Full text of "Extinct and vanishing mammals of the western hemisphere, with the marine species of all the oceans"

v *-

EXTINCT AND VANISHING MAMMALS
OF THE WESTERN HEMISPHERE

WITH THE MARINE SPECIES
OP ALL THE OCEANS

P7 U-E7

EXTINCT AND VANISHING
MAMMALS

of the

WESTERN HEMISPHERE

with the
MARINE SPECIES OF ALL THE OCEANS

by
GLOVER M. ALLEN

Special Publication No. 11

fORESTA WST1JOT1
V v MSI

STu. -'-

o205 PR-'-

cnv,

1942

AMERICAN COMMITTEE FOR INTERNATIONAL WILD LIFE PROTECTION

O very creature is better alive than dead,
men and moose and pine-trees, and he
who understands it aright will rather
preserve its life than destroy it.

HENRY DAVID THOREAU
The Maine Woods

THE INTELLIGENCER PRINTING Co.
LANCASTER, PA.

IN MEMORY OF

JOHN C. PHILLIPS

(1876-1938)

NATURALIST, CONSERVATIONIST

AND SPORTSMAN IN THE

HIGHEST SENSE

AMERICAN COMMITTEE

FOR

INTERNATIONAL WILD LIFE PROTECTION

EXECUTIVE COMMITTEE

CHILDS FRICK, Chairman

HAROLD J. COOLIDGE, Jr. DEAN SAGE, Jr.

WARREN KINNEY LAURANCE S. ROCKEFELLER

W. REID BLAIR, Secretary
New York Zoological Park

PAN AMERICAN COMMITTEE

HAROLD J. COOLIDGE, Jr., Chairman

WILLIAM G. SHELDON FREDERIC C. WALCOTT

ALEXANDER WETMORE

ADVISORY BOARD

CLINTON G. ABBOTT WILLIAM P. HARRIS. JR.

San Diego Society of Natural History Museum of Zoology, Univ. of Michigan

THOMAS BARBOUR EDWARD HOUSE

Museum of Comparative Zoology Carnegie Museum

FREDERICK R. BURNHAM EDWARD MALLINCKRODT, JR.

California Academy of Sciences Washington University, St. Louis

CHARLES M. B. CADWALADER JOHN T. MCCUTCHEON

Acad. Nat. Sciences of Philadelphia Chicago Zoological Society

JAMES L. CLARK ROBERT T. MOORE

Camp Fire Club of America California Institute of Technology

HAROLD J. COOLIDGE, JR. FAIRFIELD OSBORN

American Society of Mammalogists New York Zoological Society

STANLEY FIELD KERMIT ROOSEVELT

Field Museum of Natural History National Audubon Society

*FRANK B. FOSTER RODOLPHE M. DE SCHAUENSEE
Philadelphia Zoological Society Wilderness Club

CHILDS FRICK WTTTTATW

WlL * IAM

American Museum of Natural History

Boone and Crockett Club

*JOSEPH GRINNELL ^ TX r

Museum of Vertebrate Zoology FREDERIC C. WALCOTT

American Wildlife Institute

ALLAN HANCOCK

Allan Hancock Foundation, ALEXANDER WETMORE

University of Southern California Smithsonian Institution

* Deceased

FOREWORD

AY 1, 1942, is likely to be a significant date in the history of con-
servation, for on that day the "Convention on Nature Protec-
tion and Wildlife Preservation in the American Republics " came into
effect for the seven American countries, including the United States,
that had deposited ratifications with the Pan American Union.
Twelve other Republics have now signed the Treaty, and it is hoped
that their ratifications are soon to follow.

Publication of the present volume, devoted to the extinct and
vanishing mammals of the New World, including certain marine
mammals of all the oceans, is therefore particularly timely. Its
author, Dr. Glover M. Allen, of the Museum of Comparative Zoology
at Harvard University, who died on February 14, 1942, was one of
the great naturalists of his day. His profound knowledge of mammals
and birds was the result not only of a life-time study in museums but
also of intimate observation of animals in their native habitats in
Africa, Australia, the Caribbean, and South America, as well as in
many parts of his own country. The final preparation of this report
was one of the last projects that Dr. Allen completed before his un-
timely death. We know that he thought of it as one might think of
the clearing of a new trail into a virgin forest. It will, indeed, be the
opening of the way to a wider understanding of the genuine need for
international cooperation, if the vanishing wildlife of the Americas
is to be preserved for the benefit of future generations.

This volume is dedicated to Dr. Allen's friend Dr. John C. Phillips,
founder and first chairman of the American Committee for Interna-
tional Wildlife Protection. It was largely due to Dr. Phillips's vision
that in May 1936 a project for a comprehensive study of the recently
extinct and vanishing mammals of the world was undertaken under
the auspices of the American Committee. Information on this sub-
ject had never been assembled, and it was felt that such a report, if
carefully prepared, would be invaluable to the work of the Committee
in helping to determine those species of mammals most urgently in
need of protection and, at the same time, to estimate factors that
might have caused the extinction of species. Basic information was

Vlll FOREWORD

needed to substantiate proposals for the protection of vanishing
species in their native habitat through the establishment of adequate
national parks and reserves. The material for the preparation of
such a report was widely scattered, and assembling and evaluating it
involved a great deal of correspondence with zoologists, game
wardens, nature protection societies, and governments in many parts
of the world, as well as a survey of widely scattered articles in popular
magazines, books on sport and travel, and scientific literature.

Dr. Francis Harper, of Philadelphia, under arrangements made by
Dr. Phillips, undertook this work and devoted many months of care-
ful research to the preparation of material on the Old World mammals,
treating in his report (as yet unpublished) about 400 forms. Dr.
Allen made use of a special bibliography of references to literature
collected by Dr. Harper, and with this as a basis prepared the present
work on the New World and marine mammals, assembling much
additional material to complete it.

The Harper-Allen study could never have been brought to its
present state of completion had it not been for the most generous
financial assistance received in the form of gifts from many individuals
over a period of seven years, a special grant from the Penrose Fund
of the American Philosophical Society (1937, No. 195), and a publi-
cation grant from the Boone and Crockett Club and the Conservation
Committee of the New York Zoological Society. The American
Committee takes this opportunity to express its gratitude to Dr.
Allen and to Dr. Harper for their months of work in the preparation
of material for reports on recently extinct and vanishing mammals
of the world. We wish also to thank the forty financial contributors
of information from many parts of the world. We are indebted
further to the Academy of Natural Sciences of Philadelphia, whose
famous library proved of inestimable assistance, and to the Museum
of Comparative Zoology, at Cambridge, for the use of its library and
other facilities. Thanks are due also to Dr. John Eric Hill, of the
American Museum of Natural History, for his contributions to Dr.
Allen's volume, and to Mr. Paul H. Oehser, editor of the United
States National Museum, for the editorial supervision of this volume
through the press and for the preparation of the index. The draw-
ings for the illustrations were made especially for this work by Mr.
Earl L. Poole, of the Reading Public Museum and Art Gallery.

In these dark days, when even man is fighting to save himself from
extinction, let us hope that this book, inspired by a great sportsman
and conservationist and written by a great naturalist "whose gentle-

FOREWORD IX

ness and purity of spirit were beyond all praise, " will in some measure
be helpful in promoting a wider understanding of the need for pre-
serving the wildlife heritage of the American republics for the enjoy-
ment and wise use of generations to come. "Natural beauty," in
the words of Trevelyan, "is the highest common denominator in the
spiritual life of today. " And what a great destiny lies ahead for the
peoples of the Americas if, in winning their struggle for world free-
dom, they keep faith with their birthright of natural beauty, as well
as with the wild creatures over which they have dominion and with
the earth that sustains them!

HAROLD J. COOLIDGE, Jr.

