octave Pass, 2; Slidell, 1. 

FOSSIL SPECIES. 

FIBER NEBRACENSIS sp. nov. 

Type from Lower Pleistocene ("Equus Beds"), quarry on Niobrara 
River, near Hay Springs, Sheridan County, Nebr. No. 2702, 
American Museum of Natural History, Department of Vertebrate 
Palaeontology. Skull, adult; parts of rostrum, palate, zygomata, 
brain case, and occiput; incisors and left row of cheek teeth in situ. 
Missing parts restored. American Museum Expedition of 1897. 

Characters. — Size medium, about that of Fiber z. cinnamominus, 
the existing form in same region. Teeth as in living species. Ros- 
trum short; audital bullae peculiar, small and much flattened, without 
angular surface on basioccipital side; interorbital crest high and 
sharp, as in the boreal forms; parietals and interparietal small and 
much flattened. Jaw not peculiar. 

Measurements. — Upper molar series, crowns, in three specimens; 
12.7 (type); 12.8; 12.7. Lower molar series, crowns, five specimens: 
12.7; 13.2; 12.5; 13.5; 12.7. First lower molar, crown, seven speci- 
mens: 6.3; 6.5; 6.4; 6.5; 6.4; 6.9; 7.0. Front of crown of ml to inferior 
notch of mandible, five specimens: 21.4; 22.2; 23.2; 23.3; 23.5. 

Remarks. — The skull of this fossil species, while greatly resembling 
that of some of the living forms of Fiber, possesses a combination of 
characters not found in any other known form. In size it agrees 



1911.1 



FIBER ANNECTENS. 



33 



with living forms, thus differing widely in that respect from the other 
described fossil species. 

Material examined. — Parts of three skulls and seven lower jaws, 
all with teeth in situ, from the type locality. 

FIBER OREGONUS sp. nov. 

Type from Lower Pleistocene ("Equus Beds"), Fossil Lake (Christ- 
mas Lake), 20 miles from Silver Lake, Lake County, Oregon. No. 
8594, American Museum of Natural History, Department of Verte- 
brate Palaeontology. Young adult, right mandibular ramus; coro- 
noid process, condyle, and angle broken; m3 missing; incisor broken 
at alveolus. Cope Collection. 

Characters. — Size very small, about the size of Fiber annectens 
Brown, from Arkansas in Mid-Pleistocene, and much smaller than 
any existing species of Fiber. Anterior loop of ml with deeper 
reentrant angles than in any living species; triangles not completely 
closed in any case. From the type specimen of F. annectens, a right 
mandibular ramus also, it differs in its much more robust build (the 
type of F. annectens is rather immature, however); teeth distinctly 
larger, especially broader across crowns; triangles less nearly closed. 
The inferior notch is rounded, while in F. annectens it is sharply 
square. The root of the lower incisor extends much farther back, 
passing beyond the inferior dental foramen. 

Comparative measurements of mandibles of fossil and recent muskrats. 



orcgonux. annectens. 



osoyno- 
sensis. 



ohscurus. 



From anterior eage of crown of ml to inferior notch 

Length of first two lower molars, crowns 

Length of m 1 , crown 



19.5 
9.0 
6.0 



19 

8.2 
5.7 



25.7 
11.3 

7.1 



23 
11 
7.5 



Remarks. — Except in size, the fragment representing F. oregonus 
is almost identical with the corresponding bone in living species. It 
differs from the jaw of Fiber annectens in only a few slight characters. 

Material examined. — One right mandibular ramus, the type. 

FIBER ANNECTENS Brown. 
Fiber annectens Brown, Mem. Amer. Mus. Nat. Hist., IX, pt. 4, p. 197, February, 1908. 

Type locality. — Conard Fissure, Newton County, Ark. Middle 
Pleistocene. 

Characters. — Size very small; smallest muskrat known. The 
species is based upon one right mandibular ramus and an upper 
molar, which indicate a species close to existing forms in everything 
except size. The sharply squared inferior notch is the most striking 



34 



NORTH AMERICAN FAUNA. 



[no. 32. 



difference; in all other species this notch is well rounded. The 
anterior loop of first lower molar is more deeply cut by the reentrant 
angles than in any specimen of an existing species examined. In the 
original description the statement is made that this loop is not cut 
by reentrant angles, the deep cuts apparently leading the describer to 
believe that the loop included only that part of the pattern anterior 
to these reentrant angles. The reduction of the anteroexternal 
column of m3, a character used in his argument of intermediate 
relationship between Fiber and Neqfiber, is of little consequence; 
this column is indifferently present or absent in Fiber, this particular 
tooth showing more variation in pattern than any other. 

Measurements. — Length of jaw, condyle to anterior surface of 
symphysis, 30; from anterior edge of crown of ml to inferior notch, 
19; alveolar length of lower molar series, 12; length of first two lower 
molars, crowns, 8.2; length of first lower molar, crown, 5.7. 

Type specimen. — No. 12424, American Museum of Natural History, 
Department of Vertebrate Palaeontology. Right mandibular ramus, 
somewhat immature. Nearly complete; coronoid process broken; 
incisor broken at alveolus. Collected by Barnum Brown, 1903. 

Remarks. — This is the smallest muskrat known, living or fossil. 
In size the jaw almost matches that of Neqfiber alleni. It appears to 
be a true Fiber, in no important character connecting Fiber with 
Neqfiber. The molars are all rooted. 

Material examined. — One right mandibular ramus (the type) and 
one right upper first molar, from the same deposit. 

TABLE OF AVERAGE CRANIAL MEASUREMENTS OF MUSKRATS. 



Form. 



Locality. 





fl 


















& 


03 
O 
Eh 

,a 

a 


a 


■3 
3 
a 


03 
3 

a 


a 


s 


"3 





■5 


























03 


5 


c 


ffl 


N 


fc 


i-l 


pq 


53.2 


34.7 


34 


19.8 


7.4 


60.4 


38.8 


38 


21.2 


9.1 


65.1 


41.7 


41 


22.9 


9.9 


60.3 


40.1 


37.5 


22.3 


9.2 


58.8 


38.1 


37.1 


21 


9.1 


58.2 


38.4 


37 


20.4 


10 


60.1 


38.8 


38.4 


21.3 


9.7 


64.1 


40.5 


41.3 


23.6 


10.1 


64.1 


44 


42.5 


24.5 


10 


62.1 


40.2 


39.9 


21.4 


9.9 


54.7 


36.8 


34.7 


19.2 


8.6 


53.6 


34.6 


34.3 


18.3 


7.7 


56.3 


35.5 


35.5 


19.5 


8.9 


60.1 


38.7 


38 


20.3 


8.4 



O tn 



O hi 

> P< 

■< s 



F. obscurus . 



F. zibethicus 

F. z. macrodon.. 
F. z. aquilonius . 

F. z. albus 

F. z. spatulatus. 



F. z. zalophus. 



z. osoyoosensis . 

z. occipitalis 

z. mergens 

z. pallidus 



F. z. ripensis 

F. z. cinnamominus. 
F. rivalicius 



Newfoundland: Codroy and Bay St. 

George. 

New Brunswick 

Virginia: Dismal Swamp 

Labrador: Hamilton Inlet and Lance 

au Loup. 
Keewatin: Fort Churchill and Echi- 

mamish River. 
Alaska: Eagle; Yukon: Lake Marsh; 

British Columbia: Bennett and 

Tagisb Lake. 
Alaska: Becharof Lake, Tyoonok 

and Lake Clark. 

Washington: Oroville 

Oregon: Florence 

Nevada: Fallon and Lovelocks 

Colorado River, in Arizona and Cali- 
fornia. 

New Mexico: Carlsbad 

Kansas: Trego County 

Louisiana: Belair . . . ." 



14.2 

16.3 
17.7 
15.3 

15.3 

14.7 

14.7 

15.9 
16 
15.5 
14.5 

14.9 

15 

15.7 



1911.] BIBLIOGRAPHY. 35 

TABLE OF AVERAGE FLESH MEASUREMENTS OF MTJSKRATS. 



Form. 



F. obscurus 

F. zibethicus 

F. z. macrodon 

F. z. aquilonius 

F. z. albus 

F. z. spatulatus 

F. z. zalophus 

F. z. osoyoosensis 

F. z. occipitalis 

F. z. mergens 

F. z. pallid us 

F. z. ripensis 

F. z. cinnamominus. . 
F. rivalicius 



Num- 
ber 
aver- 



Locality. 



Newfoundland: Codroyand Bay St. George.. 

New York: Lake George and Peterboro 

Virginia: Dismal Swamp 

Labrador: Hamilton Inlet and Lance au Loup 

Keewatin 

Northern British Columbia and Alberta 

Alaska: Lake Clark and Cook Inlet 

Washington: Oroville 

Oregon: Florence 

Nevada: Fallon and Lovelocks 

Colorado River, in Arizona and California 

New Mexico: Carlsbad 

Kansas and South Dakota ' 

Louisiana: Belair 



Total 
length. 



500 
563 
620 
551 
541 
530 
533 
589 
589 
554 
431 
463 
497 
547 



Tail. 



226 
254 
274 
262 
239 
232 
228 
271 
271 
253 
191 
204 
230 
233 



Hind 
foot. 



76.0 
81.0 
88.0 
74.0 
75.0 
74.5 
69.7 
83.0 
83.5 
80.0 
66.5 
68.0 
71.3 
78.0 



1 Kansas specimens without flesh measurements of foot. Foot average from 4 specimens from South 
Dakota. 

BIBLIOGRAPHY. 

The following list of titles includes chiefly books and articles which 
contain new material, relating either to the history of the nomen- 
clature and classification of the muskrat, or to its habits and distribu- 
tion. No attempt has been made to make it complete, as the litera- 
ture on the subject is exceedingly voluminous and much of it is buried 
in old files of sporting and trapping magazines. The titles of papers 
containing notes on habits have been carefully selected, to include 
only those containing fairly complete summaries of the subject, 
gleaned from many sources and there brought together, or those recent 
papers containing original matter of special importance. 

1612. Smith, Captaine [John]. A Map of Virginia. With a Description of the 

Countrey, the Commodities, People, Government, and Religion, p. 14. 

This is apparently the first reference to the muskrat in literature. The 

name mussascus is used. 
1758. Linnaeus, C. Systema Naturae, ed. 10, I, 59. The American muskrat is here 

confounded with the Castor moschatus (= Desmana moschata) of Asia. 
1766. Linnaeus, C. Systema Naturae, ed. 12, I, 79. American muskrat recognized, 

under the name of Castor zibethicus, as a distinct species. 
1779. Severinus, Joann. Tentamen Zoologise Hungaricae, p. 107. Name emended 

to Castor zibethus. 
1788. Gmelin, Jo. Frid. Systema Naturae, ed. 13, I, 125. Muskrat transferred to 

genus Mus, and placed, with the Mus coy pus of Molina, in a starred section, 

" cauda apice compressa. " 
1792. Kerr, Robert. The Animal Kingdom, p. 225. Myocastor proposed as a sub- 
genus of Mus, to include the two species of Gmelin 's first starred section, M. 

coy pus and M. zibethicus. 
1795. Link, H. F. Beytrage zur Naturgeschichte, I, pt. 2, p. 76. Ondatra proposed 

as a genus to include the same species as Gmelin 's first starred section of Mus 

and Kerr's Myocastor, O. coypus, and 0. zibethicus. 
1798. Cuvier, G. Tableau 61ementaire de l'Histoire Naturelle des Animaux, p. 141. 

Description of l'ondatra (Mus zibethicus). 



36 NORTH AMERICAN FAUNA. I no. 32. 

1799. Lacepede, B. G. E. Tableau des Divisions, Sous-divisions, Ordres et Genres 

des Mammiferes, p. 9. Lists the genus Ondatra, with 0. zibethicus as the typical 
specific example . Presumably considered the coypu and muskrat congeneric , 
as did the previous authors. 

1800. Cuvier, G. Lecons d'Anatomie Comparee, Tome I, tabl. I, April. Proposes 

the generic name Fiber, based upon l'ondatra of the 1798 Tableau. 
1808. Tiedemann, D. Friedrich. Zoologie. Zu eeinen Vorlesungen entworfen. 

I, pp. 480-481. Ondatra americana, new name for Mus zibethicus. 
1811. Illiger, C. Prodromus Systematis Mammalium et Avium, p. 88. Mus 

zibethicus Linn. Gmel. first assigned to genus Fiber. 

1816. Oken, [Lorenz]. Lehrbuch der Naturgeschichte, Dritter Theil, Zoologie, 
Zweite Abth., Fleischthiere, pp. 386-387. Mussascus proposed as a generic 
name with Ondatra americana as only species. 

1817. Cuvier, C. Le Regne Animal, I, 192. Castor zibeticus (sic) placed in sub- 

genus Fiber of Mus. 

1817. Fischer, Gotthelp. De systemate Mammalium. <Mem. Soc. Imp. Nat. 
Moscou., V, 444. Simotes proposed as a generic name for Mus zibethicus. 

1823. Sabine, Joseph. Zoological Appendix [to Franklin's Narrative of a Journey 
to the Polar Sea.], pp. 659-660. Account of muskrats, with original descrip- 
tion of Fiber zibethicus albus. 

1827. Billberg, G. J. Synopsis Faunae Scandinavise, Tom. I, Mammalia, Con- 
spectus A. Moschomys proposed as substitute for Ondatra Lacepede. 

1829. Fischer, Joanne Baptista. Synopsis Mammalium, p. 289. Muskrat in- 
cluded in genus Lemmus as L. zibethicus. 

1829. Richardson, John. Fauna Boreali- Americana; or the Zoology of the North- 
ern Parts of British America, pp. 115-119. General account; three varieties 
recognized by name, nigra, maculosa, and alba. 

1831. Godman, John D. American Natural History, pt. 1, v. II, pp. 57-62. De- 
scription, habits, and distribution. 

1846. Audubon, John James, and Bachman, John. The Viviparous Quadrupeds 
of North America, I, 108-124. General account o£ habits. 

1857. Baird, Spencer F. Mammals of North America, pp. 560-564. General 
account; description of external characters. 

1863. Lord, J. K. Notes on Two New Species of Mammals. <Proc. Zool. Soc. 
London, March 24, pp. 95-98. Original description of Fiber osoyoosensis. 

1884. Merriam, C. Hart. The Mammals of the Adirondack Region, Northeastern 
New York. Reprinted from v. 1 and 2, Trans. Linn. Soc, New York, pp. 
275-289, September. Life history. 

1890. Mearns, Edgar A. Description of Supposed New Species and Subspecies of 

Mammals from Arizona. <Bull. Amer. Mus. Nat. Hist., II, No. 4 [author's 
edition issued February 21]. Original description of Fiber zibethicus pallidus. 

1891. Flower, William Henry, and Lydekker, Richard. An Introduction to 

the Study of Mammals, Living and Extinct, pp. 470-472. General account. 

1892. Herrick, C. L. The Mammals of Minnesota. <Bull. Geol. and Nat. Hist. 

Surv. Minnesota, No. 7, pp. 211-217. Habits of muskrats; description of 
houses. 

1894. Bangs, Outram. Description of a New Musk Rat from Codroy, Newfound- 

land. <Proc. Biol. Soc. Washington, IX, 133, September 15. Original 
description of Fiber obscurus. 

1895. Bangs, Outram. Notes on North American Mammals. <Proc. Boston Soc. 

Nat. Hist., XXVI, 541-542 [author's edition July 31]. Original description 
of Fiber z. rivalicius. 



1911.] BIBLIOGRAPHY. 37 

1897. Merriam, C. Hart. Description of a New Muskrat from the Great Dismal 
Swamp, Virginia. <Proc. Biol. Soc. Wash., XI, 143, May 13. Original 
description of Fiber macrodon. 

1899. Bangs, Outram. Notes on some Mammals from Black Bay, Labrador. <Proc. 

New England Zool. Club., I, 11-12, February 28. Original description of 
Fiber z. aquilonius. 

1900. Osgood, Wilfred H. Results of a Biological Reconnoissance of the Yukon 

River Region. North Amer. Fauna, No. 19, October 6, pp. 36-37. Original 
description of Fiber spatulatus. 

1901. Elliot, Daniel Giraud. A Synopsis of the Mammals of North America and 

the Adjacent Seas. <Field Col. Mus. Pub., Zool. ser., II, 211-214, fig. 
49. General account of muskrats; seven forms recognized. 

1901. Miller, GerritS., Jr., and Rehn, James A. G. Systematic Results of the 

Study of North American Land Mammals to the Close of the Year 1900, pp. 
130-131. Eight forms of muskrats listed. 

1902. Bailey, Vernon. Seven New Mammals from Western Texas. <Proc. 

Biol. Soc. Wash., XV, 117-120, June 2. Original description of Fiber z. 
ripensis. 
1902. Preble, Edward A. A Biological Investigation of the Hudson Bay Region. 
North Amer. Fauna, No. 22, pp. 53-54, October 31. Original description 
of Fiber z.hudsonius. 

1902. Stone, Witmer, and Cram, William Everett. American Animals, pp. 

121-127. Muskrats of Eastern North America. Fiber macrodon reduced to 
a subspecies of F. zibethicus. 

1903. Elliot, D. G. Descriptions of Apparently New Species and Subspecies of 

Mammals from California, Oregon, the Kenai Peninsula, Alaska, and Lower 
California, Mexico. ^Field Col. Mus., Zool. ser., Ill, pp. 162-164. Original 
description of F. occipitalis. 

1903. Rhoads, Samuel N. The Mammals of Pennsylvania and New Jersey, pp. 

104-106. Habits and distribution in these States. 

1904. Elliot, Daniel Giraud. The Land and Sea Mammals of Middle America 

and the West Indies. <Field Col. Mus. Pub., Zool. ser., IV, pt. 1, pp. 
306-308, fig. 55. Fiber z. pallidus as a Mexican species. 

1905. Bailey, Vernon. Report on the Mammals of Texas. <North Amer. Fauna, 

No. 25, pp. 121-122, October. Habits and distribution of Fiber z. ripensis. 

1905. Mac Farlane, R. Notes on Mammals Collected and Observed in the Northern 

Mackenzie River District, Northwest Territories of Canada, with Remarks 
on Explorers and Explorations of the Far North. <Proc. U. S. Nat. Mus., 
XXVIII, 737-738. Notes on abundance and breeding habits of muskrats. 

1906. Ingersoll, Ernest. The Life of Animals. The Mammals, pp. 436-438. 

Habits of muskrats. 

1906. Stephens, Frank. California Mammals, pp. 132-133. Habits and distribu- 

tion of Fiber z. pallidus. 

1907. Mearns, Edgar Alexander. Mammals of the Mexican Boundary of the 

United States. Bull. U. S. Nat. Mus. No. 56, pt. 1, pp. 494-498. Accounts 
of F. z. pallidus and F. z. ripensis, with notes on habits and distribution. 

1908. Brown, Barnum. The Conard Fissure, a Pleistocene Bone Deposit in North- 

ern Arkansas: with Descriptions of Two New Genera and Twenty New 
Species of Mammals. <Mem. Amer. Mus. Nat. Hist., IX, pt. 4, p. 197, 
February. Original description of Fiber annectens. 
1908. Preble, Edward A. A Biological Investigation of the Athabaska-Mackenzie 
Region. North Amer. Fauna, No. 27, October 26, pp. 191-193. Account of 
muskrats in Northern British America. F. spatulatus reduced to a subspecies 
of F. zibethicus. 



38 NORTH AMERICAN FAUNA. [xo. 32. 

1909. Howell, Arthur IT. Notes on the Distribution of Certain Mammals in the 
Southeastern United States. <Proc. Biol. Soc. Washington, XXII, pp. 62-63, 
April 17. Limits of distribution of muskrats in the Southern States. 

1909. Seton, Ernest Thompson. Life Histories of Northern Animals. An account 

of the mammals of Manitoba. I, Grass eaters, pp. 538-557. General account 
of muskrats with map of ranges of the forms. 

1910. Hollister, N. Descriptions of Two New Muskrats. <Proc. Biol. Soc. Wash- 

ington, XXIII, 1-2, February 2. Original descriptions of F. z. mergens 

and F. z. zalophus. F. osoyoosensis reduced to subspecific rank. 
1910. Lantz, David E. The Muskrat. U. S. Dept. Agric, Farmers' Bulletin 

396, pp. 1-38, April 30. General account of habits and economic relations. 
1910. Hollister, N. A New Muskrat from the Great Plains. <Proc. Biol. Soc. 

Washington, XXIII, 125-126, September 2. Original description of F. z. 

cinnamominus . 

1910. Warren, Edward Royal. The Mammals of Colorado, pp. 105-107. Original 

notes on food and breeding habits. 

1911. Evermann, Barton W., and Clark, Howard W. Notes on the Mammals 

of the Lake Maxinkuckee Region. <Proc. Washington Acad. Sci., XIII, 
19-24, February 15. Habits of the muskrat in Indiana. 
1911. Hollister, N. The Generic Name of the Muskrat. <Proc. Biol. Soc. Wash- 
ington, XXIV. Fixes type of Ondatra Link. Fiber considered the proper 
generic name for the muskrat. 



PLATE III. 

SKULLS (DORSAL VIEWS, NATURAL SIZE). 

Fig. 1. Fiber zibethicus (Linn.). o ad. Peterboro, New York. (No. 111082, U. S. 
Nat. Mus.) 

2. F. z. macrodon Merriam. $ ad. Wallaceton, Virginia. (No. 77811, U. S. 

Nat. Mus., Biol Surv. Coll.) 

3. F. z. spatulatus Osgood. $ ad. Tagish Lake, British Columbia. (No. 

108335, U. S. Nat. Mus., Biol. Surv. Coll.) 

4. F.z.zalophusB.o\\i8teT. Ad. Becharof Lake, Alaska. (No. 131483, U. S.Nat. 

Mus., Biol. Surv. Coll.) 

5. Fiber obscurus Bangs. Ad. Bay St. George, Newfoundland. (No. 76370, 

U. S. Nat. Mus., Biol. Surv. Coll.) 

6. Fiber z. albus Sabine. $ ad. Echimamish River, Keewatin. (No. 106874, 

U. S. Nat. Mus., Biol. Surv. Coll.) 

7. Fiber z. ripensis Bailey. $ ad. Carlsbad, New Mexico. (No. 109010, U. S. 

Nat. Mus., Biol. Surv. Coll.) 
40 



North American Fauna No 32, U. S. Dept. Agr. Biological Survey. 



Plate III. 




Skulls of Fiber. 

1. F. zibethicus. 4. F. z. zalophus. 

2. F. z. mac-rodon. 5. F. obscurus. 

3. F. z. spatulatus. ij. F. z. albus. 

7. F. z. ripensis. 



PLATE IV. 

SKULLS (VENTRAL VIEWS, NATURAL SIZE). 

Fig. 1. Fiber zibethicus (Linn.) 9 ad. Peterboro, New York. (No. 111082, U. S. 
Nat. Mus.) 

2. F. z. macrodon Merriam. $ ad. Wallaceton, Virginia. (No. 77811, U. S. 

Nat. Mus., Biol. Surv. Coll.) 

3. F. z. spatulatus Osgood. $ ad. Tagish Lake, British Columbia. (No. 

108335, U. S. Nat. Mus., Biol. Surv. Coll.) 

4. F. z. zalophus Hollister. Ad. Becharof Lake, Alaska. (No. 131483, U. S. 

Nat. Mus., Biol. Surv. Coll.) 

5. F. obscurus Bangs. Ad. Bay St. George, Newfoundland. (No. 76370, U. S. 

Nat. Mus., Biol. Surv. Coll.) 

6. Fiber z. albus Sabine. $ ad. Eehimamish River, Keewatin. (No. 106874, 

U. S. Nat. Mus., Biol. Surv. Coll.) 

7. F. z. ripensis Bailey. $ ad. Carlsbad, New Mexico. (No. 109010, U.S. 

Nat. Mus., Biol. Surv. Coll.) 
42 



North American Fauna No. 32, U. S. Dept. Agr. Biological Survey. 



Plate IV. 




Skulls of Fiber. 

1. F. zibethicus. 4. F. z. zalophus 

2. F. z. rnaerodon. 5. F. obscurus. 

3. F. z. spatulatus. 6. F. z. albus. 

7. F. z. ripensis. 



PLATE V. 

Fig. 1. Skull of F. z. osoyoosensis Lord. $ ad. Oroville, Washington. (No. 90505, 
U. S. Nat. Mus., Biol. Surv. Coll.) Natural size. 

2. Enamel pattern of upper molar teeth of Fiber. {Fiber z. osoyoosensis Lord. 

No. 99941, U. S. Nat. Mus. Lake Cushman, Washington.) X 3. 

3. Enamel pattern of lower molar teeth of Fiber. {Fiber z. osoyoosensis Lord. 

No. 99941, U. S. Nat. Mus. Lake Cushman, Washington.) X 3. 

4. Hind feet of Fiber. {Fiber z. macrodon Merriam. From fresh specimen col- 

lected at Cambridge, Maryland, March 16, 1909.) Natural size. 
44 



North American Fauna No. 32, U. S. Dept. Agr. Biological Survey 



Plate V. 




Skull, Tooth Enamel Pattern, and Feet of Fiber. 
1, 2, :;. F. /.. osqyoosenis. 4. P. z. macrodon. 



PLATE VI. 

Fig. 1. Fiber z. macrodon Merriam. From fresh specimen collected at Cambridge, 
Maryland, March 16, 1909. Reduced. 

2. Fiber z. cinnamominus Hollister. Right mandibular ramus of type. No. 

3724, Merriam Coll. Trego County, Kansas. Natural size. 

3. Type of Fiber annectens Brown. Middle Pleistocene, Newton County, Arkan- 

sas. No. 12424, Am. Mus. Nat. Hist., Dept. Vert. Palaeontology. Natural 
size. 

4. Type of Fiber oregonus Hollister. Lower Pleistocene, Fossil Lake, Oregon. 

No. 8594, Am. Mus. Nat. Hist., Dept. of Vert. Palaeontology. Natural size. 

5. Type of Fiber nebracensis Hollister. Lower Pleistocene, Niobrara River, 

Sheridan County, Nebraska. No. 2702, Am. Mus. Nat. Hist., Dept. of 
Vert. Palaeontology. Most of right side restored. Natural size. 
46 



North American Fauna No. 32, U. S Dept. Agr. Biological Survey. 



Plate VI. 




1. F. z. maerodon. 3. F. annectens. Type. 

•1. F. z. cinnamominus. Type. 4. F. oregonus. Type. 

5. F. nebracensis. Type. 



INDEX. 



[New names in bold-faced type; synonyms in italics.) 



Acknowledgments, 13. 
Bibliography, 35-38. 
Castor zibethicus, 16 
zibethus, 16. 
Characters, 11-12. 
Cranial measurements, 34. 
Distribution, 9-10. 
Economic relations, 10-11. 
Fiber, genus, 14. 
Fiber albus, 20-21. 

annectens, 33-34. 

aquilonius, 19-20. 

cinnamominus, 30-31. 

hudsonius, 20. 

macrodon, 18-19. 

maculosa, 16. 

mergens, 27-28. 

nebracensls, 32-33. 

niger, 16. 

nigra, 16. 

obscurus, 15. 

occipitalis, 26-27. 

oregonus, 33. 

osoyoosensis, 24-26. 

pallidus, 28-29. 

ripensis, 29-30. 

rivalicius, 31-32. 

spatulatus, 22-23. 

varius, 16. 

zalophus, 23-24. 

zibethicus, 16-18. 
Flesh measurements, 35. 
Fossil muskrats, 13, 32. 
Habits, 9-10. 
History, 7-8. 
Key to species and subspecies, 12. 



Lemmus zibethicus, 16. 

List of species and subspecies, 14-1.5. 

Localities, type, 14-15. 

Material and acknowledgments, 13. 

Measurements, cranial, 34. 

flesh, 35. 
Moschomys, 14. 
Mus zibethicus, 16. 
zibeticus, 16. 
Muskrat, Alaska Peninsula, 23. 

Arizona, 28. 

Common, 16. 

Great Plains, 30. 

Hudson Bay, 20. 

Labrador, 19. 

Louisiana, 31. 

Nevada, 27. 

Newfoundland, 15. 

Northwestern, 22. 

Oregon Coast, 26. 

Pecos, 29. 

Rocky Mountain, 24. 

Virginia, 18. 
Mussascus, 14. 
Mussascus americana, 16. 
Myocastor zibethicus, 16. 
Neofiber alleni, 34. 
Ondatra, 14. 
Ondatra americana, 16. 
zibethicus, 16. 
Pelage, 11-12. 
Simotes, 14. 
Simotes zibethicus, 16. 
Species and subspecies, key to, 12. 

list of, 14-15. 
Type localities, 14-15. 



47 



O 



U. S. DEPARTMENT OF AGRICULTURE 

BUREAU OF BIOLOGICAL SURVEY 

HENRY W. HENSHAW, Chief 



NORTH AMERICAN FAUNA 

JSTo. 33 

[Actual date of publication, August 17, 1911] 




A BIOLOGICAL SURVEY OF COLORADO 



BY 

MERRITT CARY 

ASSISTANT BIOLOGIST, BIOLOGICAL SURVEY 




WASHINGTON 

GOVERNMENT PRINTING OFFICE 

1911 



LETTER OF TRANSMITTAL. 



U. S. Department of Agriculture, 

Bureau of Biological Survey, 
Washington, D. C, February 21, 1911. 
Sir: I have the honor to transmit herewith for publication as 
North American Fauna No. 33 a report on the results of a biological 
survey of Colorado, by Merritt Gary. The report consists of three 
sections. The first characterizes the five life zones which traverse 
the State, defines their extent and limits, and discusses their agri- 
cultural and economic possibilities. The second consists of a com- 
plete list of the mammals of Colorado with brief notes on their habits, 
distribution, and economic relations. The third is a list of the 
principal trees and shrubs of Colorado observed by the assistants of 
the Biological Survey during the progress of work in the State, with 
annotations as to their distribution and abundance. 
Respectfully, 

Henry W. Henshaw, 
Chief, Biological Survey. 
Hon. James Wilson, 

Secretary of Agriculture. 



CONTENTS, 



Page. 

Introduction 9 

Effect of physiographic and climatic features of Colorado on faunal and floral 

distribution 12 

Upper Sonoran zone 14 

Great Plains division of Upper Sonoran zone 18 

Mammals 19 

Breeding birds 20 

Plants 20 

Reptiles and batrachians 21 

Great Basin division of Upper Sonoran zone 22 

Colorado River drainage 22 

Northwestern section — Colorado River drainage 22 

Mammals 22 

Breeding birds 23 

Plants 23 

Reptiles and batrachians 23 

Southwestern section — Colorado River drainage 24 

Mammals 24 

Breeding birds 24 

Plants 25 

Reptiles and batrachians 25 

Rio Grande drainage 27 

Juniper and pinyon belt 28 

Mammals 28 

Breeding birds 29 

Plants 29 

Agricultural importance of < !olorado Upper Sonoran zone 29 

Transition zone 33 

Mammals 36 

Breeding birds 37 

Plants 38 

Reptiles and batrachians 39 

Agricultural importance of the Transition zone 40 

Canadian zone 41 

Mammals 44 

Breeding birds 44 

Plants 45 

Reptiles 45 

Hudsonian zone 45 

Mammals 48 

Breeding birds 48 

Plants 49 

Arctic- Alpine zone 49 

Mammals 50 

Breeding birds 50 

Plants . „ 51 

Mammals of Colorado 51 

Principal trees and shrubs of Colorado 212 

Index 247 

5 



ILLUSTRATIONS. 



PLATES. 



Page. 

Plate I. Map of Colorado showing life zones Frontispiece. 

II. Fig. 1. — Grand River Valley near De Beque, showing strip of Upper 
Sonoran zone. Fig. 2. — Pocket of Upper Sonoran zone on southern 
slope, at 7,500 feet. Near McCoy, Grand River Valley 16 

III. Fig. 1. — Greasewood (Sarcobatus vermiculatus) in lower Snake River 

Valley, Routt County. Fig. 2. — Bench bordering valley of Ver- 
milion Creek, northwestern Routt County. The shrubbery is 
Atriplex confertifolia and Grayia spinosa, with Juniperus mono- 
sperma on bluffs 22 

IV. Fig. 1. — Spruce Tree Cliff Ruins, Navajo Canyon, Mesa Verde, with 

juniper and pinyon forest in the rough canyon country of south- 
western Colorado. Fig. 2. — Navajo Canyon. Characteristic view 
in rough canyon region of southwestern Colorado, showing dense 

growth of junipers and pinyons at 7,000 feet 28 

V. Fig. 1 . — Farms in the lower White River Valley below Rangely, at 
5,500 feet. Fig. 2. — Fruit ranches in the McElmo Valley between 
Moqui and McElmo, at 5,200 feet 30 

VI. Fig. 1. — Effect of slope exposure — yellow pines (Pinus scopulorum) 
and firs (Abies concolor) growing at same elevation (9,000 feet) in 
Wet Mountains east of Westcliffe. Fig. 2. — Characteristic view 
at upper edge of Transition zone. Undergrowth of small aspens 
(Populus tremuloides) in forest of yellow pine (Pinus scopulorum), 

on Dolores Plateau 34 

VII. Fig. 1. — Dense chaparral of oak (Quercus gambeli) in western foothills 
of West Elk Mountains, on head of Smith Fork of Gunnison River, 
altitude 8,000 feet. Fig. 2. — Yellow pines (Pinus scopulorum) near 
Elkhorn, Larimer County, showing characteristic scattering growth 

of the species on eastern foothills of Front Range 38 

VIII. Fig. 1. — Lower part of Canadian zone forest of lodgepole pines (Pinus 
murrayana) and aspens (Populus tremuloides), on North Park slope 
of Medicine Bow Range. Fig. 2. — Forest of Engelmann spruce 
(J'icea engelmanni) , in upper part of Canadian zone on summit of 
Park Range, at 10,000 feet 40 

IX. Fig. 1. — Summer cattle range, Canadian zone meadow east of Laramie 
River, 10,000 feet altitude. Fig. 2. — Canadian zone vegetation at 
9,000 feet on open summit of Uncompahgre Plateau — Frasera, Del- 
phinium, Geranium, and Lupinus 42 

X. Fig. 1. — Hudsonian zone on precipitous slopes of San Juan Moun- 
tains, southwest of Ophir. Fig. 2. — Hudsonian zone forest on 
Saguache Range near St. Elmo (timberline at about 12,000 feet).. 44 

XI. Fig. 1. — Foxtail pines (Pinus aristuta) at timberline near St. Elmo, 
Saguache Mountains, at 12,300 feet. Fig. 2. — Ascending tongue of 
Engelmann spruce at timberline on Front Range near Berthoud 

Pass (northwest slope), at 11,600 feet 46 

XII. Fig. 1. — Grays Peak group from near Berthoud Pass, showing an 
extensive area of Arctic-Alpine country. Fig. 2. — Arctic-Alpine 
zone on the Saguache Range near St. Elmo 48 



ILLUSTRATIONS. 7 

TEXT FIGURES. 

Page. 
Fig. 1. Map of Colorado showing routes and collecting localities of Merritt Cary. 10 

2. Map of distribution in Colorado of tuft-eared squirrels (Sciurus aberti 

m im us and S. a.ferreus) 65 

3. Map of distribution in Colorado of Fremont squirrel (Sciurus frcmonti). 70 

4. Map of distribution in Colorado of Hopi, Say, and Las Animas chip- 

munks (Eutamlas hopiensis, E. quadrivittatus, and E. q. animosus).. . 72 

5. Map of distribution in Colorado of golden-mantled ground squirrels 

(genus Callospermophilus) 82 

G. Map of distribution in Colorado of antelope squirrel (Ammospcrmo- 

pliilus I. cinnamomeus) 85 

7. Map of distribution in Colorado of rock squirrel (Citellus r. grammurus) . 87 

8. Map of distribution in Colorado of Wyoming ground squirrel (Citellus 

elegans) 89 

9. Map of distribution in Colorado of striped ground squirrels (Citellus t. 

parvus and C. t. pallidus) 91 

10. Map of distribution in Colorado of grasshopper mice (Onychomys I. 

brevicaudus and . I. pallescens) 100 

11. Map of distribution in Colorado of cliff mice (Peromyscus truei and P. 

nasutus) 104 

12. Map of distribution in Colorado of harvest mice (Reithrodontomys) 

except R. albescens 109 

13. Map of distribution in Colorado of bushy-tailed woodrats (Neotoma c. 

arizonse, N. c. orolestes, and N. c. rupicola) 112 

14. Map of distribution in Colorado of round-tailed woodrats (Neotoma 

fallax, N. desertorum, JV. floridana bailey i, N. albigula ivarreni, and 

N. micropus canescens) 115 

15. Photograph of nest of Neotoma desertorum on Atriplex flat near Rangely. 119 

16. Map of distribution in Colorado of Hayden and pygmy field mice ( Micro- 

tus pauperrimus and M. ochrogaster haydeni) 122 

17. Map of distribution in Colorado of yellow pocket gopher (Geomi/s 

lutescens) 129 

18. Map of distribution in Colorado of chestnut-faced pocket gopher (Cra- 

togeomys castanops) 130 

19. Map of distribution in Colorado of pocket gophers of the genus 

Thomomys 132 

20. Photograph of earth heaps of Colorado pocket gopher ( Thomomys fossor) . 135 

21. Map of distribution in Colorado of kangaroo rats (genus Perodipus) ' 139 

22. Map of distribution in Colorado of pocket mice (Perognathus fascial us 

and P. apache groups) 144 

23. Map of distribution in Colorado of Baird pocket mouse (Perognathus 

flavus) 146 

24. Map of distribution in Colorado of jumping mice (Zapus princcps and 

Z. h. campestris) 149 

25. Map of distribution in Colorado of black-tailed jack rabbits (Lepus 

cali/ornicus texianus and L. c. melanotis) 156 

26. Map of distribution in Colorado of gray fox ( Urocyon c. scotti) 177 

27. Map of distribution in Colorado of dwarf weasel (Putorius streatori 

leptus) and the black-footed ferret (P. nigripes) 184 

28. Map of distribution in Colorado of marten ( Mustela caurina origenes) . . . 189 

29. Claw marks of black bear on aspen (Populus tremuloidcs) . Lone Cone, 

San Miguel Mountains, at 10,000 feet 197 

30. Map of distribution in Colorado of lodgepole pine (Pinus murrayana) 214 



8 ILLUSTRATIONS. 

Page. 

Fig. 31. Photograph of forest of lodgepole pines (Pinus murrayana) 215 

32. Photograph of pocket of Douglas spruce (Pseudotsuga mueronata) 219 

33. Map of distribution in Colorado of common yucca ( Yucca glauca) 223 

34. Photograph of Yucca baccata in flower 224 

35. Photograph of Alpine willows in Arctic-Alpine zone. Front Range 

near Berthoud Pass : 227 

36. Photograph of desert vegetation (Atriplex nuttalli and Sarcobatus ver- 

miculatus) in lower Grand River Valley north of Mack 229 

37. Map of distribution in Colorado of. tree cactus (Opuntia arborescens) . . . 241 

38. Photograph of bilberry ( Vaccinium oreophilum) in forest on Park Range . 244 

39. Photograph of desert sagebrush (Artemisia tridentata) on plains near 

Higho, North Park 246 



No. 33. NORTH AMERICAN FAUNA. August, 1911. 



A BIOLOGICAL SURVEY OF COLORADO. 



By Merkitt Cary. 



INTRODUCTION. 

Colorado, because of its diverse and striking physical features, 
presents an interesting field for natural-history investigations, and 
has received more than ordinary attention from naturalists. Never- 
theless, prior to 1905 the distribution of mammals and birds had been 
studied in detail over only a few small areas, chiefly on the eastern 
slope of the main ranges, on the adjacent plains, and in the southern 
end of the San Luis Valley. Field work in the more accessible parts 
of the State by several field naturalists of the Biological Survey had 
accumulated important series of specimens and valuable data on 
the fauna and flora. Many perplexing problems of geographical and 
vertical distribution and relationships of species, however, remained 
to be solved. Moreover, the intimate connection and correlation 
between the natural life zones and crop zones of the country, as 
pointed out by Dr. Merriam, 1 made increasingly apparent the lack 
of information respecting Colorado. The data on the distribution 
of species in Colorado available at the beginning of 1905 proved 
entirely inadequate for a detailed study of the life zones, although 
the need of accurate delimitation of the zones constantly increases 
as vast areas are opened to agriculture and horticulture through 
modern methods of conserving and distributing the water supply. 

Accordingly, in 1905, the writer was directed to undertake a 
biological reconnoissance of certain of the least-known sections of 
northern Colorado with a view to the completion of a detailed survey 
of the State. Field operations were begun early in June of that 
year. PreUminary work in the mountains of Boulder, Jefferson, and 
Clear Creek Counties was followed by a wagon trip from Boulder 
across the range to Middle Park by way of Rollins Pass; north over 
the western end of the Rabbit Ear Mountains into North Park; thence 

» Life Zones and Crop Zones of the United States, by C Hart Merriam, Bull. 10, Biological Survey, U. S. 
Dept. Agric, 1898. 

9 



10 



NORTH AMERICAN FAUNA. 



[No. 33. 



west to Steamboat Springs, crossing the Park Range over Buffalo 
Pass, and down the Bear River Valley as far as Axial Basin. Travel- 
ing southward over the Danforth Hills to Meeker and across the 
White River Plateau, the expedition finished its season's work in the 
Grand River Valley. 

Field work was continued in northern Colorado in 1906. Again 
outfitting at Boulder, I made a wagon trip of much greater length, 
completely encircling the area covered in 1905. 1 The route was 
along the northern and western boundaries of the State, from Boulder 
north to Fort Collins; west over the Medicine Bow Range to North 
Park; across the northern end of the Park Range to Hahns Peak and 




■ RAILROAD JOU/WCyS. 



tVASOA/ AA/O PACK Tff/PS. 



+ COLLECTING LOCALITY. 



Fig. 1.— Map of Colorado showing routes and collecting localities of Merritt Cary, 1905, 1906, 1907, 

and 1909. 

Slater; down the Snake River Valley to the Escalante Hills and 
northwest to Browns Park, on Green River; thence south to Rangely, 
on White River, and over the Book Cliffs to Mack, in the Grand River 
Valley; northeast in the valleys of Grand and Eagle Rivers to Wol- 
cott; north to Egeria Park; east across the Gore Range to Middle Park; 
and back to Boulder by way of Berthoud Pass and Black Hawk. 

The field season of 1907 was devoted to a detailed study of zonal 
conditions hi all the important physiographic areas of southern Colo- 
rado, including a trip into the little known La Sal Mountains along 

» The ready resource of my camp assistant, Mr. \Y alter Blanchard, of Boulder, contributed greatly to the 
success of the wagon trips of both 1905 and 190C. Over much of the region traveling was arduous and on 
some of the high mountain passes even dangerous. 



1911.] INTRODUCTION. 11 

the boundary of eastern Utah. Owing to the extent of territory 
covered, the railroads were largely utilized for travel, but many side 
trips were made by stage and pack train. 

Knowledge was still lacking as to the distribution of a great many 
plains species along the eastern edge of the State, and further work in 
that region was deemed necessary. Consequently part of the field 
season of 1909 was spent on the Arkansas Divide and on the plains 
from Cheyenne Wells northwest to Sterling, Colorado, and Cheyenne, 
Wyoming. This trip, although of short duration, filled in an impor- 
tant gap and yielded valuable distribution data. 

In studying the life zones of Colorado it has been necessary to 
collect many mammals, birds, reptiles, and plants, illustrating 
climatic variation, and to work out in detail the geographic and 
vertical distribution of all characteristic zone species. In regions 
devoted to agriculture and stock raising, the economic relations of 
both mammals and birds have been carefully investigated. Valuable 
information bearing on the past and present ranges and abundance of 
the larger game mammals and carnivores has been acquired from 
hunters, trappers, and sportsmen of wide experience in the State. 

The present report is based chiefly on the field work prosecuted 
during 1905, 1906, 1907, and 1909. It is primarily a characterization 
of the several major distribution areas or life zones, and their geo- 
graphical and vertical boundaries as here defined are believed to be 
approximately correct and in sufficient detail for all practical pur- 
poses. (See frontispiece.) In the course of three seasons' work it 
is obviously impossible to run continuous zone lines in a rough and 
broken region like the mountainous two- thirds of Colorado. To 
insure reasonable accuracy the zone limits have been checked up 
both horizontally and vertically b}^ crossing the important mountain 
ranges, plateaus, and mesas at different latitudes; while abnormal 
variations of zone level resulting from peculiar physiographic condi- 
tions have received special attention. 

The report on life zones is supplemented by a chapter on the dis- 
tribution of trees and shrubs. Many notes on birds and reptiles are 
incorporated in the sections of life zones to which the different species 
belong. 

The list of Colorado mammals, in addition to the results of the 
four seasons' investigations conducted by the writer, contains a large 
amount of important unpublished data collected prior to 1905 by the 
following naturalists of the Biological Survey: Vernon Bailey, Dr. 
A. K. Fisher, Edward A. Preble, Arthur H. Howell, J. Alden Loring, 
and Clark P. Streator. Among the local naturalists whose work has 
contributed materially to the completeness of the mammal report 
should be mentioned E. R. Warren, of Colorado Springs, the author of 
important publications on the mammals of Colorado. 



12 NORTH AMERICAN FAUNA. [No. S3. 

New species of mammals have been described from time to time 
from the material collected by the Biological Survey in Colorado. 
Among the mammals collected in southern Colorado in 1907 were 
three new forms. Owing to unavoidable delay in the publication of 
the present report, these have been characterized by Dr. Merriam in 
the Proceedings of the Biological Society of Washington as follows: 
Eutamias minimus caryi, Neotoma albigula warreni, and Tliomomys 
talpoides agrestis. 1 

In preparing the mammal report I have freely used the collection of 
the Biological Survey and the private collection of Dr. C. Hart Mer- 
riam, both in the United States National Museum. I am indebted to 
Gerrit S. Miller, jr., curator of mammals in the National Museum, for 
access to the collections under his charge. Dr. Leonhard Stejneger, of 
the same institution, has identified the reptiles and batrachians. Dr. 
J. N. Rose and his assistants have named many of the plants. 

EFFECT OF PHYSIOGRAPHIC AND CLIMATIC FEATURES OF 
COLORADO ON FATJNAL AND FLORAL DISTRIBUTION. 

Colorado is no exception to the rule that a region of varied climatic 
and physiographic conditions possesses a correspondingly large and 
varied fauna and flora. That it is surpassed by a few other States in 
the variety of its animal and plant life is due, not so much to a lack of 
varied physical conditions as to the absence of climatic extremes. 
The comparatively rich fauna and flora of the State are largely due 
to the great range of altitude in the mountains, since the high basal 
plane lies entirely within the arid region. Colorado has neither the 
extent of latitude nor the low, hot areas of great humidity or aridity 
which contribute to the wonderfully rich fauna and flora of such States 
as California and Texas. The extensive eastern plains, the Rocky 
Mountain system traversing the central part, and the rough region of 
alternating plateaus, desert valleys, and mesas on the western slope, 
divide the State into three general topographic regions, each occupy- 
ing approximately a third of its total area. Dissimilar physical and 
climatic conditions still further divide these into smaller irregular 
areas which differ considerably in their fauna and flora. Thus, on the 
western slope, high mesas, clad with a scrubby and more or less 
scattering forest growth receiving moderate rainfall, alternate with 
lower arid desert stretches (some with less than 10 inches of annual 
rainfall) , while in the mountains are extensive belts of heavy forest with 
much greater humidity (20 to 25 inches annual rainfall) and bare alpine 
crests and summits above timberline having a truly arctic climate. 

East of the mountains are found the fauna and flora peculiar to 
the Great Plains region, with a slight admixture of Mississippi Valley 
species in the river valleys along the eastern boundary of the State. 

i Proc. Biol. Soc. Wash., XXI, pp. 143-144, June 9, 1908. 



1911.] PHYSIOGRAPHIC AND CLIMATIC FEATURES. 13 

The lower elevations on the western slope have a strong infusion 
of desert species, characteristic of the Great Basin region. The 
Rocky Mountain system has a complex fauna and flora — the foot- 
hills, the open valleys and parks, the dense forests of the central 
slopes, and the alpine summits, all having associations of mammals, 
birds, and plants, which are more or less distinctive. 

Altitudinal variation in Colorado has a pronounced effect upon 
both temperature and moisture, and accounts for great extremes in 
summer and winter temperatures in different sections, and also for 
a wide range between night and day temperatures throughout the 
year. At Grand Junction (4,500 feet) the winter and summer mean 
temperatures are 29° and 75° F.; at Gunnison (7,600 feet) they are 
11° and 59°; at Breckenridge (9,500 feet) they are 16° and 52°; 
while along the summits of the main mountain ranges the tempera- 
tures are still lower. 

As temperature is a very important factor in the distribution of 
life, Colorado is an exceptionally favorable field for illustrating 
vertical distribution. Tn fact, the wide range of elevation, from 
3,000 and 4,000 feet along the eastern boundary to considerably over 
14,000 feet on the summits of the main ranges, furnishes favorable 
conditions for characteristic species of five of the seven major life 
zones of North America. Thus, in passing from the plains of eastern 
Colorado to the summit of either the Front or Sangre de Cristo 
Ranges, or from the warm desert valleys or sage plains of the western 
counties eastward to the crest of the Continental Divide, the fol- 
lowing life zones are successively traversed: (1) Upper Sonoran, 
which includes all the basal plane of the State from which the moun- 
tains rise; (2) Transition or yellow-pine zone, occupying the foothill 
region; (3) Canadian, or zone of coniferous boreal forests, com- 
prising the lodgepole pine, aspen, fir, and spruce belts of the middle 
mountain slopes; (4) Hudsonian, or belt of dwarfed conifers 
extending to timberline; and (5) Arctic-Alpine zone, reaching from 
timberline to the summits of the highest mountain peaks. With 
the exception of the first named, all are included in full width, and 
the lower border of the Upper Sonoran zone is nearly reached in some 
of the low desert valleys of the southwest. Of the above zones the 
Upper Sonoran and Transition alone are of agricultural importance. 
Their adaptation to various crops is discussed under their respective 
headings. 

The Austral element (Upper Sonoran zone) occupies territory 
aggregating nearly half the area of the State, or about twice as much 
as the Boreal element (including Canadian, Hudsonian, and Arctic- 
Alpine zones). The region of overlapping austral and boreal species 
(Transition zone) is about equal in extent to that occupied by the 
boreal zones. 



14 



NORTH AMERICAN FAUNA. 



[No. 33. 



On the eastern slope of the front ranges, as a result of the rela- 
tively abrupt ascent from the base level of the plains, the several 
life zones are usually restricted to horizontally narrow and well- 
defined belts. West of the Continental Divide, however, they often 
cover broad and extensive areas, as the incline of the country is more 
gradual. There is, moreover, in consequence, a decided overlapping 
of species for a considerable horizontal distance, rendering the zonal 
limits less clearly defined than on the steeper slopes. The boundaries 
of the zones in western Colorado are extremely irregular and sinuous, 
owing to the broken and diversified character of the country, and 
especially to the warm desert valleys which enter the State from 
the west and deeply penetrate and parallel the westward extending 
boreal plateaus and mountains. 

The vertical breadth of the zones is much the same on both sides of 
the Continental Divide. The zone levels, however, are a few hundred 
feet higher on the warmer western slope, particularly in northern 
Colorado. In southern Colorado the combined effects of lower lati- 
tude and more elevated base level are noticeable in higher zone 
levels, which are about the same on both sides of the Continental 
Divide. 

The vertical boundaries assigned the several zones are necessarily 
approximate, since various governing factors, chief among which are 
slope exposure, steepness of slope, and deforestation, either force 
species above their normal limits or restrict their limits below nor- 
mal. On the border of each zone is a belt of varying vertical breadth 
(usually from 400 to 800 feet) in which species of the two bordering 
zones commingle. The following table gives the extreme vertical 
limits of zones in Colorado: 



Extreme vertical limits of zones in Colorado. 





Northern Colorado. 


Southern Colorado. 


Zone 


Northeast ex- 
posure. 


Southwest ex- 
posure. 


Northeast ex- 
posure. 


Southwest ex- 
posure. 


Upper Sonoran 

Transition 


Feet. 

to 5,600 

5, 600 to 7,500 

7, 500 to 10, 000 

10, 000 to 10, 900 

10, 900 to 


Feet. 
to 6,500 
6, 500 to 8,200 
8, 200 to 10, 400 
10, 400 to 11. 000 
11, 600 to 


Feet. 
to 6,500 
6, 500 to 8,000 
s. odd to 10, 500 
10,500 to 11,200 
11, 200 to 


Feet. 
to 7,800 
7. 800 to 9,000 
9.000 to 11,000 


Hndsonian 


11.000 to 12,000 
12, 000 to 







UPPER SONORAN ZONE. 



The entire base level of Colorado, comprising territory aggregating 
nearly half its area, or about 50,000 square miles, lies within the 
Upper Sonoran or arid division of the Upper Austral zone. As it is 
preeminently the agricultural zone of the State, it will be character- 
ized in considerable detail. 



1011] UPPER SONORAN ZONE. 15 

East of the mountains the Upper Sonoran zone occupies a broad 
and continuous area of nearly uniform width, comprising the plains 
region, besides a stretch of rough and broken juniper country 
south of the Arkansas River and a narrow strip along the eastern- 
most flanks of the foothills up to the point where the junipers (Juni- 
perus monosperma) and pinyons (Pinus ednlis) give way to the 
yellow pines of the Transition zone — at an elevation varying from 
5,300 to 6,000 feet north of the Arkansas Divide, and from 6,200 
to 7,300 feet farther south, according to the exposure and steepness 
of the foothill slopes. The upper (western) boundary of the eastern 
Colorado Upper Sonoran is tolerably regular from the Wyoming line 
south to the Arkansas Divide, where it bends far eastward to skirt 
this extensive Transition area. South of the Divide its regularity 
is further broken by a long, narrow arm which penetrates the moun- 
tains as far as Buena Vista, in the warm valley of the Arkansas, 
and to a less degree by westward extensions in the valleys of the 
Huerfano, Cucharas, and Las Animas Rivers still farther south. 
The following localities in the eastern foothill region well indicate 
the upper limit of the Upper Sonoran zone: Livermore, Arkins 
(2 miles west), Lyons, Boulder, Golden, wSedalia, Ramah (Arkansas 
Divide), Pikeview, Manitou (Fountain Creek Valley), Buena Vista 
(Arkansas Valley extension), Malachite (Huerfano Valley), La Veta 
(Cucharas Valley), and Weston (Las Animas Valley). 

In western and southern Colorado the Upper Sonoran element 
occupies much smaller and extremely irregular areas. It is largely 
confined to long and sinuous arms or tongues in the warm, semi- 
desert river valleys, which penetrate deeply the boreal country of 
the mountains and plateaus; but it also crosses some of the inter- 
vening watersheds at their lowest elevations and everywhere occu- 
pies the lowest slopes of mesas, plateaus, and ridges to an altitude 
of about 6,500 feet, regularly reaching 7,500 or even 7,800 feet on 
the steep southwestern slopes which many of the elevations present 
to the hottest rays of the sun. 

West of the Continental Divide the Upper Sonoran zone embraces 
two regions of irregular outline separated in Colorado by the boreal 
cap of the Book Cliffs, although connected along the Green River 
Valley in Utah. The northern and smaller of these two areas is a 
southward extension of the Red Desert Upper Sonoran of Wyoming, 
connecting across the divide south of Bear River with the White 
River Valfey extension of the Utah Upper Sonoran. Practically all 
of Routt and Rio Blanco Counties below 6,500 feet is included within 
this area. The northern half (western Routt County) is a region of 
extensive sage plains, while the rough region contiguous to the 
White River drainage is for the most part clothed with pinyons and 
junipers. The eastern and southern limits of this area are reached 



16 NORTH AMERICAN FAUNA. [No. 33. 

at or near the following localities: Dixon, Wyoming, Snake River 
Valley; Fortification, Colorado, west base Elk Head Mountains; half 
way between Hayden and Craig, Bear River Valley; north base 
Williams River Mountains; Hamilton, Williams Fork of Bear River; 
Axial, Axial Basin; north bases of Yampa and Wapiti Peaks; middle 
southwestern slopes of Danforth and Gray Hills; and northern base 
of White River Plateau, east in a rapidly narrowing tongue to a 
point on White River about 15 miles southeast of Meeker, where the 
Upper Sonoran is restricted to a narrow strip of pinyons and junipers 
occupying the warm southwestern slopes of the lowest hills bordering 
the river on the north. Narrow tongues extend southward in the 
valleys of Piceance and Bitter Creeks; thence the zone includes the 
eastern and northern bases and lower western slopes of Cathedral 
Bluffs, the northern bases of the Book Cliffs, and most of the valley 
of Evacuation Creek. In extreme western Routt County the west- 
ward continuity of the Upper Sonoran is interrupted by three arms 
of high country which extend a short distance into the State from 
Utah — the O-wi-yu-kuts Plateau and Diamond Peak on the north; 
the eastern extension of the Uinta Mountains, terminating in Mount 
Cullom, west of Green River; and, still farther south, the Yampa 
Plateau. East of Green River it is still further broken by the sum- 
mits and upper northern slopes of the Escalante Hills, which form a 
Transition and Canadian zone island separated from the Uinta 
Mountains on the west and the Yampa Plateau on the south by a 
narrow tongue of Upper Sonoran zone extending through the Ladore 
Canyon of Green River on its warm eastern side and along the north 
side of the Yampa Canyon of Bear River. 

South of the Book Cliffs the Upper Sonoran zone covers broad 
belts in the lower desert valleys of the Grand, Gunnison, Dolores, 
and San Juan Rivers and their chief tributaries. Over most of this 
region the pinyons, junipers, and other characteristic Upper Sonoran 
species of plants, birds, and mammals reach their limit of upward 
dispersion at a little over 7,000 feet elevation, but the variation due 
to slope exposure in a region so broken and incised is such as to 
prevent exact delimitation. Thus on the exposed southern faces of 
the Book Cliffs, from the Utah line east to the Little Book Cliffs, the 
upper limit conforms closely to the 7,000-foot contour; while on the 
remarkably bold and exposed slopes on the southwestern extremity 
of Grand Mesa, north of Delta, and also on the Landsend Peak, the 
most western point of the West Elk Mountains — unbroken slopes 
which rise directly out of the desert to a great height — the belt 
of junipers and pinyons can be traced to considerably over 8,000 
feet, and probably occasionally reaches 8,500 feet. Looking north 
from Delta (4,950 feet) in the bottom of the Gunnison Valley to the 
bold southwest slope of Grand Mesa, a striking view is obtained of 



North American Fauna No. 33, U. S. Dept. Agr. Biological Survey. 



Plate II 




Fig. 1.— Grand River Valley near De Beque, 
Showing Strip of Upper Sonoran Zone. 




- --» ■ " • . i~\i - . -47 «^*43^sfe3b£^r l 



Fig. 2.— Pocket of Upper Sonoran Zone on Southern Slope, at 7,500 Feet. 
Near McCoy, Grand River Valley. 



1911.] UPPER SONORAN ZONE. 17 

perhaps the greatest vertical breadth of the Upper Sonoran zone visi- 
ble at any point in the State. In contrast to the conditions on warm 
exposed slopes is the cooling effect exerted by north and especially 
northeast slopes and canyon sides, shaded during the greater part of 
the day, and also by mountain streams, which together create an 
environment unfavorable for Upper Sonoran species above 6,000 
feet. Examples of these conditions appear on the south side of 
Grand River below Glenwood Springs and also in the Grand Canyon 
a few miles above, where the Douglas spruces and other Transition 
zone forms occupy the slopes to below 6,000 feet; in constrictions of 
the Unaweep Canyon, which cuts through the Uncompahgre Plateau 
near its northern end; and in a great many other localities. 

The Grand Valley Upper Sonoran enters Colorado from the Utah 
deserts in a broad belt and penetrates the high country in a long and 
sinuous tongue for over 150 miles in an east-northeast direction. 
East of Palisade, where the Little Book Cliffs and Grand Mesa ap- 
proach each other, it is greatly narrowed, but again widens to include 
the valleys of Plateau and Roan Creeks. Between De Beque and 
Newcastle it is confined to a strip 2 or 3 miles wide along Grand 
River (see PI. II, fig. 1), and east of Newcastle occupies a still nar- 
rower strip along the north side of the valley to Glenwood Springs, 
continuing up the Roaring Fork to Basalt. The continuity of the 
Upper Sonoran in the Grand Valley is broken by the Transition 
environment obtaining through the entire length of the Grand 
Canyon, from Glenwood to Dotsero, but the zone is again encoun- 
tered in dilute form on the warm southern slopes of the lowest hills 
along the north side of Eagle River nearly to Wolcott. The Grand 
Valley between Dotsero and McCoy was not explored, but a well- 
defined Upper Sonoran pocket 4 or 5 miles in width was traversed 
on the northern side of the river at McCoy. At this point both 
pinyons and junipers were growing on the warm southern slopes up 
to 7,800 feet. (PL II, fig. 2.) The Upper Sonoran belt in the Gunni- 
son and Uncompahgre Valleys is uniformly broad, and covers a 
greater area than that in the Grand Valley, of which it is an offshoot. 
Its eastern and southern limits are marked by Somerset (North Fork 
of Gunnison River), near River Portal (Gunnison Canyon), between 
Cedar Creek and Cerro Summit, and near Portland (Uncompahgre 
Valley). 

Another extension of the Grand Valley Upper Sonoran follows 
southeast along the Dolores River, widening in the San Miguel and 
Paradox regions to include practically all the rough canyon country 
between the western escarpment of the Uncompahgre Plateau and 
the eastern flanks of the La Sal Mountains. Norwood is very near 
the eastern limit of the Upper Sonoran in the San Miguel Valley, and 
90432°— n 2 



18 NORTH AMERICAN FAUNA. [No. 33. 

Lavender marks its upper edge on Disappointment Creek, south of 
the San Miguel Mountains. South of the junction of the Disap- 
pointment with the Dolores tins area connects across an extensive 
sage desert with a northward extension of the San Juan Upper 
Sonoran area, which covers practically all the extensive desert 
region lying west of Dolores River and south of the La Plata Moun- 
tains. Continuing eastward in the San Juan drainage, arms of this 
zone extend northward in the valleys of all northern affluents — 
along the Rio Las Animas to Durango, along the Rio Pinos nearly 
to Bayfield, and on the Rio Piedra to a point opposite the Piedra 
Padre, while along the San Juan it reaches nearly to Trujillo. 

The intermountain region is penetrated by the Upper Sonoran 
zone only in the Rio Grande Valley. An arm of considerable breadth 
enters this region from New Mexico along the Rio Grande, where it is 
well characterized in a wide strip on each side of the river nearly as 
far as Alamosa. In a dilute form the Upper Sonoran element 
spreads over most of the open country as far as Del Norte, and north 
over the level plains of the San Luis Park to a little beyond Moffat. 
It includes also the J^elt of junipers and pinyons on the hot foothill 
slopes of the San Juan and Garita Mountains on the west and on the 
lower flanks of the Sangre de Cristo Range on the east side of San 
Luis Valley to an elevation of 8,500 feet, or about 1,000 feet above 
the mean level of the valley. The cold water in the many side streams 
of San Luis Valley — the Conejos, La Jara, Alamosa, Saguache, San 
Luis, Trinchera, Culebra, and the branches of the Rio Grande above 
Alamosa — carries narrow strips of Transition zone far out into the 
plain, and the immediate valleys of these streams can not be included 
in the Sonoran area. 

The Upper Sonoran zone includes within the State two well- 
marked subdivisions differing considerably in their physical charac- 
teristics. These are the Great Plains region of eastern Colorado and 
the Great Basin region of western Colorado including San Luis Val- 
ley. Since each possesses a fauna and flora in a large degree peculiar 
and distinctive, they will be treated as minor distribution areas. 

Great Plains Division of the Upper Sonoran Zone. 

With the exception of Las Animas County and parts of Pueblo, 
Otero, Bent, and Baca Counties, practically all of Colorado east 
of the foothills is a vast undulating grassy plain which varies in 
elevation from 3,000 and 4,000 feet along the eastern boundary to 
between 5,000 and 6,000 feet where it approaches the foothills. 
Throughout this region strips of sandy country alternate with areas 
of firm soils. The chief rivers are the Platte and Arkansas, which, 
with their many small tributary streams, drain most of the eastern 



1911.] UPPEE SONORAN ZONE. 19 

part of the State through shallow valleys. A large number of these 
tributaries indicated upon maps are dry most of the year, and only 
a small proportion are perennial. Rock exposures are infrequent, 
being found chiefly in the eastern tier of counties, where some of the 
streams of the Republican River drainage, in seeking a lower level, 
have worn through beds of sandstone; and in northeastern Weld 
and northwestern Logan Counties. The plains are devoid of 
natural tree growth, except along the watercourses, which are 
usually fringed with cottonwoods, various species of willows, and 
dense thickets of wild plum and cherry, with scattering clumps of 
hackberry and box elder in the neighboring gulches and arroyos. A 
little sagebrush is found toward the north, in Weld and adjoining 
counties, but over most of the region the limited shrubbery is con- 
fined to the stream valleys. Range grasses in great variety grow 
luxuriantly on the plains and form the dominant vegetation. 

The fauna and flora of this region are essentially those of the 
Great Plains, from Nebraska and South Dakota to the Panhandle of 
Texas. The following associations of mammals, birds, plants, and 
reptiles are more or less characteristic. 

MAMMALS OF GREAT PLAINS. 

Those mammals of the plains of eastern Colorado which are 
restricted to the Upper Sonoran zone are the large spotted and 
Kennicott spermophiles (Citellus spilosoma major and C. obsoletus); 
pale grasshopper mouse (Onychomys leucogaster pallescens) ; white- 
footed mouse (Peromyscus leucopus tornillo 1 ); harvest mouse (Rei- 
throdontomys nebrascensis) ; wood rats (Neotoma cinerea rupicola, 
N.floridana baileyi, and N. micropus canescens 2 ) ; Hay den vole {Micro- 
ins ochrogaster haydeni); pocket gophers (Geomys lutescens and 
Cratogeomys castanops 3 ) ; kangaroo rat (Perodipus montanus richard- 
soni) ; pocket mice (Perognatlius flavescens, P. flavus, P. hispidus 
paradoxus); black-tailed jack rabbit (Lepus californicus melanotis); 
Bailey cottontail (Sylvilagus auduboni baileyi); long-tailed skunk 
( Mephitis mesomelas varians) ; prairie spotted skunk (Spilogale inter- 
rupta) ; long-eared and hairy-lipped bats ( Myotis evotis and M. 
californicus ciliolabrum) . 

Other mammals which have their center of abundance in the 
Upper Sonoran zone, but are not entirely restricted to it, are the 
antelope (Antilocapra americana); striped spermophile (Citellus 
tridecemlineatus pallidus); prairie dog (Cynom,ys ludovicianus) ; 
white-footed mouse (Peromyscus maniculatus nebrascensis); coyote 
(Canis nebracensis) ; swift fox (Vulpes velox) ; badger (Taxidea taxus) ; 

1 In brush fringe along streams only. 2 Only in extreme southeast. 

' 3 Only in southeastern counties. 



20 NORTH AMERICAN FAUNA. [No. 33. 

black-footed ferret (Putorius nigripes); long-tailed weasel (Putorius 
longicaudus) ; and the Say and brown bats (Myotis subulatus and 
Eptesicus fuscus) . 

BREEDING BIRDS OF GREAT PLAINS. 

Among the characteristic Sonoran birds which breed on the plains 
are the following: California cuckoo (Coccyzus americanus occiden- 
talis); burrowing owl (Speotyto cunicularia hypugxa); Arkansas 
kingbird (Tyrannus verticalis); Bullock oriole (Icterus bullocM); 
western grasshopper sparrow (Ammodramus savannarum bimaculatus) ; 
western lark sparrow (Ohondestes grammaeus strigatus) ; lazuli bunting 
(Passerina amcena); western blue grosbeak (Guiraca cserulea lazula); 
Bell vireo (Vireo belli); long-tailed chat (Icteria virens longicauda) ; 
catbird (Dumetella carolinensis) ; brown thrasher ( Toxostoma rufum) ; 
and western mockingbird (Mimus polyglottos leucopterus). 

Other birds breeding commonly on the plains, which are not so 
closely restricted to the Upper Sonoran zone, are: Upland plover 
(Bartramia longicauda); long-billed curlew (Numenius americana) ; 
mountain plover (Podasocys montanus); black-crowned night heron 
(Nycticorax n. nsevius); bobwhite (Colinus virginianus) ; ferruginous 
rough-legged hawk (Archibuteo ferrugineus) ; mourning dove (Zenai- 
dura macroura carolinensis); kingbird (Tyrannus tyrannus); thick- 
billed redwing (Agelaius phamiceus fortis) ; western meadowlark 
(Sturnella neglecta); bronzed grackle (Quiscalus quiscula seneus); 
lark bunting (Calamospiza melanocorys) ; western vesper sparrow 
(Poacetes gramineus confinis) ; Brewer sparrow (Spizella breweri) ; 
white-rumped shrike (Lanius ludovicianus excubitorides) ; western 
yellowthroat (GeotJilypis trichas occidentalis) ; and yellow warbler 
(Dendroica sestiva). 

PLANTS OF GREAT PLAINS. 1 

Aside from the trees and larger shrubs along streams, including 
cottonwood (Populus occidentalis), several species of willows (Salix), 
box elder (Acer negundo), hackberry (Celtis reticulata), buffalo berry 
(Lepargyrea argentea), golden currant (Ribes longijlorum) , wild plum 
(Prunus americana), chokecherry (Prunus melanocarpa), wolfberry 
(Symphoricarpos occidentalis), sumac (Schmaltzia trilobata), and false 
indigo (Amorplia angustifolia) , the most conspicuous plants and 
smaller shrubs of the plains region are: 

1 Throughout this report, with a few exceptions, the botanical nomenclature used is that of Rydberg ; 
Flora of Colorado, Bull. 100, Colo. Agric. College Experiment Station, 1906. 



1911.] 



UPPER SONORAN ZONE. 



21 



Amorpha canescens. 
Aragallus lamberti. 

Artemisia filifolia . 
Atriplex canescens (local). 
Asclepias (several species). 
Astragalus mollissimus. 
Astragalus crassicarpus. 
( '(Kins missouriensis. 
Chrysothamn us phttensis. 
Gaura cocdnea. 
(,'ulicrrezia sarothrae. 
Glycyrrhiza lepidota. 
Helianthus lenticularis. 
Ipomaea leptophylla. 
Laciniaria punctata. 
Leucocrinum montanum. 
Lithospermum linearifolium. 
Lupinus pusillus. 
Malvastrum coccineum. 
Meriolix sen u lata. 



Oenothera (several species). 

Opuntia polyacantha. 

Opuntia arborescens (Arkansas Valley and 

southward). 
Peritoma serrulatum. 
Petalostemon villosum. 
Petalostemon oligophyllus. 
Petalostemon purpureas. 
Planfago purshi. 
Psoralea lanceolata. 
Psoralen kypogea. 
Psoralea tenuijlora. 
Prunus besseyi (in sandy country). 
Touterea stricta. 
Touterea nuda. 

Toivnsendia (several species). 
Ratibida columnaris. 
Verbena hastata. , 
Verbena bracteosa. 
Yucca glauca. 



The following list 
eastern Colorado is 
Plant Industry: 1 

Agropyron smithi. 
Andropogon furcatus. 
Andropogon halli. 
Andropogon -scoparius. 
Aristida longiseta. 
Bouteloua curtipendula. 
Bouteloua hirsuta. 
Bouteloua oligostachya. 
Buchloe dactyloides. 
Calamovilfa longifolia. 
Festuca octojlora. 



of conspicuous grasses of the Great Plains of 
taken from a recent report of the Bureau of 

Koeleria cristata. 
Muhlenbergia gracillinm . 
Muhlenbergia pungens. 
Munroa squarrosa. 
Panicum virgatum. 
Redfieldia ftexuosa. 
Schedonnardus pa n iculatns. 
Sitanion hystrir. 
Sporobolus cryptandrus. 
Stipa comata. 



REPTILES AND BATRACHIANS OF GREAT PLAINS. 

Most of the reptiles of eastern Colorado are common to the semi- 
arid part of the Great Plains. Aside from the rattlesnake (Crotalus 
confluentus) , bull snake (Pituopliis sayi), blue racer (Bascanion con- 
strictor), and hog-nosed snake (Ileterodon nasicus), lizards are the 
most noticeable reptiles, particularly hi sandy tracts. These com- 
prise, among others, the horned toad (Phrynosoma ornatissimum) , 
sand swift (IlolbrooJcia maculata), whip-tailed lizard (Cnemidoplwrus 
gularis), and Sceloporus consobrinus. A large, beautifully colored 
ring-necked lizard (probably Crotaphytus collaris) reaches the plains 
of eastern Colorado in Baca County, in the extreme southeast, 



'■ Natural Vegetation as an Indicator of the Capabilities of Land for Crop Production in the Great Plains 
Area. Bureau of Plant Industry Bull. 201, pp. 20-62, March 16, 1911. 



22 NORTH AMERICAN FAUNA. [No. 33. 

where it is reported as not uncommon. A salamander (Ambystoma 
tigrinum) from Loveland is in the Biological Survey collection. 

Great Basin Division of Upper Sonoran Zone. 

The Upper Sonoran area of Colorado lying in the drainage of the 
Rio Grande and Colorado Rivers is an integral part of the Great Basin 
region and differs from the Great Plains area in more arid climate 
and more barren and deeply eroded surface. A large proportion of 
its plants and animals are Great Basin forms, either specifically or 
subspecifically different from those of the Great Plains. Its open 
areas are generally characterized by the bunch-like growth of desert 
shrubs commonly termed sagebrush, grease brush, and rabbit brush, 
and the foothills and valley margins by a scrubby growth of junipers 
and nut pines. This area is best considered under its local subdi- 
visions. 

COLORADO RIVER DRAINAGE. 

The territory embracing the open reaches in the warm lower 
valleys from Snake River south to the San Juan possesses a fauna 
and flora characterized by a large number of Great Basin desert forms, 
few of which show appreciable differentiation in passing through the 
entire width of the western Colorado Upper Sonoran. Considered 
locally, however, with respect to both faunal and physical character- 
istics, the Colorado River drainage readily admits of division into 
two minor distribution areas — the northwestern and southwestern 

sections. 

Northwestern Section — Colorado River Drainage. 

A broad expanse of undulating sandy plains and low watersheds 
lying chiefly north of Bear River in west central Routt County is 
characterized by a dense growth of sagebrush (Artemisia tridentata). 
With an average elevation of 6,000 feet, this region is higher than the 
desert valleys to the south and has a mixture of Transition species 
in its fauna and flora. Its plants and animals, unlike those of the 
valleys of the Colorado drainage farther south, are partly derived 
from the Great Plains. The trees and slirubs, consisting of scattered 
cottonwoods, dense copses of willows, and thickets of buffalo berry, 
are confined to the banks of watercourses. A sparse growth of 
juniper covers the canyon sides and the steep faces of bluffs. The 
region is drained by the Snake and Bear Rivers, which flow through 
valleys of considerable depth — often through rugged canyons. 

MAMMALS OF NORTHWESTERN SECTION COLORADO RIVER DRAINAGE. 

The following mammals characteristic of the Routt County sage 
plains have reached that region from the west and north, and most 
of them are absent from the southwestern valleys: The least chip- 
munk (Eutamias minimus), Wortman ground squirrel (Callospermo- 



North American Fauna No. 33, U. S. Dept. Agr. Biological Survey. 



Plate III 










Fig. 1 .— Greasewood (Sarcobatus vermiculatus) 
in Lower Snake River Valley, Routt County. 




Fig. 2.— Bench Bordering Valley of Vermilion Creek, Northwestern Routt 

County. 

The shrubbery is Atriplex coiifrrtifnlia and Grayia spinosa, with Junipt rus monospi rum on bluffs. 



1911.] UPPEE SONORAN ZONE. 23 

philus lateralis wortmani), little striped ground squirrel {CiteUus 
tridecemlineatus parvus), Idaho grasshopper mouse (Onychomys 
leucogaster brevicaudus), Red Desert pocket mouse (Perognathus 
callistus), Green River pocket gopher (TJtomomys clusius ocius), and 
Fort Yuma bat (Myotis yumanensis). Species derived from the 
Great Plains region are Peromyscus maniculatus nebrascensis, Pero- 
dipus mohtanus richardsoni, Sylvilagus auduboni baileyi, Canis 
nebracensis, and Myotis californicus ciliolabrum. The long-eared bat 
(My otis evotis) occurs on the plains on both sides of the mountains. 
Species which are common over this area but which regularly range 
higher are: Lepus campestris townsendi, CiteUus elegans, Cynomys 
leucurus, Neotoma cinerea orolestes, and the gray wolf and antelope. 

BREEDING BIRDS OP NORTHWESTERN SECTION COLORADO RIVER DRAINAGE. 

The sage sparrow (Ampliispiza nevadensis), Brewer sparrow 
(Spizella breweri), and Bullock oriole (Icterus bullocki) are character- 
istic Upper Sonoran breeders in the low areas of northwestern Colo- 
rado, while the sage thrasher (Oreoscoptes montanus), white-rumped 
shrike (Lanius ludovicianus excubitorides) , desert horned lark (Otocoris 
alpestris leucolsema) , and San Diego redwing (Agzelaius phamiceus 
neutralis) breed commonly but nest also in the Transition zone. 
The sage hen (Centrocercus uropliasianus) is an occasional breeder, 
although restricted mainly to the Transition zone during the summer. 
Many Great Basin species of birds which are common breeders hi 
the valleys of southwestern Colorado are noticeably absent from the 
northwestern valleys. 

PLANTS OF NORTHWESTERN SECTION — COLORADO RIVER DRAINAGE. 

Conspicuous shrubs and plants on the sandy sage plains of Routt 
County are Artemisia tridentata, Opuntia polyacantha, Eurotia lanata, 
and several species of Chrysoiliamnus and Eriogonum; on the adobe 
soil of the first and second benches above the streams, Grayia spinosa, 
Atriplex nidtalli, A. canescens, and A. confertifolia (PI. Ill, fig. 2) ; on 
the alkaline flats in the valley bottoms, Sarcobatus vermiculatus 
(PI. Ill, fig. 1), Dondia erecta, and, in damp places, Salicornia her- 
bacea; on the rocky faces of bluffs, Juniperus monosperma, Kunzia 
tridentata (mainly Transition), Ephedra antisyphylitica, Yucca 
glauca; along streams, Populus acuminata, Lepargyrea argentea, Salix 
amygdaloides, and others. 

REPTILES AND BATRACHIAXS OF NORTHWESTERN SECTION — COLORADO RIVER DRAIN4.GE. 

Reptiles are few in number compared with those in the valleys 
farther south. Rattlesnakes (Crotalus confluentus) and horned toads 
(Phrynosoma ornatissimum) were the only reptiles noted on the sandy 
sage plains in August, 1905, although the dainty mottled Sceloporus 



24 NORTH AMERICAN FAUNA. [No. 33. 

graciosus was collected among the rock ledges along the Bear River 
bluffs near Maybell, and also on Snake River, 20 miles west of Baggs 
( Irossing. The leopard frog (Rana pipiens) is common in the streams 
of this section, a specimen having been collected on Vermillion Creek, 
near Lad ore, September 3, 1906. 

Southwestern Section — Colorado PavER Drainage. 

Although traversed by such large perennial streams as the White, 
Grand, Gunnison, Dolores, and San Juan, and their numerous tribu- 
taries, this isj throughout, a region of great aridity, and presents the 
onlv truly desert conditions found within the State. Although small 
in area it is of great importance from both a biological and a hor- 
ticultural standpoint. Most of the region is below 5,500 feet, the 
Grand Valley below Grand Junction being under 4,500 feet. In the 
extreme southwest, however, a few desert areas reach 6,000 feet. 
The larger part of the soils are to be classed as adobe. In common 
with all arid desert regions, the warm valleys of southwestern Colo- 
rado, when brought under irrigation, are remarkably productive, 
especially of fruit. 

MAMMALS OF SOUTHWESTERN SECTION — COLORADO RIVER DRAINAGE. 

The most representative of the mammals are the antelope squirrel 
(Ammospermopkilus leucurus cinnamomeus), pale grasshopper mouse 
{Onychomys leucogaster pallescens), golden-breasted canyon mouse 
(Peromyscus crinitus auripectus), big-eared harvest mouse (Reithro- 
dontomys megalotis), desert wood rat (Neotoma desertorum) / golden 
pocket gopher {Thomomys aureus), Moki kangaroo rat (Perodipus 
longipes), Apache pocket mouse (Perognathus apache), cottontail 
rabbit (Sylvilagus auduboni warreni), Texas jack rabbit (Lepus 
californicus texianus), cacomistle (Bassariscus astutus Jlavus) , Arizona 
skunk (MepMtis estor), Great Basin spotted skunk (Spilogale gracilis 
saxatilis), coyote (Canis estor), and the bats, Antrozous pallidus, 
Myotis californicus, M. evotis, Pipistrellus hesperus) and Nyctinomus 
mexicanus. 