(for the Committee)
Washington, D. C.
June 12, 1942

Publication Committee:

CHARLES M. B. CADWALADER
HAROLD J. COOLIDGE, Jr.
ALEXANDER WETMORE

CONTENTS

PAGE

FOREWORD vii

INTRODUCTION , 1

NORTH AMERICA AND THE WEST
INDIES

Order INSECTIVORA : Moles, shrews, and their relatives 5

Family Nesophontidae : Extinct Antillean insectivores 5

Family Solenodontidae : Solenodons 10

Family Soricidae : Shrews 14

Order CHIROPTERA: Bats 15

Family Phyllostomidae : Leaf-nosed bats 16

Family Natalidae : Long-legged bats 29

Family Vespertilionidae : Simple-rosed bats 30

Order EDENTATA: Edentates 34

Family Dasypodidae: Armadillos 34

Family Megalochnidae : Ground sloths 36

Order RODENTIA: Gnawing mammals 41

Family Sciuridae: Squirrels 42

Family Castoridae : Beavers 49

Family Cricetidae : Hamsterlike rats and mice 87

Family Echimyidae : Spiny rats and their relatives 99

Family Heptaxodontidae 125

Family Dinomyidae: Giant rodents 126

Family Dasyproctidae : Agoutis 128

Order CARNIVORA : Dogs, cats, and their relatives 134

Family Ursidae : Bears 135

The black bears 135

The grizzly bears 139

Family Mustelidae : Weasels and martens 165

The pine martens 165

Family Canidae: Wolves, dogs, foxes 194

The kit foxes 194

The American wolves 199

Family Felidae: Cats 233

The pumas 234

The jaguars 252

xii EXTINCT AND VANISHING MAMMALS

PAGE

Order ARTIODACTYLA : Even-toed ungulates 256

Family Cervidae: Deer 257

The wapiti, or American elk 257

The caribou 296

Family Antilocapridae : Pronghorns 322

The pronghorn antelope 322

Family Bovidae : Sheep, goats, cattle 329

The muskoxen 329

The American bison 337

The bighorn, or mountain sheep 349

SOUTH AMERICA

Order EDENTATA: Edentates 375

Family Megatheriidae : Giant ground sloths 375

Order RODENTIA : Gnawing mammals 377

Family Cricetidae : Hamsterlike rodents 377

The Galapagos rice rats 377

Family Myocastoridae : Coypus 383

Family Dinomyidae: Giant rats 386

Family Chinchillidae : Chinchillas 389

Order CARNIVORA: Dogs, cats, and their relatives 396

Family Ursidae: Bears 396

Family Canidae: Wolves, dogs, foxes 400

Order PERISSODACTYLA : Odd-toed ungulates 403

Family Tapiridae : Tapirs 404

Order ARTIODACTYLA : Even-toed ungulates 406

Family Camelidae: Camels, llamas 406

Family Cervidae : Deer 412

OCEANIC MAMMALS

Order CARNIVORA: Dogs, cats, and their relatives 417

Family Mustelidae: Weasels, martens, otters 417

Order PINNIPEDIA: Seals, sea-lions, walruses 424

Family Otariidae: Sea-lions, fur seals 425

Family Phocidae : Hair seals 447

Family Odobenidae : Walruses 469

Order CETACEA: Whales, porpoises, dolphins 477

Family Iniidae: Fresh-water dolphins 479

Family Delphinidae : Dolphins and porpoises 480

Family Ziphiidae: Ziphioid whales 481

Whaling 484

Family Physeteridae : Sperm whales 492

Family Rhachianectidae : Gray whales 495

CONTENTS Xlll

Order CETACEA Continued.

Family Balaenidae : Right whales 499

Family Balaenopteridae : Fin whales 513

Order SIRENIA : Sea-cows 528

Family Dugongidae: Dugongs 528

Family Hydrodamalidae 535

Family Trichechidae : Manatees 538

BIBLIOGRAPHY 553

INDEX . .589

ILLUSTRATIONS

PAGE

Hispaniolan solenodon (Solenodon paradoxus] 11

Jamaican hutia (Geocapromys brownii) 108

St. Vincent agouti (Dasyprocta albida) 131

Kermode's bear (Euarctos americanus kermodei) 138

Barren-ground bear (Ursus richardsoni) 164

Eastern pine marten (Maries americana americana) 173

Fisher (Maries pennanti) 180

Black-footed ferret (Mustela nigripes) 185

Long-eared kit fox (Vulpes macrotis macrotis) 198

Eastern wapiti (Cervus canadensis canadensis) 264

Newfoundland caribou (Rangifer terraenovae) 321

White-faced muskox (Ovibos moschatus wardi) 335

Wood bison (Bison bison athabascae) 348

Badlands bighorn (Ovis canadensis auduboni) 360

Peruvian chinchilla (Chinchilla lanigera brevicaudata) 395

Antarctic wolf (Dusicyon australis) 402

Vicuna (Vicugna vicugna) 411

Sea otter (Enhydra lutris) 423

West Indian seal (Monachus tropicalis) 455

Northern elephant seal (Mirounga angustirostris) 461

Lesser rorqual (Balaenoptera borealis); humpback whale (Megapiera
novaeangliae) ; Greenland whale (Balaena mysticetus) ; California

gray whale (Rhachianectes glaucus) 512

Australian dugong (Dugong australis) 531

Steller's sea-cow (Hydrodamalis stelleri) 537

Manatee (Trichechus manatus) 543

INTRODUCTION

THE accounts that follow are intended to cover, first, those
species of New World mammals that have been extermi-
nated or seriously depleted by human agency during the past
four and a half centuries; and second, the oceanic species of
world-wide distribution, mainly cetaceans, seals, and sea-lions,
that within the historic period have suffered similar reduction,
some of them over a longer term. The method followed is to
give the common name, the current Latin name, reference to
original description with type locality, the important syno-
nyms, and references to figures of the exterior or the skull, or
both. Following a brief description there are given the geo-
graphic range past and present said a short description of each
species from the viewpoint mainly of its economic or com-
mercial use, present status, protective measures, and other
facts pertinent to a better understanding of its control or
encouragement or other special needs. The introductory para-
graphs to each of the major groups were prepared by Dr. J.
Eric Hill, of the American Museum of Natural History, and
are signed with his initials.

Full acknowledgment is due Dr. Francis Harper for the large
share of the bibliographic work for the present volume that he
had already accomplished in connection with his work on the
Old World species preliminary to the preparation of the final
accounts, and which he has handed on to me. In laying out
the work Dr. Harper selected for inclusion those mammals
that have been wholly or in part extirpated or have been greatly
restricted in range or numbers by clearing and settlement or
by intensive hunting for food, fur, or other economic and com-
mercial purposes. In a search through literature and by ex-
tensive correspondence Dr. Harper had already accumulated
a considerable mass of information concerning the past and
present status of many of these species, and of this I have made

2 EXTINCT AND VANISHING MAMMALS

full use. To the list of species he selected I have added a few
others that, as the work progressed, seemed to require treat-
ment. Possibly others should have been included, such as the
three-banded armadillo, which though still found in parts of
Brazil, seems now to have an interrupted range and may have
been locally exterminated for food. Possibly, too, various
species of mammals have been in part exterminated by pre-
historic peoples, if the evidence now accumulating is to be
trusted that primitive weapons found in association with
skeletons of extinct species of bison and mammoth in parts of
our West are proof of the killing of some of the survivors of
these species by human contemporaries. There is some evi-
dence, though of a rather uncertain nature, that in southern
South America prehistoric man was contemporary with species
of ungulates, now extinct, the remains of which date from the
Pampean time of late Pleistocene. Such animals have been
omitted, however, since the present inquiry does not extend so
far back into the past.