BREEDING BIRDS OP SOUTHWESTERN SECTION — COLORADO RIVER DRAINAGE. 

A few of the most characteristic breeding birds are the California 
quail (Lophortyx californicus), 2 ash-throated flycatcher (Myiarclius 
cinerascens) , house finch {Carpodacus mexicanus frontalis) , Arkansas 
goldfinch (Astragalinus psaltria) , black-throated sparrow (Amphispiza 
bilineata) , western blue grosbeak (Guiraca cserulea lazula) , and canyon 
wren (Catherpes mexicanus conspersus). Practically all the Upper 
Sonoran birds found on the northwestern sage plains breed more 

1 Taken in White River Valley only. 2 Introduced. 



1911.] UPPER SONORAN ZONE. 25 

or less commonly in the desert valleys, as do also a number of species 
which are common to the Upper Sonoran areas on both sides of the 
mountains. 

PLANTS OF SOUTHWESTERN SECTION — COLORADO RIVER DRAINAGE. 

This region is especially marked by a large number of Upper 
Sonoran desert shrubs and plants, most of which extend into Col- 
orado only a short distance on the west and southwest, being pecul- 
iar to the desert areas of the southwestern United States. Important 
among these are: 

Berberis fremonti. Grayia spinosa. 

Coleogyne ramosissima. Lyciuin pallidum. 

Cowania mexicana. Opuntia uhipplei. 
Echinocactusivhippleispinosior (McElmo). Rhamnus smithi. 

Ephedra antisyphylitica. . Yucca baccata. 

Ephedra torreyana. Yucca harrimanise . 
Fraxinus anomala. 

Other plants and shrubs of more general distribution in the Col- 
orado Upper Sonoran which are found commonly in the south- 
western desert valleys are: 

Atriplex confertifolia. Opuntia camanchica. 

Atriplex canescens. Opuntia rhodantha. 

Atriplex nuttalli. Populus wislizeni. 

Dondia erecta. Populus acuminata. 

Eurotia lanata. Sarcobatus renniculatus. 

Echinocereus paucispinus. Schmaltzia trilobata. 

Lepargyrea argentea. Ximenesia exauriculata. 

REPTILES AND BATRACHIANS OP SOUTHWESTERN SECTION — COLORADO RIVER DRAINAGE. 

A large variety of reptiles, especially lizards, characterize the 
Upper Sonoran zone in the hot valleys and on the slopes of western 
Colorado, and several lizards which reach their greatest abundance 
in the Lower Sonoran zone of the southwest United States are rep- 
resented. Probably all have reached this region from the south- 
western deserts. Following are some of the most important reptiles, 
with brief notes on their distribution: 1 

Crotaplmjtus collaris baileyi. — This ring-necked lizard is abundant 
in the McElmo and Montezuma Valleys, east to a point 6 miles 
north of Cortez, at 6,500 feet; also in the San Miguel region ranging 
east to Coventry (6,500 feet) ; present in west Paradox Valley (west- 
ern rim at 6,500 feet), Salt Canyon, and other branches of the canyon 
of the lower Dolores River. Equally common on open deserts and 
in rocky pinyon country. I took two specimens at McElmo, June 
17, 1907, and one each at Coventry and Sinbad Valley, July 13 and 
17, 1907. 

'Many of the lizards of the desert valleys range up for some distance into the pinyon belt; but as no 
speeies appears to be restricted to the upper part of the zone, the present list will include notes bearing 
on both these areas. 



26 xoirril AMERICAN FAUNA. [No. 33. 

Ufa stansburiana. — This little desert lizard has a general distribu- 
tion from the Grand Valley southward, chiefly below 6,000 feet, but 
rarely to 7,000 feet (Spruce Tree Cliff Ruins, Mesa Verde). It was 
found in the following localities: Desert north of Mack; Plateau 
Creek; Coventry; Paradox Valley; Salt Canyon; Dolores River Can- 
yon; and McElmo. A very abundant lizard, usually occurring among 
locks, but often on the open desert. The specimens are from Mack, 
Plateau Creek, and De Beque, September 25 and 30 and October 1, 
1906, and McElmo, June 18, 1907. 

Ufa ornafa. — A rock-inhabiting species of delicate profile and slight 
build, often scarcely discernible on account of its peculiarly pro- 
tective colors and thin body. Present in most of the valleys from 
Grand River south to the San Juan. Represented by six specimens 
from Plateau Creek, September 30, 1906, and one each from Mesa 
Verde (Spruce Tree Cliff Ruins, 7,000 feet), June 13, and Coventry, 
July 29, 1907, and observed in Sinbad Valley and at McElmo and 
Arboles. 

Sceloporus elongatus. — The large gray scaly rock lizard has a wide 
distribution, north at least to Bear River, and ranging through the 
entire width of the Upper Sonoran zone, and also a short distance 
mto the Transition zone. In eastern Colorado noted only at Arkins. 
Found chiefly about rocks. Specimens from Arkins; Escalante Hills 
(7,000 feet); Meeker; Rangely; Plateau Creek; and McElmo. 

Sceloporus consobrinus. — A medium-sized species, inhabiting rock 
ledges in the pinyon and juniper country up to 7,000 feet. Smaller 
than the preceding species and not nearly so common. Specimens 
are from Douglas Spring (Escalante Hills); La Veta; and Arboles. 

Sceloporus graciosus. — A small graceful lizard, usually noted among 
greasewood (Atriplex and Sarcobatus) in open valleys and desert 
flats, chiefly below 5,500 feet. It was common at the following 
localities: Escalante Hills; valleys of Texas and Evacuation Creeks 
in extreme western Rio Blanco County; lower Grand Valley north 
of Mack; Rifle; Coventry; and McElmo. 

Phrynosoma ornatissimum. — Horned toads are common at McElmo 
and on the desert north of Mack, Mesa County, and are reported 
elsewhere between 5,000 and 6,000 feet. 

Cnemidophorus tigrls. — This large whip-tailed lizard ranges into 
the State only in the lowest and hottest desert valleys, extending up 
to about 5,500 feet. It is tolerably common in the lower McElmo 
Valley, and abundant in West Paradox and Sinbad Valleys and also 
hi the Dolores River Canyon as far down as the mouth of West Creek, 
frequenting sandy flats which are clothed with Atriplex and Sarco- 
batus. It doubtless occurs also in the lowest part of the Grand Valley 
near the Utah boundary, as I found it on Plateau Creek, 5 miles 
east of Tunnel, Mesa County. Represented by two specimens from 



1911.] UPPER SONORAN ZONE. 27 

Plateau Creek September 30, 1906, and one from McElmo June 18, 
1907. This species has its center of abundance in the arid Lower 
Sonoran zone. 

Cnemidophorus guIaris.—A medium-sized species not uncommon 
in the lowest valleys below 5,500 feet, in same situations as O. tigris. 
I collected specimens at Grand Valley October 2, 1906, near McElmo 
June 22, 1907, and at Hotchkiss August 8, 1907. It was observed 
also in Salt Canyon and in the canyon of Dolores River near the 
mouth of West Creek. 

Bascanion tseniatum. — This handsome and graceful snake enters 
the State, so far as known, only in the Grand River Valley, where a 
specimen was taken on Plateau Creek September 30, 1906, and another 
at Morris, west of Rifle, August 13, 1907. The Plateau Creek indi- 
vidual climbed bushes with great ease, while the one from Morris 
was discovered among beds of prickly pear on a sandy knoll. Both 
localities are below 5,300 feet. 

Thamnopliis elegans vagrans. — This garter snake was taken at 
Meeker, Rio Blanco County, August 11, 1905. 

Crotalus confluentus. — Rattlesnakes are not at all common, but are 
reported in most localities. 

Pituophis sayi. — A bull snake was taken on Dry Creek, 8 miles 
west of Naturita, July 20, 1907. 

Scaphiopus liammondi. — Numbers of these toads were caught in 
my mousetraps among beds of prickly pear on a sandy knoll at 
Morris, west of Rifle, August 15, 1907, at 5,200 feet. 

Bufo lentiginosus woodhousei. — -This toad was taken at Rangely 
September 12, 1906, and also at Rifle August 15, 1907. 

Chorophilus triseriatus. — A specimen of this little frog was collected 
at Rangely September 13, 1906, at 5,500 feet. 

RIO GRANDE DRAINAGE. 

The Upper Sonoran element which follows up the Rio Grande 
Valley and in dilute form spreads over the level expanse of San Luis 
Park is mainly characteristic of the arid valleys of western Colorado. 
It includes among mammals Antilocapra, Eutamias minimus cari/i, 1 
Citellus tridecemlineatus parvus, Onychomys Uucogaster pallescens, 
Reithrodontomys montanus, 1 Thomomys aureus pervagus, Perodipus 
montanus, 2 Perognaihus jlavus and P. apache, Sylvilagus auduboni 
warreni, and Lepus californicus texianus ( ?) . Among plants and 
shrubs the various greasewoods (Sarcobatus and Atriplex), rabbit 
brush {Chrysothamnus) , and sages {Artemisia and Eurotia) predomi- 
nate, while Gutierrezia, Grindelia, Helianthus, Ximenesia, Peritoma 
(especially P. sonorx) , . Yucca, and Opuntia are conspicuous genera 

1 Restricted to San Luis Park and Rio Grande Valley, so far as known. 

2 San Luis Park, Rio Grande Valley, and intermountain valleys of northern New Mexico. 



28 NORTH AMERICAN FAUNA. [No, 33. 

on the arid plains. The Rio Grande, Conejos, and other streams are 
fringed in many places with thickets of Ribes longiflorum and 
Schmaltzia trilobata, while willows and broad-leaved cottonwoods 
are abundant. 

JUNIPER AND PINYON BELT. 

A belt of country clothed with juniper and piny on, uniformly 
rough and broken in configuration, marks the higher part of the 
Upper Sonoran zone on the basal flanks of all the mountains in Colo- 
rado except the Front Range. 1 These two species (Pinus edulis and 
Juniperus monosperma 2 ) form the characteristic tree growth on the 
slopes between the yellow pine belt of the higher foothills and the 
open plains or desert valleys at their bases. They also densely 
clothe extensive areas partially removed from the main mountain 
ranges, comprising in western Colorado practically all the lower 
elevations and the rough country forming the watersheds between 
the valleys (see PI. IV), and east of the mountains a region of alter- 
nating canyons and ridges extending from southeastern Pueblo and 
eastern Huerfano Counties southeast to western Baca County. The 
pinyons grow to a higher elevation than the junipers, extending a 
short distance into the Transition zone. They occasionally reach 
8,500 or 9,000 feet on hot slopes, as on the western side of the Sangre 
de Cristo Range, the eastern side of the Arkansas Valley between 
Salida and Buena Vista, and the bold southwest exposures of Grand 
Mesa and Sierra Blanca. They are not found as low as the junipers, 
however, and are usually absent from the rough areas below 5,500 feet. 

Most of the mammals, birds, and plants of the juniper and piny on 
country in Colorado are more or less characteristic of this belt over 
much of the Great Basin region. Few of the species are found on 
the Great Plains, and the fauna and flora as a whole belong to the 
Great Basin division of the Upper Sonoran zone. 

MAMMALS OF JUNIPER AND PINYON BELT. 

The following mammals represent the Upper Sonoran element in 
the juniper and piny on belt in different parts of the State. On both 
slopes of the mountains: Rock squirrel (Citellus variegatus gram- 
murus), cliff mice (Peromyscus truei and P. boylei rowleyi), and gray 
fox (Urocyon cinereoargenteus scotti). In western Colorado only: 
Hopi chipmunk (Eutamias hopiensis), golden-breasted canyon mouse 
{Peromyscus crinitus auripectus), Arizona wood rat (Neoioma cinerea 
arizonae), cacomistle (Bassariscus astutus jiavus) , Great Basin spotted 

1 North of the Arkansas Divide the yellow pines of the Transition zone usually extend down to the edge 
of the plains, the juniper belt being but slightly indicated on a few outlying ridges and talus slopes by a 
very sparse growth. 

" A third, Juniperus scopulorum, is common in the upper part of this belt, but also extends up through 
the Transition zone. 



North American Fauna No. 33, U. S. Dept. Agr. Biological Survey. 



Plate IV. 




Fig. 1.— Spruce Tree Cliff Ruins, Navajo Canyon, Mesa Verde, with Juniper 
and Pinyon Forest in the Rough Canyon Country of Southwestern 
Colorado. 





**Sy^ 




1 fiib*^ 


" Lx£ "'" -•< -". .:-''«^- I — V 


•^•-.•" a -r^Kaey 










sir 


?** -v.. ^j | 


* f \ t f. - - - w ; . 


&» ■ ■ . ■ / 




} .*. X 







Fig. 2.— Navajo Canyon, Mesa Verde. 

Characteristic view in rough canyon region of southwestern Colorado, showing dense growth of 
junipers and pinyons at 7,01)0 feet. Scattering Douglas spruces are on cool slope at left. 



1911.] UPPER SONORAN ZONE. 29 

skunk (Spilogale gracilis saxatilis), and Arizona skunk (Mephitis 
estor). In the Escalante Hills: Utah chipmunk {Eutamias dorsalis 
utdhensis). In southeastern Colorado: Warren wood rat (Neotoma 
albigula warreni). 

BREEDING BIRDS OP JUNIPER AND PINYON BELT. 

Breeding birds of general distribution in this belt are the Wood- 
house jay {Aphelocoma woodhousei) , pinyon jay (Cyanoceplialus 
cyanocephalus), gray titmouse (Bseolophus inornatus griseus), lead- 
colored bush tit (Psaltriparus plumbeus), western gnat catcher 
(Polioptila cserulea obscura), and Baird wren (Thryomanes bewicki 
bairdi). The canyon towhee (Pipilo fuscus mesoleucus) , scaled quail 
(Callipepla squamata), and the road runner (Geococcyx californianus) 
are common breeders in the juniper country of southeastern Colorado, 
but the quail and road runner breed also in the open valleys of the 
region and in the Lower Sonoran zone. 

PLANTS OF JUNIPER AND PINYON BELT. 

The vegetation of the juniper belt shows a great preponderance of 
Upper Sonoran species. Conspicuous among them are: 

Juniperus monosperma. Celtis reticulata (rare). 

Juniperus utahensis (west Colorado). Ephedra antisyphylitica (west Colorado). 

Pinus edulis (also lower Transition). Ephedra torreyana (southwestern Colo- 

Fendlera rupicola . rado). 

Philadelphus micro phyllus. Echinocereus aggregatus (southeastern I !< >lo- 

Fallngia acuminata (San Luis Valley re- rado). 

gion). Echinocereus viridiflorus. 

Cowania mexica?ia (southwestern Colo- Cactus radiosus. 

rado). Opuntia arborescens (southeastern Colo- 
Peraphyllum ramossissimum (also lower rado). 

Transition). Opuntia whipplei (southwestern Colo- 
Cercocarpus parvifolius. rado). 

A triplex canescens (locally common). Yucca baccata (southern Colorado). 

Schmaltziatrilobata. Yucca harrimanize (southwestern Colo- 
Quercus (several species). rado). 

Agricultural Importance of Colorado Upper Sonoran Zone. 

Although entirely within the arid region and embracing areas of 
great aridity, the Upper Sonoran zone in Colorado is nevertheless of 
great agricultural importance. It is the only zone winch affords 
suitable physical and temperature conditions for extensive and 
varied agriculture. With a light and insufficient annual rainfall, 
which varies from 15 or 20 inches on the eastern plains to less than 
10 inches on some of the western desert tracts, and a remarkably 
dry atmosphere (average relative humidity about 50 per cent), the 
rich agricultural and fruit districts on both sides of the mountains 
have necessarily been developed largely by irrigation. It is true that 



30 NORTH AMERICAN FAUNA. [No. 33. 

a certain measure of success with corn, cane, and cereal crops is 
attained through dry farming in many sections on the eastern plains 
where no water for irrigation is available; and as the various systems 
of soil culture come into more general use and are made more effective 
in conserving rainfall, the hard and loam soil tracts should become 
increasingly productive. The great agricultural wealth of the State 
has been built up mainly, however, along the base of the foothills 
and in the valleys of the larger streams, and this area of greatest 
production will always be limited by the water supply. 1 

The distribution of Upper Sonoran crops is at present local; 
and so dependent are many of the crops upon natural protection, 
adequate water supply, and suitable soils, entirely aside from tem- 
perature, that they can not be grown over the whole of a region so 
varied as the Upper Sonoran of Colorado. The State is noted for 
its thrifty, scattered agricultural communities, some of which have 
become famous as producers of particular crops. Thus the produc- 
tion of potatoes on the plains of the Greeley region is enormous, as 
is also the yield of watermelons and cantaloupes at Rocky Ford, in 
the warm Arkansas Valley, and of peaches in the hot desert valley 
of Grand River near Grand Junction and Palisade. Sugar beets are 
a staple crop in the Platte and Arkansas Valleys, where a number of 
large sugar factories take care of the product, and they are grown 
extensively in the Grand Valley below Grand Junction. The raising 
of a great variety of vegetables on a large scale for canning is an 
important industry fostered by the establishment of several canneries 
in the region between Longmont and Fort Collins. Wheat and oats 
are important crops east of the mountains, and yield heavily under 
irrigation. Even the moderate yield now secured under dry farming 
warrants a large acreage under improved methods of handling the 
soil. Both of these cereals are successfully grown on moderate slopes 
and benches bordering many of the desert valleys in western Colorado. 
(See PL V, fig. 1.) A small acreage of corn is grown on the eastern 
plains, but it is not a sure crop west of the mountains, owing to the 
prevalence of cool nights during the growing season. Among the 
hay and forage crops alfalfa easily takes the lead, yielding two cuttings 
in the upper part of the zone, above 6,000 feet, and three or four in 
the warmer parts. 

The leading fruit districts lie at the eastern base of the foothills 
and in the desert valleys of the western slope. The region from 
Boulder north to Fort Collins is noted for apples and small fruits, 
and most varieties of strawberries, raspberries, blackberries, currants, 
plums, and cherries thrive in this section. The Florence and Canon 
City district, in the Arkansas Valley, also is a heavy producer of berry 

1 Vitally essential to the agricultural interests of the State are the mountain forests. Many authorities 
maintain that they induce precipitation; they also conserve moisture by holding the snows of winter, thus 
insuring a regular supply of water in the streams well into the growing season. 



North American Fauna No. 33, U. S. Dept. Agr. Biological Survey. 



Plate V. 







Fig. 1.— Farms in the Lower White River Valley 
Below Rangely, at 5,500 Feet. 




Fig. 2.— Fruit Ranches in the McElmo Valley Between Moqui and McElmo, 

at 5,200 Feet. 



1911.] UPPEE SONORAN ZONE. 31 

fruits and apples. The western desert valleys are peculiarly adapted 
to a groat variety of Upper Sonoran fruits, as apples, pears, peaches, 
prunes, apricots, and nectarines, while even almonds and peanuts 
have been successfully raised in a few of the warmer localities. (See 
PI. V, fig. 2.) Many hundred carloads of peaches, apples, and pears 
are shipped annually from the lower Grand, Gunnison, and Uncom- 
pahgre Valleys, the chief shipping points being Fruita, Grand Junc- 
tion, Palisade, Delta, Hotchkiss, Paonia, and Montrose. The adobe 
soils which prevail in most of the desert tracts prove wonderfully 
productive under irrigation, and the transformation wrought in a 
few years by water is little short of marvelous. 

At present most of the eastern plains region and the sage plains 
of Routt County are used for grazing, but all the hard and loam soil 
tracts will doubtless eventually be brought under cultivation. The 
extensive sandy areas, with their luxuriant growth of range grasses, 
are well suited to grazing, but are also especially adapted to certain 
crops. 1 The short-grass plains which form the watershed between 
the Arkansas and Cimarron Rivers, in the extreme southeast, are 
arable throughout, and are being gradually reclaimed by dry-farming 
methods. 

The intimate connection between the natural life zones and the 
crop zones has already been referred to. The same isotherms which 
limit the upward or northward dispersion of certain associations of 
native plants, birds, and mammals are found to limit effectively the 
successful production of certain fruits and other crops. The varie- 
ties of fruits, cereals, and miscellaneous crops which have proved 
best adapted to the Upper Sonoran zone in various parts of the 
western United States have been listed by Merriam. 2 In this zone 
the standard varieties of apples, plums, and cherries, the sugar beet, 
most of the cereals, and many other fruits and vegetables, reach 
their highest development and productiveness. It is not likely that 
all the varieties listed will prove an unqualified success in the Upper 
Sonoran of Colorado, but the tables furnish valuable aid in selecting 
suitable varieties. Presupposing suitable soil and water conditions, 
any intelligent fruit grower or farmer with a fair knowledge of the 
distribution and character of the Upper Sonoran element should be 
reasonably certain whether the proper temperature conditions for 
the maturing of particular fruits and crops are present in a given 
locality. Such knowledge is of special value along the upper edge 
of the zone, where, among the lower foothills and in regions of broken 
configuration, agriculture and horticulture, following the line of 
least resistance, are now very largely confined to the stream valleys. 
These are frequently descending tongues of the Transition zone, and 

1 Natural Vegetation as an Indicator of the Capabilities of Land for Crop Production, Bureau of Plant 
Industry Bull. 201, pp. 75-78, 1911. 

2 Life Zones and Crop Zones, Bull. 10, Biological Survey, pp. 37-40. 



32 NORTH AMERICAN FAUNA. [No. 33. 

the cooling influence of the streams very often causes the entire 
valley bottoms to be filled with Transition vegetation. From the 
nature of the case most of the Upper Sonoran fruits and crops tested 
in these valleys have proved either a total failure or only a partial 
success, and after long experimentation have been replaced by the 
hardier varieties of the Transition zone. Many of the experiments 
might have succeeded had they been tried in the warm pockets or on 
the moderately inclined slopes along the warm sides of the valleys, 
and had the rainfall been utilized by cross furrowing and ditching. 
These warm exposures are covered with Upper Sonoran vegetation, 
are occasionally open, but more often clothed with scattering pinyons 
and junipers, and in many cases, with a little clearing of rocks and 
shrubbery, are quite capable of cultivation. Many ranchmen have 
already discovered the advantages of such locations for fruit raising, 
but along the bases of the mountains are considerable areas of such 
land, still unreclaimed, which under present methods of conserving 
and handling the water supply are capable of producing the finest 
Upper Sonoran fruits. 

Although the larger part of the open San Luis and Rio Grande 
Valleys is included in the Upper Sonoran on the zone map, this region 
is very nearly on the border line between the Upper Sonoran and 
Transition zones, with a slight preponderance of Upper Sonoran 
species. This area is generally considered too cold for Upper Sonoran 
crops and fruits, and few have been grown thus far. This is not 
strange, however, since the part reclaimed for agriculture is very 
largely in valleys of streams, whose cooling effect precludes the suc- 
cessful cultivation of any but the hardy vegetables and cereals of the 
Transition zone. Warmer conditions are indicated on all the bor- 
dering foothill slopes by a belt of junipers, pinyons, and other Upper 
Sonoran vegetation, which extends from 1,000 to 1,500 feet above the 
level of the valley and continues out into the open country for some 
distance along the the warm sides of ridges. The warmest and most 
protected slopes are at the base of the foothills, and many of 
them are open and, with little clearing, suitable for cultivation. 
That certain of the hardier Upper Sonoran fruits, particularly apples, 1 
can be successfully grown on these warm, protected slopes, wherever 
water is obtainable, seems highly probable, and the advantage of 
fruit growing in an extensive isolated agricultural region like the 
San Luis Valley is apparent. 

An advantage of fruit raising in the foothills is the protection they 
afford from the cold winds which sweep the plains in winter. An 

1 In various parts of the southwest the pinyon belt has proved especially adapted to apples. Perhaps 
the best example within the State is in Montezuma Valley and the neighboring country, where in the 
natural openings and clearings among the pinyons the very finest apples are grown. When water from 
the several large ditches now under construction is available, apple growing will become one of the most 
important industries of southwestern Colorado. 



1911.] TEANSITION ZONE. 33 

added advantage is that, owing to the altitudinal difference in the 
progress of the season, fruit in the foothill districts often escapes 
injury from a late spring frost which catches the fruit of the lower 
country in full blossom. This was notably the case in 1907, when 
late frosts were general over the fruit districts on both sides of the 
mountains, greatly diminishing the crop. 

The protection afforded in the foothill valleys of western Colorado 
is especially favorable to peach growing. The lower edge of the 
pinyon belt appears to limit the successful growing of peaches, nec- 
tarines, tomatoes, and melons over most of the region, but under 
favorable conditions peaches have been grown somewhat higher. In 
eastern Colorado peach growing is not carried on to any great extent. 
At Canon City the average is about two crops every five years. This 
failure is not due to coldness of climate, for the mean temperature is 
unusually high, but to long spells of warm weather in winter, which 
cause the buds to start, whereupon a sudden cold snap freezes them. 

The many long arms or extensions of the Upper Sonoran zone in 
western Colorado have a special value in that, deeply penetrating 
the high country of the Transition and Boreal zones, they enable the 
cultivation of Sonoran fruits and crops to be carried far within 
regions devoted to mining and stock raising, and thus render their 
production more lucrative than elsewhere. An excellent example is 
the narrow, semidesert Grand Kiver Valley, between Grand Junction 
and Glenwood Springs. 

With the completion of the Gunnison Tunnel (through which 
water from the Gunnison River is to reach 150,000 acres of desert 
land in the Uncompahgre Valley) and many other private irriga- 
tion projects of less scope, the ranchmen of western Colorado in 
particular are rapidly awakening to the great possibilities of irrigation 
and to the fact that fruit raising is far more remunerative, acre for 
acre, than ranching and hay raising. 

TRANSITION ZONE. 

In general this may be said to be the foothill zone of Colorado, 
with its lower limit marked by the edge of the plains on the east and 
by the approach to desert conditions along the western bases of the 
mountains and plateaus. It is a neutral distribution area of consid- 
erable breadth lying between the Boreal (Canadian) and Austral 
(Upper Sonoran) regions, and elements of both zones enter about 
equally into its composition. Although a number of species of mam- 
mals, birds, and plants are wholly or mainly restricted to the Transi- 
tion zone and characterize it locally, it is best marked on the whole as 
a region of overlapping boreal and austral species. 1 

1 See Merriam, Proc. Biol. Soc. Wash., VII, p. 31, 1892. 
90432°— 11 3 



34 NORTH AMERICAN FAUNA. [No. 33. 

In addition to occupying continuous belts in the broken foothill 
country which flanks both sides of the main ranges, the Transition 
zone covers large areas in North Park, the upper Bear River Valley, 
the Wet Mountain Valley, Archuleta County south of the San Juan 
Mountains, on the Arkansas Divide with the adjacent region border- 
ing the South Platte, and elsewhere. Many outlying elevations in 
west cin Colorado, chiefly secondary plateaus descending toward the 
Colorado River, are capped with this zone. Among these are the 
Mesa Verde, in Montezuma County, and the Yampa Plateau and 
Escalante Hills in western Routt County. It covers also most of the 
mountains west of the Ladore Canyon of Green River, forms a narrow 
belt around the Canadian zone summit of the O-wi-yu-kuts Plateau, 
occupies the summits of the higher eastward projecting flanks of the 
La Sal Mountains in western Montrose County, and is present in 
dilute form (indicated by pockets of Pseudotsuga) on the upper north- 
east faces of the Rabbit Hills, in western Rio Blanco County. On 
the east side of the mountains the only Transition zone island of 
consequence is on the summit of the Mesa de Maya, in southern Las 
Animas County. The total area in Colorado covered by the Tran- 
sition zone is in the neighborhood of 25,000 square miles, or approxi- 
mately one-fourth of the area of the State. 

Warm slopes in various parts of the State having an unusual 
exposure to the direct rays of the sun carry Transition species any- 
where from 500 to 1,000 feet above the mean zone level. A few 
examples are a southwest slope 1 mile northeast of Lake San Cris- 
tobal, near Lake City, where scattering yellow pines are encountered 
up to 10,000 feet; at Bath, on the summit of Trout Creek Pass, and 
along the western slopes of the Trout Pass Hills, where the same tree 
occurs regularly up to 9,500 feet and sparingly among the Douglas 
firs for another 300 feet; and on the eastern slope of the Sangre de 
Cristo Range in the Mosca Pass region, where the pines are common 
at 9,500 feet. Transition zone vegetation is carried to an abnormal 
elevation also on the remarkably exposed southwest slopes of Sierra 
Blanca and Grand Mesa, and on Badito Peak, the southernmost point 
of the Wet Mountains. (See PL VI, fig. 1.) 

An excellent illustration of warmer environmental conditions 
created through deforestation, and a consequent upward extension of 
Transition zone species, is afforded on the mountains north of and 
adjacent to Clear Creek, just east of Fall River. The southwest side 
of a mountain between Fall River and Russell Gulch is clothed to 
the summit, at 9,000 feet, with a dwarfed growth of Pinus scopulorum, 
intermixed with Pseudotsuga mucronata and Pinus Jlexilis, and such 
shrubs as Acer glabrum and Cercocarpus parvifolius. The upper 500 
feet of this slope is said to have been forested with Pinus murrayana 
in early days before mining activities exhausted the best forests of 



North American Fauna No. 33, U. S. Dept. Agr. Biological Survey. 



Plate VI. 




Fig. 1.— Effect of Slope Exposure. 

Yellow pines (Pinus scopulorum) and firs (Abies concolor) growing at same elevation (9,000 feet) 
on hot and cold slopes in Wet Mountains east of Westcliffe. 










Fig. 2.— Characteristic View at Upper Edge of 
Transition Zone. 

Undergrowth of small aspens (Popidus tremuloides) in forest 
of yellow pine (Pinus scoptdorum) on Dolores Plateau. 



1911.] TRANSITION ZONE. 35 

the region, and many stumps still remain. At present the summit 
and upper northern slopes of this mountain are clothed with a young 
and dense growth of lodgepole pines, but they apparently can not 
recover a foothold on the upper southwest slope, where they formerly 
abounded. At no other point in northern Colorado was Pinus 
scopulorum observed above 8,500 feet, and it reaches above 8.000 
feet in only a few places. (PI. VI, fig. 2.) 

The Transition zone varies greatly over the State, not only in 
respect to its vertical and horizontal breadth, but also in its charac- 
terization, being sometimes strongly marked and again dilute and 
ill-defined. It is most uniform in width along the eastern flanks of 
the front ranges, where the chief irregularities are due to the eastward 
extending Arkansas Divide and the deeply penetrating valley of the 
Arkansas River. On the western slope of the Continental Divide the 
Transition belt is extremely irregular in outline, because of the 
many plateaus which project westward from the main ranges and the 
intervening desert valleys. It completely fills the upper valleys and 
skirts the lower valleys of the Snake, Bear, White, Grand, Gunnison, 
Dolores, and San Juan Rivers, and their affluents. Two independent 
arms of the Transition zone enter the State along its northern bound- 
ary in the intermountain region — one along the North Platte, a 
narrow tongue which widens to include practically all the extensive 
sage plains of North Park at an average elevation of about 8,000 feet; 
the other, a narrow strip of yellow pine and sage country in the 
Laramie River region. Two belts of this zone likewise penetrate the 
intermountain region of Colorado from the south, occupy the foothills 
of the San Juan and Culebra Ranges, skirt the great San Luis Valley, 
and connect with the upper Arkansas Transition through a narrow 
tongue across Poncha Pass. 

As already observed, the Transition zone varies greatly in breadth. 
The variation in horizontal breadth, due entirely to difference in 
slope incline, is well illustrated by the following examples: On steep 
slopes such as are found north of Grand River near Glenwood Springs, 
and between De Beque and Rifle, the aspens of the high Canadian 
zone country are often less than half a mile from the junipers of the 
warm Upper Sonoran valley; while on the western side of the Elk 
Head Mountains, where the slope is very gradual, the Transition zone 
merges almost insensibly into the Upper Sonoran sage plains, and has 
a width of 8 or 10 miles or even more. Again, in a few instances, the 
Transition element is almost lost or crowded out by an unusual 
upward extension of Upper Sonoran zone. For example, on the 
north side of the Grand River Valley at McCoy, Eagle County, 
pinyons and junipers (Juniperus monosperma) follow up the warm 
south slopes to 7,800 feet, while lodgepole pines and aspens clothe the 
summits and northern slopes at 8,400 feet. The intervening 600 feet, 



36 NORTH AMERICAN FAUNA. [No. 33. 

or upper part of the southern slopes, supports a scattering growth of 
Pinus scopulorum, Pseudotsuga mucronata, Amelancliier alnifolia, 
Symphoricarpos oreophilus, and other Transition species. 

Narrow extensions or arms of the Transition zone penetrate the 
mountains for a considerable distance on both. sides of the main 
ranges, following along the warm north sides of the valleys. Under 
such conditions the upper limit of the yellow pines is very materially 
lowered where they penetrate an extensive region of high altitude — 
that is, the highest elevation at which the pines grow is considerably 
less at the head of a valley where the surrounding environment is 
uniformly boreal than on the slopes and ridges near the mouth of the 
valley, where the bordering high country is much broken and incised 
and thus permits a more general diffusion of warm air currents. This 
difference in the upward extension of the pines amounts to nearly 500 
feet in the Fountain Creek Valley. In the region about Manitou the 
upper limit on warm slopes is approximately 9,000 feet, while at 
Woodland Park, in the upper valley, it is about 8,500 feet. 

MAMMALS OF COLORADO TRANSITION. 

Both boreal and austral mammals are represented in the Transition 
zone in Colorado, while only six species appear to be restricted to 
it. These are two squirrels of the Sciurus aberti group, three 
pocket gophers of the genus Thomomys, and a small brown bat, 
Myotis lucifugus longicrus. Sciurus aberti ferreus inhabits the east- 
ern foothills of the front ranges from Loveland and Arkins south 
to the Mosca Pass region, and Sciurus aberti mimus occurs in the 
stately yellow pine forests on the southern slopes of the San Juan and 
La Plata Mountains. Thomomys clusius is found on the sage plains 
of North Park, on the high grassy plateau in northeastern Weld 
County, and on the western (higher) end of the Arkansas Divide. 
Thomomys talpoides agrestis, an isolated mountain form of the 
common Thomomys talpoides of the northern plains, is known only 
from the San Luis Valley meadows. Another gopher, Thomomys 
fulvus, appears to enter the State only on the Raton Mesa, in the 
Trinidad region. Myotis I. longicrus is common in the Transition 
zone at several points in western Colorado, and appears to be a 
characteristic zone species. 

The following mammals, while not entirely restricted to the Transi- 
tion zone, have their center of abundance there, and characterize it 
locally in various parts of the State: The Say chipmunk (Eutamias 
quadrivittatus) , in the foothills of the eastern and southern moun- 
tains; Wyoming ground squirrel (Citellus elegans), on the high sage 
plains of North Park and in northwestern Colorado; Estes Park cliff 
mouse (Peromyscus nasutus) and Gale wood rat (Neotoma fallax) , in 
the eastern foothills of the front ranges; pygmy vole (Microtus 






1911.] TRANSITION ZONE. 37 

pauperrimus) , on the higher sage plains of North and Middle Parks 
and in Routt County; white-tailed jack rabbit (Lepus campestris), 
east of the Continental Divide, from North Park south to the Arkansas 
Divide, and in San Luis Valley; Townsend jack rabbit (Lepus campes- 
tris townsendi), of the higher open country west of the Continental 
Divide; Rocky Mountain cottontail (Sylvilagus nuttalU pinetis), 
mountain coyote (Cards Testes), and mountain wildcat (Lynx uinta), 
found throughout the mountainous districts; and the northern plains 
skunk (Mephitis hudsonica), which occurs on both slopes of the 
northern mountains. 

The following mammals of wider zonal range have been found in 
this zone in various parts of the State: 

Mammals Common to Transition and Upper Sonoran. 

Antilocapra americana. Castor canadensis fro ndutor. 

Odocoileus virginianus macrourus. Canis occidental-is. 

Onychomys brevicaudus. Spilogale tenuis. 

Eptesieus fuscus. 

Mammals Common to Transition and Canadian. 

Cervus canadensis. Microtus nanus. 

Callos pernio philus lateralis. Microtus pennsylvanicus modestus. 

Eutamias amoenus operarius. Zapus princeps. 

Eutamias minimus consobrinus. Vulpes macrourus. 

Ereihizon epixanthum. Ursus americanus. 

Mammals Common to Transition, Canadian, and Upper Sonoran. 

Odocoileus hemionus. Felis oregonensis hippolestes. 

Cynomys gunnisoni. Lutreola vison energumenos. 

Cynomys leucurus. Putorius arizonensis. 

Taxidea taxus. 

BREEDING BIRDS OF COLORADO TRANSITION. 

Birds which breed chiefly in the Transition zone in Colorado are 
the sage hen (Centrocercus urophasianus) / saw- whet owl (Crypto- 
glaux acadica), sharp-shinned hawk (Accipiter velox), Rocky Mountain 
hairy woodpecker (Dryobates villosus monticola), Lewis woodpecker 
(Asyndesmus lewisi), Wright flycatcher (Empidonax wrighti), spurred 
towhee (Pipilo maculatus montanus), green-tailed towhee (Oreospiza 
cJdorura), white- throated swift (Aeronautes melanoleucus) , plumbeous 
vireo (L,anivireo solitarius plumbeus) , Macgilhvray warbler (Oporomis 
tolmiei), Rocky Mountain nuthatch (Sitta carolinensis nelsoni), p}'gmy 
nuthatch (Sitta pygmsea), and chestnut-backed bluebird (Sialia 
mexicana bairdi). 2 Among the birds occupying restricted areas in the 
Colorado Transition during the breeding season may be mentioned 

1 Mainly in northwestern Colorado. 2 Southern Colorado. 



38 



NORTH AMERICAN FAUNA. 



[No. 33. 



the Grace warbler (Dendroica gracise) , a common breeder in the yellow- 
pine forests on the southern slopes of the San Juan Mountains. 

Other birds conspicuous during; the breeding season in the Transition 
belt in different parts of the State are: 



Bubo virginianus palhscens. 
Calamospiza melanocorys. 
Colaptes cafer collaris. 
Dendragapus obscurus. 
Dendroica auduboni. 
Euphagus cyan ocephalus. 
Falco sparverius phalxna. 
Myiochanes richardsoni. 
Oreoscoptes montanus. 
Otocoris alpestris leucolsema. 
Passerculus sandwichensis alaudinus. 
Penthestes gambeli. 
Phalxnoptilus nuttalli. 



Pica pica hudsonia. 

Planesticus migratorius propinquus. 

Pocecetes gramineus confinis. 

Salpinctes obsoletus. 

Sayornis saya. 

Sialia currucoides. 

Spizella breweri. 

Spizella passerina arizonse. 

Sturnella ncglecta. 

Tachycineta thalassina lepida. 

Troglodytes aedon parkmani. 

Vireosylva gilva swainsoni. 

Xanthocephalus xanthocephalus. 



Several genera of the Anatidas and at least one genus of the Linii- 
cohie are found on Lake John, North Park, during the summer, and 
doubtless breed there. These include Branta canadensis, Anas 
platyrhynchos, Spatula clypeata, Querquedula discors, and Steganopus 
tricolor. Other breeders in the same region are the coot (Fulica 
americana) and the great blue heron (Ardea lieroclias). 



PLANTS OF COLORADO TRANSITION. 

In Colorado, as in most sections of the western United States, the 
yellow pine (Pinus scopulorum) (see PI. VII, fig. 2) is the character- 
istic Transition tree, and the zone is practically coextensive with these 
pines wherever they occur. Their distribution in the Colorado foot- 
hill§ is quite general, except in the northwest, where a very sparse 
and scattering growth occurs in a few widely separated localities. 
Over much of western Colorado north of the San Juan Mountains 
and Uncompahgre Plateau the Transition zone is a partially open 
region, sage-covered slopes and parks alternating with brushy slopes 
and ridges of chaparral, including such shrubs as oaks (Quercus gam- 
oeli (PI. VII, fig. 1), Q. fendleri, and others), chokecherry (Prunus 
melanocarpa) , Juneberries (AmelancJiier alnifolia, A. baJceri, and 
others), and Ceanothus velutinus, PerapJiyTlum ramosissimum , Sym- 
phoricarpos oreophilus, Kunzia tridentata, Cercocarpus parvifolius, 
and in the Escalante Hills Cercocarpus ledifolius. The most promi- 
nent forest tree in the Transition belt of northwestern Colorado is the 
Douglas spruce (Pseudotsuga mucronata) , which occurs chiefly on 
sharp declivities and on the exposed crests of the plateaus above 
7,000 feet. 



North American Fauna No. 33, U. S. Dept. Agr. Biological Survey. 



Plate VII. 




Fig. 1.— Dense Chaparral of Oak (Quercus gam- 
beli) in Western Foothills of West Elk Moun- 
tains, on Head of Smith Fork of Gunnison 
River, Altitude 8,000 Feet. 




Fig. 2.— Yellow Pines iPinus scopulorumi near 
Elkhorn, Larimer County, Showing Charac- 
teristic Scattering Growth of the Species 
on Eastern Foothills of Front Range. 



1911.] TRANSITION ZONE. 39 

The vegetation of the sage plains of North Park, in addition to 
Artemisia tridentata, consists chiefly of Sarcobatus vermiculatus and 
Chrysothamnus, with Kunzia tridentata added in sandy strips of 
country. 

Along streams throughout the State this zone is best indicated by 
the narrow-leaved cottonwood (Populus angustifolia) . Some of the 
shrubs found commonly in the neighborhood of streams are the 
alder (Alnus tenuifolia) , Rocky Mountain birch (Betula fontinalis) , 
dogwood (Svida stolonifera riparid), hazel (Corylus rostrata), haws 
{Crataegus soligna, C. wheeUri, and others), willow (Salix nuttalli), 
Distegia involucrata, 1 and Opulaster ramaleyi. Another willow 
(Salix perrostrata) grows in dense clumps in bogs and around spring 
holes. A large variety of shrubs and plants are characteristic of the 
rocky slopes, among which are several species of June berry (Amelan- 
chier), mountain holly (Cercocarpus parvifolius) , 2 ninebark (Opulaster 
monogynus), flowering raspberry (Oreobatus deliciosus) , New Jersey 
tea (Ceanothus pubescens) , Oregon grape (Berberis aquifolium), bear- 
berry (Arctostaphylos uvaursi), currants (Ribes inebrians and R. 
pumilum), Kunzia tridentata, Edwinia americana, and Holodiscus 
dumosus. A large species of bearberry (Arctostaphylos pungens 
platyphylla) forms a dense chaparral on the western slopes of the 
Uncompahgre Plateau and on the opposite (eastern) slopes of the 
La Sal Mountains, between 8,000 and 8,500 feet. The only cactuses 
found with any regularity in the Colorado Transition are the little 
Opuntia fragilis, which is common in the yellow pine forests near 
Pagosa Springs and on the Uncompahgre Plateau, and was found 
also along the North Fork of the Gunnison River above Somerset; 
the unique snake cactus (Echinocactus simpsoni), most abundant in 
the Wet Mountain Valley and adjacent region, so called because of 
the peculiar snake-like growth occasionally formed; and Echinocereus 
viridijlorus, also of the Wet Mountain region. Cactus missouriensis 
occurs occasionally in the yellow pine belt of the eastern foothills. 
Chrysothamnus elegans, C. bigelovi, and several other species of 
rabbit brush are common and characteristic shrubs in the Wet 
Mountain Valley and in all the open valleys and parks between 
7,000 and 9,000 feet. 

REPTILES AND BATRACHIANS OF COLORADO TRANSITION. 

A few reptiles occur with more or less regularity in the Transition 
zone, but it is doubtful whether any are restricted to it. Among the 
lizards, the large gray rock lizard (Sceloporus elongatus) 3 is found in 
rocky situations along the lower edge of the zone in many parts of 
western and southwestern Colorado. One of the whip-tailed lizards, 

1 Also lower Canadian. ■ Also Upper Sonoran. 3 Most abundant in Upper Sonoran zone. 



40 NORTH AMERICAN FAUNA. [No. 33 

(Cnemidophorus gularis) 1 seen in the foothills near Golden in June, 
1905, was in a rank growth of grass on a yellow pine slope, at 6,500 
feet. Horned toads (Phrynosoma ornatissimum) are tolerably com- 
mon in the sandy yellow pine country on the northern end of the 
Uncompahgre Plateau. In July, 1907, several were seen on the head 
of Dominguez Creek, at 8,500 feet, and one individual was collected. 
Another, not over a third grown, was taken in late September, 1906, 
at the same elevation on the chaparral-covered summit of the Book 
Cliffs at Baxter Pass near the Utah boundary. A garter snake 
( TJiamnophis elegans vagrans) 2 is not uncommon in the pine belt and 
is found also in the lake region of North Park. A specimen taken at 
Higho, North Park, in August, 1906, was crawling through the sage- 
brush at some distance from water. Another was collected along 
Snake River east of Slater, Routt County, in August, 1906, at a little 
over 7,000 feet. The little green snake (Liopeltis vernalis) was taken 
but once, on the bank of the Rio Pifios, a few miles below Vallecito, 
June 6, 1907. A small frog (Chorophilus triseriatus) was secured on 
Snake River, about 10 miles east of Slater, August 21, 1906. Edward 
A. Preble collected the larva of a salamander {Amby stoma tigrinum) 
in Estes Park in 1895. 

AGRICULTURAL IMPORTANCE OF THE TRANSITION ZONE. 

The rough and broken character of much of the Transition area in 
Colorado precludes agriculture on an extensive scale, although the 
climatic conditions are favorable for many of the hardier vegetables, 
cereals, and fruits. The most important of its natural resources are 
the yellow pine forests, which in some localities are very extensive. 
The few areas of any size which are sufficiently open for cultivation 
are North Park, the Wet Mountain Valley, San Luis Valley, and the 
western (higher) end of the Arkansas Divide, together with certain 
of the larger foothill valleys. Of these, the first two are devoted 
largely to ranching and hay raising, the Arkansas Divide to grazing, 
and the San Luis Valley alone to extensive agriculture. Wheat, 
oats, and rye are the leading cereals raised in the Colorado Transition, 
while timothy is an important hay crop. A great deal of alfalfa is 
grown along the lower edge of the zone, where it usually yields two 
cuttings. Higher up it is not a success. Potatoes and Canada field 
peas are important crops in the San Luis Valley. A great variety of 
vegetables are grown in the San Luis Valley, and also in the stream 
valleys of the higher foothills over the State, where a good market is 
furnished by the neighboring mining camps. Comparatively little 
fruit is raised, but with a proper selection of hardy varieties horticul- 
ture might be made an important industry in the protected foothill 

i Most abundant in Upper Sonoran zone. 2 Also in lower Canadian zone. 



North American Fauna No. 33, U. S. Dept. Agr. Biological Survey. 



Plate VIII. 




Fig. 1.— Lower Part of Canadian Zone Forest of 
Lodgepole Pines (Pinus murrayana) and Aspens 
(Populus tremuloides) on North Park Slope 
of Medicine Bow Range. 




.•:'-< • : it 




Fig. 2.— Forest of Engelmann Spruce (Picea engelmannp in Upper Part of 
Canadian Zone, on Summit of Park Range, at 10,000 Feet. 



1911.] 



CANADIAN ZONE. 41 



valleys wherever water is available for irrigation. A variety of berry 
fruits, such as June berries, strawberries, raspberries, currants, and 
gooseberries, are native to the Transition zone in Colorado. 

The Transition zone is suited to the growth of practically all the 
vegetables, including the cabbage, lettuce, turnip, radish, potato, beet 
(both table and sugar), pea, bean, onion, carrot, parsnip, and early 
sweet corn; all hardy cereals; and the hardier varieties of apples, 
cherries, and small fruits. The crab apple reaches its best develop- 
ment in this zone. 1 

CANADIAN ZONE. 

This, the more extensive of the boreal zones, occupies the middle 
slopes on the main ranges and extensive areas in the mountain parks, 
and caps all of the higher western plateaus, thus including the larger 
part of the coniferous forests of the State. Broadly speaking, the 
Canadian zone is characterized in the mountains of Colorado by 
extensive forest belts of aspens (Populus tremuloides), lodgepole pines 
(Pinus murrayana), and the lower, heaviest part of the Engelmann 
spruce belt (see PI. VIII, fig. 2). White firs (Abies concolor) are added 
to these in the southern mountains. On the western plateaus it is 
marked by either a mixed forest of aspens and Engelmann spruces, 
or else a partially open country of grassy parks and aspen groves. 
In the park region of central Colorado considerable areas of open 
grass land are in this zone. 

Chief among the elevated areas in western Colorado extensively 
capped with Canadian zone are the White River, Book, and Uncom- 
pahgre Plateaus; Grand, Battlement, Lone, and Blue Mesas; the 
Danforth and Huntsmans Hills; and the Cathedral Bluffs. The 
Gore, Elk Head, Rabbit Ear, Williams River, and West Elk Moun- 
tains are largely or wholly Canadian, while perhaps the largest areas 
of this zone are found in the Middle, Egeria, and South Park regions. 
Small Canadian zone islands cap the summits and upper northern 
slopes of many small outlying peaks and plateaus, especially in western 
and southern Colorado. Among these may be mentioned Diamond 
and Zenobia Peaks, Mount Cullom, and Yampa and O-wi-yu-kuts 
Plateaus, all in extreme western Routt County; Ute Peak, the 
highest point in the Sierra el Late, in Montezuma County; isolated 
table mountains south of the San Miguel and San Juan Mountains, 
and peaks between the upper forks of the South Platte; and Raton 
Mesa, southeast of Trinidad. 

In the mountains of the northern half of the State the lodgepole 
pine and aspen forests are regularly entered at between 7,500 and 
8,000 feet, and continue up to a little over 10,000 feet. (See PI. 

1 Some of the varieties of cereals and fruits which have proved adapted to the Transition zone (arid sub- 
division) in Idaho, eastern Washington, and Utah, are listed by Merriam in Life Zones and Crop Zones, 
Bull. No. 10, Biological Survey, pp. 25-27, 1898. 



42 NORTH AMERICAN FAUNA. [No. 33 

VI II, fig. 1.) Passing south to the latitude of Colorado Springs, the 
Canadian zone level is raised 500 to 1,000 feet, and continues high 
over most of the southern mountains, with the vertical breadth about 
as in the north. The vertical boundaries are subject to great local 
variation, according to physiographic conditions, entirely apart 
from the regular elevation and depression due to slope exposure. 
Striking examples are the upper Arkansas Valley and Gunnison 
regions, on opposite sides of the Continental Divide. On the moun- 
tain slopes bordering the upper Arkansas from the Royal Gorge to 
Buena Vista and beyond, Canadian zone species are seldom encoun- 
tered much below 8,500 or 9,000 feet, even on cold exposures, and 
along the east side of the valley in the vicinity of Trout Creek Pass 
they are crowded on warm slopes above 9,500 feet. Over much of 
the upper Gunnison country lodgepole pines and aspens grow regu- 
larly as low as 8,000 feet on cold exposures and 8,500 feet on warmer 
slopes. The comparatively low elevation reached in the Gunnison 
country is undoubtedly due to the influence of the great mass of 
boreal country which practically surrounds the region. The unusu- 
ally high temperature and consequent high zone levels which pre- 
vail in the upper Arkansas region are probably due to several factors. 
The southward trend of this narrow valley admits the hottest rays 
of the sun, which shine with great directness upon very abrupt 
bordering slopes; while the region to the east and south is neither 
so high nor so extensively boreal as that which surrounds the Gun- 
nison country. The western slopes of the Sangre de Cristo Range 
and the eastern slopes of the Cochetopa Hills and Garita Mountains, 
bordering the northern two-thirds of the San Luis Valley, and the 
eastern slopes of the Sangre de Cristo Range near Mosca and Sand 
Hill Passes, are other notable examples of abnormally high zone levels, 
where on warm slopes the Canadian element is often forced above 
9,500 feet. This is probably the result of the high, 7,500-foot base 
level of San Luis Valley. South from Mosca Pass along the eastern 
slope of the range the lower boundary of the Canadian zone gradually 
drops to about 8,000 feet, which level is uniformly maintained on the 
slopes of the high Culebra Range west of Trinidad. The low 6,000- 
foot base level of the Trinidad plains would account for low zone 
levels on the Culebras. 

The effect of slope exposure on zone level is well shown along the 
lower edge of the Canadian zone, both on individual slopes and on 
opposite sides of mountain ranges, especially on those which rise 
abruptly with few flanking foothills. On the high narrow Sangre de 
Cristo Range the difference in mean elevation between the east and 
west slopes amounts to fully 500 feet, and it is nearly as great on 
some of the other ranges. Throughout the mountains, cool shaded 
north and northeast slopes and gulches carry the aspens, lodgepole 



North American Fauna No. 33, U. S. Dept. Agr. Biological Survey. 



Plate IX. 




Fig. 1.— Summer Cattle Range, Canadian Zone 
Meadow East of Laramie River, 10,000 Feet 
Altitude. 




Fig. 2 —Canadian Zone Vegetation at 9,000 Feet 
on Open Summit of Uncompahgre Plateau- 
Frasera, Delphinium, Geranium, and Lupinus. 



1911.] CANADIAN ZONE. 43 

pines, firs, and other Canadian zone vegetation to a low elevation. 
Some of the cold streams on the eastern slope of the Front Range 
carry quite pronounced tongues of Canadian zone as low as 7,000 feet. 
These consist of thickets of aspens, a scattering fringe of blue spruces 
(Picea parryana), and occasionally an Engelmann spruce. These 
descending boreal tongues are invariably embraced by warm slopes 
clothed in different parts of the State with yellow pines, oaks, sage- 
brush, and other Transition zone vegetation. At Honnold, Routt 
County, lodgepole pines and dense thickets of scrubby aspens crowd 
down the steep northern face? of the Elk Head Mountains to the 
bank of Snake River, at 7,000 feet, alternating with ascending 
tongues of sagebrush (Artemisia tridentata) on all the warm Transi- 
tion slopes up to 7,500 or even 8,000 feet. 

The boreal forest belts are best shown on sharply inclined slopes, 
like the steep north side of the Clear Creek Valley between Silver 
Plume and Graymont, where in summer a beautiful tricolored 
appearance is presented, due to the three shades of green of the 
aspen, lodgepole pine, and Engelmann spruce forest. The slopes 
up to 10,500 feet are densely clothed with the first two trees in about 
equal abundance, and present a patched appearance, thickets and 
ascending tongues of the light-green aspens being scattered through 
the more uniformly distributed, dark, yellowish-green pines. (See PI. 
VIII, fig. 1.) Joining the upper edge of this forest is a regular and 
well-marked belt of greenish-black Engelmann spruces, while still 
higher up the slope may be seen the straggling growth of foxtail pines 
(Pinus aristata) of the timberline region — of the same hue as the 
Engelmann spruces, but distinguishable by their ragged appearance. 

In Colorado extensive forests cover a very large percentage of the 
total area occupied by the Canadian zone, which is poorly adapted to 
crops. The only open land at all suited to agriculture is found along 
the lower edge, between 8,000 and 9,000 feet, and comprises narrow 
strips along the streams in the larger mountain valleys and parks 
and on the western plateaus. At present this agricultural land is 
largely utilized for wild hay and timothy. As most of the park and 
plateau region supports a luxuriant growth of grasses, affording an 
excellent summer range for cattle (see PI. IX, fig. 1), there would 
probably be no economic advantage in cultivating the hardy cereals 
grown in the warmer part of this zone in various sections of the 
United States and Canada. A very limited acreage of rye, oats, 
and wheat is grown in Middle Park and elsewhere in the mountains, 
but as yet their culture is scarcely beyond the experimental stage. 
The hardier vegetables also, as beets, parsnips, lettuce, turnips, 
potatoes, cabbages, and carrots, are raised successfully along the 
lower edge of the zone. The fact that wild strawberries, raspberries, 
currants, and blueberries grow in profusion in this zone strongly 



44 NORTH AMERICAN FAUNA. [No. 33. 

suggests that cultivated varieties might do well in favored spots. 
The roughness of much of the region in this zone precludes the possi- 
bility of extensive agriculture even under favorable climatic condi- 
tions. The chief resource of the Canadian zone, however, is its tim- 
ber, especially the extensive forests of lodgepole pines. On some of 
the northern mountains, as the Park and Medicine Bow Ranges, 
where mining operations have made the least inroad on the forests, 
these pines grow very closely and to maximum size, many of the 
trees reaching a diameter of 2 feet or more. With their remarkably 
uniform straight growth and close stand, these pines form a beau- 
tiful and stately forest and yield valuable lumber. (See fig. 31.) 

MAMMALS OF COLORADO CANADIAN ZONE. 

The following boreal mammals have their center of abundance in 
the Canadian zone of Colorado, and rarely range much lower, although 
most of them extend up into the Hudsonian zone: Fremont chick- 
aree (Sciurus fremonti), woodchuck (Marmota engelhardti) , Rocky 
Mountain red-backed mouse (Evotomys gapperi galei), Preble phe- 
nacomys (Phenacomys preblei), mountain phenacomys (Phenacomys 
orophilus), Rocky Mountain field mouse (Microtus mordax), tawny 
white-footed mouse (Peromyscus maniculatus rufinus) , snowshoe rabbit 
(Lepus hairdi), Colorado pocket gopher (Thomomys fossor), black 
bear (Ursus americanus), western fox (Vulpes macrourus), Rocky 
Mountain jumping mouse (Zapus princeps), dwarf weasel (Putorius 
streatori leptus), Canada lynx (Lynx canadensis), marten (Mustela 
caurina origenes), wolverine (Gulo luscus), shrews (Sorex vagrans 
dobsoni, S. tenellus nanus, S. obscurus, and S. personatus), water 
shrew (Neosorex palustris navigator) , and silver-haired bat (Lasionyc- 
teris noctivagans) . 

BREEDING BIRDS OF COLORADO CANADIAN ZONE. 

The following birds appear to be very commonly found in the 
Canadian zone in Colorado during the breeding season: Rocky Moun- 
tain jay (Perisoreus canadensis capitalis), olive-sided flycatcher 
(Nuttallornis borealis), Alpine three-toed woodpecker (Picoides ameri- 
canus dorsalis), Lincoln sparrow (Melospiza lincolni), Dendragapus 
obscurus, Cyanocitta stelleri diademata, Loxia curvirostra minor, 
Spliyrapicus thyroideus, S. varius nuchalis, Spinus pinus, Junco 
phseonotus caniceps, Iridoprocne bicolor, Sitta canadensis, Cinclus 
mexicanus unicolor, Myadestes townsendi, Hylocichla guttata auduboni, 
Ih ndroica auduboni, and Carpodacus cassini. Some of the above 
species breed more or less commonly from the foothill region nearly 
to timberline, but their center of abundance during the breeding 
season is in the Canadian zone, between 8,000 and 10,000 feet. 



North American Fauna No. 33, U. S. Dept. Agr. Biological Survey. 



Plate X. 




Fig. 1.— Hudsonian Zone, San Juan Mountains, Southwest of Ophir. 




Fig. 2.— Hudsonian Zone Forest on Saguache Range Near St. Elmo (1 2,000 Feet.) 



1911.] HUDSONIAN ZONE. 45 

PLANTS OF COLORADO CANADIAN ZONE. 

The coniferous forest of the Canadian zone is composed chiefly of 
lodgepole pines in the northern two-thirds of the State and of white 
firs in the southern third; but it includes also the heaviest and lowest 
part of the Engehnann spruce belt and a scattering growth of balsam 
firs (Abies lasiocarpa) along its upper edge; and in the lower third 
a fringe of blue spruce along the streams, more or less Rocky Moun- 
tain white pine (Pinus Jlexilis) , especially in the southern mountains, 
and an admixture of Douglas spruce along its lower edge throughout 
the State. The aspen (Populus tremuloides) marks this zone in all 
the mountainous sections. 

The most characteristic shrubs and plants are the Canadian buffalo 
berry (Lepargyrea canadensis) on dry slopes, chiefly of the Front, 
Park, and Saguache Ranges; alder (Alnus tenuifolia), and willows 
(Salix geyeriana and others), which fringe the cold streams and bogs 
between 10,000 and 11,000 feet; elderberries (Sambucus melanocarpa 
and S. microbotrys); blueberry (Vaccinium caespitosum); currant 
(Ribes wolfi); mountain juniper (Juniperus sibirica); mountain maple 
(Acer glabrum); columbine (Aquilegia cserulea); twinflower (Linnxa 
americana); Pachystima myrsinites; Rubacer parvijlorus; Viburnum 
pauciflorum; Viola canadensis neomexicana; Frasera speciosa (see 
PI. IX, fig. 2) ; Cytherea bulbosa; Actsea viridijtora; Dasiphora fruti- 
cosa; Rosa manca; and several species of Pyrola, Epilobium, Castil- 
leja, and Delphinium. The little dwarf birch (Betula glandulosa) 
forms a conspicuous fringe along cold streams and bogs in the upper 
part of the zone in many of the mountain ranges. 

REPTILES OF COLORADO CANADIAN ZONE. 

The only reptile observed in this zone is a garter snake (Thamno- 
pMs elegans vagrans), collected hi a cold bog near Pearl, North Park, 
in August, 1906, at an altitude of a little over 8,000 feet. It is not 
known, however, that this snake occurs regularly hi the Canadian 
zone. It is not uncommon over the State in the Transition zone. 

HUDSONIAN ZONE. 

On all the higher mountains of Colorado a dark, somber-colored 
forest belt of varying width is prominent just below timberline. 
Traced along the upper slopes of the main ranges this belt is seen to 
be heaviest and broadest in gulches and basins, while on jutting 
shoulders and exposed ridges of steep incline it often contracts to a 
narrow black line. (See PL X, fig. 1.) This is the upper forest of 
Engehnann spruce (Picea engelmanni) and balsam fir (Abies lasio- 
carpa), and marks in a general way the limits of the Hudsonian zone. 
It is a neutral or transition area combining both Arc tic-Alpine and 
Canadian zone elements. Very few species are restricted to it, and 
it is therefore not so well characterized as either bordering zone. 



46 NORTH AMERICAN FAUNA. [No. 33. 

Especially is this lack of character manifest along the lower edge of 
the zone, where in the heavy forests on slopes of moderate incline 
Hndsonian conditions merge almost insensibly into those of the 
Canadian zone. Well-defined belts of the Hndsonian zone are found 
onlv on such mountain ranges as are sufficiently high to afford alpine 
conditions on their summits and highest spurs, as the Front, Park, 
Saguache (see PI. X, fig. 2) and Sangre de Cristo Ranges, and the 
Sun Juan Mountains. The lower stretches of the above ranges and 
many connecting and outlying mountains with a near approach to 
timberline conditions on their highest elevations are capped with 
this zone. In dilute form it may be traced also along the crests of 
some of the outlying ridges and spurs which do not reach timberline. 

The vertical width of the Hudsonian zone on the Colorado moun- 
tains is not far from 1,000 feet. It narrows to 600 feet in places, and 
again widens to as much as 1,200 feet, and in extreme cases to 1,500 
feet. The lower boundary is seldom clearly defined, being usually 
on steep slopes at between 10,500 feet and 10,800 feet on warm 
exposures, and several hundred feet lower on cool slopes and espe- 
cially in gulches. The upper part of the Hudsonian belt is well indi- 
cated by timberline conditions. On normal warm slopes the spruces 
and firs begin to dwarf at a little over 11,000 feet. They rapidly 
become stunted and much dwarfed, then semiprostrate, and finally 
as dense prostrate mats from 1 to 3 feet in height and often many 
feet in diameter, crowd far up the exposed ridges in narrow ascending 
tongues, giving place to the bare Alpine slopes at a mean elevation 
of about 11,500 feet, but on the warmest slopes from 11,800 feet to 
12,300 feet. 

The upper limit of this zone (timberline) is the best marked of all 
zonal boundaries. It is very sinuous, and affords perhaps the most 
striking example of one zone lapping past another. Traced along 
the upper slopes of a mountain range, it is seen regularly to dip 
down several hundred feet to skirt the lower edges of basins and 
gulches occupied by descending arms of the Arctic- Alpine zone; and 
again to reach far above its normal elevation on the exposed crests 
of the bordering ridges. In this manner tongues of Arctic-Alpine 
and Hudsonian zone regularly lap past each other at timberline for 
a vertical distance of 400 or 500 feet and in extreme cases for twice 
that distance, or about the width of the zone. (See PL XI, fig. 2.) 

The following are a few of the limits of the Hudsonian zone in 
different parts of the State that will serve to illustrate the range of 
variation as to vertical position: Rollins Pass, Front Range (east 
slope), 10,000 to 10,900 feet; near Fremont Pass (east slope), 11,000 
to 11,600 feet; St. Elmo, Saguache Range (southwest slope), 10,800 
to 12,300 feet; Ophir, San Juan Mountains (north slope), 10,500 to 
11,500 feet; Continental Divide, southeast of Lake City (northwest 
slope), 10,800 to 12,000 feet. 



North American Fauna No. 33, U. S. Dept. Agr. Biological Survey. 



Plate XI. 




Fig. 1.— Foxtail Pines (Pinus aristata) at Timber- 
line Near St. Elmo, Saguache Mountains, at 
12,300 Feet. 




Fig. 2. —Ascending Tongue of Engelmann Spruce 
at Timberlineon Front Range Near Berthoud 
Pass (Northwest Slope), at 11,600 Feet. 



1911.] • HUDSONIAN ZONE. 47 

The Engelmann spruce predominates in the Hudsonian forest belt 
throughout Colorado, and in certain sections is the only tree present. 
Usually, however, thickets of balsam fir are scattered through the 
spruce forest in bogs and on cool slopes, and occasionally the firs 
form a dense growth, as on the eastern slope of the Front Range in 
the vicinity of Rollins Pass, on the western slope of the same range 
above Arrow and Idlewild, and west of Alpine Pass on the Saguache 
Range. Both the spruce and fir extend up to extreme timberline. 
They reach their maximum size, and the former its densest growth, 
along the upper edge of the Canadian zone. Above 11,000 feet 
dwarfed growth is the rule, but there are marked exceptions. In 
crossing the two high ridges which form the divide between the Lake 
Fork of the Gunnison and Cebolla Creek, southeast of Lake City, at 
elevations of 11,350 and 11,550 feet, respectively, I found the spruce 
belt continuous and heavy, the trees averaging from 60 to 80 feet in 
height. On a southwest slope in the Saguache Mountains north of 
St. Elmo, at an altitude of 11,700 feet, Engelmann spruces were 
growing commonly to a height of 30 or 40 feet. 

On the Front Range south of James Peak, throughout the length 
of the Saguache Range, and on the Sangre de Cristo Range south at 
least to Crestone Peak, another tree occupies much of the timberline 
region to the partial exclusion of the spruce and fir This is the 
foxtail pine (Pinus aristata), which crowds up the warm sides and 
along the crests of exposed gravelly ridges in a ragged, one-sided, 
wind-beaten growth from 5 to 15 feet in height. This pine appears 
to be local in its distribution, being present on some mountains 
and entirely absent on neighboring peaks of the same range. It 
is largely confined to the front ranges south of the latitude of Den- 
ver and was not noted on the mountains of western and extreme 
northern Colorado. It is the characteristic timberline tree on the 
mountains bordering South Park on the west, and also on that part 
of the Saguache Range known as the University Range — from Salida 
north to Buena Vista. Warm south and southwest slopes are 
chiefly occupied by the foxtail pines. Timberline on the cool sides of 
the mountains is usually formed by spruces and firs. But at St. 
Elmo, where foxtail pines form the highest recorded timberline in the 
State (12,300 feet), the gnarled and twisted, often nearly prostrate, 
trunks of the uppermost of these evidence in a most striking manner 
the stern contest for existence waged against a rigorous and adverse 
climate. (See PI. XI, fig. 1.) 

Although characteristic of the Hudsonian zone, the Engelmann 
spruce, balsam fir, and foxtail pine are by no means restricted to 
it. The spruce forms regularly a heavy growth and the fir a scatter- 
ing growth along the upper edge of the Canadian zone, while the pine 
occurs sparingly on exposed gravelly points and ridges as low as 9,500 
feet, particularly in the South Park region. 



48 



NORTH AMERICAN FAUNA. 



[No. 33. 



The mossy floor of the Iludsonian forest is saturated with moisture 
throughout the summer from the melting snows of the higher slopes, 
and cold bogs abound. Flowering plants grow in profusion, although 
the number of species is not large — the flora being mostly a mixture 
of overlapping Arctic-Alpine and Canadian zone species. 

The Hudsonian and Arctic- Alpine regions are not only the sources, 
but also the conservers, of much of the water supply, so vitally 
essential to successful agriculture throughout the State, and the 
importance of the regions in this regard can scarcely be overesti- 
mated. The forests of Engelmann spruce and balsam fir, although 
yielding lumber of very inferior quality, are used extensively for 
raining timbers in the higher mountains. 



MAMMALS OF COLORADO HUDSONIAN ZONE. 

Among mammals, no species appears to be restricted to the Hud- 
sonian, but the pika, marmot, and mountain sheep are characteristic 
of the timberline region. 

The following are found in greater or less abundance, as either 
residents or stragglers: 



Callos pernio philus lateralis. 

( 'anis lestes. 

Cervus canadensis. 

Erethizon epixanthum . 

Eutamias amcenus operarius. 

Eutamias minimus consobrinus. 

Evotomys gap peri galei. 

Felis oregonensis hippolestes. 

Gulo luscus. 

Lepus bairdi. 

Lepus campestris toivnsendi. 

Lynx canadensis. 

Lynx uinta. 

Marmota engelhardti. 

Microtus mordax. 



Microtus nanus. 

Mustela caurina origenes. 

Neotoma cinerea orolestes. 

Ochotona saxatilis. 

Odocoileus hemionus. 

Ovis canadensis. 

Peromyscus maniculatus rufinus. 

Putorius arizonensis. 

Putorius streatori leptus. 

Sciurus fremonti. 

Sorex obscurus. 

Thomomys fossor . 

Ursus americanus. 

TJrsus horribilis. 

Vulpes maerourus. 



BREEDING BIRDS OF COLORADO HUDSONIAN ZONE. 

The following birds are restricted mainly in their breeding range to 
the Hudsonian zone: Pinicola enucleator montana, Certhia familiaris 
montana, and Regulus satrapa. 

The following birds occur commonly in summer in Hudsonian and 
Canadian zones, some of them breeding early in a lower zone : 

Picoides americanus dorsalis. Wilsonia pusilla pileolata. 

Nucifraga columbiana. Regulus calendula. 

Perisoreus canadensis capitalis. Hylocichla guttata auduboni. 

Myadestes townsendi. Planesticus migratorius propinquus. 

Zonotrichia leucophrys. Sialia currucoides. 

Junco phxonotus caniceps. Colaptes cafer collaris. 
Spinus pinus. 



North American Fauna No. 33, U. S. Dept Agr. Biological Survey. 



Plate XII. 



^ti&AJ^ £" 


', '~*il 












^ 


yr^^Ptepq^- - . . ■*•> 


• * 


.--ft-' ''^Sy 


jr- V -" -;* 






" - ■ Mi- 




**^„ ■ *! 



Fig. 1.— Grays Peak Group from Near Berthoud Pass, Showing an Extensive 
Area of Arctic-Alpine Country. 




Fig. 2.— Arctic-Alpine Zone on the Saguache Range Near St. Elmo. 



1911.] ARCTIC-ALPINE ZONE. 49 

PLANTS OF COLORADO HUDSONIAN ZONE. 

The distribution of the Engelmann spruce (Picea engelmanni), 
balsam fir {Abies lasiocarpa), and foxtail pine {Pinus aristata) in the 
Hudsonian zone has already been discussed in detail. Conspicuous 
plants and shrubs of the timberline region are: Caltha leptosepala , 
Trollius albiflorus, Ranunculus unguiculatus, Erysimum radicum, 
Sedum stenopetalum, Polemonium (several species) , Trifolium (several 
species), Androsace (several species), Clementsia rhodantha, Rhodiola 
polygama and R. integrifolia, Mertensia alpina and other species, 
Epilobium, Vaccinium erythrococcum, Kalmia micropTiylla , Betula 
glandulosa, Salix glaucops, S. pseudolapponum, S. cMorophylla , and 
S. saximontana. 

ARCTIC- ALPINE ZONE. 

Truly arctic conditions characterize this zone, which occupies the 
summits of the liighest peaks and the crests of all the higher moun- 
tains above the limit of tree growth. Small areas are found on the 
Park Range from Buck Mountain and Mount Zirkel south nearly to 
Buffalo Pass; on the southern half of the Medicine Bow Range; on 
the higher eastern end of the Rabbit Ear Mountains; and on the 
Vasquez, Williams, and Gore Mountains of the Middle Park region.' 
A belt which is continuous, or at most is broken only by narrow gaps 
dipping to timberline or a trifle below, extends from Longs Peak 
south along the crest of the Front Range to Grays Peak (see PL XII, 
fig. 1) and the Leadville region. Extensive areas of the Arctic- Alpine 
zone are found on the mountains of southern Colorado, especially on 
the San Juan Mountains, and on the Saguache (see PI. XII, fig. 2), 
Sangre de Cristo, and Culebra Ranges. The La Plata, San Miguel, 
Elk, Tarryall, and Wet Mountains, and the Kenosha Range all have 
many summits reaching far above timberline. Small islands of the 
Arctic- Alpine zone also cap many of the more or less isolated peaks 
of the Gunnison country, Pikes Peak, and a few of the highest eleva- 
tions at the eastern end of the White River Plateau known collec- 
tively as the Flat Top Peaks. 

The altitude of extreme timberline, which marks the lower bound- 
ary of the Arctic- Alpine zone, varies over the State according to slope, 
exposure, and latitude from a trifle under 11,000 feet to 12,300 feet. 
It is lowest in the northern mountains and in the San Juans, 1 where 
the mean elevation is 11,500 feet and the extremes are 10,900 feet 
(east side of Rollins Pass and south of Ophir), and 11,800 feet (moun- 
tains southeast of Lake City) ; and liighest on the Front and Saguache 
Ranges west of South Park and the Arkansas Valley, with a mean 

i The reason for the low average timberline elevations in the San Juan Mountains is the lowness of the 
base level (5,000 feet), where their latitude and their proximity to a hot desert region on the west would 
lead us to expect a high timberline. 

90432°— 11 4 



50 NORTH AMERICAN FAUNA. [No. 33. 

elevation of nearly 11,800 feet, and extremes of 11,000 feet and 12,300 
feet (north of St. Elmo). Under normal conditions the Arctic-Alpine 
zone usually dips lowest in basins and gulches with northeast expo- 
sures, where it is often embraced for 500 feet or more by ascending 
tongues of the Hudsonian zone, which follow up the bordering ridges. 
The exceptional cases, where timberline is lowest on west or south- 
west slopes, will usually be found to be due to unnatural conditions. 
Thus at Boreas Pass, in the mountains west of South Park, timber- 
line is below 11,600 feet on a southwest slope of slide rock, while on 
the northeast slope opposite, with favorable soil conditions, it reaches 
11,900 or 12,000 feet. In approximating the elevation of normal 
timberline due allowance must always be made for the effects of 
rock slides and avalanches, which, viewed from a distance, appear to 
be descending tongues of Arctic- Alpine zone reaching far below tim- 
berline. These slides, by carrying away the soil and leaving in its 
place great masses of slide rock, create conditions unsuitable for tree 
growth, yet do not remove the climatic barrier which prevents the 
downward dispersion of Alpine species. 

The Arctic- Alpine area is a bare and bleak region covered with snow 
for the greater part of the year, and with more or less remaining on 
the ground in summer. Naturally the variety of life is here reduced 
to a minimum. 

MAMMALS OF COLORADO ARCTIC-ALPINE ZONE. 

No mammals are restricted to the Arctic-Alpine area in Colorado. 
The following species range at times to considerably above timber- 
line in different parts of the mountains: Mountain sheep (Ovis 
canadensis), grizzly bear (Ursus Tiorribilis), coyote (Canis Testes), 
marten (Mustela caurina origenes), western fox (Vulpes macrourus), 
porcupine (Erethizon epixanthum), snowshoe rabbit (Lepus bairdi), 
two chipmunks (Eutamias amaznus operarius and E. minimus conso- 
brinus), marmot (Marmota engelhardti) , Colorado pocket gopher 
(Thomomys fossor) , pika (Ochotona saxatilis), and two held mice (Mi- 
crotus mordax and M. nanus). Of these the marmot, pika, pocket 
gopher, and field mice apparently live in tins bleak region throughout 
the year. 

BREEDING BIRDS OF COLORADO ARCTIC-ALPINE ZONE. 

Three species of birds are restricted to the Alpine zone during the 
breeding season. These are the white-tailed ptarmigan (Lagopus 
leucurus), the pipit (Anthus rubescens), and the brown-capped rosy 
finch (Leucosticte australis), none of wlrich are known to breed much 
below 12,000 feet. The pileolated warbler ( Wilsonia pusilla pileolata) 
and the white-crowned sparrow (Zonotrichia leucophrys) breed regu- 



1011.] MAMMALS. 51 

larly in the thickets of alpine willows for 500 feet above timberline, 
but their breeding range also extends down through the Hudsonian 
zone. Other birds observed as stragglers above timberline are Buteo 
boreaMs calurus, Chordeiles virginiunus Jienryi, Corvus corax sinuatus, 
J unco pJiseonotus caniceps, and a humming bird (probably Selasphorus 
platycercus). 

PLANTS OF COLORADO ARCTIC-ALPINE ZONE. 

The only shrubs able to withstand the rigorous climate on the 
wind-swept slopes of this region are several species of alpine willows, 
among which are Salix petropMla, S. cMorophylla, and S. glaucops, 
with its smooth variety glabrata. Dense thickets of these willows 
from a few inches to 3 feet in height occur from timberline to 13,000 
feet, chiefly in ascending tongues through alpine bogs on the slopes 
of gulches and basins (see fig. 35) ; wlule S. petrophila, at least, is met 
with sparingly on the highest summits, at 14,000 feet. Some of the 
plants characteristic of the region above timberline are Silene acaulis, 
Saxifraga debilis, Leptasea austromontana, Micranthes rhomboidea, 
Rhodiola polygaina, Clementsia rhodantha, Mertensia alpina, Myosotis 
alpestris, Pedicularis grcenlandica, Polemonium confertum, Polygonum 
viviparum, Swertia palustris, Sieversia turbinata, Phlox condensata, 
Besseya alpina, Trifolium nanum, TKlaspi purpurascens, and Macro- 
nema discoideum. 1 Most of the vegetation present above 13,000 feet 
consists of mosses and lichens. 

MAMMALS OF COLORADO. 

The following list is believed to include all species of mammals 
known to occur in Colorado, and aims to furnish accurate knowledge 
of their geographical and vertical distribution within the State. As 
already stated, it is based primarily upon investigations conducted 
in Colorado by the writer during the field seasons of 1905, 1906, 1907, 
and 1909. In addition, the data secured by other members of the 
Biological Survey during the past 20 years have been incorporated, 
and all important articles bearing on the subject, both old and recent, 
have been freely quoted. 

The first accurate information regarding Colorado mammals was 
obtained on Maj. Long's expedition in 1820. Several species new to 
science were obtained within the State, and were described by Thomas 
Say in the Report of Long's Expedition to the Rocky Mountains, 
v. 2, 1823. 

In the summer of 1871 Dr. J. A. Allen spent four weeks in Park 
County, securing data which later formed the basis of an important 

1 As indicative of the arctic environment of this region, note the occurrence of Pedicularis granlandica, 
Myosotis alpestris, and Polygonum viviparum— species which are abundant in Greenland and Alaska. 



52 NORTH AMERICAN FAUNA. [No. 33. 

paper on the mammals of the South Park region, published in the 
Bulletin of the Essex Institute, VI, pp. 53-58, 1874. 

Many Colorado mammals collected in the course of the United 
States Geographical Surveys West of the One Hundredth Meridian 
are recorded by Coues and Yarrow in v. 5 (Zoology) of the Report of 
the Survey, 1875. Other articles and notes of less scope pertaining to 
the subject, which have appeared from time to time in various pub- 
lications, need not be detailed. Four important recent publications 
on Colorado mammals, however, often quoted in the present report, 
are here cited in full to avoid frequent repetition of the complete 
reference. 

Allen, J. A. List of Mammals Collected by Mr. Charles P. Rowley in the San Juan 
Region of Colorado, New Mexico, and Utah, with Descriptions of New Species. 
<Bull. Am. Mus. Nat. Hist., V, pp. 69-84, 1893. 

Warren, Edward R. The Mammals of Colorado. ^Colorado College Publica- 
tions, gen. ser. no. 19, Science ser. no. 46, XI, pp. 225-274, 1906. 

Warren, Edward R. Further Notes on the Mammals of Colorado. ^Colorado 
College Publications, gen. ser. 33, Engineering ser. 1, No. 4, pp. 59-90, 1908. 

Warren, Edward R. The Mammals of Colorado. Pp. 300, with maps and numerous 
text figures. New York and London, 1910. 

Didelphis virginiana Kerr. Virginia Opossum. 

An apparently authentic instance of the occurrence of the 
opossum in Colorado was given me by Mr. George Heckler, a ranchman 
living in Shell Rock Canyon, in the northwest corner of Baca County. 
Mr. D. A. Rhinehart, of Lamar, who was familiar with the opossum 
farther east, afterwards verified Mr. Heckler's statement. 

Mr. Heckler states that during the winter of 1903-4 he was hauling 
supplies from Las Animas to the sheep camps in southern Bent and 
Las Animas Counties, and one stormy evening camped in an old cabin 
among the cottonwoods near the head of Caddoa Creek, some 12 miles 
north of Gaume's ranch. Soon after kindling a fire in the fireplace, 
he saw a peculiar animal skulk in at the open door, and knocked it 
over with a stick of wood. After examining it and noting that it was 
an animal with which he was unacquainted, Mr. Heckler left it lying 
near the door, apparently dead. Happening to glance in its direction 
a little later, however, he saw the eyes slyly open for a look around, 
and the tales of "playing 'possum" which he had read came to his 
mind. Mr. Heckler kept the animal alive and gave it to a gentleman 
residing a few miles north of Springfield, where Mr. Rhinehart saw it 
soon after its capture. 

The opossum is known to occur commonly in west-central Kansas, 
in western Oklahoma, and along certain of the streams on the Texas 
Panhandle. The individual captured on Caddoa Creek was probably 
a straggler, as careful inquiry in the Arkansas Valley and elsewhere in 
the region elicited no information respecting the presence of opossums. 
It may have reached the head of Caddoa Creek either by way of the 



1911.] MAMMALS. 53 

Arkansas, which is heavily fringed with cottonwoods from middle 
Kansas westward, or from the southeast, through the valley of the 
Cimarron River. Either route is a logical one for Didelphis to follow 
westward. Caddoa Creek is a southern affluent of the Arkansas River 
and has a scattering fringe of cottonwoods along most of its length. 
The region at the head of Caddoa Creek is wild and unsettled, and the 
cabin where the opossum was captured is the only one in many miies. 

Cervus canadensis Erxleben. Elk; Wapiti. 

The elk is now exterminated over much of its former range in 
Colorado, and the few bands which remain in the wildest parts of the 
western plateaus and mountains are small and widely scattered. The 
huge piles of antlers at many of the ranches in the northern mountains 
are a mute testimony to the former abundance of this noble animal. 
Estimates in 1898 placed the number of elk in Colorado at 7,000; in 
1902 at 3,000. In 1909 their numbers were reduced to considerably 
less, and were divided about equally between northern and southern 
Colorado. 1 Conservative estimates of the number in Routt and Rio 
Blanco Counties varied from 200 or 300 to twice that number. Mr. 
Andrew R. Hodges, a well-known game warden with wide experience 
in the elk range, thinks there were fully 400 and possibly 500 elk in 
the mountains of Gunnison County up to 1909. In addition there are 
known to be small bands in the San Juan and La Plata Mountains 
and elsewhere. 

In 1905, when I began work in Colorado, a few elk were still found in 
the Rabbit Ear Mountains, ranging on both their north and south 
slopes near the heads of Troublesome Creek (Middle Park) and 
Arapahoe Creek (North Park). A small band was said to range the 
Vasquez Mountains, and also on the headwaters of the Williams Fork 
of Grand River, while several other bands were reported on the White 
River Plateau, southeast of Meeker, probably at one time the best elk 
range in the State. In August, 1905, I found the nearly complete 
skeleton of a large bull elk in a- narrow box canyon along the East 
Fork of Rifle Creek, 20 miles northeast of Rifle, and saw many 
mounted heads in Meeker, Newcastle, and Glenwood Springs. Mr. 
A. G. Wallahan, of Lay, reported that, unless they had been killed 
since 1904, a fair number of elk still remained in the Flat Top country 
at the eastern end of the Williams River Mountains. 

In 1906 I learned of a few elk in the Danforth Hills (northwest of 
Meeker) , in the mountains west of Green River, and along the crest of 
the Book Cliffs. A small band reported on the head of Elk River, on 
the western slope of the Park Range, is said to cross occasionally to 
the neighboring Elk Head Mountains. A mounted head at Baggs 
Crossing, Wyoming, belonged to a large bull elk which a trapper 

i Careful estimates made by the Forest Service officers in the spring of 191 1 show a total of about 2,100. 



54 NORTH AMERICAN FAUNA. [No. 33. 

named Criss shot in the region between Sunny Peak and the Vermillion 
Bluffs in 1902. During the winter of 1905-6 two elk were seen at the 
Sand Springs, west of Sunny Peak, Routt County. Apparently none 
remain on the Medicine Bow and Front Ranges, or on the divide east 
of Laramie River, since I could get no definite information regarding 
their presence in that region in 1906. Dr. A. K. Fisher learned of a 
few on the south side of Longs Peak in 1894. 

In 1907, 12 or 15 elk were reported in the Cochetopa Hills west of 
Saguache, and forest rangers estimated that there were nearly 100 
remaining on the San Juan Mountains. These are much scattered, 
and are usually encountered singly or in twos or threes. One of the 
largest bands, consisting of about a dozen individuals, is said to range 
near the summit of the San Juans south of Wagon Wheel Gap. The 
same year fully 50 elk were reported in the La Plata Mountains, their 
favorite range being on the head of Hermosa Creek, northwest of 
Durango. Warren states that a few elk remain in Delta and Pitkin 
Counties. 1 They must have been at one time very abundant in the 
West Elk Mountains, as at Cebolla, in 1907, I saw large piles of antlers 
in front of the hotel. 

Allen says that elk were becoming rare in Park County in 1871 ; 2 
and Brewer, writing of South Park at the same period, records but 
three individuals observed by his party. 3 Trippe has recorded elk 
from the higher parts of Clear Creek County. 4 

Forest Supervisor H. N. Wheeler thinks the region on the head of 
Hermosa Creek, in the Montezuma National Forest, would be an 
ideal location for a national game preserve, with special reference to 
elk, since so many are living in that region. State and national pride 
should demand that this noble game animal, which formerly ranged 
the Colorado mountains by thousands, be given the protection which 
only a game preserve can afford. The close season on elk in Colorado 
extends to November 1, 1924 (amendment of 1909). It has hitherto 
been very difficult, however, to obtain in the wildest mountain dis- 
tricts a strict enforcement of the game laws. 

Mr. Barrett Littlefield, of Slater, Routt County, had about 100 
elk in captivity on his ranch on the northern slope of the Elk Head 
Mountains, a short distance south of that point, in 1906, and had 
raised calves and steer elk for a number of years, readily selling the 
dressed meat for 20 cents a pound at the nearest railroad point, 
Rawlins, Wyoming. The steer elk are said to weigh from 500 to 
600 pounds when well grown. Gray wolves have killed a number of 
elk from Mr. Littlefield's herd during the past few years. 

i Mammals of Colorado, Colo. College Pub., Sci. ser. no. 46, XI, p. 236, 1906. 

2 Bull. Essex Inst., VI, p. 56, 1874. 

3 Am. Nat., V, p. 220, 1871. 

4 See Coues, Birds of the Northwest, p. 22-1, 1*74. 



1011.] MAMMALS. 55 

Odocoileus virginianus macrourus (Rafinesque). White-tailed Deer. 

The white-tailed deer is at present uncommon, being largely 
restricted to the foothills and eastern slopes of the front ranges, where 
it occurs sparingly across the entire width of the State. A very few 
have been killed on the mountain slopes bordering the San Luis Val- 
ley, but it appears to be absent from other sections west of the main 
ranges. In early days this deer was found pretty generally over the 
plains region of eastern Colorado, where it frequented the thickets 
and cottonwood growth which fringed the Platte, Cache la Poudre, 
Boulder, Arkansas, and other streams. 

In 1905 Mr. Walter Blanchard informed me of a pah of white-tailed 
deer which had been on his ranch 5 miles west of Boulder for a num- 
ber of years, and he has several times seen fawns. I saw deer tracks 
in the gulches hi that vicinity several times dining June, 1905. Near 
the summit of Floyd Hill, in the eastern edge of Clear Creek County, 
a fine large buck jumped from a dense thicket of dwarfed aspens 
directly in front of me, June 23, 1905. Settlers in that vicinity report 
small numbers of deer in the heavy forests several miles south of Clear 
Creek. A few white-tailed deer were reported in the yellow-pine belt 
west of Arkins, Larimer County, and in the Laramie River region. 
Mr. T. J. McKenna, of the Stevens mill, on Mount McClellan, states 
that this deer formerly occurred in the Grays Peak region, but that 
none have been seen recently. A specimen from the Cache la Poudre 
River was in the collection of mammals which Mrs. M. A. Maxwell 
exhibited at the International Exposition at Philadelphia in 1876. 

White-tailed deer were reported at a number of localities in southern 
Colorado in 1907. A few are said to have been killed in the Wet 
Mountains east of Westcliffe during recent years. Mr. J. W. Frey, 
of Salida, considers the species of rare occurrence in that vicinity, 
but states that a few have heen seen recently in Pleasant Valley, on 
the east side of the Sangre de Cristo Range. This deer is reported 
rare at both La Veta and Bradford, in western Huerfano County, but 
several have been killed on the San Luis Valley slope of the Sangre de 
Cristo Range, in the vicinity of the Mosca and Medano Passes, within 
the past 10 years, according to the cowboys of the Medano Springs 
ranch. There is a skull in the National Museum from Rio Grande 
County, on the western side of the San Luis Valley. 

Most of the hunters with whom I conversed had never seen the 
white-tailed deer in the juniper country of Las Animas and Baca 
Counties, but Mr. D. A. Rhinehart, of Lamar, stated that a very few 
were killed there in early times. The species was formerly common 
among the cottonwoods and willows along the Arkansas River, but 
apparently none remain. A ranchman living at Prowers Station, 
west of Lamar, is said to have a few in captivity, but I was unable to 
learn where they were secured. Warren gives the following data: 



56 NORTH AMERICAN' FAUNA. [No. 33. 

"C. E. Aiken says there are a few in the foothills west of Monument, 
El Paso County, unless recently killed off. Dr. W. H. Bergtold tells 
me it is still found near Trinidad and southward, and also in parts of 
the Arkansas Valley, between Pueblo and the State line." 1 Trippe 
says "Cervus leucurus" was formerly common in Clear Creek County. 2 

Odocoileus hemionus (Rafinesque). Mule Deer. 

The mule deer has a much more general distribution in Colorado 
than 0. macrourus, and ranges from the lowest foothills occasionally 
to timberline. It is found in every county west of the Continental 
Divide, being probably most abundant in Routt and Rio Blanco 
Counties. East of the main ranges a few are left in the rough juniper 
country of Las Animas and Baca Counties, in the higher parts of El 
Paso, Teller, and Jefferson Counties, and in 1907 they were reported 
tolerably common in the foothills of Custer and Huerfano Counties. 
Mr. Edward A. Preble reported a few mule deer in the Estes Park 
region in 1895, but I heard of none in the foothills of Boulder and 
Larimer Counties in 1906. Apparently none remain on the plains 
east of the mountains, where they were common in early times. 

Abundant signs of deer were noted on all the mountains and plateaus 
west of the Front and Medicine Bow Ranges in August and Septem- 
ber, 1905 and 1906. The aspen forests on the Rabbit Ear Mountains, 
White River Plateau, and in the Hahns Peak country were especially 
frequented by them, and large piles of antlers were seen at ranches 
throughout the region. The Flat Top country of the Williams 
River Mountains, west of Egeria Park, was said in 1906 to be the best 
deer country in the State. All the many hunting parties which I 
met returning from this region or the White River Plateau early in 
October of that year had been successful. 

Mr. A. G. Wallahan, of Lay, Routt County, states that prior to 
1900 large herds aggregating many hundreds of mule deer passed his 
ranch on the Lay game trail each fall in regular migration from their 
summer home in the Elk Head and Williams River Mountains to 
the winter range in the rough juniper and pinyon country bordering 
the lower Snake and Bear Rivers. Since 1900 their numbers have 
been greatly depleted, and during the winter of 1904-5 Mr. Wallahan 
saw only 17 in the vicinity of his ranch. These semiannual migra- 
tions performed by the mule deer from high to low country in the 
fall, and back again to the mountains in the spring, are now scarcely 
perceptible in most sections where 5 or 10 years ago they were impor- 
tant events. 

In 1907 this deer was reported as tolerably common in the San 
Juan and La Plata Mountains, in Archuleta, La Plata, and Monte- 
zuma Counties. Formerly there was a regular movement down into 

i Mammals of Colorado, p. 237, 1906. 2 See Coues, Birds of the Northwest, p. 224, 1874. 



1011.] MAMMALS. 57 

the pinyon country on the Indian reservation along the southern 
border of the State, but during the past few years the Ute, Navajo, 
and Apache Indians have been ruthlessly slaughtering deer, and the 
greater number now remain throughout the year in the high country, 
where they are comparatively safe. Forest Ranger E. E. Chapson 
states that at present coyotes are the worst enemies of deer in the 
San Juan Mountains north of Pagosa Springs, as they kill a great 
many fawns in summer and numbers of adults in winter, when the 
snow crust impedes the deer's progress, but is strong enough to bear 
the weight of the coyotes. Mr. Chapson has often found the car- 
casses of deer thus killed, and states that the coyotes, usually hunt- 
ing in bands of five or six, first hamstring the deer, after which they 
easily kill it. Mountain lions also kill a large number in the San 
Juan and La Plata Mountains. 

During the same season (1907) deer were reported common at 
various other points. On the Uncompahgre Plateau tracks were 
seen in abundance, both in the sandy yellow pine country at the 
head of Dominguez Creek and in the trails leading through the aspen 
groves along the crest of the plateau near Uncompahgre Butte. In 
the Coventry region a good many deer were inhabiting the pinyon 
country during July, and tracks were common also in the aspen 
country on Lone Cone. Before sunrise July 24 I saw a doe browsing 
in the willow brush in the bottom of the Naturita Canyon, near Cov- 
entry. Mule deer were quite common late in June in the oak and 
aspen thickets of the Lone Mesa region, where they were feeding 
extensively on acorns. During the winter they are said to range 
chiefly in the yellow pine country on the Dolores Plateau. Just 
before sunset June 26, as I was crossing a ridge on the southwest 
flank of Beaver Mountain, southeast of Lone Mesa, my eye was 
arrested by a beautiful sight. Standing in a small grassy opening 
among the dense oak chaparral a hundred feet below me and in the 
shade of the ridge, with head thrown back over his shoulders, nose 
pointing exactly in my direction, and every sense alert, stood a large, 
sleek-coated, 4-point buck. I stopped in plain sight and silently 
watched the old fellow for fully five minutes, and he apparently 
regarded me with equal interest. His curiosity was fully satisfied, 
however, the moment I quietly and slowly concealed myself among 
the oak brush, and lowering his head he loped off through the 
chaparral. 

Deer were reported as more abundant than usual on the mountains 
bordering the San Luis and upper Arkansas Valleys, and nearly all 
the hunters outfitting at Buena Vista, Salida, and Saguache during 
the open season of 1907 met with success. A locomotive on the 
Denver & Rio Grande Railroad, running light, killed a fine buck 
November 8, in Browns Canyon, 6 or 8 miles north of Salida. On 



58 NORTH AMERICAN FAUNA. [No. 33. 

the western slopes of the Sangre de Cristo Range the mule deer is 
said to be usually found in the pinyon belt between 8,000 and 9,000 
feet, although occasionally met with at timberline in summer and 
early fall. I was informed that deer were scarce in 1907 on the head 
of Smiths Fork, in the West Elk Mountains, and this was the case 
also in western Montezuma County. 

Allen found mule deer, " Cervusmacrotis," common near timberline 
in Park County in 1S71; 1 while Trippe recorded the species as for- 
merly common in Clear Creek County. 2 

The Colorado game law in force during 1907 and 190S allowed the 
killing of one deer "with or without horns" by any person during the 
open season. This inclusion of deer without horns worked incal- 
culable injury to the deer of Colorado during the two years it was 
operative, as a great many does and fawns were killed by unscrupu- 
lous hunters and particularly by novices. The pernicious results 
were not fully manifest in 1908, when it was estimated a total of 
about 2,500 deer were killed during the open season, but reports show 
that a great scarcity of deer throughout the mountains marked the 
season of 1909. 

Antilocapra americana (Ord). Antelope. 

Antelope are now comparatively scarce even in the thinly settled 
parts of the eastern plains region, and few remain on the sage plains 
of North Park and Routt County, where formerly there were thou- 
sands. A small number are still present in San Luis Valley. Most 
of the antelope of eastern Colorado are now in three areas — on the 
plains of western Baca and southern Otero, Bent, and Prowers Coun- 
ties; on the Arkansas Divide; and in northwestern Logan and north- 
eastern Weld Counties. 

That this most graceful game mammal is doomed to early extinc- 
tion in many sections seems probable despite the protection afforded 
by law, but the increase in numbers during the past few years, both 
on the Arkansas Divide and in northeastern Weld County, is very 
perceptible. The decrease of antelope in the State at large during 
the past 10 years, however, has been very great. In 1898 the State 
game warden placed the number at 25,000, while in 1908 the game 
commissioner estimated not over 2,000. A conservative estimate 
based on data collected by the Biological Survey would be not over 
1,200 in 1909. Many cowboys and even ranchmen in the outlying 
districts kill antelope whenever they can. It is gratifying, however, 
to find that many large ranch owners have the public interests and 
the preservation of game sufficiently at heart to afford this fine 
mammal rigid protection on the ranges under their control. 

i Bull. Essex Inst., VI, p. 56, 1874. 2 See Coues, Birds of the Northwest, p. 224, 1874. 



1911.] MAMMALS. 59 

Mr. A. G. Wallahan, of Lay, the well-known photographer of big 
game, estimated in 1905 that 300 would fully eover the number of 
antelope remaining in Routt County. He reported a band of about 
30 on the Iron Springs Divide, east of Godiva Ridge, and another 
larger band near Junction Mountain, between the Snake and Bear 
Rivers. In 1906 I was informed by Mr. John Criss, of Baggs Crossing, 
Wyoming, a trapper of many } T ears' experience in Routt County, that 
only a few hundred antelope are left in that region; whereas as re- 
cently as 1898, immense herds, aggregating thousands of individuals, 
wintered there. During the summer the antelope were scattered over 
the sage plains in small bands. Mr. Criss thinks that persistent 
hunting in that region drove most of the antelope northward to the 
Wyoming plains. In 1906 antelope were reported in small numbers 
in North Park; on the sage flats bordering Muddy Creek, in north- 
western Middle Park; on the divide between the Snake and Green 
Rivers, north of the Escalante Hills; and in the parks on the O-wi- 
yu-kuts Plateau. In August, 1906, I saw the tracks of small bands 
at the eastern end of Godiva Ridge (Routt County) and near Iligho 
(North Park) , while five or six were watering at Elk Springs, 8 miles 
south of Lily. I saw none in Mesa County, and Warren states, on the 
authority of Mr. W. P. Ela, of Grand Junction, that but two or three 
have been killed in that region during the past 20 years. 1 

Numerous data on the distribution of antelope in the San Luis 
Valley and in other parts of southern Colorado were gathered during 
the summer of 1907. None appear to remain in Montezuma County, 
and hunters in that region state that few were ever found there. 
Mr. George J. Ashbaugh, living in the McElmo Canyon, told me 
that the last antelope seen by him were three or four on the water- 
shed between Yellow Jacket and McElmo Creeks about 1897. A 
conservative estimate of the number of antelope in the San Luis 
Valley would be from 50 to 75. The cowboys of the Medano Springs 
ranch, who range the eastern side of the valley from Crestone south 
to Garland, state that there are about 30 or 40 in that region, the 
largest band, consisting of about 25, being found on the Luis Maria 
Baca grant, south of Crestone. Others are scattered over the plains 
from the west base of Sierra Blanca north to the northern end of 
the sand dunes, west of the Medano Pass. Antelope are reported 
between Garland and San Luis, at the west base of the Culebra 
Range, and Vernon Bailey heard of a few in the unsettled region 
between Antonito and the Rio Grande in 1904. Mr. C. H. Auld, 
of Colorado Springs, reports seeing an antelope from the train be- 
tween La Jara and Antonito September 18, 1907. 

An antelope is said to have been killed, in the fall of 1907, within 
sight of Salida, presumably on the extensive flats in the Arkansas 

1 Mammals of Colorado, p. 237, 1900. 



60 NORTH AMERICAN FAUNA. [No, 33. 

Valley northwest of that point. This appears to be the only recent 
record in the upper Arkansas Valley. Antelope were not uncom- 
mon in South Park in 1871, according to Allen. 1 They were formerly 
abundant in the Wet Mountain Valley, but none have been seen 
in that region for several years. 

Conservative estimates made by residents of Baca County in 
the fall of 1907 placed the number of antelope in that region at 
between 50 and 100. The majority were said to range on the level 
short-grass plains of western Baca County, just east of the rough 
juniper country. Very few are reported in eastern Baca County, 
which is well settled. Prof. D. E. Lantz reports possibly 12 ante- 
lope on the plains north of Higbee, in southeastern Otero County, in 
1910. 

Antelope were reported by Mr. C. P. Streator as common on the 
high plains of the Arkansas Divide near Flagler and Limon in 1894, 
and Prof. Lantz says their numbers were increasing near Hugo in 
1905, owing to rigid protection. In May, 1909, I saw a band of 
fully 20 from a Rock Island train about 10 miles east of Limon, and 
in crossing tjie Arkansas watershed from Cheyenne Wells northwest 
to Seibert I saw several small bands. One band of nearly 200 is 
said to have been seen near Agate, on the northern slope of the 
Arkansas Divide, northwest of Limon, in December, 1908. Several 
antelope were run over by trains in that vicinity the same winter, 
which was one of excessive snowfall. This was due to the fact that, 
in feeding along the railroad right of way, which was kept fairly 
clear of snow, the antelope became quite unsuspicious and accus- 
tomed to passing trains. 

In driving from Sterling northwest to Grover early in June, 1909, 
several small bands of antelope were encountered on the flats border- 
ing Horsetail Creek south of the Chimney Cliffs, while on the undu- 
lating grassy plateau northwest of Pawnee Buttes, where the antelope 
seemed most numerous, 19 were counted in the course of a short 
morning's drive. From observations made during the winter of 
1908-9 residents variously estimate at from 200 to 300 the antelope 
in the region between the Burlington Railroad from Cheyenne to 
Sterling and the Nebraska boundary. A point 10 miles west of the 
Platte River at Iliff indicates the present eastern limit of range in 
winter, while the summer range is still more restricted. 

A very few antelope are found in Yuma County, according to Mr. 
W. E. Wolfe, of Wray. Mr. H. G. Smith, of Denver, states that 
they were last seen in Denver County in the late seventies. 

Bison bison (Linn.). Buffalo; American Bison. 

The buffalo was formerly present over much of the State, even rang- 
ing in summer to timberline in certain sections of the mountains, 

' Bull. Essex Inst., VI, p. 56, 1874. 



1911.] MAMMALS. 61 

as is proved by the bleached and weathered skulls occasionally 
found at that elevation. While most numerous on the plains east 
of the mountains, they nevertheless must have been common in the 
larger mountain parks, especially on the sage plains of North Park, 
where the bleached skulls, now rapidly disintegrating after more 
than 20 years' exposure, may still be seen in considerable numbers. 
A favorite range of the buffalo was the extensive region of sage 
plains in western Routt County, where in sections least frequented 
by range cattle the deeply worn trails can still be distinguished. 
Few, if any, well-preserved skulls can be found at the present time. 
On my trips in North Park and in Routt County in 1905 and 1906 
I met with only one skull at all complete, and in this case the teeth 
were lacking. This skull was found embedded in a bog north of 
Higho, North Park, and had been little exposed to the weather. 
In most of the skulls found the nasals and horn sheaths, as well as 
the teeth, are missing. - ■ 

The South Park plains were much frequented by buffalo in early 
times, and nearly all the travelers passing through the region men- 
tioned them. They were by no means confined to the open grassy 
plains of the park, since Brewer records skulls found on the bordering 
mountain slopes as high as 11,000 feet in both openings and forests. 1 
The same writer states that the mountain animals were unlike the 
buffalo of the eastern plains, being smaller, with longer, shaggier, 
and blacker hair. 2 Allen mentions bleached skulls found far above 
timberline on Mount Lincoln, by Mr. Bennett and others on the 
"extremest sources" of the Platte River. 3 Warren records a skull 
which he found in the mountains near Irwin, Gunnison County, at 
nearly 11,000 feet, and states that he has seen skulls at other locali- 
ties in the West Elk Mountains. 4 

The data at hand respecting the time of the buffalo's disappear- 
ance in most sections of the mountains are scanty. According to 
Brewer (1. c.) it was abundant in South Park previous to 1862, and a 
few were shot in 1867, while he "heard of none in 1869." Allen 
(1. c.) states that a small band was seen along the Platte River in 
South Park in June, 1871, while Coues and Yarrow, writing of the 
South Park region, state that a few were seen and two were killed in 
1873. 5 Mounted specimens in the Maxwell collection of Colorado 
mammals exhibited at the Philadelphia Exposition in 1876 were shot 
in September, 1873, near Mount Whiteley, Middle Park, according 
to Coues, who also states that a small band of buffalo still lingered 
in North Park in 1876. 6 The last buffalo in the Pikes Peak region 

i Am. Nat, V, p. 221, 1871. 

2 In this connection, see Allen, Bull. Essex Inst., VI, p. 55, 1874. 

3 Bull. Essex Inst., VI, p. 55, 1S74. 

* Mammals of Colorado, p. 239, 1906. 

a Explorations W. of 100th Meridian, V, p. 07, 1875. 

6 Dartt, On the Plains and among the Peaks, p. 221, 1879. 



62 NOKTH AMERICAN FAUNA. [No. 33. 

seems to have been a bull, killed in 1879 near the Seven Lakes, and 
mounted by Mr. C. E. Aiken, of Colorado Springs, as recorded by 
Warren (1. c). The famous and somewhat mythical "Lost Park 
Herd" appears on good authority to have lived in the fastness of 
its mountain retreat in comparative safety for many years after its 
fellows in other sections of the mountains had been exterminated. 
The remnant of this herd — two bulls, one cow, and one calf — were all 
killed in February, 1897, in Lost Park, near Bison Peak, Park County. 1 
The last buffalo in Routt County appears to have been one killed by 
Ute Indians at Cedar Springs, 6 miles west of Craig, in 1884. 2 

The buffalo of the plains east of the mountains appear to have 
made their last stand in Baca County, in the extreme southeast, 
where one was killed near Springfield as late as 1889. This is the 
same year that the remnant of the Texas herd, driven to the north- 
west corner of the Panhandle, was estimated by Dr. W. T. Hornaday 
at 25. 3 As the Baca County plains are practically a northern con- 
tinuation of the Llano Estacado, and but a short distance north of 
the Panhandle, it seems altogether likely that the buffalo killed in 
Baca County, a cow heavy with calf, had strayed from the Texas 
herd, as the animals remaining on the eastern Colorado plains had 
been exterminated several years before. 

Ovis canadensis Shaw. Mountain Sheep; Bighorn. 

A few bands of mountain sheep live on nearly all the high moun- 
tain ranges of Colorado. On the main ranges they are usually seen 
at or near timberline and seldom below the Canadian zone. On the 
plateaus and in the rough country of western and southwestern 
Colorado, however, they occur at much lower elevations. 

Mountain sheep show a gratifying increase during the past few 
years over the State at large. Although they have been protected 
by law in Colorado since 1885, the marked increase at the present 
time is the result both of a more efficient game- warden service and 
of local protection afforded by an aroused public sentiment. A 
danger which threatens mountain sheep in Colorado, as well as in 
other Western States, is the introduction of scab from domestic 
sheep allowed to graze on the higher mountain slopes. An instance 
is given by Warren, as follows: 

C. F. Frey tells me they suffer much from scab in the "West Elk Mountains, and that 
a party told him in 1902, at one place near the head of Sapinero Creek, 75 head were 
counted which had died of scab. Domestic sheep have been run in that locality, and 
the wild sheep doubtless contracted it from them. 4 

In the northern ranges mountain sheep were tolerably common in 
1905 and 1906. Lumbermen at Fraser, Middle Park, reported two 

■ See American Field, p. 273, Mar. 19, 1910. 

2 Felger, Univ. of Colo. Studies, VII, No. 2, p. 143, 1910. 

3 Extermination of the American Rison, Rept. U. S. Nat. Mus., 1S89. 
* Mammals of Colorado, p. 23S. 1906. 



1911.] 



MAMMALS. 63 



bands, each numbering about 20, in the neighboring mountains. One 
of these bands is found on the Front Range, west of Berthoud Pass, 
and thence north on the Vasquez Mountains to Vasquez Peak; the 
other is usually seen near Arapahoe Peak, but ranges along the crest 
of the Front Range from James Peak as far north as Longs Peak. 
The sheep reported in the vicinity of Silver Plume and Graymont 
doubtless belong to one of these bands. Another large band was said 
to be found on the Medicine Bow Range between Lulu Pass and 
Chirks Peak. At Spicer, North Park, I saw the heads and capes of 
two fine rams which had been killed near Clarks Peak in the spring 
of 1905. These heads measured between the horn tips 15 and 15^ 
inches, respectively. This band is stated to range near timberline 
even in winter. A few sheep are still found in the mountains near 
Estes Park, but they are more common on the western slope of the 
range. A band of about 30 was seen by Nelson in the summer of 
1883 on the summit of the range west of Estes Park. They were 
reported in 1905 as tolerably common on the Park Range between 
Buffalo Pass and Mount Zirkel. Other small bands were reported 
in the Elk Head Mountains and near Pyramid Peak in the Williams 
River Mountains. Mr. John Criss, of Baggs Crossing, Wyoming, states 
that one or two sheep have been killed in the rough country north 
of Snake River, in the vicinity of Sunny Peak, since 1900. A very 
few were also reported on Junction Mountain (between Snake and 
Bear Rivers), and on Zenobia Peak and Mount Cullom, near the 
Ladore Canyon of Green River. In 1902 Dall De Weese estimated 
their number as probably 700 in the State. A large ram was killed 
among the Snake River bluffs north of Lily during the winter of 
1905-6. 

On October 16, 1906, while far above timberline in the mountains 
south of James Peak, and on the edge of a small lake in a deep basin 
surrounded by rather precipitous walls, the crashing of a number 
of loosened rocks on the slope opposite broke the deep silence. No 
sheep were visible on the rocky ledges far above, but upon climbing 
the steep slope I discovered the fresh tracks of a large ram and a ewe 
hi the snow on the crest of the divide at 13,000 feet. 

In 1907 sheep were reported on most of the mountain ranges of 
southern Colorado, and seemed to be on the increase. The sharp, 
jagged peaks and ridges of the high and narrow Sangre de Cristo 
Range afford a fine home for them, and they are said to be found 
along its entire length, being in greatest numbers hi the Sierra Blanca 
group at the southern end. Cowboys riding for the Medano Springs 
ranch often meet with sheep among the pinyons along the lower 
western slopes of the Sangre de Cristos between Mosca Pass and the 
southwest flank of Sierra Blanca. A very few are reported also on 
West Spanish Peak, south of La Veta, and on Pikes Peak. Forest 



64 NORTH AMERICAN FAUNA. [No. S3. 

Ranger E. E. Chapson, of Pagosa Springs, says that a good many 
sheep are killed by snowslides in the San Juan Mountains north of 
that point. In one instance he discovered the carcasses of three or 
lour sheep which apparently had been so closely shut in by a snow- 
slide as to be at the mercy of a band of coyotes. The snow crust was 
sufficiently strong to bear the light-footed coyotes, but too weak to 
support the heavy-bodied sheep, and unmistakable signs in the 
snow showed that the sheep had been lulled and partly eaten. Mr. 
J. P. Galloway, of Norwood, states that sheep are often seen on 
Dolores Mountain, but seldom range west across the valley to Lone 
Cone, the westernmost peak of the San Miguel Mountains. Some 
years ago a band of about 20 crossed to the Cone, but soon returned 
to the more extensive range on Dolores and neighboring mountains. 

Mountain sheep are not at all uncommon in the rough Upper 
Sonoran country of extreme western Montrose and Mesa Counties. 
Most of the region on both sides of the Dolores River below Paradox 
Valley is very precipitous and quite inaccessible, affording an excel- 
lent sheep range. Most of the sheep in this region live on the east 
side of the Dolores, between Salt Canyon and the mouth of West 
Creek, chiefly on the very broken southwest face of the Uncompahgre 
Plateau. A very few are reported among the rocky ridges just 
below the steep northern rim of West Paradox Valley, and Mr. 
William Boren, of Norwood, informed me of a lone ram which has 
been seen several times on the rocky points along the east side of the 
Dolores Canyon, just south of Paradox Valley. Sheep are said to 
be tolerably common in the West Elk Mountains, and occasionally 
range down to the cliffs along the Gunnison, southwest of Crawford, 
during the winter. 

Brewer found sheep common on the mountains surrounding South 
Park in 1871, and states that they were particularly abundant above 
timberline on the highest peaks. 1 

[Sciurus niger rufiventer Geoffroy. Western Fox Squirrel. 

Fox squirrels are not indigenous to Colorado, but have been intro- 
duced at Greeley, where they are increasing in a gratifying manner, 
and also, according to Young, at Denver. 2 ] 

Sciurus aberti ferreus True. Northern Tuft-eared Squirrel. 

Sciurus aberti concolor True, Proc. U. S. Nat. Mus., XVII, p. 241, 1894. Type 
from foothills west of Loveland, Larimer County, Colorado. (Not of Blyth.) 
Sciurus aberti ferreus True, Proc. Biol. Soc. Wash., XIII, p. 183, 1900. 

This interesting form of Sciurus aberti has a much interrupted range 
in the yellow pine belt of the eastern foothills, between 6,000 and 
8,000 feet in northern Colorado, and chiefly between 7,000 arid 9,500 
feet farther south. Judging from the meager data at hand its range 

•Am. Nat., V, pp. 220-221, 1871. 2 Proc. Acad. Nat. Sci. Phila., p. 406, 1908. 



1911.] 



MAMMALS. 



65 



was formerly continuous or nearly so, but now it is largely restricted 
to the higher foothills west of Loveland, eastern Park and southern 
Jefferson Counties, the Arkansas Divide between Monument and 
Eastonville, and the eastern slopes of the Sangre de Cristo Range 
between the Medano and Mosca Passes. (See fig. 2.) 

Although the species is commonly reported by hunters and ranch- 
men in the foothills of southern Larimer County, my experience in 
the pine belt west of Arkins (practically the type locality) in July, 
1906, leads me to believe that it is now uncommon, if not rare, in 
that section. Several days were spent among the heavy forests of 
yellow pines between 6,000 and 7,500 feet in a systematic search for 




^ SC/URUS ABERT/ M/fitUS. %^, « "»•] rerRfteos. 

Fig. 2.— Distribution in Colorado of tuft-eared squirrels (Sciurus aberti mimus and S. a. fcrrcus). 

these squirrels, but not a single one was seen or heard. One inform- 
ant stated that the squirrels come down into the valleys in the autumn 
and feed extensively upon the wild plums. In October, 1894, Mr. 
C. P. Streator collected two squirrels 12 miles west of Loveland, but 
found the animals scarce. He attributed this scarcity to a shortage 
in the crop of pine cones, upon winch the squirrels are said to feed. 
In the fall and winter months of 1894, 1895, and 1896 Mr. R. S. 
Weldon collected 12 specimens 3 miles northwest of Arkins and 1 at 
Bellevue. Possibly this squirrel lives at a higher elevation in sum- 
mer than in winter, but there are no data on this point. The type 
90432°— 11 5 



66 NORTH AMERICAN FAUNA. [No. 33. 

and one topotype in the National Museum were collected west of 
Loveland by Mr. W. G. Smith. 

In 1905 Mr. Edward Allensby, living in the South Boulder Canyon, 
14 miles southwest of Boulder, reported a few of these squirrels in 
the pine forests of that vicinity, stating that they were usually observed 
during the winter months. Mr. Walter Blanchard, of Boulder, saw 
a melanistic specimen which was killed near Sugar Loaf Mountain, 10 
miles west of Boulder, about 1903. Mr. T. J. McKenna, of Denver, 
states that many years ago, while located at Jamestown, Boulder 
County, a few black squirrels were found in that vicinity. In 1894 
Streator learned of five squirrels which had been killed in the vicinity 
of Gold Hill. 

This squirrel was reported at a number of localities in southern 
Colorado in 1907, where it seems to be more common than farther 
north. It was said to be not at all uncommon in the pine forests of 
eastern Park County between 8,000 and 8,500 feet, especially on 
Craigs Creek and along the North Fork of the Platte near Bailey. 
Prof. D. E. Lantz found it common among the pines at Cascade in 
July, 1910. Mr. J. W. Frey reports a very few in the Arkansas Hills 
east of Salida, and Mr. C. E. Aiken, of Colorado Springs, has a summer 
skin from the Cripple Creek region, which is gray with a trace of red- 
dish in the middle of dorsum, thus greatly resembling S. a. mimus 
in faded pelage. Sciurus ferreus has been seen on the San Luis 
Valley side of the Sangre de Cristo Range, according to cowboys of 
the Medano Springs ranch, Mr. William King having seen a few gray 
individuals among the pines at the west end of Mosca Pass, and 
others reporting it from Medano Pass. The species is quite generally 
distributed along the eastern slopes of the Sangre de Cristos north at 
least to Westcliffe, and it is said to occur also in the Wet Mountains. 
In the region around Bradford, in the northwestern corner of Huer- 
fano County, tins squirrel is said to be tolerably common, both on 
the slopes east of Medano Pass and in the pine-clad gulches of Prom- 
ontory Bluffs, the vertical range here being from 8,000 to 9,500 feet. 
The inclement weather prevailing in this region during mid-November 
doubtless accounted for the inactivity of squirrels, as none were seen, 
but signs of their presence were found in large nests of pine needles 
in the upper branches of pines and also in fresh cuttings scattered 
here and there over the forest floor. 

The species reaches its greatest abundance in the yellow pine 
forest on the Arkansas Divide between Eastonville and Monument. 
In the forest southwest of Eastonville the squirrels were active during 
the bright crisp days of early December. The forest floor for a con- 
siderable radius around the nest trees was generously sprinkled with 
fresh green cuttings of branch tips, pieces of gnawed bark, and freshly 
peeled twigs, while a few of the upper limbs of the trees were gnawed 



1911.] MAMMALS. 67 

and peeled in places, much as if by porcupines. The squirrels them- 
selves were very quiet and shy and were seldom observed, and not 
a single sound was heard which could be attributed to them. I 
hunted carefully for the larger part of a day before catching sight of 
one, and then secured two line brown males, two black males, and 
one black female. These specimens are in beautiful long and silky 
winter pelage, and the ear tufts are of maximum length. The fur of 
melanistic specimens is of fine quality, and were the squirrels numerous 
it would undoubtedly be in great demand. 

The nests are nearly always constructed of pine needles and lined 
with strips of the inner bark of the pine. In the Eastonville region 
I found S. ferreus living entirely in open nests, but near Bradford it 
is said often to take up its abode in hollow pines and even in hollow 
cottonwoods (Populus angustifolia) . 

Tins squirrel is one of the most striking examples of extreme mel- 
anism among mammals. Of the 24 specimens in the Biological Sur- 
vey and National Museum collections, including the type, 19 are 
melanistic (either black or dark brown) and 5 are gray. A man wdio 
had hunted in the Estes Park region for a great many years informed 
Streator in 1894 that out of nearly 100 of these squirrels which he 
had seen or killed only 1 was gray. This ratio is, of course, much 
too high, but it is certain that melanistic individuals preponderate in 
most sections. Near Eastonville I saw no squirrels in the gray phase, 
nor could I learn of any having been seen there. Air. C. E. Aiken 
states, however, that in a batch of nearly 150 skins of S. ferreus 
wiiich he purchased many years ago from the region between Monu- 
ment and Eastonville, the black phase was most common and the 
brown phase next, while there were a very few gray skins. An odd 
skull, which I secured at Bailey, belonged to a black squirrel. On 
the slopes of the Sangre de Cristo Range the gray phase seems to be 
more common than elsewhere, and no black squirrels were reported 
from the Mosca and Medano Passes. 

Sciurus aberti mimus Merriam. Tuft-eared Squirrel. 

So far as known, this large, handsome squirrel occurs in Colorado 
only in the extensive yellow pine forests which clothe the low^er 
slopes of the San Juan and La Plata Mountains, in Conejos, Archu- 
leta, La Plata, Montezuma, and Dolores Counties, in the Transition 
zone. (See fig. 2.) The center of abundance appears to be the 
Pagosa Springs region, Archuleta County, between 7,000 and 7,500 
feet. 

Tuft-eared squirrels are commonly reported from the upper valley 
of the Los Pifios and along Vallecito Creek, in northeastern La Plata 
County, and are found sparingly to the lower edge of the pine belt a 
few miles north of Bayfield, but are said never to range down into 
the pinyon country. The species has been recorded from Florida, 



68 NORTH AMERICAN FAUNA. [No. 33. 

La Plata County, by Allen. 1 One was noted at X,200 feet in the 
upper edge of the yellow pine belt on the southwest slope of Pagosa 
Peak. Mr. Steve Elkins, of Mancos, states that a few of these 
squirrels inhabit the pine forests at the western base of the La Plata 
Mountains, and their range extends a short distance out on the 
Dolores Plateau, northeast of Dolores. In crossing this plateau 
between Dolores and the Lone Mesa late in June, 1907, I saw no signs 
of squirrels, but learned from Forest Ranger James Lowell that they 
are present in small numbers. The Dolores Plateau appears to mark 
the northern limit of the range, as none were found in the yellow pine 
forests which clothe much of the San Miguel Plateau, the northern 
slopes of the San Miguel Mountains, and the Uncompahgre Plateau. 
Vernon Bailey says this squirrel was reported to him as common in 
the eastern foothills of the San Juan Mountains, 10 miles west of 
Antonito, Conejos County, in August, 1904. This is the only record 
of its occurrence in Colorado east of the Continental Divide. On 
the east side of the San Luis Valley, and along the eastern slopes of 
the front ranges, this form is replaced by S. a.ferreus. 

This squirrel is characteristic of the stately yellow pine forests near 
Pagosa Springs and in the open vistas can be seen at a considerable 
distance. It is often first detected on the ground, moving about 
among the pine cones which carpet the forest floor in many places. 
When alarmed, it lopes leisurely up to the base of a pine, usually the 
nest tree, which it seems reluctant to climb, barking and scolding at 
the intruder until approached somewhat closely. When thoroughly 
frightened, it betakes itself to the higher branches, and its claws 
make a very audible sound on the dry bark. When seated motion- 
less on an exposed limb far up in a big pine, this squirrel presents an 
odd appearance, due to its long hairy ear tufts. Once safely within 
the confines of the nest tree it will occasionally scamper part way 
down the trunk in a daring fashion, chattering excitedly. In 
climbing up or down a tree it spreads its feet far apart and by its flat 
appearance reminds one strongly of a flying squirrel. 

The nest tree — usually a large dead pine with a hollow sufficiently 
large for the squirrels' home — is generally located in the heaviest 
forest, and very few of the animals live in small timber or along the 
outskirts of the forest. A few nests, composed largely of dry pine 
needles, were seen in the upper branches of large pines, but most of 
the squirrels appeared to be living in hollow trees. A stomach 
examined at Pagosa Springs contained a mass of finely masticated 
green material which could not be identified with certainty, but 
probably consisted of the inner bark of the terminal branches of the 
yellow pine. One squirrel was seen gnawing the bark from a good-sized 
limb, apparently feeding. The many freshly cut tips of terminal 

1 Bull. Am. Mus. Nat. Hist., V, p. 83, 1893. 



1011.] MAMMALS. 69 

branches beneath the pines in the neighborhood of the nest trees attest 
to the squirrels' activities. 

In the silence of the vast forest reaches, the calls of this squirrel 
are at times the only sounds which reach the ear. During rainy or 
inclement weather, however, the squirrels are inactive and the calls 
rarely heard. The soft bark, sometimes sounding like wuh, wuh, 
wuh, and again like chuck, chuck, chuck, is usually repeated three 
or four times at short intervals, and each call is accompanied by a 
jerk of the tail. These squirrels are occasionally kept in confinement 
and are said to make desirable pets. 

In a series of four males and two females collected a few miles west 
of Pagosa Springs, May 29, 1907, two were only about two-thirds 
grown. Fully half the squirrels noted here during the last week in 
May were immature, easily distinguishable on a tree by a dark- 
colored area in the middle of the dorsum. In adults this area is 
generally reddish. 

Scrums fremonti Aud. and Bach. Fremont Squirrel; Chickaree. 

Sciurus fremonti Audubon and Bachman, Quad. N. Am., Ill, p. 237, 1853. Type 
from " Rocky Mountains " (probably from the Park region of central Colo- 
rado). 1 

The Fremont squirrel lives in the coniferous forests throughout 
the mountains of Colorado, and ranges from the upper edge of the 
yellow pine belt to timberline. (See fig. 3.) It is typical in the 
Canadian zone of the main ranges, but on the western plateaus 
becomes slightly reddish on the rump and upper surface of the tail, 
as shown by a large series of specimens from various localities. This 
squirrel is usually common wherever found, and its lively, cheering 
chatter is one of the few sounds which break the silence of the high 
mountain forests. On the main ranges it occurs chiefly in the forests 
of lodgepole pine and Engelmann spruce, but in the southern moun- 
tains is perhaps more common in the forests of white fir {Abies 
concolor), while on the western plateaus it frequents the more open 
growth of Douglas spruce on the summits and upper northern slopes. 
In the Escalante Hills this squirrel is reported to come down among 
the yellow pines, at 7,000 feet, during the winter, and in a great many 
sections I have found it along the extreme upper edge of the Transi- 
tion zone, at 8,500 feet, where pockets of balsam and aspen are scat- 
tered among the yellow pines. I shot one individual in the Upper 
Sonoran zone, at 6,000 feet, on Tabeguache Creek, 8 miles north of 
Nucla, Montrose County. It was in almost the last yellow pine found 

1 So considered by Dr. J. A. Allen, who treats the matter in detail in Bull. Am. Mus. Nat. Hist., X, pp. 
289-290, 1S98. Dr. Allen's conclusion that the type specimen of fremonti (in the collection of the Phila- 
delphia Academy of Natural Sciences) was not collected near South Pass, Wyo., as had been previously 
supposed, is confirmed by specimens in the Biological Survey collection. Large series of squirrels from 
the Wind River Mountains, near South Pass, and also from the Green, Ferris, and Laramie Mountains to 
the east and south, are quite distinct from fremonti, being referable to the northern S. hudsonicus group. 



70 



NORTH AMERICAN FAUNA. 



[No. 33. 



along the creek, at a point where the canyon sides were clothed with 
junipers and pinyons, but really within a short distance of the 
Canadian zone forests. 

Like the common northern red squirrel, which it greatly resembles 
in all respects except color, S. fremonti feeds chiefly upon pine and 
spruce cones, which are hoarded in large caches at the bases of trees, 
beneath logs, and among rocks. I have never found it living in a 
hollow tree, although it may do so occasionally. The nests of pine 
or spruce needles and fine strips of bark are usually constructed in 
the fork of a branch well out from the main trunk, at from 20 to 40 
feet above the ground, and in the densest forest. I have found the 
nests occupied by the squirrels in both summer and winter. This 




Fig. 3.— Distribution in Colorado of Fremont squirrel {Sciurus fremonti). 

squirrel is not at all shy, and may be coaxed to within a few feet by 
making a nondescript " screeping " noise. One seen by Mr. Morris M. 
Green, near Almont, in August, 1909, was laboriously ascending a 
tree, carrying a large cantaloupe rind, which had been left by a 
camper. In some localities it is called the little gray squirrel, which 
is, of course, a misnomer, as it is subgenerically different from the 
gray squirrels and its color is olive brown. 

(?) Sciurus fremonti neomexicanus Allen. New Mexico Chickaree. 

It is not certain that neomexicanus occurs within the State, but 
it has been taken at Costilla Pass and Bear Canyon, New Mexico, 
both localities within a few miles of the Colorado boundary. A. H. 



1011.] MAMMALS. 71 

Howell thinks that a squirrel which he saw among the pinyons near 
Trinidad in September, 1903, was this reddish form of S. fremonti; 
while others seen by Mr. J. H. Gaut on the Conejos River, 12 miles 
west of Antonito, may have been neomexicanus. Unfortunately, 
there are no specimens from along the southern edge of the State, 
and the identity of the squirrels found in the southern San Juan and 
Culebra Mountains can be decided only by future investigations. 

Eutamias quadrivittatus (Say). Say Chipmunk. 

Stiurus quadriinttatus Say, in Long's Exped. Rocky Mts., II, p. 45, 1S2:>. Type 
from Arkansas River, about 26 miles below Canon City, Colorado. 

This large chipmunk, the four-lined squirrel of Say, was described 
from a specimen taken by Maj. Long's party at a point on the Arkan- 
sas River about 30 miles below the place where the river leaves the 
mountains, July 17 or 18, 1820. Tins point, as Dr. Merriam has 
already shown, 1 was probably about 26 miles below Canon City. 
The species has its center of abundance in the yellow pine belt of the 
Transition zone in the eastern foothill region, where it ranges nearly 
across the State from north to south. In Larimer County its known 
range is restricted to the lower eastern slopes of the Medicine Bow 
Mountains south of Arkins. Farther south, E. quadrivittatus has a 
wide distribution within its zonal limits, and is occasionally taken 
in the Canadian and Upper Sonoran zones. In the southern tier of 
counties it extends east to the western edge of Baca County 2 and 
west to La Plata County. The western limits may be roughly indi- 
cated by Florida and Bayfield, La Plata County; Silverton; Sapinero; 
St. Elmo, Saguache Mountains; near McCoy, Eagle County; Grand 
Lake; and Medicine Bow Range. (See fig. 4.) A large series of 
specimens from Canon City (near the type locality) and from a great 
many localities over the range of this species are in the Biological 
Survey collection. 

West of the Front and Saguache Ranges I have taken this species 
at only one locality, Sapinero, in the valley of the Gunnison, at 7,300 
feet, but it has been collected at Sulphur Springs, Grand Lake, near 
Sheephorn Pass, and near McCoy, as recorded by Warren. 3 In 
extreme southern Colorado quadrivittatus is found in the foothills bor- 
dering the San Luis Valley and is common west of the Continental 
Divide, being especially numerous in the yellow pine forests of 
Archuleta and La Plata Counties, on the southern slope of the San 
Juans. In this region it is not uncommon in the pinyon belt, as at 
Arboles and south of Bavfield, and was not seen above 7,500 or 8,000 
feet north of Pagosa Springs or in the Vallecito region. This 

i Proc. Biol. Soe. Wash., XVIII, p. 103, 1905. 

2 The Baca County chipmunks have been recently separated as a pale race by E. R. Warren ( Proc. Biol. 
Soc. Wash., XXII, p. 105. 1909,. 

3 Further Notes on the Mammals of Colorado, Colo. College Pub., gen. sit. no. 33, p. 08, 1908. 



72 



NORTH AMERICAN FAUNA. 



[No. 33. 



chipmunk has, however, been taken at Silverton (9,000 feet). West 
of the L;i Plata Mountains, and apparently west of their southern 
extension, it is replaced by the Hopi chipmunk (E. hopiensis). The 
limits of these two species along the San Juan River have not been 
ascertained, but it is quite possible the ranges meet, as at McCoy, 
on the upper Grand River. 

At Sapinero, in mid-October, 1907, E. m. consobrinus was the only 
chipmunk seen alive, but October 15 three specimens of quadrivittatus 
got into my mousetraps along the rocky ledges of a gulch extending 
from the Gunnison River up through the sagebrush slopes southwest 
of the town. The species doubtless reaches the Gunnison region 




$J £Ur/>M/AS HOP/EAJS/S. 



Y/.C. QCADWV/TTATUS. 



£■ Q AA//MOSUS. 



Fig. 4.— Distribution in Colorado of Hopi, Say, and Las Animas chipmunks ( Eutamias hopiensis, E. quad- 
rivittatus, and E. q. animosus). 

from San Luis Valley by way of Cochetopa Pass, rather than across 
the high Saguache Range, since at St. Elmo, on the east side of this 
range, it was not observed in the Hudsonian and Arc tic- Alpine zones, 
although tolerably common in the Chalk Creek Valley, at 10,000 feet. 
In the Upper Arkansas Valley it is abundant to a short distance 
above Buena Vista. At Como, in South Park, several were seen on 
the timbered ridges at 10,000 feet, one being noted in the upper 
branches of a foxtail pine (Pinus aristata). 

E. quadrivittatus is the largest and best known of the Colorado 
chipmunks, and is the one commonly seen along the eastern base of 
the foothills. At several localities, notably at Colorado Springs and 



1911.] MAMMALS. 73 

Boulder, the ranges of this species and the smaller operarius of the 
boreal zones overlap for a vertical distance of nearly 3,000 feet. The 
two chipmunks are easily confused in life, as the coloration is very 
similar, but quadrivittatus is considerably larger and slightly brighter 
in color. Like most chipmunks, this species is out in greatest abun- 
dance during the early morning hours or late in the afternoon, and may 
be seen frisking about the rocks and stumps of trees on the sides of 
canyons or along fences or busily feeding in the thickets of wild cherry 
and June berry so abundant in the canyon bottoms. It is usually 
shy, and when surprised hastily takes refuge among the rocks, utter- 
ing high-pitched, chippering notes. The ordinary note, however, is 
a soft chuck, chuck, usually uttered when the animal is at a distance 
from the observer and either sitting on the summit of a large rock 
far up the canyon side or on a tree stump in the silence of the yellow 
pine forest. 

The food consists chiefly of seeds of various weeds and grasses and 
of Cercocarpus parvifolius, the fruit of the prickly pear (Opuntia), 
and also juniper berries, currants, wild cherries, and June berries. 
In autumn this chipmunk gathers a winter's supply of food, hoards 
it in crevices and under rocks, and goes into at least partial hiberna- 
tion over most of its range, according to the length and severity of 
the winter. It is often out in mild winter weather. 

Eutamias quadrivittatus animosus Warren. Las Animas Chipmunk. 
Eutamias quadrivittatus animosus Warren, Proc. Biol. Soc. Wash., XXII, pp. 
105-106, June 25, 1909. Type from Irwin's ranch, Las Animas County, 
Colorado. 

The range of tins large pale race of E. quadrivittatus is imperfectly 
known. Warren based his description upon three specimens, all 
taken by himself in the rough Upper Sonoran juniper country of 
northwestern Baca and northeastern Las Animas Counties, at eleva- 
tions of about 5,000 feet. The first specimen he secured in the 
spring of 1905, at Gaume's ranch, Shell Rock Canyon, in the north- 
west corner of Baca County. No more specimens were known until 
April, 1909, when Warren made another trip to the region and col- 
lected two more at Irwin's ranch, in Las Animas County, some 12 
miles west of Gaume's ranch. (See fig. 4.) 

In late November, 1907, I made a trip to Gaume's ranch, but failed 
to see any chipmunks, as they were already in hibernation. They 
were reported to be common, however, in the juniper country of that 
section. Although the two localities at which specimens have been 
taken are close together, it seems probable that this pale chipmunk 
will be found to occupy most, if not all, of the rough juniper and 
pinyon country of Las Animas and western Baca Counties, locally 
known as The Cedars. It is reasonably certain that animosus is 
confined to the Upper Sonoran area east of the foothills. Chipmunks 



74 NORTH AMERICAN FAUNA. [No. 33. 

which I saw at Walsenburg were fully as dark as typical quadrivittatus, 
while Trinidad specimens more nearly agree with those from Archu- 
leta and La Plata Counties in being more brightly colored than 
typical specimens and having the white dorsal stripes wider and purer 
white. 

Eutamias hopiensis Merriam. Hopi Chipmunk. 

This handsome species is represented from Colorado by a series of 
specimens from White River, near Angora; Rangely, 20 miles south- 
west; Evacuation Creek, north base of "Book Plateau; Rifle; Dotsero, 
Grand River Canyon ; McCoy; Somerset; Coventry; Uranium, Sinbad 
Valley; and Spruce Tree Cliff Ruins, Mesa Verde. 

I first met with the Hopi chipmunk among the rocky bluffs at the 
ford of White River, near Angora, September 12, 1906. At this point 
fully a dozen were actively engaged in carrying oats from a stack 
along the river and storing them in caches among the rocky ledges 
near by. Numbers were seen among the White River bluffs as far 
west as Rangely, and thence the range is continuous across the rough 
juniper country to Dragon Junction, Utah, and southeast in the 
valley of Evacuation Creek to the north base of the Book Plateau, 
at 7,000 feet. This chipmunk doubtless occurs in all of the country 
between the Book Plateau and White River and west of Cathedral 
Bluffs. On the Book Plateau it was again found on the southern 
slope near Atchee, at 7,000 feet, which is very near the upper limit 
of junipers and pinyons. Thence it ranges south to Grand River and 
northeast in the Grand Valley uninterruptedly to the western end 
of Gore Canyon. It appears not to extend through this rugged gorge, 
however, as it was not found at Kremmling, Middle Park. It is 
very abundant in the lower valley of Plateau Creek, and occurs 
sparingly on the warm juniper slopes north of the Eagle River Valley 
nearly to Wolcott, where the Transition zone country proves an 
effective barrier to its eastward extension. 

South of Grand River the Hopi chipmunk reaches its eastern limit 
at the following localities: Basalt, at the junction of the Frying Pan 
and Roaring Fork Rivers; Somerset, on the North Fork of the Gunni- 
son; Crawford, on the Smith Fork of the Gunnison; phtyon slopes 
north of Dallas Creek, 5 miles west of Ridgway; Placerville, 1 on the 
San Miguel River; Cortez, in the Montezuma Valley; and the Spruce 
Tree Cliff Ruins, at the head of Navajo Canyon, which breaks to the 
south from the Mesa Verde into the canyon of Mancos River, a tribu- 
tary of the San Juan. It was not found in the Durango region or at 
Arboles, and apparently does not follow the San Juan Valley east 
of the low divide extending south from the La Plata Mountains. 
In the warm valleys and canyons of western Mesa, Montrose, and 

1 Warren, Further Notes on the Mammals of Colorado, p. G8, 1908. 



1011.] MAMMALS. 75 

San Miguel Counties hopiensis is abundant up to 7,000 feet, and was 
noted at the following localities: Coventry; Naturita Valley, 3 miles 
east of Naturita; Tabeguache Canyon, north of Nucla; west rim of 
West Paradox Valley up to the lower edge of the yellow pines, at 7,000 
feet; Sinbad Valley and surrounding rim; Salt Canyon; Dolores 
Canyon to mouth of West Creek; and 4 miles up the valley of West 
Creek. It apparently does not extend through the northern end of 
the Uncompahgre Plateau in the Unaweep Canyon; and since at all 
other points the Canadian zone cap of this plateau is an effectual 
barrier, it must reach the Uncompahgre Valley, on the east side of 
the plateau, from the north, through the Grand and Gunnison 
Valleys. 

The Hopi chipmunk is common in western and southern Montezuma 
County, where I have seen it in the McElmo Canyon at Ashbaugh's 
ranch and Moqui, and in the southern borders of the Mesa Verde. 
At the head of Navajo Canyon these chipmunks were abundant 
the middle of June. They frequented the cliffs along the warm 
side of the canyon during the early morning hours, and were usually 
seen running along the rocky ledges in the bright sunlight in family 
groups of four or five, the young being from half to two-thirds grown. 
About 9 o'clock the chipmunks usually disappeared and were not again 
seen until the following morning. A nursing female in worn pelage 
and a male in bright fresh summer coat were collected at this point. 

Eutamias hopiensis was described from Keam Canyon, Painted Des- 
ert, Arizona, and is known also from Bluff City, Utah, and a num- 
ber of localities in the San Juan Valley of northwestern New Mexico. 
Its range has not been worked out in detail, but is known to be in the 
desert canyons and pinyon and juniper country bordering the Colorado 
River and its tributaries from northern Arizona and northwestern 
New Mexico north to northwestern Colorado, mainly in the Upper 
Sonoran zone. It occurs in Colorado at few points above 7,000 feet. 

As far as known the ranges of E. hopiensis and quadrivittatus meet 
at only one point in Colorado (see fig. 4). E. R. Warren collected 
both species in May, 1907, at Yarmany Creek, near McCoy, Eagle 
County, at an elevation of 6,900 feet, 1 and the specimens, which I 
have examined, show no evidence of intergradation. The two species 
occur in close proximity elsewhere. I have taken E. quadrivittatus 
at Sapinero and have seen hopiensis at Crawford. 

In size and general appearance the Hopi chipmunk resembles E. 
quadrivittatus of the eastern foothills, but its movements are more 
deliberate and its colors much brighter and richer. The long tail 
is carried more nearly horizontally, even when the animal is running. 
This striking habit, together with the graceful downward curve of 
the tail near the tip, serves to distinguish it, even at a distance, from 

1 Further Notes on the Mammals of Colorado, p. G8, 1908. 



76 NORTH AMERICAN FAUNA. [No. 33. 

the small K. consobrinus, with which it commingles in the higher parts 
of its range. In the juniper country south of White River its habit 
of leaping up a tree when alarmed and hiding on the opposite side 
of a branch may cause it to be confused with the gray utaJiensis, 
which is found in similar country in the Escalante Hills. The Hopi 
chipmunks appear equally at home among the hot rocks in the pre- 
cipitous canyons and in the dense juniper and piny on growth which 
clothes the bordering mesas. They feed extensively upon the berries 
of Juniperus monosperma throughout their range. 
Eutamias amcenus operarius Merriam. Colorado Chipmunk. 

Eutamias amcenus operarius Merriam, Proc. Biol. Soc. Wash., XVIII, p. 164, 
June 29, 1905. Type from Gold Hill, Boulder County, Colorado. 

This chipmunk was formerly confounded with quadrivittatus, with 
winch it occurs at many points in the eastern foothills of Colorado; 
but while the two are nearly alike in coloration, operarius is decidedly 
smaller, and its short, round skull and short nasals indicate affinities 
with a different group. The rump is usually ashy gray, while that of 
quadrivittatus is olivaceous. From consolrinus of the western moun- 
tains operarius may be distinguished by its more robust skull, although 
in summer pelage the two are sometimes almost indistinguishable 
externally. It is the highest ranging chipmunk in the State, occurring 
regularly far above timberline along the crest of the front ranges. 
It is found also as low as 6,500 feet at several localities in the eastern 
foothills, and E. R. Warren has recently shown me specimens taken 
by Robert B. Rockwell along the South Platte, 3 miles south of Little- 
ton, at about 5,300 feet. 

Although most abundant on the higher slopes of the front ranges, 
operarius is common in the mountains bordering the San Luis Valley 
and extends westward in the San Juan Mountains to Silverton and 
Lake City, where fairly typical specimens have been taken. A June 
specimen from the southeast base of Lone Mesa, west of the La Plata 
Mountains, seems referable to operarius, although the only other 
chipmunk taken at that locality is nearer consobrinus. A very 
brightly colored July specimen taken in the yellow pine country 
at the head of Dominguez Creek, on the Uncompahgre Plateau, is 
intermediate, resembling consobrinus externally. E. operarius was 
abundant in the high La Sal Mountains, Utah, west of the Paradox 
Valley, in July, 1907, at 11,000 feet, and was seen considerably above 
timberline. The chipmunks of the La Sal Mountains are entirely 
isolated from those of the Colorado mountains by a broad belt of 
Upper Sonoran desert country occupied, so far as known, by only 
E. hopiensis. 

The extent of country over which the ranges of operarius and 
consobrinus overlap has not been worked out with precision. Both 



1911.] MAMMALS. 77 

forms have been taken at Canadian Creek, at the west base of the 
Medicine Bow Range in North Park; at Coulter, Middle Park; and at 
Lone Mesa (9,000 feet), west of the La Plata Mountains. Chip- 
munks from Sapinero are typical consobrinus, while others from Lake 
City, 35 miles south, are operarius. Specimens collected above 
timberline on the Saguache Mountains at St. Elmo are, strangely 
enough, consobrinus, indicating that operarius is not found on the 
west side of the Arkansas Valley. Chipmunks taken by Warren on 
the head of Eagle River (near Tennessee Pass) are intermediate but 
nearest operarius. 

Average measurements of five skulls of adult male topotypes of 
E. operarius are as follows: Occipito-nasal length, 32; basilar length, 
24.6; zygomatic breadth, 18. Average of four skulls of adult male 
consobrinus from Canadian Creek, North Park: Occipito-nasal length, 
31; basilar length, 23.5; zygomatic breadth, 17.2. 

There are specimens of operarius in the Biological Survey collection 
from the following localities: Gold Hill, type locality; Estes Park; 
Longs Peak; Boulder, 5 miles west; Nederland; Golden; Idaho 
Springs; Cascade; Mount Kelso; Elkhorn; Livermore; Berthoud Pass; 
Canadian Creek; Coulter; Como; Lone Mesa; La Sal Mountains, Utah; 
Silverton; Lake City* Hermit; Cumbres; Antonito; and Fort Garland. 
Warren has specimens from Colorado Springs; Florissant; Tarryall 
Creek; Salida; Boreas Pass, 11,470 feet; Breckenridge ; Poncha Pass; 
Herard; Querida; Crestone; and Tercio, Las Animas County. 

Eutamias minimus (Bachman) . Least Chipmunk. 

Chipmunks from the Snake River Valley and the adjacent sage 
plains and from the Browns Park region along Green River agree 
well with typical E. minimus from Green River, Wyoming. In the 
region between Bear River and the Danforth Hills, minimus grades 
into the dark form consobrinus of the high mountainous country on 
the south and east. Four August and September specimens from 
Snake River (20 miles west of Baggs Crossing), Sunny Peak, and 
Ladore are typical minimus. Others from Lay, Axial Basin, and 
Lily show an approach to consobrinus. 

The least chipmunk is one of the most characteristic mammals of 
the sage plains and appears not to range at any point much above 
6,000 feet. During the latter part of August, 1906, numbers were 
observed in the Snake River Valley near Sunny Peak, Routt County. 
Some were in dry arroyos or on the level sage plain, but the majority 
were busily engaged in gathering a winter's supply of buffalo berries 
(Lepargyrea argentea), which were fully ripe in the dense thickets 
along the river. Near Ladore, in the Green River Valley, this chip- 
munk was often seen in the tops of Sarcobatus bushes. On the north 
slopes of the Escalante Hills it ranges up to the edge of the junipers 
and pinyons inhabited by the large gray Eutamias utahensis, but 



78 NORTH A.MEBICA3S FAUNA. [No. 33. 

apparently does not enter the timber growth. This chipmunk may 
often be seen in the top of a sagebush, but unlike E. consobrinus is 
very wild and difficult to capture. When running, it holds its tail 
at right angles to the back. The tail then appears very long, but 
this may be partly due to the small size of the body. 

Eutamias minimus caryi Merriam. San Luis Chipmunk. 

Mitamias minimus caryi Merriam, Proc. Biol. Soc. Wash., XXI, p. 143, June 9, 
1908. Type from Medano ranch, San Luis Valley, Colorado. 

So far as known, this handsome ashy gray chipmunk is restricted 
to the eastern and northern central parts of San Luis Valley, where 
it inhabits the open Sarcobatus and Chrysothamnus plains. The 
original series of specimens from the Medano Springs ranch, east of 
the San Luis Lakes, consists of 8 males and 5 females collected Octo- 
ber 24-29, 1907. In the summer of 1909 Warren found it at Moffat, 
Hooper, Mosca, and Crestone, thus considerably extending the 
known range. Chipmunks reported from the greasewood plains be- 
tween Moffat and Saguache are doubtless the same form, and sug- 
gest a very general dispersion over the central part of the valley. 
At the time of my stage trip between these two points, November 6, 
1907, the chipmunks had apparently retired to winter quarters. 
I heard of none at Alamosa or at any point in the open valley south of 
Mosca. 

This chipmunk strongly resembles E. minimus of the Routt County 
sage plains, its nearest relative, as regards both wariness and appear- 
ance. When the animal is running, the tail is carried at a right angle 
to the back, and appears much longer in proportion to the body than 
measurements show it to be. At the Medano ranch these chipmunks 
were usually seen among the low sandy hummocks and ridges bor- 
dering the meadows, where the chico brush or chaparral of Sarcobatus 
vermiculatus, Chrysothamnus patens, and Atriplex occidentalis reaches 
its maximum growth, often from 6 to 8 feet in height, and affords 
suitable homes and retreats among the gnarled and many branched 
roots and basal stems which the drifting sands have left partially ex- 
posed. Most of my specimens wer.e shot, but a few were caught in 
mousetraps. The chipmunks were out chiefly during the forenoon, 
and any time after sunrise on bright mornings could be seen in the 
tops of the chico brush busily storing the seeds in their cheek pouches, 
afterwards descending and caching them in the burrows, which were 
usually in the sand at the bases of the shrubs. One usually noted 
my approach at a distance of fully 50 yards, and, after hastily climb- 
ing to the top of a tall bush for a good look, would descend to the 
ground, sometimes silently, and again uttering excited, high-pitched 
notes. When the little fellow next appeared above the level of the 
chico it was generally in a tall bush fully 30 yards farther on. This 
maneuvering would be kept up for some time until a very large bush 



1911.] MAMMALS. 79 

would conveniently intervene and afford sufficient cover to approach 
within range. The chipmunks were silent for the most part, and the 
alarm notes were heard only a few times. The notes are not loud, 
but are very high pitched, and quite unlike those of .quadrivittatus 
and operarius of the foothills and mountains on both sides of the 
valley. The size, coloration, and habits readily distinguish this 
chipmunk from either of the above forms. Cowboys at the Me- 
dano ranch state that, in addition to the chico seeds, these chipmunks 
are very fond of the large roundish seeds of a honey plant (Peritoma 
sonorse), locally known as skunk weed, which grows rankly on the 
dry sandy ridges extending through the meadows. 

Eutamias caryi is characterized as follows (from original descrip- 
tion, 1. a): 

"Characters. — Similar to minimus but paler and grayer. In fall 
pelage (late October) pale gray, most marked on neck and rum]), 
and almost as clear en inner pair of light stripes; outer pair of white 
stripes purer white than in minimus; pale face stripes whitish, in 
striking contrast with the alternating dark stripes. 

"Measurements. — Type: Total length, 194; tail vertebrae, S7; hind 
foot, 30. Average of 10 specimens from type locality: Total length, 
194; tail vertebras, 89; hind foot, 30.2." 

Eutamias minimus consobrinus (Allen). Wasatch Chipmunk. 

This is the small, brightly colored chipmunk so abundant in the 
Canadian and Transition zones on the mountains and plateaus west 
of the Front and Medicine Bow Ranges and north of Grand River. 
It is present also on some of the plateaus and mesas of the south- 
western counties, where it frequents the oak chaparral, ranging at 
least as far south as the Mesa Verde and Ute Peak. In eastern 
Middle and North Parks and on the western slope of the La Plata 
Mountains its range meets and slightly overlaps that of operarius 
and, as noted under operarius, both forms have been taken together 
at several localities. On some of the western plateaus consobrinus 
ranges a short distance into the pinyon belt of the Upper Sonoran 
zone, where over small areas in the Grand, White, and North 
Gunnison Valleys, and in the San Miguel region, it commingles with 
the large Hopi chipmunk (E. Iwpiensis). Toward the north, on the 
sage plains of Routt County, it grades into minimus. Its center of 
abundance is in the heavy forests of the Canadian zone, but it is 
common on the high sage plains of North Park, and even ranges into 
the greasewood in some of the warm Upper Sonoran valleys in the 
extreme western part of the State. It is especially numerous in the 
Gunnison region, on the White River Plateau, and in the mountains 
bordering North and Middle Parks, except on the east side. 

E. consobrinus generally remains below the Hudsonian zone and, 
broadly speaking, it is not such a high ranging form as operarius. 



80 NORTH AMERICAN FAUNA. [No. 33. 

Near St. Elmo, in the Saguache Mountains, however, it was tolerably 
common at timberline October 9, 1907, while down in the Chalk 
Creek Valley at St. Elmo (10,000 feet) none were noted. Most of 
the chipmunks seen were running actively over the snow banks in 
the ragged growth of foxtail pines (Pinus aristata) just below tim- 
berline. Seen at a distance on the bright snow banks, with their 
tails held erect, these chipmunks are striking little objects. Three 
specimens were shot at an altitude of 12,000 feet. 

The sharp notes of this chipmunk are characteristic sounds in the 
depths of the aspen and spruce forests. Deserted cabins are espe- 
cially frequented by them, and near a camp the chipmunks soon 
become tame and unsuspecting. In the latter part of September, 
1906, from 20 to 30 could often be counted within sight of our camp 
at Baxter Pass, at 8,000 feet, on the Book Plateau, where they were 
busily gathering acorns. They were rarely observed during the 
middle of the day, being out chiefly in the early morning hours. 
In the high country the food consists largely of wild cherries, June 
berries, and snowberries (Symphoricarpos oreopliilus) . In the White 
River Valley in September they were feeding extensively upon buffalo 
berries {Lepargyrea argentea) . 

A large series of specimens from the following localities indicates 
the range of consobrinus in Colorado: Coulter; Sulphur Springs; 
Kremmling; Mount Whiteley; Arapahoe Pass; Canadian Creek; 
Pearl; Elk Head Mountains; Meeker; White River Plateau; Rangely; 
Evacuation Creek; Baxter Pass, Book Plateau; Gypsum; Sapinero; 
Somerset; Mesa Verde; and Lone Mesa. 

Allen has recorded (as Tamias quadrivittatus) a series of 16 chip- 
munks collected by W. W. Granger at Three Forks (forks of Snake 
River, near Honnold) in 1895. 1 Specimens in the Biological Survey 
collection from 5 miles south of Honnold are consobrinus, and that 
Mr. Granger's series is referable to the same form is indicated by the 
small hind foot measurements given— 29.5 to 30.7. 

Eutamias dorsalis utahensis Merriam. Utah Chipmunk. 

Fifteen specimens of this handsome gray chipmunk were collected 
in the dense juniper and pinyon growth on the northern slopes of 
the Escalante Hills, at Douglas Spring, early in September, 1906, but 
it was not taken elsewhere in western Colorado, although much of 
the region seems favorable for it. These chipmunks were abundant 
hi the heavy growth of junipers and pinyons between 6,000 and 
6,500 feet, and were found also in rocky ledges among the scattered 
yellow pines on the summits at 7,000 feet. Warren collected this 
species on Cross Mountain, east of Snake River, a few miles northeast 
of Lily, in 1907. This locality is some 20 miles east of Douglas Spring. 

i Bull. Am. Mus. Nat. Hist., VIII, p. 256, 1S96. 



1911.] MAMMALS. 81 

It is reported also from the eastern end of Yampa Canyon, a few 
miles west of Lily. 

Eutamias utahensis was described from Ogden, Utah, and has not 
been previously taken much east of Provo in that State. It is an 
Upper Sonoran form and probably ranges eastward along the southern 
foothills of the Uinta Mountains, entering Colorado in the region of 
the Yampa Plateau, south of Bear River. Future work in northeast- 
ern Utah will determine whether there is continuity of range or the 
Colorado colony is entirely isolated. 

These chipmunks at Douglas Spring were remarkably wild, and 
it required much perseverance to secure my series of specimens, all 
of which were shot. The favorite feeding time was in the early 
morning and again just before sunset, when they were usually in the 
tops of junipers busily feasting upon the berries. I seldom man- 
aged to approach nearer than 30 yards without alarming them. 
When frightened they uttered a series of high-pitched notes, and 
after a hasty descent to the ground fled precipitately, rarely stopping 
within sight. The bushy tail is very prominent and gives the animal 
the appearance of a small squirrel, and this resemblance is heightened 
by the ease and rapidity with which it climbs trees and keeps on the 
opposite side from the observer. The seeds of the juniper berry 
appear to be the chief food, and they filled the cheek pouches of the 
specimens collected. Several caches of these berries were found in 
the hollow branches of junipers in which the chipmunks appeared 
to be living. 

Callospermophilus lateralis (Say). Say Ground Squirrel. 

S[ciurus] lateralis Say, in Long's Exped. Rocky Mts., II, p. 46, 1823. Type from 
Arkansas River (a few miles below present site of Canon City), Colorado. 

The Say ground squirrel is a common and characteristic mammal 
throughout the mountains of Colorado in the Transition and Cana- 
dian zones. It is found from the eastern bases of the foothills to 
timberline and west of the main ranges occurs on all the timbered 
hills and plateaus to the Utah boundary. The Biological Survey 
collection is rich in specimens of this species from a great many 
localities over its range. C. lateralis is replaced in the rocky canyons 
and juniper country along Bear and Snake Rivers by a pale form, 
C. I. wortmani. Specimens from the Transition zone summits of 
the Escalante Hills, in extreme western Routt County, are nearest 
lateralis. Others from the White River Valley near Rangely are 
noticeably paler than typical specimens from the eastern foothills 
and approach wortmani. (See fig. 5.) 

In the foothills near Boulder C. lateralis is most abundant in the 
yellow pine belt between 5,500 and 8,500 feet, but in crossing the 
mountains from Fort Collins to North Park it was first noted in the 
90432°— 11 6 



82 



NORTH AMERICAN FAUNA. 



[No. 33. 



lower edge of the lodgepole pine belt a few miles northeast of Bald 
Mountain at 8,500 feet. From this point it was common across the 
Laramie Divide, Laramie Valley, and Medicine Bow Range to the 
edge of the sage plains of North Park. At Arapahoe Pass (Rabbit 
Ear Mountains), and on the White River Plateau, southeast of 
Meeker, lateralis was common in the aspen and lodgepole pine forests 
at 9,000 feet, and was especially numerous about clearings and 
among the charred stumps and logs in fire-swept tracts. Farther 
west, on the Book Plateau, it was common on the open rocky south 
slope as far down as Carbonera, at the edge of the Grand Valley, and 




SSSSJJJSJ CAl l OSP£/?MOPH/LUS i. WOftrMAM. WW// C. /LArfftAC/S- 

Fig. 5.— Distribution in Colorado of golden-mantled ground squirrels (genus Callospcrmophilus). 

was also observed in the valley of Evacuation Creek at the north base 
of the plateau. 

In the southern part of the State the Say ground squirrel is most 
abundant in the Transition zone — in the extensive yellow pine for- 
ests at the southern base of the San Juan Mountains from Pagosa 
Springs west to Vallecito; on the Dolores and Uncompahgre Pla- 
teaus; and at the north base of the San Miguel Mountains, south of 
Norwood. It has a wide zonal range, however, from the lower edge 
of the pines to timberline, and at Mancos, at least, is abundant in the 
upper edge of the pinyon country. One was noted on a pinyon-clad 
slope a few miles west of Ridgway, Ouray County. 



1911.] MAMMALS. 83 

This ground squirrel usually lives in rocky ledges and among piles 
of bowlders, but in the heavy forests often digs its own burrow or 
takes up its abode in a hollow log or deserted cabin. In the sandy 
yellow pine country at the head of Dominguez Creek on the Uncom- 
pahgre Plateau it appeared to be living entirely in burrows, and in 
some places in the higher parts of its range it uses so constantly the 
tunnels of the mountain pocket gopher (Thomomys fossor) that it is 
continually getting into traps set for gophers. Near Cochetopa Pass 
Loring once found one living in a burrow which it had excavated in 
the earth roof of a cabin. It is fond of sunning itself in exposed 
situations during the warmer part of the day, and may often be seen 
sitting upright and motionless on a point of rocks, tree stump, or 
ridge pole of a cabin. The animals are rarely observed on cloudy 
days, and they do not come out so early in the morning as the chip- 
munks do, but await the warming rays of the sun. Several families 
of this species were living in the slide rock near the Stevens Mill, 
at timberline on Mount McClellan, in June, 1905. Each morning 
during my stay, at about 9 o'clock, they were busily feeding, in com- 
pany with chipmunks and white-throated sparrows, on the oats left 
hi the trail in front of the mill by the ore teams. At Estes Park, in 
August, 1894, Dr. A. K. Fisher found this species feeding extensively 
upon currants (Ribes cereum). In southern Colorado acorns form a 
part of its food. Warren states that near Querida it has been seen 
to kill a young bluebird in a nest in the bank of a gulch, apparently 
with the intention of eating it. 

Near Georgetown, June 25, 1905, I saw a female and two young 
about a third grown romping among some loose rocks on the bank of 
Clear Creek. When the old squirrel first saw me she ran to the 
little ones and pushed them back into a hole among the rocks with 
her forefeet. As soon as she had left them the youngsters came out 
and began playing again. The mother returned and again pushed 
them into their safe retreat, appearing much excited at my presence. 
This was continued for a number of times, until I tired of watching 
the performance. Near Vallecito, June 5, 1907, the young were half 
or two-thirds grown, and two weeks later a great many young of 
about the same age were seen north of Dolores. 