Although few modern species of New World or marine
mammals have been directly exterminated by man, it is never-
theless true that almost any of those having a considerable
value for food, fur, oil, or other products, as well as those
having a less tangible interest as game animals for trophies,
are likely to be endangered increasingly with the more intensive
methods of modern times and with the growth of populations
and consequent reduction of areas available for their support.
With the settlement of this hemisphere by Europeans, the
control measures they imposed on the native mammals were
naturally directed first against the larger predators wolves,
pumas, bears that menaced their meager flocks or even at
times endangered their own lives. Clearing of forests for
agricultural purposes had an immediate effect in reducing the
available food and shelter for sylvan species and thus indirectly
in driving out or eliminating some of these. Probably this
factor has been a potent one in the extinction of various fruit-
eating bats of the West Indian islands. The growing demands
of the fur trade, with its increasing inroads on the fur bearers,
have resulted in great reduction of native stocks in many re-
gions, even those remote from civilization where professional
trappers have for long periods pursued the more valuable
species unremittingly. The large game animals, which fur-

INTRODUCTION 3

nished dependable supplies of meat and hides, were drawn
heavily upon during pioneer stages and later were intensively
pursued, in part for sport and in part for incidental food pur-
poses. Finally the oceanic species, particularly the larger
cetaceans and seals, have been hunted with increasing vigor
and improved devices the world over, until at the present day
many of the regions formerly productive now no longer repay
the effort or expense of pursuit.

Protective legislation in aid of certain species of commercial
or economic importance, began early, as in colonial New
England, but in most cases did not come till much later, some-
times too late to be of much avail. On the other hand, there
have been some mammals that have proved more adaptive or
have responded more quickly to a release of the pressure put
upon them, and with proper management may prove a de-
pendable asset for a long time to come. Any facts that will
serve to bring out the needs of those species worth conserva-
tional treatment should be of value as a guide to future manage-
ment of these natural resources. The last decade has seen much
progress in these matters and a wider awareness that the time
has come when immediate steps <are necessary to preserve such
assets before their depletion has gone too far.

A long step forward in the preservation and protection of
American wildlife was made when, on October 12, 1940,
through the agency of the Pan American Union, representa-
tives of the nations of the Western Hemisphere affixed their
signatures to the Convention on Nature Protection and Wild-
life Preservation in the American Republics, thus pledging
their countries "to adopt measures for the protection of useful,
harmless and ornamental species of plant and animal life.
They have thus given formal recognition to the fact that many
such species know no national boundaries, and that true con-
servation of the gifts of Nature should begin before these
resources have been dissipated by thoughtless or selfish de-
struction" (from the statement of Dr. Hector David Castro,
Minister of El Salvador). The Senate of the United States
ratified this convention on April 7, 1941. In the "Annex" to
the Convention, three mammals are included the protection
of which is declared to be of special urgency and importance in
the United States and is to be as nearly complete as possible,
namely, the woodland caribou, the sea otter, and the manatee

4 EXTINCT AND VANISHING MAMMALS

(see Congressional Record Senate, for April 7, 1941). Similar
lists are to be submitted by the other countries. 1

GLOVER M. ALLEN
February, 191+2.

1 For a full account of this and other related matters see HAYDEN, SHERMAN STRONG,
The International Protection of Wild Life, an examination of treaties and other agree-
ments for the preservation of birds and mammals, 246 pp. Columbia University Press,
1942.

NORTH AMERICA AND THE
WEST INDIES

Order INSECTIVORA: Moles, Shrews, and Their Relatives

T NSECTIVORES are primitive mammals of moderate or
1 small size, with long pointed muzzles and sharp-cusped
chewing teeth. They feed chiefly on insects and other inverte-
brates, but some prey on small mammals or fish. The New
World insectivores belong to the suborder Lipotyphla and
include two groups differing in important dental characters.

The archaic insectivores, with V-shaped molar teeth, are
represented in the West Indies by two families; all are either
extinct or threatened with extinction:

(1) Nesophontidae, five primitive species from the islands of
Puerto Rico, Haiti, and Cuba. These are extinct, so far as
known.

(2) Solenodontidae, two species of long-snouted heavy -bodied
insectivores, found in the islands of Haiti and Cuba.

The more modern insectivores, with W-shaped molar teeth,
are represented by two families, both of which are widely
distributed in the Old World and North America, but only one
extends into Central or South America:

(3) Talpidae, moles. None of these are discussed here.

(4) Soricidae, shrews. A single species is considered to be
in danger of extinction because of rarity. J. E. H.

Family NESOPHONTIDAE: Extinct Antillean Insectivores

NESOPHONTES EDITHAE Anthony
Nesophontes edithae Anthony, Bull. Amer. Mus. Nat. Hist., vol. 35, p. 725, 1916

("Cueva Catedral, near Morovis, Porto Rico").
FIGS.: Anthony, H. E., 1916, pi. 23 (cranium); 1918, pp. 366-375, figs. 17-23 (skull

and larger bones).

Our first knowledge of this remarkable genus and species is
due to Dr. H. E. Anthony, who, during his investigation of cave
deposits in Puerto Rico, brought to light abundant remains of

6 EXTINCT AND VANISHING MAMMALS

the species that he named Nesophontes edithae, including many
nearly complete skulls and other parts of the skeleton from
caves at Morovis and Utuado. It apparently lived on the
island until comparatively late, perhaps post-Columbian,
times.

This is the largest of the species at present known from the
Greater Antilles, but since no living examples have been found,
only the skeletal characters have been described. It was
probably about the size of a chipmunk. In life, the snout was
probably elongate and flexible, the head long and slender, the
eye small, the limbs only moderately long and probably with
five clawed toes on each; the thorax seems to have been narrow
instead of widened as in Solenodon, and the tail was probably
about as long as the body. The tubular skull lacked the jugal
bone. The anterior pair of incisors is not greatly enlarged as
it is in that genus; the canine is double-rooted and retains its
primitive large size, while the upper molar teeth have high
cusps, forming a V-shaped pattern. Only one of the original
four premolars is missing so that the tooth formula is: if, CT,
pmf , ml = 40. The humerus shows the entepicondylar fora-
men, the pubic bones are not united below, and the tibia and
fibula are nearly separated, becoming slightly united distally
in adults. "No living insectivore has such an assemblage of
generalized characters," a fact that makes the group of unusual
interest. It was made by Anthony the type of a separate
family, Nesophontidae, and is regarded by Winge as somewhat
ancestral to the moles. The skull length is about 40-44 mm.
The abundant remains found by Anthony represent a larger
and a smaller size, the former of which are taken to be of males,
the latter of females, an unusual difference among the insecti-
vores.

Among the rugged and tree-covered hills of Puerto Rico, this
mammal must once have been a common species and doubtless
was preyed upon by owls, through the agency of which its
remains were probably brought into the limestone caves where
they have been found. Why this large species should be found
on this island alone, while on Cuba and Hispaniola only small
forms occur, or how it originally reached these islands at all, are
questions still unsolved. Nor is it clear that the introduction
of rats from Europe, or the burning of forests with intensive
cultivation, is alone responsible for its disappearance. Pre-

NORTH AMERICA AND THE WEST INDIES 7

sumably it lived up till the earlier post-Columbian times.
Hitherto no representative of the genus is known from Jamaica
or other West Indian Islands except Cuba and Hispaniola.
That so primitive and interesting a type should have wholly
died out within Recent times is to be regretted.

NESOPHONTES MICRUS G. M. Allen

? Nesophontes micrus G. M. Allen, Bull. Mus. Comp. Zool., vol. 61, p. 5, pi., fig. 14,
Jan. 1917 ("Cavern in the Sierra of Hato-Nuevo, Province of Matanzas, Cuba").

FIGS.: Allen, G. M., 1917, pi., fig. 14 (jaw); 1918, pi., figs. 7-10; Anthony, H. E., 1919,
pi. 37, figs. 8, 9, 11, 12 (crania).