These ground squirrels go into winter quarters during October — at 
the higher elevations early in the month. 1 At McCoy, on Grand 
River, I noted only one, October 9, 1906, and on my return through 
Middle Park and over Berthoud Pass, from October 12 to 18, saw 
none. In 1907 the latest records were: One at St. Elmo (10,500), 
October 10; one at Sapinero (7,300 feet), October 16; and one east 
of Lake San Cristobal, San Juan Mountains (9,500 feet), October 18. 

i Warren writes me that he shot one at Colorado Springs Nov. 11, 1909, which is an unusually late 
record. 



84 NORTH AM E RICAN I 'A 1 1 X A. [No. 33. 

Streator found the species had hibernated at Gold IIill 7 October 27, 
1894; while Warren states that it had disappeared for the winter at 
Crested Butte before October 8, 1905. 1 They are said to come out 
in spring before the snow is gone. On this point Warren observes: 
"While it disappears, in the Elk Mountains at least, with the first 
snowstorms in early October, it comes out in the spring before the 
snow is gone. I have known it to tunnel through 3 feet of snow to 
get to the surface. A specimen taken early in April under such 
circumstances was very fat.'.' 2 

Callospermophilus lateralis wortmani (Allen). Wortman Ground 
Squirrel. 

This pale form, described from Kinney ranch, Sweetwater County, 
Wyoming, inhabits the rough bail-land region which borders the lower 
Snake and Bear Rivers in western Routt County, and, so far as known, 
is restricted to tins low arid Upper Sonoran country and the adjoin- 
ing desert areas of Wyoming. (See fig. 5.) Specimens from the 
plateaus and mountains to the east, south, and west, in the Transi- 
tion zone, are referable to C. lateralis; but, as noted under that 
species, specimens from near Rangely, in the Upper Sonoran area 
of the W T hite River Valley, although referred to C. lateralis, are paler 
and approach the present form. 

A specimen of C. wortmani from the northern edge of Routt 
County, 20 miles southwest of Baggs Crossing, Wyoming, August 26, 
1906, was taken in a trap set for wood rats among the scattering 
junipers which clothed the steep southern face of a rocky bluff on 
the north side of Snake River. This was the only one seen on my 
journey down the Snake River Valley in 1906, and the animals ap- 
peared to be rare. Another very pale individual, which appeared 
to be this form, was seen a few days later, however, among the bluffs 
on the north side of Bear River, near the mouth of Sand Creek, a 
few miles below Maybell. Near Lily, at the confluence of the Snake 
and Bear Rivers, this squirrel was said to be abundant. Several 
specimens from the Snake River bluffs, 7 miles north of Lily, collected 
by Warren in the summer of 1907 and sent to the Biological Survey 
for identification, are clearly referable to wortmani. Warren says 
the animals were tolerably common at this point, but I saw none 
while encamped at the same place the previous year. 

Ammospermophilus leucurus cinnamomeus (Merriam). Antelope 
Squirrel. 

The antelope squirrel is found in the warm desert areas of western 
and southwestern Colorado below 6,000 feet, chiefly in the valleys 
of the streams tributary to the Colorado and Green Rivers. There 

i Mammals of Colorado, p. 241, 1900. 2 Ibid, pp. 240-241, 1900. 



1011.] 



MAMMALS. 



85 



are specimens at hand from White River, 20 miles east of Rangely; 
Rangely; Fruita; Grand Junction; Hotchkiss; Coventry, 6,400 feet; 
and Ashbaugh's ranch, near McElmo. The northward dispersion is 
limited by the Yampa Plateau, between White and Bear Rivers. 
(See fig. 6.) 

In 1906 antelope squirrels were first met with in the White River 
Valley at the ford east of Angora, and from this point to the Utah 
boundary they were common among rock ledges along the river. In 
crossing the country between White and Grand Rivers we saw them 
until we were 10 miles southwest of Rangely, and not again until 
we reached Carbonera, at the southern base of the Book Cliffs. On 
the desert between Carbonera and Mack, and thence up the Grand 




Fig. 6. — Distribution in Colorado of antelope squirrel (A mmospermophilus leucurus cinnamomeus). 

River Valley to Palisade, the antelope squirrel is an abundant and 
characteristic mammal. It was not noted at any point in the White 
and Grand Valleys above 5,500 feet. 

East of the Uncompahgre Plateau the species ranges over a con- 
siderable area in the Gunnison and Uncompahgre Valleys. Mr. C. H. 
Smith, of Coventry, states that he has seen several at the eastern 
base of the plateau, southwest of Montrose.- In the valley of the 
North Fork of Gunnison River it occurs east to Hotchkiss and 
probably to Paonia. In August, 1907, I found these squirrels 
tolerabl} r common on the open rocky slopes north of the river at 
Hotchkiss. but did not see any near Crawford, at the base of the West 



86 NORTH AMERICAN FAUNA. [No. 33. 

Elk Mountains. A skin in the National Museum, collected by Capt. 
James Stevenson September 18, 1873, labeled "Elk Mts.," probably 
came from some point at their western base. 

The antelope squirrel is rather generally distributed in the warm 
valleys of southwestern Colorado. In Montezuma County it was not 
found at Mancos, but a single individual was seen on an Atriplex flat 
at the north base of Mesa Verde, a little west of Point Lookout, 
which appears to be the eastern limit. It is abundant along McElmo 
Creek, both among the rocky ledges in the canyon and on the lower 
bordering mesas. In the region of the lower San Miguel and Dolores 
Rivers the species is restricted to the hottest valleys and slopes, and 
was noted as follows: Naturita west to Dry Creek; Salt Canyon, 
outlet of Sinbad Valley; canyon of Dolores River, Salt Canyon to 
mouth of West Creek; and West Creek Valley to 4 miles above 
mouth. It does not extend through the Unaweep Canyon to con- 
nect with its range in the Gunnison Valley. Mr. William Boren, of 
Norwood, says antelope squirrels are common in Dry Creek Basin in 
Gypsum Valley, and on lower Disappointment Creek in western 
San Miguel County. The above localities indicate a general distri- 
bution below 6,000 feet. 

Antelope squirrels frequent sandy arroyos and are striking objects 
as they frisk about in the morning sunshine with the pure white 
under surface of the upraised tail showing prominently. They are 
easily alarmed and retreat precipitately to the burrows, which are 
usually in the sandy bank of a dry desert wash or beneath sage or 
Atriplex bushes. In a few moments the animal may be watching the 
intruder from the mouth of a burrow or from behind a pile of rocks, 
but it disappears at the slightest noise or movement. At Fruita, 
Mesa County, several were living in the cemetery, and one or two 
burrows were found beside gravestones. None of the squirrels 
uttered a sound while under my observation, but one winch J. Alden 
Loring heard near Grand Junction had a note described as "loud, 
shrill, and rattling, and gradually dying out like a policeman's whis- 
tle." Mr. Loring states that the antelope squirrel has from four to 
six young in a litter. Mr. George J. Ashbaugh and other ranchmen 
between Moqui and McElmo, Montezuma County, state that antelope 
squirrels do much damage in the spring by digging up newly planted 
corn. During my stay at Mr. Ashbaugh's ranch, June 18-23, 1907, 
the young were about two-thirds grown, and proved a nuisance by 
getting into traps placed at the mouths of Perodipus burrows on the 
small sandy flats at the base of the rocky canyon walls. The species 
is more or less active in winter, as Mr. C. H. Smith, of Coventry, has 
taken it in January. 



1011.] 



MAMMALS. 



87 



Citellus variegatus grammurus (Say). Rock Squirrel. 

Sciurus grammurus Say, in Long's Exped. Rocky Mts., II, p. 72, 1823. Type 
from Purgatory River, near mouth of Chacuaco Creek, Las Animas County, 
Colorado. 

The large gray rock squirrel is common among the rock ledges of 
the eastern foothills, and also throughout the warm valleys of southern 
and southwestern Colorado north to the southern base of the Book 
Cliffs mainly in the Upper Sonoran zone. East of the mountains it is 
not known to extend north to the Wyoming line, but is found in the 
foothills west of Fort Collins. Along the eastern slope of the moun- 
tains it occurs regularly in open, rocky situations up to 7,000 feet, 




Fig. 7.— Distribution in Colorado of rock squirrel (Citellus varkgatus grammurus). 

and toward the south is occasionally found to 8,000 feet. In the 
Grand River Valley the upper limit appears to be at 6,500 feet, a little 
below the upper edge of the juniper and pinyon belt. In some of the 
southwestern valleys, however, it ranges considerably higher, and on 
the southwestern slopes of the Uncompahgre Plateau extends well 
into the yellow pine belt wherever suitable rocky ledges occur. (See 
fig. 7.) 

The Biological Survey has specimens of grammurus from Higbee, 
Otero County, winch are practically topotypes. Specimens from 
both eastern and western slopes are fairly typical, those from western 
Colorado not differing sufficiently to be referred to C. v. utdh, described 
by Dr. Merriam from Ogden, Utah. Rock squirrels which I saw in 



g8 NORTH AMERICAN FAUNA. fNo, 33. 

the Grand River Canyon above Glenwood Springs in October, 1906, 
appeared quite reddish on the back, however, in this respect suggest- 
ing the Utah form. 

Rock squirrels nearly always live in rocky situations, the ledges 
and bowlder-strewn sides of canyons, the bare rock}' - slopes along the 
base of the foothills, and the rim rock of outlying mesas and buttes 
being especially frequented. In the piny on country near Bayfield, 
La Plata County, their burrows were often found along the margins 
of fields in a nearly level country. As a rule, however, the burrows 
are located beneath bowlders at the base of a rocky canyon rim or in 
rock slides. Rock squirrels are quite shy and wary, and when one 
is surprised in the bottom of a canyon, as is often the case, it invari- 
ably runs up the slope and takes refuge among the rocks above. If 
the observer remains perfectly quiet, he may at length detect the 
animal peering silently over the top of u large bowlder, but it gener- 
ally vanishes at the slightest noise or motion. I watched one of 
these squirrels dusting itself near Bayfield. Apparently it was 
unaware of my presence and at intervals would run to a dusty spot 
in a path, throw the dust up with its fore feet, turn on its back, and 
wriggle and squirm along the ground in the greatest enjoyment. 
This performance was repeated a number of times, when suddenly 
the little fellow spied me and raced off through the brush. 

The food of rock squirrels consists chiefly of pinyon nuts, acorns, 
and juniper berries, and consequently over much of their range the 
animals do little damage. In some sections, however, they are 
reputed to show a fondness for young chickens. Rock squirrels are 
abundant in the McElmo Valley, Montezuma County. Mr. George 
J. Ashbaugh states that they destroy many apricots on the trees for 
the sake of the seeds, of which they are especially fond; 1 they eat 
holes in cantaloupes and watermelons on the vines in search of the 
seeds, which they carry into the rocks to be eaten at leisure; and they 
also dig up and eat much newly planted corn. 

While at Ashbaugh's ranch in June, 1907, I often heard the sharp 
alarm notes of rock squirrels in the orchard back of the house. Near 
Coventry in July they were feeding extensively upon pinyon nuts. 
In Grand Valley, near Glenwood Springs, in October, 1906, numbers 
Were seen in the tops of large pinyons busily feasting upon the nuts, 
and so common is this habit in that section that the animals are 
locally known as gray tree squirrels. 

There are few data at hand on the breeding of C. grammurus. An 
old female collected by Loring at Lyons, May 28, 1893, was heavy 
with three large fetuses; while the young squirrels were about half 
grown at Ashbaugh's ranch, June 19, 1907. 

1 Warren (The Mammals of Colorado, p. 103, 1910) states that a rock squirrel which was killed at Ash- 
baugh's ranch in the fruit season had 50 or more apricot pits in its pouches. 



1911.] 



MAMMALS. 



89 



Citellus elegans (Kennicott). Wyoming Ground Squirrel. 

This large gray ground squirrel is common in the mountain parks 
and on the sage plains of the northwestern part of the State. It does 
not appear to range south of the valley of Grand River, and it occurs 
east of the main ranges only in the Laramie River country. (See 
fig. 8.) It is abundant in Middle and North Parks, ranging across the 
Rabbit Ear Mountains at Arapahoe Pass (9,000 feet), and doubtless 
at other points where there is sufficient open country. In crossing 
the Medicine Bow Mountains west of Glendevey I did not meet with 
this squirrel, and assume that its range extends from North Park 
around the north end of the Medicine Bows into the Laramie Valley. 




Fig. 8.— Distribution in Colorado of Wyoming ground squirrel ( Citellus elcgans). 

A specimen in the National Museum from Cameron Pass, at 10,000 
feet, was probably taken on the North Park side. The species was 
observed in the small parks along the heads of watercourses on the 
high forested divide east of the Laramie River as high as 10,300 feet, 
and extends east to Log Cabin and Fish Creek, 1 Larimer County, 
which points are near its eastern limits. In Routt County it is 
found everywhere except in the higher Elk Head and Williams River 
Mountains, the Escalante Hills, the Yampa and O-wi-yu-kuts Plateaus, 
and the higher western slope of the Park Range. Several of these 
ground squirrels were observed about 10 miles northwest of Hahns 

> Bailey, Spermophiles of Mississippi Valley, Bull. No. i, Biological Survey, p. 60, 1S93. 



90 NORTH AMERICAN FAUNA. [No. S3. 

Peak, which appears to be near its eastern limit in the Snake River 
country. It ranges to the headwaters of Bear River, in Egeria Park, 
thence south to McCoy, on Grand River, and south across the Piney 
Divide to Wolcott. The heavily forested Gore Range, east of Egeria 
Park, apparently proves a barrier, although the species is common 
in the parks on its western slope 8 miles east of Toponas. In the 
Eagle River Valley I found C. elegans tolerably common from Gypsum 
to Wolcott, and Warren reports it present to a point 3 miles east of 
Minturn, while in the valley of White River it occurs from Buford 
west as far as Rangely. From Meeker the range is north across the 
Danforth Hills to Axial Basin, south in the valley of the Piceance to 
Rio Blanco, 1 and from Angora north to Lily, on Bear River. 

The sage flats in Middle and North Parks are densely populated 
with these ground squirrels, and ranchmen consider them very 
injurious to the cattle range and to small grain. Judging from my 
own observations the damage inflicted is by no means slight, and 
when the large territory inhabited by them is considered, it must be 
very considerable. During July I often saw numbers in the rye 
fields eating the green stalks, and not a vestige of grass remained near 
their burrows. Ranchmen in the Snake River Valley between 
Honnold and Slater claim that this species destroys fully a third of the 
rye crop, pulling down the stalks to get at the heads, and the appear- 
ance of the small fields in August, 1906, fully sustained their state- 
ment. In the summer of 1904 a 5-acre field of oats on Little Bear 
Creek, a tributary of Fortification Creek, is said to have been utterly 
ruined by ground squirrels. In North Park I often saw them in the 
hay meadows, whither they resort in the early morning, busily engaged 
in pulling down and eating the tall grass stems. 

This species hibernates very early in the autumn. In 1905 none 
were observed at Lav August 3, and Mr. A. G. Wallahan informed me 
that they retired to winter quarters about the middle of July, although 
in the lower Snake River country they usually remain active until 
about August 1. August 17, 1906, a few were out near Honnold, but 
the majority were already in hibernation. An immature specimen 
was taken in a trap set in the underground tunnel of a pocket gopher 
near Baggs Crossing August 24, but none were noted above ground. A 
very few were seen near Meeker August 12, 1905. At Sulphur Springs, 
Middle Park, Warren found these squirrels running about in fresh snow 
on stormy da} T s in April, and they probably come out of hibernation 
somewhat earlier at the lower elevations. I noted a great many bur- 
rows which had been opened by badgers, which appear to feed 
extensively upon this species — a fact well worth noting by farmers 
and stockmen. 

1 Specimen in U. S. National Museum. 



1911.] 



MAMMALS. 



91 



Citellus tridecemlineatus pallidus (Allen). Pale Striped Ground 

Squirrel. 

This is the common striped ground squirrel of the plains, but it 
is found also in some of the mountain parks on the eastern slope 
of the mountains at an elevation of over 9 ; 000 feet. It occurs also 
sparingly in North Park. It is replaced on the sage plains and in 
the mountain parks west of the Continental Divide, and also in the 
San Luis Valley, by another form of the same group — the little 
C. t. parvus. (See fig. 9.) Specimens from Loveland, Sterling, 
Pawnee Buttes, Tuttle, Eureka Hill (Cheyenne County), Las Animas, 




C. T. fiHRVVS. 'MM. C- T. PAL UDUS . 

Fig. 9.— Distribution in Colorado of striped ground squirrels ( Citellus t. parvus and C. t. pallidus). 

and other localities indicate a general distribution over the plains 
of eastern Colorado. 

At Golden the pale striped ground squirrel is common both on 
the plains and in the parks of the foothills at 7,300 feet. One was 
found dead in the trail near Elkhorn, in the foothills of Larimer 
County, at 7,500 feet. In North Park small numbers were occu- 
pying an isolated strip of sandy country east of Canadian Creek, 
at the west base of the Medicine Bow Range. C. 'pallidus has 
doubtless reached this region from the north, as the Medicine Bow 
Range is an effective barrier on the east. It was not observed on 
the sage plains of central and western North Park. It is common 
in the Wet Mountain and Huerfano Valleys, and near La Veta in 



92 NORTH AMERICAN FAUNA. [No. 33. 

the Cuch&ras Valley; and on August 23, 1007, I found it abundant 
and active on the high grassy plains of South Park near Como, at 
an elevation of 0,800 feet. It is reported also from the upper Arkan- 
sas Valley near Buena Vista, but I did not find any at Salida. The 
National Museum has a specimen from Twin Lakes. The species 
probably reaches the upper Arkansas Valley from South Park 
through some of the open mountain passes. 

Over most of its range this ground squirrel is injurious to truck 
gardens and grain fields. Its worst habit is that of digging up newly 
planted corn and eating the kernels. Near Valmont, Boulder 
County, in June, 1005, it was reported very injurious. Because 
of its depredations several farmers were obliged to plant their sweet 
corn a second time. In this section it was living in burrows along 
the grassy margins of gardens and cultivated fields. 

Citellus tridecemlineatus parvus (Allen). Little Striped Ground 

Squirrel. 

This ground squirrel, the smallest member of the group, has its 
center of abundance in the semidesert areas of western Routt and 
Rio Blanco Counties. (See fig. 0.) This region was traversed so 
late in the summer, in both 1905 and 1906, that most of the squir- 
rels were already hibernating; therefore their numbers could not be 
ascertained. The few observed were living in deserted burrows of 
white- tailed prairie dogs, with the exception of three in the White 
River Valley, which were occupying small burrows apparently exca- 
vated by themselves. The squirrels collected were sluggish in their 
movements and extremely fat, which indicated that they were ready 
for hibernation. They were usually sitting at the mouths of the 
burrows enjoying the warm autumn sunshine, but near Rangely 
one was seen in the top of a Sarcobatus bush. At Mud Springs, on 
the White River Plateau at 9,000 feet, 0. parvus was occupying 
deserted burrows of Tltomomys fossor. Two adults and two young 
were observed in a park at this point August 18, 1905. This squirrel 
is reported as common in Lily Park, at the confluence of the Snake 
and Bear Rivers; in Browns Park, near the Utah boundary; and 
on the Iron Springs Divide, between the Snake and Bear Rivers. 
In the Snake River Valley it is found as far east as the mouth of 
Four-mile Creek, where one was seen August 22, 1006. Warren 
states that he saw it at Big Beaver Creek, Rio Blanco County, in 
1907. x During my investigations six specimens were collected, 
as follows: Axial Basin, August 8, 1905; Mud Springs, White River 
Plateau, August 18, 1905; Escalante (7 miles west), August 31, 
1906; and Rangely, September 13 and 17, 1906. 

This little ground squirrel is the form occurring over most, if not 
all, of the San Luis Valley. Loring found it not uncommon at 

1 Further Notes on the Mammals of Colorado, p. 71, 1908. 



1911.] MAMMALS. 93 

Fort Garland in July, 1892, and collected three specimens, while 
Mr. J. II. Gaut secured two at Antonito, Conejos County, August 
30 and 31, 1904. Bailey has observed this ground squirrel at La 
Jara, and it was reported. to me as abundant at the Medano Springs 
ranch, near the San Luis Lakes, in 1907. While in the San Luis 
Valley in 1909 Warren found it at Mosca, San Luis Lakes, and at 
Moffat. 

Nothing is known of the distribution of 0. parvus in the region 
between the White River Plateau and the San Luis Valley aside 
from a specimen in the U. S. National Museum labeled "Elk Mts.," 
collected by Capt. Stevenson September 6, 1873. 

Citellus obsoletus (Kennicott). Kennicott Ground Squirrel. 

This is the small grayish ground squirrel inhabiting the sandy areas 
on the plains north of the Arkansas Divide. It does not appear to be 
present on the higher plains, being found principally in the valleys of 
the Platte and Republican Rivers and their tributaries. The western 
limits of range have not been ascertained. A specimen in the Colo- 
rado Historical and Natural History Society collection was taken at 
Sand Creek, near Denver, and the Biological Survey has the species 
from Greeley. Specimens from Hugo, Tuttle, Wray, Sterling, Avalo, 
and Greeley are referable to C. obsoletus, although there is consider- 
able variation. Specimens from Wray and Tuttle have the coloration 
of obsoletus, but in size approach major. The Sterling series agrees 
best with specimens of obsoletus from Cherry County, Nebr. (assumed 
to be typical), in small size and in coloration, most of the specimens 
having the characteristic coal-black edgings of the indistinct spots on 
posterior part of dorsum. 

Like most other forms of spotted ground squirrels, 0. obsoletus is 
largely restricted to sandy country. In starting northwest across the 
plains from Cheyenne Wells in 1909, I first saw the species on the level 
plain a few miles northwest of that point May 10. No others were 
found until I reached the South Fork of Republican River, where an 
adult male was shot 5 miles east of Tuttle May 18. The following 
day another was collected on the hard soil divide 2 miles south of 
Wray, and one was noted in the sand 2 miles east of the same point. 
Mr. W. E. Wolfe, of Wray, informed me that in the valley east of the 
town these squirrels are more numerous than the striped ground 
squirrels. May 24 a spotted ground squirrel was seen in the sand 
hills midway between Wray and Yuma. I saw none in crossing the 
hard soil watershed between Yuma and Sterling. The species was 
encountered 6 miles southeast of Sterling May 28, one being trapped 
in a gravelly arroyo just below the bluffs, and others were taken in 
the sand along the east side of the Platte at Sterling a day or so later. 
The only one found northwest of Sterling was dug out of its burrow 



94 NORTH AMERICAN FAUNA. [No. 33. 

on a hard soil flat along Horsetail Creek, 13 miles east of Avalo, June 
3. Prof. Lantz found these ground squirrels common south of the 
Big Sandy near Hugo. A burrow which he dug out in sandy soil 
was 12 feet in length, but at no place more than 18 inches below 
the surface. This burrow had three entrances and terminated in a 
small round chamber which contained a slight nest of grass, in which, 
instead of a squirrel, a large bull snake was coiled. 

Citellus spilosoma major (Merriam). Large Spotted Ground Squirrel. 

Spotted ground squirrels from the Arkansas Valley and southward, 
though not typical, are referred to this form. Specimens from Las 
Animas and La Junta (18 miles south) in the Biological Survey col- 
lection agree in size with typical major from Albuquerque, New Mex- 
ico, but are grayer. Others collected by Warren at Lamar and at 
Monon, Baca County, are best referred to this form. These are said 
to be the common ground squirrels in the Arkansas Valley at Pueblo. 1 
Spotted ground squirrels from the Arkansas Divide and northward 
are nearest obsoletus, indicating that the tw T o forms may intergrade 
in southern Colorado. Prof. D. E. Lantz, who collected specimens 
south of La Junta, thinks this form comes out of hibernation about 
April 20. 
Cynomys ludovicianus (Ord). Prairie Dog. 

This is the large brown prairie dog of the eastern plains of Colorado. 
There is probably not a county east of the foothills in which it is not 
present in considerable numbers, and colonies are found in some of 
the broader foothill valleys to an elevation of 6,000 feet. The western 
limit of range may be roughly indicated by Livermore, Larimer 
County; Lyons; Boulder; Rockvale, Fremont County; 2 Badito, Huer- 
fano County; and Trinidad. 

At several points in southern Colorado the range of this species 
almost meets that of the smaller C. gunnisoni, but usually the two 
species are separated by a vertical distance of from 1,000 to 2,000 
feet — ludovicianus occupying the valleys and the flat tops of the 
lowest mesas, while gunnisoni lives in the parks of the highest foot- 
hills and in the mountains. Near Badito colonies of both species 
are found within a mile or so of each other — ludovicianus occupying 
the flat along Huerfano River, and gunnisoni the open parks among 
the pinyons on the first benches south of the river and only a few 
hundred feet above it. At Gardner, 12 miles above Badito, gunni- 
soni is found in the Huerfano Valley. 

Prairie dogs are especially abundant along the Santa Fe Railroad 
between Trinidad and La Junta, and in Baca County in the extreme 
southeast corner of the State. In 1909 I found them common on the 

i Warren, Mammals of Colorado, p. 242, 1906. 

s Field Col. Mus. Pub. 115, Zool. ser., VIII, p. 181, 1907. 



1911.] MAMMALS. 95 

eastern end of the Arkansas Divide near Cheyenne Wells, and again 
from a point 15 miles northwest of Sterling west to Grover. Their 
numbers are decreasing in Yuma County and elsewhere, owing to 
rapid settlement. Numerous grass-grown mounds between Yuma 
and Sterling indicate that formerly the animals were abundant, 
but in 1009 I saw very few inhabited colonies in tins section. In the 
thinly settled grazing country dog towns often cover large areas of 
excellent cattle range and are thus a source of loss to stockmen. 

In 1802 Dr. A. K. Fisher found this species abundant at Trinidad 
in open places among the pinyons on the table mountains, and 
reports that at Las Animas nine prairie dogs were drowned out of a 
single burrow. In 1005 Prof. Lantz reported a colony at Byers in 
which albinos are common, stating that "a dozen pure white speci- 
mens can sometimes be seen at a time." Near Higbee, Otero County, 
Prof. Lantz reports that all the females had apparently given birth 
to young before April 12, 1010. In six instances the number of young 
was six; in two instances it was four. 

Cynomys gunnisoni (Baird). Gunnison Prairie Dog. 

Spermophilus gunnisoni Baird, Proc. Acad. Nat. Sci. Phila., VII, p. 334, 1855. 
Type from Cochetopa Pass, Saguache County, Colorado. 

The type of Cynomys gunnisoni was collected by F. Kreuzfeldt on 
Capt. E. G. Beckwith's expedition in 1853. The species is consider- 
ably smaller than C. ludovicianus, with very short ears and tail, the 
latter bordered and tipped with white. Like C. leucurus of north- 
western Colorado, it does not seem to be of marked gregarious habits, 
the burrows being scattered here and there over valleys, mesas, and 
even steep slopes. It is more of a mountain animal than either of 
the above species, but is equally at home in the Upper Sonoran, 
Transition, and even lower Canadian zones. 

This is the prairie dog so abundant in South Park, in parts of the 
upper Arkansas, San Luis, and Rio Grande Valleys, and in the open 
country and valleys south and west of the San Juan and La Plata 
Mountains and the Uncompahgre Plateau. It crosses the Cochetopa 
Pass into the Gunnison country and ranges as far west as the Black 
Mesa. The northern and western limits are indicated by localities 
as follows: Como (South Park), Twin Lakes, Leadville, near Crested 
Butte, 1 Black Mesa, Cerro Ridge, and the Uncompahgre Plateau. 
The ranges of gunnisoni and leucurus do not seem to meet at any 
point, although separated by only a very narrow strip of country in 
the Cimarron region. C. gunnisoni is known to inhabit the lower 
western parts of Montezuma, Dolores, San Miguel, and Montrose Coun- 
ties, and doubtless extends westward some distance into eastern Utah, 
possibly in that region also approaching the range of C. leucurus. 

i Further Notes oil the Mammals of Colorado, p. 71, 1908. 



96 NORTH AMERICAN FAUNA. [No. 33. 

Prairie dogs are reported abundant on the extensive sage flats and 
deserts south of the La Sal Mountains, but the species is not known. 

In crossing the Continental Divide between Lake City and Creede, 
C. gunnisoni was not noted south of Lake City, but was again seen 
on the flats just below Lake Santa Maria, at 9,400 feet, and it may 
occur still farther up the Rio Grande Valley. The range is continuous 
across the Poncha Pass, between the San Luis and Arkansas Valleys, 
and also across Trout Creek Pass, between the Arkansas Valley and 
South Park. To the east this species is common in the small moun- 
tain parks near Divide, Teller County, at 9,500 feet, and extends 
down the Fountain Creek Canyon to Cascade. It follows the Arkan- 
sas River down to Texas Creek and thence extends south across the 
Wet Mountain Valley to the Huerfano, and down this valley to Badito. 
It occupies the divide between the Huerfano and Cucharas Rivers, 
and from near La Vet a. apparently follows southeast around the 
Spanish Peaks, as it is again met with on the parks and plateaus of 
the Trinidad region. In open sections like South Park, it reaches 
its upper limit at between 9,500 and 10,000 feet, but the vertical 
range is governed entirely by the character of the country. Thus on 
the southern slopes of the San Juan Mountains in Archuleta and La 
Plata Counties, where there is an extensive area of 3^ellow pine forest 
between 7,000 and 8,500 feet, prairie dogs are not found so high, 
along the Rio Pinos occurring only as far up as Vallecito, at about 
8,000 feet. 

As the higher part of its range is a region of very limited agricul- 
ture and supports an abundance of wild grasses, this species can not 
there be considered very injurious, but in some of the low irrigated 
valleys of the southwestern counties it is very destructive to grain 
and alfalfa fields. On the South Park plains, just east of Como, in 
August, 1907, the heads of prairie dogs could be seen sticking out of 
the tall grass in all directions, but grass was so abundant that they 
were making no visible impression upon it. The greatest damage 
caused by prairie dogs was reported from Coventry, the McElmo 
Valley, and near Bayfield. At the last two localities the ranchmen 
successfully drown them out of all the low ground which can be 
reached with ditch water; but on the neighboring benches and mesas, 
and even on the largest dry knolls in the fields, the prairie dogs more 
than hold their own, as comparatively few ranchmen take the trouble 
to poison those left on high ground. Several ranchmen at Bayfield 
have for several years kept their land clear of prairie dogs by the use 
of carbon bisulphide or wheat soaked in strychnine, but a lack of 
cooperation permits the animals to maintain their present abundance. 
In a single pasture in the Rio Pinos Valley 181 prairie dogs are said to 
have been drowned out and killed in a single day. Old residents at 
Bayfield state that 15 years ago the animals were unknown in the 



1011.] MAMMALS. 97 

valley, but that they have gradually worked northward with the 
cultivation of the land. According to Mr. C. G. Bates, of Bayfield, 
a peculiar disease killed the larger part of the prairie dogs on Florida 
River some years ago. The affection is said to have caused excessive 
weakness, and in its later stages loss of hair. 

This species appears to hibernate throughout the winter. It was 
out in abundance at Divide, and also on the South Park plains 
between Howbert and Hartsel, October 5, 1907, and all of the ani- 
mals appeared fat, sleek, and well furred. At Sapinero two were seen 
out of their burrows October 16, but none thereafter, although I was 
in good prairie-dog country until mid-November. 

Cynomys leucurus Merriam. White-tailed Prairie Dog. 

This handsome species replaces C. ludovicianus and gunnisoni on 
the sage plains of northwestern Colorado, where it occupies much of 
the open country west of the Park and Gore Ranges and north of the 
lower Gunnison Valley. It occurs also in North Park, but I did not 
find it in Laramie Valley, east of the Medicine Bow Range, nor does 
it range across the Rabbit Ear Mountains into Middle Park and Blue 
River Valley. In Snake River Valley it is found east to Hoimold, 
and in White River Valley it is common as far up as the mouth of 
South Fork. Prairie dogs occur throughout the Bear River 
region, and follow this stream to its headwaters in Egeria Park; 
thence, sparingly, south across the divide to McCoy on Grand River, 
and again across Piney Divide to Wolcott, on Eagle River, and west 
in the Grand Valley to Gypsum. They do not extend through the 
Grand Canyon above Glenwood, nor do they pass around it, and they 
are absent from the Grand Valley between Glenwood and Grand 
Junction. On the desert areas between Grand Junction and the 
Utah boundary, prairie dogs are common, doubtless coining in from 
the west, where the range is probably continuous around the west- 
ern end of the Book Cliffs in Utah. They range from the Axial 
Basin south across the lowest passes of the Danforth Hills to the 
White River Valley at Meeker, but apparently do not cross the 
White River Plateau or its western extension, the Book Plateau, at 
any point in the State. 

Instead of extending northeast from Grand Junction in the nar- 
row Grand Valley, C. leucurus ranges to the southeast in the broad 
Gunnison and Uncompahgre Valleys, and occurs over a wide area 
between the Grand Mesa and Uncompahgre Plateau. In the Uncom- 
pahgre Valley it was noted south to a point on Dallas Creek, a few 
miles west of Ridgway. East of Montrose it was abundant along 
the railroad at Cedar Creek, and a few were seen almost to the sum- 
mit of Cerro Ridge, between Cedar Creek and Cimarron. None were 
observed at Cimarron, and the divide between the Cimarron and 
90432°— 11 7 



98 NORTH AMERICAN FAUNA. [No. 33. 

Uncompahgre Rivers appears to mark the eastern limit of range in 
this region. The species extends east along the North Fork of the 
Gunnison to Ilotchkiss and Paonia, and was abundant at the west 
base of the West Elk Mountains, between Hotchkiss and Crawford. 
The majority observed in this section were on the dry adobe flats, 
where the only vegetation worthy of mention was the prostrate, 
scrubby, desert-growing Atriplex nuttalli and a sparse growth of 
Dondia (probably D. moquini) in damp alkaline spots. 

C. leucurus is not unlike gunnisoni in size and general coloration, 
but may be readily distinguished from the latter by its white tail 
and by the broad dusky patch which covers the eye and extends down 
over the cheek. It is not extensively colonial, the burrows being 
scattered here and there over the sage plains. The burrows are 
apparently occupied for many years, and the ejected earth accumu- 
lates into very large mounds, often as much as 3 feet in height and 
8 or 10 feet in diameter. These prairie dogs are not very shy and 
often sit at the mouth of the burrow until approached within a rod. 
The usual note is a peculiar querulous cry, very unlike the short, 
sharp bark of ludovicianus. Chattering alarm notes also are occa- 
sionally heard as one walks through a colony. 

Wherever white-tailed prairie dogs live in the neighborhood of 
cultivated ground they are very injurious to green crops. Loring 
states that in the vicinity of Grand Junction the burrows are usually 
in the dry banks of irrigating ditches, and the prairie dogs inflict 
considerable damage on the adjacent truck farms by eating cabbages, 
cantaloupes, and other crops. While eating, they sit erect on their 
hind legs, but if disturbed run to the burrows, carrying the food in 
their mouths. They destroy considerable areas of range grasses and 
feed extensively in alfalfa fields and hay meadows in the river valleys 
throughout their range. 

Marmota engelhardti Allen. Engelhardt Woodchuck; Marmot. 

The Colorado marmots are tentatively referred to this species, 
which was described from Beaver Mountains, Utah. The scanty 
material at hand is insufficient for accurate comparison, and definite 
conclusions are impossible in view of the present condition of the 
group. 

The marmot is one of the most characteristic mammals of the 
mountains, and occurs from 6,000 feet in the foothills to the rock}' 
summits of the highest peaks, at over 14,000 feet. It has, therefore, 
a vertical range of about 8,000 feet. Marmots are much more abun- 
dant above than below 8,000 feet, and are especially numerous in 
the slide rock near timberline. They are reported present on the 
summit of Grays Peak, at an elevation of 14,341 feet, and have been 
observed also on the summit of Longs Peak, 100 feet lower. The 
clear, shrill whistle of the marmot is one of the few sounds that break 



1911.] MAMMALS. 99 

the silence of the high altitudes, and the large reddish-brown animals 
may often be seen sunning themselves on the warm, flat surfaces of 
rocks during the middle of the day. In the IT alms Peak region 
marmots were abundant, and were usually seen around abandoned 
prospect holes and mining shafts, while at Arkins, Larimer County, 
one was living in an old sandstone quarry. On the Bear River 
meadows east of Hayden, Routt County, in August, 1905, I saw a 
marmot run into its burfow beneath a pile of brush, while a week 
later four were observed feeding on grass in a meadow along Good 
Spring Creek, on the north slope of the Danforth Hills. I have 
found marmots more or less common at Coulter; Mount Whiteley; 
Mclntyre Creek, Medicine Bow Mountains; Elk Head Mountains; 
Mount Kelso; near Boulder; Sapinero; and Georgetown. They are 
reported common in the following localities: St. Elmo; Lone Cone; 
Pagosa Springs; La Plata Mountains, northeast of Mancos; San Juan 
Mountains, north of Vallecito; Silverton; Cochetopa Pass; Estes 
Park ; and Longs Peak. 

In Park County, in 1871, Allen found the marmot abundant from 
the Platte Valley to above timberline, and states that black speci- 
mens occur frequently in that region. 1 

Epimys norvegicus (Erxleben). Norway Rat; Wharf Rat. 

There is very little information available on the distribution in 
Colorado of this noxious species, which is usually known as barn rat 
or house rat. Doubtless by this time it has reached most of the 
larger towns at least, as Warren says it is found in Denver, Colorado 
Springs, Pueblo, and Greeley; 2 and in 1905 I was informed that it was 
common in the Boulder warehouses. Rats are uncommon away from 
towns, and I have met very few ranchmen whose buildings are 
infested with them. Mr. C. II. Smith, of Coventry, Montrose County, 
reports a very few on his ranch, so they are probably more or less com- 
mon over the western part of the State, although Warren did not hear 
of them at Grand Junction. Near Valmont, Boulder County, I caught 
a large Norway rat in a trap set at a ground squirrel burrow in a 
prairie dog town, a quarter of a mile from the nearest farm. 

Mus musculus Linnseus. House Mouse. 

I have not seen the common house mouse in the higher mountains, 
and have no information regarding its presence above the Transition 
zone. It is tolerably common in cultivated districts on the plains, 
in towns, and around farm buildings, and even in fields and meadows, 
but is not such a pest as in the older settled States. House mice are 
by no means restricted to the vicinity of railroads, for I have found 
them in sparsely settled districts 40 or 50 miles from the nearest rail- 
road. At Ashbaugh's ranch, near McElmo, Montezuma County, one 

1 Bull. Essex Inst., VI, p. 57, 1874. - .Mammals of Colorado, p. 244, 1000. 



100 



NORTH AMERICAN FAUNA. 



[No. 33. 



was taken in a tule marsh along McElmo Creek, and I have elsewhere 
observed that marshes are much frequented by this species. A 
house mouse was killed in one of the buildings at Gaume's ranch, 
northwestern Baca County, a region where ranches are many miles 
apart. Another was caught in the barn at the Medano Springs ranch', 
near the San Luis Lakes. Prof. Lantz found them abundant near 
Higbee, in southeastern Otero County. White-footed mice of several 
•species take the place of house mice over much of the timbered and 
canyon country, and in many places are associated with the latter 
about buildings and sheds. 

Onychomys leucogaster pallescens Merriam. Pale Grasshopper Mouse. 

The large pale grasshopper mouse is generally distributed over 

the LTpper Sonoran plains and deserts on both sides of the Con- 




\0/VKCf/OMKS L£OCQGAST£R BREWCAUOUS. 



^%% O.L. PALLESCENS. 



Fig. 10.— Distribution in Colorado ol grasshopper mice {Onychomys I. brcvicaudus and 0. 1, pallescens). 

tmental Divide, except in the northwest corner of the State, 
where the smaller 0. brevicaudus replaces it. (See fig. 10.) Speci- 
mens from the southwestern desert region and also from the San 
Luis Valley are larger and paler than those from the northeastern 
plains, where an approach to 0. leucogaster is suggested by the 
darker color. 

This species apparently occupies a considerable area in western 
Montezuma, Dolores, San Miguel, Montrose, and Mesa Counties, 
but its range is imperfectly worked out. The distribution in the lower 



1911.] MAMMALS. 101 

Grand River Valley appears to be limited, the only specimens thus 
far taken coming from Fruita, and the Uncompahgre Plateau must 
certainly separate the Grand Valley range from the more extensive 
area occupied farther south. The Book Cliffs limit its northward 
dispersion and separate it from brevicaudus. In the San Luis Valley 
it undoubtedly occurs throughout the Upper Sonoran zone. The 
grasshopper mice of the Upper Arkansas Valley above the Grand 
Canyon would appear to be separated from those of the eastern 
plains by the Royal Gorge, but nothing is known of the distribution 
in this section aside from the specimens taken on the sandy slopes 
along the east side of the Arkansas Valley just above Salida. Warren 
has taken 0. pattescens in the Wet Mountain Valley. It may have 
reached this high mountain valley from either the Arkansas Valley 
on the north or the Huerfano Valley on the south. 

Grasshopper mice are nocturnal. They especially frequent sandy 
areas, and are often taken in traps set at the burrows of kangaroo 
rats, ground squirrels, and pocket mice. Their carnivorous propensity 
is one of the chief obstacles the collector meets hi trapping the rarer 
desert mice, and often after nights of trapping without success he 
is chagrined to find in one of his traps the partly devoured and 
mangled remains of a rare pocket mouse. Sometimes in regions 
where grasshopper mice are plentiful, a miscellaneous catch of 
other species will be almost ruined by them. Much of the food of 
grasshopper mice consists of soft-bodied insects, such as grass- 
hoppers and crickets. The name scorpion mice, sometimes applied 
to these rodents, is due to a marked fondness for scorpions, which prob- 
ably form part of their food in Colorado, particularly in the south- 
west. Vegetable food also is eaten. At the Medano Springs ranch, 
near the San Luis Lakes, these mice proved a nuisance by eating 
carrots, potatoes, and cabbages in a vegetable cellar. 

Because of the nature of their food, grasshopper mice decompose 
much more rapidly than mice which feed chiefly on seeds and vege- 
table matter. 

The pale grasshopper mouse is represented in the Biological Survey 
collection by a large series of specimens from Loveland, Pawnee 
Buttes, Sterling, Greeley, Golden, Hugo, Limon, Loco, Canon City, 
Salida, Burlington, Las Animas, La Junta (18 miles south), Gaume's 
ranch (Baca County), Antonito, Medano Springs ranch, and Conejos 
River, in eastern and southern Colorado; and others from Fruita, 
Coventry, and Xaturita, in the southwestern part of the State. I 
have taken immature specimens at Arkins, in the foothills of Larimer 
County, and at Ashbaugh's ranch, near McElmo, Montezuma County. 
A skull in the Merriam collection is from Roggen, Weld County. 
Specimens from additional localities, Springfield and Monon (Baca 
County), Moffat, Hooper, San Luis Lakes, Crestone, and Westcliffe, 
are in the Warren collection. 



102 NORTH AMERICAN FAUNA. [No. 33. 

Onychomys leucogaster brevicaudus Merriam. Idaho Grasshopper 
Mouse. 

This small species apparently does not range south of Bear River, 
in northwestern Colorado. (See fig. 10.) On the sage plains of 
North Park it occurs as high as 8,500 feet in the upper part of the 
Transition zone, but in the region between Snake and Bear Rivers, 
in western Routt County, it inhabits a much lower altitude. It is 
usually met with in sandy strips of country, but at no point is it 
abundant, being always greatly outnumbered by white-footed mice. 
This species is represented by a series of eight specimens from Cana- 
dian Creek, east of Walden, North Park; Snake River, south of 
Sunny Peak, Routt County; and Bear River, south of Lay. 

Allen records a specimen from Three Forks (forks of Snake River), 
in northeastern Routt County. 1 Warren mentions specimens taken 
by himself at Craig, and on Snake River 7 miles north of Lily. 2 
Peromyscus leucopus tornillo Mearns. Texas White-footed Mouse. 

This form represents P. leucopus in the arid region from the Rio 
Grande at El Paso, Texas, north to the Arkansas Valley. An immature 
pair collected in the brushy bottom along the Arkansas River at 
Canon City by Loring, and specimens taken by Warren at Lamar, and 
in Baca County at Monon, Springfield, and Gaume's ranch, indicate 
the known distribution of this form within the State. Much work 
remains to be done in working out the distribution of this and many 
other small mammals on the eastern plains. Judging from its habits 
farther south, it should be found in the brushy stream bottoms of 
most of the region lying south of the Arkansas Valley and east of the 
foothills, particularly in the piles of drift and rubbish left after 
freshets. Regarding its habits in Baca County, Warren says: "The 
specimens taken in Baca County were mostly found among the sand- 
stone bluffs along the water courses, although a few were taken about 
some ranch buildings." 3 

Peromyscus maniculatus nebrascensis (Mearns). Nebraska White- 
footed Mouse. 

This is the common white-footed mouse of the Upper Sonoran 
plains region of both eastern and northwestern Colorado, as shown 
by series of specimens from a great many localities. It is found also 
in the sandy strip of country along the eastern side of San Luis Valley. 
Along the edge of the foothills and the higher western plateaus it 
grades into the dark reddish P. rufinus of the mountains. Like that 
form, it inhabits all conceivable situations, but it is normally more 

i Bull. Am. Mus. Nat. Hist,, VIII, p. 253, 1896. 

2 Further Notes on the Mammals of Colorado, p. 72, 1908. 

3 Mammals of Colorado, p. 245, 1900. 



1011.1 MAMMALS. 103 

abundant, and sometimes becomes exceedingly numerous despite 
coyotes, hawks, and owls. Such was the case on the sage plains of 
western Routt and Rio Blanco Counties in 1906; also at the Medano 
Springs ranch, in the San Luis Valley, in October, 1907, where in a 
single night 38 were caught in 1 acre by 60 traps, and their excessive 
numbers all but prevented my securing topotypes of Reithrodontomijs 
montanus. On several occasions two were taken in a small trap at 
one setting. In the Green River Valley near Ladore white-footed 
mice were often seen in the dusk darting about the bases of cotton- 
wood trees. They were everywhere a great nuisance, gnawing har- 
ness, foraging in our provisions, and, above all, depleting our much- 
needed supply of oats. In the rocky juniper and pinyon country of 
western Colorado this race is found over considerable areas with two 
other forms of white-footed mice, P. truei and auripectus, but every- 
where outnumbers them. 

Peromyscus maniculatus luteus Osgood. Yellow White-footed Mouse. 
Five specimens in the collection of the Colorado Agricultural Col- 
lege have been identified by W. H. Osgood as P. m. luteus. They 
were collected in Spring Canyon, 4 miles southwest of Fort Collins. 
Others taken by Warren at Wray, Yuma County, seem referable to 
this beautiful yellowish species. There are specimens from Haigler, 
Nebraska, just east of the Colorado line, hi the Biological Survey col- 
lection. Future collecting will doubtless show that luteus is present 
over much of the sandy area of the northeastern counties. 

Peromyscus maniculatus rufinus (Merriam). Tawny White-footed 
Mouse. 

The wide range of this form in the Colorado mountains is shown 
by a large series of specimens. The center of abundance is in the 
boreal zones on the main ranges. In the foothills east of the front 
ranges, and on the slopes of the plateaus in northwestern Colorado, 
it shades off almost imperceptibly into the pale form nebrascensis of 
the Upper Sonoran plains and valleys. Specimens from Estes Park, 
Boulder, Gold Hill, and Nederland show a decided approach to P. 
nebrascensis. P. m. rufinus is almost omnipresent in the mountains 
'and lives indiscriminately under logs and brush piles in the heavy 
forests, among rocks, in cabins along streams, in mountain bogs, and 
among the sagebrush in mountain parks. In the southwest it occurs 
at lower elevations and inhabits semidesert areas. In the sand hills 
at the 'west base of the Medicine Bow Range, in North Park, this 
mouse was abundant among the CJirysothamnus bushes. 

These mice were very numerous in the sandy flats along the eastern 
side of the Arkansas Valley just north of Salida, in November, 1907. 
Nearly all trapped at this point were in soiled pelage, the dorsal fur 
being of a dark olive-greenish cast, while the underparts were heavily 



104 



NORTH AMERICAN FAUNA. 



[No. 33. 



tinged with dirty plumbeous. Mr. J. W. Frey, of Salida, accounts 
for this soiling by the statement that the prevailing westerly winds 
carry the smoke of the Salida smelter to the east side of the valley, 
where much of the time it hangs in a dense pall, soiling the vegetation 
and, indirectly, the mice which come in contact with it. Kangaroo 
rats living on the same sand flats were somewhat smoke-soiled, but 
not to the same extent. 

P. m. ruftnus appears to be as strictly nocturnal as the other deermice. 
In the little cabin on the edge of the chfTs just above the Spruce Tree 
Cliff Ruins, at the head of Navajo Canyon, on the Mesa Verde, I 




Fig. 1 1. — Distribution in Colorado of cliff mice (Pcromyscus trim and P. nasutus). 

heard these mice squeaking about midnight of June 13. A faint 
squeak in one end of the cabin elicited answers from other parts of 
the building, and the noise was kept up for some time. 

Peromyscus truei (Shufeldt). True Cliff Mouse. 

This large-eared species has a wide distribution in the warmer parts 
of western Colorado, where in common with many other Upper 
Sonoran mammals its range is practically coextensive with the 
juniper and pinyon belt. (See fig. 11.) It has been taken also in 
the juniper country of the southeastern corner of the State, and 
Warren has it from Salida and Parkdale, in the upper Arkansas Val- 
ley. Over much of its range in northwestern Colorado this species 



1911.] 



MAMMALS. 105 



and the smaller-eared nehrascensis occur together, and in the south- 
west it is usually associated with P. auripectus and rowleyi. 

The True cliff mouse almost always lives in the cliffs and rocky 
ledges along canyons or in hollow junipers and pinyons. In the 
dense juniper growth at the northern base of the Escalante Hills it 
was very abundant and was taken in traps placed in and around 
hollow junipers. 

In the Grand River Valley it ranges east to McCoy, Eagle County. 
Although taken in Montezuma County, both on the Mesa Verde and 
in the McElmo Canyon, it does not appear to be so common there as 
auripectus. At Coventry it was abundant in rocky ledges among the 
pinyons and along the cliffs bordering Xaturita Creek, where it was 
associated with both rowleyi and auripectus. Most of those taken at 
this point in July, 1907, were from half to two-thirds grown. 

I found P. truei common in ths rocky canyons and juniper country 
near Gaume 's ranch in northwestern Baca County, and also collected 
a specimen in the rocky ledges at Rhinehart 's Stage Station, on the 
plains of southern Prowers County, 20 miles south of Lamar. Prof. 
Lantz has taken it in southeastern Otero County, 18 miles south of La 
Junta. 

Although this species almost invariably inhabits the Upper Sonoran 
zone, I have taken it once in the aspen forest in the Canadian zone. 
This was at the Club ranch, near Uncompahgre Butte, at the extreme 
head of Mesa Creek,- at nearly 9,000 feet. An adult female was 
caught in a trap set beneath a log lying in a boggy aspen thicket 
July 17, 1907. There was none of the normal environment of truei — 
rocky pinyon or juniper slopes — within a mile or within a vertical dis- 
tance of fully 1,500 feet, and this individual must have reached tins 
high elevation from the lower country by following up one of the 
numerous canyons heading on the upper western slopes of the 
Uncompahgre Plateau. 

In September, 1906, a favorable opportunity was afforded for 
observing these interesting mice on White River a few miles west of 
Rangely. My tent was pitched in a dense Lepargyrea tlricket, and a 
sack of oats proved a great attraction to white-footed mice. All 
night long they were darting noisily over the tarpaulin, climbing the 
tent walls, and investigating the provisions. In the early morning 
numbers of both P. truei and nehrascensis were seen running about, all 
intent upon transferring a winter's supply of oats from the sack to 
their nests in the thicket. I obssrved that one of the large-eared 
mice, meeting an individual of the other species, invariably retreated 
precipitately, and gained the oat sack only by the exercise of great 
caution. The True cliff mouse presents a most peculiar and striking- 
appearance in life with its large protruding eyes and immense ears. 
It is remarkably agile and takes long leaps with the greatest ease. 



106 noivni aatertcan fauna. tNo,&. 

Peromyscus nasutus (Alien). Estes Park Cliff Mouse. 

Vesperimus nasutus Allen, Bull. Am. Mus. Nat. Hist., Ill, p. 299, 1891. Typo 
from Estes Park, Larimer County, Colorado. 

This large plumbeous species has been taken at a number of locali- 
ties in the eastern foothills of Colorado, but its range has not been 
satisfactorily worked out. (See fig. 11.) While it has been found as 
high as 8,400 feet, its center of abundance appears to be in rocky 
situations along the lowest edge of the foothills. It has not been 
taken in Wyoming, but is common in many places in the foothills of 
the New Mexico mountains. It belongs to the large-eared group of 
white-footed mice and bears a close resemblance to P. truei. It may 
be distinguished from the latter species, however, by its somewhat 
smaller ears and grayer or more plumbeous coloration, while the 
skull is a little larger and has longer nasals and much smaller audital 
bullae. 

At Estes Park (the type locality), Edward A. Preble secured speci- 
mens in a rocky gulch near the summit of a hill bordering the park, 
at 8,400 feet. Dr. Fisher found the species common in crevices in 
the rocky rims of the table mountains near Trinidad. Other mem- 
bers of the Biological Survey have taken it at Boulder, Gold Hill, and 
( 'anon City. Warren reports it as common in the vicinity of Colorado 
Springs, ''both in the foothills and in the bluffs to the north and east 
of the city." 1 

Peromyscus boylei rowleyi (Allen) . Rowley Cliff Mouse. 

Specimens of this large long-tailed species from Arboles, Mesa 
Verde, and Coventry in the Biological Survey collection, and others 
from Cortez, Salida, and Irwin's ranch (Las Animas County), identi- 
fied for Warren, represent all the information available on its dis- 
tribution within the State. Noland's ranch, Utah (the type locality), 
is situated on the San Juan River within a few miles of the Colorado 
line. The Rowley cliff mouse is an Upper Sonoran species inhabiting 
rock ledges and cliffs in the juniper belt and probably occurs over 
most of the lower country in southwestern Colorado. It probably 
reaches Salida, in the upper Arkansas Valley, from the south and 
east, since Warren found it in the juniper country of northeastern 
Las Animas County. There are no specimens from the San Luis 
Valley, and hence it seems unlikely that it has reached the upper 
Arkansas region over the Poncha Pass. 

At Coventry, Montrose County, where a small series representing 
both sexes was collected in July, 1907, I found P. rowleyi and P. truei 
about equally abundant in some low rocky ledges in a dense growth 
of pinyons. A long line of traps placed along the high and nearly 
naked cliffs bordering Naturita Creek yielded fewer specimens of 
rowleyi and more of truei. Most of the cliff mice taken at this season 

i Mammals of Colorado, p. 247, 1906. 



1911.] MAMMALS. 107 

were half or two-thirds grown. Near the Spruce Tree Cliff Ruins, on 
Mesa Verde, at an elevation of 7,000 feet, an adult male rowleyi was 
trapped in a Microtias runway among the fallen leaves in an oak 
thicket. Tracks of a large cliff mouse were abundant in the fine 
chalk-like dust among all the cliff ruins examined on the Mesa Verde, 
but, as I had no time for careful trapping, it was impossible to deter- 
mine which of the two species, tinn i or rowleyi, was more abundant. 
The Rowley cliff mouse was not taken at Ashbaugh's ranch, in the 
lower McElmo Canyon, but is doubtless present there, since Warren 
has two specimens from Cortez, near the head of McElmo Canyon. 
Arboles appears to be near the eastern limit of dispersion in the 
valley of the San Juan. Traps placed among the dry rock ledges on 
the north side of the San Juan at this point, June 7, 1907, yielded 
two immature specimens and one adult. 

Peromyscus crinitus auripectus (Allen). Golden-breasted Canyon 
Mouse. 

The type locality of this beautiful mouse is Bluff City, Utah, on 
the San Juan River, 40 miles west of the Colorado line. It inhabits 
the rock ledges and cliffs in much of the rough canyon and mesa 
country of western and southwestern Colorado, in the Upper Sonoran 
zone, becoming increasingly numerous toward the extreme south- 
west. It ranges northeast in the Grand River Valley to the Grand 
Canyon, east of Glenwood Springs, but was not found on the lower 
White River or at any point north of the Book Cliffs, and this high 
escarpment doubtless forms the northern boundary of its habitat. 
This mouse is usually associated with P. truei, but does not range 
so high. It appears to be restricted to the warmest valleys below 
6,500 feet. 

I first met with the golden-breasted mouse in the canyon of Plateau 
Creek, 5 miles east of Tunnel, Mesa County, where a female was 
trapped among the rocky ledges. It was again found at three widely 
separated localities in Montezuma and Montrose Counties. A speci- 
men was secured among the Spruce Tree Cliff Ruins on Mesa Verde, 
and five specimens, representing both sexes, were collected in the 
rock ledges and cliffs bordering McElmo Creek at Ashbaugh's ranch, 
20 miles west of Cortez. At Coventry, Montrose County, P. auri- 
pectus is greatly exceeded in numbers by both truei and rowleyi. I 
trapped a single specimen at this point in the rocky cliffs north of 
Naturita Creek. Mr. C. H. Smith, of Coventry, has collected a 
number of the golden-breasted mice among the same cliffs and does 
not consider them at all rare. Warren has this species from Grand 
Junction. Thus far it has not been found at any point in the Gun- 
nison Valley, but this is doubtless owing to lack of careful collecting 
in that region. 



108 NORTH AMERICAN FAUNA. [No. 33. 

Reithrodontomys montanus (Baird) . Mountain Harvest Mouse. 

Reithrodon montanus Baird, Proc. Acad. Nat. Sci. Phila., p. 335, L855. "Vicinity 
nf the Rocky Mountains, lat. 38°." Type locality restricted by Allen (Bull. 
Am. Mus. Nat, Hist., VII, p. 125, 1895) to "upper part of the San Luis Valley." 

The type specimen of this interesting little harvest mouse in the 
National Museum was collected by F. Kreuzfeldt on Capt. E. G. 
Beckwi th's expedition to the Pacific coast in August, 1853, and 
until recently remained unique. The exact locality is indeterminate, 
but from the itinerary appears to have been on the east side of the 
San Luis Valley about one-third the distance between old Fort 
Massachusetts (near Fort Garland) and Cochetopa Pass — probably 
not far from the San Luis Lakes. 

Vernon Bailey collected a harvest mouse at Del Norte in Septem- 
ber, 1903, but it was so immature as to throw little additional light 
on the status of Baird's montanus. The Del Norte specimen was 
trapped "under a bunch of Sarcobatus at the edge of an alfalfa 
field." . 

Considerable areas of swamp land and wet meadows in various 
parts of the San Luis Valley furnish an ideal environment for harvest 
mice. When in the region, in October, 1907, in quest of specimens 
of montanus, I visited the Medano Springs ranch, 15 miles northeast 
of Mosca and a short distance northeast of the San Luis Lakes, as 
this seemed very near where the type specimen must have been 
taken. Here the extensive hay meadows and marshes along Medano 
Creek appeared well suited to harvest mice; but after trapping care- 
fully for a week in wet situations without success and taking mean- 
while one adult and three young in the rank grass beneath Sarco- 
batus bushes on higher ground, I decided that montanus must be 
sought in dry situations. Accordingly, during the first week in 
November, all the traps were placed in a grassy weed patch on a 
broad sand ridge extending through the meadows and perhaps 

6 feet above their level. Most of the specimens collected at this 
time were caught in traps placed in the dense growth of grass beneath 
the bushes of rabbit brush (Chrysoihamnus patens), which occupied 
one end of the sand ridge. A very few were taken among the scat- 
tering dry stalks of Peritoma sonorse. The series of 13 males and 

7 females collected contains only 4 or 5 adults. Doubtless many 
more of this species would have been taken had the majority of the 
Imps not been filled each night with voles and white-footed mice, 
which were exceedingly numerous in both wet and dry situations. 

This is probably an isolated intermountain species. Its known 
range is indicated by Del Norte, Medano ranch, and Crestone, Warren 
having taken it at the last-named place in early October, 1909. 
(See fig. 12.) 



1911.] 



MAMMALS. 



109 



A. II. Howell, of the Biological Survey, who has in preparation a 
revision of the genus Reithrodontomys, has compared with related 
forms the topotype series of R. montanus from Medano Springs 
ranch. The following characterization is from his manuscript: 

"General characters. — Size small (about the size of humilis); ears 
and tail short; colors pale. 

"Color. — (Fresh whiter pelage, October and November). Pinkish 
buff, clearest on sides and face, much mixed with blackish on dorsal 
surface. Black hairs, most pronounced on hinder back. Ears much 
as hi megalotis, usually clothed on inner surface with ochraceous 
buff hairs (a little darker than body hairs), sometimes, though not 




X.MZGALOT/S. ^/?.W/VH/y^. ^^ /?. OYCH£J A/£S/MSC£MS/S 

Fig. 12.— Distribution iu Colorado of harvest mice (Reithrodontomys) except R. albescens. 

always, with a distinct blackish area on lower margin. Tail dis- 
tinctly bicolor — grayish brown above, white beneath. Compared 
with megalotis, the general tone is paler and less brownish; dorsal 
area more distinct from color of sides; sides pinkish buff instead of 
ochraceous buff; tail grayer (less brownish). 

"Cranial characters. — Skull about the same size as that of humilis, 
but with narrower rostrum and interorbital region; zygomata nar- 
rowing anteriorly; nasals longer; molars heavier. Compared with 
megalotis the resemblance is very close; montanus averages smaller, 
with narrower and relatively higher brain case. 

" Measurements. — Average ol 10 specimens from Medano Springs 
ranch: Total length, 126 (118-139); tail vertebra?, 58 (51-64); 



110 NORTH AMERICAN FAUNA. [No. 33. 

hind foot, 17 (16-17.5). Average of three skulls: Occipito-nasal 
length, 20.2; breadth of brain case, 9.4: length of nasals, 7.0." 

R. montanus is a small, short-tailed species related to R. albescens 
and R. griseus, and, like these, frequents dry grassy situations. 
Compared with R. megalotis, it may be distinguished by its smaller 
size, paler coloration, and shorter ears and tail. From the pale 
grayish form albescens of the Nebraska sand hills, montanus differs 
m darker (less gray) coloration, and somewhat larger ears, while the 
skull is relatively longer and narrower and the rostrum longer than 
in albescens. 

Reithrodontomys montanus albescens Cary. Pallid Harvest Mouse. 

Four specimens of harvest mice from Loveland in the Miller col- 
lection seem referable to R. in. albescens, although somewhat inter- 
mediate in characters between albescens and griseus. The relative 
abundance and distribution of harvest mice on the plains of eastern 
and northeastern Colorado will not be known until more thorough 
collecting is done. Possibly albescens or some other form of the 
short-tailed group occurs with R. nebrascensis over much of the plains 
region, as in western Kansas and Nebraska. 

Reithrodontomys dychei nebrascensis Allen. Nebraska Harvest 

Mouse. 

The Nebraska harvest mouse doubtless occurs over most of the 
plains region east of the foothills in the northern two-thirds of the 
State. (See fig. 12.) There are specimens from Denver, Golden, 
Yalmont, Boulder (5 miles west), Loveland, Greeley, and Canon City, 
but none have been taken on the plains south of the Arkansas Divide. 
The species is usually taken in wet meadows, in alfalfa fields, or 
along the grassy margins of irrigation ditches, but may be found 
also in smaller numbers on dry uplands where there is a sufficiently 
heavy growth of grass and weeds. Harvest mice range a short 
distance into the foothills, along some of the streams, following the 
widest and warmest valleys. I trapped one specimen at Blanch- 
ard's ranch, on Middle Boulder Creek, 5 miles west of Boulder. 
It was taken in a grassy swale along the stream at 5,600 feet, in the 
Transition zone. At Wray one was seen in a tangle of willow brush 
in the boggy bed of a canyon, where in moving about in the brush 
it often wrapped its tail around a twig by way of assistance. 

Reithrodontomys megalotis (Baird). Big-eared Harvest Mouse. 

This large-eared desert harvest mouse ranges into the State from the 
west and southwest, following the low Upper Sonoran valleys. It 
is not known to occur north of the Grand River Valley. (See fig. 12.) 

At Arboles, in southwestern Archuleta County, near the New Mexico 
boundary, a female was taken in a trap set in the short grass beneath 
a dense growth of willows and buffalo berry {Lepargyrea argentea) on 



1911.] MAMMALS. Ill 

the bank of the San Juan River. Two males taken from six traps set 
in a tule marsh along McElmo Creek at Ashbaugh's ranch, Montezuma 
County, during; the night of June 19, 1907, indicate that harvest mice 
are tolerably common in that locality. I collected one at Hotchkiss 
August 9, beneath a low rocky ledge bordering a small tule marsh 
along the North Gunnison River. This specimen was nearly eaten 
up by other mice, but the skin of back and rump which I preserved 
is grayish and does not at all resemble the usual fulvous summer 
pelage of megalotis. This species is not by any means restricted to 
marshes and damp situations, as was shown by two specimens trapped 
among beds of Opuntia on a high sand knoll in the Grand Valley near 
Morris, 7 miles west of Rifle, August 13 and 14. These specimens, 
a male and a female, are gray, like the Hotchkiss example. The 
female contained six fetuses. A series of 25 specimens taken by 
Howell at Grand Junction early in November, 1895, are in early 
winter pelage; while another individual which Preble secured at the 
same locality August 25, 1895, is in the bright fulvous summer coat. 
Warren has a specimen from Cortez, Montezuma County. 

Neotoma cinerea orolestes Merriam. Colorado Bushy-tailed Wood 
Rat. 

Neotoma orolestes Merriam, Proc Biol. Soc. Wash., IX, p. 128, July 2, 1894. 
Type from Saguache Valley (20 miles west of Saguache), Colorado. 

Bushy-tailed wood rats are common throughout the mountains and 
also in the rough juniper and pinyon country of western Routt and 
Rio Blanco Counties. On the eastern slopes of the main ranges they 
rarely occur below 6,000 feet, and are largely replaced in the lower 
foothills by the smaller round-tailed N. fallax. They have been 
found from an elevation of only 4,500 feet at Grand Junction to tim- 
berline on Mount McClellan and near Silverton, thus having a verti- 
cal range of approximately 6,000 feet. (See fig. 13.) Specimens 
from the lower arid western portions of Routt and Rio Blanco Counties 
are usually much paler than those from the main mountain ranges. 

In 1905 and 1906 I observed N. orolestes in all the mountainous coun- 
try traversed. It was least common in the Middle and North Park 
region, where a few stick nests were noted in the cliffs along Grand 
River and in the Rabbit Ear Mountains, and one rat was seen in an 
old cabin near Fraser. In the rough country bordering the lower 
Snake and Bear Rivers the favorite homes of this species are in hollow 
junipers, which are often completely filled with sticks, cactus pads, 
bones, and other nest material. In this region the rats live also 
among the rocky bluffs, in hollow cottonwoods along the streams, in 
the adobe banks of arroyos, and occasionally in thickets of buffalo 
berry. One nest noted on Snake River was in a buffalo berry bush 
fully 10 feet above the ground. The few old cabins which I entered 
in the Snake River Valley below Baggs Crossing were inhabited by 



112 



NORTH AMERICAN FAUNA. 



[No. 33. 



numbers of rats, in addition to two or three species of bats and innu- 
merable white-footed mice. The cupboard, stoves, shelves, and 
l)i inks were filled with trash carried in by the rats, and an immense 
pile of rubbish had been piled up in a corner of one of the cabins. 
I often saw wood rats running about in the cabins during the day- 
lime, and on Bear River shot one among the rocks at midday. At 
one of our Snake River camps wood rats were very abundant and 
all night long could be heard climbing the tent walls, investigating 
our supplies, and running back and forth across our blankets. One 
was bold enough to nip my companion's ear. 




A/£OTOMA C.Aff/ZONAE 



ty/ty/. A/.C.O/fOCSSr£S 



£ /V.C.tf(/P/COLA. 



Fig. 13.— Distribution in Colorado of bushy-tailed wood rats (Neotoma c. arizonx, X. c. orolestes, and N. c. 

rupicola). 

Supposed nests of N. orolestes were seen at Sapinero, Somerset, and 
on the Uncompahgre Plateau in 1907, and the species was reported 
at St. Elmo, in northwestern Huerfano County, and at many other 
localities. Bailey reports it common in the cairyon of Conejos River, 
west of Antonito. It is represented in the Biological Survey collec- 
tion by specimens from a wide range of localities in the Colorado 
mountains. 

In its propensity for carrying rubbish into deserted cabins this 
species resembles other members of the genus, and it also shares with 
them the habit of hiding in its nest all manner of portable articles 
which attract its fancy. The industry of these rats is remarkable, 



1911.] MAMMALS. 113 

and a pair can accomplish a surprising amount of work in one night. 
On the lower Snake River a stick nest of considerable size was built 
near the oat sack in our camp wagon while we slept, and a wood rat 
was found early the next morning domiciled therein. 

The nests are found in a variety of situations. In the mountains 
they are usually placed in cracks and crevices of rocky ledges and 
cliffs or in deserted cabins. Miners at the Stevens Mill, at timberline 
on Mount McClellan, reported a few rats living in the mine several 
hundred feet from the entrance, and stated that the animals often 
passed them on the ladders. 

Neotoma cinerea arizonse Merriam. Arizona Bushy-tailed Wood Rat. 

Bushy-tailed wood rats from Ashbaugh's ranch (near McElmo), 
Montezuma County, and from Coventry, Montrose County, are refer- 
able to this small southwestern species. Warren has taken it at 
Cortez, Montezuma County, and at Bedrock, Montrose County, and 
it appears to be the species so common among the cliff ruins on Mesa 
Verde. It is mainly Upper Sonoran in distribution, and probably does 
not occur north of the San Miguel Valley (see fig. 13), since N. 
orolestes is the species found in the Grand Valley at Grand Junction. 
Intergradation between arizonse and orolestes probably takes place 
in the region of the Uncompahgre Plateau, but there are no speci- 
mens to decide this point. 

The Arizona wood rat is uncommon at Coventry, being greatly 
outnumbered by the little gray round-tailed N. fallax in the cliffs 
along Naturita Creek. In the cliffs- at Ashbaugh's ranch, also, I 
found both species occurring together, but here the bushy-tailed 
rats were more numerous. The nests of the two are apparently 
indistinguishable, and both species were taken in traps set at the 
same nest. One of the yellowish bushy-tailed rats was seen running 
along a rocky ledge about 4 o'clock one afternoon. Its course led 
from one sheltering rock to another and obliged it to cross short 
intervening spaces of sunlight. The rat. took advantage of all the 
shadows and dark recesses, and when forced to cross a bright open 
space its movements were so quick that the eye could scarcely follow. 
Other rats were dimly seen moving about in nests far back under 
rocks. 

Many wood rat tracks which I saw in the fine dust of the Spruce 
Tree Cliff Ruins (see PI. IV, fig. 1) on Mesa Verde, as well as fresh 
stick nests found in the neighboring rock ledges, indicated an abun- 
dance of the animals at this point — doubtless both fallax and ari- 
zonse being present. Large piles of well-preserved rat excrement 
found in many of the rooms among the ruins were apparently as old 
as the ruins themselves, being blackened with the same smoke which 
begrimed the interior of the rooms and caverns centuries ago. 
90432°— 11 8 



114 NORTH AMERICAN FAUNA. [No. 33. 

Neotoma cinerea rupicola Allen. Pallid Bushy-tailed Wood Rat. 

Eight specimens from Pawnee Buttes and the Chimney Cliffs (30 
miles northwest of Sterling) agree closely with topotypes of N. 
rupicola from Corral Draw, Big Bad Lands, South Dakota. This* 
interesting series greatly extends the known range of the species and 
gives the first record for Colorado. (See fig. 13.) 

Near the Chimney Cliffs two nursing females, an adult male in 
badly worn pelage, and two young about quarter grown, were col- 
lected June 4 to 8, 1909, on the talus slopes and rocky buttes in the 
open valley a mile or so south of the cliffs. The rats probably in- 
habit all the buttes scattered here and there over the upper drainage 
of Horsetail Creek east to within 15 miles of Sterling. They appear 
to be uncommon, however, among the precipitous white badland 
cliffs which form the northern boundary of the Horsetail Valley. 
This may be because there are fewer natural crevices and recesses 
for nesting sites in the soft white rock of the badland formation. 
Rat nests were numerous beneath the bowlders and rock ledges on the 
buttes in the valley, as they were also in similar situations at Pawnee 
Buttes, some 25 or 30 miles southwest. The species appeared in- 
active at both localities, however, and I found it difficult to secure 
specimens. The two nursing females from the Chimney Cliffs are in 
fresh light yellowish summer pelage, while all the males taken are in 
worn and faded whitish winter pelage. 

N. rupicola is the palest known member of the genus. It is a small 
form of the bushy-tailed group, inhabiting the rough areas of badland 
bluffs and buttes from western South Dakota south to northeastern 
Colorado, in the Upper Sonoran zone. I secured it in the bluffs along 
Laramie River at Uva, Wyoming, and it occurs also in the rocky 
bluff region of extreme western Nebraska. 

Neotoma floridana baileyi Merriam. Bailey W'ood Rat. 

This is the only representative of the Neotoma floridana group in 
Colorado. It enters the State from the east, in the valWs of the 
Republican and Arkansas Rivers and tributary streams, and ranges 
westward along the Arkansas to Pueblo and farther north to Flagler 
and Wray. (See fig. 14.) It is restricted to the Upper Sonoran zone. 

At Olney, in December, 1894, Mr. C. P. Streator found this wood 
rat abundant in thickets of tree cactus (Opuntia arborescens) and 
collected a series of 12 specimens. He was informed by stockmen 
that a few are found 10 miles north of Arlington, Kiowa County, 
and also along the Arkansas River south of Chivington. Concerning 
the distribution of N. baileyi in the Arkansas Valley, Mr. Streator 
reports that it "evidently occurs wherever there are rocks, tree 
cactus, or hollow trees." He found it common also in the cliffs 
bordering the valley of the South Fork of the Republican River near 



1911.] 



MAMMALS. 



115 



Flagler. It is abundant in the sandstone ledges bordering the South 
Fork Valley in the vicinity of Tuttle, where I captured specimens, 
and found the nests composed of sticks, dried cow manure, bones, 
and rubbish of all sorts, with cactus and yucca spines for protection. 
Prof. D. E. Lantz reports pack rats common in hollow cottonwoods 
along the Big Sandy near Hugo, and I found old nests among the 
rocks on the northern face of Cedar Point, northwest of Limon. 
This should be the form represented at both the above localities. 
At Wray, where I secured specimens, this rat was inhabiting sand- 




YsM.FXLLAX. 



\A/.DES£KTORUM. ^^.W.frBA/L£Y/. 



-A/.A.WAXREA//. 



/VA7. C/W£SC£A/S. 



Iig. 14. —Distribution in Colorado of round-tailed wood rats ( Ncotoma fallax, N.descrtorum, JV. f.bailcyi, 
N. a. warrcni, and JV. m. canescens). 

stone ledges along the valley, much as at Tuttle. There is a specimen 
from Pueblo, and the rats are reported by Mr. H. W. Nash as com- 
mon at that point. 1 ' The American Museum of Natural History has 
a specimen from Fort Lyon, on the Arkansas River, collected by 
Capt. P. M. Thorne February 4, 1885. This specimen has been 
recorded as N. campestris. 2 

Neotomaf. baileyi is a large, long-tailed, gray form of the round- 
tailed group, and quite unlike any of the other Colorado species. 

Neotoma micropus canescens iVllen. Hoary Wood Rat. 

The hoary wood rat has been found only in the southeast corner of 
the State, in southeastern Otero County and in Baca County, this 

i Mammals of Colorado, p. 247, 1906. 

2 Allen, Bull. Am. Mus. Nat. Hist., VI, p. 322, 1894. 



HG NORTH AMERICAN FAUNA. [No. 33. 

being the northern known limit. (See fig. 14.) A specimen taken 
by Warren at Monon, and sent to the Biological Survey for identi- 
fication, proves referable to this form. Mr. Warren found this wood 
rat also at Springfield, and states that he found it living chiefly 
among rocks, but that some were living in the buildings of an 
unoccupied ranch. Near Higbee, Otero County, where Prof. Lantz 
secured a specimen in April, 1910, these rats were living among 
rocks and in hollow junipers on the mesas and in clumps of tree cactus 
(Opuntia arborescens) in the open Purgatory Valley. On the deserts 
of western Texas and over the greater part of its range the species 
lives in open country, the nests being usually in bunches of cactus 
or other spiny shrubbery. 

Neotoma albigula warreni Merriam. Warren Wood Rat. 

' Neotoma albigula warreni Merriam, Proc. Biol. Soc. Wash., XXI, pp. 143-144, 
June 9, 1908. Type from Gaume's ranch, northwest corner of Baca County, 
Colorado. 

This medium-sized wood rat closely resembles N. m. canescens in its 
gray coloration, but its relationships are with albigula. It inhabits 
the rough juniper country of southeastern Colorado, and also occurs 
in northeastern New Mexico, where it doubtless grades into A T . 
albigula. It is abundant hi Shell Rock Canyon, in extreme north- 
western Baca County, and Warren found it at Irwin's ranch, in eastern 
Las Animas County. It may be the species inhabiting the rock 
ledges at Rhinehart's stage station, 20 miles south of Lamar, in 
southern Prowers County, but no specimens were taken at that point. 
Doubtless it occurs throughout the juniper country of western Baca 
and eastern Las Animas Counties. (See fig. 14.) 

At Gaume's ranch, in Shell Rock Canyon, I found these wood rats 
living among rocks along the canyon walls or in hollow junipers on 
the upper rims of canyons, and occasionally in large stick houses 
reared against the bases of junipers in the dense growth well back 
from the canyon rims. Whether among the rocks or in the 
junipers, the nests were fortified with a varied assortment of spines 
and thorns, the sharp spiny bundles of the tree cactus (Opuntia 
arborescens) always predominating. The stick houses averaged about 
2 feet in height and often contained several bushels of dead juniper 
branches. Judging from the signs observed at the nest entrances, 
the rats were subsisting largely at this time of year (November) upon 
the berries of Juniperus monosperma. 

According to Mr. E. J. Gaume, these rats seldom take up their 
abode in abandoned houses as do the "short-tailed blue rats" (un- 
doubtedly referring to N. m. canescens) of eastern Baca County. Re- 
garding the habits of this species at Gaume's ranch, Warren says: ' ' It 
did not seem to breed as early as N. micropus [N. m. canescens] at 
Motion, for half-grown young of the latter species were taken the 1st 



1911.] MAMMALS. 117 

of May, while the present species was apparently just beginning to 
breed after the middle of the same month." 1 

Neotoma fallax Merriam. Gale Wood Rat. 

Neotoma fallax Merriam, Proc. Biol. Soc. Wash., IX, p. 123, July 2, 1894. Type 
from near Gold Hill, Boulder County, Colorado. 

The type specimen of this small gray round-tailed wood rat was 
taken by the late Mr. Denis Gale on a tributary of Boulder Creek, 
near Gold Hill, at the extreme upper limit of the range of the animal. 
Other specimens in the Biological Surve}^ collection are from the 
following localities, chiefly in the lower eastern foothills: Boulder; 
Arkins; Loveland; Canon City; Trinidad; Fisher Peak; Martinsen, 
Las Animas County; Arboles; near McElmo; and Coventry. Spe- 
cimens from Colorado Springs, Spring Canyon (near Fort Collins), 
and Grand Junction have been sent to the Biological Survey for 
identification. Other localities from which Warren has specimens are 
Cortez, Salida, Van Andert's Spring (on Little Fountain Creek, south- 
west of Colorado Springs), and Howard. 

N. fallax is common among the eastern foothills north across the 
State as far as Fort Collins. It occurs from the lowest outlying 
talus slopes at the edge of the plains to an altitude of about 7,500 feet, 
mainly in the Transition zone. (See fig. 14.) In the Canadian 
zone it is replaced by the larger bushy-tailed N. orolestes. The two 
species occur together at a number of localities (including the type 
locality of N. fallax) between 6,000 and 7,000 feet; and at Blan- 
chard's ranch, 5 miles west of Boulder, both have been taken at 
5,800 feet. Throughout this region of vertical overlapping of ranges 
(about 2,000 feet), however, fallax is the commoner species. Warren 
has taken both species at Grand Junction, below 5,000 feet. 

From Trinidad this species probably extends south in the Upper 
Sonoran and Transition zones around the southern ends of the Culebra 
and San Juan Ranges, thus reaching southwestern Colorado from 
the south, rather than from the east through the lowest mountain 
passes. At Pagosa Springs, Arboles, and Bayfield it inhabits rock 
ledges — at Pagosa Springs in gulches in the yellow-pine forest and 
at the last two localities down among the pinyons. It is also com- 
mon with N. arizonse in the cliffs along McElmo Creek at Ashbaugh's 
ranch, Montezuma County, and on Naturita Creek at Coventry, at 
the last locality being occasionally found about ranch buildings. 
This seems to be the species commonly known in the lower parts of 
western Mesa, Montrose, and San Miguel Counties as the sleek-tailed 
rat. Fresh nests were found beneath the rocky ledges on Dry Creek, 
west of Naturita, and others in the Sinbad Valley, near Uranium. 
A very few nests seen in the bluffs along the North Gunnison River 

i Mammals of Colorado, p. 248, 1906. 



118 NORTH AMERICAN FAUNA. [No. 33. 

near Hotchkiss appeared to belong to fallax rather than to orolestes 
It is not known from the San Luis Valley, but careful collecting on 
the lower slopes of the bordering mountains may prove its presence 
in that region. 

In its habits fallax does not differ materially from the other wood 
rats. It is equally at home in deserted mines, prospect holes, cabins, 
and among rocks; while a talus slope seems peculiarly attractive to 
it. The type specimen was taken in a stamp mill. Near Canon 
City Bailey found several nests in hollow junipers, and in south- 
western Colorado I have often seen nests in junipers and in pinyons. 
The piles of sticks, stones, and other trash accumulated by these rats 
among the rocks and on talus slopes are not so large as in old cabins. 
The stove in a cabin which I entered near Gold Hill was completely 
filled with sticks and rubbish, while the nest in the oven was com- 
posed of softer materials. Among the varied assortment of nest 
materials used by this wood rat may be mentioned juniper branches, 
sticks, stones, bones, dried horse dung, pads of prickly pear cactus 
(Opuntia), and the branches of a variety of thorny and spiny plants, 
such as roses, haws (Cratsegus) , Ceanothus fendleri, and yuccas. The 
inner nest is often composed of the soft inner bark of the juniper. At 
Walsenburg and Badito, in Huerfano County, and also in the bluffs 
along the San Juan River at Arboles, these rats often construct theii 
nests entirely of the spiny branches of tree cactuses (Opuntia arbores- 
cens and 0. whipplei). 

A female wood rat which I surprised in her nest among the rafters 
of a cabin near Boulder, July 23, 1906, glided away like a shadow in 
the semidarkness, with two young about quarter grown clinging to 
her teats. Her movements were rapid but perfectly noiseless. A 
female collected by Loring at Loveland contained three small fetuses 
May 1, while an immature individual from near McElmo, June 18, 
is about two-thirds grown. Acorns, pinyon nuts, and juniper berries 
are the chief food of fallax hi southern Colorado, judging from the 
quantities found in and around some of the nests. 

This wood rat appears to be chiefly nocturnal, but as we rode down 
the bed of a dry desert arroyo north of Nucla, Montrose County, on 
a hot day in July, one was seen dodging about among the bordering 
rocks at midday. 

Neotoma desertorum Merriam. Desert Wood Rat. 

The small desert wood rat reaches Colorado in the extreme lower 
White River Valley, in the vicinity of Rangely (see fig. 14), where it 
meets the range of the larger, bushy-tailed N. orolestes. A number 
of the nests of desertorum were found on the first bench south of Wliite 
River, 5 miles west of Rangely, at 5,300 feet, where four of the rats 
were collected September 15 and 16, 1906. These nests, or houses, 
were 2 or 3 feet in height, and were constructed of dried cow manure 



inn.] 



MAMMALS. 



119 



and the pads of prickly pear. Scattered here and there over the 
cactus flat they looked curiously like muskrat houses on a marsh. 
(See fig. 15.) The rats were trapped at the entrances to the nests, 
of which there were usually two above ground, and often a third 
opening beneath a bush a yard or so distant. The largest of the 
houses were in thickets of Atriplex confertifolia, but the smaller struc- 
tures were reared in bunches of Opuntia. Specimens of N. desertorum 
and orolestes were taken within a few feet of each other, the latter 
species living in the banks of the dry arroyos which border the flat 
occupied by the small desertorum. 

I did not find the species in the desert areas of southwestern Colo- 
rado in 1907, although much of the region seems suitable. 

Phenacomys oro- 
philus Merriam. 
Mountain Phen- 
acomys. 

An adult female 
Pit t itacom ys which 
Warren collected 
December 11, 1906, 
at Lake Moraine, El 
Paso County, and 
forwarded to the 
Biological Survey 
for identification, 
proves referable to 
P. oropMlus of the 
northern Rocky 
Mount ains, and is 
quite distinct from P. preblei, taken on Longs Peak. The Lake 
Moraine specimen differs from the type of preblei in its larger size, 
more robust skull, and particularly in its much grayer pelage. The 
fur is full and long and of a beautiful frosted gray color, while the feet 
and tail are almost white. The measurements are: Total length, 140; 
tail vertebra?, 30; hind foot, 18. Warren secured this specimen near 
Ruxton Creek, at 10,250 feet, " in a hole near creek bank." An imper- 
fect skull in the National Museum which belongs to an alcoholic speci- 
men (Xo. 59691, Fairplay, Park Co., July 11, 1873) is large and robust 
as compared with the skull of the type of P. preblei, and accords well 
with the. Lake Moraine specimen. The skin has lost its color value 
through long immersion in alcohol. The altitude of Fairplay is about 
9,900 feet. ' 

Little is known of the habits of these interesting sliort-t ailed mice. 
The}' appear to be very rare in Colorado and restricted to the moun- 




Fig. 15.— Nest of Ncotoma desertorum o.i Atriplex flat near Rangely. 



120 NORTH AMERICAN FAUNA. [No. 33. 

tains above 9,000 feet. At present they are known only from the 
higher eastern slopes of the Front Range, in the Canadian zone. 

Phenacomys preblei Merriam. Preble Phenacomys. 

Phenacomys preblei Merriam, Proc. Biol. Soc. Wash., XI, p. 45, March 16, 1897. 
Type from Lilie Mountain, near Longs Peak, Colorado. 

The type of this small pale ochraceous species, an adult male, has 
until recently remained unique. It was collected August 12, 1895, 
by Mr. Edward A. Preble on Lilie Mountain (known also as Twin 
Peak), not far from Longs Peak. Preble states that the exact local- 
ity was on a southwest slope of the mountain not far from Lamb's 
ranch, at an elevation of approximately 9,000 feet. The specimen 
was taken in a trap set for white-footed mice among fallen logs on a 
dry slope supporting scanty vegetation. This slope had been cov- 
ered by trees, most of which had fallen. 

Another specimen, an adult female, from North Boulder Creek, 
near Nederland, has been recorded recently by Mr. R. T. Young, 1 who 
secured it on the Silver Lake trail September 19, 1900, at an elevation 
of between 9,000 and 10,000 feet. This specimen, which is hi the 
Philadelphia Academy of Natural Sciences, I have been permitted 
to examine through the courtesy of Mr. Witmer Stone. It is a well- 
made skin and has a perfect skull, which is more robust than the skull 
of the type of P. preblei, with the interorbital ridges well developed 
(inclosing a median sulcus). The skin measurements are: Total 
length, 143; tail, 33; and hind foot, 18; as compared with 130-30-17 
of the type of P. preblei (a male), and 146-38-19 of the type of P. 
oropJiilus (a female). In coloration the Nederland specimen is nearest 
P. preblei, it being much too yellowish for oropliilus, with the back 
and sides strongly suffused with ochraceous. 

Evotomys gapperi galei Merriam. Rocky Mountain Red-backed 

Mouse. 

Evotomys galei Merriam, N. Am. Fauna No. 4, p. 23, pi. II, fig. 3, 1890. Type 
from Ward, Boulder County, Colorado (9,500 feet). 

In Colorado red-backed mice are restricted to the boreal forests of 

the high mountains. Their center of abundance is in the lodgepole 

pine forest belt of the Canadian zone, but they range up to at least 

11,000 feet, in the Hudsonian zone, well within the Engelmann spruce 

belt. Cold, mossy, heavily forested slopes having a great deal of 

fallen timber are most frequented by red-backed mice, which live 

beneath the rotten logs and brush piles. They seem to be largely 

nocturnal, although one July afternoon I saw one running along 

under a log at Petersons Lake, west of Eldora, Boulder County. In 

most parts of the mountains they are difficult to trap owing to the 

greater abundance of field mice and white-footed mice, which get 

» Proc. Acad. Nat. Sci. Phila., p. 400, 190S. 



1011.] 



MAMMALS. 121 



into the traps early in the night. For this reason it is often impos- 
sible to judge whether red-backed mice are common or rare, but it 
seems certain that in most sections they are greatly outnumbered 
by other species. At the type locality, however, Preble found them 
abundant and caught nothing else in his traps. 

During the field seasons of 1905 and 1906 eleven specimens were 
taken at four widely separated localities in northern Colorado. In 
July, 1905, five were trapped beneath fallen logs in dry lodgepole pine 
forest at Coulter, Middle Park (8,500 feet); three more were caught 
August 16 in a thicket of balsam (Abies lasiocarpa) in the aspen forest 
on the White River Plateau, 25 miles southeast of Meeker, at the same 
altitude; August 10, 1906, an immature male was secured in a bog 
near Pearl, in the northern end of North Park; and in late September 
three others were captured in the Douglas spruce forest at Baxter 
Pass, on the summit of the Book Plateau, within a few miles of the 
Utah boundary. In October, 1907, I found Evotomys tolerably 
common in mossy spruce woods and thickets near St. Elmo in the 
Saguache Mountains, at 10,000 feet. Most of the individuals trapped 
at this point were half or two-thirds grown. Other specimens are 
from Gold Hill, Longs Peak, Lake City, and Silverton. 

The type of E. galei, which is in the Merriam collection, was taken 
by Mr. Denis Gale at Ward, July 13, 1889. The National Museum 
has specimens from the Silver Lake trail, near Nederland, Boulder 
County; and others labeled Twin Lakes and Del Norte. 1 Warren 
records this species from Crested Butte and Irwin, Gunnison County; 
Divide, Teller County; Lake Moraine, and mountains near Colorado 
Springs, El Paso County. 2 

Microtus pennsylvanicus modestus (Baird). Saguache Meadow Mouse. 

Arvicola modesta Baird, Mamm. N. Am., p. 535, 1857. Type from "Sawatch 
Pass, Rocky Mountains" (Cochetopa Pass, Cochetopa Hills, Colorado). 

This large, dark meadow mouse is common on the plains at the 
eastern base of the foothills and in the San Luis Valley region of 
southern Colorado, but is apparently rare in the northern mountains. 
Small colonies are found in marshes at Golden and Valmont, and at 
Medano Springs ranch, near the San Luis Lakes, the species was 
extremely abundant, not only in marshes and wet meadows, but 
also on grassy uplands and wherever any cover was afforded. A 
small colony of meadow mice found in an alkaline marsh near the 
mouth of Four-mile Creek, Routt County, and others reported in 
the McElmo Valley, in western Montezuma County, may have been 
of this species, but no specimens were secured at either locality. 

In August, 1892, Loring collected a large series of topotypes at 
Tevebaugh's ranch, 9 miles south of Cochetopa Pass, where he found 

i Probably from some point in mountains near Del Norte. 
2 Mammals of Colorado, p. 219, 1900. 



122 



NORTH AMERICAN FAUNA. 



[No. 33. 



well-beaten runways ramifying through the tall grass near a creek. 
Six, seven, and even eight young were found by Loring in single 
nests, which were composed of fine dried grass and were said to 
resemble the nest of the ovenbird. Bailey found this meadow 
mouse very abundant in both dry and wet meadows along the Rio 
Grande at Del Norte, where it was doing slight damage in clover and 
alfalfa fields. At Wray, 1 in May, 1909, these meadow mice were 
feeding extensively upon the blossoms and leaves of the false Solo- 
mon's seal (Vagnera stellata), and fragments of the blossoms of a 
species of Senecio were also found scattered along the runways. 




^^ M/CffOTOS (LAGVfiOSj PAUP£Aff//*t/S. 'W/yM/CPOTUS OCHPOG/1S7E?? WWfiV/ 

Fig. Mi.— Distribution in Colorado of Ilayden and pygmy field mice (Microtus paupcrrimus and M. ochro- 

gaster haydeni). 

In addition to the series mentioned above, the Biological Survey 
has specimens of M. modestus from Denver, Loveland, Antonito, 
and Fort Garland. A specimen in the National Museum is from 
Twin Lakes; some from Fort Collins have been identified for S. 
Arthur Johnson; and Warren has others from Colorado Springs, 
Westcliffe, and Divide. The species is recorded from Estes Park. 2 

Microtus ochrogaster haydeni (Baird). Upland Mouse. 

The upland mouse has been taken in Colorado at only a few locali- 
ties along the eastern base of the foothills, and at Wray and Tuttle, 

' Several specimens collected in the cool canyon bogs at Wray are abnormally dark, but can be matched 
by a few specimens from elsewhere in the range of modestus. 
a Mammals of Colorado, p. 250, 1906. 



1011.] MAMMALS. 123 

near the eastern border. (See fig. 16.) It is doubtless present, how- 
ever, over most of the dry uplands and prairies east of the moun- 
tains and north of the Arkansas Valley. Unlike M. modestus, which 
inhabits damp meadows and marshes almost exclusively, haydeni is 
usually, though not always, found on high grassy plains. On my 
trip across the plains from Cheyenne Wells to Sterling in 1909, I col- 
lected a young individual at Wray in a cold bog inhabited by a large 
colony of modestus; while at Tuttle four upland mice were taken in a 
marsh along the South Fork of the Republican River, and signs of them 
were seen on a grassy flat on the dry upland 8 miles south of Seibert. 
Other localities from which the Biological Survey has specimens are 
Loveland, Canon City, and Fort Collins. The species is reported 
from Greeley. 1 

Microtus mordax (Merriam). Rocky Mountain Field Mouse. 

This large, long-tailed field mouse has been taken from an eleva- 
tion of 4,600 feet (Grand Junction) to considerably over 12,000 feet 
(Mount Kelso), 2 and thus has the widest vertical range among the 
Colorado species of Microtus. It is abundant throughout the Cana- 
dian and Hudsonian zones of the mountains, and follows down cold 
streams in places through the Transition zone. It is of little eco- 
nomic importance, owing to the very limited agriculture carried on 
within its range. 

This species is fond of forests and of cool, damp situations where 
the vegetation is rank. On Mount Kelso, near Grays Peak, several 
large colonies were discovered in dense thickets of alpine willows 
(Salix cMorophylla and S. glaucops) considerably above timberline. 
Most of my specimens were taken under logs in heavy forests or in 
cold mountain bogs grown up with willows. In many localities the 
field mice were so abundant that all the traps would be thrown by 
them during the early evening hours, and thus my chances for the 
rarer mammals were spoiled. I found a small colony at Rangely in 
a marsh bordering White River at 5,300 feet, and another on a small 
tributary of the San Miguel River near Coventry at about 5,500 feet. 
Tins species is represented in the Biological Survey collection by speci- 
mens from a wide range of localities in the mountainous parts of the 
State. 

Microtus nanus (Merriam) . Dwarf Field Mouse. 

In Colorado this small field mouse is largely an inhabitant of the 
boreal zones and is common far above timberline on some of the 
mountain ranges. It is especially numerous in the high mountain 
parks of the northern two-thirds of the State, in dry, grassy sage- 

i Mammals of Colorado, p. 251, 1906. 

'Warren (The Mammals of Colorado, p. 100, 1910) states that he has seen a specimen from the Summit 
llou.se on Pikes Peak, at 14,147 feet. 



124 NORTH AMERICAN FAUNA. [No. 33. 

brush country. It does not appear to enter the surrounding forests 
to any extent, but occasionally may be taken along streams and in 
mountain meadows. I found several large colonies on the Front 
Range just south of James Peak, at an elevation of 12,500 feet, 
in October, 1906. An interminable labyrinth of runways extended 
through the moss in copses of an alpine willow (Salix glaucops), and 
many of the field mice could be seen darting from cover to cover. 
This species was very abundant in the spruce belt on Mount Kelso, 
at about 11,000 feet, and in places the mossy slope was almost honey- 
combed by the burrows. It was impossible to catch pocket gophers 
(Thomomys fossor) on Mount Kelso because of these mice, which used 
the gopher tunnels and were continually getting into the traps. 

M. nanus is found as low as 6,000 feet in the meadows along White 
River, east of Meeker, and also near Coventry, Montrose County. 
In the last locality it is very abundant in the irrigated alfalfa fields, 
where it is considered very injurious. In the Hahns Peak region 
and in the parks on the western slope of the Gore Range these field 
mice were frequently observed in their runways in the daytime. At 
Coulter, Middle Park, I often watched them feeding on grass among 
the willows before the tent in the middle of the forenoon, and they 
seemed quite oblivious of my presence. On the grassy plains east 
of Como, South Park, this field mouse was using the abandoned 
burrows and tunnels of Thomomys fossor to a large extent. 

Mr. Morris M. Green writes of a colony winch he observed at Al- 
mont (8,000 feet), Gunnison County, in August, 1909: "There was 
a small colony of these animals in a little marsh, supporting a luxuri-* 
ant growth of young cottonwoods, spearmint, and succulent, tender 
swamp grasses. Their runways in several places had little piles of 
tender grass stalks, cut to a length of 2 or 3 inches, which the little 
animals could doubtless manipulate easily in their paws. Two 
adults and three half- grown young were trapped. One runway led 
to a ball-shaped nest of dead grass under a dead log. The nest had 
not been frequented lately, and was doubtless a winter abode." 

There are specimens from other localities as follows: Arrowhead, 
Mount Whiteley, Arapahoe Pass, Rabbit Ear Mountains, Estes Park, 
Cochetopa Pass, and Ruby Lake. Bailey records the species from 
Twin River and Twin Lakes. 1 Specimens taken by Warren at Crested 
Butte and Irwin, Gunnison County, have been identified by the Bio- 
logical Survey. 

Microtus (Lagurus) pauperrimus (Cooper) . Pygmy Field Mouse. 

In July, 1905, while traveling across the sage plains of eastern 
North Park, I detected evidences of a small species of Microtus 
(presumably a Lagurus) and later secured three specimens in the 
sand hills at the west base of the Medicine Bow Range, east of Walden. 

i N. Am. Fauna, No. 17, p. 31, 1900. 



1911.] MAMMALS. 125 

In 1906 an adult female was secured at the same locality; six speci- 
mens were collected at Elk Springs, 8 miles south of Lily, Routt 
County; eight were taken near Toponas, in Egeria Park; and the 
bleached anterior part of a skull was found in the nest of a wood 
rat near Douglas Spring, at the north base of the Escalante Hills. 
The above specimens compared with typical M. pauperrimus from 
northeastern Oregon show no differences. 

This small, gray, short-tailed field mouse appears to have a some- 
what interrupted distribution in northwestern Colorado. (See fig. 
16.) Although much of western Routt County consists of sandy 
sagebrush country, well suited to the species, it was found in but two 
localities. In the sand hills east of Walden, North Park, the rim- 
ways of these field mice were beneath the prostrate lowest branches 
of large clumps of CJirysoihamnus, and that this is the chief food 
plant was attested by the many neat little piles of the smaller steins 
and leaves, cut into short lengths, here and there along the runways. 
At Elk Springs, on the watershed between Bear and White Rivers, 
a large colony occupied the grassy swale in which the springs are 
located, and extended for a considerable distance into the surround- 
ing sandy sage plain. This colony probably numbered many hun- 
dreds, and the numerous runways ramifying in all directions formed 
a perfect network. The small colony near Toponas was located 
on the sage plain at the west base of the Gore Range, at an altitude 
of 8,000 feet, and was apparently subsisting largely upon range 
grasses. The members of this colony were very active during the 
early evening hours. In less than ten minutes, just after sunset 
one frosty October evening, three entered traps within a few feet 
of where I stood. Warren has taken this species at Hot Sulphur, 
in Middle Park. It has probably reached the open sage parks of 
Grand County from North Park, as the Gore Range would seem 
to be an effectual barrier on the west. 

Allen has recorded this mouse from Kinney ranch, Sweetwater 
County, Wyoming, 1 which locality is within 30 miles of the Colorado 
boundary. 

Fiber zibethicus osoyoosensis Lord. Rocky Mountain Muskrat. 

Muskrats are reported in most of the streams of central and west- 
ern Colorado below 9,000 feet, but I have found them common 
only in the marshes and lakes of the intermountain parks. In 
the chain of bogs and sloughs in the low meadows along Tomichi 
Creek, between Gunnison and Parlin, in October, 1907, muskrats 
were living in tule houses, and the many open trails ramifying 
through the moss and other vegetation on the surface of the ponds 
were evidence of their abundance. Muskrats were abundant in 

1 Bull. Am. Mus. Nat. Hist., VIII, p. 248, 1896. 



12 G NORTH AMERICAN FAUNA. [No. 33. 

the San Luis Valley, especially in the San Luis Lakes. They were 
common in marshes along Grand River, hi Middle Park, where I 
collected three specimens; two were taken near Hebron on the 
upper waters of the Platte, in North Park; and two more in a small 
tributary of the San Miguel River, near Coventry. Other specimens 
in the Biological Survey collection are from Tevebaugh's ranch, near 
Cochetopa Pass, secured by Loring in 1892. 

Fiber zibethicus cinnamominus Hollister. Great Plains Muskrat. 

The muskrats of the eastern Colorado plains are referable to the 
pale reddish plains form F. z. cinnamominus, recently described 
from Kansas. 1 They are smaller and paler and have smaller skulls 
and teeth than the mountain form of central and western Colorado, 
F. 2. osoyoosensis. Specimens from the eastern foothills are con- 
sidered by Hollister to be cinnamominus; 2 others from Middle and 
North Park localities he refers to osoyoosensis. 3 I have no records 
of muskrats above 9,500 feet, and thus the high front ranges appear 
to be an effective barrier between the two forms in Colorado. 

On the plains scores of muskrat houses may often be seen on a 
single marsh or lake. This is especially noticeable at Barr and 
other points in the lake region northeast of Denver. Although 
muskrats are present in most of the streams on the plains, their 
numbers are small compared with those inhabiting lakes and marshes. 
They are very troublesome in irrigated sections, as they are con- 
tinually burrowing in the banks of ditches and reservoirs, often 
causing serious leaks. 

This form is represented in the Biological Survey collection by 
specimens which I secured at Wray, Yuma County, and a female 
which contained eight small fetuses, taken by Loring in a small 
snow-fed lake at 9,500 feet, near Ward, Boulder County, June 8, 1893. 

Castor canadensis frondator Mearns. Broad-tailed Beaver. 

The identity of the Colorado beaver can not at present be deter- 
mined, as no satisfactory specimens are available for study. Prob- 
ably the beaver of the southern and western parts of the State is, 
frondator, but those from the higher mountains in northern Colorado 
may prove referable to canadensis. Skulls from Lake Moraine and 
Crested Butte, sent to the Biological Survey for identification by 
Warren, are referred to frondator. 

Very few beavers remain in the streams of the eastern plains 
region, where they were abundant in early times. Mr. A. E. Beards- 
ley reports a few in the Platte River, 20 miles east of Greeley, accord- 
ing to Warren. 4 Throughout the mountainous parts of the State 

i Proc. Biol. Soc. Wash., XXIII, pp. 125-126, Sept. 2, 1910. 

2 N. Am. Fauna No. 32, p. 31, 1911. 

3 Ibid., p. 26, 1911. 

* Mammals of Colorado, p. 244, 1906. 



1911.] 



MAMMALS. 127 



they are holding their own very well and in many sections appear to 
be increasing. This is due far more to the protection afforded them 
by ranchmen than to protective laws, which are often disregarded 
by trappers in the unsettled sections. 

In 1905 and 1906 beavers were reported in fair numbers on the head- 
waters of the Laramie River, the streams of Middle and North Parks, 
and on the upper Snake, Bear, and White Rivers and their affluents. 
The} T were said to be common also in the Ladore Canyon of Green 
River and in the Yampa Canyon of Bear River, many being trapped 
each winter. At Blanchard's ranch, 5 miles west of Boulder, in June, 
1905, I found an aspen sapling freshly cut by a beaver, but was 
told that very few of the animals remained on Middle Boulder Creek. 
A thriving colony is located on Beaver Creek, a few miles above its 
junction with the South Boulder. When I visited this colony in 
October, 1906, it was being carefully protected by ranchmen. It 
extended along the creek for several miles, and all the dams were in 
good repair. A great many holes were discovered under the banks, 
and the majority appeared to be inhabited. Only one beaver lodge 
of recent construction was observed, and two or three dilapidated 
structures were in neighboring ponds. It was evident that the 
members of this colony were largely bank beavers. 

In 1907 a few beavers were reported in the Cucharas River, a mile 
or two east of La Veta, and others in the same stream 4 miles south of 
La Veta. They were said to have formerly been abundant. Forest 
Ranger E. E. Chapson says a colony of four or five are living in the 
San Juan River on his ranch, 12 miles northeast of Pagosa Springs, 
but that beavers are uncommon on the upper waters of the San Juan. 
They have always been scarce in the streams heading on the western 
slopes of the La Plata Mountains, according to Mr. Steve Elkins, of 
Mancos, but were more common in some of the deep box canyons 
along the Dolores River, south of Paradox Valley. A few are still 
found on the lower Dolores, although a great many have been caught 
during the past 5 or 10 years. Along the Los Pinos, from Vallecito 
down nearly to Ignacio, beavers were reported as quite common, 
but I did not get an opportunity to examine any of their colonies, 
two or three of which are said to be located within a few miles of 
Bayfield. The beavers on this stream are trapped to a considerable 
extent despite protective laws. Mr. E. G. Bates, of Bayfield, con- 
siders them a nuisance in the lower Los Pinos Valley, as they are 
continually throwing dams across the large irrigation ditches, thus 
flooding much land and preventing the proper utilization of the 
water. 

In 1892 Loring found beaver signs on Sangre de Cristo Creek, 
Costilla County, but reported that none of the animals were left in 
the vicinity of Fort Garland. In 1893 he found a large colony of 



128 NORTH AMERICAN FAUNA. [No. 33. 

beavers on Fall River, in Estes Park, counting 10 dams in a distance 
of 2 miles. Bailey reported a protected colony on the Alamosa 
River in 1904, and saw beaver work in the canyon of Conejos River, 
west of Antonito. 

Allen reported beavers common in 1871 on the South Platte and 
its tributaries, in Park County, 1 and again in 1892 on the Florida, 
Animas, Mancos, and San Juan Rivers of southwestern Colorado. 2 
In 1895 a colony of 8 or 10 beavers was living in a tule marsh within 
a stone's throw of the Union Station at Pueblo. 3 The Fur Trade 
Review (p. 242, 1901) quotes the Denver Times as stating that D. D. 
Finch, a ranchman living near Trinidad, found it necessary to apply 
to the State game commissioner for a permit to kill a number of 
beavers which had established themselves in the creek on his ranch. 
It is stated that the beavers cut down many of his fruit trees and 
dammed the creek so that it flooded the first floor of his house. 
Warren mentions a colony in Grand River, below Grand Junction, 4 
and writes that in 1909, while passing through Hardscrabble Canyon, 
in the Wet Mountains, he saw considerable beaver work. 

The habits of the beaver are too well known to require extended 
description. The quaking aspen is the tree most used by beavers 
in the Colorado mountains, both in the construction of dams and for 
food. The engineering skill in controlling water by means of dams 
displayed by a large colony of beavers on the Slate River at Crested 
Butte has been described in an extremely interesting article by 
Warren. 5 The work of beaver colonies on the Grand and White 
Rivers has been described by Barber. 6 The results of a close study 
of the work of a colony of beavers in the South Platte River near 
Littleton is given by Rockwell m the Denver Post of August 9, 1908. 

Geomys lutescens Merriam. Yellow Pocket Gopher. 

This interesting pocket gopher, the only representative of the 
genus in Colorado, has a wide range over the plains of the eastern 
half of the State. (See fig. 17.) It is very abundant in some sec- 
tions and uncommon in others, the abundance and scarcity appar- 
ently depending largely upon the nature of the soil. It is most 
abundant in sandy areas, where the soft soil favors tunneling. 

This species is the most injurious of the Colorado pocket gophers. 
It seriously damages the alfalfa crop by eating the roots, while the 
numerous mounds of earth thrown up by the animals in alfalfa 
fields and meadows also dull the mowing machine sickles and cause- 
great loss of time to ranchmen during the haying season. At Wray, 



1 Bull. Essex Inst., VI, p. 5(1, 1S74. « Mammals of Colorado, p. 244, 1900. 

2 Bull. Am. Mus. Nat. Hist., V, p. 81, 1S93. » Proc. Wash. Acad. Sci., pp. 429-43S, 1904. 

3 See Forest and Stream, pp. 43-44, 1895. « Am. Nat., XI, pp. 371-372, 1877. 



1011.] 



MAMMALS. 



129 



Yuma County, this gopher is reported very destructive to orchards 
planted in sandy soil, as it cuts the roots of a great many young 
trees and frequently kills even those of large size. 

On the western end of the Arkansas Divide, in northern Weld 
County, and perhaps elsewhere along the eastern base of the foot- 
hills, G. lutescens occupies considerable territory adjacent to the 
range of Thomomijs clusius. Both gophers were taken on the grassy 
plateau north of the Chimney Cliffs in northwestern Logan County 
within 2 miles of the Nebraska boundary at 5,100 feet. Geomys is 
prevalent on the Arkansas Divide as far west as Eureka Hill, wlule 
Thomomys is the gopher found south of Seibert. Over a wide area 




Fig. 17.— Distribution in Colorado of yellow pocket gopher {Geomys lutescens). 

on the southeastern plains G. lutescens occurs with Cratogeomys 
castanops. In the vicinity of Lamar it is abundant in sandy country, 
while Cratogeomys is largely restricted to hard-soil flats. Streator 
found lutescens common in sandy river bottoms at both Pueblo 
and Limon. 

There are specimens of G. lutescens from Loveland, Valmont, 
Sterling, Avalo, Seibert, Pueblo, Limon, Burlington, Hugo, Kit 
Carson, Twin Buttes, Chivington, and Las Animas. Other localities 
represented in the Warren collection are Monon (Baca County) 
and Colorado Springs. A specimen collected at Denver by Mr. 
W. D. Hollister has been identified by the Biological Survey. 
90432°— 11 9 



130 



NORTH AMERICAN FAUNA. 



[No. 33. 



Cratogeomys castanops (Baircl). Chestnut-faced Pocket Gopher. 

Pseudostoma castanops Baird, Rept. Stansbnry's Exped. to Great Salt Lake, p. 
313, 1852. Type from near Bents Fort, Colorado (near present site of Las 
Animas). 

The large chestnut-faced gopher has not been taken much north 
of the type locality, and the Arkansas Valley marks in a general 
way the northern limit of its range. (See fig. 18.) Mr. C. E. Aiken, 
of Colorado Springs, has a mounted specimen taken near the reser- 
voirs several miles north of Lamar, but the species appears not to 
reach Arlington and Chivington on the line of the Missouri Pacific 
Railroad. From this latitude (about 38° 15') the species ranges 
southward to Chihuahua, Mexico. 




Fig. 18.— Distribution in Colorado of chestnut-faced pocket gopher ( Cratogeomys castanops). 

The type specimen in the United States National Museum was 
collected on the "prairie road to Bents Fort," which would be near 
the present site of Las Animas, whence the Biological Survey has a 
series of topotypes. Other specimens are from Olney and from 
La Junta (18 miles south). Warren has the species from the follow- 
ing localities: Lamar; Monon, Baca County; Irwin's ranch, Las 
Animas County; and 3 miles west of Pueblo. 

C. castanops may be distinguished at once from the yellow pocket 
gopher (Geomys lutescens), which is found with it over most of south- 
eastern Colorado, by its much larger size and unisulcate upper 
incisors. It is usually found on hard-soil flats, while G. lutescens 
prefers sand}' strips of country and soft soils. The characteristic 



1911.] MAMMALS. 131 

large flat earth heaps of Cratogeomys were seen in abundance on the 
high plains from Lamar south to Springfield, in Baca County, and 
thence northwest to Gaume's ranch, in Shell Rock Canyon, and 
north to Caddoa Station, Bent County. This gopher came under 
my observation in a cattle and sheep grazing region where there 
is little land under cultivation and where it is not the pest that it is 
in agricultural sections. 

Thomomys talpoides agrestis Merriam. San Luis Pocket Gopher. 

Thomomys talpoides agrestis Merriam, Proc. Biol. Soc. Wash., XXI, p. 144, June 
9, 1908. Type from Medano ranch, San Luis Valley, Colorado. 

This large pale gopher was found in abundance in the hay meadows 
bordering Medano Creek at the Medano Springs ranch, near the San 
Luis Lakes, in October, 1907. The piles of earth thrown up average 
larger than those of any other species of TJiomomys found in Colorado, 
a few being noted fully 4 feet in diameter and a foot high, and they 
are so numerous as to give a dotted appearance to the closely mown 
meadows. These large earth heaps prove a great hindrance in 
harvesting hay, as they clog and dull the mowing machine sickles very 
rapidly. A white, frosted appearance is presented by many of the 
gopher hills which have been thrown up by animals tunneling in 
extensive alkaline deposits, but over most of the meadows the piles 
are of rich black loam. While most abundant on low ground at 
Medano ranch, a few gophers inhabit the low, sandy, cactus-covered 
hummocks which surround the marshes and meadows. I found 
them very difficult to catch, as they nearly always covered my steel 
traps with earth and stopped up the tunnels for some distance from 
the traps. 

In traveling northeast from Mosca to the Medano ranch we detected 
no signs of gophers until we reached the meadow lands southeast 
of the San Luis Lakes. The characteristic large hills were not seen 
near Hooper nor in staging from Moffat northwest to Saguache, 
but were tolerably common in the alkaline soil along the railroad 
near Moffat and to a point 12 miles north of there. In 1909 Warren 
secured specimens of T. agrestis at Crestone, at the western base of 
the Sangre de Cristo Range, and others on Mosca Creek. The species 
thus appears to have a restricted range in the northeastern part of the 
open San Luis Valley (see fig. 19), but the limits and area of its dis- 
persion are as yet unknown, as are also its food habits. Large 
sandy areas in the valley appear to be entirely uninhabited by pocket 
gophers. It is highly interesting to find in the large intermountain 
San Luis Valley a gopher with relationships nearest talpoides of the 
northern Great Plains. Careful collecting in the Rio Grande Valley 
of northern New Mexico and on the high mountain barriers on the 
east and north sides of San Luis Valley make it reasonably certain 
that agrestis is entirely isolated. 



132 



NORTH AMERICAN FAUNA. 



[No. 33. 



The following characterization will serve to distinguish agrestis 
(from original description, 1. c): 

''Characters. — Size and general characters much as in talpoides, 
but color very different — pale drab as in ocius instead of dark brown 
as in talpoides; skull also different. 

' ' Color. — Upperparts uniform drab, sometimes with a pale reddish 
(dull ochraceous) cast on top of head and neck; ear spots conspicu- 
ously dusky; underparts soiled whitish, the plumbeous underfur 
showing through. 

" Cranial characters. — Skull similar to talpoides, but averaging 
longer and narrower; zygomata less spreading; nasals less regular, 




5$$5$$$ T.C/.US/US. 
TFUi-t/US. 



Fig. 19. — Distribution in Colorado of pocket gophers (genus Thomomys). 

less truncate posteriorly, less straight on outer side — tending to spread 
outward on posterior third; premaxillse longer and broader poste- 
riorly; maxillary root of zygomata longer and broader and swollen to 
articulation with jugal so that the jugal part of arch is abruptly and 
conspicuously narrower; bullse slightly more swollen. 

"Measurements. — Type: Total length, 220; tail vertebra?, 57; 
hind foot, 30. Average of 4 females from type locality: Total length, 
212: tail vertebras, 55; hind foot, 29." 

Thomomys clusius Coues. Coues Pocket Gopher. 

This species extends into North Park from the north, and east of 
the mountains is common on the high plains of northeastern Weld 



1911.] MAMMALS. 133 

and northwestern Logan Counties. Toward the south it is again 
common on the higher western end of the Arkansas Divide, from a 
point 8 miles south of Seibert west to the base of the mountains near 
Colorado Springs. (See fig. 19.) The distribution on the plains 
between the Arkansas Divide and northeastern Weld County is 
unknown, there being no specimens from the intervening region. 
Specimens from the lower foothills of Jefferson, Boulder, and Larimer 
Counties are apparently intermediate in some respects between 
T. clusius and fossor of the higher mountains, which indicates a 
probable continuity of range on the plains at the eastern base of the 
foothills from the AVyoming boundary south to the Arkansas Divide. 
Throughout its range T. clusius is mainly an inhabitant of high plains 
and open foothill country in the Transition zone. 

This gopher was found sparingly in the sand hills east of Canadian 
Creek, North Park, at the west base of the Medicine Bow Mountains. 
In July it was feeding largely upon a species of lupine (Lupinus 
alpestris?) , and large caches of the stems and leaves of this plant 
were found hi some of the burrows. North Park specimens accord 
weU with typical clusius from Bridgers Pass, Wyoming. This gopher is 
apparently rare in North Park, since it was found at no other point. 

The Coues gopher is the dominant species on the high plains of 
northeastern Weld and northwestern Logan Counties. At Chimney 
Canyon, some 30 miles northwest of Sterling, where I collected five 
specimens on the grassy plateau which forms the summit of the 
Chimney Cliffs, at 5,100 feet elevation, Geomys also was sparingly 
present. Thomomys was the only gopher found at Pawnee Buttes, 
some 30 miles southwest of Chimney Canyon, and thence westward 
and northward to the Wyoming boundary north of Grover. It is prob- 
ably the dominant gopher over all the high grassy watershed between 
Horsetail and Lodge Pole Creeks, including a small area in extreme 
western Nebraska. 

On the Arkansas Divide the eastern limit of its range appears to be 
8 miles south of Seibert, where a specimen was taken from a small 
colony in hard soil on the north slope, at about 4,500 feet. From 
this point the species has a continuous range westward on this high 
watershed to the base of the foothills. Streator found it abundant 
at Limon and Flagler, and Warren states that it is tolerably common 
on the plains at Colorado Springs, occurring there with Geomys 
lutescens. 1 

Thomomys clusius ocius Merriam. Green River Pocket Gopher. 

On the sage plains of western Routt and Rio Blanco Counties this 
pale grayish form is the only pocket gopher present. It is abundant 
in the lower valleys of the Snake, Bear, and White Rivers, and in 
Browns Park, on Green River. On the watersheds between these 

i Mammals of Colorado, p. 252, 1906. 



134 NORTH AMERICAN FAUNA. [No. 33. 

valleys, however, it occurs only in scattered colonies, usually in 
sandy strips of country. At no point was this gopher found much 
above 6,000 feet. Baggs Crossing appears to be the eastern limit of 
its range in the Snake River Valley, while the boreal-capped escarp- 
ment of the Book Plateau forms an effective barrier to its southward 
dispersion. (See fig. 19.) In all the high mountainous country on 
the south and east of its range T. ocius is replaced by T. fossor. The 
ranges of the two do not appear to meet, however, and there is a 
region of varying width in which no gophers were found. 

Gopher hills were abundant on the divide southwest of Rangely, 
and a few were seen near the Utah boundary in the valleys of Texas 
and Evacuation Creeks. A specimen taken on Bear River, south of 
Lay, August 8, and another from Snake River (15 miles northeast of 
Sunny Peak), August 24, are in short reddish summer pelage. Sep- 
tember specimens from Ladore, Lily, Elk Springs (8 miles south of 
Lily), and Rangely are in the full grayish winter coat. The anterior 
part of a bleached skull was found in a wood-rat nest at Douglas 
Spring, at the north base of the Escalante Hills. 

In the river valleys, where ocius is most abundant, ranchmen do 
not consider it very injurious, since it prefers sandy and waste soil 
and greasewood flats- to meadow land and alfalfa fields. 

Thomomys fossor Allen. Colorado Pocket Gopher. 

Thomomys fossor Allen, Bull. Am. Mus. Nat. Hist., V, p. 51, 1893. Type from 
Florida, La Plata County, Colorado. 

The general distribution of this dark-brown, medium-sized species 
in the Colorado mountains is indicated by specimens from a wide 
range of localities. It is the only gopher found in the higher moun- 
tains and occurs regularly from 7,000 feet to 13,000 feet, both in 
heavy forests and on the open slopes far above timberline. It is 
common as low as 7,000 feet in the yellow pine forests of Archu- 
leta, La Plata, and other southwestern counties, but at few points 
does it range much below the Canadian zone. (See fig. 19.) The 
greatest numbers are found in the aspen belt of the Canadian zone, 
where the numerous fresh hills of earth thrown out each night attest 
to the great activity of the animals in the rich black soil. (See fig. 
20.) The lower limits of range are roughly indicated by Florida, 
Pagosa Springs, Sapinero, Hayden, Meeker, Elkhorn (Larimer 
County), and foothills near Boulder and Golden. Gophers from the 
lower eastern foothills of the Front and Medicine Bow Ranges are 
not typical, and they show an approach toward clusius of the higher 
plains. 

Ranchmen in the foothill valleys and mountain parks suffer con- 
siderable loss through the depredations of these animals, and every 
year a large acreage of alfalfa is killed by gophers cutting the roots 
just beneath the surface of the ground. In the spring of 1905 Mrs. 



1911.] 



MAMMALS. 



135 



Blanchard, living 5 miles west of Boulder, discovered, 3 inches below 
the surface, a cavity in which a gopher had a store of nearly 50 
tiger-lily bulbs, evidently gathered, the previous fall. The cavity 
was nearly full and the bulbs were scattered through loose earth, 
which had been thoroughly worked over. A tunnel led directly from 
the cache to the flower bed a rod or so distant. Near Golden this 
gopher is said to make itself a nuisance by burrowing in the banks 
of irrigation ditches and reservoirs, and this is probably true in other 
sections along the lower edge of its range. The numerous hills of 
earth and stones thrown up in hay meadows and grain fields dull the 
sickles of mowing and harvesting machines. 
As an offset to the 




injury inflicted upon 
agricultural inter- 
ests along the lower 
edge of its range, T. 
fossor is an impor- 
tant agent in the con- 
servation of forests 
and moisture in the 
higher mountains , 
where it is most 
abundant. The thor- 
ough and continual 
working and enrich- 
ing which the soil re- 
ceives through the 
activities of gophers 
is highly beneficial 
to forest growth, and 
at the same time a 
large amount of 
moisture which 
would otherwise run 
off the mountain slopes is retained in the numerous burrows and 
underground tunnels which might properly be termed natural water 
traps. 

On the higher open mountain slopes, particularly above timber- 
line, one often sees peculiar long serpentine ridges of earth, some- 
times dry and hard packed, but more often partially disintegrated 
through the action of moisture. These are formed by gophers during 
the winter when snow covers the ground to a considerable depth. 
The loose earth thrown out is packed into the ramifying tunnels 
which the animal has made through the snow on the surface of the 
ground. 



Fig. 20.— Earth heaps of Colorade pocket gopher ( T ho mom ys fossor) 
on Book Plateau, at 8,000 feet. 



136 NORTH AMERICAN FAUNA. [No. 33. 

Thomomys fulvus (Woodhouse). Fulvous Pocket Gopher. 

A female from Fisher Peak, southeast of Trinidad, is referable to 
this species. It was collected September 15, 1903, at an elevation of 
8,000 feet, by Mr. A. H. Howell, who states that gophers were toler- 
ably common in the parks on the upper slopes of Fisher Peak, in the 
Transition zone. This gopher probably reaches Colorado only in the 
Trinidad region (see fig. 19), although it has been taken at a number 
of points in the foothills of northern New Mexico just south of the 
Colorado line. 
Thomomys aureus Allen. Golden Pocket Gopher. 

The golden pocket gopher is found in the low valleys of south- 
western Colorado from the lower Grand River Valley southward in 
the Upper Sonoran zone. In a general way it occupies the valleys 
and lower flats south and west of the San Juan and La Plata Moun- 
tains and the Uncompahgre Plateau (see fig. 19), but as it has been 
taken at Grand Junction, it may cover a considerable area in the 
lower Gunnison Valley. In 1907, however, I saw no signs of pocket 
gophers at any point in either the Gunnison or Uncompahgre Valley. 
Gophers seem to be absent or very scarce in the Grand Valley also, as 
in following along Grand River from the Utah line to Rifle, a dis- 
tance of 100 miles, I neither saw nor heard of any. 

There are in the Biological Survey collection specimens of T. 
aureus from Arboles; Bayfield; Mesa Verde, northern end, 8,100 feet; 
Ashbaugh's ranch, near McElmo; Coventry; and Grand Junction. 
The Grand Junction specimen, taken November 7, 1895, by Mr. 
A. H. Howell, is an adult female in beautiful golden gray winter 
pelage. The others from Colorado are all June and July specimens, 
a June male from Arboles being darker than typical aureus. 

At Bayfield, Ashbaugh's ranch, and Coventry these gophers are 
abundant, and are reported to be very injurious in grain and alfalfa 
fields. They are stated to be very destructive also to young orchards 
near Bayfield, as they cut off the roots of the trees. They reach 
their upper limit along the Los Pinos in the pinyon country near 
Bayfield at 6,500 feet, all the yellow pine belt north of that point 
being occupied by T. fossor. Along the San Juan River they are 
found as far east as Arboles, where they are common in sandy soil 
along the railroad track. In the Coventry region these gophers are 
very numerous on the cultivated flats between the San Miguel River 
and Naturita Creek, extending up to about 6,500 feet. They are 
a great nuisance in this irrigated section, tunneling in the banks of 
ditches and reservoirs and causing numerous and often serious leaks. 
In the arid desert region along the lower San Miguel and Dolores 
Rivers gophers are very scarce, the only signs noted being at the 
Dolores River crossing in Paradox Valley and at Uranium in Sinbad 



1911.] MAMMALS. 137 

Valley, and these workings were not fresh. They are a pest at 
Ashbaugh's ranch in the valley of McElnio Creek, but apparently 
do not inhabit the bordering mesas. Many ranchmen keep a number 
of cats purposely for killing gophers, and a great many of the animals 
are drowned out when the fields are irrigated. Mr. George J. Ash- 
baugh attributes to pocket gophers the holes which he occasionally 
finds eaten into watermelons from beneath. 

Gopher mounds were abundant on the northern end of Mesa Verde, 
between 7,500 and 8,100 feet, in the Transition zone, and a specimen 
collected June 14 at 8,100 feet is referred to aureus. This seems to 
be the only Colorado record for this species in the Transition zone. 
The workings were in soft soil, either among sagebrush or in the 
small grassy parks scattered here and there among the dense chap- 
arral of oak and June berry. No signs of gophers were found among 
the pinyons on the southern end of Mesa Verde below 7,500 feet. 
Gopher hills seen in alfalfa fields just west of Mancos, and also in the 
bottom of the Dolores Canyon, 2 miles east of Dolores, were the char- 
acteristic large earth mounds of aureus. 

Thomomys aureus pervagus Merriam. Espanola Pocket Gopher. 

This large species is the most richly colored of the Colorado pocket 
gophers. It was described from Espanola, Santa Fe County, N. Mex., 
and is known to range north as far as Salida, in the lower end of 
the upper Arkansas Valley. The distribution between these two 
points has not been worked out in detail, but specimens from the 
Huerfano Valley at Gardner and from Antonito and the Conejos 
River Canyon, 10 miles west of Antonito, indicate a range in the 
foothill valleys along the east sides of both the Sangre de Cristo and 
San Juan Ranges, chiefly in the Upper Sonoran zone. (See fig. 19.) 

It is impossible to state from present knowledge how this species 
has reached the upper Arkansas Valley, or whether, indeed, continuity 
of range exists. It is reasonably certain, however, that it does not 
cross the 9,000-foot Poncha Pass from the San Luis Valley, as the 
Poncha Pass region, which is practically hi the Canadian zone, is 
occupied by T. fossor. The characteristic large hills of T. pervagus 
were seen in abundance on the sandy flats in widenings of the Arkansas 
Canyon from Salida down to Texas Creek Station; but the Grand 
Canyon of the Arkansas, beginning a few miles below Texas Creek 
and culminating in the stupendous chasm of the Royal Gorge, must 
certainly separate pervagus from the gophers inhabiting the Upper 
Sonoran country around Canon City, at the lower end of the canyon. 
It remains to determine the species occurring in the Arkansas Valley 
near Canon City, and thence south along the eastern base of the Wet- 
Mountains, but logically it should be pervagus, as this is the gopher 
abundant in the Huerfano Valley from Badito west to Gardner. 



138 NORTH AMERICAN FAUNA. [No. 33. 

Apparently, also, this is the form whose large hills were so common 
in the Cucharas Valley from Walsenburg up to La Veta, and between 
Trinidad and Weston in the upper Las Animas Valley. In going 
from Texas Creek to Westcliffe and thence overland south to Gardner, 
the small hills of fossor were the only signs of gophers noted, and 
these were found chiefly in the high Wet Mountain Valley. Over a 
wide strip of mixed yellow pine and pinyon country south of the 
Arkansas River at Texas Creek, and also on the southern slopes of 
the divide connecting the Wet Mountains with the Sangre de Cristo 
Range, no gophers appear to be present. 

In the San Luis Valley, T. pervagus is known only about its southern 
end. Just how far north it extends along the eastern base of the San 
Juan Range is not known. The large gopher hills which Bailey 
found abundant along the Rio Grande at Del Norte may have been 
either pervagus or the pale T. agrestis described from Medano Springs 
ranch, near the San Luis Lakes, the range of which also is very imper- 
fectly known. I saw no gopher work near Saguache. 

Throughout its range T. pervagus revels in the rich mellow soil of 
the stream valleys, or occupies the sand flats along the lower border- 
ing slopes, as in the Arkansas Valley above Salida. In the Cucharas, 
Huerfano, and Las Animas Valleys these gophers frequent alfalfa 
fields, throwing out the rich brown or reddish soil in large hills, often 
3 feet or more in diameter, which must prove a great hindrance in 
harvesting the crop, aside from the serious injury which the animals 
inflict upon the alfalfa itself by eating the roots, of which they are 
very fond. 

The Salida specimens were collected at Sand Park, the sandy east- 
ern slope of the Arkansas Valley a mile north of the town. The gopher 
hills were common in this strip of country, both along the embank- 
ments of the Denver & Rio Grande Railroad near the river, at a little 
over 7,000 feet, and up to the lower edge of the pinyons 400 or 500 
feet higher, but they were not noted in the adjoining adobe soil. 
They were usually thrown up among the bunches of prickly pear 
(Opuntia polyacantha) or beneath Chrysoihamnus &nd~Atriplex bushes. 
One of these shrubs *at least, Atriplex canescens, forms a part of the 
animal's food, as in One of the underground tunnels was a cache of its 
stems and leaves. 

Specimens from Salida and Gardner, taken in early November, are 
in beautiful full reddish-brown winter pelage, with a considerable 
coal black area on chin, lips, and throat. An immature Gardner 
specimen has the sides of the throat strongly suffused with reddish- 
orange, and this suffusion is more or less indicated in all the adults. 



1011.1 



MAMMALS. 



139 



Perodipus montanus (Baird). San Luis Kangaroo Rat. 

Dipodomys montanus Baird, Proc. Acad. Nat. Sci. Phila., VII, p. 334, 1855. 
Type from old Fort Massachusetts (near Garland), Colorado. 
This peculiar Upper Sonoran mammal inhabits the sandy areas in 
the San Luis Valley (see fig. 21), and thence extends south into the 
high valleys and parks of northern New Mexico. It is one of the 
highest ranging species of the genus, in the San Luis Valley occurring 
up to 7,900 feet and possibly a little higher, and throughout its range 
it is restricted to the mountain valleys. The type of montanus, in 
the National Museum, was collected in 1853 by F. Kreuzfeldt, on Capt. 
Beckwith's expedition. In addition to a large series of topotypes 




= P£ftOD/PUS lOA/G/P£S : ^^ f- MO/VTAMUS. gtffrj^M. fi/CHAPDSOW. 

Fig. 21.— Distribution in Colorado of kangaroo rats (genus Perodipus). 

from Fort Garland, the Biological Survey has specimens from Anto- 
nito; Conejos River, 10 miles west of Antonito; and Medano Springs 
ranch, near the San Luis Lakes. Warren has taken the species at 
Crestone. 

The characteristic burrows are tolerably common along the rail- 
road at Moffat, and thence northwest to a point 6 miles southeast of 
Saguache, and kangaroo rats are said to be occasionally plowed out 
in the fields at Saguache. . In going north from Moffat the burrows 
were often noted in the gravelly soil along the railroad near the hot 
springs, but the only signs of kangaroo rats seen farther north were 
two fresh burrows in sandy soil a mile south of Villa Grove, at an 



140 NORTH AMERICAN FAUNA. [No. 83. 

elevation of 7,900 feet — a point doubtless very near the northern 
limit of the species. The high Poncha Pass country effectually sepa- 
rates montanus from P. richardsoni of the upper Arkansas Valley. 

At Fort Garland Loring found this species abundant in a cultivated 
field, but absent from the sagebrush country. Bailey reported it 
very common in the Rio Grande Valley at Alamosa and Del Norte 
and in the Conejos River Canyon. A specimen taken on Conejos 
River had its cheek pouches filled with the heads of gramma grass. 
Del Norte is apparently very near the western limit of montanus along 
the Rio Grande, as I saw no signs above that point. 

Kangaroo rats are stated to have been very numerous at Mosca 
some years ago when only small areas were under cultivation. At 
that tune they were very injurious to crops, digging up large quanti- 
ties of newly planted grain and caching it in their burrows along the 
sandy margins of the fields, and also feeding extensively on tender 
green stems of wheat. Of late years the rats have been largely 
driven out of the central part of the San Luis Valley by extensive cul- 
tivation and irrigation, and now are gathered in the sandy unculti- 
vated parts, particularly in the sandhills along the western base of 
the Sangre de Cristo Range from Garland north to Crestone. Accord- 
ing to cowboys they are so numerous in the Luis Maria Baca grant, 
south of Crestone, that their burrows and tunnels completely under- 
mine the sandy ground in many places and make riding difficult and 
even dangerous. 

At the Medano ranch kangaroo rats were found chiefly in the sandy 
hummocks just southeast of the headquarters. The burrows were 
usually beneath Sarcobatus or Atriplex bushes, or in beds of prickly 
pear (Opuntia), and more rarely under the large rabbit brush (Chryso- 
thamnus patens). As with other species of kangaroo rats, there are 
usually from three to six entrances to a nest, each entering the 
ground at an angle of less than 45 degrees, sometimes nearly hori- 
zontally, and usually from different directions. Kangaroo rats are 
nocturnal as a rule, but I caught one individual in a trap during the 
daytime at the Medano ranch, and early one morning before sunrise 
saw another dart into its burrow, leaving a tiny cloud of dust behind. 

Perodipus montanus richardsoni (Allen). Richardson Kangaroo Rat. 
This interesting species is represented by a large series of speci- 
mens, chiefly from east of the mountains on the plains, where it is 
generally distributed. It occurs also over parts of the sandy sage 
plains of northwestern Colorado, in Routt County, reaching that 
region from the Wyoming plains on the north. (See fig. 21.) It 
follows some of the warm Upper Sonoran valleys into the eastern 
foothills of the front ranges, the upper limits being indicated by 
Arkins, Larimer County: Twin Lakes; Salida; 8 miles west of Gard- 
ner, Huerfano County; and La Veta. 



1911.] MAMMALS. 141 

In northwestern Colorado kangaroo rats occur sparingly in the 
sandy bottoms along Bear River south of Lay; 'and also near the 
Wyoming boundary, a few miles southwest of Baggs Crossing. In 
the lower valley of Snake River and throughout the sandy sage 
plains north of Bear River they were found in abundance. At no 
point were they noted much above 6,000 feet. The greatest num- 
bers were found on the east side of Snake River, near Sunny Peak. 
Here a large colony occupied a sandy strip of country abounding in 
blow-outs, and the burrows were either in the banks of these or be- 
neath bushes of Atriplex confertifolia or Grayia spinosa. Specimens 
from northwestern Colorado are referable to P. richardsoni, although 
they average a little smaller than the typical form and have weaker 
maxillary arches — in these two respects approaching P. montanus 
of the San- Luis Valley. 

During a trip from Cheyenne Wells northwest across the plains to 
Sterling and Grover I found kangaroo rats generally distributed in 
all sections having sandy and soft soils. Our night camp, June 2, 
some 20 miles northwest of Sterling, was near an old sod corral, the 
sides of which had tumbled down and partially disintegrated. The 
soil here was soft and easily excavated and had attracted hundreds 
of kangaroo rats, whose burrows fairly honeycombed the ground. 
A small number of traps put out here secured 12 specimens. 

Kangaroo rats are abundant at Gardner, in the Huerfano Valley, 
both on sandy flats along the river and on the low adjoining benches. 
The burrows at this point were usually beneath chico brush (Atriplex 
canescens) , and after a 2-inch snowfall during the night of November 
18 I saw the peculiar round tracks made by a few which had come 
out of their burrows during the night and skipped about on the fresh 
snow. The leaps taken when running are long, often 3 or 4 feet. 
Kangaroo rats are reported along Muddy Creek, 8 miles west of 
Gardner. In the Cucharas Valley they are common at Walsenburg, 
and a very few burrows were found at La Veta. 

In November, 1907, I found kangaroo rats exceedingly numerous 
on the gently inclined sandy slopes lying along the east side of the 
Arkansas River, just north of Salida; and I saw from the train near 
Howard and Cotopaxi, farther down the river, burrows which I attrib- 
uted to this species. Mr. J. W. Frey states that the rats are found north 
to Browns Canyon, 7 miles above Salida, and west to Poncha Springs. 
Kangaroo rats were most abundant at Salida among the CJiryso- 
thamnus bushes, Russian thistles, and in beds of Opuntia polyacaniha, 
on the upper sand slopes just below the edge of the pinyons, between 
7,100 and 7,400 feet. The Salida series is not typical P. richardsoni, 
but is much nearer to that species than to montanus of the San Luis 
Valley. A Twin Lakes specimen recorded by Coues and Yarrow 1 

lExpl. W. Of 100th Mer., V,p. 109, 1875. 



142 NORTH AMERICAN FAUNA. [No. 33. 

as " Dipodomys phillipsi ordi" was collected by Dr. J. T. Rothrock 
in August, 1873, and if correctly labeled is an important record 
because of the high elevation. The Royal Gorge, and in fact much 
of the Grand Canyon of the Arkansas, would seem to prevent con- 
tinuity of range from the plains; but it is evident that the kangaroo 
rats of the upper Arkansas Valley reached that region by way of 
the Arkansas drainage area, apart from all physiographic considera- 
tions, since their relationship is clearly with ricliardsoni. 

The typical home of the kangaroo rat is in sandy river bottoms or 
on the numerous sand ridges scattered here and there over the 
plains. It is seldom found living in hard soils, but often takes up 
its abode in cultivated fields. The more or less horizontal burrows 
are excavated beneath bunches of prickly pear, yucca, and sage- 
brush, or in the banks of blow-outs and railroad embankments. 
The animals are nocturnal and most active during the latter part 
of the night. During the day the burrows are often closed from 
within, but early in the morning they are usually found open, with a 
quantity of freshly ejected sand at the entrances. Their food con- 
sists of various seeds, and the stems of grass and wheat are often 
found in the capacious external cheek pouches. The pouches of a 
specimen collected at Salida early in November contained about 
equal quantities of the leaves of Atriplex canescens and the seeds of 
a species of Chrysothamnus . 

Perodipus longipes (Merriam). Mold Kangaroo Rat. 

Kangaroo rats from Fruita, Grand Junction, Hotchkiss, and Ash- 
baugh's ranch (McElmo Valley), show no departure from typical P. 
longipes from the Painted Desert, Arizona. Others from Coventry and 
Naturita in the San Miguel region, at a somewhat higher elevation, 
average darker than typical specimens, but, as they accord well in 
other essentials, are not considered separable. 

The Mold kangaroo rat is a creature of the sandy desert areas of the 
southwest and is normally restricted to the lowest and warmest parts 
of the Upper Sonoran zone. It reaches Colorado along the Colorado 
River Valley, the northern limit of range being marked by the Book 
Cliffs (see fig. 21), which separate it from P. ricliardsoni of the northern 
plains. 

By reason of the extensive cultivation of the lower Grand Valley 
during recent years, kangaroo rats now occupy less ground than for- 
merly. They are found chiefly in the narrow strip of rough unculti- 
vated country lying at the southern base of the Book Cliffs, and on 
the desert northwest of Mack. In the arid valley of West Salt Creek, 
between Mack and Carbonera, the characteristic burrows were often 
noted in the banks of arroyos and on sand flats. In a large colony 
on the sandy plain 3 miles northwest of Fruita, the entrances to the 
burrows were usually in large bunches of prickly pear (Opuntia), 



1911.] MAMMALS. 143 

but were occasionally beneath bushes of Atriplex confertifolia. During 
the time this colony was under my observation, toward the end of 
September, the animals seemed inactive, and very few piles of fresh 
earth were noted at the burrows. No signs of kangaroo rats were 
seen in the Grand Valley above Grand Junction, and they probably 
do not occur east of Palisade, where the valley narrows into a more or 
less continuous canyon. They extend southeast in the Gunnison 
Valley for some distance, apparently reaching their eastern limit near 
Hotchkiss, on the North Gunnison River, where I collected an imma- 
ture specimen in August. Burrows were often seen on the arid waste 
between Rogers Mesa and Delta, but none were noted in the Montrose 
region. 

Kangaroo rats are reported abundant on the lower San Juan River, 
and they extend as far east as Moqui in the McElmo Valley. In 
June I found a few burrows on sandy flats along McElmo Creek at 
Ashbaugh's ranch, midway between Moqui and McElmo, but the 
animals were inactive here or else the burrows were deserted. Warren 
has taken the species at Ashbaugh's ranch. 

Near Coventry, at 6,800 feet, scattered colonies of kangaroo rats 
are found in the rather hard clayey soil on the sagebrush flats, and 
the rats become increasingly numerous toward the west, as the ele- 
vation becomes lower, the soil sandy, and the region more desertlike. 
At Dry Creek, 5 miles west of Naturita, and thence down the East 
Paradox Valley and across the Dolores River to the head of West 
Paradox Valley, they are very numerous on the sandy flats and hum- 
mocks. On my return trip from the La Sal Mountains through the 
Sinbad Valley and thence down the Dolores River to the mouth of 
West Creek, however, no signs of kangaroo rats were seen, although 
much of the region is suitable for them. North of San Miguel River 
they are found at Nucla and nearly to Tabeguache Creek. 

The burrows of this species average somewhat larger than those of 
P. richardsoni, although the animals themselves are of nearly the 
same size. The deserted burrows are often used by cottontails 
(Sylvilagus a. warreni) in the lower San Miguel region, and occasionally 
both animals are found living in the same colony. 

Perog-nathus hispidus paradoxus Merriam. Kansas Pocket Mouse. 

This large pocket mouse occurs sparingly on the plains of eastern 
Colorado, from the base of the foothills in Boulder and Jefferson 
Counties east and southeast to Baca County, in the extreme south- 
eastern corner of the State. It has been taken at a few widely 
separated localities over this region in the Upper Sonoran zone. 

Specimens were collected at Sterling by Dr. A. K. Fisher, who 
reports the species as tolerably common at that locality. Another 
was secured at Hugo by Prof. Lantz, who found it common along 



144 



NORTH AMERICAN FAUNA. 



[No. 33. 



Big Sandy Creek. In the Merriam collection there is a specimen from 
Boulder County, collected in 1889 by the late Denis Gale, of Gold 
Hill, at 5,500 feet. The precise locality is not given on the label, 
but was probably at the base of the foothills not far from Boulder. 
I noted several of the characteristic auger-hole burrows of this pocket 
mouse on the lowest foothill slopes a mile southwest of Golden, and 
found the species common in southern Prowers and Baca Counties. 
Specimens from Monon, Baca County, are recorded by Warren. 1 




P.AP/ICHE. = P.CAIUSWS. lilllHIIII P- FASCIATUS /A/FAALUTEUS. ^^ P- FLAVESCEA/S. 
Fig. 22.— Distribution in Colorado of pocket mice (Perojnatkus fasciatus and P. apache groups). 

Perognathus fasciatus infraluteus Thomas. Buff-bellied Pocket 
Mouse. 

Perognathus infraluteus Thomas, Ann. and Mag. Nat. Hist., ser. 6, XI, p. 406, 
May, 1893. Type from Loveland, Larimer County, Colorado. 

Known only from the type locality, from which place the Bio- 
logical Survey has a series of 10 specimens, collected in October, 
1894, by Mr. Clark P. Streator. The type specimen in the British 
Museum was taken by Mr. W. G. Smith April 4, 1892. (See fig. 22.) 

The Loveland animal appears to be geographically separated from 
its near relative, P. fasciatus of the northern plains, from which it 
is distinguishable by the strong buffy suffusion on the underparts 
in marked contrast to the pure white underparts of P. fasciatus. 

> Mammals of Colorado, p. 253, 1906. 



1911.] 



MAMMALS. 145 



Perognathus flavescens Merriam. Plains Pocket Mouse. 

All the records for this pocket mouse are from the plains region 
north of the Arkansas Valley, where it appears to occur only in sandy 
strips of country, being most abundant in the northeastern counties. 
(See fig. 22.) It has not been taken in the southeast, where P. 
flavus replaces it. At Sterling Dr. A. K. Fisher found it common in 
sunflower patches on sandy soil, and Loring found it at Greeley 
occupying sandy strips of country with P. flavus and living in bur- 
rows beneath Opuntia and yuccas. Specimens were secured by 
Streator on the sandy bottoms along the Arkansas River at Pueblo 
in December, when the animals were gathering their winter's supply 
of seeds. At Tuttle I trapped a specimen beneath a yucca in a 
sand blow-out on the north side of the valley of the South Fork of 
the Republican River. A specimen from Boulder County was 
found by the late Mr. Denis Gale in the nest of a long-eared owl 
(Asio wilsonianus), at 5,500 feet, May 12, 1890. This record proves 
that the range of this pocket mouse extends at least to the eastern 
base of the foothills, and the animal may be found even in some of 
the warm foothill valleys. Warren has taken specimens at Colorado 
Springs, but thinks the species is uncommon at that point. 

Perognathus flavus Baird. Baird Pocket Mouse. 

This beautiful little pocket mouse is common on the Upper Sonoran 
plains of eastern and southern Colorado between 4,000 and 7,500 
feet elevation, the highest altitude being reached in the San Luis 
Valley. (See fig. 23.) At a number of localities on the northeastern 
plains it is associated with P. flavescens. The burrows of the two 
species are not readily distinguished, but those of flavus average 
smaller. A large series from eastern plains localities are typical 
flavus, but specimens from Ashbaugh's ranch, Montezuma County, 
are considerably darker, and approach P. f. bimaculatus. 

At Antonito and along the Conejos River, Bailey found this species 
abundant on the sandy sage plains, and I found it inhabiting similar 
areas at the Medano Springs ranch, near the San Luis Lakes, and also 
in weed patches and dry meadows. Most of the older gopher hills 
near the Medano ranch had been tunneled more or less by pocket 
mice, but I seldom found the burrows inhabited. These tunnels 
usually entered the soft dirt of the gopher hill from one side, passing 
horizontally through, and often connected with other horizontal 
tunnels. One such burrow was inhabited by an immature pocket 
mouse, which I caught in my hands. The hole reached a depth of 
12 inches in the soft earth, ending in a small chamber 1 by 2 inches, 
in which were stored a few grass seeds. Although pocket mice are 
usually nocturnal, I caught one in a trap at the Medano ranch in the 
daytime, and also saw another individual running in the grass near 
90432°— 11 10 



146 



NORTH AMERICAN FAUNA. 



[No. 33. 



the same burrow at 2 p. m. on a bright day. They appear to be 
inactive on damp or rainy nights. 

At Ashbaugh's ranch two specimens were trapped beneath Atriplex 
bushes on sandy flats. Mr. Ashbaugh thinks this is the little pinkish 
mouse which does so much damage in the McElmo Valley at corn- 
planting time by enlarging the hole left by the corn planter and 
taking out the kernels; and also in the autumn, when it eats much 
grain beneath the shocks. 

Near Greeley Loring found P. fiavus abundant over a sandy strip 
of country, living among clumps of yucca and prickly pear. Streator 
collected a series of specimens in sunflower patches on waste land 




Fig. 23.— Distribution in Colorado of Baird pocket mouse (Perognathus fiavus). 

near Loveland in October, baiting his traps almost exclusively with 
sunflower seeds, which appeared to be the chief food of these mice at 
the time. Prof. Lantz found a large store of sunflower seeds in a 
burrow which he dug out in southeastern Otero County in April, 1910. 
He observes that the tail of this species is to some extent prehensile. 
In the case of several which he carried alive in his hand, the tail at 
times clasped a finger. In Shell Rock Canyon, northwestern Baca 
County, I usually found the burrows beneath tree cactus (Opuntia 
arborescms) in the sandy bed of the canyon or on the adjoining 
benches. 

There are specimens from Burlington, and Streator reported this 
species at Olney and Flagler. I saw numerous signs of small pocket 



1911] MAMMALS. 147 

mice at Gardner and Walsenburg among bunches of prickly pear 
(Opuntia polyacantha) on the sandy benches bordering the Huerfano 
and Cucharas Valleys. Warren has specimens from Colorado 
Springs, Lamar, Springfield, and Texas Creek between Rito and 
Hillside. A specimen in the collection of the Colorado Agricultural 
College, taken by Mr. S. Arthur Johnson in Spring Canyon, 4 miles 
southwest of Fort Collins, has been identified by the Biological 
Survey. 

The distribution of P. Jlavus in the upper Arkansas Valley has not 
been worked out. It may occur in the sand as far up as Buena 
Vista, since it is common on the sandy slopes just above Salida. 
This region is doubtless reached by way of the Arkansas drainage 
area, although the Royal Gorge must prevent continuity of range in 
the immediate river valley. 

Perognathus apache Merriam. Apache Pocket Mouse. 

Eight specimens from Rifle, Fruita, Coventry, and Medano Springs 
ranch (San Luis Valley) are provisionally treated as P. apache, 
although none of them are quite typical. All are darker than typical 
P. apache, in this respect tending toward P. a. melanotis. The 
Coventry specimen in particular is very richly colored, like melanotis, 
but is larger, as are also three specimens from Bedrock, Montrose 
County, identified for Warren. It is probable that the Coventry and 
Bedrock specimens could be safely referred to melanotis, but, since 
dark, richly colored specimens are of irregular occurrence throughout 
the range of apache, it seems best to include all Colorado specimens 
under this species. (See fig. 22.) A specimen from the Grand River 
Valley near Rifle is very large for apache, but is equaled in size by one 
from Espanola, New Mexico. More material from southern and 
southwestern Colorado is needed before the status and distribution 
of the apache group of pocket mice within the State can be satis- 
factorily determined. 

Little is known concerning the habits of these handsome, medium- 
sized pocket mice. In the lower Grand River Valley in 1906 I found 
them among the prickly pears on the sandy desert north of Fruita, 
and also on a sandy piece of waste land near Morris, 7 miles west of 
Rifle. A fair-sized colony was occupying a sandy knoll near Morris, 
and the many freshly ejected sand piles showed its members to be 
active. However, when I again # visited the locality the following 
year, I noted very few signs. The distribution in the Grand Valley 
appears to be very local, as no signs of pocket mice were seen between 
Fruita and Rifle. 

In the region of the lower San Miguel and Dolores Rivers the 
characteristic burrows of pocket mice were noted on a sandy sage 
flat near Uranium, in the Sinbad Valley. The extreme abundance of 
white-footed mice at this point prevented my taking specimens, but 



148 NORTH AMERICAN FAUNA. [No. 33. 

the large size of the sand heaps thrown out from the burrows pointed 
to apache rather than to the small P. flavus. As is usual with pocket 
mice., the entrances to the burrows were closed during the daytime. 

In early November, specimens of apache were caught in traps set 
for harvest mice in a sandy weed patch on the Medano ranch, near 
the San Luis Lakes. They were taken beneath bushes of Chrysotham- 
nus patens, but no burrows were found attributable to the species. 
One specimen got into a trap between 9 a. m. and 4 p. m. on a bright 
day. This individual evidently had been out foraging, as each of 
its cheek pouches contained nearly a thimbleful of the seeds of a 
honey plant (Peritoma sonorse). These seeds numbered 164, and 
averaged about the size of No. 4 shot. 

(?) Perognathus callistus Osgood. Red Desert Pocket Mouse. 

While encamped on Snake River, southeast of Sunny Peak, Routt 
County, in August, 1906, numerous signs of a medium-sized pocket 
mouse were found on the first bench south of the river valley. 
Characteristic small heaps of dry earth had been recently ejected 
from most of the burrows, which were usually beneath bunches of 
prickly pear (Opuntia polyacantha) . However, a large number of 
traps kept out for several nights failed to yield a specimen, owing 
chiefly to the abundance of white-footed mice, which were continually 
getting into the traps during the early evening hours. Signs of 
pocket mice were not observed elsewhere in northwestern Colorado, 
or at any point north of the Grand River Valley, although much of 
the region seems well suited to their needs. 

The type locality of P. callistus is Kinney ranch, Wyoming, 40 
miles northwest of Sunny Peak. The character of the country at 
both localities is similar, and it seems reasonable to treat the Snake 
River pocket mice as of this species. (See fig. 22.) 

Zapus hudsonius campestris Preble. Prairie Jumping Mouse. 

This Great Plains representative of the common northern jumping 
mouse enters Colorado along the South Platte Valley. (See fig. 24.) 
Thus far it has been found in this State only on the plains at the 
eastern base of the foothills. In July, 1895, Preble trapped two 
specimens in a dense growth of weeds along an irrigating ditch at 
Loveland. One in the Merriam collection was taken at Denver by 
A. W. Anthony, September 13, 1885. Warren says it is reported 
from Greeley by A. E. Beardsley. 1 

Zapus princeps Allen. Rocky Mountain Jumping Mouse. 

Zapus princeps Allen, Bull. Am. Mus. Nat. Hist., V, p. 71, 1893. Type from 

Florida, La Plata County, Colorado. 

In the Colorado mountains this large jumping mouse is chiefly an 

inhabitant of the Canadian zone, but it also follows down some of the 

streams into the Transition zone. It frequents the dense growth of 

i Mammals of Colorado, p. 254, 1906. 



1911.1 



MAMMALS. 



149 



Heracleum lanatum and other rank vegetation bordering cold moun- 
tain bogs and streams, but is also occasionally taken beneath logs in 
heavy forest. The species has a wide distribution in the mountainous 
parts (see fig. 24), but is nowhere abundant. Bailey says it is found 
in some of the marshes of the San Luis Valley. Near Del Norte he 
noted long lengths of grass stems which had recently been cut by 
jumping mice, while a specimen from east of Antonito was taken 
"in the grassy woods along the Conejos River, where most of the 
plants are Transition zone species which follow the river bottoms 
clown into Upper Sonoran zone." Ranchmen near Meeker, Rio 




zapvs p/i/zvctPS^ W^k z/ipus v. cjAtPtsrp/s. 

Fig. 24.— Distribution in Colorado of jumping mice (genus Zapus). 

Blanco County, state that during the haying season jumping mice 
are often seen in the meadows bordering White River. The vertical 
distribution of this species is indicated by specimens from Meeker 
(6,000 feet) and Arapahoe Pass (over 9,000 feet). 

Erethizon epixanthum Brandt. Yellow-haired Porcupine. 

Yellow-haired porcupines are more or less abundant throughout 
the mountains, but are most often seen in the coniferous forests of 
the Canadian zone. They occur regularly to timberline, and in the 
Grays Peak region are said to be found occasionally among the 
rocks far above the limit of trees. They are reported sparingly from 
the yellow pine belt of the eastern and southern foothills, and even 



150 NORTH AMERICAN FAUNA. [No. 63. 

extend down into the juniper and pinyon country in some of the 
sou thwestern counties . 

Porcupines are reported to be common in the following localities: 
Middle and North Parks; Elk Head Mountains; White River Plateau; 
St. Elmo, Saguache Mountains; San Miguel Mountains; Lake City; 
San Juan Mountains, north of Pagosa Springs and Vallecito; La 
Plata Mountains, northeast of Mancos; and Culebra Mountains, near 
La Veta. I saw parts of a skin, but recently removed, at Highbridge, 
on Berthoud Pass, in October, 1906, and a dead porcupine was found 
in the trail east of Lake San Cristobal in the San Juan Mountains at 
10,000 feet, while I have found quills in a mountain rat's nest near 
Mount Whiteley in northwestern Middle Park. In the foothills 
west of Antonito, Conejos County, Bailey saw numerous Douglas 
spruces and pinyons which had been partly divested of bark by 
porcupines, and also found a great many porcupine pellets in caverns 
beneath the broken lava rock; while farther north in the San Luis 
Valley two or three porcupines are said to have been killed by cow- 
boys in the open valley on the Medano Springs ranch, near the San 
Luis Lakes. 

The food of porcupines consists largely of the bark of coniferous 
trees, and the lodgepole pine seems to be preferred to firs and spruces. 
Occasionally such large areas of bark are gnawed from a tree that it 
dies. A porcupine was seen in an Engelmann spruce on Lone Cone, 
at an elevation of 11,000 feet, July 27, 1907, and I saw many spruces 
between 10,000 and 11,000 feet from which the bark had been partially 
stripped between 20 and 30 feet above the ground. Most of the 
yellow pines seen along the railroad in the valley a mile south of 
Vance Junction, San Miguel County, July 1, showed evidences of 
porcupine work in large sections divested of bark, on both the main 
trunks and the larger branches. The injury thus inflicted upon the 
coniferous forests throughout the mountains must be considerable. 
The animals feed to a small extent, at least, upon the aspen, since 
in 1905 I saw a number of these trees, both in the Rabbit Ear Moun- 
tains and on the White River Plateau, from which the bark had been 
gnawed at a height of 10 or 12 feet. 

Owing to their sluggish movements, porcupines fall an easy prey 
to some of the larger predaceous mammals. On this point Warren 
remarks: "In spite of its protecting quills, it is eaten by coyotes, 
mountain lions, and bobcats, though possibly only in winter when 
other food is scarce, that being the only season when the writer has 
found remains of the animal so killed. 1 

Trippe records the porcupine as an inhabitant of Clear Creek County 
in the early days; 2 while Allen says it was common from foothills to 

i Mammals of Colorado, p. 254, 1906. 2 See Coues, Birds of the Northwest, p. 225, 1874. 



1911.1 MAMMALS. 151 

timberline in Park County in 1871, and also reports it from the 
regions bordering the headwaters of the Arkansas and Del Norte 
[Rio Grande] Rivers. 1 

Ochotona saxatilis Bangs. Rock Cony; Pika. 

Ochotonasaxatilis'BsLngs, Proc. N. Eng. Zool. Club, I, p. 41, 1899. Type fromnear 
timberline, Snowy Range, Park County, Colorado. 

The peculiar little rock cony is one of the characteristic mammals 
in the timberline region of the high mountain ranges. It is most 
abundant in the rock slides at or near timberline, but has been 
found also near the summits of the highest peaks, and on some of 
the western plateaus as low as 8,500 feet. 

Conies are very abundant at 12,000 feet in the Grays Peak region 
and on Rollins Pass, and a little lower on Berthoud Pass. I heard 
of a small colony in slide rock near Arapahoe Pass, in the Rabbit Ear 
Mountains, at 9,000 feet. Frank Hayes, a taxidermist of Glenwood 
Springs, secured several specimens in rock slides near the head of 
Noname Creek, Garfield County, at an elevation of only 8,500 feet. 
Warren mentions seeing conies as low as 9,300 feet near Crested 
Butte. 2 They were reported from the San Juan Mountains, north of 
Pagosa Springs and Yallecito; La Plata Mountains, northeast of 
Mancos; Saguache Mountains, near St. Elmo; Lone Cone, San Miguel 
Mountains; and on the Sierra Blanca group. Prof. Lantz found them 
abundant on Pikes Peak between 12,000 and 1.3,500 feet. 

The habits of conies are most interesting. As far as my observation 
goes, they live entirely in slide rock, usually on steep slopes, but near 
Silverton Loring found their characteristic haystacks in the crevices 
of lumber and slab piles near an abandoned sawmill; while Mr. D. 
Costello, of Gardner, tells of a cony which took up its abode beneath 
the floor of a cabin in the mountains north of Crested Butte. 3 The 
haystacks of these industrious little animals, comprising their winter 
food, are composed of many species of grasses and weeds, cut and 
gathered in summer, and allowed to dry among the rocks. Thistles 
are found in most of the stacks, and seem to be a favorite food. 
Well-worn runways lead from one stack to another and extend to 
neighboring rock slides. Conies are usually quite shy and would 
be seldom observed were it not for the odd, complaining notes which 
they utter continually when alarmed. The grayish color of the 
animal closely matches the dull-colored rocks in which it is found, 
and the notes often appear to come from a distant pile of rocks when 
in reality the motionless animal is within a few feet; or, again, the 
reverse may be true. 

There are specimens from Mount Kelso; Longs Peak; Bald Moun- 
tain, 5 miles west of Ward; Sand Mountain, near Hahns Peak; Lake 

1 Bull. Essex Inst., VI, pp. 57 and 06, 1874. 3 See under Putorius streatori leptus, p. 188. 

2 Mammals of Colorado, p. 254, 1900. 



152 NORTH AMERICAN FAUNA. [No. 33. 

City; Cumbres; and Silverton. The type and a large topotype series 
of 0. saxatilis were collected by an expedition of the Museum of 
Comparative Zoology on the Snowy Range near Montgomery, Park 
County, in 1871. Mr. E. Thompson Seton has collected the species 
on Pagoda Peak, in eastern Rio Blanco County. 