NESOPHONTES LONGIROSTRIS Anthony

Nesophontes longirostris Anthony, Bull. Amer. Mus. Nat. Hist., vol. 41, p. 633, 1919

("Cave near the beach at Daiquiri, Cuba").
FIGS.: Anthony, 1919, pi. 37, figs. 10, 13 (cranium).

These two Cuban Nesophontes are at first glance very similar,
though N. longirostris has a slightly longer rostrum, and the
upper premolars are spaced instead of being in contact. They
are both known from skeletal remains only, the former from
both western and eastern Cuba, the latter from the type locality
in the eastern part of the island.

These closely similar species differ remarkably from N.
edithae of Puerto Rico in their very much smaller size. Other-
wise they are much alike, although a number of minor charac-
ters separating them have been pointed out by Anthony (op. cit. 9
p. 633), such as the greater size of the first as compared with
the second upper premolar, whereas in N. edithae these are
equal; again, in the lower jaw, the second premolar is only
about half as large as the two others instead of equaling them.
Length of palate from postpalatal notch to alveolus of first
incisor, about 14.1-15.3 mm. in N. micrus; 14.7 in N. longi-
rostris, contrasted with 20.5 in N. edithae. Cranial length of
N. micrus, 28 mm.

Nothing is known of N. longirostris except that it occurred
in eastern Cuba up to relatively recent times. Since this end
of the island still harbors various species that are absent from
the western end, it is likely that this was a local species, differ-
ing only slightly from N. micrus, the remains of which are now
known from recent cave deposits on the Isle of Pines as well as
from both eastern and western Cuba itself. No doubt, as

8 EXTINCT AND VANISHING MAMMALS

Dr. Anthony has suggested, the remains of these insectivores
in the cavern deposits are due to the agency of owls, such as
the Cuban barn owl, which even at the present time is building
up in some of these caves deposits of bones of introduced rats
as well as of bats. These bones, especially skulls, are more or
less broken so that complete crania are rare among the deposits.
Anthony adduces some evidence that the Nesophontes were
abundantly represented in the deposits prior to the deposition
of rat bones, so that the introduction of the rats may have
been the decisive factor in extermination of the insectivores.
Thus the latter may have been extinct only since white occupa-
tion.

NESOPHONTES PARAMICRUS Miller

Nesophontes paramicrus Miller, Smithsonian Misc. Coll., vol. 81, no. 9, p. 3, 1929

("Large cave near St. Michel, Haiti").
FIGS.: Miller, 1929a, pi. 1, fig. 1 (skeletal remains); 1930, pi. 2, figs. 2, 3 (rostrum).

NESOPHONTES HYPOMICRUS Miller

Nesophontes hypomicrus Miller, Smithsonian Misc. Coll., vol. 81, no. 9, p. 4, 1929

("Deep cave near the Atalaye plantation [near St. Michel], Haiti").
FIGS.: Miller, 1929a, pi. 1, fig. 2 (skeletal remains); 1930, pi. 2, fig. 1 (rostrum).

NESOPHONTES ZAMICRUS Miller

Nesophontes zamicrus Miller, Smithsonian Misc. Coll., vol. 81, no. 9, p. 7, 1929 ("Large

cave near St. Michel, Haiti").
FIG.: Miller, 1929a, pi. 1, fig. 3 (skeletal remains).

These three insectivores, apparently extinct and known only
from skeletal remains found in Hispaniolan caves, may be dis-
cussed together. They have no common names.

They are differentiated from one another largely on size
characters. In body dimensions they approximated mice or
small rats.

"Insectivores of the genus Nesophontes are abundantly
represented in the Haitian caves. They have not previously
been recorded from the island of Hispaniola. In the superficial
layer of the cave floors the bones of these animals occur in un-
disturbed material along with remains of Epimys rattus [roof
rat] and Mus musculus [house mouse]. This association is so
intimate that there appears to be no reason to doubt the
simultaneous occurrence of the insectivores and the introduced
rodents. Some of the jaws of Nesophontes are more fresh in

NORTH AMERICA AND THE WEST INDIES 1)

appearance than some of the jaws of Epimys near which they
were found. Whether or not Nesophontes now exists alive is
a question which for the present cannot be answered. No bones
of insectivores have been found in any of the numerous fresh
owl pellets which I have examined. It seems not improbable,
however, that if any part of the island remains uninvaded by
the roof rat, the native animal might now be found to exist
there" (Miller, 1929a, p. 3).

Wetmore and Swales write (1931, p. 238) : "It seems probable
that the deposits [made by the extinct owl in the Haitian Caves]
were accumulated over a long period of years extending perhaps
from four hundred to two thousand or more years ago."
They record the finding of skulls of Nesophontes in a supposed
nesting site of this owl in a sink-hole on La Selle, Haiti.

Miller also records (1929c, p. 4) remains of N. paramicrus
and N. hypomicrus from a deposit made by the extinct owl,
Tyto ostologa, in a cave on the islet San Gabriel in the Samana
Bay region of the Dominican Republic.

The occurrence of remains of these insectivores with those
of the introduced roof rat and house mouse indicates their
survival to post-Columbian times, while their usual absence
from fresh owl pellets suggests their extinction.

In the following year, Miller (1930) found bones of both
these species in material collected earlier in 1930 about 10
kilometers southwest of Constanza, Dominican Republic.
This consisted of a mass of owl pellets, apparently those of the
small living barn owl, found "in a shelter under an overhanging
ledge about 100 feet up the northern flank of Monte Culo de
Maco" in the rain-forest region where disintegration may take
place much sooner than in the drier parts of the island. Some
of the bones were so fresh as to retain bits of dried tissue while
the brain case of one of the skulls was found to be "packed full
of hair by the action of the owl's stomach." From this evidence
Miller concludes that these insectivores as well as two associ-
ated genera of rodents may still exist in that region. It is
hoped that future exploration may confirm this suggestion.

10 EXTINCT AND VANISHING MAMMALS

Family SOLENODONTIDAE : Solenodons

HlSPANIOLAN SOLENODON
SOLENODON PARADOXUS Brandt

Solenodon paradoxus Brandt, Mem. Acad. Imp. Sci. St. Petersbourg, ser. 6, vol. 2, p.

459, 1833 (Haiti).
FIGS.: Brandt, 1833, pis. 1, 2; Allen, G. M., 1910, pis. 1-3 (colored), 4-9 (anatomy).

First made known by the Russian zoologist Brandt, a little
over a century ago, this large insectivore was supposed to be
nearly or quite extinct in its island habitat until 1907, when A.
Hyatt Verrill found that it was still living in the interior of
northeastern Hispaniola.

It is a stoutly built animal, with a total length of about 22
inches, of which the strong tapering tail constitutes about 10
inches. Males and females are alike in size and in color, which
varies from a mixed blackish and buff dorsally, with clearer
yellowish sides, to a deep ferruginous tint, darkest and clearest
on the lower throat. Characteristic is a small squarish spot
of white in the middle of the nape. The tail is nearly naked
and dusky in color except at the base, which is flesh-color. The
feet and distal part of the limbs are nearly hairless and pale-
tinted. The muzzle is provided with a long cartilaginous
snout, supported at its base by a small round bone (the os
proboscidis). The feet are armed with strong, slightly curved
claws for digging, five on each foot. In its long tubular skull
it somewhat resembles Nesophontes and the Madagascar tenrec,
but the teeth show wide divergence from the former, in that
the anterior upper pair of incisors and the second lower incisors
are remarkably enlarged, the latter deeply grooved on the
inner side. This enlargement is suggested by the generic name
("sword tooth"). The upper and lower canines are somewhat
reduced, the former about equaling the second incisor. Except
for the loss of one premolar, the full placental tooth formula is
present: it, ci, pml, ml = 40.