Lepus campestris Bachman. White-tailed Jack Rabbit. 

This fine species is rather generally distributed over the eastern 
plains, except south of .the Arkansas River, where it appears to be 
absent in some sections and sparingly present in others. It occurs 
also in considerable numbers in the mountain parks on the eastern 
slope of the mountains, to an elevation of 10,000 feet, but not west 
of the Continental Divide, being replaced there by the grayer form, 
L. c. townsendi. It is far more abundant in northern than in southern 
Colorado. At no point is it more numerous than in North Park, and 
in tliis region and the San Luis Valley I found it far more abundant 
than on the plains east of the mountains. 

The white-tailed jack rabbits of the eastern plains are typical L. cam- 
pestris, but those of the higher elevations are less yellowish and more 
grayish, showing an approach toward townsendi. This departure is 
well indicated in specimens from the immediate eastern slopes of the 
Continental Divide, as in the upper Arkansas Valley and in the 
northern and western parts of the San Luis Valley. Two females 
which Warren collected at 12,000 feet on the summit of Boreas Pass 
in early August are very gray, and are clearly referable to townsendi; 
while a male which I collected August 22 at Como, on the grassy 
South Park plains just below the eastern end of the Boreas Pass, at 
9,800 feet, is nearest the campestris type. 

In North Park white-tailed jack rabbits were very abundant in 
1905. Early one morning in July, in a half hour's ride along Grizzly 
Creek near Hebron, I counted 19, most of which were feeding in 
alfalfa fields on the bottoms. Formerly jack rabbits were go numer- 
ous and so destructive to crops in the San Luis Valley that prize 
hunts were held each year and many thousands were killed. They 
are still found throughout the valley, but are common only in 
the sandy strip of country lying along the west base of the Sangre 
de Cristo Range. At the Medano Springs ranch, near the San Luis 
Lakes, I found them abundant in October, 1907, and shot six speci- 
mens among the Sarcobatus and Chrysothamnus bushes on the sand 
ridges. All of these, as well as others examined at the Medano ranch, 
had the upper central area of the tail heavily shaded with plumbeous, 
but the cowboys report that a lighter colored rabbit with the tail en- 
tirely white is occasionally killed in that region. 

White-tailed jack rabbits are said to be not uncommon near West- 
cliffe, in the Wet Mountain Valley, and a few are reported near Brad- 



1011.1 



MAMMALS. 153 



ford and La Veta, in western Huerfano County. One was collected 
December 5, 1907, in the dense yellow pine forest west of Eastonville. 
In some sections on the plains this species is becoming scarce of late 
years, and the numbers are rapidly decreasing throughout the plains 
region as the country becomes more settled. This is well illustrated 
by the fact that in May and June, 1909, I saw only one individual in 
the course of a wagon trip of over 300 miles, from Cheyenne Wells 
northwest to the Wyoming line, north of Grover. At certain locali- 
ties L. melanotis appears to be replacing campestris. I saw a white- 
tailed jack rabbit just west of Fort Collins in 1906, and heard of a few 
at Wray, Yuma County, in December, 1907, although the predominant 
species at the last locality is melanotis. L. campestris was said in 1907 
to occur very rarely along the east edge of Baca County, but I was 
unable to verify the report. In 1909 I saw one on the Arkansas 
Divide near Resolis. Among the eastern foothills of the Front Range 
it ranges as high as Estes Park and Gold Hill. 

Allen found this species common in the parks of Park County in 
1871/ while Trippe recorded it as common in Clear C^eek County. 2 

Lepus campestris townsendi Bachman. Western White-tailed Jack 
Rabbit. 

Tins is the western gray form of the white-tailed jack rabbit, and 
in Colorado, at least, is more an inhabitant of the mountains than its 
eastern relative. Though not unlike L. campestris, and scarcely dis- 
tinguishable in the Held, L. townsendi is much grayer, and a blackish 
or plumbeous area on the upper central part of the tail is almost 
always well indicated. It replaces L. campestris in the mountains 
west of the Continental Divide, and may occur regularly along its 
crest, since it has been taken at extreme timberline on Boreas Pass, 
at 12,000 feet. Specimens from just east of the Rocky Mountain 
watershed, in South Park and the upper Arkansas and San Luis 
Valleys, are best referred to campestris, though evidently intergrades. 3 

This rabbit occurs sparingly in the sagebrush country at Norwood 
and Coventry between 6,500 and 7,000 feet, but Mr. C. H. Smith says 
that at both localities it is outnumbered fully 10 to 1 by the black- 
tailed species L. c. texianus. During July, 1907, I saw only one at 
Coventry, but another was seen at the west base of Lone Cone, in the 
San Miguel Mountains, July 26. This was in a grassy opening among 
the dense oak chaparral at about 9,000 feet. Mr. J. P. Galloway, of 
Norwood, states that white-tailed jack rabbits are not uncommon on 
the lower slopes of Lone Cone. Tracks of jack rabbits were seen in 
the sandy, yellow pine country near the head of Dominguez Creek, 
on the Uncompahgre Plateau, at 8,500 feet, July 15. The form 

1 Bull. Essex Inst., VI, p. 58, 1874. 3 See Nelson, N. Am. Fauna No. 29, p. 81, 1909. 

2 See Coues, Birds of the Northwest, p. 225, 1874. 



154 NORTH AMERICAN FAUNA. [No. 33. 

represented on this plateau is undoubtedly L. townsendi, as the region 
is too high for ttxianus. 

The white-tailed jack rabbits inhabiting the sage plains of western 
Routt County probably reach that region from the Wyoming plains 
on the north. As no specimens from that region are at hand, it can 
not be stated with certainty whether they are townsendi or campes- 
tris. From geographical considerations alone they might be referred 
equally well to either. In 190G one was noted in the Snake River 
Valley west of Baggs Crossing, and tracks were noticed south of Sunny 
Peak. Another rabbit was seen at Douglas Spring at the northern 
base of the Escalante Hills, and the species was reported present 
throughout the Snake and Bear River region. Rabbits reported from 
the White River Valley above Meeker, 1 Egeria Park, and on the Gore 
Range, and others seen in Middle Park, can be referred without much 
question to L. townsendi. 

Warren gives the following data regarding this form in the higher 
mountains : ' ' Two females killed near Boreas Pass, in Summit County, 
each contained foetuses, one four, the other five; one lot would 
probably have been born within a day or two, the other in about a 
week. This was on the 5th day of August. It seems very late in 
the season for young to be born at such high altitude, where winter 
sets in so early, one might say in mid-October frequently, and these 
rabbits are reported to live in these high regions the year round. 
Mr. H. L. Curtiss writes me he has seen them in winter on Fairview 
Mountain, near Pitkin, Gunnison County, at 12,000 feet." 2 Warren 
told me that the Boreas Pass specimens were jumped from beneath 
the dwarfed and matted Engelmann spruces at extreme timberline 
on Baldy Mountain, and that other hunters had found this rabbit in 
similar situations along the crest of the Front Range, where the ani- 
mals obtain in the stunted and matted conifers protection from the 
icy winds which sweep the bleak and inhospitable summits. In the 
high mountain districts both townsendi and campestris assume a 
beautiful whitish winter coat, but do not become as white as the 
snowshoe rabbit (L. bairdi). 

Lepus bairdi Hayden. Rocky Mountain Snowshoe Rabbit. 

The large furry-footed snowshoe rabbits are found throughout the 
higher mountains from the lower edge of the Canadian zone at 8,500 
feet to considerably above timberline. Along the lower edge of their 
range they meet the mountain cottontail rabbit (Sylvilagus n. pinetis), 
but above 9,000 or 9,500 feet L. bairdi is often the only rabbit present. 
I have seen very few of these rabbits on my trips through the moun- 
tains, but the tracks of their large furry feet and the w r ell-worn trails 
in the snow which wind in and out of the dense willow copses in the 

i Felger. Univ. of Colo. Studies, VII, p. 144, 1910. 
2 Further Notes on Mammals of Colo., p. 79, 1908. 



1911.] MAMMALS. 155 

gulches just below timberline show that considerable numbers are 
present in most sections. 

Snowshoe rabbits are adepts at hiding in the cover of forest vege- 
tation and easily escape notice. In the gloaming of an August 
evening one was discovered feeding in the willows which fringed a 
spring on the White River Plateau, 25 miles southeast of Meeker, at 
8,500 feet, and it allowed a close approach before hopping uncon- 
cernedly into a denser thicket. Another got into a trap set under a 
cabin near Coulter, Middle Park, in the middle of October. The 
weather was cold and wintry, but this specimen had just begun to 
assume the white winter pelage on the legs and flanks. Lumbermen 
at Fraser, Middle Park, report snowshoe rabbits abundant, and often 
find them living under log piles and brush heaps. While following 
a logging trail near Fraser I discovered the remains of a rabbit which 
had been captured by a bobcat during the previous night. Bobcats 
and coyotes living in the higher mountains feed quite extensively 
upon this species, and on Berthoud Pass their tracks were often seen 
in the snow where they had been hunting rabbits among the willow 
copses near timberline. In June, 1905, rabbit signs were abundant 
among the alpine willows far above timberline on Mount Kelso, near 
Grays Peak. The species is said to be common in the forests of the 
Gore Range, east of Toponas; at Hahns Peak; in the San Juan 
Mountains, north of Pagosa Springs and Vallecito; on Lone Cone, 
San Miguel Mountains; and on Veta Pass. I found it very rare in 
the Saguache Mountains, near St. Elmo, Chaffee County. Accord- 
ing to Preble it was tolerably common on Longs Peak in August, 
1894, while Loring secured a fine series of 28 specimens at Silverton 
in the latter part of October, 1893. The majority of the Silverton 
specimens were collected in underbrush in the canyon, but a few were 
taken on the mountain sides near timberline. All were changing 
from brown summer pelage to the white winter coat, and Loring 
states that the white pelage was farthest advanced on individuals 
taken near timberline. Allen says snowshoe rabbits were reported 
as common in the timbered parts of Park County in 187 1. 1 

Lepus californicus melanotis Mearns. Kansas Jack Rabbit. 

Over the entire plains region of eastern Colorado black-tailed jack 
rabbits are found in varying abundance. (See fig. 25.) Formerly 
they were outnumbered by L. campestris in many sections on the 
northern plains, but the reverse is now true. They become increas- 
ingly numerous toward the south, where campestris is rare or absent. 
In Huerfano County, and doubtless elsewhere, they follow some of 
the widest valleys into the foothills, and may be found occasionally 
among the junipers and pinyons as high as 7,500 feet. The species 
seems to prefer open grassy plains to the foothill valleys grown up 

i Bull. Essex Inst., VI, p. 58, ls74. 



156 



NORTH AMERICAN FAUNA. 



[No. 33. 



with Atriplex and Chrysothamnus and is most abundant at some 
distance from the mountains. 

Although present in much smaller numbers than formerly, bkek- 
tailed jack rabbits are still sufficiently numerous in certain sections, 
particularly in the rich agricultural region lying along the Arkansas 
River, to injure seriously such crops as alfalfa, grains, cabbages, and 
sugar beets. In early days jack rabbits were extremely abundant 
in the Arkansas Valley, and for a number of years annual hunts were 
organized at Lamar to lessen the pest. Many thousand jack rabbits 
were sometimes killed in a single hunt. 1 




5 l fPe/s cal /rofiAt/CL/s rsx/^A/as. 



Wtft, L . C AfSL/IA/OT/S. 

Fig. 25.— Distribution in Colorado of black-tailed jack rabbits (Lepus californicus teiianus and L. c. 

melanotis). 

In driving from Cheyenne Wells northwest across the plains to 
Cheyenne, Wyoming, in May and June, 1909, I found these jack rab- 
bits common only in the sandy bunch grass country of Yuma County. 
Elsewhere they appeared to be greatly reduced in numbers. A very 
few were seen on the eastern end of the Arkansas Divide, and farther 
west at River Bend and Ramah, while two immature individuals 
were encountered near Pawnee Buttes, in northeastern Weld County. 
Prof. Lantz found them abundant in the Purgatory Valley south of 
La Junta in April, 1910. 



1 For details of the Lamar rabbit hunts see Jack Rabbits of the United States, by T. S. Palmer, Bull. 
No. 8, Biological Survey, pp. G3-G4, 1897. 



1911.] MAMMALS. 157 

Lepus calif ornicus texianus Waterhouse. Texas Jack Rabbit. 

The black-tailed jack rabbits inhabiting the Upper Sonoran desert 
valleys of western and southwestern Colorado (see fig. 25) are refer- 
able to the Texas form, although intermediate in coloration between 
texianus and melanotis} In some sections they are abundant, in 
others very scarce, and their local abundance varies much from year 
to year, as in other parts of their range. 

In the summer of 1907 I found jack rabbits common at Bayfield, 
La Plata County, and at Coventry, Montrose County. Small num- 
bers were reported also on Mesa Verde, in McElmo Valley; near 
Mancos and Dolores in Montezuma County; and between Naturita 
and Paradox in western Montrose County. Early in June the 
rabbits were quite common along the lower edge of the pinyons and 
in the adjoining open sagebrush valleys and slopes southwest of 
Bayfield, at 6,500 feet. During the heat of the day they were 
usually resting quietly in the shade of the pinyons, but in the early 
morning and toward sundown could be seen actively moving about 
in the open, either nibbling at the short grass in the openings among 
the sagebrush or more often feeding in the grain and alfalfa fields. 
Several ranchmen southwest of Bayfield have small patches of grain in 
the openings along the lower edge of the pinyons, and these suffer 
most from the depredations of jack rabbits. One forenoon at 10 I 
surprised an entire family of rabbits eagerly feeding in one of these 
small isolated fields of young grain among the pinyons. This 
family, consisting of two adults and three or four young about two- 
thirds grown, had been levying heavy tribute upon the tender grain 
shoots, and the field was in a fair way to be entirely destroyed. A 
very little time and effort spent by the owner of this piece of grain 
in shooting jack rabbits would in all probability have saved it. Mr. 
E. G. Bates, of Bayfield, states that a diet of young alfalfa produces 
the same bloating effect on jack rabbits as on cattle and usually 
results fatally. No rabbits thus affected have come under my obser- 
vation, unless it be in the case of a much bloated individual which I 
found lying dead in the sagebrush a short distance from a large alfalfa 
field near Coventry. In the Bayfield region jack rabbits are said to 
be very injurious to orchards in winter, when green food is scarce 
and the animals are forced to subsist by browsing and by eating the 
tender bark of young trees. 

Black-tailed jack rabbits are not common in eastern Montezuma 
County, but toward the Utah boundary and thence west to the 
Abajo (Blue) Mountains in eastern Utah, their numbers are said to 
increase rapidly. I saw a single individual on a fhit in the McElmo 
Canyon, near Moqui, June 22. North of the Montezuma Valley 

i See Nelson, N. Am. Fauna No. 29, p. 145, 1909. 



158 NORTH AMERICAN FAUNA. [No. 33. 

texianus follows the pinyon and sagebrush country around the western 
ends of the Dolores and San Miguel Plateaus, chiefly west of the 
Dolores River, and reaches the San Miguel and Naturita Valleys 
through the Paradox Valley and Dry Creek Basin. It ranges east- 
ward along San Miguel River to Coventry and Norwood and follows 
Naturita Creek nearly to its head at the northwest base of Lone 
Cone, where I saw one in a grassy opening in the oak chaparral at 
8,500 feet, July 27, 1907. At the last three localities the range of 
this species overlaps that of the white-tailed L. c. townsendi, texianus 
being the common form in the lower elevations at Coventry and 
Norwood, and townsendi predominating up around the base of Lone 
Cone, in the Transition zone. I saw no jack rabbits while crossing 
the Paradox Valley on my way to the La Sal Mountains in July, but 
tracks were common in East Paradox Valley. Mr. C. H. Smith, of 
Coventry, reports them very injurious to growing grain, alfalfa, and 
vegetables, especially cabbages. Most of the cultivated land there 
is surrounded by a dense growth of sagebrush, which affords the 
rabbits effective concealment during the day and ample protection, 
as I can testify after several days of hunting with poor success. 

A few black-tailed jack rabbits are reported in the desert areas of 
the lower Grand River Valle} r between Grand Junction and the 
Utah boundary. They are said to occur in small numbers near 
Douglas Spring, at the north base of the Escalante Hills, in western 
Routt County, but this report lacks verification. Mr. J. H. Gaut saw a 
black-tailed jack rabbit east of Antonito, in the Rio Grande Valley, 
in September, 1904, but as it was not secured its species can not be 
determined. On geographic grounds it should be texianus. 

Sylvilagus floridanus similis Nelson. Nebraska Cottontail. 

This small gray form of the small-eared floridanus group of cotton- 
tail rabbits conies into the State from the northeast, along the valleys 
of the South Platte and Republican Rivers and their tributaries, and 
ranges to the eastern base of the foothills, as indicated by the follow- 
ing localities at which specimens have been collected: Arvada, Jef- 
ferson County; Littleton; Barr; Orchard, Morgan County; Masters, 
Weld County; Dry Willow Creek, Yuma County; and Sterling. Mr. 
W. L. Burnett states that a specimen has been taken near Loveland. 

Little is known of the habits of this cottontail in Colorado aside 
from the fact that most of the specimens have been taken in brushy 
thickets along watercourses, just as farther east. S. a. baileyi occurs 
on the plains on both sides of the Platte from the base of the foothills 
to the Nebraska line; so similis doubtless inhabits the wild plum 
thickets and the willow and cottonwood fringe along the river banks. 

Cottontails w r ere very scarce at Wray in December, 1907, but I 
saw a few tracks in the dense plum and hackberry thickets in the 



1911.] MAMMALS. 159 

gulches south of Chief Creek, where similis should be found. A 
skull, however, found at a hole beneath ;i rocky ledge along one of 
the gulches had the characteristic large audital bullae of haileyi. 
S. similis can be distinguished readily from haileyi by its much 
smaller and shorter ears and small audital bullae. The only other 
Colorado cottontail which shares these characters is S. pinetis, a 
species restricted to the mountains. At Sterling, in June, 1909, I 
found this cottontail restricted to the immediate valley of the Platte, 
where one was occasionally jumped in the dense thickets of wild 
cherry and snowberry along the river. They were very wild, and 
the only one collected was a nursing female shot in an alfalfa meadow 
along the roadside June 2. 

Sylvilagus nuttalli pinetis (Allen). Rocky Mountain Cottontail. 

This is the cottontail of the mountain districts of Colorado, where 
it is generally distributed, mainly in the Transition zone. It is most 
abundant in the yellow pine forests of the eastern foothills of the 
front ranges, but is often found along the lower edge of the aspen 
belt. In the Pikes Peak region it occurs as high as 11,500 feet. 1 
It is common on some of the pinyon-clad ridges and mesas of western 
Colorado as low as 6,000 feet along the upper edge of the Upper 
Sonoran zone ; and has been taken on the higher sage plains of Routt 
County with haileyi. The favorite abode of the mountain cottontail 
is on the cool north slopes in the upper part of the pine belt, where 
it finds abundant cover in the creeping juniper (Juniperus sihirica) 
and in dense thickets of aspens, as well as in and among fallen logs. 
Ledges of rock and hollow logs are favorite retreats. This cotton- 
tail, like grangeri, belongs to the short-eared nuttalli group, and can 
not be easily confused with the long^-eared cottontails, haileyi and 
warreni, of the surrounding Upper Sonoran plains and valleys. The 
rich dark winter pelage of pinetis is usually strongly tinged with 
vinaceous. 

Sylvilagus nuttalli grangeri (Allen). Black Hills Cottontail. 

E. W. Nelson, in his recently published monograph of the rabbits 
of North America, 2 refers to this form specimens of cottontails from 
the Escalante Hills and Lay, in western Routt County, and from 
Meeker. This is a northern member of the nuttalli group closely 
related to pinetis, with which it intergrades in southern Wyoming. 
S. a. haileyi, a member of the long-eared auduhoni group, is also found 
over much of western Routt County. The cottontail rabbits of north- 
western Colorado are not typical of any race, but appear to be inter- 
grades — grangeri, pinetis, and possibly haileyi, being involved. A 
specimen I shot on the sage plains near Lay in August, 1905, seems 
to be intermediate between haileyi and grangeri. The cottontails of 

i Warren, The Mammals of Colorado, p. 49, 1910. 2 N. Am. Fauna No. 29, p. 207, 1909. 



160 NORTH AMERICAN FAUNA. [No. 33. 

the northwestern sage plains thus present a most perplexing problem. 
To quote Mr. Nelson (1. c, p. 206): "One specimen in the Biological 
Survey collection (No. 139098) from Lay, Colorado, is indistinguish- 
able in external characters from three specimens of S. a. haileyi from 
the same place, but its skull is that of grangeri, to which it has been 
referred. Several other specimens of cottontails, some haileyi and 
some grangeri, from northwestern Colorado are extremely puzzling, 
and much more material from there and elsewhere in this State is 
needed before the relationships and ranges of the several cottontails 
can be satisfactorily determined." 

In the Escalante Hills I found the short-eared cottontails in small 
numbers at the edge of the yellow pines at about 7,000 feet, and 
others were seen down among the dense pinyon and juniper growth 
on the northern slopes. 
Sylvilagus auduboni baileyi (Merriam). Plains Cottontail. 

The common long-eared cottontails of the eastern plains of Colo- 
rado are haileyi, the only other form occurring there being the short- 
eared similis, which inhabits the brush patches and fringe along the 
South Platte and other streams of the northeastern counties. These 
two cottontails are quite unlike, and besides their structural differ- 
ences have dissimilar habits. Thus haileyi inhabits the open grassy 
plains, where it lives in abandoned prairie dog and badger holes, or 
else takes up its abode in the rock ledges and bluffs bordering the 
valleys; while similis, so far as known, does not dwell in the open 
but in dense thickets along streams and in the bottoms of connecting 
gulches. The plains cottontail has a wide range from Montana south 
to the edge of the Llano Estacado of northern Texas. It occupies 
practically all the plains of Wyoming, extending south in northwest- 
ern Colorado on the sage plains and in the valleys of Routt and Rio 
Blanco Counties, where the high escarpment of the Book Cliffs and 
the White River Plateau separates it from warreni of the Grand Val- 
ley and southward. The range of haileyi meets that of the mountain 
cottontail (S. pinetis) at Meeker, Craig, and other points along the 
bases of the Elk Head Mountains and of the White River Plateau, 
as well as on the slopes of the higher divides between the river valleys. 
This species follows up the drainage of the Arkansas Valley to Salida, 
and probably occurs on the extensive Upper Sonoran flats farther up 
the valley, but in other sections is not known to penetrate the eastern 
foothills for any distance. 

Throughout its range the plains cottontail is preeminently an in- 
habitant of the semiarid Upper Sonoran plains, where it lives in holes 
along the steep-cut banks of dry arroyos, in the deserted burrows of 
prairie dogs and badgers, and often in holes beneath sagebrush, Atri- 
plex, or prickly pear (Opuntia polyacantha) on the open plain. In 



1911] MAMMALS. 161 

rough and broken regions, as along the Snake, Bear, and White Riv- 
ers, and in the juniper country of Baca and Las Animas Counties, it 
lives chiefly along the rocky rims of canyons. 

In August, 1905, cottontails were abundant in the Bear River Val- 
ley south of Lay. In the evening and early morning, and often at 
midday, numbers usually fed on the small grassy flats between the 
river and the base of the bluffs. They were quite wild, despite their 
abundance, and when alarmed scampered up the arroyos and dry 
rocky slopes in all directions to their retreats in the ledges far above 
the river. Sometimes fully 20 individuals, adults and young, were 
in sight at once. 

In traveling down the Snake River Valley in August, 1906, few 
cottontails were observed as compared with the great numbers seen 
on Bear River in 1905, and Mr. John Criss, of Baggs Crossing, Wyo- 
ming, informed me that a disease which he termed cholera had been 
thinning out their numbers very perceptibly throughout the region. 
In common with other rabbits haileyi is subject to a periodical disease, 
as yet very little understood, which invariably follows excessive 
abundance. This disease seems to be nature's check to abnormal 
increase, and did it not prevail, at least among the plains cottontails, 
much of the cattle range in the vicinity of streams and gulches would 
undoubtedly be ruined. As it is, the injury to the range is very con- 
siderable when the rabbits reach their maximum numbers, as on Bear 
River in 1905, and again on McElmo Creek and along the lower San 
Juan River in 1907, where S. a. warreni was so abundant that scarcely 
a spear of grass remained in the vicinity of the streams. 

This species is fairly free from parasites and grubs and, except in 
the years when it is suffering from disease, is excellent food. On the 
sage plains of Routt County young cottontails were an important 
item on our bill of fare, and they were well-flavored and tender. 

Sylvilagus auduboni warreni Nelson. Colorado Cottontail. 

Sylvilagus auduboni warreni Nelson, Proc. Biol. Soc. Wash., XX, p. 83, July 22, 
1907. Type from Coventry, Montrose County, Colorado. 

This is the cottontail of the warm Upper Sonoran valleys and 
lowest mesas of southwestern Colorado south of Grand River Valley. 
The cottontails of the open plains of the San Luis Valley, although 
not typical, are referable to this form and, especially toward the 
northern end of the valley, approach haileyi in general paleness of 
coloration. The ranges of these two closely related forms appear 
to meet in this region, warreni doubtless extending around the 
southern end of the San Juan Mountains in New Mexico and then 
north into the San Luis Valley, while haileyi probably reaches the 
region from the south and east by following around the southern 
end of the Culebra Range. Under the Circumstances it seems best 
90432°— 11 11 



162 NORTH AMERICAN FAUNA. [No. 33. 

to refer all of the San Luis Valley cottontails to warreni, since this 
type predominates. The northern boundary of the dispersion of 
warreni is marked by the White River and Book Plateaus, the cot- 
tontails found in the Grand Valley as far east as Rifle being fairly 
t vpical, while those occurring in the White River Valley from Meeker 
west are baileyi. 

This form, like the other long-eared cottontails, is an inhabitant 
of sage plains, Sarcobatus valleys, and Atriplex flats, and is often 
found also among the rocks and pinyons. Cottontails were uncom- 
* mon in the Grand Valley in both 1906 and 1907. Near Rifle, several 
were seen in the greasewood along arroyos and irrigation ditches, 
and north of Fruita, Mesa County, the rabbits were often started 
from their forms beneath Atriplex confertifolia on the open desert, 
or from the dense thickets of Chrysothamnus along the irrigation 
ditches. They invariably took refuge in the deserted burrows of 
badgers and white-tailed prairie dogs, in which apparently they 
were living. They were abundant at Hotchkiss in August, 1907, 
and one seen just east of Crawford, at the west base of the West 
Elk Mountains, was probably this form. At Coventry I found 
them abundant during July, both among the pinyons and out in 
the sagebrush, and collected several topotypes. On my trip to the 
La Sal Mountains they were seen in abundance in the Dry Creek 
and Paradox Valleys, but were scarce in Sinbad Valley and thence 
down the Dolores River to the mouth of West Creek. My camp 
assistant shot 11 young cottontails from the tent while cooking 
supper at our Dry Creek camp, 3 miles west of Naturita, July 19, 
and there seemed to be fully as many more among the rocks and 
sagebrush after his fusillade. 

These rabbits were very abundant in the McElmo and San Juan 
Valleys in June, 1907, where they were reported quite injurious. 
Ranchmen stated that their numbers near Bluff City, Utah, were so 
great that nearly all the range grasses in the vicinity of canyons had 
been eaten by them before the middle of June. Mr. George J. Ash- 
baugh, who lives in the McElmo Canyon west of Moqui, says the 
cottontails are very injurious to his fruit trees during the winter. 
On the Mesa Verde these rabbits are scarce, only two being seen 
among the pinyons at 7,000 feet. They were fairly common at Bay- 
field, La Plata County, early in June, and a single individual noted 
in a willow copse along the San Juan River at Arboles may have 
been this form. 

The type of this interesting cottontail, a female in winter pelage, 
measures: Total length, 375; tail vertebrae, 51; hind foot, 102; length 
of ear from notch, in dried skin, 70. It was collected at Coventry 
by Mr. C. H. Smith, January 4, 1907, and is in the Biological Survey 
collection. Regarding the characters of warreni, Nelson says (1. c): 



1911.] MAMMALS. 163 

"Similar to baileyi in size, length of ears, and abundant pelage, 
but darker colored with more distinct gray rump patch and darker 
rufous on nape and legs." Its distribution is given as "south- 
western Colorado and adjacent parts of Utah, New Mexico, and 
Arizona." 

Felis oregonensis hippolestes Merriam. Cougar; Mountain Lion. 

The mountain lion was formerly present over at least all the rough 
parts of the State, and in early times it was occasionally seen even 
well out on the plains along the more heavily brush-fringed streams. 
At present it is becoming rare east of the Continental Divide, although 
holding its own fairly well in the rough canyon and mesa country 
of the west and southwest. It is now most numerous in the pinyon 
country of Montezuma and Dolores Counties, and in western Kio 
Blanco and Routt Counties, the latter region being to-day perhaps 
the best lion country in the United States. 

Probably F. hippolestes is the only form represented in Colorado, 
but a skin from Montezuma County, in the possession of Mr. Steve 
Elkins, of Mancos, is considerably paler and less reddish than lions 
from the Meeker region. Unfortunately there are no skulls or skins 
from extreme southwestern Colorado available for study. A series 
of 12 skulls from Meeker in the Biological Survey collection, collected 
by Mr. Theodore Roosevelt in January and February, 1901, have 
been referred to F. hippolestes. 1 

In some of the southern and western counties mountain lions are 
sufficiently numerous to be very destructive to stock, especially 
young colts. Near Lily Park, in western Routt County, calves also 
are said to be often killed by lions. Through Mr. James Lowell, of 
Dolores, a forest ranger, I learned that 16 colts had been killed by 
mountain lions in that region during the spring of 1907 — 11 in the 
yellow pine country between Plateau and Beaver Creeks, 20 miles 
north of Dolores, and 5 in the high aspen and spruce country on Bear 
Creek, an affluent of the Dolores River, about 20 miles east of Dolores. 
In addition to colts and calves, the lions prey much upon wandering 
bands of s'heep in the yellow pine forests of Archuleta County. 

Among game mammals deer appear to suffer most from mountain 
lions, and that they form the chief prey is evident from the fact that 
the lions move up into the mountains or down into the low country 
with the migration of the deer. Mr. E. E. Chapson, a forest ranger 
in the San Juan National Forest, thinks the lions and coyotes are the 
most important factors in the destruction of deer in the San Juan 
Mountains. The lion tracks seen in winter are almost invariably 
following deer trails, and, as the bodies of deer which have been 
killed by mountain lions are often found, Mr. Chapson thinks they 
kill many more deer than they require for food. Mr. Steve Elkins 

i Proc. Wash. Acad. Sci., Ill, p. 586, 1901. 



164 NORTH AMERICAN FAUNA. [No. 33. 

states that lions are very destructive to deer in the Montezuma 
National Forest. Mr. J. P. Galloway, of Norwood, relates that a 
lion, the track of which he followed several years ago on Wild Steer 
Mesa, south of East Paradox Valley, had dragged the carcass of a 
freshly killed deer to a secluded rocky place among the pinyons and 
left it covered with a pile of pinyon needles and cones fully 4 feet 
high. This habit of caching carcasses for future consumption has 
been noted by a number of Colorado hunters. 

Mountain lions are much hunted with dogs in the regions of their 
greatest abundance, as in the Keystone country northwest of Meeker 
and in the Mancos region. When pursued by dogs, they are readily 
treed, usually after a short dash, seeking refuge among the upper 
branches of pinyons or junipers, where they are at the mercy of the 
hunters. The Meeker region has long been a famous lion country, 
and it was here that Roosevelt had his well-known hunt in the winter 
of 1901, bringing out a fine series of specimens, besides gathering 
important data on the habits of the species. 

Although most abundant in the broken rocky pinyon and juni- 
per country on the lower western slope of the Continental Divide, 
lions are nevertheless found occasionally above timberline. On the 
Saguache Mountains, 2 miles north of St. Elmo, I followed the track 
of a medium-sized lion through the snow for some distance, October 
9, 1907. It was first crossed at 12,000 feet, and as far as followed, 
kept along the high wind-swept crests of the mountains above 
timberline. 

Mountain lions were reported in varying numbers at the following 
localities in the northern mountains in 1905 and 1906: 

Gore Range and mountains surrounding Middle Park: Small 
numbers reported. 

Park Range (headwaters of Grand Encampment River) : Reported 
by lumbermen as not at all uncommon. Tracks seen in the trail 
near the tie camps August 13, 1906. 

Snake River region (Baggs Crossing to Escalante) : Formerly com- 
mon, but none remain. Mostly poisoned by professional wolf 
trappers some years ago, when there was a large bounty on wolves 
in this region. 1 

Browns Park: Fair numbers reported in mountains south of Green 
River, near Mount Cullom. 

Escalante Hills: Three killed on south slope, 10 miles west of Lily, 
in the winter of 1905-6. 

Lily Park: Reported common in surrounding pinyon country in 
winter. 

1 This suggests that a liberal use of poison in a region where lions are troublesome would be the best 
means of reducing their numbers. Old trappers and wolf hunters state that lions readily eat of poisoned 
carcasses. 



1911.] MAMMALS. 165 

Rangely (lower White River): Not uncommon. Several killed 
each winter. 

Book Plateau (near Baxter Bass) : Reported as occasional in winter. 

The following data on the abundance of lions in southern Colorado 
were secured in 1907: 

Culebra Range (south of La Veta) : Reported not uncommon in 
wilder parts, but none recently killed. One killed on Cucharas 
River several years ago. 

Sangre de Cristo Range (Sierra Blanca group) : Reported rare. 
Track seen on Sierra Blanca winter of 1906-7. 

Sangre de Cristo Range (head Huerfano River) : A few reported. 

Cochetopa Hills (near Saguache) : Said to be scarce. 

Mancos region : Three killed by Mr. Steve Elkins during past year- 
I saw a young lion's track in Navajo Canyon, on the Mesa Verde, 
June 13. 

Sierra el Late : A few reported around Ute Peak. 

Lower San Miguel and Dolores Rivers: Becoming scarce. 

Lone Cone (San Miguel Mountains) : Has never been common. 

Coventry: Scarce. Several have been killed along the San Miguel 
River, but none recently. 

Uncompahgre Plateau: Rare. 

Yallecito: Occasionally met with. 

Over much of the eastern slope mountain lions are very rare, where 
they were formerly common. Allen says, regarding its former presence 
in Park County: "Not uncommon. Its cry was once heard near 
our camp at Montgomery." l Trippe 2 records it as an early inhabit- 
ant of Clear Creek County. It was reported from Estes Park during 
the early nineties of last century. Coues mentions two mounted 
specimens in the collection of Colorado mammals exhibited at the 
Philadelphia Exposition in 1876 by Mrs. M. A. Maxwell, stating that 
"one was killed near Boulder by poisoning the carcass of a young 
horse which the panther had destroyed." 3 

Lynx canadensis Kerr. Canada Lynx. 

The Canada lynx inhabits the Canadian zone forests of the higher 
mountains in Colorado, but in most sections its numbers are rapidly 
decreasing. The scattering records at hand indicate a former gen- 
eral distribution over the central and northern mountainous parts of 
the State, while a few are still left in the San Juan and La Plata 
Mountains of the southwest. At present the animal occurs chiefly in 
the heavy forests of the Park and Gore Ranges, the Rabbit Ear and 
Vasquez Mountains, and in southern Pitkin and Eagle Counties. It 
is said seldom to wander below 8,000 feet, 4 even in the heaviest snows 
of winter. 



i Bull. Essex Inst., VI, p. 53, 1874. 

2 See Coues, Birds of the Northwest, p. 224, 1S74. 

3 Dartt, On the Plains and Among the Peaks, p. 218, 1879. 

4 A lynx reported to have been killed at Bayfield (0,500 feet) is an exception. 



166 NORTH AMERICAN FAUNA. [No. 33. 

In 1905 lynxes were reported to be tolerably common in the moun- 
tains surrounding Middle Park. Alpert & Co., of Kremmling, pur- 
chased several skins taken in the winter of 1904-5 in the Williams 
Mountains, near the headwaters of the Williams Fork of Grand 
River; and Mr. Fred Selak, a fur buyer living near Coulter, annually 
handles a small number of skins from both the Grand Lake region 
and the Rabbit Ear Mountains. Lynxes are said to leave the higher 
mountains in February and March and come down into the forested 
country of the Grand Lake region, following the downward move- 
ment of the grouse and ptarmigan. The North Park slope of the 
Rabbit Ear Mountains is very good lynx country, and Mr. W. H. 
Graham, of Spicer, informed me that he and his brother usually trap 
from 10 to 15 each winter near the head of Arapahoe Creek. On 
the Medicine Bow Range and in the lodgepole pine forests east of 
the Laramie River lynxes are said to be uncommon. A few are 
reported by lumbermen on the Park Range along the headwaters of 
the Grand Encampment River, but little trapping appears to be 
done in that section. Dr. Kerneghan, of Steamboat Springs, has 
three fine lynx skins which were taken on the west slope of the Park 
Range during recent years, and states that in the winter of 1904-5 
he saw tracks of a lynx in the aspen thickets on a mountain 2 miles 
south of Steamboat Springs at about 7,500 feet. Mr. J. R. Carron, 
storekeeper at Columbine, near Hahns Peak, usually buys two or 
three skins each year, but in 1906 none were brought in. A lynx 
taken in the Elk Head Mountains in the winter of 1905-6 was sold 
to Mr. Robert Mcintosh, of Slater. According to Mr. A. G. Walla- 
han, a few are still found in the Williams River Mountains. Mr. 
Dall DeWeese, of Canon City, has a mounted specimen from the 
South Fork of White River, where he says lynxes were not at all 
uncommon some years ago. Mr. Frank Hayes, a taxidermist of 
Glenwood Springs, states that in the winter of 1903-4 he saw a lynx 
track at Mud Springs, on White River Plateau, and in 1905 purchased 
five skins which had been taken the preceding winter near Mount 
Jackson, at the northern end of the Saguache Range. While located 
at Aspen just previous to 1900, Mr. Hayes purchased six or eight 
lynx skins each winter, taken in the three following regions: Italian 
Mountain and Taylor Park, which are on the headwaters of Taylor 
River; region about Snow Mass Peak; and Independence Pass, at the 
head of the Roaring Fork of Grand River. 

A very few lynxes were reported in 1907 in the mountains north 
of Pagosa Springs, and Mr. Don C. Coulson, of Bayfield, has handled 
a few skins from the high country in the Vallecito region, and also 
one skin which was taken in the winter of 1905 on a ranch adjoining 
the town of Bayfield (6,500 feet). Mounted specimens seen at Sil- 
verton and Ourav were doubtless killed in the neighboring moun- 



1011.] MAMMALS. 167 

tains, but I could learn nothing definite concerning their history. 
Apparently tins lynx is more numerous in the La Plata Mountains 
than in the San Juans. Mr. Steve Elkins, of Mancos, has trapped 
several in the spruce belt, and states that his hounds occasionally 
tree one while following a bear trail. This species is known as the 
snowshoe lynx in the La Plata Mountains. 

A winter skin from Grand Lake, in northeastern Middle Park, is in 
the Biological Survey collection. 

On the occurrence of Lynx canadensis in Park County, Allen says: 
"Represented as common. Saw skins of this species in the posses- 
sion of hunters, taken in the vicinity of Mount Lincoln. " 1 Warren 
mentions a skin which Air. C. E. Aiken, of Colorado Springs, received 
from Beulah, and which is supposed to have come from either the 
Wet Mountains or the Sangre de Cristo Range. 3 

Lynx baileyi Merriam. Plateau Wildcat. 

The bobcats of the lower parts of southern and eastern Colorado 
are referable to L. baileyi. The few Colorado specimens at hand do 
not permit an accurate outline of the distribution, but this wildcat 
appears to be most abundant in the Upper Sonoran zone. Along the 
eastern slope the species ranges a short distance into the foothills, 
and in the southwest, is found commonly over an extensive area of 
rocky pinyon and juniper country. Bobcats are rare or entirely 
absent over much of the plains region east of the mountains, where 
the brush fringe and the few rocky ledges and bluffs along some of 
the streams furnish the only suitable environment. A few have been 
killed in the rocky canyons along Chief Creek, near Wray, Yuma 
County, during recent years, and in May, 1909, I saw tracks of a 
bobcat among the sandstone ledges near Tuttle, on the South Fork 
of Republican River. The animals have never been common in 
the Chimney Cliffs, northwest of Sterling, according to residents. 
The extensive juniper country in western Baca, Las ^Animas, and 
southern Otero and Bent Counties, however, is an ideal habitat, and 
bobcats are abundant throughout that region. They are reported 
common in the lower foothills from Fort Collins and Arkins south to 
Gardner, La Veta, and Trinidad, and in the southwest I have seen 
tracks at Arboles, Bayfield, on the Mesa Verde, near McElmo, East 
Paradox Valley, Sinbad Valley, south of Grand Valley, Plateau 
Creek, and on the desert north of Mack, lower Grand River Valley. 3 
Bobcats reported from the cottonwood-fringed streams along the east 
side of the San Luis Valley are probably baileyi, as a specimen has 
been taken on Conejos River at the southern end of the valley. 

Bobcats prey much upon rabbits, wood rats, and other small cliff- 
dwelling mammals, and in the open valleys sometimes subsist to a 

1 Bull. Essex Inst., VI, p. 53, 1S74. * Some of the above notes may refer to L. uinta. 

2 Mammals of Colorado, p. 25S, 190ti. 



168 NORTH AMERICAN FAUNA. [No. 33. 

large extent upon prairie dogs and pocket gophers. Locally they are 
probably to be considered useful animals, the balance in their favor 
being due to the destruction of so many noxious rodents. On the 
other hand, wherever poultry is within reach they commit serious 
depredations, and near Pagosa Springs they are said to be very 
destructive to sheep. 

Lynx uinta Merriam. Mountain Wildcat. 

Although represented by specimens chiefly from the northwestern 
part of the State, this appears to be the species of the higher foothills 
and mountains generally, where it replaces haileyi of the lower foot- 
hills and canyons. It is a more robust animal than haileyi, with the 
cranium proportionally larger, and in size approaches L. canadensis, 
with which species it occurs at some of the higher elevations. 

Bobcats are abundant in the rough pinyon and juniper country 
of Routt, Rio Blanco, and Garfield Counties. In 1906 the greatest 
numbers were reported northwest of Meeker and west of Snake 
River between Baggs Crossing and Escalante. Dogs are often used 
in hunting bobcats in the Meeker region, as the animals are readily 
treed after a short run. A series of skulls from near Meeker was 
collected by Mr. Theodore Roosevelt in 1901 and added to the Bio- 
logical Survey collection. 

Cottontails and snowshoe rabbits probably form most of the food 
of this species. Just below timberline in the Saguache Mountains 
near St. Elmo, and in other localities ranging between 10,000 and 
11,000 feet elevation, I have followed the tracks of bobcats in the 
snow as they crossed and recrossed the dense willow copses in the 
trails of snowshoe rabbits, which they had evidently been hunting. 
In the lower part of its range, as in western Routt County, prairie 
dogs are next in importance on the summer bill of fare. 

While walking along the railroad track near Rogers Mesa, in the 
North Gunnison Valley, in October, 1907, I had an excellent oppor- 
tunity to watch the method by which the bobcat hunts prairie dogs. 
I was just emerging from a deep cut when I saw a large reddish bobcat 
at a distance of not over 40 feet. It was sneaking through the 
scattering greasewood bushes flat upon its belly, its short tail twitch- 
ing nervously, and the excited chattering of prairie dogs on a neigh- 
boring flat showed that its approach had been noted by the alert 
animals. One large old prairie dog in particular, apparently the 
cat's intended victim, was seated at its burrow on the edge of the 
town, chattering in a bantering maimer and appearing less frightened 
than the rest. The burrow was within leaping distance (about 10 
feet) of the edge of the greasewood, and in making its approach the 
cat took advantage of every bush, stopping in the cover of each for 
a few moments. When it reached the last bush and was gathering 



1911.] MAMMALS. 169 

itself for the final leap, the old prairie dog disappeared, but only just 
in time, as in another moment the cat landed on the rim of the bur- 
row. Rapid, nervous jerks of the tail showed the cat's disappoint- 
ment as it glared about in different directions. Up to tins time my 
presence had not been noted, and not until I had thrown several 
stones did the cat see me, whereupon it bounded away across the 
dog town in long leaps. The section men working along the railroad 
stated that they often saw a cat near tins colony, and it doubtless 
had its den in the neighboring rock ledges along the North Gunnison 
River, living easily on the fat denizens of the town. 

On September 7, 1906, I saw two bobcats in the Sarcobatus brush 
along Snake River, a mile north of Lily, and tracks were often seen 
in the dry sandy beds of arroyos in the valleys of northwestern 
Colorado. 

Canis occidentalis Richardson. Gray Wolf. 

Gray wolves were formerly abundant over practically the entire 
State, except possibly the highest mountains, and were especially 
numerous on the eastern plains, where large bands preyed upon the 
buffalo. From this habit of hanging on the flanks of the large herds, 
they were generally known as buffalo wolves. The mountain animals 
are said to average much darker than those of the plains. Unfor- 
tunately there are no specimens available from the mountains to 
settle this point, but it is unlikely that two forms occur in the State. 
Wolves are still found in considerable numbers in North Park and 
in Routt and Rio Blanco Counties, where they kill a great many 
range cattle. A few are probably found throughout the mountains 
west of the main ranges, and small numbers are still present over 
the more unsettled parts of the eastern plains region, particularly in 
Baca and eastern Las Animas Counties, in the extreme southeast, 
where, in 1907 and in 1910, they were said to be common and to kill 
a great many sheep. 

In 1906 wolves were common over most of Routt County, notwith- 
standing the bounty of $15 authorized by the local stock associa- 
tion, the additional $10 offered by the county, and the efforts of 
several professional wolf trappers employed by the association. 
The heaviest losses of stock were at that time incurred on the Iron 
Springs Divide and south of the Elk Head Mountains, although 
wolves were reported as unusually abundant in Browns Park on 
Green River. In the latter region the stock association hired three 
or four trappers to reduce their numbers, and about fifty were killed 
during the winter of 1905-6, the majority being trapped. Mr. John 
Criss, a trapper of many years' experience in the Snake River coun- 
try, informed me that the wolves have been so persistently hunted, 
trapped, and poisoned that they will now rarely come to a scent of 



170 NORTH AMERICAN FAUNA. [No. 33. 

any description and seldom to a baited trap, while poisoning is unsuc- 
cessful. He has had the best success with traps set blind and placed 
in trails or near water holes in the badland country, several miles 
back from the Snake River Valley. A method of fastening wolf 
traps, used successfully by Mr. Criss in cold weather, seems worthy 
of mention. The trap is securely chained to a bush or stake at the 
edge of a steep-walled gully or wash so that the wolf in its struggles 
to escape will leap over the edge and hang half suspended and 
helpless, unable to regain the top of tlu? bank. Wolves thus trapped 
in severe weather usually freeze to death in a few hours. 

An impression prevails among stockmen in northwestern Colorado 
that wolves retire to the mountains to whelp, but I find no evidence to 
support this theory. 1 In Dixon, Wyoming, I saw a nearly adult black 
wolf in captivity, which had been captured as a cub in a den among 
the Snake River bluffs, 20 miles west of Baggs Crossing, in the spring 
of 1905. This individual was kept in a large cage in the back yard of 
its owner in Dixon. A boy of 3 years was petting and stroking its 
head through the bars, and the wolf's every movement betokened its 
pleasure in the companionship of the little fellow. All playfulness 
immediately left it, however, on the approach of a man, when the 
wild, untamable wolf nature was revealed in bared fangs, curling 
lips, and glaring eyes. The mother of this wolf was gray, as was 
also one of the three cubs captured in the den. According to trappers 
both black and white wolves occur, but white ones are said to be 
extremely rare. 

In the Lily Park region, on the lower Bear River, Mr. F. C. Barnes 
states that wolves were numerous until 1902, but during the two years 
following a trapper named Snyder killed 61. Since that time few 
cattle have been killed in that section by wolves. In 1905 wolves were 
reported in considerable numbers in the White River country, partic- 
ularly in the valley of the Piceance, but were scarce near Rangely 
in 1906. During the winter of 1904-5, 7 were killed out of a band of 
nearly 25 which was ranging in North Park, but in 1906 wolves 
were reported scarce in that region. I often saw wolf tracks in the 
trail as we traveled through the parks on the divide east of the Lara- 
mie River, in August, 1906, and the animals were then said to be very 
troublesome in that section. Tracks were observed as high as 10,000 
feet. Wolves are of rare occurrence in Middle Park, but two are said 
to have been seen on the stage road near Coulter during the winter of 
1903-4, and another near Grand Lake the following winter. One of a 
band of three which ranged on the head of Willow Creek, in the 
northern part of Middle Park, was killed early in the summer of 1906. 
In Egeria Park and on the Gore Range wolves are reported as of rare 

1 On this point, see Bailey, Wolves in Relation to Stock, Game, and the National Forest Reserves, 
Bull. 72, U. S. Forest Service, 1907. 



1011.] MAMMALS. 171 

occurrence. They were uncommon over most of southern Co.orado 
in 1907, particularly in the San Luis Valley, the Pagosa Springs 
region, and in Montezuma County, where they are considered very 
rare. According to Mr. Steve Elkins, of Mancos, none have been 
reported in that region since the winter of 1904-5, when four or five 
were seen between Cortez and Mancos. In the region contiguous to 
the upper waters of the Vallecito and Los Pinos, in northeastern La 
Plata County, they are said to be increasing during the past few years, 
but no serious damage is reported. Forest Supervisor E. W. Shaw, 
of Durango, states that a band of 12 was seen near Vallecito in the 
winter of 1906-7. A few wolves were reported from the western part 
of San Miguel and Montrose Counties, a large male having been 
killed in the Dry Creek Basin in the winter of 1906-7, and a female 
with four whelps was stated to be ranging the same region in the 
summer of 1907. According to Warren, wolves were reported in 
the fall of 1906 to be increasing on the Black Mesa, south of the West 
Elk Mountains; 1 

Dr. A. K. Fisher reported wolves as common near Las Animas in 
1892 and in the Estes Park region in 1894, and according to Streator, 
numbers were to be found the same year on the Republican River, 
north of Burlington, and in the vicinity of Olney. Prof. Lantz reports 
that a band of three was often seen in the vicinity of Hugo during the 
winter of 1904-5. The rough canyon country of Las Animas, Baca, 
and southern Otero and Bent Counties was in early days resorted to 
by large numbers of wolves for breeding purposes, and many still 
breed in that region. 

Ranchmen living in northwestern Logan and northeastern Weld 
Counties stated in the summer of 1909 that wolves were very scarce 
in that section, only one being known to inhabit the Horsetail Basin 
south of the Chimney Cliffs. This is said to be a female, and is sup- 
posed to be the mother of eight whelps which were dug out of a den in 
the rough country on the head of Deadman Creek, 20 miles northeast 
of Avalo, in the spring of 1909. In the spring of 1908 a litter of six or 
seven was dug out in the same canyon, two of which were taken alive 
to Nebraska, and another one was kept on a ranch north of Sterling 
until it became vicious, when it was killed. In 1908 a cowboy named 
Frank Jordan is stated to have roped an old male wolf on the open 
plains in the same vicinity. 

Allen states that "Canis lupus" was comparatively scarce in Park 
County in 1871, although formerly abundant there. 2 As " Cani 
occi&entdlis'''' Trippe records the wolf as an early inhabitant of Clear 
Creek County. 3 

i Further Notes on Mammals of Colorado, p. S2, 1908. 

2 Bull. Essex Inst., VI, p. 54, 1874. 

» See Coues, Hirds of the Northwest, p. 224, 1874. 



172 NORTH AMERICAN FAUNA. [No. S3. 

Canis lestes Merriam. Mountain Coyote. 

This large, dark, richly colored species occurs more or less abun- 
dantly throughout the mountains, ranging from the lower foothills to 
above timberline. Its habits are not unlike those of the coyotes of 
the plains and valleys, but its prey is somewhat different, consisting 
chiefly of dusky grouse, cottontails (Stjlvilagus n. pinetis), and snow- 
shoe rabbits, and including also many deer and fawns. A small band 
of these coyotes hunting together has even been known to kill several 
mountain sheep which had been shut into a small pocket by an 
avalanche. 1 Fawns are of course preyed upon in summer and early 
fall, bat the adult deer only in winter when the crust will sustain the 
coyotes but not the deer. I am informed by forest rangers who have 
seen coyotes pursuing deer in this manner that a band of five or six 
will overtake a deer and hamstring it very quickly on a weak crust of 
snow. Many calves also are killed by coyotes in the mountain parks, 
and in certain localities it is almost impossible to raise chickens and 
turkeys because of their depredations. 

At the Medano Springs ranch, in the San Luis Valley, coyotes 
were unusually abundant and destructive in October, 1907. Num- 
bers were seen each morning on the broad hay meadows west and 
north of the ranch buildings, where they mixed freely with the cattle, 
and evinced little fear of man unless he carried a gun. Several were 
noted lying quietly on the tops of haystacks, from which they could 
detect anyone approaching. During the day one was in almost every 
extensive weed patch or growth of rank marsh grass, ready to pick 
up the turkeys and chickens which strayed too far from the ranch 
buildings. On the meadows near Saguache coyotes have been seen 
catching meadow mice and playing with them like a cat. In the 
yellow pine forests of Archuleta County coyotes are very destructive 
to sheep, notwithstanding the night fires kept burning by the Mexican 
herders. 

On the Saguache Mountains near St. Elmo I saw many fresh coyote 
tracks in the snow above 11,000 feet, October 9, 1907. One track 
followed a high ridge above timberline, at 12,500 feet. I have noted 
tracks also on the summit of Berthoud Pass, and coyotes are commonly 
reported on Grays Peak and in other sections of the high mountains. 

Canis nebracensis Merriam. Plains Coyote. 

The coyotes of the eastern plains region are referable to C. nebracen- 
sis, the type of which is from Johnstown, Nebraska, and specimens 
from the sage plains of North Park and Koutt County seem nearest 
this species. This light-colored coyote inhabits the lower levels, 
being replaced in the higher foothills and mountains by C. lestes, a 
much darker animal. 

1 See note under Ovis canadensis, p. f>4. 



1911.] MAMMALS. 173 

On my trips through the northern parts of the State I found co}^otes 
most numerous on the extensive sage plains bordering the Snake and 
Bear Rivers, in Routt County, although they are abundant in North 
Park and at many points east of the mountains, especially in the 
Wray region and northwest of Sterling. Wherever found in any 
numbers in settled districts, coyotes kill a great many sheep, young 
calves and pigs, and much poultry. They are heartily hated by 
stockmen and fanners alike, who never lose an opportunity to kill 
them. The stock association in Routt County was offering a bounty 
of $1 on coyotes in 1906, but most of the professional wolf trappers 
claimed that this was not sufficient remuneration, and hence coyotes 
were increasing rapidly at that time. A few trappers, however, were 
at work on coyotes in the Snake River region in 1906. Mr. John 
Criss, of Baggs Crossing, Wyoming, who makes his winter head- 
quarters near Sunny Peak, is said to trap and poison about 200 coy- 
otes each winter. While I was traveling down the Snake River 
Valley coyotes were numerous and bold, and were heard at all hours of 
the day and night. 

Coyotes partly compensate for the damage to live stock and poultry 
by catching great numbers of noxious rodents, such as pocket gophers, 
prairie dogs, ground squirrels, rabbits, and mice — particularly meadow 
mice, of which they appear very fond. One noted in a hay meadow 
near Lake Jolm in the western part of North Park, August 9, was 
so busy hunting for meadow mice in the rank grass that it appeared 
oblivious of the presence of eight teams at work in the field and did 
not become alarmed until my companion shot at it. 

Mr. W. O. Potter, of Avalo, states that once on the plains of north- 
eastern Weld County he saw an eagle flying heavily with a captured 
prairie dog, and under the eagle a coyote following along with the 
apparent intention of pouncing on the quarry when the eagle became 
wearied and alighted to eat it. 

Canis estor Merriam. San Juan Coyote. 

Doubtless all the coyotes inhabiting the warm Upper Sonoran 
deserts and valleys of southwestern Colorado should be referred to 
C. estor, which was described from Noland's ranch, San Juan River, 
Utah. The Book Cliffs appear to separate this small pale desert 
species from C. nebracensis of the Routt County sage plains. 

Coyotes are abundant near Mancos, on the Mesa Verde, in the 
McElmo Valley, in the region of the lower San Miguel and Dolores 
Rivers, and in the lower Grand Valley. The chief damage appears to 
be to sheep and poultry, although in the McElmo Valley they are 
said to eat a great many watermelons and cantaloupes on the vines. 
A coyote seen near McElmo June 20 was small and light-colored, and 
in all the lower valleys the small pale coyotes are distinguished 
locally from the large dark animals of the mountains. 



174 NORTH AMERICAN FAUNA. [No. 33. 

Vulpes macrourus Baird. Western Fox. 

This species, which is larger and darker than the eastern red fOx 
(V. fulvus), is common in the Colorado mountains. It is subject to 
a wide range of color variation, covering the red, cross, silver-gray, 
and very rarely the black phase. The proportion of dark individuals 
is much greater than with fulvus. This fox is found at the present 
time chiefly in the forest belt of the higher mountains, above 8,000 
feet. Its distribution in this region is general, and in some sections 
considerable numbers are present. I have no data respecting its 
range on the plains in early days, but, judging from its former occur- 
rence on the plains of Nebraska and Wyoming, it was probably present 
sparingly over eastern Colorado. In the canyon country in parts of 
southwestern Colorado this fox is fairly common as low as 6,000 feet. 
The species yields a fur of considerable value and is extensively 
trapped in winter. 

In 1905-6 it was reported as tolerably common in the Rabbit Ear, 
Williams, and Williams River Mountains, on the Front, Park, and 
Medicine Bow Ranges, and on the White River Plateau. Small 
numbers were reported on the Gore Range and in the mountains 
around Mount Cullom and Zenobia Peak, in western Routt County. 
The red or tawny phase appears to predominate in the Elk Head 
Mountains, since in a series of 17 skins handled by Mr. Robert 
Mcintosh, of Slater, in 1906, 15 were red and only 2 were of the cross 
phase. Mr. L. Wallace, of Granby, Middle Park, states that the ma- 
jority of skins taken in the Grand Lake region are cross foxes. Ac- 
cording to Warren, four out of six foxes killed near Crested Butte, 
Gunnison County, were of the cross phase. 1 Foxes were reported in 
small numbers in the mountains of Clear Creek County in 1905. At 
the Stevens Mill, on Mount McClellan, I learned of a young fox 
which had been captured in the spring by Italian miners on Grays 
Peak, at about 12,000 feet, and kept in confinement for a time. 
Skins which Loringsaw at Grand Junction in 1893 probably came from 
the Grand Mesa region, and during the same year he reported foxes 
from Estes Park. 

In the southern and southwestern mountains this fox appears to 
have fully as wide a distribution as father north. In the San Juan 
Mountains north of Pagosa Springs and Vallecito the cross phase is 
most common, according to the best-informed hunters and trappers. 
There are a few silver-grays, and very rarely a black fox is seen. 
Cross foxes are said to predominate also in the Saguache Mountains 
near St. Elmo, and a black fox has been seen near Tin Cup. Foxes 
are reported in small numbers from the Cochetopa Hills, La Plata 
Mountains, near Lone Cone in the San Miguel Mountains, and the 
Uncompahgre Plateau. Mr. Case, a trapper, is said to have taken 

i Mammals of Colorado, p. 259, 1906. 



1911.] MAMMALS. 175 

19 foxes along the Dolores River Canyon between the mouths of 
Disappointment and Paradox Creeks during the winter of 1906-7. 
Of these, several were cross foxes, while 2 were graded by the furrier 
as silver-cross and brought $50 apiece. The western canyon country 
seems to be especially frequented by foxes, although the altitude is 
low. Both silver-gray and red foxes are reported sparingly from the 
mountains on each side of the Wet Mountain Valley, and a black fox 
is said to have been seen at the eastern base of the Sangre de Cristo 
Range, west of Westcliffe, many years ago. Mr. E. W. Scott, of La 
Veta, has seen a fox skm from Veta Pass, and several silver foxes are 
stated to have been trapped on Sierra Blanca during recent years. 

At Ward, Boulder County, in June, 1893, Loring found a fox den 
among the rocks on a mountain side, and collected the female, whose 
fur was badly worn, and four half-grown young. Scattered around this 
den was a varied assortment of bones (including a fish bone), chicken 
wings, bird feathers, and a pair of old buckskin gloves. Each time 
the den was approached, the old fox barked and ran away a few rods 
in an attempt to lure the intruder from its vicinity. 

Allen states that in a series of 40 winter skins which he examined at 
Montgomery, Park County, in 1871, nearly half were cross and one 
individual was black. 1 The only published color description of the 
black phase of V. macrourus relating to a Colorado specimen appears 
to be that given by Coues and Yarrow of a melanistic skin from Los 
Pinos (now Bayfield), La Plata County, as follows: 2 

"A specimen, which we are inclined to refer to this species on 
account of its great size and especially large tail, is jet black all over, 
with a pure white tip to the tail; one of the finest examples of complete 
melanism we have seen. The purity of the black is only interrupted 
by a slight gray grizzle on the face and rump." 

Warren has published the description of a dark silver skin from 
Cumbres Pass, Conejos County, in the Colorado Museum of Natural 
History. 3 Coues says there were many specimens of this fox in Mrs. 
M. A. Maxwell's mounted collection of Colorado mammals which was 
on exhibition in Washington during the winter of 1876-77. 4 

Vulpes velox (Say). Kit Fox; Swift. 

[Canis] velox Say, Long's Exped. to Rocky Mts., I, p. 487, 1823. Type from South 
Platte River (in Logan County?), Colorado. 

The small swift or kit fox was formerly common over the plains of 
eastern Colorado, but has become rare in most sections. In 1892 
Dr. A. K. Fisher says it was reported tolerably common in the 
vicinity of Sterling (near type locality), and Mr. Edward A. Preble 
states that it was considered very rare in the region about Love- 

i Bull. Essex Inst., VI, p. 54, 1S74. 

2 Expl. W. of 100th Mer., V, p. 55, 1875. 

3 The Mammals of Colorado, p. 238, 1910. 

* Dartt, On the Plains and Among the Peaks, p. 219, 1879. 



176 NORTH AMERICAN FAUNA. [No. 33. 

land in 1895. On the plains of Boulder County very few, if any, 
remain at the present time. The only record for this region in 
recent years appears to be that of two which were killed on the 
farm of Mr. Samuel Hays, 3 miles northeast of Boulder, in 1903. 
Mr. W. II. Graham, of Spicer, says he has met with this fox but once 
during a long residence in North Park. In 1893 he shot two near 
their den, which was on an open sandy slope near Arapahoe Creek, 
in the southern part of the park. Prof. Lantz found a dead swift 
on the prairie near Cheyenne Wells in 1903, and was informed that 
they were not uncommon in that region. In 1 907 swifts were reported 
as common on the plains of Baca County and also in southern Bent 
and Prowers Counties. In all probability they are now more com- 
mon in southeastern Colorado than elsewhere in the State. A few 
were reported in 1909 on the eastern end of the Arkansas Divide and 
near Tuttle on the South Fork of the Republican River. There is 
said to be a colony of them near Keota, southwest of Pawnee Buttes, 
in northeastern Weld County. 

Urocyon cinereoargenteus scotti Mearns. Gray Fox. 

The range of the gray fox in Colorado is in the juniper and pinyon 
foothills on both sides of the mountains, chiefly in the Upper Sonoran 
zone. On the east side of the mountains it is known from Loveland 
and the Estes Park region, the foothills from the Arkansas Valley 
southward, and the rough country of Las Animas, southern Bent, 
and western Baca Counties. (See fig. 26.) It is more common in 
the rough pinyon country of western and especially of southwestern 
Colorado, where it ranges north sparingly to the Escalante Hills in 
western Routt County. Specimens examined from both sides of 
the mountains are referable to this form. 

Most of my notes are from west of the mountains. Mr. A. G. 
Wallahan, of Lay, reports seeing two skins of gray foxes in Lily 
Park in March, 1905, and Mr. F. C. Barnes, of Lily, killed one near 
there the following winter. Members of a railroad surveying party 
working in the Yampa Canyon of Bear River during the winter of 
1904-5 are reported to have killed several gray foxes. In the Esca- 
lante Hills, north of Bear River, as well as in the southwestern 
counties, the animal is known as the pinyon fox. Ranchmen report 
an occasional gray fox at Rangely, in the White River Valley, and 
also at Mack, in the lower Grand River Valley. A few are said to 
have been killed near Rifle during the past few years, and at Grand 
Junction in 1893 Loring saw the skin of one which had been killed 
in Mesa County. 

The gray fox reaches its greatest abundance in the region from 
Montezuma County north to the San Miguel River. Mr. Steve 
Elkins states that in the Mancos region they are usually found in the 
pinyon belt, but occasionally also among the yellow pines at the 



1911.] 



MAMMALS. 



177 



west base of La Plata Mountains. He often hunts foxes with dogs 
and kills a good many in this manner. A ranchman living on McElmo 
Creek, in western Montezuma County, is said to have caught 15 gray 
foxes in one winter. I found a den near McElmo in June, 1907, which 
had apparently just been vacated by a family of foxes. It was in 
loose shaly earth near the summit of a bare hill in plain view of a 
house and near a traveled road. There were a number of entrances 
to this den several rods apart, and the ranchman living near by stated 
that he often saw the old fox and three young ones frisking about 
during the day. This fox is reported as common in the pinyon 
country bordering the lower San Miguel and Dolores Rivers, and Mr. 




Fig. 26. — Distribution in Colorado of gray fox ( Urocyon c. scotti). 

C. H. Smith, of Coventry, traps a number each winter along the 
Naturita Canyon just back of his ranch. In the Uncompahgre Valley 
near Montrose these foxes are termed swifts. A mounted specimen 
which I saw at Montrose was killed within 3 miles of that point and 
a number of skins and rugs seen in stores there indicate that gray 
foxes are common in that section. 

Mr. J. W. Frey, of Salida, has a mounted specimen taken in that 
vicinity and says a number have been killed in the pinyon hills 
bordering the upper Arkansas Valley during recent years. In 
Huerfano County gray foxes are found in the pinyon country at both 
Gardner and La Veta, and Mr. E. W. Scott, of La Veta, has a mounted 
specimen from the upper Cucharas Valley. A mounted specimen 
90432°— 11 12 



178 NORTH AMERICAN FAUNA. [No. 33. 

seen at Buffalo, Jefferson County, was probably killed in the neigh- 
boring hills. In 1904 Bailey reported a few gray foxes among the 
lava buttes east of Antonito, in the southern end of the San Luis 
Valley, but I could learn of none in the northern and central part of 
the valley in 1907. Loring saw two gray foxes among the rocks in 
Estes Park in 1893 and shot one of them. 

Mephitis hudsonica Richardson. Northern Plains Skunk. 

This species, the largest of the Colorado skunks, comes into the 
State from the north and inhabits chiefly the higher mountainous 
sections. It is not known to occur south of Colorado Springs and 
Salida. Unfortunately the specimens examined represent very few 
localities, and therefore the distribution can not be accurately given. 
At Arkins, Larimer County, judging from a large series of skulls, M. 
hudsonica and the long-tailed skunk of the plains, M. m. varians, 
occur in about equal abundance, apparently without intergradation. 1 
The ranges of the two species may overlap over a considerable area 
along the lower eastern foothills, since Warren has taken both at 
Colorado Springs. A series ol 19 skulls from Arkins and 1 skull from 
Spicer, North Park, in the Biological Survey collection, are referable 
to hudsonica. Specimens in the Warren collection from Colorado 
Springs and Salida are doubtfully referred to this species, although 
they are by no means typical. 

All the records of skunks secured in the mountains of northern and 
central Colorado are tentatively referred to hudsonica. In 1905 and 
1906 I noted tracks near Coulter, on Grand River above Kremmling, 
and along the upper Snake River at Honnold. A den was found among 
the rocks on the north slope of the Elk Head Mountains, southeast 
of Slater, and the remains of a skunk were seen at Hayden, on the 
upper Bear River. Skunks were reported in Middle, North, and 
Egeria Parks, and in the vicinity of Meeker and Glenwood Springs. 
In 1893 Loring found skunks common in Estes Park, where they 
were said to kill much poultry. 

Allen found skunks common in Park County in 1871, and says they 
range above timberline in that section. 2 

Mephitis mesomelas varians Gray. Long-tailed Skunk. 

The long-tailed skunk is found on the eastern plains, and also enters 
the State from the south in the Rio Grande Valley. It occurs at a 
few points along the edge of the foothills with M. hudsonica, as 
already stated, but over the lower eastern part of the State appears 
to be the only large skunk present. The skunks reported at Brad- 
ford, Gardner, and La Veta, in the foothills of Huerfano County, and 
also along the pinvon-clad foothills bordering the San Luis Valley, 
are doubtless M. varians. This species can be readily distinguished 

i See Howell, N. Am. Fauna No. 20, p. 25, 1901. - Bull. Essex Inst., VI, p. 54, 1874. 



1911.] MAMMALS. 179 

from M. Jiudsonica by its smaller size, relatively longer tail, and by 
the usual absence of the pencil of white hairs at the tip of the tail. 

In addition to a large series of odd skulls and several skins from 
Arkins, Larimer County, the Biological Survey has specimens of 
varians taken at Loveland, Sterling, Canon City, Chivington, and 
Antonito. The Merriam collection contains a female from Boulder 
County. Howell records a specimen from Conrow (Chaffee County) , 
and two from Costilla County. 1 A skull in the National Museum 
from Cripple Creek may have been taken in the low Oil Creek Valley 
west of that point, as varians is not known to occur at the higher 
elevations or much above the upper edge of the Upper Sonoran zone. 
Specimens from Colorado Springs and from Gaume's ranch and 
Springfield, Baca County, have been identified by the Biological 
Survey for Warren, and he has recently recorded the species from 
Wray, Yuma County. 2 

The food of skunks consists chiefly of insects and the smaller 
rodents. Near Higbee, Otero County, Prof. Lantz found them feeding 
extensively upon the larvae of tiger beetles (Cicindela), which they 
dug out of the sand on the banks of Purgatory River. 

Mephitis mesomelas estor Merriam. Arizona Skunk. 

The large skunks reported in the warm southwestern valleys, from 
Grand River southward, are probably M. m. estor, since three speci- 
mens from the pinyon country around Coventry are referable to this 
form. Although veiy generally reported, the large skunks appear 
to be less common in most sections of the southwest than the small 
spotted species (Spilogale). Tracks of skunks were seen along Pla- 
teau Creek, near Tunnel, Mesa County; in the mud along McElmo 
Creek, Montezuma County; and along an irrigation ditch at Nucla, 
western Montrose County. Loring saw the tracks of a skunk at 
Silverton. 

The distribution and relationships of the skunks of western Colo- 
rado are imperfectly known, and additional specimens from many 
localities in this region are greatly desired. It is not likely that estor 
occurs north of the Book Cliffs. The skunks of the Routt County 
sage plains are probably M. Jiudsonica, but there are no specimens 
at hand from that region. 

Spilogale interrupta (Rafinesque). Prairie Spotted Skunk. 

This handsome dark species of spotted skunk ranges a shore dis- 
tance into the central eastern part of the plains region of Colorado, 
the only Colorado specimen known being one taken at Wray, Yuma 
County, by Warren, and identified by the Biological Survey. While 
at Wray in December, 1907, I was informed that small numbers of 

1 N. Am. Fauna No. 20, p. 32, 1901. 2 Further Notes on Marnmals of Colorado, p. 83, 1908. 



180 NORTH AMERICAN FAUNA. [No. 33. 

spotted skunks frequented the sandstone bluffs along the south side 
of the Chief Creek Valley. 

Prof. Lantz states that spotted skunks are reported common in 
the vicinity of Hugo, Lincoln County. The specific identity of the 
Hugo animal is uncertain, since there are no specimens available 
from that section, and since S. tenuis, the species found along the 
eastern foothills, probably ranges a short distance out on the plains 
in the region of the Arkansas Divide. The prairie spotted skunk 
may be distinguished from the other forms occurring in the State, 
S. tenuis and S. g. saxatilis, by its larger size and darker coloration, 
the white spots and bars being at a minimum. The tail also is usu- 
ally wholly black except the white tip. 

Spilogale tenuis Howell. Rocky Mountain Spotted Skunk. 

Spilogale tenuis Howell, Proc. Biol. Soc. Wash., XV, p. 241, Dec. 16, 1902. Type 
from Arkitis, Larimer County, Colorado. 

Little is known regarding the range of this species, all the infor- 
mation at hand indicating a scattering distribution in both the Tran- 
sition and Upper Sonoran zones among the lower foothills at the 
eastern base of the mountains. Most of the records are from north 
of the Arkansas Divide, in Larimer, Boulder, Jefferson, and Douglas 
Counties. It has been collected in northeastern New Mexico, how- 
ever, and no doubt ranges across the entire width of Colorado along 
the eastern foothills. 

In June, 1905, I saw the tracks in a dry gulch a mile or two south- 
west of Golden, while at Arkins (the type locality) the animals were 
reported tolerably common in the valleys below 6,000 feet. Mr. 
Berry, a ranchman, is said to have killed one a mile southwest of 
Arkins in July, 1906. Mr. Vernon Bailey caught one among the 
rocks in the foothills 2 miles west of Boulder in October, 1903, but 
it escaped. Several specimens have been collected at Boulder by 
Mr. R. T. Young. 1 Warren records a specimen from Sedalia, Doug- 
las County, in the collection of Colorado College, 2 and has an imma- 
ture specimen from Colorado Springs in his own collection. A speci- 
men from Estes Park is in the American Museum of Natural History, 
and one from Loveland is in the National Museum. 

In 1907 I was informed by Mr. William King, of the Medano 
Springs ranch, near the San Luis Lakes, that he has seen a number 
of spotted skunks among the foothills on the San Luis Valley side of 
the Sangre de Cristo Kange, near Mosca Pass. The species may have 
crossed the mountains through this low pass, as it is reported present 
in the Muddy and Huerfano Valleys, on the east side of the range. 
In the rough juniper country of Las Animas and western Baca Coun- 
ties, in the extreme southeast, spotted skunks are reported much 
more common than Mephitis. 

1 Proc. Acad. Nat. Sci. Phila., p. 406, 1908. 2 Further Notes on Mammals of Colorado, p. 83, 1908. 



mil.] MAMMALS. 181 

Spilogale gracilis saxatilis Merriam. Great Basin Spotted Skunk. 

The small spotted skunks of the warm Upper Sonoran valleys west 
of the Continental Divide are referable to saxatilis, although speci- 
mens from Coventry, in western Montrose County, are not typical. 
Over most of their range Mephitis occurs with them but in smaller 
numbers. The spotted skunks are much more active and agile than 
the large ones and readily climb pinyons and junipers. Rock ledges 
along canyons are much frequented by them. 

None were collected in northwestern Colorado during 1905 and 1906, 
but they were reported as occurring in all the valleys entering the State 
from the west. According to Mr. John Criss, a wolf trapper with head- 
quarters at Baggs Crossing, Wyoming, the spotted skunk is tolerably 
common on the lower Snake River, where it is the only skunk present. 
It occasionally gets into his wolf traps in winter, and he has trapped 
it as far east as the old Edwards sheep camp, 35 miles below Baggs 
Crossing. Spotted skunks are reported common at Escalante, Routt 
County; and in Lily Park, at the confluence of the Snake and Bear 
Rivers, ranchmen often find them beneath houses and in cellars. At 
Rangely, in the valley of White River, they are said greatly to out- 
number the large striped skunks. They are reported also from the 
upper White River country, according to Felger. 1 The range of sax- 
atilis probably extends in the Grand Valley as far east as Glenwood 
Springs. I occasionally heard of it west of Rifle, Garfield County, 
and Mr. Fred Baker, a taxidermist of Glenwood Springs, reports that 
he has handled several skins taken during the past few years in the 
Grand Valley east of Newcastle. 

In 1907 spotted skunks were reported as common and greatly out- 
numbering Mephitis in the Mancos and McElmo Valleys (Montezuma 
County), in the region of the lower Dolores and San Miguel Rivers, 
and at Coventry. A few were reported also in the vicinity of Bay- 
field, La Plata County. At Ashbaugh's ranch, in the McElmo 
Canyon, a female got into a small trap which I had baited with a 
piece of potato and set among the rocks for wood rats. Mr. C. H. 
Smith has caught a number of these skunks in the pinyons back of 
his ranch at Coventry, one being taken in a tree trap set for gray 
foxes 3 feet up on the trunk of a large pinyon standing on the upper 
rim of Naturita Canyon. A male in the Biological Survey collection 
was taken by Howell among rocks 2 miles south of Grand Junction. 

Taxidea taxus (Schreber). Badger. 

The badger is one of the most widely distributed mammals in the 
State, occurring throughout the plains and deserts of the lower parts, 
and in the mountains ranging with more or less regularity nearly to 
timberline. Two forms may be present, there being a possibility of 

! Univ. of Colo. Studies, VII, No. 2, p. 145, 1910. 



182 NORTH AMERICAN FAUNA. [No. S3. 

T. berlandieri in extreme southern and southwestern Colorado, but 
this can not be determined from the scanty material now available. 

The numerous holes which badgers are continually digging on the 
plains in search of their rodent prey make horseback riding in certain 
sections somewhat hazardous. Aside from this, however, badgers 
are most beneficial mammals, since their food consists chiefly of 
prairie dogs, ground squirrels, and other noxious rodents, which they 
secure by digging down to the nests. On the sage plains and in the 
mountain parks of northwestern Colorado the badger feeds exten- 
sively upon the white- tailed prairie dog (Cynomys leucurus), the 
Wyoming ground squirrel (Citellus elegans), and pocket gophers 
(Thomomys). Badgers are voracious, and one animal will open up a 
large number of burrows in a night. 

In 1905 and 1906 badgers were abundant in Middle, North, and 
Egeria Parks and in the valleys of the Bear, Snake, Green, White,Grand, 
and Eagle Rivers. A specimen shot among the White River bluffs, 
20 miles east of Meeker, was running up the steep side of a canyon. 
One was found dead in the trail near the head of Grand Encampment 
River, on the Park Range, at an elevation of 10,000 feet. Warren 
mentions that he killed a specimen in Gunnison County at about 
11,500 feet. 1 In 1907 badgers were reported common in southern 
and southwestern Colorado, where they prey largely upon prairie 
dogs {Cynomys gunnisoni). Loring secured a specimen near Coch- 
etopa Pass, and Bailey reports them common in the Rio Grande 
Valley near Antonito. The burrows are common on the South Park 
plains east of Como, at 9,800 feet. Other members of the Biological 
Survey have found badgers common at Sterling, Loveland, Estes 
Park, Las Animas, Burlington, and Olney. 

Lutra canadensis (Schreber). Otter. 

Otters seem to have been always rare in the State, although the 
reason is not apparent. The country is well watered, and nearly all 
the lakes and streams are well stocked with fish and should offer a 
satisfactory habitat. My notes refer for the most part to western 
Colorado, and the otter of that region may be the southwestern form 
described by Rhoads as L. canadensis sonora. This can not be con- 
firmed at present, as there are no specimens available. 

During the winter of 1902-3 four otters were trapped on Snake 
River within a few miles of Slater, Routt County, by two trappers, 
Messrs. James Coates and James Parsons, living at Slater. In 1905 
Mr. A. G. Wallahan, of Lay, Routt County, reported a few still 
present in the Yampa Canyon on Bear River. 

Warren thinks the rarity of otters in mountain streams may be due 
to the "freezing of their sources of supply" in winter. 2 There are, 

i Mammals of Colorado, p. 201, 1906. 2 Ibid, p. 265, 1906. 



1911.] MAMMALS. 183 

however, many fine lakes throughout the mountains well stocked 
with fish and furnishing a most favorable environment for otters. 
The following records are given by Warren (1. c): "Grand Junction 
and Big Dolores River, a few (Dr. E. F. Eldredge); Platte River, 
east of Greeley, one specimen (A. E. Beardsley) ; Julesburg, occasional 
(H. G. Smith)." In a more recent publication Warren gives the 
following data: "Mr. Henry Lehman, of Grand Lake, tells me there 
are a few otter in the Grand River, in Grand County. Herman W. 
Nash writes me that friends of his saw two in the Gunnison River, near 
Sapinero, in August last. 1 Felger records this species from the White 
River Valley on the authority of Ball. 2 Trippe records the otter 
from Clear Creek County; 3 and Coues mentions a specimen from 
Boulder County, which he saw in the collection of Mrs. M. A. Maxwell. 4 
A skull in the National Museum is from Pueblo. 

Lutreola vison energumenos (Bangs). Mink. 

Minks occur in varying abundance on nearly all the larger streams 
of Colorado, and in the unsettled parts the skins bring considerable 
revenue to trappers. They are more abundant in the mountains 
than on the plains, and are especially common in Middle and North 
Parks, where their fondness for poultry makes chicken raising 
most unsatisfactory on many ranches near streams. The minks of 
the higher mountains yield a rich dark fur of good quality, but the 
plains animal is said to be considerably paler and consequently less 
valuable. As no specimens of minks from the eastern part of the 
plains region are at hand, it is not certain that they are of the same 
form as the mountain animal, but specimens from along the eastern 
base of the foothills are identical with those from the higher moun- 
tains. 

A female which I collected on Boulder Creek, 5 miles west of 
Boulder, in June, is very dark, the fur being full and heavy, even at 
that late date. Other dark minks were observed on Mclntyre 
Creek, Larimer County, and in the Elk Head Mountains, during 
August, and I saw two rich dark skins from the upper Los Pinos in 
a store at Bayfield. Mink tracks were observed on Snake River, 
at Honnold; on Bear River, at Hayden and Lily; on White River, 
at Meeker and Rangely; on Grand River, at Glenwood Springs, 
Kremmling, and Hot Sulphur; on Plateau Creek, near Tunnel; on 
Good Spring Creek, near Axial; and on Green River, near Ladore. 
Minks were reported abundant on the Cucharas River at La Veta 
and on the San Juan River at Pagosa Springs. Specimens from 
Loveland, Cochetopa Pass, Pagosa Springs, and Arkins (Larimer 

1 Further notes on Mammals of Colorado, p. 84, 1908. 

2 Univ. of Colo. Studies, VII, No. 2, p. 146, 1910. 
s See Coues, Birds of the Northwest, p. 224, 1874. 
4 Fur-bearing Animals, p. 312 (footnote), 1877. 



184 



NORTH AMERICAN FAUNA. 



[No. 83. 



County) are in the Biological Survey collection. Others in the 
Warren collection from Crested Butte and Colorado Springs have 
been examined by the Biological Survey. 

In 1871 Allen found minks common along the streams of Park 
County as high as 10,000 feet. 1 

Putorius nigripes And. and Bach. Black-footed Ferret. 

This rare and little known mammal has been recorded from a 
number of localities on the plains of eastern Colorado (see fig. 27), 
but here, as elsewhere over its range, its numbers are small. Usually 




PurOff/OS S7ff£ATOft/ J.SPTVS 



PCW/f/OS A//6ff/P£S. 



Fig. 27.— Distribution in Colorado of dwarf weasel (Putorius streatori leptus) and black-footed ferret 

(P. nigripes). 

it is found in prairie-dog towns, where it takes up its abode in an 
abandoned burrow, and from this convenient base preys upon the 
defenseless inhabitants of the colony. These ferrets are most 
beneficial mammals, because their prey consists largely of prairie dogs. 
In December, 1894, black-footed ferrets were reported to Streator 
as present, but rare, at three points in the Arkansas Valley: Olney, 
Otero County; Arlington; and Chivington, Kiowa County. Prof. 
D. E. Lantz heard of a very few in the region about Hugo in 1905, 
where they are known as prairie-dog ferrets. In 1907 I heard of 
this species in the dog towns of Baca and southern Prowers Counties, 

1 Bull. Essex Inst., VI, p. 54, 1874. 



1911.] MAMMALS. 185 

and it seems to be more generally known in that region than in other 
sections on the plains. I did not hear of ferrets in the prairie-dog 
country of northwestern Colorado, but Warren records two mounted 
specimens at Meeker, said to have been taken near there, and men- 
tions a specimen from the Laramie River, 12 miles south of the 
Wyoming boundary, in the University Museum at Boulder. 1 

Two specimens in the Warren collection indicate a most remarkable 
vertical range for this plains mammal. Warren says: "One speci- 
men in my collection came from Divide, Teller County, at an ele- 
vation of 9,800 feet, another was found dead in Lake Moraine, El 
Paso County, altitude 10,250 feet. It is a mystery how the animal 
came there, and when skinned there were no marks on its body to 
indicate the cause of death. In the spring of 1004 C. E. Aiken 
mounted one which came from near Clyde Station, El Paso County, 
altitude 9,440 feet." 2 

Others have been recorded by Coues, 3 as follows: A specimen 
taken in the Cache la Poudre Valley by Dr. Law and brought to 
Washington by Dr. Hayden; another reported by Dr. Hayden as 
having been kept in confinement at Greeley for a considerable period; 
and two or three specimens procured by Mrs. M. A Maxwell in the 
vicinity of Denver and forming a part of her exhibit of Colorado 
mammals at the Philadelphia Exposition. All the above ferrets 
were taken in prairie-dog towns, and one of Mrs. Maxwell's speci- 
mens had been drowned out of a prairie-dog hole and captured alive. 
This individual was kept in confinement for some time. "It became 
quite tame, readily submitting to be handled, though it was furious 
when first caught. It was kept in a wire cage and fed on beef. 
When irritated it hissed and spat like an angry cat. It used to hide 
by covering itself over with the material of which its nest was com- 
posed, but at times, especially at night, it was very active and 
restless." 4 

Putorius longicauda (Bonaparte). Long-tailed Weasel. 

The large long-tailed weasel of the plains should be present over 
practically all of the State east of the mountains, but until thorough 
collecting is done there, its abundance and distribution can not be 
indicated. It is not known from the mountains of Colorado. Weasels 
from the lower edge of the foothills are referable to the slightly 
smaller P. arizonensis, so it is quite probable that P. longicauda is 
restricted to the plains well out from the mountains. It probably 
occurs over a considerable area in western Routt County, since a male 
in the Warren collection from the sage plains at Lay, examined by 
the Biological Survey, is referable to this species. Another Colorado 
specimen of P. longicauda examined by the Biological Survey was 

i The Mammals of Colorado, p. 194, 1910. 3 Fur-bearing Animals, pp. 150-151, 1877. 

s Ibid., p. 2G4, 1906. < Dartt, On the Plains and Among the Peaks, p. 220, 1879. 



186 NORTH AMERICAN FAUNA. [No. 33. 

taken at Wray, Yuma County, and is in the collection of the Colorado 
Historical and Natural History Society in Denver. Weasels reported 
present in Baca County are undoubtedly of the long-tailed species. 
Warren, on the authority of A. E. Beardsley, 1 records P. longicauda 
from Platte River and Greeley. 

Putorius arizonensis Mearns. Mountain Weasel. 

This weasel is tolerably common in the mountainous parts of the 
State, and replaces P. longicauda of the plains from the eastern base 
of the foothills westward. It has a wide vertical range and occurs on 
both slopes of the Continental Divide from 5,000 feet to timberline. 
The only other weasel found in the Colorado mountains is the much 
smaller P. s. leptus. 

In 1005 and 1906 four specimens were collected at Steamboat Springs; 
Meeker; Higho, North Park; and Tunnel, Mesa County. Another 
weasel, which was not secured, ran across the trail in front of us at 
Boulder Falls, on Middle Boulder Creek, and took refuge in a pile of 
rocks. Mr. Frank Hayes, a taxidermist of Glenwood Springs, showed 
me several skins from the head of Noname Creek, Garfield County. 
Large weasels which can without much question be referred to this 
species were reported in Middle and North Parks; Escalante Hills; 
valleys of the Green, White, and Grand Rivers; Medano Springs 
ranch; Westcliffe; Bradford; Gardner; La Veta; Pagosa Springs; 
Rogers Mesa, near Hotchkiss; Placerville; and Lone Cone, San Miguel 
Mountains. The lumbermen at Fraser, in eastern Middle Park, 
report that a great many weasels live among the log piles, where 
they prey upon chipmunks. Mr. T. J. McKenna, of the Stevens Mill, 
at timberline on Mount McClellan, states that large weasels are 
common and tame about the mill and often come into the cabins 
during the heavy snows of winter. Mr. Edward A. Preble saw one 
at timberline on Longs Peak in 1895. There are two white winter 
specimens from Coventry. Warren has taken P. arizonensis at 
Colorado Springs, and states that one was killed at Crested Butte, 2 
while another specimen which he sent to the Biological Survey 
for identification came from Sapinero. A winter specimen from 
Semper, Jefferson County, has been identified for Mr. W. D. Hol- 
lister, of Denver, and another from the San Luis Valley (between 
Monte Vista and Del Norte, November 24, 1903) for the Colorado 
Historical and Natural History Society. The last specimen is a male 
in nearly full winter pelage, and measures: Total length, 425; tail 
vertebrae, 155; hind foot, 46.5. 

This weasel frequents the piles of large bowlders and debris in 
canyon bottoms and along mountain streams, where it preys chiefly 
upon mice, chipmunks, and Say spermophiles. When surprised in 

i Mammals of Colorado, p. 264, 1906. 2 Ibid., p. 265, 1906. 



1011.] MAMMALS. 187 

the open, it immediately seeks refuge among the nearest rocks, but 
once in this safe retreat its curiosity overcomes its fear, and it is 
seldom out of sight for more than a moment. It frisks in and out 
among the rocks, stopping now and then to crane its long neck at the 
observer, and even stands erect on its hind legs to get a better view 
of the object of its curiosity. Occasionally it is found at a distance 
from rocks. At Steamboat Springs I shot one as it ran past my 
camp among the alders on the bank of Bear River, and one of two 
specimens from near Cochetopa Pass was shot by Loring on a fence 
rail in a cultivated field. One seen by Prof. Lantz near Edlowe, 
Teller County, took refuge in a prairie-dog burrow. A weasel shot 
among the White River bluffs east of Meeker was carrying a large 
Callospermo'philus in its mouth. 

This is doubtless the weasel noted by Allen in Park County in 1871 
and recorded as P. ermineus / and also the species recorded by Coues 
as P. longicauda from the "mountains of Colorado." 2 A summer 
skin from Fort Garland, recorded by Coues and Yarrow as P. longi- 
cauda, 3 is not available for examination, but may be referred to 
arizonensis on geographic grounds. 

Putorius streatori leptus Merriam. Dwarf Weasel. 

Putorius streatori leptus Merriam, Proc. Biol. Soc. Wash., XVI, p. 76, 1903. Type 
from Silverton, San Juan County, Colorado. 

The meager data bearing on this beautiful little weasel indicate 
a general distribution in the higher mountains over the State. (See 
fig. 27.) Very little seems to be known of its habits. Judging 
from verbal descriptions this is the small weasel which occasionally 
proves such a nuisance to trappers in the heavy forests on the higher 
mountain ranges by getting into marten traps in winter. 

I saw a mounted specimen in winter pelage at Steamboat Springs 
in 1905, but beyond the fact that it was killed in the neighboring 
mountains the proprietor could tell nothing of its history. At Glen- 
wood Springs, in August, 1907, Mr. William Cross, a taxidermist, 
told me that he had recently seen one of these diminutive weasels in 
brown summer pelage peering from beneath the sidewalk, but failed to 
secure it. Mr. Cross showed me a winter skin which he had obtained 
from a trapper in the Glen wood region. Mr. Anton Stark, agent 
of the Colorado & Southern Railroad at St. Elmo, in the Saguache 
Mountains, reports the dwarf weasel as not uncommon thereabouts. 
One which he captured in winter and kept in confinement became 
so tame that it was finally allowed its freedom, but remained about 
the house for some time and proved an expert mouser. 

Mr. D. Costello, of Gardner, Huerfano County, relates an incident 
which occurred many years ago while he was prospecting in the 

i Bull. Essex Inst., VI, p. 54, 1874. a Explorations W. of 100th Meridian, V, p. 5% 1875. 

2 Fur-bearing Animals, p. 141, 1877. 



188 NORTH AMERICAN FAUNA. [No. 33. 

mountains of northern Gunnison County, back of Crested Butte. 
Soon after locating in a cabin adjacent to a large rock slide just 
below timberline he discovered that a cony was occupying a large 
grass nest beneath the cabin floor. It often appeared in the cabin, 
coming up through a broken board in the floor, and in time became 
very friendly. Finally a day came when the cony did not make 
its usual appearance, but a tiny weasel w T as seen at the hole in the 
broken board, peering in all directions and craning its long slim 
little neck with the bold curiosity so characteristic of the larger 
weasels. Fearing for the welfare of the cony, Mr. Costello killed 
the tiny cutthroat, but apparently too late, as he saw no more of 
his interesting companion. It seems probable that the cony is 
often preyed upon by this weasel, as the same rock slide often harbor:; 
both animals. P. leptus is only about 6 inches in length, exclusive 
of the short black-tipped tail of 1 or 2 inches, and the body is little 
larger around than that of a small mouse. 

Mr. Walter Blanchard has presented the Biological Survey with 
a brown summer specimen of P. leptus, which was killed on his 
ranch 5 miles west of Boulder in June, 1902. The type specimen 
from Silverton, in beautiful white winter pelage, was shot by Loring 
October 20,- 1893, as it was peering from under a log. Another 
immature specimen from the same locality was caught in a trap 
set in the underground tunnel of a pocket gopher (Thomomys fossor) , 
which suggests that this gopher may form part of the weasel's bill 
of fare. Another specimen in the Biological Survey collection 
was obtained at Crested Butte February 17, 1902, by Warren, 
wno gives the following observations on this species: "About Crested 
Butte, judging from the tracks one sees after a fresh fall of snow, 
it is quite common. It often burrows under the surface of the 
light snow, and runs beneath for quite a distance, then reappears 
on top, having been hunting down a mouse." 1 Recently P. leptus 
has been recorded from Coventry, Montrose County. 2 A male 
winter specimen from Larimer County is in the collection of the 
State Agricultural College at Fort Collins. One from Boulder County, 
in the Field Museum of Natural History, is recorded by Elliot. 3 A 
specimen from near Boulder in the mounted collection of Colorado 
mammals, exhibited by Mrs. M. A. Maxwell in Washington in 1876-77, 
has been recorded by Coues as "the least weasel, Putorius vulgaris,"* 
and was very likely the present form. 

i Mammals of Colorado, p. 264, 1906. 

* Warren, The Mammals of Colorado, p. 198, 1910. 

3 Field Col. Mus. Pub., 115, VIII, p. 449, 1907. 

4 Dartt, On the Plains and Among the Peaks, p. 220, 1879. 



1011.] 



MAMMALS. 



189 



Mustela caurina origenes Rhoads". Rocky Mountain Marten. 

Mustela caurina origenes Rhoads, Proc. Acad. Nat. Sci. Phila., p. 458, 1902. 
Type from Marvine Mountain, Garfield County, Colorado. 

In the dense forests of lodgepole pine and spruce which clothe 
the upper slopes of the higher mountain ranges of northern Colorado 
the marten is still present in considerable numbers. It appears 
to be uncommon on all the southern ranges except the San Juan 
Mountains, where from a point northeast of Pagosa Springs west 
to Silverton and Telluride it is reported in good numbers. Martens 
are rarely observed below 8,000 or 8,500 feet, or the lower edge of the 
Canadian zone forest belt. They range regularly to timberline, how- 
ever, and have been seen 1,500 feet above timberline near Silverton. 




Fig. 28.— Distribution in Colorado of marten ( Mustela caurina origenes). 

Throughout their range (see fig. 28) martens are hunted and 
trapped extensively, and consequently are not nearly so abundant 
as formerly. Ski are often used in hunting them in winter, when 
snow covers the mountains to a depth of several feet, and when 
pursued in this manner the animals quickly take refuge in trees, 
where they are easily shot. Hunting martens on ski is said to be 
very exciting sport, and at times hazardous, owing to the roughness 
of the country. This is a favorite method of hunting in Middle 
Park and in the San Juan Mountains. Most of the martens secured, 
however, are taken in either steel traps or deadfalls. Although 
Colorado martens are somewhat paler than those farther north, 
they nevertheless yield a valuable fur. 



190 NORTH AMERICAN FAUNA. [No. 33. 

Most of my information on the present distribution and abundance 
of martens relates to the northern half of the State. During the 
period from 1900 to 1905 the following were considered the best 
regions for martens: The mountain ranges surrounding Middle and 
North Parks ; the Williams River Mountains and the eastern part of 
the White River Plateau, west of Egeria Park; and the mountains 
south of Aspen, Pitkin County. Mr. Fred N. Selak, of Coulter, 
Middle Park, informed me in 1905 that he handled from 25 to 30 
marten skins each year, taken chiefly in the Rabbit Ear and Williams 
Mountains and along the western slope of the Front Range near 
Arapahoe Peak. Alpert & Co., of Kremmling, receive a few skins 
each year from the Gore Mountains. Mr. W. H. Graham, of Spicer, 
North Park, states that he usually traps between 10 and 20 martens 
each winter in the Rabbit Ear Mountains, near the head of Arapahoe 
Creek. The animals are reported common on the Park Range but 
very rare on the Medicine Bow Range. Many trappers market the 
fur in Denver. A conservative estimate of the annual catch in the 
Middle and North Park region would be 100 skins. 

In August, 1905, Mr. Frank Hayes, a taxidermist of Glen wood 
Springs, showed me skins of four martens trapped the previous 
winter at Bennett's Well, on the head of Noname Creek, 8 miles north- 
east of that point. He had also purchased a number of skins taken 
on Divide Creek, south of Newcastle, and at Mud Springs, on the 
White River Plateau. Mr. Hayes was located at Aspen about 1900, 
and handled between 40 and 50 marten skins each year, nearly all of 
which came from near the Montezuma mine, on the east slope of Hay- 
den Peak. Two other fur buyers were located in Aspen at the same 
time, and Mr. Hayes estimates that fully 100 skins were marketed there 
annually. This region is probably at present the best marten country 
in Colorado. Mr. Robert Mcintosh, of Slater, received 10 marten 
skins from the Elk Head Mountains during the winter of 1905-6. 
Martens are reported rare on the high timbered divide east of Laramie 
River and on the headwaters of Grand Encampment River. 

Mr. Steve Elkins, of Mancos, reports a few in the spruce belt of the La 
Plata Mountains, and I am informed they are tolerably common in the 
extensive forests of Engelmann spruce on the mountains west of Rico. 
They seem to have been always rare on the Sangre de Cristo Range. 
Mr. E. W. Scott, of La Yeta, has the skin of a marten killed in the 
spruce belt on East Spanish Peak in the winter of 1904-5, and states 
that he has seen three other marten skins from the same mountain. 
Mr. T. J. McKenna, of Denver, states that 30 years ago he trapped a 
great many martens near the Tin Cup mine, and on the headwaters 
of the Cimarron River, in Gunnison County. Skins which Loring 
saw at Grand Junction in 1893 probably came from the Grand Mesa 
country. 



1911.] MAMMALS. 191 

The type and one topotype of M. origenes in the National Museum 
were collected in Garfield County, near Mount Marvine, in Septem- 
ber, 1901, by Mr. E. Thompson Seton. The skulls may be readily dis- 
tinguished from those of M. americana by the small, flattened, rec- 
tangular bullse and the peculiar saddle-shaped upper posterior molar. 

(?)■ Mustela pennanti Erxleben. Fisher. 

In his list of the mammals of Park County, Allen records the fisher 
as " more or less common." 1 I have made careful inquiry of old 
hunters and trappers throughout the heavily forested region of the 
northern Colorado mountains and have yet to meet one who is 
familiar with the fisher or who has even heard of the animal within 
the State. Warren does not record it and it seems probable that 
Allen's record was based on erroneous information. 

Gulo luscus (Linnaeus). Wolverene. 

In the high Canadian zone forests of the wilder parts of the moun- 
tains the wolverene is still occasionally seen. While never common 
in the State, it was formerly of general occurrence. At present it 
appears to be restricted largely to the San Juan and La Plata Moun- 
tains, the mountains of northern Gunnison County, and the ranges 
surrounding North and Middle Parks. 

Mr. W. H. Graham, of Spicer, North Park, states that wolverenes 
are occasionally reported from the Rabbit Ear Mountains, but that 
very few have been killed during recent years. He saw a track on 
the head of Arapahoe Creek during the winter of 1904-5, and in the 
fall of 1903 saw a wolverene which a trapper caught on Owl Mountain, 
in southeastern North Park. The wolverene was considered very 
rare in Middle Park in 1905. Mr. Fred Selak, of Coulter, reported 
that one was killed in the fall of 1903 near the head of Ranch Creek, 
on the western slope of the Front Range. Two skins, which were 
purchased by Alpert & Co., of Kremmling, in 1903, were said to 
have been trapped the previous winter in the heavy forests on the 
head of the Williams Fork of Grand River. Mr. T. J. McKenna, of 
Denver, states that while at the Tin Cup mine, in Union Park, Gun- 
nison County, in the summer of 1883, he saw a wolverene which 
had been killed in that vicinity. He says it was not at all rare in 
the mountains of Gunnison County 25 or 30 years ago. 

Mr. Wood Galloway, of Norwood, states that for many years his 
father had a wolverene skin, taken in Antelope Park, Mineral County, 
and says the animals were not at all uncommon in that section 30 
years ago. A wolverene shot at the east base of Pagosa Peak in the 
winter of 1905 was mounted and on exhibition in Pagosa Springs 
until 1907, when it was sent to Denver with a miscellaneous mounted 
collection. Mr. H. N. Wheeler, formerly supervisor of the Monte r 

« Bull. Essex Inst., VI, p. 54, 1874. 



192 NORTH AMERICAN FAUNA. [No. 33. 

zunia National Forest, has a wolverene skin said to have been taken 
in the spruce belt near Calico Mountain, west of Rico, in 1905, and 
through other sources I learn that these animals are of occasional 
occurrence in the La Plata Mountains. One is said to have been killed 
near the Silver Picket mine, on Mount Wilson, in the vSan Miguel 
Mountains, about 1895, and according to Mr. D. Costello, of Gardner, 
one was killed on the head of Huerfano River many years ago. 

Allen says the wolverene was not uncommon in Park County in 
1871, and he saw the skin of one taken near Montgomery; 1 while it 
formerly occurred in Clear Creek County, according to Trippe. 2 
Coues mentions a specimen in the collection of Colorado mammals 
exhibited by Mrs. M. A. Maxwell at the Philadelphia Exposition. 
It was caught in a steel trap in the mountains near Boulder. 3 In 
an interesting general account of the wolverene Mr. C. A. Cooper 
describes his capture of an old male in the heavy forest on the summit 
of Gore Pass, between Middle and Egeria Parks, during the winter 
of 1883, and also mentions four beautifully marked wolverene skins 
which he examined at Trappers Lake, Garfield County, in the winter 
of 1889, 4 while a companion of Cooper's reported seeing a wolverene 
in Rock Creek Canyon, Egeria Park, in 1888. 5 Warren records a 
wolverene taken near Irwin, Gunnison County, about 1890, and 
states that he saw fresh tracks in the snow near Irwin in October, 
1905, at an elevation of 11,000 feet. 6 A specimen in the Carter 
collection in the Colorado Museum of Natural History was taken 
neat' Breckenridge, Summit County. 

Bassariscus astutus flavus Rhoads. Civet Cat; Cacomistle. 

The data at hand bearing on this handsome and interesting animal 
are rather meager and unsatisfactory, and probably do not well 
indicate its distribution and abundance in Colorado. The species 
appears to be restricted to the Upper Sonoran zone, and with a single 
exception all the records are from the rough canyon country in the 
lower southwestern part of the State from the Grand River Valley 
southward. 

The civet cat is tolerably common on Mesa Verde and thence north 
to the southern base of the Uncompahgre Plateau. In June, 1907, 
I saw its small, round, cat-like tracks in the dust beneath many of 
the overhanging rock ledges along Navajo Canyon, 25 miles south- 
west of Mancos; while among the Spruce Tree Cliff Ruins at the 
head of this canyon, and particularly in the darkest recesses of the 
cavern behind the ruins, the footprints of civet cats, leading here 

i Bull. Essex Inst., VI, p. 54, 1874. 

2 See Coues, Birds of the Northwest, p. 224, 1874. 

3 Dartt, On the Plains and Among the Peaks, p. 219, 1879. 
* Big Game of North America, pp. 479-501, 1890. 

6 Ibid., p. 492, 1890. 

e Mammals of Colorado, p. 262, 1906. 



1911. J MAMMALS. 193 

and there among the numerous smaller tracks of cliff mice and wood 
rats, were plainly discerned in the thick layer of fine rock dust. 
However, none came to meat-baited traps which I scattered along 
the ledges. Mr. Steve Elkins, of Mancos, states that he has seen a 
civet cat which was killed in Mancos Cairyon. This species is not 
reported in McElmo Canyon, and seems to be little known to the 
hunters of Montezuma County, but this is not strange, since it is 
nocturnal, and in hunting for cliff mice and wood rats rarely leaves 
the caves and deeper recesses in the rocky walls of canyons. A 
mounted specimen which I saw in Durango probably came from 
either Montezuma County or southern La Plata County. 

The civet cat is more generally known in the region bordering the 
lower San Miguel and Dolores Rivers. Mr. Henry Huff, an Indian 
living at Norwood, who trapped in the Dry Creek Basin, in western 
San Miguel County, during the winter of 1906-7, states that the 
" ring-tails" are not at all uncommon among the ledges of that region. 
A fine adult civet cat caught by him and an immature individual 
taken the same winter in the canyon of San Miguel River, a few miles 
north of Coventry, are in the Warren collection. Civet cats are 
reported from the Tabeguache Canyon, north of the San Miguel, and 
Mr. J. P. Galloway, of Norwood, informed me of five which were killed 
some years ago at the Sunrise copper mine in West Paradox Valley. 

There are three Mesa County records of civet cats. In June, 1893, 
Loring saw a skin in a fur buyer's store at Grand Junction winch was 
said to have been taken the previous fall about 4 miles from that place. 
A specimen which the National Museum obtained from Conductor 
Tuttle of the Colorado Midland Railway, was caught by a trapper in 
Mesa County a few years ago, and was mounted by Mr. C. E. Aiken, 
of Colorado Springs. The history of tins specimen is given by Dr. 
W. W. Arnold in Outdoor Life. 1 Warren records Bassariscus from 
Delta on the authority of A. E. Beardsley. 2 An animal answering 
its description was reported to Mr. C. E. Aiken from Beaver Creek, 
Fremont County, some 25 miles south of Colorado Springs, in 1904. 

Procyon lotor (Linnaeus). Raccoon. 

The data on the distribution and abundance of the raccoon within 
the State are somewhat limited and probably do not indicate the 
extent of its range. It is tolerably common on the eastern plains, 
along some of the larger foothill streams, and in the extreme south 
from the San Luis Valley west to La Plata County. According to 
Warren, the animal has been taken in Grand County, west of the Front 
Range. 3 It has not been found in the lower northwestern part of the 
State. Whether the raccoons which reach the southern counties from 
the south and southwest, along the Rio Grande and the San Juan River 

1 Outdoor Life, p. 933, Nov., 1905. 3 Further Notes on Mammals of Colorado, p. S3, 1908. 

2 Mammals of Colorado, p. 260, 1906. 

90432°— 11 13 



194 NORTH AMERICAN FAUNA. [No. 33. 

and tributary streams, are distinct from those east of the mountains 
is not certain, as no specimens are available for comparison. It 
seems probable, however, on geographic grounds, that the raccoons 
occurring west and south of the San Juan Mountains, in Archuleta 
and La Plata Counties, are P. mexicanus. The raccoons which were 
reported in 1907 from the San Luis Lakes and the cottonwood-fringed 
streams at the western base of the Sangre de Cristo Range may have 
reached the San Luis Valley from the east by crossing over the low 
Mosca Pass from the head of the Huerfano River. 

Raccoons are stated to be common at a number of points on the 
plains east of the mountains, and probably occur along all the larger 
streams, which are well fringed with cottonwoods, and follow them 
some distance into the foothills. There is a skull from the foothills 
near Arkins, Larimer County, and the distribution in the foothills is 
further indicated by reports from Gardner and La Veta, Huerfano 
County. A specimen hi the Warren collection from the mouth of 
the Platte Canyon, Douglas County, has been examined by the Bio- 
logical Survey. Raccoons are said to be tolerably common in the 
marshes and along streams in the Loveland region. Loring, who 
secured a specimen at Loveland in 1897, states that he captured it in 
a cat-tail marsh which was bordered by a fringe of cottonwoods, and 
he thinks the animals were living in holes in the banks at that point, 
rather than in hollow trees. At Las Animas in July, 1892, Dr. A. K. 
Fisher noted along the Arkansas River and on the banks of irrigation 
ditches a great many tracks made by raccoons searching for frogs. 
Prof. D. E. Lantz reported a few raccoons on Big Sand} T Creek, near 
Hugo, in 1905. Warren says they are found at Watervale, in south 
central Las Animas County. 1 

In May, 1909, raccoons were abundant among the sandstone 
ledges along the South Fork of the Republican River at Tuttle and 
in similar situations at Wray, while many tracks were seen on sand- 
bars in the South Platte River 2 miles northeast of Sterling. Several 
dens were located among the rocks near Tuttle, at one of which; 
within 50 yards of our camp, a large, dark-colored female raccoon 
was trapped. Dismal howls and barks, not unlike those of a dog, 
advised us of her capture one night about 10 o'clock. The raccoons 
at Tuttle were feeding extensively upon grasshoppers, and the excre- 
ment found near the dens consisted very largely of the remains of 
th se insects. 

Raccoons are said to follow up the Los Pinos as far as Bayfield, 
La Plata County, and are found in small numbers along the San 
Juan River to Pagosa Springs, but I was informed at both localities 
that prior to 10 or 15 years ago none were present. At Arboles I saw 
tracks in a dry arroyo extending north from the San Juan River. 

i The Mammals of Colorado, p. 219, 1910. 



1911.] MAMMALS. 195 

Ursus americanus Pallas. Black Bear. 

The black bear is still tolerably common in the wilder parts of the 
mountains, especially in the ranges surrounding North, Middle, and 
Egeria Parks, and in the Elk, Saguache, San Juan, and La Plata 
Mountains, where the forests are most extensive. It is dichromatic, 
as in other parts of the Rocky Mountains, and the numerous skins 
examined and the statements of hunters and trappers indicate that 
black, brown, and cinnamon animals frequently occur in the same 
localities. The typical black phase is least common in the southern 
mountains and at the lower elevations, but is more frequent in the 
northern part of the State. 

In a series of nine very large skins which I examined in the store 
of Fred Selak, near Coulter, Middle Park, five were brown and four 
black. They had been purchased from a trapper living in the 
Vasquez Mountains, near the headwaters of the Williams Fork of 
Grand River. Most of them had been taken in large log traps, and 
as a consequence had mutilated their claws by their efforts to escape. 
A number of bears are killed each spring in the lodgepole pine and 
spruce forests on the mountain ranges bordering Middle Park, and 
skins from this section brought about $25 each in 1905-6. In the 
aspen forests on the White River Plateau, and near Arapahoe Pass, 
in the Rabbit Ear Mountains, many of the larger trees were badly 
scarred to a height of 8 feet or more by the claw marks of bears. In 
October, 1906, I saw the tracks of a medium-sized bear in the snow 
above timberline near Berthoud Pass. Bears were reported in vary- 
ing numbers in 1905-6 on the Medicine Bow, Park, and Gore Ranges, 
Elk Head and Williams River Mountains, Danforth Hills, and 
Mount Cullom, west of Green River. Tracks are said to have been 
seen among the pinyons near Douglas Spring, in the Escalante Hills, 
in the winter of 1905-6. At Glenwood Springs I saw many skins 
from the high country south of Grand River and others from the 
White River Plateau. 

Bears are now becoming scarce on the eastern slopes of the Front 
Range. In 1893 Loring reported them common in the Estes Park 
region, one trapper having killed 14 in three years; but the animals 
are stated to be rare at present in the higher parts of Clear Creek 
County (Floyd Hill and Grays Peak). 

Bears were not considered uncommon in the mountains of southern 
Colorado in 1907, being reported in greatest numbers in the Cochetopa 
Hills, in the San Juan Mountains north of Pagosa Springs and Yalle- 
cito, in the La Plata Mountains northeast of Mancos, and in the 
San Miguel Mountains from Mount Wilson west to Lone Cone. 
They occur more or less commonly the entire length of the Sangre de 
Cristo and Culebra Ranges, particularly on the more heavily forested 
eastern slopes, and also in the West Elk Mountains and on the 



196 NORTH AMERICAN FAUNA. [No. 33. 

Uncompahgre Plateau. They are rare in the rough parts of Las 
Animas County. Prof. Lantz heard of one which was killed 20 miles 
south of Higbee about 1908. 

In the Pagosa Springs region, as elsewhere, bears are usually found 
in the aspen and spruce belt, but they occasionally come down from 
these elevations into the yellow pine forests and kill many sheep. 
The winter dens are left about May 1, or sometimes earlier, and, as 
sufficient snow for tracking remains in the higher country for an- 
other month, May is considered an excellent time for bear hunting. 
Most of the hunting is done with hounds, and large packs of bear dogs 
are owned at both Pagosa Springs and Mancos, which are probably 
the best two outfitting points for bear hunts in the southern moun- 
tains. A number of hunting parties were located in the mountains 
north of Pagosa Springs at the time of my visit, late in May, 1907. 
A large black bear was killed on Pagosa Peak May 30, and a little 
earlier in the same month an old female and cub were killed on 
Middle Mancos River, 10 miles east of Mancos. A brown cub was 
roped at the Evans sawmill, near Vallecito, a few days before I 
reached there, June 5. Mr. Steve Elkins, of Mancos, one of the most 
successful bear hunters in southwestern Colorado, states that the 
largest black bear which he has killed in the La Plata Mountains 
weighed very nearly 500 pounds. 

In the San Miguel region bears appear to occur regularly at a lower 
altitude than elsewhere in Colorado, and are not uncommon 
down into the piny on country. One is said to have been killed 
within a mile of Placerville in the spring of 1907, while in the well- 
settled country between Coventry and Norwood a large brown bear 
was seen at the carcass of a cow in an open field a few days before 
my arrival, July 1. The animals are occasionally seen along Naturita 
Creek near Coventry. In the pinyon country of western San Miguel 
and Montrose Counties the cinnamon phase prevails, and very light 
colored individuals are not uncommon. Mr. Henry Huff, an Indian 
living at Norwood, showed me the skin of a remarkably light colored 
yearling cub which he had captured in the Dry Creek Basin by crawl- 
ing into the den after killing the mother, an old cinnamon, and the 
other cub. This skin was a pale creamy yellowish white throughout, 
with the exception of the face and nose, which were very light brown. 

Bear signs were abundant in the aspen belt on Lone Cone late in 
July. A great many of the aspens were scratched and clawed, many 
of the marks being recent. Two trees showed plainly where the 
animals had climbed to the upper branches — sure evidence that they 
were not grizzlies. The claw marks on these trees were very distinct 
to a height of 30 feet or more. (See fig. 29.) Other bear signs — 
chiefly overturned logs and rocks where the animals had been search- 
ing for ants and beetles — were, abundant on Lone Cone, and indicated 
that the region was one of bruin's favorite ranges. 



1911.] 



MAMMALS. 



197 



The Biological Survey has a series of 10 bear skulls which Mr. 
Theodore Roosevelt secured on Divide Creek, Garfield County, in 
April, 1905; also three from Pagosa Springs, taken in the summer of 
1907. Skulls in the National Museum were collected by Mr. E. 
Thompson Seton in the Rifle region and near Mount Marvine. Allen 
found black and cinnamon bears in about equal numbers in Park 
County in 187 1. 1 Coues and Yarrow mention a specimen obtained 
by Lieut. Marshall at Pagosa Springs in 1874. 2 

Ursus horribilis Ord. Grizzly Bear; Silver-tip. 

At present grizzly bears are uncommon, if not rare, in the northern 
mountains, but are 
occasionally seen in 
the wilder moun- 
tains of southern 
Colorado, parti c u- 
larly in the San 
Juan, La Plata, and 
San Miguel Ranges. 
Many of the data 
respecting the griz- 
zly (or silver-tip, as 
it is generally known 
to hunters) within 
the State, past and 
present, are unsat- 
isfactory and some- 
what conflicting, 
many of the reports 
undoubtedl} T refer- 
ring to black bears, 
or more often to 
large cinnamon 
bears. 

In 1905-6 the best 
informed hunters 
and trappers in the 
northern mountains considered the grizzly rare. The reports which 
follow seem to refer beyond question to this species. A very large 
old silver-tip was reported in the region about Strawberry and Grand 
Lakes, in northeastern Middle Park, in 1905. This old fellow is said 
to have ranged that part of the western slope of the Front Range 
for a number of years, and is well known to the hunters of the region 
as Old Saddleback — so called because of an area of light-colored fur 




Fig. 29. — Claw marks of black bear on aspen (Populus tremulohles), 
Lone Cone, San Miguel Mountains, at 10,000 feet. 



i Buil. Essex Inst., VI, p. 54, 1874. 



Explorations W. of 100th Meridian, p. 07, 1875. 



198 NORTH AMERICAN FAUNA. [No. 33. 

near the middle of the back. 1 Another very large bear was reported 
the same year from Mount Baker, at the eastern end of the Rabbit 
Ear Mountains, and according to well-informed hunters a few silver- 
tips were yet to be found in the vicinity of Pyramid Peak, in the 
mountains west of Egeria Park. The skin of a two-thirds-grown 
silver-tip killed near Sleepy Cap, Williams River Mountains, in 
October, 1S90, by Mr. Oscar Lampton, of Montrose, has front claws 
3 inches in length, following the curve. Lumbermen at the tie 
camps on the headwaters of Grand Encampment River, at the 
northern end of the Park Range, stated that a large silver-tip often 
came about the camps during the spring of 1906. Mr. William Cross, 
the Glenwood Springs taxidermist, showed me a photograph of the 
skin of a very large silver-tip killed in October, 1906, on the northern 
side of the Book Plateau, in extreme western Garfield County, by 
Mr. Harry Payne Whitney, of New Haven, Connecticut. This bear, 
judging from the photograph, was a remarkably large one, and the 
claws on the forefeet were very long and regular. The longest fore 
claw (measured by Mr. Cross) was exactly A\ inches along the curve. 
The skull in Mr. Cross's shop was very large and massive and indi- 
cated an old animal. Mr. Cross stated that at least two silver-tips 
have been killed in the mountains near Gypsum during recent years — 
a large individual in the spring of 1903 on the head of Gypsum Creek 
by Mr. Muckey; and another bear, nearly grown, in the same region 
in the spring of 1907 by Mr. Jake Borah, the veteran Meeker hunter. 

I have no data respecting the recent occurrence of this bear along 
the eastern slopes of the Front Range, and it appears to be now 
extremely rare or entirely absent. In the early seventies of last 
century grizzlies were not uncommon along the higher crests of the 
Front and Saguache Ranges. Brewer mentions six which he saw 
above 13,000 feet in the mountains in close proximity to Grays Peak 
and near Mount Yale, and says: "Judging from the few seen and 
from skins examined in Denver, they are smaller than those of Cali- 
fornia, the hair not so long and shaggy, the color more silvery, or 
truly grizzled." 2 

In the summer of 1907 considerable information was secured on 
the present range and abundance of the silver-tip in the southern 
mountains. Small numbers are found in the San Juan Mountains 
north of Pagosa Springs and Vallecito, and according to Forest Ranger 
E. E. Chapson, an average of one or two had been killed north of 
Pagosa Springs each spring until 1907, when none were killed. 
Silver-tips kill a number of cattle in the southern foothills of the 
San Juan some years, one having killed four head in the country 
to the west of Pagosa Peak in May, 1907. This is doubtless the same 

1 Doubtless an extreme of the ord inary grizzling or silver tipping of the hairs, which is normally heaviest 
on the saddle and between the shoulders. 

2 Am. Nat., V, p. 221, 1871. 



1911.] MAMMALS. 199 

individual which was reported as killing cattle on the middle fork of 
Piedra River a few weeks later. Henshaw states that grizzlies were 
quite common in the higher San Juan Mountains as late as 1873. As 
far as could be learned, the silver- tip has a higher range than the black 
bear, and is rarely met with as low as the yellow pine belt. 

Mr. Steve Elkins, of Mancos, states that the silver-tip is encoun- 
tered rather often in the La Plata Mountains and that he has 
killed several during the past 10 years. Among these and other skins 
from the La Plata region examined by Mr. Elkins, much individual 
color variation was exhibited, some being a faded dirty yellowish, 
with the hair long and shaggy, while others were dark clove brown 
or blackish, with hair well silvered at the tip — sleek, short-haired, and 
beautiful animals. The skin of a large silver-tip which was trapped 
on West Dolores River in the spring of 1906 is owned by Mr. Harry 
Pyle, of Dolores. 

A silver-tip was shot on North Mesa, just west across the Naturita 
Valley from Lone Cone, San Miguel Mountains, October 2, 1905, by 
Mr. Oscar Lampton, of Montrose. In Montrose Mr. Lamp ton showed 
me the skin of this bear hi the form of a beautiful full-head rug with 
the skull inside. It was a good-sized animal with the longest fore 
claws measuring 4 inches, but the unworn teeth indicated that it was 
young. The foreclaws on the skin of a large old cinnamon hear from 
the Uncompahgre Plateau, 15 miles west of Montrose, are markedly 
smaller and shorter than those of the smaller of Mr. Lampton's two 
silver-tips (from the Williams River Mountains) . The silver-tips are 
both in good fur and agree precisely in color, being a uniform grizzled 
black-brown on the back; the legs, a rich dark shade of the same 
color without the grizzling; face, plain brown; nose, clay brown; 
claws, black. 

Mr. Henry Huff, an Indian, and Mr. J. P. Galloway, both of Nor- 
wood, killed a very large old male silver- tip among the Engelmann 
spruces just below timberline on the west slope of Lone Cone, San 
Miguel Mountains, May 26, 1907, after a long chase from the lower 
north slope of the mountain. A female bear seen at the same time 
escaped, and is said still to range the Lone Cone country. The snow 
was reported to have been 8 feet deep where the old male was dis- 
patched, and hence it was with some misgivings that we set out on 
July 27 to secure the skull. The Indian led us straight to the spot, 
however, in a dense spruce thicket where the down timber was heaviest. 
The snow had but recently melted, and the immense carcass had 
settled across a large log. Mr. Galloway estimated the weight of the 
bear when killed at between 800 and 1,000 pounds, and states that 
the skin at the time of its removal from the body was so heavy that 
he and his companion had great difficulty in packing it upon a horse. 
The stomach of this bear, examined on the day it was killed, contained 



200 NOBTH AMERICAN FAUNA. [No. 33. 

nothing but a double handful of ants and the larvae of wood-boring 
beetles. Fortunately, the skin was at a near-by ranch at the time 
of my visit to Lone Cone, and I measured and photographed it. 
The pelt was in prime condition, with a uniformly short, dense pelage, 
shortest on the back between the shoulders, where the hair averaged 
about 3 1 inches in length, and longest on the sides of the upper fore- 
legs, where it averaged 6 1 inches. The Indian aptly likened the flaps 
of long shaggy hair on the upper forelegs to a cowboy's chaps, and 
said the resemblance in life was most striking. The fur was a dark, 
rich clove brown at base throughout, shading into plain brownish 
black on sides and on fore and hind legs; ears and face, dark brown; 
nose, clay brown; dorsal area, from rump to between ears, a beautiful 
silver}" clove brown, brightest between shoulders, owing to greater 
length of silvered area at tips of hairs. The ears were comparatively 
short, but could not be accurately measured on the dried skin. The 
following skin measurements indicate the large size of this bear, 
although they are probably somewhat exaggerated on account of 
stretching: Distance between tips of fere claws, outstretched skin, 
96^ inches; length from tip of tail to end of nose, outstretched skin, 
87 inches; longest fore claw, along curve, 4J inches; across angle 
(straight), 3 J inches; hind claws (badly worn), under 1 inch. Mr. 
Galloway's flesh measurements of hind foot were: Heel to tip of claws, 
12^ inches; across base of toes to edge of hair on side, 1\ inches. The 
old age of this bear was indicated by the much-worn teeth and hind 
claws. One of the massive upper incisors was a mere stub, having 
been broken off for some time. Several rifle balls had penetrated 
the posterior part of the skull and badly shattered it, but the anterior 
part was in good condition. 

Mr. E. R. Warren, of Colorado Springs, has in his collection the 
skull of a medium-sized silver-tip which was killed some years ago 
in the Dry Creek Basin, west of the San Miguel Mountains, by Mr. 
Jack Watson, of Norwood. The Dry Creek Basin is Upper Sonoran 
country, the altitude being not over 6,000 feet — an unusually low 
elevation for the silver- tip. 

While on Lone Cone, July 27, I photographed a large aspen which 
plainly showed fairly fresh claw marks of a large bear, apparently a 
silver-tip, to. a height of 13 feet. Several other aspens on the upper 
slopes of Lone Cone were claw-marked by silver-tips, but most of 
the scratching was old. 

Grizzlies are becoming very scarce on the Sangre de Cristo and 
Culebra Ranges, but are still occasionally encountered among the 
Cochetopa Hills west of Saguache. A large individual is said to have 
been taken in a trap on the head of Trinchera Creek in the spring of 
1907 by Mr. George Wheeler, of Fort Garland. That the grizzly 
was occasionally met with east of the mountains in early times is 



1911.] MAMMALS. 201 

shown by an account of a bear which killed Lewis Dawson, a mem- 
ber of Jacob Fowler's party, at the mouth of Purgatory River in 182 1. 1 
The National Museum has the skin and skull of a grizzly killed 
near Twin Lakes July 28, 1876; a skull from Burro Mountain, near 
Elwood, in the San Juan Mountains; and two skulls of bears killed 
on Miller Creek, in the Glenwood Springs region, in either 1896 or 
1897. The Biological Survey has skulls from Pagosa Springs and 
Lone Cone, and Coues and Yarrow mention a skin secured near 
Pagosa Springs in 1874 by Lieut. Marshall. 2 

Scalopus aquaticus intermedins (Elliot). Plains Mole. 

Moles enter the northeastern part of the State from the plains. 
Thus far they are known only from Yuma County, but very likely 
occur throughout the sand hill region from the Arickaree Valley 
north at least to Holyoke. I found them abundant at Wray in 
December, 1907, in the valley along Chief Creek, and especially 
numerous in the young apple orchard of ]\lr. W. E. Wolfe, a mile east 
of the town. The loose sandy loam soil of this orchard of several 
acres was furrowed with a network of mole runways. Notwithstand- 
ing the great abundance of moles on his ranch, Mr. Wolfe has detected 
no injurious effects from their presence, and considers that they 
confer a benefit by thoroughly working over the surface soil in search 
of worms. A male was trapped in Mr. Wolfe's orchard December 
13, when the ground was partially frozen and covered with a 3-inch 
fall of snow, which is good evidence that moles are active in loose 
soil during winter weather. I again visited Mr. Wolfe's orchard in 
May, 1909, but found only one fresh mole runway. Several older 
runways seen, however, indicated that the animals either were 
inactive or were working at a considerable depth. In driving from 
Wray to Yuma I saw mole runways in the sand at two points, 5 and 
12 miles, respectively, west of Wray. At Yuma a small colony was 
working in soft soil on the embankment along the Burlington Rail- 
road a mile east of town. No signs of moles were seen in driving 
northwest from Yuma to Sterling. 

The Wray specimen agrees very closely with topotypes of S. a. 
intermedins from Alva, Oklahoma, in small size, pale coloration, and 
particularly in small size of skull. The skull of this specimen meas- 
ures: Total length, 33.9; mastoid breadth, 17.4; palatal length, 15. 
Skin measurements are: Total length, 145; tail vertebra?, 28; hind 
foot, 23. 

An alcoholic specimen from Dry Willow Creek, 12 miles southeast 
of Wray, in the collection of the Colorado Historical and Natural 
History Society, has been recorded by Warren. 3 

i Journal of Jacob Fowler. New York, 1898. 

2 Explorations W. of 100th Meridian, V, p. 66, 1875. 

3 Further Notes on Mammals of Colorado, p. 84, 1908. 



202 NORTH AMERICAN FAUNA. [No. 33. 

Sorex personatus I. Geoff. St. Hilaire. Masked Shrew. 

This species is apparently far less common and widely distributed 
in the Colorado mountains than S. obscurus, although at one point — 
St. Elmo, at ](),()()() feet in the Saguache Mountains — it was the 
more common of the two. I have found it at only two localities in 
the mountains of northern Colorado — an adult male being taken at 
9,000 feet on Arapahoe Pass, Rabbit Ear Mountains, in a damp, 
mossy place on a cold north slope; and a female being trapped in a 
cold bog near Pearl, in the northern end of North Park, at 8,500 feet. 
At St. Elmo, in October, 1907, eight specimens of 8. personatus were 
taken in three nights' trapping, chiefly beneath mossy rotten logs 
in the damp thickets of Engelmann spruce along Chalk Creek. The 
above specimens are in all respects typical S. personatus, but one 
from Loveland, at the eastern base of the foothills, shows an approach 
to the paler and grayer plains form 8. p. haydeni. 

Several specimens of S. personatus taken by Mr. R. T. Young in the 
vicinity of Boulder, and at Buchanan Pass, Boulder County, are in the 
National Museum. Others from Irwin, Gunnison County, at 10,700 
feet; Lake Moraine, El Paso County, at 10,250 feet; and Mud Springs, 
on the White River Plateau, at 8,850 feet, are in the Warren collec- 
tion. Warren states that he has recently identified specimens from 
the Summit House on Pikes Peak, 14,147 feet altitude. 1 A specimen 
from Marvine Lodge, Rio Blanco County, is recorded by Felger. 2 

Sorex obscurus Merriam. Rocky Mountain Shrew. 

This is the common shrew of the high Colorado mountains in the 
Canadian and Iludsonian zones. Specimens from timberline on 
Longs Peak and Mount McClellan indicate its upper limits, while on 
cold slopes and along streams, where the conditions are semiboreal, I 
have taken it as low as 5,800 feet, and it has recently been recorded 
from Boulder (5,400 feet). 3 It is usually trapped in mountain bogs 
and beneath moss-covered rotten logs in spruce and aspen forests. In 
a grassy mountain park near Gore Pass, between Egeria and Middle 
Parks, one was secured in a runway of Microtus nanus, at some dis- 
tance from the forest; while Loring trapped a specimen in a cabin 
near Silverton. Three of these shrews got into my traps set among 
rank vegetation in an aspen bog near Uncompahgre Butte, on the 
Uncompahgre Plateau, at 9,000 feet, July 17, 1907. At St. Elmo, in 
the Saguache Mountains, at 10,000 feet, I collected two specimens in 
October in the damp spruce thickets along Chalk Creek. This shrew 
was not so common as 8. personatus at this point. The mountain 
shrew is chiefly nocturnal, but Loring caught one at Gold Hill during 
the daytime, and one afternoon at Baxter Pass, on the Book Plateau, 

i The Mammals of Colorado, p. 263, 1910. s Young, Proc. Acad. Nat. Sci. Phila., p. 407, 1908. 

2 Univ. of Colo. Studies, VU, No. 2, p. 146, 1910. 



1911.] MAMMALS. 203 

I saw one running about under a spruce log. Specimens of this spe- 
cies from a great many Colorado localities are in the Biological Survey 
collection. 

Sorex vagrans dobsoni Merriam. Dobson Shrew. 

A specimen of the little Dobson shrew from Lake Moraine, El Paso 
County, which Warren sent to the Biological Survey for determina- 
tion in 1905, is the only Colorado record. It was collected at an 
altitude of 10,250 feet. 

Sorex tenellus nanus Merriam. Dwarf Shrew. 

Sorex tenellus nanus Merriam, N. Am. Fauna No. 10, p. 81, 1895. Type from Estes 
Park, Larimer County, Colorado. 

This little shrew must be very rare, as not one was captured in my 
three seasons' work on all the higher mountain ranges of the State. 

In addition to the type, an adult female taken in Estes Park 
August 3, 1895, by Mr. Edward A. Preble, the Biological Survey has a 
skull from Westcliffe, Custer County. A third specimen from Colo- 
rado, in the Warren collection, has been identified by the Biological 
Survey. It was collected on Bear Creek, in the mountains near 
Colorado Springs. 

Neosorex palustris navigator Baird. White-bellied Water Shrew. 

In the Colorado mountains the large water shrew is found chiefly 
between 7,500 and 10,500 feet, in the Canadian zone, but I have 
taken it at 6,400 feet, and it has been reported once from as low as 
6,000 feet. The scattered localities at which specimens have been 
taken indicate a general distribution over the mountainous sections. 
The species is always found near water, and is usually taken in traps 
set in the moss at the edge of mountain streamlets, or in the dense 
vegetation of cold bogs or mountain meadows. Little is known of 
its habits. 

At St. Elmo, in the Saguache Mountains, I found these shrews 
common along some small streams tributary to Chalk Creek, in 
October, 1907. Three out of four specimens collected at this point 
were taken in traps placed on moss-covered rocks behind a small 
waterfall, where the vegetation was saturated by the dashing spray. 
Mr. J. W. Frey, of Salida, reports water shrews as common on a 
small stream heading on the east slope of Round Mountain, at the 
northern end of the Sangre de Cristo Range; and they are reported 
from the head of Muddy Creek, in northwestern Huerfano County. 
In a long line of traps set along Maverick Creek, 2 miles northeast of 
Coventry, the night of July 4, I secured three water shrews, all being 
taken in traps beneath waterfalls. The elevation is 6,400 feet, excep- 
tionally low for this shrew, and the pinyon and juniper clad bluffs 
bordering the Maverick gave anything but a boreal environment to 
this species, which is so closely associated with cold dashing mountain 



204 NORTH AMERICAN FAUNA. [No. 33. 

streams, mossy jocks, and towering spruces. At Coventry I was 
informed water shrews are occasionally seen at the Stephens ranch, 
in the wesl era end of West Paradox Valley, at a little under 6,000 feet. 
There are a number of cold springs, bordered with a dense growth of 
water cress at this point, which is at the eastern base of the La Sal 
Mountains. 

Specimens in the Biological Survey collection are from Gold Hill; 
Elkhorn, Larimer County; Cochetopa Pass; St. Elmo; Almont; Her- 
mit ; Rico ; and Coventry. Loring saw the skin of one of these shrews 
at Silverton. The National Museum has specimens from Black Hawk; 
Middle Park; Mount Elbert; and old Fort Massachusetts, near Fort 
Garland. Others from Crested Butte and Lake Moraine are in the 
Warren collection. 

Corynorhinus macrotis pallescens Miller. Big-eared Bat. 

Very little information is at hand regarding the range of the big- 
eared bat in Colorado. Thus far it appears to have been taken at only 
five localities, all of them along the eastern base of the foothills. A 
specimen in the collection of the Colorado Agricultural College, 
secured at Fort Collins by Mr. S. Arthur Johnson, has been identified 
by the Biological Survey. Miller records another specimen from 
Larimer County, 1 and Mr. A. E. Beardsley has taken two of these bats 
at Trinidad, according to Warren. 2 More recently Mr. R. T. Young 
has recorded a specimen collected in Boulder Canyon, at an approxi- 
mate elevation of 7,000 feet. 3 Junius Henderson, curator of the 
museum of the University of Colorado, Boulder, informs me of the 
capture of a specimen in a tunnel at Crisman, Fourmile Canyon, 
Boulder County, at an altitude of 7,000 feet, by John J. Blanchard, 
November 1, 1909. 

Nyctinomus mexicanus Saussure. Free-tailed Bat. 

Four males of this austral species from Newcastle, Garfield County, 
have been identified for Warren, who states that he secured them 
along with a number of brown bats (Eptesicus fuscus) , July 16, 1907, 
"behind a sheet-iron shutter on a building where they had gone for 
shelter during the day." 4 The elevation of Newcastle is 5,374 feet. 

Nyctinomus has not been taken elsewhere within the State, or 
indeed anywhere in the neighboring parts of New Mexico, Arizona, 
and Utah. The Newcastle record is therefore a very great northward 
extension of known range for the Rocky Mountain region. The animal 
must reach this point from the southwest through the warm Grand 
Valley extension of the Colorado River Sonoran area; and it is note- 
worthy that this Lower Sonoran species should be found near the 
extreme upper edge of a narrow tongue of the Upper Sonoran zone, 

i X. Am. Fauna No. 13, p. 53, 1897. 3 Proc. Acad. Nat. Sci. J'hila., p. 407, 1908. 

2 Mammals of Colorado, p. 267, 1906. ' Further Notes on Mammals of Colorado, p. 85, 1908. 



1911.] MAMMALS. 205 

surrounded on all sides except the west by broad belts of boreal 
country. The occurrence of the free-tailed bat at Newcastle sug- 
gests that it may be found in the lowest and warmest valleys of the 
southwestern counties, particularly in the Grand Valley between 
Grand Junction and the Utah boundary, which is the lowest and 
most truly desert-like part of the Upper Sonoran area in western 
Colorado. Hitherto the northern known limit of range in the Colo- 
rado Valley has been Grand Falls, Arizona, in the valley of the Little 
Colorado. 

Nyctinomus depressus Ward. Tacubaya Free-tailed Bat. 

Through Mr. E. R. Warren, of Colorado Springs, the Biological 
Survey has examined a Colorado specimen of this large free-tailed 
bat. Dr. S. M. Bradbury, of Grand Junction, who owns the speci- 
men, says it was killed by some boys at Grand Junction about 1900. 
Warren has already placed this specimen on record. 1 

The type locality of N. depressus is Tacubaya, Federal District, 
Mexico. A few individuals have been captured in the desert areas 
of the southwestern United States, but the Grand Junction record 
extends the known range of the species far to the east and north. 

Antrozous pallidus (LeConte). Pale Bat. 

The pale bat occurs in several of the lowest and warmest Upper 
Sonoran valleys of extreme southwestern Colorado, but has not been 
reported from north of the Grand River Valley. An extreme record 
is that of a National Museum specimen taken at Pueblo, east of the 
mountains, since the species is largely restricted to the desert regions 
of the Southwest. 

An excellent opportunity for observing this species was afforded 
at Ashbaugh's ranch, in the Ale Elmo Canyon, 20 miles west of Cortez, 
in June, 1907. At dusk each evening numbers of these large bats 
appeared about the cliffs immediately north of the ranch, coursing 
in great circles above the upper rim rock in quest of insect prey. 
The large size, rapid sailing flight, and slow wing beats made these 
bats most conspicuous in contrast with the hosts of small Myotis and 
Pipistrettus which darted about in the gloaming in jerky, erratic 
flight. The majority of the pale bats flew so high as to be out of gun 
range from the base of the cliffs, but I managed to shoot two females 
on the evening of June 21. These bats were rarely observed about 
the ranch buildings, and invariably appeared first over the cliffs in 
the early twilight. Although none were seen actually emerging from 
the cliffs, the numerous cracks and crevices doubtless formed their 
retreat during the day. 

In the deep canyon of Tabeguache Creek, north of Nucla, Montrose 
County, six or eight pale bats flew about the cliffs at a considerable 

1 Mammals of Colorado, p. 2(58, 1900. 



206 NORTH AMERICAN FAUNA. [No. 33. 

height above our camp in the gathering twilight, July 19. They 
appeared in companies of two and three and coursed about with 
characteristic steady, rapid flight. I have observed this species at 
the following two points in the Grand River Valley, a single individual 
at each locality: In the rock-walled canyon of Plateau Creek, 5 miles 
east of Tunnel, Mesa County, September 30, 1906; and among the 
cliffs along Grand River, 7 miles west of Rifle, August 14, 1907. 
Neither bat was collected, but identification was reasonably positive. 
A large light-colored bat, apparently Antrozous, is commonly reported 
from the Grand River Valley near Crevasse, Mesa County, where it is 
said to live in ranch buildings. 

The first recorded instance of the capture of Antrozous in Colorado 
is that given by Coues and Yarrow of a specimen taken at Pueblo by 
W. D. Wheeler in October, 1S74, and deposited in the United States 
National Museum. 1 Dr. F. W. True informs me that this specimen, 
which appears to have been mummified, was destroyed about 1905, 
but that the skull is preserved. 

Myotis subulatus (Say). Say Bat. 

Vespertilio subulatus Say, Long's Exped. to Rocky Mts., II, p. 65, 1823. Type 
from Arkansas River, near La Junta, Otero County, Colorado. 

Warren records two specimens of the Say hat from Colorado 
Springs. 2 These and the type appear to be the only Colorado speci- 
mens of this widely distributed species known at the present time. 
The species should be present over much of eastern Colorado. 

Myotis lucifugus longicrus (True). Long-legged Bat. 

This western form of the little brown bat is not uncommon in west- 
ern and southern Colorado. One shot at Steamboat Springs as it 
was flying over Bear River at dusk August 1, 1905, is fighter colored 
than normal M. longicrus. Two others in grayish pelage were se- 
cured on the White River meadows, a few miles east of Meeker, 
August 12. At the last locality large numbers of these bats were 
flying over White River late in the evening. A dark female was 
collected at Coventry August 1, 1907, in my bedroom at 2 a. m. 
This form is also represented in the Biological Survey collection by a 
specimen from Grand Junction, collected June 23, 1893, by Mr. J. 
Alden Loring; and another collected on Conejos River, west of An- 
tonito, by Mr. James H. Gaut, September 4, 1904. 

Three specimens from the eastern part of the San Luis Valley have 
been identified recently for Warren. Two were collected above 
Herard, in Madenos Canyon, Saguache County, at an elevation of 
8,700 feet, July 12, 1909. A third, from the Medano Springs ranch, 
east of Mosca, June 22, 1909, has distinct whitish edgings to the 

» Explorations W. of 100th Meridian, V, p. 85, 1875. 2 Mammals of Colorado, p. 267, 1906. 



1911.] MAMMALS. 207 

uropatagium, and is altogether paler than normal longicrus. Young 
records a mutilated skin in the collection of the Academy of Natural 
Sciences, Philadelphia, 1 from Eldora, in the high foothills of Boulder 
County. 

Myotis evotis (H. Allen). Long-eared Bat. 

The long-eared bat has been taken on both sides of the mountains 
in the Upper Sonoran zone, and a specimen recently recorded by 
Mr. R. T. Young from an elevation of nearly 7,000 feet, in the Steam- 
boat Springs region, 1 shows that the species occasionally ranges into 
the Transition zone. 

A series of 11 specimens was collected at the old L7 ranch, a few 
miles southeast of Sunny Peak, Routt County, in August, 1906. 
The bats had their abode in the deserted ranch house, and numbers 
could be seen flying in and out of the open doorway at dusk. I se- 
cured the specimens by entering the house after nightfall with a 
candle. Some were caught with an insect net, while others were 
knocked down with a hat. The above series presents considerable 
color variation, due largely to age. Three are immature and in gray 
pelage, but the majority are light yellowish brown. In 1907 this 
species was tolerably common at Ashbaugh's ranch, in the McElmo 
Canyon, Montezuma County, from June 20 to 23, where two were 
caught in the house after dark. One which flew into a house at Do- 
lores the evening of June 27, 1 captured with my hand. At Coventry 
this was the most abundant bat about the ranch buildings in July. 
Five were caught in a house after dark July 24 as they flew in at 
the open window, attracted by the light. I have not taken this 
bat about rocky ledges and cliffs, and it appears to frequent mainly 
houses and outbuildings. Nearly all the above specimens are females. 
Specimens from Loveland have been recorded by Miller. 2 

Myotis yumanensis (H. Allen). Fort Yuma Bat. 

This pale southwestern species is at present known in Colorado 
only from western Routt County, but eventually may be found at 
other points along the western border of the State, in the Upper 
Sonoran zone. It is represented by two females from Snake River, 
south of Sunny Peak, August 28, 1906; and a male from near Lily, 
at the confluence of the Snake and Bear Rivers, September 9, 1906. 
I secured these bats in deserted ranch buildings after nightfall, 
where they were not at all common, being greatly outnumbered by 
M. evotis. 

The above specimens accord well in color with typical yumanensis 
from Fort Yuma, California, but have a somewhat longer forearm 
and foot. 

Proc. Acad. Nat. Sci. Phila., p. 408, 1908. 2 N. Am. Fauna No. 13, p. 80, 1897. 



208 NORTH AMERICAN FAUNA. TNo. 33. 

Myotis calif ornicus (And. and Bach.). Little California Bat. 

This small bat has been taken thus far at several widely separated 
localities in southern and western Colorado, and east of the moun- 
tains is known from two points in the Arkansas Valley and from 
Boulder Canyon. Its zonal distribution appears to be mainly Upper 
Sonoran. 

In the McElmo Canyon, Montezuma County, it was apparently 
common in June, 1907, and I shot two at dusk as they were flying 
about the rocky cliffs north of Ashbaugh's ranch and caught another 
by lamplight in the house after dark. Another was captured in a 
house in the Grand Valley near Morris, 7 miles west of Rifle, August 
14. An adult male from the San Luis Valley, 7 miles east of Anto- 
nito, Conejos County (8,000 feet), September 1, 1904, was taken by 
Mr. James II. Gaut. Small bats which I saw at Pagosa Springs 
May 28, 1907, and at the Paradox crossing of Dolores River July 7 
were probably M. calif ornicus. Specimens in the Warren collection 
from Bedrock (Montrose County), Salida, and Van Andert's Spring 
on Little Fountain Creek (El Paso County) have been examined by 
the Biological Survey. 

Specimens of M. californicus from the highest elevations — Salida, 
Van Andert's Spring, Antonito, and Rifle — are almost as dark as typical 
californicus.. Others from Ashbaugh's ranch and Bedrock, in the low 
desert valleys of southwestern Colorado, are considerably paler, agree- 
ing precisely in coloration with specimens in the Biological Survey 
collection from the deserts of Arizona and Nevada. They are not as 
pale, however, as M. c. ciliolabrum. 

Since the above was written Mr. Junius Henderson has written to 
me as follows: "Two specimens of Myotis californicus were taken in 
Marchioness Tunnel, Boulder Canyon, altitude 6,200 feet, by John J. 
Blanchard, December 22, 1909, thus making a winter record. I did 
not see these two bats until some ten minutes, perhaps, after he took 
them, when one of them was squealing viciously, but the other was 
quiet. In the warm cabin the former soon began to fly about the 
room. Supposing the winter habits of the animals here to be well 
known, Blanchard made no particular examination of them when he 
found them, but says he believes they were dormant, which is my 
impression from the condition of one of them when it reached me. 
The further fact that this tunnel, unlike some, does not seem to be 
infested with insects or other food for bats would tend to confirm the 
idea of hibernation. Possibly the temperature may give you some 
light on that subject. At the breast of the tunnel, where the bats 
were taken, about 350 feet back into the mountain, the temperature 
was soon after measured and found to be 46° F." 






1911.] MAMMALS. 209 

Myotis calif ornicus ciliolabrum (Merriam). Hairy-lipped Bat. 

The only western Colorado specimens of this small pale form of 
M. californicus appear to be two females from the old L7 ranch on 
Snake River, a few miles southeast of Sunny Peak, Routt County, 
taken August 28 and 29, 1906. Both the above bats I secured in the 
deserted ranch buildings after nightfall. This form was uncommon 
at that locality, being greatly outnumbered by M. evotis. In June, 
1909, I found the hairy-lipped bat abundant in the badland cliffs at 
Chimney Canyon, some 30 miles northwest of Sterling. It appeared 
to be the common bat at this point, outnumbering Eptesicus, the 
only other bat seen, fully 10 to 1. Two were shot as they issued from 
the cliffs at dusk to feed on insects in the bottom of the canyon. 

Pipistrellus hesperus (H. Allen). Western Bat. 

These diminutive black-eared bats — the smallest species found in 
the State — inhabit the Upper Sonoran zone in the western and south- 
western valleys. So far as my observations go, they live only about 
cliffs and in rock-walled canyons, where soon after sunset they issue 
in large numbers from the rocky ledges, and with rapid, erratic flight 
dart about in their nightly quest for insects, appearing in the gloaming 
like large moths. Aside from its small size, P. hesperus in flight may 
be easily distinguished by the very narrow wings. 

In the lower McElmo Canyon, Montezuma County, I found this the 
most numerous species in June, 1907, and it was common also in the 
canyon of Tabeguache Creek, north of Nucla, Montrose County, in 
July. One was seen flitting about the cliffs on Grand River, 7 miles 
west of Rifle, August 14. At early dawn, October 1, 1906, a very 
small bat , which I took to be this species, flew over our camp beneath 
the steep cliffs in the canyon of Plateau Creek, 5 miles east of Tunnel, 
Mesa County. 

There are seven Colorado specimens of P. hesperus in the Biological 
Survey collection, all females: One each from Ashbaugh's ranch (near 
McElmo), June 21, and Tabeguache Creek, July 19, 1907; four from 
Grand Junction, June 22 and 23, 1S93; and another from Rifle, August 
25, 1908. Warren has this bat from Bedrock, Montrose County. 

Eptesicus fuscus (Beauvois). Brown Bat. 

The common brown bat has a wide distribution and probably 
occurs over the whole State, except in the higher mountains above 
the Transition zone. One was shot near Steamboat Springs, Routt 
County, in August, 1905, as it was flying over Bear River; while 
later in the same month another was secured on White River, a few 
miles east of Meeker, Rio Blanco County. Bats which appeared to 
be fuscus were seen flying in McElmo Canyon, Montezuma County, 
in June, 1907, and others over the Dolores River in Paradox Valley, 
July 7. 

90432°— 11 14 



210 NORTH AMERICAN FAUNA. [No. 33. 

This bat has been recorded by Miller from Loveland, 1 and by Allen 
from Florida, La Plata County. 2 According to Warren, Mr. A. E. 
Beardsley reports it common at Greeley. 3 More recently Warren 
mentions taking a specimen at Douglas Spring, Routt County, 1 
at Colorado Springs, and 21 at Newcastle, and says: "Only one of 
the brown bats [Newcastle series] was a male, all the rest being 
females, one of the latter having a small young one attached to a 
teat. This was on July 16, 1907. The Douglas Spring specimen, 
taken June 26, contained a single good-sized embryo." 4 E. fuscus 
has also been taken in Boulder Canyon, at about 7,000 feet, according 
to Young. 5 All the above localities are in the Upper Sonoran zone 
or in the Transition zone. 

Considerable color variation is shown by specimens of E. fuscus 
from different parts of Colorado, some being as dark as normal and 
others much paler. A specimen in the Biological Survey collection 
from the Chimney Bluffs, some 30 miles northwest of Sterling (June 
7, 1909), is the only one examined, however, which shows a near 
approach to the extremely pale coloration of E.f. pallidus. Although 
nearly as pale as pallidus, this specimen has the measurements of 
fuscus, to which it is referred. Another specimen from Steamboat 
Springs has the dark coloration of fuscus, but in large size approaches 
pallidus. 

Eptesicus fuscus pallidus Young. Pale Brown Bat. 

Eptesicus pallidus Young, Proc. Acad. Nat. Sci. Phila., p. 408, 1908. Type from 
Boulder, Colorado. 

The exact status of this peculiar large pale bat will not be known 
until much additional material is received from the plains region of 
eastern Colorado. The type from Boulder in the National Museum 
is larger than fuscus, and is very pale. It measures: Total length, 
127; tail, 50; hind foot, 12. Other specimens from Boulder, the 
measurements of which are given in the original description (1. c), 
are nearly as large as the type. The skull of the type is like that of 
fuscus, only larger, being about the size of E. miradorensis. The type 
of pallidus can be closely matched in size by a dark specimen of 
fuscus from Steamboat Springs, and in pallid coloration by a smaller 
individual from the Chimney Cliffs, northwest of Sterling. 

Mr. Junius Henderson, curator of the museum of the University 
of Colorado at Boulder, writes me of the recent capture of three, 
additional specimens of E.f pallidus at Boulder. Warren, who has 
examined these bats, says that they are paler than his specimens of 
E. fuscus and agree in measurements with the type of pallidus. 

> X. Am. Fauna No. 13, p. 98, 1897. * Further Notes on Mammals of Colorado, p. 85, 1908. 

"■ Bull. Am. Mus. Nat. Hist., V, p. 83, 1893. » Proc. Acad. Nat. Sci. Phila., p. 409, 1908 (footnote). 
3 Mammals of Colorado, p. 268, 1906. 



1911.] MAMMALS. 211 

Lasionycteris noctivagans (LeConte). Silver-haired Bat. 

During the breeding season the silver-haired bat is probably 
restricted largely to the Canadian zone in Colorado, as it is elsewhere 
in North America, but specimens taken in both the Transition and 
the Upper Sonoran zones indicate that at certain seasons it performs 
vertical migrations of some extent. 

One evening in August, 1905, while encamped at the Widows Corral, 
on the White River Plateau, 25 miles southeast of Meeker, I obtained 
frequent glimpses of a small bat flitting about in the aspen forest. 
It flew so near the ground that its course was very difficult to follow 
in the fast-gathering twilight, but it was shot and proved to be a 
specimen of the silver-haired bat. This locality is in the Canadian 
zone, at 8,500 feet. The Biological Survey has a specimen of this 
bat from Rifle, Garfield County, collected by Loring in 1893; while a 
specimen taken by Dr. Elliott Coues in North Park at 10,000 feet, 
September 16, 1876, is in the National Museum. 

Warren has specimens from Green Mountain Falls, Newcastle, and 
Salida, and records the species from Glen Eyrie (near Colorado 
Springs) and Greeley. 1 

Mr. Junius Henderson informs me by letter of the capture of a 
specimen at Boulder June 6, 1909. 

I did not meet with this bat in southern Colorado, but it has been 
recorded from Florida, La Plata County, where Charles P. Rowley 
collected two specimens in 1892. 2 

Nycteris cinereus (Beauvois). Hoary Bat. 

The large hoary bat has been taken only a few times in the State, 
and consequently its local distribution is not well known. The 
normal breeding range is in the Canadian zone, and therefore it may 
be expected to breed in the mountains. Thus far, however, speci- 
mens have been collected only in the eastern foothills and along their 
immediate eastern bases and in the lower valley of Grand River. A 
mounted specimen owned by Dr. S. M. Bradbury, of Grand Junction, 
taken at that locality, has been examined, and there is a specimen 
in the Merriam collection from Boulder County, secured by the late 
Mr. Denis Gale September 16, 1889. Miller records three specimens 
from Larimer County. 3 Warren quotes the statement of Mr. A. E. 
Beardsley that this bat is "frequent at Greeley/' 4 and states in a 
recent letter that it has been taken at Salida and Boulder. 

Nycteris borealis (Muller). Red Bat. 

Biological Survey collectors have not met with this species in 
Colorado, although it should be found over the eastern plains region. 
The only record seems to be that of Mr. A. E. Beardsley, who says 
the red bat is rare at Greeley. 5 

1 The Mammals of Colorado, p. 277, 1910. * Mammals of Colorado, p. 268, 1906. 

3 Allen, Bull. Am. Mus. Nat. Hist., V, p. .83, 1893. 5 Warren, Mammals of Colorado, p. 268, 1906. 

» N. Am. Fauna No. 13, p. 114, 1897. 



212 NORTH AMERICAN FAUNA. [No. 33. 

PRINCIPAL TREES AND SHRUBS OF COLORADO. 

The explorations of the Biological Survey in Colorado have resulted 
in the accumulation of valuable notes on the distribution of many 
species of plants, and those relating to the trees and shrubs are brought 
together in the following briefly annotated list. No attempt is made 
to give a complete list of Colorado trees and shrubs, and the species 
included are chiefly those whose known ranges in the State have been 
considerably extended by the work of the Biological Survey. Several 
of those r.pporcntly have not been recorded previously from the 
State. Most of the specimens collected have been identified by the 
botanists of the United States National Museum. 

The nomenclature followed is mainly that of Dr. Rydberg in his 
Flora of Colorado (Bull. 100, Colo. Agr. Exper. Station, 1906). Dis- 
tribution notes relative to a number of species of restricted range are 
taken from this publication and, to avoid repetition, are unaccom- 
panied by references. 

Pinus aristata. Foxtail Pine. 

The foxtail pine is a characteristic timberline tree on parts of the 
Front, Kenosha, Saguache, and Sangre de Cristo Ranges, but was not 
observed on the mountains of western or extreme northern Colorado. 
On the Sangre de Cristo Range the southern limit appears to be near 
Crestone Peak, but I again encountered it on the high Culebra Range 
southwest of La Veta. On the Saguache Range it was found south 
to a point northwest of Villa Grove, and scattering trees were observed 
at timberline on the San Juan Mountains northeast of Pagosa Springs. 
None of these pines were seen on the Front Range north of James 
Peak. Foxtail pines are most abundant on the Saguache Range near 
St. Elmo and in the Grays Peak region. A straggling and dwarfed 
growth fringes the exposed ridges on both sides of Clear Creek Valley 
above Silver Plume, between 11,000 and 11,500 feet. 

Although largely confined to the timberline region between 11,000 
and 12,000 feet, and perhaps more nearly restricted to the Hudsonian 
zone than any other tree in the State, the foxtail pine is occasionally 
found on bare exposed ridges as low as 9,500 feet. Near Como it is 
common on many of the ridges bordering South Park at 10,000 feet. 
Between Clyde Station and Cheyenne Mountain on the Cripple Creek 
Short Line Railway the species forms a considerable forest. At this 
point the pines are very regular in shape, few are branched near the 
base, and many are 30 feet high. In the timberline region the trees 
are usually under 15 feet in height, dwarfed and ragged, and the 
majority are much branched. (See PL XI, fig. 1.) At a distance 
they greatly resemble a ragged growth of juniper (Juniperus mono- 
sperma), and are in marked contrast to the more symmetrical Engel- 
mann spruces, which usually share the bleak timberline slopes with 



1911.] TREES AND SHRUBS. 213 

them. Foxtail pines usually grow on slopes having warm south or 
west exposures and very rarely on cool slopes. This species forms the 
highest recorded timberline in the State (12,300 feet), near St. Elmo, 
on the Saguache Range. 

Pinus flexilis. Rocky Mountain White Pine. 

This small, bushy, much branched phie has a general but very scat- 
tering distribution in the mountains of northern Colorado, at eleva- 
tions between* 7,000 and 10,000 feet. Like the foxtail pine of the 
timberline regions, which it greatly resembles, P. flexilis usually grows 
along the crest of bare, outlying, gravelly ridges, and owing to its 
wind-swept location usually grows straggling and one-sided. It is 
most abundant on the open mountain slopes east of Fall River, Clear 
Creek County, on the ridges of the South Park region, and on the 
mountains east of Laramie River, but was noted at other localities as 
follows: Mclntyre Creek, east slope of Medicine Bow Range; outly- 
ing ridges north of Higho and east of Canadian Creek, North Park; 
Halms Peak (8,500 feet) ; bluffs north of Snake River, 8 miles east of 
Slater; Grand River Canyon, east of Glen wood Springs; Empire; 
canyon on Grand River, west of Hot Sulphur; and valley of South 
Boulder Creek. The species does not attain its maximum size in 
northern Colorado, and no trees more than 30 feet high were observed. 

P. flexilis was found at but few localities in the southern mountains 
in 1907. Groves of considerable extent are on the partially open 
slopes at the head of Wahatoye Creek, between the Spanish Peaks 
and the Culebra Range, at about 9,000 feet elevation, and the species 
is common throughout the La Veta region. A much larger growth is 
attained here than farther north, many of the trees reaching a height 
of 50 feet. This pine is not uncommon at Pagosa Springs, between 
Needleton and Silverton, on the Rio Grande bluffs below Wagon Wheel 
Gap, and at Divide. 

Pinus scopulorum. Rocky Mountain Yellow Pine. 

The yellow pine is a characteristic tree in the foothills of all parts 
except northwestern Colorado. The eastern foothills of the Medicine 
Bow and Front Ranges are clothed with a scattering growth from 
their bases to the lower edge of the lodgepole pines at 8,000 or 9,000 
feet, the heaviest forests being west of Loveland, on the western end 
of the Arkansas Divide, and in the South Platte region. Farther 
south there is a good stand on the Wet Mountains, becoming heaviest 
in northwestern Huerfano County artd continuing southward in a 
well-defined belt along the eastern foothills of the Sangre de Cristo 
and Culebra Ranges. On the lower mountain slopes bordering San 
Luis Valley the pines are restricted to narrow interrupted belts, which 
converge at its northern end and connect over Poncha Pass with the 
pine country along the Upper Arkansas. 



214 



NORTH AMERICAN FAUNA. 



[No. 33. 



West of the Continental Divide, P. scopulorum is largely restricted 
to the region south of the Grand and Gunnison Rivers. Stately 
forests of great extent occupy the southern and western slopes of the 
San Juan, La Plata, and San Miguel Mountains, and the eastern slopes 
of the La Sal Mountains, ranging between 6,500 and 8,500 feet, and a 
moderately heavy growth extends to the northwestern end of the 
Uncompahgre Plateau. In the Gunnison country yellow pines are 
very local and scattering, except in the valley of the Lal^e Fork, where 
they are tolerably common. In northwestern Colorado they are 
rarely seen, and occur in widely separated areas of small extent as 
follows: Medicine Bow Mountains, east of Canadian Creek, North 




Fig. 30.— Distribution in Colorado of lodgepole pine {Pinus murrayana). 

Park, lowest outlying spurs, 8,800 feet; north slope of Elk Head 
Mountains, 20 miles southeast of Slater, rocky exposed ridges, 7,800 
feet; Grand River Canyon to 6 miles west of Glenwood Springs; 
southwest slope 8 miles north of McCoy, Eagle County, 7,800 feet; 
and on the summits and northern slopes of the Escalante Hills, in 
western Routt County, where there is a tolerably heavy growth 
between 6,400 and 7,500 feet. 