Although this solenodon is said by Verrill to go under various
names in Hispaniola, as orso, hormigero, juron, these are merely
Spanish equivalents for "bear," "anteater," and "ferret."
Hearne (1835, p. 105), however, in presenting a specimen to the
Zoological Society of London in 1835, declared that in Haiti it
was known to the people as agouta. Verrill writes that in the
Dominican Republic it is called milqui.

NORTH AMERICA AND THE WEST INDIES

Following its rediscovery by Verrill in 1907, five living
animals were sent to the zoological gardens at Washington and
New York from the Dominican Republic, and others including
four living adults were received by the Museum of Comparative
Zoology at Cambridge. The last came from a locality in the
interior of the northeastern part of Hispaniola known as La
Vega. Since that time a German collector, Paul Thumb, by
the aid of a well-trained dog, succeeded in 1935 and 1936 in
securing a number, about 20 in all, which he shipped alive to
the zoological garden at Hamburg, Germany. Of these,
several survived the voyage and formed the subject of three
short papers by Dr. Erna Mohr (1936-37) concerning their
ways in captivity, with a number of photographic illustrations.
Herr Thumb secured his solenodons mainly in rocky and
wooded country in the vicinity of San Jose de las Matas. In
life the solenodon is a rather slow-moving creature, constantly
on the move with a shuffling gait, sniffing everywhere with its
long nose, and scraping and scratching here and there with its
long claws, exploring for food. It is capable of delivering a
sharp bite if too much disturbed, however.

According to Dr. Mohr (1937 A), Herr Thumb found soleno-
dons in dens among broken rock for the most part, or occa-
sionally in hollow logs. In the former case the depth varied
according to conditions of the rocks. The holes he dug out in
1935 were packed with the empty shells of a snail, but this was
not the case with those examined in the next year; while at the

Hispaniolan solenodon (Solenodon paradoxus)

12 EXTINCT AND VANISHING MAMMALS

Hamburg gardens, Dr. Mohr found the solenodons quite
indifferent to live snails presented to them. Thumb found as
many as eight solenodons in a single burrow. In captivity it
was ascertained that the female may have one to three, usually
two, young at a birth and may bring forth twice in a single
year, though showing no particular breeding season. The
group of eight in a den may thus be accounted for as the two
parents with their young of two successive litters. Three of
the families dug out by Thumb consisted of the adult pair and
their single young. A family secured for the Museum of Com-
parative Zoology in 1937 by W. J. Clench was of similar consti-
tution. These were from southwest of Sabana, Dominican
Republic.

At the present time it appears that this solenodon is in no
immediate danger of extinction, but is still present in areas of
stony forest in northeastern Dominican Republic. If, however,
such areas are extensively encroached upon through clearing
and cultivation, it will no doubt be reduced in numbers.

CUBAN SOLENODON; "AYRE"; "ALMIQUI"

SOLENODON CUBANUS Peters

Solenodon cubanus Peters, Monatsb. Konigl. Preuss. Akad. Wiss. Berlin, 1861, p. 169,

1862 (mountains near Trinidad and Bayamo, Cuba).
FIGS.: Peters, 1863, pis. 1-3 (exterior and anatomy); Mohr, 1937a, figs, on pp. 7, 8.

The Cuban solenodon was probably the animal first made
known by Oviedo as the ay re in his famous "Historia General y
Natural de las Indias," 1535, but it remained for the Cuban
naturalist Felipe Poey to bring it to the notice of modern
scientists, in a newspaper report in El Plantel in 1838, a
century ago. It was not till nearly 25 years later, however,
that Peters was able to compare a specimen with the Hispanio-
lan species and point out its distinctive characters.

In size this is slightly smaller than the species S. paradoxus,
about 20 inches long, tail about 7 inches, and is quite different
in color. In place of the blackish and buffy to ferruginous back,
the Cuban solenodon is blackish brown in color, with a varying
amount of white, or buffy, which may include only the base of
the muzzle, the cheeks, and a mark on each shoulder, as in a
specimen in the Museum of Comparative Zoology, or may be

NORTH AMERICA AND THE WEST INDIES 13

more extensive, to include the entire head and lower surfaces,
and extend back along the side of the neck to the shoulder,
whence it continues in scattered long white hairs to the
haunches. There is no distinct white mark on the nape.

The first definite knowledge of this species is due to Poey,
who obtained specimens in the mountains east of Bayamo,
where it was said to be well known, and he published an account
of it with a colored plate in his "Memorias sobre la Historia
Natural de Cuba" in 1851. He, however, supposed it to be the
same as the species of Hispaniola, and after search in the earlier
accounts of Cuban animals, failed to identify it with any known
to the early historians of the country. For this reason, he
proposed that it be called the almiqui, a name derived from
that of one of the mountains near where his specimens were
taken, in the region east of Bayamo. Gundlach subsequently
obtained it near Trinidad, about halfway of the length of the
island of Cuba, to the westward. In 1886, he sent three to the
United States National Museum that he secured from the high
mountains about 30 miles from Bayamo. One of these was
exhibited alive at a meeting of the Biological Society of Wash-
ington (Proc. Biol. Soc. Washington, vol. 4, p. x, 1886). It was
the surviving adult male of a family of three, of which the adult
female and her young one succumbed during the journey from
Cuba, and was said by Dr. F. W. True (1886) to be destined
for the zoological garden at Philadelphia. Even at that time
the animal seems to have become rare, for Gundlach wrote
that it had been secured only after "the promise of many
years" by the same person who had supplied specimens pre-
viously to himself and Poey. No more are known to have been
captured until 1909, when two Swedish engineers secured a
live one in the mountains of eastern Cuba. This animal died
soon after, and its remains were sent to the Riksmuseum at
Stockholm. Two photographs taken before its death are repro-
duced by Dr. Erna Mohr (1937a) and are apparently the only
ones that show the living animal. At about the same time a
second specimen (or possibly the same one) was in captivity
in the possession of M. Bofell, director of the Municipal
Museum of Santiago. According to Paul Serre (1910) it had
been captured by laborers near Baracoa, in extreme eastern
Cuba, in the course of road construction. Its owner is said to
have refused various offers to buy it, but its subsequent fate

14 EXTINCT AND VANISHING MAMMALS

is unknown. In the 30 years that have since elapsed I have
no knowledge of the capture of any others. Efforts of various
naturalists to discover it in eastern Cuba have been unavailing,
and even Herr Thumb, who was so successful in finding S.
paradoxus in the Dominican Republic with the aid of his dog,
failed to find it when in 1937 he hunted for it in that region.

Very little is known of its habits. It was believed to have
lived in caves, but, curiously, the researches of Dr. H. E.
Anthony and others in cave excavations have revealed very
little even in the way of recent remains, although bones of the
much smaller Nesophontes are common. The Museum of
Comparative Zoology, however, has a lower jaw from a cave
at San Lucas, near Maisi, eastern Cuba. This paucity of cave
fragments may indicate that the animal was little preyed upon
by owls, perhaps on account of its size and strength, but was
rather secure in holes among rocks such as the Hispaniolan
species makes use of. Captive animals are said to have eaten
the meat of chickens, either raw or cooked, while the one
captured by the Swedish engineers ate freely of raw beef given
in small pieces. With the introduction and spread of the Bur-
mese mongoose in Cuba, it is very possible that this voracious
carnivore may already have exterminated the last of the
solenodons. The loss of either of the two species is greatly to
be deplored, since they are in the New World the last living
representatives of the group of insectivores with triangular
instead of squarish molars, the Zalambdodonta, found else-
where at the present time only in Africa (the potamogale) and
on the island of Madagascar (the Madagascar hedgehog and
its relatives).

Family SORICIDAE: Shrews

WHITE MOUNTAINS DWARF SHREW

SOREX MYOPS Merriam

ftorex teneUus myops Merriam, Proc. Biol. Soc. Washington, vol. 15, p. 76, Mar. 22,
1902 (Pipers Creek [Cottonwood Creek], near main peak of White Mountains,
altitude 9,500 feet, Mono County, California).

FIGS.: Jackson, H. H. T., 1928, pi. 3, fig. B'; pi. 6, tig. R (skull).

Few specimens of this tiny shrew are known. In color the
back is drab, the lower parts pale smoke gray tinged with pale

NORTH AMERICA AND THE WEST INDIES 15

olive-buff (Jackson); tail buffy brown to tawny olive above,
below nearly "tilleul buff, darkening toward tip." It is said
to be thus paler than its near relatives of the ornatus group, and
its skull is smaller than in any except Sorex tenellus and S.
nanus of Owens Valley, California, and Colorado, respectively.
Total length, 98 mm., tail, 41; hind foot, 12; length of skull,
15.2.

This small shrew is regarded by Dr. H. H. T. Jackson (1928)
as a distinct species, though its relationship to S. tenellus is
doubtless close. For many years it was represented only by
the two specimens in the collection of the U. S. Biological
Survey. Dr. Jackson believed it was probably confined to the
White Mountains of California, "where it is seemingly rare
and may even be exterminated by the destruction of its habitat
through sheep grazing." More recently, however, Dr. William
H. Burt (1934) has discovered it in the Charleston Mountains
of southern Nevada, where he considers it common "along the
small streams at altitudes of 8,000 feet or above" up to 10,000
feet, not far from water. The possibility that grazing has
already much restricted its habitat in the White Mountains
makes its inclusion here excusable.

Order CHIROPTERA: Bats

The bats form a large complex order; they are the only
mammals that truly fly. There are two suborders :

Megachiroptera, fruit bats or flying-foxes. There is a single
family, found in southern Asia, Africa, Australia, and many of
the oceanic islands. They are characterized by blunt cheek
teeth and a second claw in the expanded wing-hand.

Microchiroptera, insectivorous bats (several are secondarily
fruit-eating or carnivorous and one family is specialized for a
diet of blood). Sixteen families are usually recognized, of
which six are restricted to the New World and two are com-
mon to both hemispheres. The New World families follow:

(1) Noctilionidae, primitive bats, found in tropical America.
They are related to the Old World Rhinopomidae and Embal-
lonuridae. Two genera are recognized; neither is here con-
sidered endangered.

(2) Phyllostomidae, New World leaf-nosed bats. These do
not appear to be closely related to any Old World group, but

16 EXTINCT AND VANISHING MAMMALS

probably developed in South America. Most of the species of
this large family have lancetlike nose-leaves; those that do not
show this character have cutaneous growths on the lower jaw.
Fifty-two genera are recognized, of which seven (including 13
species) are discussed in this volume, as either extinct or
extremely rare.

(3) Desmodontidae, vampire bats. The blood-drinking bats
are restricted to the American tropics. They are related to the
preceding family and have a rudimentary nose-leaf. None is
discussed in this work.

(4) Natalidae, long-legged bats. These small, delicately
built bats are found in the American tropics, north to the
Bahama Islands and central Mexico. One species is believed
to have become extinct in historical times.

(5) Furipteridae. Two genera, closely related to the long-
legged bats, are found in tropical South America. Neither
genus is discussed here.

(6) Thyropteridae. A single genus, with large suction disk
on the thumb, is restricted to tropical America. This bat is not
considered endangered.

(7) Vespertilionidae, common bats. They are found in both
hemispheres to the limit of tree growth and on the oceanic
islands to Samoa and Hawaii. Three species are here con-
sidered in danger of extinction.

(8) Molossidae, free-tailed bats. Found in the warmer parts
of both hemispheres. None of these bats is thought to be a
vanishing form. J. E. H.

Family PHYLLOSTOMID AE : Leaf-nosed Bats
PUERTO RICAN LONG-NOSED BAT

MONOPHYLLUS PRATER Anthony

MonophyUus frater Anthony, Bull. Amer. Mus. Nat. Hist., vol. 37, p. 565, 1917 (cave

near Morovis, Puerto Rico).
FIGS.: Anthony, H. E., 1917c, pi. 56, figs. 5, 6; 1918, p. 349, fig. 8 (skull).

Of the various genera of bats confined to the West Indies
and unknown from the neighboring mainland, the long-tongued
members of MonophyUus are remarkable, for they are found in
almost all the islands of the entire Antillean chain. Because
those on the several islands can usually be distinguished from

NORTH AMERICA AND THE WEST INDIES 17

their neighbors by minute characters, they have been given
distinctive names. Thus we have as the oldest named form,
M . redmani of Jamaica; others are M . cubanus of Cuba, M .
cubanus ferreus of Haiti, M. portoricensis of Puerto Rico, an
allied form on Antigua, M. luciae on Santa Lucia, M. plethodon
on Barbados, and M. clinedaphus from an unknown locality.
All are living, and no doubt vary in abundance on the several
islands. On Puerto Rico, however, in recent years, Dr. H. E.
Anthony has discovered a second and larger species, which he
has distinguished as M. f rater; on each of the other islands only
one form is known to occur.

Members of this genus of the subfamily Glossophaginae are
likely to be confused with Glossophaga of the mainland (and
some of the Antillean islands). They are small species with
long snouts surmounted by a small lancet-shaped nose-leaf
and are provided with a long extensible tongue for feeding on
fruit juices and perhaps in part upon nectar or pollen of
flowers. The teeth are peculiar in that the lower incisors are
minute and the cheek teeth are slightly elongate and spaced.
The cusps of the latter are low and the outer ones are placed
close to the edge of the crown "-so that the usual W-shaped
pattern of the cusps is obscured. The dentition includes, on
each side, two incisors above and below, a canine, two upper
and three lower premolars, and three molars above and below.
Externally it is easily distinguished from Glossophaga by the
longer tail, which is about half as long as the femur and pro-
jects beyond the edge of the narrow interfemoral membrane,
whereas in the latter genus the tail extends barely to the middle
of the wide interfemoral membrane.

The species M. f rater is known only from five fragmentary
skulls excavated in the large Cathedral Cave near Morovis,
Puerto Rico, by Dr. H. E. Anthony some years ago. With
these was associated a skull of the living species still found in
Puerto Rico, M. portoricensis, showing that the two were con-
temporaries and represented distinct species. M . f rater is
considerably larger than the latter, with a narrow elongate
rostrum. This larger size is obvious from the comparative
measurements: Length of the upper cheek teeth (alveoli),
6.8-7.1 mm. as against 5.3; in the smaller species, length of
palate 12.4-12.7 mm. as against 9.3. In spite of much collect-
ing done in Puerto Rico, no living examples of this larger

18 EXTINCT AND VANISHING MAMMALS

species have yet come to light. While it is quite possible that
it may eventually be found, it seems more likely that it is now
extinct, perhaps as a result in part of extensive clearing and
agriculture. As Anthony points out, the interesting thing
about this species is that it affords another instance of two
closely related species from the same region, differing most
obviously in little but size. A similar case is afforded by the
bats of the genus Chilonycteris, of which three species occur in
Cuba.

JAMAICA LONG-TONGUED BAT

REITHRONYCTERIS APHYLLA Miller

Reithronycteris aphyUa Miller, Proc. Acad. Nat. Sci. Philadelphia, 1898, p. 334

("Jamaica").
FIGS.: Miller, 1898, figs. 2b (head), 3 (skull), 4 (section of palate), 5a (tongue).