In the eastern and northern mountains the yellow pines seldom 
attain large size, and in rocky situations they are often quite scrubby. 
In the southwest, however, particularly in Archuleta County, there 
are as large trees and handsome forests as can be.found in any of the 
Rockv Mountain States. Lumbermen have exhausted some of the 



1911.] 



TREES AND SHRUBS. 



215 



best pine forests of southwestern Colorado, but there still remain 
large areas, and fortunately these are now largely included in the San 
Juan and Montezuma National Forests. 

Pinus murrayana. Lodgepole Pine. 

A broad belt of magnificent lodgepole pine forest with a vertical 
breadth averaging 2,000 feet occupies the middle slopes of the high 
mountain ranges of northern Colorado between 8,000 and 10,000 feet. 
(See fig. 30.) This forest clothes the summits of the Laramie Divide 
and of the lower parts of the Medicine Bow, Park, and Gore Ranges. 
Farther south this pine covers an extensive area on the Saguache 
Range and on the mountains of the Gunnison country. The southern 
limit is reached on the Continental Divide at the head of the Saguache 
River. On the Sangre de Cristo Range a narrow belt extends south 
along the eastern 
slope as far as Cres- 
tone Peak, but the 
species appears to be 
entirely absent from 
the western slope of 
the same range. A 
few of these pines are 
found on the summit 
of Veta Pass — the 
southernmost point 
at which the species 
is known to occur 
within the State. 

The heaviest and 
purest forests of 
lodgepole pines were 
traversed on the high 
divide east of Lara- 
mie River and on the Park and Gore Ranges, at an elevation of about 
9,500 feet. (See fig. 31.) On most of the ranges, however, the pines 
are mixed with aspens below 9,500 feet and with Engelmann spruces 
above 10,000 feet. 

To the west P. murrayana is the common forest tree in the Elk 
Head Mountains and in the region bordering Egeria Park, but on the 
White River Plateau it does not extend far west of the South Fork of 
White River. A scattering growth is reported on Diamond Peak 
and Mount Cullom, on opposite sides of Green River in extreme 
northwestern Routt County. The lowest elevation reached within 
the State appears to be on the northern slope of the Elk Head Moun- 
tains near Honnold, where narrow tongues descend to the bank of 




Fig. 31.— Forest of lodgepole pines (Pinus murrayana), eastern slope of 
Park Range, west of Pearl, at 10,000 feet. 



216 NORTH AMERICAN FAUNA. [No.:::;. 

Snake River at 7,000 feet. The extreme upper limit on the Boreas 
Pass is on a southwest slope at a little over 1 1 ,000 feet. This pine 
appears to be absent from the mountains of southwestern Colorado. 
The lodgepole pine is very largely restricted to the Canadian zone. 
In the mountain districts of central Colorado much of the best growth 
has been used for mining timbers, and farther north lumbering opera- 
tions have depleted large tracts. At present a young and dense 
growth covers much of these areas, which are practically all included 
in National Forests. 

Pinus edulis. Pinyon; Nut Pine. 

The pinyon is found in abundance in the rough country of western 
and southern Colorado at elevations varying from 5,000 to 7,500 
feet. It clothes the rocky slopes and bluffs bordering the river 
valleys and forms a dense growth on most of the rough intervening 
watersheds and mesas. In the mountains of the San Luis Valley 
region the pinyon belt reaches an elevation of 9,000 feet on warm 
slopes. 

This small pine is fully as common in the Upper Sonoran zone as 
Juniperus monosperma, and marks this zone over extensive areas. 
In many localities, however, it enters the lower edge of the Transition 
zone, where it commingles to a certain extent with yellow pines and 
Douglas spruces. It is nearly always associated with the juniper 
{Juniperus monosperma) below 7,000 feet. 

A very dense growth of pinyon covers the Mesa Verde and prac- 
tically all the broken country from Montezuma County north to 
Mesa County. A heavy and continuous belt is found on the lowest 
flanks of all the mountains from the San Juans north to the Book 
Cliffs, and in the Grand River Valley the species extends eastward 
as far as McCoy, Eagle County. North of the Book Cliffs the distri- 
bution is more restricted, as follows: Evacuation Creek Valley up 
to 7,000 feet on north slope of the Book Cliffs; heavy growth in 
Pinyon Valley and on the Rabbit Hills; divide between Bear and 
White Rivers, south of Lily, to 6,000 feet; Escalante Hills, dense 
growth on north slopes at 6,400 to 7,000 feet; scattering growth 
southwest slope of Cross Mountain, and also on south slope of O-wi- 
yu-kuts Plateau. Pinyons are found on- most of the mesas and 
ridges of the San Luis Valley region, and form heavy belts on the 
lower bordering mountain slopes north as far as Villa Grove. On 
the eastern slope of the mountains they occur regularly north to 
Manitou, and follow up the warm Arkansas Valley to considerably 
above Buena Vista, the growth being especially dense and extensive 
on the head of Huerfano River and in the adjacent country. The 
species reaches its eastern limit in Colorado, and probably in the 
United States, in Las Animas County, where a considerable growth 



1911.] TREES AND SHRUBS. 217 

is found on Mesa de Maya and on other high points in that rough 
region. 

Picea engelmanni. Engelmann Spruce; White Spruce. 

The stately Engelnaann spruce forms a heavy forest belt just 
below timberline on all the higher mountain ranges, and in a dwarfed 
state is common at timberline. At this elevation it occurs in pros- 
trate mats 1 to 3 feet in height and often 8 or 10 feet in diameter. 
Its greatest development is reached along the upper edge of the 
Canadian zone, where the spruce forest is either pure or mixed with 
balsam firs, many of the trees being 3 or 4 feet in diameter and 100 
feet in height. 

The heaviest spruce forests were on the Park, Medicine Bow, and 
Saguache Ranges, in the San Juan Mountains, and on both slopes 
of the Front Range from Rollins Pass south to Berthoud Pass. A 
moderate growth of Engelmann spruce is found on the Uncompahgre 
and White River Plateaus, on Grand Mesa, and on other high plateaus 
of western Colorado. 

The upper limit of the Engelmann spruce varies with timberline 
from 11,000 to 12,000 feet. The vertical width of the belt depends 
much upon the steepness of slope, averaging between 1,000 and 1,500 
feet on gradual slopes, but often narrowing to 500 feet, as is the 
case just below timberline on a steep southwest exposure south of 
Berthoud Pass. Below 10,000 feet P. engelmanni occurs only in 
damp situations on cold slopes and in descending tongues along 
streams, usually embraced by heavy forests of lodgepole pine or 
aspen. It is found as low as 8,200 feet along Pass Creek, on the 
eastern slope of the Park Range; while on the mountains east of 
Laramie River it occurs at 8,500 feet. Scattering trees occur on 
Middle Boulder Creek, 5 miles west of Boulder, at 6,000 feet, but 
this low elevation is abnormal. 

Picea parryana. Blue Spruce. 

The blue spruce has a very scattering distribution in the lower 
part of the Canadian zone on both slopes of the main ranges. In 
northern Colorado in 1905 and 1906 small clumps and single trees were 
noted here and there along streams at elevations of from 7,000 to 8,500 
feet, as follows: West of Log Cabin, Larimer County, 7,500 feet; 
Nederland, 8,200 feet; South Boulder Creek; northeast base of Floyd 
Hill; Empire; Idaho Springs; Fall River, Clear Creek County; south 
slope of Park Range; north of Hahns Peak, 8,500 feet; Snake River 
bluffs; Honnold to 8 miles east of Slater; north slope of Piney Divide, 
south of McCoy, 7,500 to 8,000 feet; Pass Creek, northwest of Kremm- 
ling; Eagle River, Dotsero to Wolcott; and canyon of the Grand east 
of Glenwood Springs. In the mountains of southern Colorado the 



218 NORTH AMERICAN FAUNA. [No. 33. 

blue spruce has a general distribution between 8,500 and 9,500 feet, 
being much more common than farther north. 

This species is apparently never found away from the immediate 
vicinity of streams, 1 all the trees that I have observed growing either 
on the banks of watercourses or on the nearest benches. It proba- 
bly belongs to the Canadian zone, but in common with many other 
boreal species finds a suitable environment in the cool conditions 
which obtain along streams well down into the Transition zone. 
The characteristic scattering distribution of the blue spruce along 
watercourses is best seen along Eagle River between Wolcott and 
Dotsero, on the Frying Pan River, and on the many streams of Clear 
Creek County. 

Pseudotsuga mucronata. Douglas Spruce; Red Fir. 

The Douglas spruce is of comparatively small size in Colorado. 
The largest trees are found along watercourses, where a height of 1 00 
feet and a diameter of 3 or 4 feet are occasionally attained. The 
average growth, however, is less than a third of these dimensions, 
especially among the eastern foothills. 

The species has a general distribution in the mountain districts, 
being most abundant on the eastern slopes of the Medicine Bow and 
Front Ranges and on the plateaus and higher mesas of western Colo- 
rado. It is mainly a Transition zone tree, but occurs commonly in 
the lower Canadian zone also, particularly in the western mountains. 

On the eastern slopes of the main ranges the Douglas spruce is 
occasionally found along streams as low as 5,800 feet. In the upper 
part of the yellow pine belt it occupies most of the steep north slopes 
between 6,000 and 8,000 feet. Near Fall River, in Clear Creek 
County, it extends up southwest slopes to the 9,000-foot summits 
of hills, whose northeast exposures are clothed with lodgepole pines 
and aspens. The Douglas spruce thus occupies a position interme- 
diate between the yellow and lodgepole pine belts east of the Conti- 
nental Divide. Over most of western Colorado it is found just above 
the pinyon belt, largely replacing the yellow pine in the Transition 
zone, and with the aspen clothing the Canadian zone summits and 
upper northern slopes of the plateaus and higher mesas. (See fig. 
32.) On the crest of the Uncompahgre Plateau and elsewhere it 
grows at an elevation of over 10,000 feet. 

The following localities indicate the distribution of Pseudotsuga 
in northwestern Colorado: North slope of Piney Divide, south of 
McCoy, 7,000 to 8,000 feet; south slopes of mountains north of 
McCoy, 8,000 to 8,500 feet; mountains south of Eagle; canyon of the 
Grand above Glenwood Springs; Battlement Mesa; Great Hog Back; 

1 At Saguache and elsewhere these handsome spruces grace many dooryards, where they have been 
planted. As an ornamental tree the blue spruce has few equals. 



1911.] 



TREES AND SHRUBS. 



219 



southern slopes of White River Plateau; Grand Mesa; high country 
on both sides of Grand River between Glenwood Springs and Grand 
Junction, at elevations ranging from above 6,000 feet on cold slopes 
to 7,000 or 8,000 feet on warm slopes; Book Cliffs, heavy growth on 
north slope near Baxter Pass, 7,000 to 8,500 feet; hills bordering 
Evacuation Creek Valley; sparse growth on northeast upper slopes 
of Rabbit Hills; and Zenobia Peak, scattering growth reported. In 
southwestern Colorado I observed the species as follows: Southern 
slopes of La Plata Mountains between Durango and Mancos in the 
Canadian zone; upper slopes of Ute Peak; steep upper rims of Sinbad 




Fig. 32.— Pocket of Douglas spruce (Pscudotsuga mucronata) on northern escarpment of Mesa Verde, 

7,500 to 8,000 feet. 

and West Paradox Valleys (cool exposures); Unaweep Canyon; 
Cimarron; Vernal Mesa; West Elk Mountains east of Crawford; 
Roaring Fork of Grand River, Basalt to Aspen; Sapinero; and Lake 
City. 

Abies lasiocarpa. Balsam Fir. 

The balsam fir is found throughout the Hudsonian zone and 
along the upper edge of the Canadian zone. It is not so uniformly 
distributed as the Engelmann spruce and does not form as extensive 
forests. Occasionally, however, its growth is heavy, as on the west- 
ern slope of the Front Range in the vicinity of Rollins Pass and on 
the west side of the Saguache Range below Alpine Tunnel, where 
extensive forests were traversed between 10,000 and 11,000 feet. 



220 NORTH AMERICAN FAUNA. [No.33. 

In a dwarfed state this fir often extends up to extreme timberline 
with the Engelmann spruce. Its characteristic growth is in clumps 
and thickets scattered here and there through the spruce forest, 
and the smooth, light-colored trunks of the firs are very conspicuous 
among the dark spruces. Occasionally the species is encountered 
as low as 8,500 feet, and on the eastern slope of the Park Range 
northwest of Kremmling I found it growing along streams at 8,200 
feet. It formed heavy thickets at 10,000 feet on the divide east of 
the Laramie River, and also on the Medicine Bow Range; and on 
thi 1 northern part of the Park Range it was common both on the 
Buffalo Pass and on the headwaters of Grand Encampment River. 
On the White River Plateau it was growing in dense thickets in the 
aspen forests between 8,500 and 9,000 feet, chiefly on northern 
exposures. 

This tree appears to have fully as wide a range in southern as in 
northern Colorado, but the growth is smaller and more scattered 
toward the south. The species was observed in southern Colorado 
on t;ffe summit of the Uncompahgre Plateau, at Rico and Ophir, 
fromS-Needleton to Silverton, and between Vance Junction and 
Sawpit. 

Abies concolor. White Fir. 

The white fir is found in the southern mountains, where it is 
usually common between 8,500 and 10,000 feet, in the lower Canadian 
zone. It does not occur far north of Colorado Springs on the eastern 
slope, and Ouray marks the northern limit west of the Continental 
Divide. The white fir forms considerable forests in the Wet Moun- 
tains and on the southern slopes of the San Juan Mountains north 
of Pagosa Springs, but usually the growth is somewhat scattering, 
and is restricted either to the vicinity of streams or to cool exposures, 
where the firs mingle with aspens. The largest trees observed were 
at 9,000 feet on the southwest slope of Pagosa Peak, where a height 
of 75 feet was not uncommon. 

The white fir was noted at the following localities: Wet Mountains, 
east of Westcliffe, 9,000 feet; mountains between Canon City and 
Cripple Creek, above 8,500 feet; northern exposures on eastern slope 
of Sangre de Cristo Range near Sand Hill Pass, 9,000 to 9,500 feet; 
head of Wahatoye Creek, 9,000 feet; upper northern slope of Fisher 
Peak, south of Trinidad; mountains near Saguache; Animas Canyon, 
near Silverton; and lower mountain slopes surrounding Ouray. 

Juniperus scopulorum. Rocky Mountain Juniper. 

This juniper is common in the Upper Sonoran and Transition 
zones throughout the mountains, and also often forms the only 
coniferous growth in gulches and on rocky ridges and buttes on the 
higher plains of northeastern Colorado. It has an extreme vertical 



1911.] TREES AND SHRUBS. 221 

range within the State from 4,500 to 10,000 feet, and is most abun- 
dant in the higher foothills between 6,000 and 8,000 feet. I observed 
it at the following localities: Boulder; Golden; Platte Canyon Sta- 
tion; Pawnee Buttes; Bailey; Colorado Springs; La Veta; hills 
north of Canon City; Promontory Bluffs; bluffs east of Grover; 
Buena Vista; St. Elmo; Weston; Chimney Cliffs, northwestern 
Logan Count}'; Book Plateau, sparingly above 7,500 feet; Glen- 
wood Springs; Wolcott; McCW; Ohio City; Gunnison; Sapinero; 
Lake City; Lone Cone; Beaver Mountain, Dolores County; Man- 
cos; Mesa Verde, north escarpment; and Pagosa Springs. 

Juniperus monosperma. Juniper. 

This is the most abundant juniper in the State, and is a charac- 
teristic Upper Sonoran species. It forms a well-defined belt cover- 
ing most of the lowest foothill slopes of western and southern Colo- 
rado and is prominent on the eastern foothills from the Arkansas 
Valley southward. A heavy grow T th covers a large area of rough 
canyon country in Las Animas and western Baca Counties, and 
many outlying ridges «and buttes in Otero County are clothed with 
junipers. The species reaches a large size and dense stand on the 
Escalante Hills and other low elevations of western Routt County, 
and is abundant on Mesa Verde, in Montezuma County. 

The vertical position of J. monos'perma is immediately below the 
pinyon belt, although scattering junipers occur with the pinyons as 
high as 7,500 feet in southern Colorado. It is most abundant 
between 5,000 and 7,000 feet on arid slopes. The resinous one- 
seeded berries of this juniper are much used as food by chipmunks, 
wood rats, and other small rodents. 

During my explorations in Colorado I have observed the species 
as follows: Slater; bluffs bordering lower Snake River Valley; 
Godiva Ridge; Cross Mountain; Escalante Hills; O-wi-yu-kuts 
Plateau, southern slopes; Vermilion Bluffs; watershed between Bear 
and White Rivers south of Lily; 5 miles southwest of Rangely to 
base of Rabbit Hills; northern and southern slopes of Book Plateau 
below 7,500 feet; hills south of Mack, Mesa County; lower slopes of 
Grand Mesa; northeast slope of Little Book Cliffs; De Beque to 
Glen wood Springs; bluffs along north side of Eagle River from 
Dotsero nearly to Wolcott; McCoy; Basalt; Somerset; Ridgway; 
Coventry; Placerville; western Montrose and San Miguel Counties; 
Mesa Verde; McElmo Valley; Salida; Canon City; Walsenburg; 
Pueblo: and Gaume's ranch, northwestern Baca County. 

Juniperus sibirica. Low Mountain Juniper. 

This beautiful procumbent shrub is conspicuous among the under- 
growth of the forests hi the Canadian and upper Transition zones, 
where it is almost omnipresent. It reaches its greatest abundance 



222 NORTH AMERICAN FAUNA. [Ho. 33. 

along the upper edge of the yellow pine belt at 8,500 or 9,000 feet. 
I have found it particularly common in the higher eastern foothills of 
the Front Range; on the forested ridges near Como, South Park; in 
the Wet Mountains east of Westcliffe; on the Sangre de Cristo Range 
in western Huerfano County; and at St. Elmo, Saguache Range, 
10,000 feet. 

Juniperus prostrata. Creeping Juniper. 

The creeping juniper or savin appears to be uncommon in the 
Colorado mountains. I collected it on Lone Mesa, Dolores County, 
at 9,200 feet, June 26, 1907; and saw dense patches of it on open 
rocky ridges along Pass Creek, northwest of Kremmling, Middle 
Park, in October, 1906. Rydberg records it from North Cheyenne 
Canyon, Parlin, and Owl Canyon. 

Ephedra antisyphylitica. Joint Fir. 

The joint fir is a characteristic shrub on the warm rocky Upper 
Sonoran slopes of western Colorado, extending eastward with the 
junipers and pinyons for some distance into the mountains along the 
warm north sides of the river valleys. E. antisyphylitica is the com- 
mon species of widest range, but two other forms are present in the 
southwestern valleys. 

I have observed this joint fir at the following localities: Escalante 
Hills, 7,000 feet; White River bluffs east of Rangely; pinyon country 
southwest of Rangely; juniper-covered hills at north base of the Book 
Cliffs; Glen wood Springs; Basalt; Mesa Verde; McElmo; lower San 
Miguel and Dolores River regions (omnipresent below 7,000 feet) ; 
Placerville; and Coventry. 

Ephedra torreyana. 

I have taken tins species only on the warm juniper slopes along the 
north side of North Gunnison River at Somerset, 6,000 feet. It is 
recorded by Rydberg from Deer Run, Mesa County. 

Yucca glauca. Yucca. 

This yucca is one of the most characteristic Upper Sonoran plants 
in the State, being almost omnipresent below 6,000 feet, and extend- 
ing often to 8,000 feet on exceptionally warm slopes in the foothills. 
It occurs in abundance on both sides of the mountains (see fig. 33), 
but the densest growth is on the eastern plains, particularly in the 
southeastern counties. The tall spikes of greenish white flowers are 
very prominent on the plains during June. 

Following are some of the localities at which Yucca glauca was 
observed on my trips over the State: Gaume's ranch, northwest 
Baca County; Limon; Cheyenne Wells; Tuttle; Wray; Sterling; 
Pawnee Buttes; Grover; east of Boulder; near Fort Collins; Bailey; 
Gardner; La Veta; Buena Vista; Wet Mountain Valley; Poncha Pass, 



1911.] 



TREES AND SHRUBS. 



223 



summit; Salida; Cascade; Pueblo to Walsenburg; Saguache and San 
Luis Valleys generally; Gunnison; Hotchkiss; Montrose region; 
Placerville; McElmo Valley; lower Dolores River region up to 7,000 
feet; Unaweep Canyon; Rifle; De Beque; Plateau Creek; and hills 
between Carbonera and Mack. 

Yucca harrimanise. Harriman Yucca. 

Locally common at several points in the juniper belt of southwestern 
Colorado below 7,000 feet. I found it in 1907 near Ridgway; on the 
slopes of Cerro Ridge east of Montrose; and on the rocky slopes along 




Fig. 33. — Distribution in Colorado of common yucca ( Yucca glauca). 

the west side of Sinbad Valley. Rydberg records the species from 
Cimarron and Durango. 

Yucca baccata. Spanish Bayonet. 

This large-leaved yucca is a characteristic Upper Sonoran plant in 
parts of southwestern Colorado, growing chiefly among rocks on 
warm juniper slopes and in the lowest valleys. It was flowering on 
Mesa Verde at 7,000 feet June 13, 1907, the spikes of large greenish 
white flowers dotting the rocky rims of Navajo Canyon just above 
the Spruce Tree Cliff Ruins. (See fig. 34.) I observed this yucca at 
Arboles, Bayfield; McElmo; Coventry; on slopes bordering Paradox 
and Sinbad Valleys; and along the canyon of Dolores River between 
Salt Canyon and the mouth of West Creek. There appears to be 



224 



NORTH AMERICAN FAUNA. 



[No. 33. 



only one Colorado record for Y. baccata east of the mountains — 
Rydberg recording it from Trinidad. 

Populus tremuloides. Aspen; Poplar. 

Extensive aspen forests clothe the summits and northern slopes 
of nearly all the western plateaus and mountains between 8,000 and 
10,000 feet, but on the eastern slope of the main ranges the trees are 
usually small, and form dense thickets rather than open forests. 
This boreal poplar is restricted to the Canadian zone, the center of 
abundance being at about 9,000 feet. Small thickets sometimes 
occur on cold northeast slopes as low as 7,000 feet, while on warm 
southern slopes a dwarfed growth usually extends to at least 10,500 




Fig. 34. — Yucca baccata in flower, Navajo Canyon, Mesa Verde. 

feet. The best aspen forests were encountered on the White River, 
Book, and Uncompahgre Plateaus, and on the north slope of the 
Rabbit Ear Mountains near Arapahoe Pass, where trees fully 2 feet 
in diameter and 50 feet in height were not at all uncommon. On the 
crest of the Uncompahgre Plateau, near its northern end, beautiful 
aspen groves alternate with reaches of open grassy country, the 
coniferous element so common elsewhere being very largely absent. 
The aspen forest is often tolerably clear of large undergrowth, but 
is mixed here and there with thickets of balsam firs or lodgepole 
pines. Throughout the mountains are areas which formerly sup- 
ported coniferous forests, but have been devastated by forest fires. 



1911.] TREES AND SHRUBS. 225 

These are now mostly covered with a young growth of aspens, which 
seem to secure a foothold much more quickly than any of the conifers. 

Populus angustifolia. Narrow-leaf Cottonwood. 

There is scarcely a foothill stream in the State which is not fringed 
with more or less of these cotton woods. It is a characteristic Transi- 
tion zone species, with a vertical range from 5,500 to about 9,000 feet. 
Its vertical position is intermediate between the aspen of the Canadian 
zone and the broad-leaved cottonwoods of the Upper Sonoran zone. 
It attains its largest size along the lower edge of its range at the base 
of the foothills, where it often mingles for a short distance with the 
broad-leaved species. Scattered clumps of large size are found the 
entire length of the Snake River Valley, in northwestern Colorado. 

Populus occidentalis. Broad-leaf Cottonwood. 

This is the cottonwood which fringes most of the streams on the 
eastern plains, the heaviest and largest growth being found along 
the Platte and Arkansas Rivers and on many of the streams near 
the base of the foothills from Denver north to Fort Collins. The 
growth is particularly large in the Arkansas Valley between Las 
Animas and the Kansas boundary. It is found on only a few streams 
in the foothills, but near Livermore, Larimer County, I found it up 
to 6,000 feet along Lone Pine Creek, and a single tree was growing 
on Middle Boulder Creek at 5,800 feet. On the dry plains of Baca 
County the beds of the streams, few of which are perennial, are 
fringed by a gnarled and stunted growth of cottonwoods. 

Populus wislizeni. Southwestern Cottonwood. 

This is the predominant broad-leaved cottonwood of southwestern 
Colorado, where it occurs along most of the watercourses below 6,000 
feet. I found it common at Arboles; Grand Junction; along McElmo 
Creek; on San Miguel River below Naturita; Dolores River between 
Paradox Valley and the mouth of West Creek; Smith Fork below 
Crawford; and North Gunnison River to 5 miles below Somerset; 
and it was probably the species growing along the Rio Grande at 
Alamosa. East of the mountains it is found on the plains at Colorado 
Springs, according to Rydberg. 

Populus acuminata. Smooth-bark Cottonwood. 

During my explorations in Colorado I observed this cottonwood 
only in the Upper Sonoran stream valleys of the northwestern 
counties from the Wyoming boundary south to Mesa County, as 
follows: Scattering fringe along Snake River for 8 miles above Lily; 
dense growth of large size on Green River from Browns Park to 
northern end of Ladore Canyon; lower Vermilion Creek, western 
Routt County; scattering fringe along White River from Utah 
90432°— 11 15 



226 NORTH AMERICAN FAUNA. [No. 33. 

boundary east to Angora; Plateau Creek, Mesa County; and Dolores 
lliver, at the mouth of West Creek (growing with P. wislizeni). 
This cottonwood occurs east of the mountains, as Rydberg records 
it from Fort Collins, Denver, and Walsenburg. 

Salix amygdaloides. Peach-leaved Willow. 

A common willow along streams in the Upper Sonoran zone on 
both sides of the mountains. Observed at Boulder; Livermore; on 
Green River near Ladore; and in the lower valleys of Snake and 
White Rivers. 

Salix perrostrata. 

Taken in June, 1905, in the foothills near Golden, where small 
clumps were growing in bogs on the higher slopes between 6,500 
and 7,500 feet in the upper Transition zone. 

Salix nuttalli. Nuttall Willow. 

Scattered clumps were growing with S. perrostrata, between 6,500 
and 7,500 feet, around springs and in bogs among the foothills near 
Golden. Willows noted in September, 1906, at springs on the north- 
ern slope of the Book Plateau, between 7,500 and 8,000 feet, were 
probably S. nuttalli. 
Salix geyeriana. 

Dense thickets of this low willow fringe most of the streams and 
bogs in Middle and Egeria Parks and near Halms Peak. It was 
the prevailing species near Coulter in Middle Park and at the eastern 
end of the Elk Head Mountains near Columbine. It was not observed 
in the mountains of southern Colorado. 

Salix glaucops. 

Tins is perhaps the most abundant alpine willow in the mountains. 
(See fig. 35.) In June, 1905, I found dense copses on the higher 
slopes of Mount Kelso between 11,000 and 12,500 feet, and in 1906 it 
was common on the streams of Egeria Park and the Gore Range. 
The usual growth of tins willow varies from about 4 feet in the lower 
part of its range in the Canadian zone to 2 feet or even less above 
timberline. The glabrous-leaved form (S. g. glabrata) is common 
in bogs on the alpine slopes of Mount Kelso. 

Salix chlorophylla. 

I have found tins alpine willow only on Mount Kelso, where it 
grows on the boggy slopes between 11,000 and 12,500 feet in dense 
copses from 2 to 4 feet high. It was more abundant above than 
below timberline. 

Betula fontinalis. Rocky Mountain Birch. 

This handsome birch forms a conspicuous fringe along foothill 
streams throughout tile State, and often extends out on the plains 



TREES AND SHRUBS. 



227 



for some distance. It is usually associated with narrow-leaved 
cottonwoods and willows, and I have seen it only near stream banks. 
In July, 1907, I found several large clumps fully 20 feet in height on 
Dolores River near the mouth of West Creek, at 5,000 feet. The 
species occurs regularly up to 8,000 and occasionally to 9,000 feet 
through the entire width of the Transition zon'e. I did not observe 
it in the Canadian zone. 

I found B. fontinalis common at the following localities : Streams 
of Routt County; North Park; Plateau Creek; Bailey; Lake George; 
St. Elmo; Frying Pan River, Basalt to Peachblow; head of Smith 
Fork, West Elk Mountains; Unaweep Canyon; Placerville; Durango; 
Rico; Mancos; Bay- 
field ; and La Veta. 

Betula glandulosa. 
Dwarf Birch. 

The habitat of the 
dwarf birch is along 
the borders of cold 
bogs and streams in 
the higher moun- 
tains, .between 9,000 
and 11,000 feet, in 
the upper Canadian 
and lower Hudson- 
ian zones. It has 
been found on prac- 
tically all the higher 
mountain ranges of 
the State. The nor- 
mal growth is from 
3 to 5 feet high, but 
in the Hudsonian 
zone at about 11,000 
feet the species is 
dwarfed, rarely exceeding 2 feet. It was particularly abundant in 
1906 in the mountain meadows on the Park Range and on the 
mountains east of Laramie River, where a low, dense growth fringed 
most of the bogs between 9,000 and 10,000 feet. I saw it on Grand 
River, 5 miles east of Hot Sulphur; on the San Juan Mountains near 
Ophir, at 10,500 feet; and southeast of Lake City, 9,000 to 11,000 
feet. The leaves were falling near Lake City, October 18, 1907. 

Alnus tenuifolia. Alder. 

Alders form a dense fringe along cold streams throughout the 
Colorado mountains. Thev are most abundant in the Canadian 




Fig. 35. — Alpine willows in Arctic-Alpine zone, Front Range, near Ber- 
thond Pass. 



228 NORTH AMERICAN FAUNA. [No. 33. 

zone up to 10,000 feet, but occur regularly as low as 7,000 feet and 
occasionally to 6,000 feet. The species attains its maximum size 
along the smaller streams flowing from the Medicine Bow Mountains 
into North Park, a tributary of Canadian Creek being fringed with 
alders fully 20 feet in height. 

The following localities at which I have noted alders indicate a 
wide and uniform distribution: Log Cabin, Larimer County; North 
Park; Halms Peak; Snake River, Honnold to 10 miles above Slater; 
Gore Mountains; Fraser River, Middle Park; Smith Fork of Gunni- 
son River, West Elk Mountains; Frying Pan River; Gunnison; 
Crested Butte; Pitkin; vSapinero, Lake Fork of Gunnison River; 
Wagon Wheel Gap; Poncha Pass; Creede; Manitou; Bailey, South 
Platte River; Lake George; Buena Vista; St. Elmo; Lone Cone; 
Unaweep Canyon; Rico; Silverton; Bayfield; Pagosa Springs; Pagosa 
Junction; La Veta; and Las Animas River, Segundo to Weston. 

Corylus rostrata. Beaked Hazelnut. 

Small thickets of hazel are common on Middle Boulder Creek, at 
Blanchard's ranch, 5 miles west of Boulder, at an elevation of nearly 
6,000 feet. I did not observe the beaked hazel elsewhere in Colorado, 
but Rydberg records it from other points in the eastern foothills 
from North Cheyenne Canyon north to Larimer County. 

Celtis reticulata. Hackberry. 

The hackberry has a scattering distribution in the Upper Sonoran 
zone on both sides of the mountains, being most common on the 
eastern plains in gulches leading back from stream valleys. Small 
clumps are found in most of the canyons of the lowest eastern foot- 
hills from Boulder north to Fort Collins. I have not found it above 
5,500 or 6,000 feet. Its growth in Colorado is uniformly scrubby, 
and only occasionally does it attain the stature of a tree. Hack- 
berries are common on the bare foothill slopes near Platte Canyon 
Station, at Golden, Boulder, Arkins, and at a point 8 miles west of 
Fort Collins. wSeveral large clumps grow in Shell Rock Canyon, north- 
western Baca County. At Wray, Yuma County, the species is very 
common in the gulches leading back from Chief Creek Valley. 

Few hackberries were met with in western Colorado. Scattering 
trees grow on the slopes along Plateau Creek, 5 miles east of Tunnel, 
Mesa County, and others on West Creek, near its junction with Dolores 
River, in southwestern Mesa County. 

Atriplex canescens. Orache; Gray Saltbush. 

This species has a wide range in the Upper Sonoran zone on both 
sides of the mountains, but is most abundant in western and southern 
Colorado. I have not observed it much above 7,000 feet. It is a 
characteristic shrub in the warmer valleys of the eastern foothills 
south of the Arkansas Valley, where it is locally termed "chico brush." 



11)11.] 



TREES AND SHRUBS. 



229 



tffetei 



It often grows to a height of several feet on sandy or alkaline flats in 
the bottoms of valleys. Pocket gophers appear to feed extensively 
upon its leaves, which I have often found in their tunnels. 

It was noted at the following localities : Lily Park ; Midland Basin, 
Routt County; southwest of Rangely; Escalante Hills; desert north 
of Alack; Hotchkiss; Montrose; Paradox Valley; Dolores River, near 
mouth of West Creek; McElmo Valley; Medano Springs ranch, San 
Luis Valley; Salida; Gardner; La Veta; Limon; and 30 miles northwest 
of Sterling. 
Atriplex confertifolia. Round-leaved Saltbush. 

The round-leaved saltbush is a characteristic Upper Sonoran slirub 
of the desert stretches of western Colorado, growing principally upon 
dry alkaline flats, but often on sandy areas. Occasionally it forms 
a dense growth like sagebrush, as on the plains of extreme western 
Routt County, be- 
tween the Escalante 
Hills and Vermilion 
Bluffs. (See PI. Ill, 
fig. 2.) I observed 
the species at locali- 
ties as follows: Lower 
Snake River Valley 
east to Baggs Cross- 
ing ; plains north of 
Escalante Hills; 
Browns Park; Lily 
Park; Midland Basin; 
M a y b e 1 1 ; badlands 
near Rangely; Evac- 
uation Creek Valley; desert north of Mack; southern slopes of Book 
Cliffs to Atchee, 7,000 feet; Plateau Creek; Fruita; De Beque; New- 
castle; Hotchkiss; Montrose; Paradox Valley; and McElmo Valley 
east to north base of Mesa Verde at Point Lookout. This species 
occurs also east of the mountains, as Rydberg records it from Denver 
and Pueblo. 

Atriplex nuttalli. Saltbush. 

This small shrub, often called salt sage, is a characteristic Upper 
Sonoran species, found chiefly on alkaline flats in the desert valleys of 
western Colorado. In some of the dry desert basins it forms the 
principal shrubby vegetation, although it is of small size and usually 
more or less prostrate. It is the most conspicuous shrub in Midland 
Basin, western Routt County; on the desert north of Mack (see fig. 
36); and on the alkaline stretch of country between Hotchkiss and 
the West Elk Mountains. I found it abundant in Browns Park; 




^toui 






.**£•*§ 



" :0„>- 




Fig. 30. — Desert vegetation (Atriplex nuttalli and Sarcobatus vcrmkulatus) 
in lower Grand River Valley, north of Mack, at 4,500 feet. 



230 NORTH AMERICAN FAUNA. I No. 33. 

along Snake River west of Baggs Crossing; southwest of Rangely; 
and in McElmo and Uncompahgre Valleys. 

Grayia spinosa. Common Grayia. 

This low, spiny, mealy shrub is a characteristic plant in the sandy 
and alkaline Upper Sonoran tracts of northwestern Colorado, where it 
grows in profusion in river valleys and on the lowest bordering 
benches up to a little over 6,000 feet. Together with Atriplex con- 
fertifolia, it forms the principal shrubby growth in some of the river 
valleys of western Routt and Rio Blanco Counties (see PL III, fig. 2). 
I found it common in the Snake River Valley from 20 miles west of 
Baggs Crossing to Lily; on the flats between the Escalante Hills 
and Browns Park; in Midland Basin; in White River Valley from 
Angora westward to the Utah boundary; on rocky slopes along Bear 
River at Maybell; on the desert north of Mack, Mesa County; at 
Rifle, Hotchkiss, and Somerset. In the Browns Park region the leaves 
had nearly all fallen by the 1st of September. 

Grayia brandegei. 

The shrubby growth observed on the tops of low mesas along the 
McElmo Valley near Moqui was probably G. brand er/ei, recorded by 
Rydberg from the McElmo Valley. 

Sarcobatus vermiculatus. Greasewood. 

The zonal range of the greasewood is mainly Upper Sonoran. Its 
wide range over*the lower parts of the State is well indicated by the 
following localities at which it was observed during my explorations: 
North Park in alkaline situations ; Snake River Valley to 5 miles east 
of Slater at 6,800 feet; Bear River Valley east to Craig and Steamboat 
Springs; Browns Park (very rank growth, sometimes 12 feet in 
height); Lily Park; Midland Basin; Rangely; Texas Creek; Evacua- 
tion Creek; north of Atchee, southern slope Book Plateau, to 7,500 
feet; desert north of Mack (see fig. 36); Plateau Creek; De Beque; 
Rifle; Dotsero to Eagle; Kremmling and Muddy Creek Valley, Mid- 
dle Park; Hotchkiss; Montrose region; Dallas Creek near Ridgway; 
lower valleys of San Miguel and Dolores Rivers; McElmo Valle} 7 ; 
Montezuma Valley east to north base of Mesa Verde at Point Lookout; 
Bayfield; Salida; Gardner; San Luis Valley generally — Alamosa, 
Saguache, Hooper, Villa Grove, Mosca, and Medano ranch. 

Berberis fendleri. Barberry. 

In Colorado the barberry appears to be confined to dry rocky slopes 
and ridges in the region south and west of the San Juan and La 
Plata Mountains. I found a scattering growth on rocky ridges in 
the yellow pine forest at Pagosa Springs (7,000 feet), and later met 
with the species on dry, open slopes at Durango. Rydberg records 
B. fendleri from Durango, Arboles, Mancos, and Mancos Canyon. 






1911.] TREES AND SHRUBS. 231 

Odostemon aquifolium. Oregon Grape. 

The Oregon grape has a wide distribution in the Colorado moun- 
tains from the foothills to about 10,000 feet. I found it most abun- 
dant on cool, shaded slopes in the upper part of the yellow pine belt, 
where with Arctostaphylos it forms the characteristic lower under- 
growth. The ripe clusters of bluish berries remain on the shrubs for 
some time in the autumn. The Oregon grape is abundant at Boulder; 
Coulter, Middle Park; Floyd Hill; Honnold; northern slopes of Elk 
Head Mountains; northern slopes and summit of Escalante Hills, 
7,000 feet; canyon of the Grand above Glen wood Springs; Unaweep 
Canyon; Cerro Ridge; Somerset; West Elk Mountains, head of Smith 
Fork; Sapinero; Dolores; Mesa Verde; Bayfield; Pagosa Springs; 
Wet Mountains east of Westcliffe; and hills near La Veta. 

Odostemon fremonti. 

In August, 1907, I found this large prickly -leaved shrub abundant 
along the Smith Fork of Gunnison River, at the western base of West 
Elk Mountains, a few miles east of Crawford, Delta County. At this 
point it was growing among junipers and pinyons on the warm rocky 
slopes near the stream. Many of the shrubs were 6 or 8 feet in height. 
Rydberg records 0. fremonti from Smith Fork Canyon, and it is not 
known to occur elsewhere in the State. 

Edwinia americana. 

I have not observed this species west of the Continental Divide, 
but it is a conspicuous shrub on cliffs and rock ledges on the eastern 
slope across the State from north to south, chiefly in the Transition 
zone. It was in full bloom in the foothills west of Boulder June 10, 
1905, and the white-flowered cymes were very handsome. It is par- 
ticularly common in the hills along the South Platte near Bailey, 
and between Manitou and Woodland Park. 

Fendlera rupicola. 

In Colorado this low shrub is confined to the southwestern coun- 
ties, where it is tolerably common on dry slopes and low mesas in the 
Upper Sonoran and lower Transition zones. It was abundant on 
the pinyon slopes along the Los Pinos at Bayfield, La Plata County 
(6,500 feet), and a few of the shrubs were in flower June 5, 1907. It 
was common in July, 1907, on the low benches along Dolores River 
between Salt Canyon and the mouth of West Creek, at 5,500 feet, and 
I noted it for some distance up the valley of West Creek. Ryd- 
berg records the species from Durango, Mancos, Cerro Summit, 
Dolores, Hotchkiss, and Los Pinos. 

Ribes cereum. Red Currant. 

This is a common currant on dry rocky slopes up to at least 10,000 
feet, but is most abundant in the yellow pine belt. Observed at 



232 NORTH AMERICAN FAUNA. [No. 33. 

Boulder; Golden; Pawnee Buttes, northeastern Weld County; 
Bailey; Manitou; Buena Vista; Como; St. Elmo; Thomasville; 
Wagon Wheel Gap; Coulter; Escalante Hills; and Pagosa Junction. 

Ribes wolfi. 

This high-ranging currant is not uncommon in the Canadian and 
Hudsonian zones, growing even on rocky slopes to a short distance 
above timberline. It was not observed at any point below 8,000 feet. 
The black viscid berries were still on the bushes on the summit of the 
Book Plateau September 22, 1906. It is a common shrub on Lone 
Cone, San Miguel Mountains; at Thomasville; and on McClellan 
Mountain between 11,000 and 12,000 feet. 

Ribes longinorum. Golden Currant. 

The zonal range of this flowering currant is mainly Upper Sonoran. 
Its yellow bloom was conspicuous at Boulder June 8, 1905, and along 
the San Juan River at Pagosa Springs May 27, 1907. This currant 
was observed also at Golden and at Wray. Vernon Bailey reports it 
common along streams in the southern end of San Luis Valley. 

Opulaster intermedins. Nine bark. 

In early June, 1905, I found this handsome flowering shrub in full 
bloom on Middle Boulder Creek at 5,800 feet, in the Transition zone. 
Rydberg records it from other points in the eastern foothills of the 
Front Range. 

Opulaster monogynus. 

Common in June, 1905, in the foothills near Boulder and Golden 
between 6,500 and 7,500 feet. This species does not grow along 
streams like 0. intermedins, but on dry rocky slopes. Rydberg gives 
O. monogynus a wide distribution in the eastern and central mountain 
districts. 

Rubacer parviflorus. Salmonberry. 

The large-leaved salmonberry is conspicuous on cool forested slopes 
in the Canadian zone throughout the mountains of western and 
central Colorado, but is uncommon east of the Continental Divide. 
The large juicy red fruit has a delicious flavor and is usually ripe by 
the end of August. Salmonberries were found in great abundance 
on the western slopes of the Park Range east of Steamboat Springs; 
near Halms Peak, 8,000 to 8,500 feet; on the Book Plateau; on Lone 
Cone; and at Thomasville. 

Oreobatus deliciosus. False Raspberry. 

This is a characteristic shrub in the eastern foothills up to 9,000 or 
10,000 feet, but was not observed in the western mountains. In the 
foothills west of Boulder the large showy white flowers were out June 6, 
1905, and the species was flowering at Georgetown June 20. It is 



1911.] TREES AND SHRUBS. 233 

particularly abundant in the yellow pine belt at Bailey and on the 
slopes near Buena Vista. The dark purplish fruit of tins shrub is 
inedible. 

Rubus strigosus. Red Raspberry. 

Red raspberries were fully ripe on cool north slopes in the yellow 
pine belt near Arkins July 26, 1906. At this locality they were grow- 
ing in profusion at 6,000 feet. The species is common at Ophir, St. 
Elmo, Thomasville, and on the eastern slopes of the Sangre de Cristo 
Range in western Huerfano County. It is usually found in slide rock 
or in areas which have been swept by forest tires. 

Dasiphora fruticosa. Shrubby Cinquefoil. 

The shrubby cinquefoil is one of the most showy flowering shrubs 
in mountain meadows and on open slopes between 8,000 and 11,000 
feet. It occurs in greatest abundance in wet meadows in the high 
mountain parks at 9,000 or 10,000 feet in the Canadian zone. The 
large yellow flowers dotted the upper slopes of Mount McClellan 
between 10,500 and 11,000 feet during the middle of June, 1905, and 
the species was flowering at 10,000 feet on the South Park plains at 
Como as late as August 21, 1907. Above 11,000 feet Dasiphora oc- 
curs in a dwarfed state not over a foot high — about half its normal 
growth. It is an abundant shrub near Pearl, in northwestern North 
Park, at 8,700 feet, where it was flowering August 8, 1906. I observed 
the species on the White River and Uncompahgre Plateaus; at Divide, 
Teller County; in Slate River Valley between Almont and Crested 
Butte; and in the meadows along the headwaters of Cebolla Creek, 
southeast of Lake City, up to 11,000 feet. 

Holodiscus dumosus. 

This handsome flowering shrub is conspicuous among rocks in the 
foothill districts up to 9,000 feet on both slopes of the mountains. 
The species occurs mainly in the Transition zone. It was in flower 
July 12, 1905, in the canyon along Grand River just west of Hot 
Sulphur. I have found it common in the canyon of the Grand 
above Glen wood Springs; in Unaweep Canyon; on the Uncompahgre 
Plateau, head of Dominguez Creek; at Ouray; Buena Vista; and in 
the Wet Mountains east of Westcliffe, 9,500 feet. 

Kunzia tridentata. 

Tins species is generally distributed over the mountainous sec- 
tions of the State from 6,000 to 9,500 feet. It is usually present 
on dry open hills and was not observed in heavy forests. It forms 
dense thickets in the sand dunes along the western base of the Medi- 
cine Bow Range in North Park, and in the sandy yellow pine country 
on the head of Dominguez Creek, at the northern end of Uncompah- 
gre Plateau. At a little distance this species bears a close resemblance 



234 NORTH AMERICAN FAUNA. [No. 33. 

to sagebrush (Artemisia tridentata), but is darker green and seldom 
grows to a height of more than 2 feet. It is usually in full bloom by 
the 1st of July. 

K. tridentata was observed at the following localities: Nederland; 
Elkhorn, Larimer County; bare hillsides along Snake River at Slater; 
Canadian Creek, North Park; Coulter, Middle Park; river bluffs near 
Baggs Crossing; watershed between Snake and Bear Rivers; O-wi-yu- 
kuts Plateau ; Escalante Hills ; Book Plateau (a little on 8,500 foot 
summit at Columbine) ; canyon of the Grand above Glenwood Springs ; 
Somerset; Dillon; Sapinero; Arboles; East Paradox Valley (rare); 
and Uncompahgre Plateau. 
Cowania mexicana. Cliff Rose. 

The handsome cliff rose is restricted in Colorado to the warm 
valleys and lower mesas of the southwestern counties, where it has 
an irregular distribution from the Mesa Verde north to Unaw T eep 
Canyon. The growth is usually from 4 to 6 feet. The species forms 
the most conspicuous shrubbery on the warm open rocky slopes along 
the north side of Unaweep Canyon, at the northern end of the Uncom- 
pahgre Plateau, where it occurs up to 7,000 feet. By the end of 
July, 1907, it had ceased blooming in Unaweep Canyon and also on 
the lower slopes of Salt Canyon, between Sinbad Valley and Dolores 
River, at 5,500 feet. Cliff roses were a mass of yellow bloom June 
13, 1907, along the rocky rims of Navajo Canyon, Mesa Verde, at 
7,000 feet. Here they were growdng commonly among junipers and 
pinyons, often on rocks where there was scarcely any soil. 

Cercocarpus parvifolius. Mountain Holly. 

The mountain holly is almost omnipresent on the warmer foothill 
slopes from the lowest edge of the pinyon belt to the upper edge of 
the yellow pine belt, on both slopes of the mountains. It forms a 
dense growth, covering many of the open slopes along the lowest edge 
of the foothills, and it is common on the higher rocky ridges on the 
plains of northeastern Colorado at a little over 5,000 feet. 

The following localities show the wide range of C. parvifolius: 
Livermore; Pawnee Buttes, Weld County; Platte Canyon Station; 
Bailey; Salida; Golden; Boulder; Manitou; Eastonville; Walsen- 
burg; Gardner; Trinidad; Slater; Godiva Ridge; Escalante Hills; 
O-wi-yu-kuts Plateau; south of Lily; southwest of Rangely; Plateau 
Creek; De Beque; Meeker; West Elk Mountain^ east of Crawford; 
Somerset; Cerro Summit; Ouray; Placerville; Naturita; Sinbad 
Valley, above 6,000 feet; Dolores; Ute Peak, lower slopes; Mancos; 
Mesa Verde; Arboles; Bayfield; and Wagon Wheel Gap. 

Cercocarpus ledifolius. Mountain Mahogany. 

The mountain mahogany was noted on the 7,000-foot crest of the 
Escalante Hills near Douglas Spring, in western Routt County, where 






1911.] TREES AND SHRUBS. 235 

I found it growing commonly on rocky ridges among the yellow 
pines in September, 1906. Apparently it does not occur elsewhere 
in Colorado. 

Rosa manca. Rose. 

This handsome rose grows in profusion in openings along the crest 
of the Uncompahgre Plateau at about 9,000 feet. The bushes were 
a mass of pink bloom near Uncompahgre Butte July 16, 1907. 

A great variety of wild roses grow in the Colorado mountains up to 
10,000 feet, and are in flower nearly all summer. Since few speci- 
mens were collected, I am unable to correlate my notes with the vari- 
ous species and can give no data of consequence on their distribution. 

Amelancliier bakeri. June Berry. 

This June berry was collected on West Creek, near its junction with 
Dolores River, in western Mesa County. Tt is a common shrub in 
the Unaweep Canyon and also on the summit and upper slopes of 
Mesa Verde, where it forms a dense chaparral. This is probably 
the species so abundant over most of the mesas from western Mon- 
tezuma County north to Mesa County, in the Transition zone. 

Amelanchier oreophila. 

An Amelanchier which I collected on Lone Mesa, Dolores County, 
in June, 1907, proves to belong to this species. Scattering shrubs of 
small size were growing among the oak chaparral on the dry upper 
slopes at 9,400 feet, in the lower Canadian zone. I did not collect it 
elsewhere, but Rydberg gives it a wide range on the western slope. 

Amelanchier alnifolia. Common June Berry. 

As few specimens of June berries were preserved, the distribution 
data given below doubtless refers to several species. A. alnifolia is, 
however, the widest ranging species in the State. June berries are 
abundant on the dry, partially open Transition slopes throughout the 
mountains, and on many of the western mesas they often form a 
dense chaparral. Some years they bear an abundance of fruit, but 
usually it is rather scanty and of poor quality. The berries are 
eagerly eaten by birds and chipmunks. I observed the shrubs in 
abundance at the following localities: Halms Peak, below 8,500 feet; 
slopes along Snake River between Honnold and Baggs Crossing; 
Godiva Ridge; O-wi-yu-kuts Plateau; Escalante Hills, above 6,400 
feet; south of Lily; southwest of Rangely; Book Plateau, every- 
where above 6,500 feet; Plateau Creek; south of De Beque; canyon 
of the Grand above Glenwood Springs; bluffs between Dotsero and 
Wolcott; summit of Piney Divide, 8,500 feet; Transition slopes 
bordering Sinbad Valley; Lone Cone; Dolores; Mancos; Durango; 
Bayfield; Pagosa Springs; Pagosa Junction; and La Veta. 



236 NORTH AMERICAN FAUNA. [No. 33. 

Peraphyllum ramosissimum. Dwarf Apple. 

I found this shrub in great abundance on dry open hillsides south 
and west of the San Juan and La Plata Mountains, chiefly between 
6,500 and 7,500 feet. The dry slopes along the Los Pinos at Bay- 
field were covered with a mass of its pale roseate blossoms the first 
week in June, 1907, and it was in flower up to the middle of the 
month on the Mesa Verde. It was noted as follows: South slope 
of Book Plateau above Atchee, 8,000 feet; Newcastle; East Fork 
of Rifle Creek; canyon of the Grand above Glenwood Springs; Basalt; 
Somerset; rocky slopes near Hotchkiss; Montrose region above 
7,000 feet; north base of Lone Cone at 7,000 feet; Sinbad Valley 
rim above 6,000 feet; Dolores; McElmo Valley; Arboles; and Bay- 
field. 
Prunus americana. Wild Plum. 

This species occurs sparingly in gulches on the eastern plains and 
at the eastern base of the foothills in the Upper Sonoran zone. Thick- 
ets of wild plum were observed at Wray; Gaume's ranch, north- 
western Baca County; Arkins; and Boulder. Rydberg records the 
species from a number of localities extending across the State from 
north to south at the base of the foothills. 

Prunus pennsylvanica. Red Cherry. 

Rydberg gives the red cherry a wide range in the eastern foothills 
of the Front Range up to 9,500 feet. I found it common in the foot- 
hills at Boulder and Manitou. 

Prunus melanocarpa. Chokecherry. 

This species appears from my observations to be the common 
wild cherry of the mountains, particularly in western Colorado. 
The growth is uniformly scrubby, often forming a dense chaparral 
on the Transition zone summits and upper slopes of plateaus and 
mesas in the western counties. It is sparingly present in gulches 
on the eastern plains. The following are localities at which P. 
melanocarpa was observed: Northern slopes Elk Head Mountains, 
7,800 feet; Slater to Baggs Crossing; White River below Angora; 
Book Plateau, dense chaparral 7,000 feet to summit; Glenwood 
Springs; summit of Piriey Divide, 8,000 feet; Pitkin; Somerset; 
lower slopes of Lone Cone; Uncompahgre Plateau; Placerville; 
Ouray; Vernal Mesa; West Elk Mountains east of Crawford; Frying 
Pan River, Basalt to Thomasville; Dolores; Mancos; Hermosa; 
Arboles; Pagosa Springs; Sterling; Wray; and Tuttle. 

Robinia neomexicana. Locust. 

A very few of these locusts were observed on the rocky banks of 
Grand River near Tunnel, Mesa County, at about 5,000 feet. Re- 
corded by Rydberg from Denver, Walsenburg, La Veta, and Trinidad. 



1911.] 



TREES AND SHRUBS. 237 



Rhus rydbergi. Poison Ivy. 

The poison ivy is common in canyons on the eastern plains and 
also in the lower eastern foothills of the Front Range. It was 
observed at Tuttle; Wray; Pawnee Buttes; in gulches along the 
southern escarpment of the Chimney Cliffs, 30 miles northwest 
of Sterling; in the foothills at Boulder and Golden; and it was col- 
lected in Navajo Canyon, Mesa Verde, southwest Colorado. 

Schmaltzia glabra. Sumac. 

This sumac I have found sparingly up to 6,000 feet in the foothills 
west of Boulder and Golden, and also near Livermore and Platte 
Canyon Station. Its brilliant reddish autumnal foliage was very 
conspicuous on the slopes along Middle Boulder Creek October 20, 
1906, when the leaves had just commenced to fall. 

Schmaltzia trilobata. Skunk Bush. 

The zonal range of this sumac is mainly Upper Sonoran, and it is 
equally abundant on both sides of the mountains. In the foothills 
it usually grows on the warm sides of canyons and along streams. 
In the warm desert valleys of western Colorado it is often found 
with sagebrush in the open, but also forms a good growth along 
watercourses, where it sometimes attains a height of 8 or 10 feet. 
East of the mountains this sumac occurs chiefly in gulches in the 
rougher parts of the plains. 

S. trilobata was observed as follows: Tuttle; Wray; bluffs east of 
Sterling; Chimney Cliffs; Pawnee Buttes; Gaume's ranch, north- 
western Baca County ; Livermore, dry slopes up to 6,500 feet; Boulder; 
Golden; Platte Canyon Station; Segundo; Buena Vista, up to 8,500 
feet; bluffs along Snake River below Baggs Crossing; Plateau Creek; 
Basalt; Hotchkiss; Naturita; Sinbad Valley; West Creek; McElmo; 
and Pagosa Springs. 

Pachystima myrsinites. 

This little evergreen shrub is abundant throughout the moun- 
tains, growing in dense clumps in wooded gulches and on shaded 
north slopes between 6,000 and 9,000 feet. It was found in abun- 
dance at the following localities: Northern slopes of Elk Head 
Mountains at 7,800 feet; Park Range up to 9,000 feet; canyon of the 
Grand above Glenwood Springs; summit of Escalante Hills; Book 
Plateau, 8,000 feet; Navajo Canyon, 7,000 feet; and northern escarp- 
ment of Mesa Verde. 

Acer glabrum. Mountain Maple. 

The mountain maple is common and widely distributed in the 
mountains of Colorado, growing in greatest profusion on damp 
shaded slopes between 5,500 and 9,000 feet and forming dense 
clumps on the borders of streams and bogs. The bright yellow 



238 NOKTH AMERICAN FAUNA. [No. 33. 

autumnal foliage gives a brilliant coloring to the mountain slopes 
in late September. The average height is not over 8 feet, although 
12 feet is sometimes attained. I have found this species especially 
common among the eastern gulches of the Front Range and on 
both slopes of the Saguache Range. It occurs on the northern 
slope of the Book Cliffs and is abundant at Aspen and in most of the 
mountains of southern Colorado. 

Acer negundo. Box Elder. 

Box elders are common trees on most of the streams of the eastern 
plains and scattered clumps are often met with on foothill streams 
to elevations of 7,000 or 8,000 feet on both sides of the mountains. 
The zonal distribution is mainly Upper Sonoran. In my trips over 
the State I have found box elders at the following localities: Snake 
River, from 5 miles east of Slater to Baggs Crossing ; southern slopes 
of Book Cliffs near Atchee at 7,500 feet; Rifle Creek; canyon of 
Plateau Creek ; canyon of the Grand above Glen wood Springs ; Frying 
Pan River, Basalt to Peachblow; Ouray; Unaweep Canyon; Wray; 
Platte Canyon Station; Las Animas River, Segundo to Weston; and 
Shell Rock Canyon, northwestern Baca County. 

Rhamnus smithi. Buckthorn. 

This species is known apparently from only two localities in the 
State. It was abundant and in flower May 27 on the banks of San 
Juan River at Pagosa Springs (7,000 feet), and was again encoun- 
tered in the hot canyon of Dolores River between Salt Canyon and 
the mouth of West Creek at about 5,000 feet. Dense spreading 
thickets of buckthorn, often 10 or 12 feet high, fringed Dolores River 
for most of this distance and extended eastward in the West Creek 
Valley to an elevation of 5,500 feet. The black-green foliage of 
R. smithi was in marked contrast to the predominant gray-green 
desert vegetation along the Dolores and on the bordering canyon 
sides. The fruit of this shrub had nearly all turned to a rich pur- 
plish black, July 13, 1907, but some was still green or only partially 
colored. This species has been recorded from Pagosa Springs by 
Rydberg. 

Ceanothus velutinus. Mountain Balm. 

This shrub forms a dense chaparral about 2 feet in height on the 
central slopes of most of the mountain ranges of northern Colorado, 
extending south to Thomasville on the western slope of the Saguache 
Range. I did not see it on any of the mountains of southern Colorado. 
The rankest growth was observed in the Middle Park region, on the 
Park Range, and on the hills around Steamboat Springs, where in 
places the dense thickets were well-nigh impassable. The species is 
occasionally found as low as 6,500 feet, but usually grows between 



1911.] TREES AND SHRUBS. 239 

7,500 and 9,500 feet. It is most abundant on dry, partially open 
slopes which have been burned over by forest fires. The oval bright 
green leaves are remarkably shiny and glabrous, giving ; i peculiar 
brilliance to this chaparral on a bright day. 

Other points where I observed C. velutinus are: Northern slopes of 
Elk Head Mountains; bluffs along Snake River, 10 miles east of 
Slater; Escalante Hills, 7,000 feet; watershed between Bear and 
White Rivers, north of Midland Basin, 6,500 feet; juniper slopes at 
McCoy, 7,000 feet; northern slopes of Piney Divide, 8,000 feet; Coul- 
ter, Middle Park; northern slopes of Book Plateau; and Dillon. 

Ceanothus pubescens. 

This is a common shrub on dry rocky slopes in the eastern foothills 
of the front ranges, chiefly in the Transition zone. I found it abun- 
dant also in sandy yellow pine forest near the head of Dominguez 
Creek on the Uncompahgre Plateau, at 8,000 feet. It is common in 
the foothills near Boulder and La Veta. 

Ceanothus fendleri. 

This species is common in the Transition zone over most of the 
State, according to the range given by Rydberg. I have met with it 
only in the southwest on the summit of Lone Mesa, 0,400 feet, and 
on the open gravelly benches along McElmo Creek in western Mon- 
tezuma County, 5,500 feet. 

Cactus missouriensis. Ball Cactus. 

The common ball cactus is abundant on the high plains from the 
Arkansas Divide near Cheyenne Wells northwest to Weld and Logan 
Counties. I have found it common on the South Park plains near 
Como up to 10,000 feet, in the yellow pine belt at Bailey, and on the 
sage plains of western Routt County. 

Cactus radiosus. 

This ball cactus is tolerably common among rocks in the pinyon 
belt at Coventry, in western Montrose County, where I collected 
specimens in July, 1907, at an elevation of 6,500 feet. 

Echinocactus simpsoni. Snake Cactus. 

This peculiar cactus is found chiefly at the higher elevations. In 
November, 1907, I found it common in Wet Mountain Valley; in the 
Wet Mountains east of Westcliffe between 9,000 and 9,500 feet; and 
on the eastern slopes of the Sangre de Cristo Range near Mosca Pass — 
both in yellow pine forest and in the open. 

Echinocereus viridiflorus. Cereus. 

This greenish flowered cereus is abundant in the eastern foothills 
of the front ranges across the State from north to south. It ascends 



240 NORTH AMERICAN FAUNA. [No. 33. 

to at least 9,000 feet in the Wet Mountains east of WestclifTe, where 
in November, 1907, I found it in rocky soil among the yellow pines. 
It was common at Walsenburg, 7,000 feet; and at Gaume's ranch, 
in Shell Rock Canyon, northwestern Baca County, it was abundant 
among rocks at 4,600 feet. The species was not observed in western 
Colorado. 

Echinocereus paucispinus. 

This cereus is not uncommon in the rocky piny on and juniper 
country of southwestern Colorado, chiefly below 7,500 feet. It was 
taken north of Dolores at 7,500 feet, and was very common at Cov- 
entry. It is recorded from Durango by Rydberg. 

Opuntia polyacantha. Prickly Pear. 

This is an Upper Sonoran species, occasionally growing to 7,000 
feet on warm slopes in the foothill valleys. It is of general occurrence 
on the plains east of the mountains, where it is the common prickly 
pear, and it is present on the sage plains of Routt County. The yel- 
low flowers usually dot the plains during the first two weeks in June. 
It is common at Slater; Snake River Valley; Maybell; Boulder; Fort 
Collins; Sterling; Pawnee Buttes; Platte Canyon Station; Limon; 
Wray; Cheyenne Wells; Pueblo; Walsenburg; Salida; and Gaume's 
ranch, northwestern Baca County. 

Opuntia rhodantha. 

This handsome red-flowered prickly pear is abundant in the warm 
valleys of extreme southwestern Colorado. At McElmo it was in 
bloom during the middle of June, 1907. Rydberg records it from 
Grand Junction and Boulder. 

Opuntia camanchica. 

A large-jointed species, chiefly of southwestern Colorado — McElmo, 
Cortez, Dolores, Coventry, and Paradox Valley. It was in flower in 
McElmo Valley June 15 to 22, 1907. Rydberg records it from 
Colorado Springs. 

Opuntia fragilis. 

This small-jointed species is not uncommon in the foothill districts 
between 6,000 and 8,500 feet, in the Transition zone. It was growing 
in rocky situations in the yellow pine forest near Pagosa Springs, and 
also on the northern end of the Uncompahgre Plateau, and was 
occasionally noted on the high plains near Pawnee Buttes, in north- 
eastern Weld County. It was observed on rocky juniper slopes along 
the head of Smith Fork, West Elk Mountains; at Somerset; Buena 
Vista; Plateau Creek, Mesa County; and Unaweep Canyon. Ryd- 
berg records it from Denver and Boulder. 



1011.] 



TREES AND SHRUBS. 



241 



Opuntia arborescens. Tree Cactus. 

The tree cactus is a characteristic Upper Sonoran species from the 
Arkansas Valley southward (see fig. 37), and over much of south- 
eastern Colorado is the most prominent shrub on the level plains. 
It extends some distance into the foothills along the warmest slopes 
of the valleys, reaching its western limit in the Arkansas Valley at a 
point 5 miles east of Salida, and in the Huerfano Valley a short 
distance above Gardner. It reaches at Fountain its northern limit 
along Fountain Creek, and it occurs near Trinidad in the Las Animas 
Valley, and at Walsenburg in the Cucharas Valley. It is particularly 
abundant in the canyons of Las Animas and western Baca Counties. 




Fig. 37.— Distribution in Colorado of tree cactus {Opuntia arborescens). 

Lepargyrea argentea. Buffalo Berry. 

Dense thickets of buffalo berry are present along many of the 
streams of the plains on both sides of the Continental Divide, and 
the species extends into the foothills along some of the watercourses 
to an elevation of from 6,000 to 7,000 feet. The zonal distribution 
is mainly Upper Sonoran. Sandy river banks are especially suited 
to the growth of the buffalo berry. The brilliant scarlet clusters of 
berries are usually ripe by the middle or end of August, and are 
eagerly eaten by sage thrashers and many other birds. In late 
August, 1907, chipmunks (Eutamias minimus) were feeding exten- 
90432°— 11 16 



242 NORTH AMERICAN FAUNA. [No. 33. 

sively upon the ripe buffalo berries in the Snake River Valley. We 
found the berries when cooked and used as sauce a welcome addition 
to our camp fare, but uncooked they are extremely acid. Observed 
at La Porte; Snake River Valley, Slater to Lily; Browns Park; White 
River Valley, Angora to Rangely; Dotsero to Eagle; Newcastle; 
Dallas Creek, near Ridgway ; Basalt; Hotchkiss; near Montrose; Natu- 
rita; Pagosa Junction; Arboles; and along most of the streams on 
the eastern plains. 

Lepargyrea canadensis. Canadian Buffalo Berry. 

This is a characteristic Canadian zone shrub in the mountains of 
northern and central Colorado, becoming less common toward the 
south. It is conspicuous in the undergrowth of the dry lodgepole 
pine forests of the central mountain slopes between 8,000 and 10,000 
feet, but I have not observed it much lower. Above 10,000 feet it 
becomes decumbent, often not more than a foot high, and is of rare 
occurrence above 11,000 feet. The bright red berries of this buffalo 
berry have an attractive appearance, but are unpalatable and very 
bitter. They are ripe by the 1st of August, and usually remain on 
the bushes during most of that month. The leathery leaves had 
nearly all fallen at St. Elmo, in the Saguache Range, at 10,000 feet, 
October 9, 1907. 

I found L. canadensis at the following localities: Eight thousand 
five hundred to ten thousand feet on the Medicine Bow and Laramie 
Mountains ; Park Range, west of Pearl, 9,000 to 10,000 feet; Ophir, to 
10,500 feet; Culebra Range, near La Veta; Thomasville; Floyd Hill; 
Como ; Dillon ; and St. Elmo. Rydberg mentions its occurrence in the 
southwestern mountains — near Ouray and on Bear Creek Divide in 
the West La Plata Mountains. 

Svida stolonifera riparia. Red-osier Dogwood. 

The dogwood or cornel is a prominent shrub along streams in the 
Transition zone nearly throughout the mountains, and its clusters 
of white berries are very conspicuous in autumn. Rydberg records 
another species (S. interior) from several points in the foothills, 
but in my trips through the mountains I have met with only the 
present species. I observed the cornel on Plateau Creek, east of 
Tunnel; along White River at Meeker; on the upper reaches of 
Smith Fork, in the West Elk Mountains; along streams heading in 
Book Cliffs, above 7,o00 feet; and on most of the streams on the 
eastern slopes of the Front Range. 

Arctostaphylos uvaursi. Red Bearberry. 

This handsome bearberry is one of the most widely distributed 
mountain shrubs in Colorado, growing on practically all the ranges 
from 6,000 to 10,000 feet. It grows luxuriantly on dry shaded 
slopes beneath lodgepole and yellow pine forest, and the trailing 



1911.] TREES AND SHRUBS. 243 

mats of rich dark green are usually very dense and of considerable 
extent. 

I found the red bearberry abundant at Thomasville; Dillon; 
Como; St. Elmo; Escalante Hills, summit at 7,000 feet; Wet 
Mountains, east of Westcliffe; Bradford, Huerfano County; foot- 
hills west of Boulder; Coulter, Middle Park; Divide; and in the 
yellow pine forest on the Arkansas Divide at Eastonville. 

Arctostaphylos pungens platyphylla. Manzanita. 

The manzanita is found on the dry slopes of the Uncompahgre 
Plateau and on the eastern slopes of the La Sal Mountains in western 
Montrose County. It appears not to have been recorded previously 
from the State. 

In July, 1907, 1 noted a scattering growth on the head of Dominguez 
Creek, at the northern end of the Uncompahgre Plateau, and in 
descending the steep southwestern escarpment north of I s abeguache 
Creek a dense chaparral of this manzanita was traversed on the 
rocky slopes immediately below the aspen belt, at about 8,000 feet. 
I found it a common undershrub in the yellow pine forest just above 
the western rim of West Paradox Valley, between 7,000 and 8,000 
feet. The species came under my observation only in the Transition 
zone and appears to grow principally on dry, partially open slopes. 

Vaccinium csespitosum. Huckleberry; Blueberry. 

This blueberry is common from Yankee Doodle Lake (10,500 
feet) to timberline on Rollins Pass; at Ophir (10,500 feet) to timber- 
line; and on the upper slopes of Lone Cone. It is most abundant 
on the mossy floor of the Canadian zone forests. 

Vaccinium oreophilum. Bilberry. 

The bilberry forms low, dense carpets 8 or 10 inches high in the 
lodgepole pine and Engelmann spruce forests between 8,000 and 
11,000 feet, and is particularly abundant on the Front, Park, and 
Gore Ranges. (See fig. 38.) I found it on Mount McClellan and 
near Berthoud Pass at 11,000 feet altitude. 

Vaccinium erythrococcum. Small-leaved Bilberry. 

This diminutive small-leaved bilberry occurs on Mount McClellan 
between 11,000 feet and timberline, and was observed on Berthoud 
Pass, Buffalo Pass, and on the Park Range along the headwaters 
of Grand Encampment River. Rydberg records it from a number 
of localities on the Front, Saguache, Park, and Sangre de Cristo 
Ranges. 

Fraxinus anomala. Ash. 

This small Sonoran species I saw only in the warmer parts of 
southwestern Colorado, where it was observed at several localities 
from Mesa County south to Montezuma County at elevations varying 



244 



NORTH AMERICAN FAUNA. 



[No. 33. 



from 5,000 to 5,500 feet. It usually grows to a height of from 6 to 
10 feet, and single trees or small clumps are scattered here and there 
in gulches and on the warm rocky slopes in canyons. It was observed 
in the following places: Slopes bordering McElmo Valley, between 
Moqui and McElmo; Salt Canyon, the outlet of Sinbad Valley; 
canyon of Dolores River down to the mouth of West Creek; western 
(lower) part of Unaweep Canyon; and lower slopes bordering West 
Paradox Valley. Rydberg records this species from Grand Junction, 
Deer River, and between Hotchkiss and Smith Fork. 

Lycium pallidum. Matrimony Vine. 

This species is restricted in Colorado, so far as known, to the low 
arid Upper Sonoran stretches of the extreme southwest. I found a 

scattering growth in 
the partially open 
sage and pinyon coun- 
try between Bayfield 
and Ignacio, La Plata 
County. It is re- 
corded by Rydberg 
from McElmo Creek 
and San Juan Valley. 
Sambucus micro- 
botrys. Elder. 
This elderberry is 
abundant in the Ca- 
nadian zone forest 
throughout the 
mountains between 
8,000 and 10,500 feet. 
I have seldom ob- 
served it in the Tran- 
sition zone, but occa-' 
sionally it occurs 
nearly to timberline. 
The red berries are usually ripe by the middle of August or the first 
of September. S. microbotrys is abundant on both slopes of the Front 
Range; at Pearl, North Park; Buffalo Pass, Park Range; summit of 
Continental Divide; north of Ilahns Peak; Arapahoe Pass, Rabbit 
Ear Mountains; Thomasville; Dillon; St. Elmo; headwaters of Cebolla 
Creek southeast of Lake City; and Lone Cone. 

Sambucus melanocarpa. 

This is a lower ranging species than S. microbotrys, and grows more 
in the open and chiefly in the Transition zone. I met with it only in 




Fig. 38. — Vaccinium orcnphilum on the iloor of lodgepole pine forest, 
Park Range, west of North Park, 10,000 feet. 



1911.] TREES AND SHRUBS. 245 

western Colorado, but Rydberg records it from several points in the 
eastern foothills of the Front Range. It was noted as follows : West- 
ern slope of Continental Divide north of Halms Peak, 9,000 feet; 
Halms Peak; Escalante Hills, 6,400 to 7,000 feet; and Uncompahgre 
Plateau, 8,000 feet. It was growing in profusion on the rocky north- 
ern slopes and along the crest of the Escalante Hills, where the black 
berries were fully ripe September 3, 1906. 
Viburnum pauciflorum. Few-flowered Viburnum. 

This small viburnum was observed only at Arapahoe Pass, Rabbit 
Ear Mountains, at an elevation of 9,000 feet. Rydberg records it 
from Grand Lake, Minnehaha, and Clear Creek. 
Linnsea americana. Twinflower. 

The twinflower occurs on the higher slopes of the front ranges 
nearly across the State, according to Rydberg. I collected it at 
10,000 feet near Idlewild on the Middle Park slope of the Front Range, 
and observed it near Mount Whiteley in northwestern Middle Park, 
at about 8,500 feet — both localities being in the Canadian zone. 
Near Idlewild this small trailing vine was abundant on mossy slopes 
in the damp Engelmaim spruce forest. 
Symphoricarpos occidentalis. Wolfberry. 

The wolfberry is a very abundant shrub on the banks of the South 
Platte River near Sterling, where it forms dense thickets about 2 
feet high. I found it also at Golden. R} T dberg records it from a 
number of localities along the eastern base of the foothills. 
Symphoricarpos oreophilus. Snowberry. 

A dense scrubb}" growth of this small-leaved snowberry covers 
many open mountain slopes in the Transition zone, particularly in 
western Colorado. It is rarely found in forests or damp situations, 
but grows rampant on dry, rocky hillsides and mesas. It occurs 
in abundance in the following localities: Hahns Peak region below 
8,500 feet; Escalante Hills; divide between Bear and White Rivers; 
southwest of Rangely; Book Plateau above 6,500 feet; canyon of 
the Grand above Glen wood Springs; Mount Whiteley; West Elk 
Mountains east of Crawford; Vernal Mesa; Somerset; Uncompahgre 
Plateau; Lone Cone; and Ute Peak. 
Distegia involucrata. Involucred Fly Honeysuckle. 

Throughout the mountains this honeysuckle is a common and 
conspicuous undershrub in the forests of the middle slopes. It usu- 
ally grows in damp situations and is particularly common along 
streams, where it reaches its rankest growth. I have found it as 
low as 7,000 feet along cold streams in various parts of the moun- 
tains, and it was common at 10,500 feet on the upper reaches of 
Cebolla Creek in the San Juan Mountains southeast of Lake City. 
The dark purplish or blackish berries usually fall in late August. 



246 



NOETH AMERICAN FAUNA. 



[No. 33. 



Artemisia tridentata. Sagebrush. 

The common sagebrush is almost omnipresent on the higher plains 
of western Colorado and also in most of the higher mountain parks 
up to 10,000 feet, but was not noted on the plains east of the moun- 
tains. It grows on an average about 2 feet high, but under favor- 
able conditions attains a height of 6 feet or more. The rankest 
growth I have observed was on .the banks of sandy arroyos near Lay, 
Routt County, where many of the shrubs were 8 or 10 feet in height. 
A. tridentata is abundant at the following localities: Glendevey, Lara- 
mie Valley; Livermore; North Park plains (see fig. 39); Ilahns Peak 
to Slater; Snake River Valley; Iron Springs Divide; Lily Park; south- 
west of Rangely; Plateau Creek; De Beque to Glenwood and Dot- 
sero, Grand River Valley; Wolcott; Piney Divide; McCoy; Egeria 
Park; parks in Gore Mountains; Middle Park; Roaring Fork Valley 




Fig. 39.— Desert sagebrush (Artemisia tridentata) on plains near Higho, North Park. (Park Range in the 

distance. ) 

to Aspen; Uncompahgre Plateau; Lone Cone; Lone Mesa; Naturita; 
Coventry; Cerro Ridge; Somerset; Sapinero; Gunnison; Creede; 
Poncha Pass; Buena Vista; Leadville; Hotchkiss; Saguache; Bay- 
field; Arboles; and McElmo. 
Artemisia cana. 

This sage occurs in Colorado at a somewhat higher average eleva- 
tion than A. tridentata, although at many points the two species 
grow together. It was found in abundance at Coulter, Middle Park; 
near Toponas, Egeria Park; on the Uncompahgre Plateau; and on 
the lower slopes of Lone Cone, in the San Miguel Mountains. Ryd- 
berg records it from Breckenri'dge; Marshall Pass; Hay den Flats, 
Routt County; Hebron, North Park; and Timnath. 



INDEX. 



[Synonyms in italics; pages containing the principal reference to a species in bold-faced figures.] 



Abies concolor, 41, "220. 

lasiocorpa, 45, 40, 121, 219-220. 

Accipiler velox, 37. 
Acer glabrum, 34, 45, 237. 

negundo, 20, 288. 
Acknowledgments, 11-12. 
Actaca viridifiora, 45. 
Aeronautes melanoleucus, 37. 
Agelaius phceniceus fortis, 20. 

phceniceus neutralis, 23. 
Agriculture, Canadian zone, 43. 

Transition zone, 40-41. 

Upper Sonoran zone, 20-33. 
Agropyron smithi, 21. 
Alder, 39, 45, 227-228. 
Alnus tenuifolia, 39, 45. 227-228. 
Ambystoma tigrinum, 22. 40. 
Amelanchier alnifolia, 36, 38, 235. 

bakeri, 3S, '2:55. 

oreophila, 235. 
Ammodramus savannarum bimaculatus, 20. 
Aminospermophilus leucurus cinnamomeus, 24, 

84-86. 
Amorpha angnstifolia, 20. 
Amphispiza bilineata, 24. 

nevadensis, 23. 
Anas platyrhynchos, 38. 
Andropogon furcatus, 21. 

halli, 21. 

scoparius, 21. 
Androsace, 49. 
Antelope, 19, 23, 58-60. 
Anthus rubescens, 50. 
Antilocapra, 27. 

ampricana, 19, 37, 58-60. 
Antrozous pallidus, 24, 205 206. 
• Apheloeoma woodhousei. 29. 
Apple, dwarf, 236. 
Aquilegia cferulea, 45. 
Aragallus lamberti, 21. 
Archibuteo ferrugineus, 20. 
Arctostapbylos pungens platyphylla, 39, 243. 

uvaursi, 39,242-24:',. 
Ardea herodias, 38. 
Aristeda longiseta, 21. 
Artemisia, 27. 

cana, 246. 
filifolia, 21. 

tridentata, 22, 23, 39, 43, 234, 246. 
Asclepias, 21. 
Ash, 243-244. 
Asio wilsonianus, 145. 
Aspen, 35, 41, 45, 224-225. 
Astragalinus psaltria, 24. 
Astragalus crassicarpus, 21. 
mollissimus, 21. 



Asyndesmus lewisi, 37. 
Atriplex, 27, 86, 140, loo. 

canescens, 21, 23, 25, 29, 138, 141, 228-229. 

confertifolia, 23, 25, 1 10, 141. 143, 162, 229. 

nuttalli, 23, 25, 98, 229-230. 

occidentalis, 78. 

Badger, 19, 90, 162, 181-182. 
I3a?olophus inomatus griseus, 29. 
Balm, mountain, 238-239. 
Balsam, 45, 47, 48, 49, 121, 219-220. 
Barberry, 230.' 
Bartramia longicauda, 20. 
Bascanion constrictor, 21. 
tseniatum, 27. 
Bassariscus astutus flavus, 24, 28, 192-193. 
Bat, 24. 

big-eared, 204. 
brown, 20, 209-210. 
Fort Yuma, 23, 207. 
free-tailed, 204-205. 
hairy-lipped, 19, 209. 
hoary, 211. 
little California, 208. 
long-eared, 19, 23, 207. 
long-legged, 206-207. 
pale, 205-206. 
pale brown, 210. 
red, 211. 
Say, 20, 206. 
silver-haired, 44, 211. 
small brown, 36. 
Taeubaya free-tailed, 205. 
western, 209. 
Batrachians, Transition, 39. 

Upper Sonoran, 21, 23, 25. 
Bayonet, Spanish, 223-224. 
Bear, black, 44, 195-196. 
gr