The present species seems most closely related to Phyllonyc-
teris, which it may represent in Jamaica. Externally and in
its dentition it resembles Phyllonycteris poeyi but is at once
distinguished by the vestigial nose-leaf, which instead of
forming an erect lancet is reduced to almost nothing and ap-
pears as a short piglike snout, set off by an encircling groove.
The ears, too, are shorter, and the tragus shows four notches
on its outer margin; the hind foot is large, three-fourths as long
as the tibia, and with strong claws; calcar absent, as in Phyllo-
nycteris; and tail shorter than tibia. Color light yellowish
brown (in alcohol) . The genus differs remarkably from Phyllo-
nycteris, however, in that the floor of the brain case is elevated
out of its usual position in such a way that the roof of the
posterior nares is formed "by two longitudinal folds given off
probably by the pterygoids and nearly meeting in the median
line." The skull has the rostral portion relatively broader,
and there is less contrast in both diameter and height of the
upper incisors. Total length, 88 mm.; tail, 12; tibia, 17;
forearm, 48; skull length, 26.

The only known specimen of this genus and species is the
type, a male preserved in the Museum of the Institute of
Jamaica. Its further history is not recorded, but it was taken
in that island some time prior to 1898; no fragments were found
by Dr. H. E. Anthony in the course of his explorations for
cave fossils in Jamaica, so that nothing is known of its haunts,
habits, or present status. Very likely, however, it is on the
way to extinction if it has not already gone.

NORTH AMERICA AND THE WEST INDIES 19

DOMINICAN HAIRY BAT

ARDOPS NICHOLLSI (Thomas)

Stenoderma nichollsi Thomas, Ann. Mag. Nat. Hist., ser. 6, vol. 7, p. 529, 1891 (Island
of Dominica, Lesser Antilles).

MONTSERRAT HAIRY BAT
ARDOPS MONTSERRATENSIS (Thomas)

Stenoderma montserratensis Thomas, Proc. Zool. Soc. London, 1894, p. 133 (Island of
Montserrat, Lesser Antilles).

ST. LUCIA HAIRY BAT
ARDOPS LUCIAE (Miller)

Stenoderma luciae Miller, Proc. Acad. Nat. Sci. Philadelphia, 1902, p. 407 (Island of
St. Lucia, Lesser Antilles).

GUADELOUPE HAIRY BAT

ARDOPS ANNECTENS Miller

Ardops annedens Miller, Proc. Biol. Soc. Washington, vol. 26, p. 33, Feb. 8, 1913
("Island of Guadeloupe, Lesser Antilles'^.

So far as known, the bats of this genus are confined to the
islands of the Lesser Antilles, where they appear to be rare, for
few specimens seem to be preserved in collections. They are at
present known to inhabit St. Lucia, Dominica, Guadeloupe,
and Montserrat only, and on each seem to be represented by a
slightly differing race. In time, with clearing of wooded areas,
their habitat is likely to be more and more restricted, wtth the
possibility of their final extermination; hence they may be
considered here collectively.

All the four known "species" are of similar appearance and
of medium size, about 3 inches long from nose to tip of tail,
with rather long loose pelage of a light-brown color. The fur
of the body extends thickly over the hind limbs to the ankles
and out along the border of the lateral membrane for a short
distance; on the arms the hair extends thickly out to the wrist.
The interfemoral membrane is very narrow, and there is a
short calcaneum. The head is short and blunt; the ears are
not much longer than the distance from their base to the end
of the nose. The genus is closely related to Phyllops and
Ariteus, of Cuba and Jamaica, respectively. It is smaller than

20 EXTINCT AND VANISHING MAMMALS

the species of Artibeus found in the same islands and differs
conspicuously in the skull by having the palate deeply emar-
ginate posteriorly and in the great development of the post-
glenoid process. In each jaw are two incisors, a canine, two
premolars, and three molars. The crown of the inner upper
incisor is short and thick, scarcely or not higher than long.
Additional details of the teeth are given fully by Miller (1907,
p. 151).

The first-described species, A. nichollsi, of Dominica, is
small, with a forearm of about 46 mm.; A. montserratensis, of
Montserrat, is slightly larger, with a forearm of 51.5 mm.; the
Guadeloupe form, A. annectens, is of intermediate size, with a
forearm of 48 mm.; while the most southerly one, A. luciae,
of St. Lucia, is very little larger than A. nichollsi, forearm 47
mm., and has a small white spot on the shoulder and distinctly
bifid inner upper incisors. If, as seems likely, the genus once
occurred on Martinique, it has probably now been extirpated
with clearing and volcanic destruction.

These bats are evidently tree-living, rather than cave-
haunting, which implies not only that they are somewhat
solitary and so easily escape observation, but also are more
difficult to find on account of living among the leaves by day.
Of the Montserrat species, Thomas records that it is said to
hang by day underneath branches of trees, and to do much
damage to the cocoa plantations. It seems likely, however,
that it can hardly be common enough to be very destructive,
and that the larger Artibeus is the real culprit. Of the Guade-
loupe form, Miller mentions five specimens, implying a small
group captured. Dr. G. K. Noble, who secured an adult
female and well-grown young in Guadeloupe in 1914, writes
me that he found these hanging together directly over the
path he was following through the woods near Sainte Rose.
Later his guide caught another one, so that they must occur in
some numbers still. Ardops is a member of the fruit-eating
group of leaf -nosed bats; hence it must depend for a living
upon the soft fruits of various forest trees, and these in turn
must occur in sufficient variety to provide a continuous supply
throughout the year. It is evident, therefore, that any impor-
tant change in the forest cover, whereby such trees are de-
stroyed or replaced by other kinds that do not produce the
desired fruit, must inevitably affect these local species. One

NORTH AMERICA AND THE WEST INDIES 21

may surmise that as already indicated, such changes have
taken place on Martinique and possibly too on St. Vincent,
where otherwise one might expect the genus to occur. Of its
status on the other Lesser Antillean islands practically nothing
is known.

FALCATE-WINGED BAT

PHYLLOPS FALCATUS (Gray)

Arctibeus falcatus Gray, Ann. Nat. Hist., vol. 4, p. 1, 1839 ("Cuba").

SYNONYM: Stenoderma albomaculatum Gundlach, Monatsber. Konigl. Preuss. Akad.

Wiss. Berlin, 1861, p. 155.
FIGS.: Dobson, 1878, pi. 28, figs. 3, 3a (dentition); Anthony, H. E., 1917b, pi. 34, fig.

3 (skull).

LESSER FALCATE-WINGED BAT
PHYLLOPS VETUS Anthony

Phyllops veins Anthony, Bull. Amer. Mus. Nat. Hist., vol. 37, p. 337, 1917 ("Cave at

Daiquiri, Province of Oriente, Cuba").
FIGS.: Anthony, 1917b, pi. 34, figs. 4-6 (skull).

HISPANIOLAN FALCATE-WINGED BAT

PHYLLOPS HAITIENSIS (J. A. Allen)

Ardops haitiensis J. A. Allen, Bull. Amer. Mus. Nat. Hist., vol. 24, p. 581, Sept. 11,

1908 (Cafta Honda, Dominican Republic).
FIGS.: Anthony, H. E., 1917b, pi. 34, fig. 2 (skull).

The small fruit-eating bats of this genus much resemble
those of the genus Ardops of the Lesser Antilles, but so far as
known they are confined to Cuba and Hispaniola. Their
superficial resemblance to the larger Artibeus, common in the
same islands, may at times have caused them to be overlooked,
yet from the paucity of known specimens they are probably
actually fewer in numbers, or even, in the case of P. veins, are
already extinct.

Similar in size to the members of the genus Ardops, and like
them with a short blunt rostrum, hairy limbs and wing-border,
they may be at once distinguished by the fact that the deep
emargination at the posterior part of the palate is continued
forward in a converging V-shaped outline, which extends about
to the level of the middle of the second molar, whereas in
Ardops this indentation is a narrow, parallel-sided arch. In
Phyllops, also, the crown of the inner upper incisor is slender,

22 EXTINCT AND VANISHING MAMMALS

noticeably higher than long, whereas in Ardops and in Ariteus
the crown of this tooth is short and thick, scarcely or not higher
than long (Miller, 1907). In contrast to the related genus
Stenoderma, the rostrum of the skull rises above the level of the
low supraorbital ridges and the nasal opening extends much
less than halfway back to the point of union of these ridges. In
addition, the third upper molar is still more minute, and the
second upper molar is scarcely three-fourths the size of the
first instead of nearly equaling it in crown area. In P. falcatus
the head and body measure about 1.9 inches; the tail is absent;
forearm, 1.65 inches; tibia, 0.6 inch; hind foot, 0.4 inch. The
fur is described as gray-brown, with dark tips; the lower side is
slightly paler.

The species described from Cuba as P. vetus is smaller than
P. falcatus in its skull, the palatal emargination is narrower,
and the small third upper molar is nearly circular instead of oval.
The Hispaniolan species, P. haitiensis, is of about the same
size as P. vetus but differs in the sudden contraction of the V--
shaped palatal emargination near its tip and in the apparent
lack of the shallow pits in the basioccipital.

Very little is known of the habits or status of these three
bats. Both P. falcatus and P. vetus are Cuban, while P.
haitiensis is at present known from Hispaniola only. The
first was sent by MacLeay to the British Museum, in 1838 or
1839, with the note, "Killed in my bedroom." Gundlach
mentions specimens from Matanzas and Cardenas, and in 1914
I examined two skulls taken from owl pellets in eastern Cuba
by Dr. Charles T. Ramsden. In 1917, Dr. H. E. Anthony, in
the course of excavations in a cave at Daiquiri, eastern Cuba,
secured seven fragmentary skulls and four mandibles. Al-
though he did not find the species alive, he mentions "an almost
perfect skull taken from an owl pellet collected by Mr. Barnum
Brown ... at the 'Cueva de los Machos' near Cienfue-
gos." "The material," he writes (1919), "is, for the most part,
fresh in appearance and not deeply discolored, some of it very
recent in fact." In addition, there is a specimen in alcohol in
the U. S. National Museum from Santiago, Cuba. Of P. vetus,
discovered by Dr. Anthony in the cave deposits at Daiquiri,
eastern Cuba, about 40 skulls were obtained. He states that
it was "found as a fossil and judging from the condition of the
specimens evidently has not been frequenting this region since

NORTH AMERICA AND THE WEST INDIES 23

the more recent animal remains, bats of the present day, were
deposited. Therefore, while it is quite possible that this bat
may be discovered living on some part of the island I am led to
believe that it is truly extinct, a fossil of an earlier period than
the very recent." Since, like its living relative, this was in all
probability a tree-dwelling and fruit-eating species, it may have
had a rather precarious foothold in this drier eastern part of
Cuba. In a note in his paper of 1919, Anthony adds that a
further comparison of the specimens of the two species, P.
falcatus and P. veins, confirms his belief that the latter was an
older inhabitant of the region for the bones "are all more
ancient in appearance, more deeply stained and discolored";
while the fact that the owls that prey upon these bats have not
brought them in to add to the more recent deposits is indicative
of the same thing.

Of the Hispaniolan species still less is known, for beyond
the original specimen from Cana Honda, Haiti, it has elsewhere
been recorded only from owl deposits. These are: near Con-
stanza, in the mountainous interior of the Dominican Republic,
where "in a shelter under an overhanging ledge about 100 feet
up the northern flank of Monte* Culo de Maco," Herbert W.
Krieger secured a broken skull and a mandible (Miller, 1930,
p. 6); in a large cave near St. Michel, ten skulls and several
mandibles; one skull from the deep cave and a mandible from
the crooked cave near the same place; and a skull from owl
pellets in the cave at Diquini, Haiti. These were found at all
levels from the surface to a depth of two feet (Miller, 1929a).

These small, heavy -bodied, fruit-eating bats may have been
easily taken by their enemy the barn owl. Probably the
species still persists in wooded regions of this island.

" DUSKY NASEBERRY BAT"

ARITETJS ACHRADOPHILUS (Gosse)

Artibeus achradophilus P. H. Gosse, Naturalist's Sojourn in Jamaica, p. 271, footnote

1851 (Content, Jamaica).
SYNONYMS: Artibeus sulphureus P. H. Gosse, op. cit., p. 271, footnote, 1851; Adieus

flavescens Gray, Proc. Zool. Soc. London, 1866, p. 117.
FIGS.: Gosse and Hill, 1851, pi. 6, fig. 4 (nose leaf); Peters, 1867, pp. 433-434, pi. 2.

For practically all we know of this bat, we are indebted to
P. H. Gosse who first described it in his "Naturalist's Sojourn
in Jamaica," the island to which it is confined.

24 EXTINCT AND VANISHING MAMMALS

The species is one of the group of West Indian fruit-eating
bats, of which the genera Ardops, Phyllops, and Stenoderma
are related members. Of about the same size and appearance
as these, it may at once be distinguished by the possession
of only two instead of three upper molars (the minute third
molar of the other genera having been lost in Ariteus) and by
the presence of a "minute though evident metaconid" in the
first lower molar. As in the related genera, the nose-leaf is
lanceolate with a distinct midrib; the short interfemoral mem-
brane is concave behind and like the legs is covered with
rather long hair, forming a fringe along the posterior margin.
It is said to be light reddish brown in color, paler beneath,
and with a small white patch on each shoulder. There are no
facial stripes, and the tail is lacking. Length of head and body,
2.2 inches; forearm, 1.6 inches; tibia, 0.6 inch; hind foot, 0.4
inch.

This genus may be regarded as the representative in Jamaica
of the genus Phyllops, from which the above characters dis-
tinguish it. Gosse (Gosse and Hill, 1851) writes that the first
specimen he secured came in at the open window of a house at
the Vineyard, near Black River, where he was staying. He
supposed that, like other bats that flew in and out, it was in
search of insects attracted by the lights. Three others he
succeeded in shooting in the early evening at Content, and
these are presumably the specimens listed by Dobson as in the
collection of the British Museum. Here a large and fruitful
naseberry tree (the nispero of the Spanish colonists) Achras
sapota attracted many bats by its large fruits, resembling
"a very rough russet apple, firm and fleshy, of a rich sugary
sweetness." At about a quarter of an hour after sunset the
bats began to visit the tree. "First one comes, takes a rapid
flight around the tree, darts once or twice through the dense
foliage, and winging away is lost in the light of the sky. Another
and another comes immediately, and performs the same evolu-
tions . . . By carefully following the flight of an indi-
vidual with the eye, we perceive that now and then he alights
for a moment on some object at the extremity of a bunch of
leaves ; but no sooner has the eye rested on the spot than the
sooty wings are again spread, and he is pursuing his giddy
course with his fellows. The object of his visit is a ripe nase-
berry, nestled in the midst of that rosette of leaves. Occa-

NORTH AMERICA AND THE WEST INDIES 25

sionally the weight of the suspended Bat dislodges the ripe
fruit, and it falls to the ground, splitting with the shock. On
picking it up we see that it has been just bitten, not gnawed as
by the rodent incisors of a mouse, but nibbled in a ragged
manner." No doubt several species of fruit-eating bats visit
these trees for the feast.

What may be the present status of this bat does not seem to
be known. However, it is evidently rare in collections, whether
from actual scarcity or through failure of visitors to capture it.
On account of its island habitat and the intensive agriculture
carried on in Jamaica, it is likely to become reduced in numbers
as time goes on.

STENODERMA RUFUM Geoffroy

Stenoderma rufum E. Geoffroy, Description de 1'Egypte, vol. 2, p. 114, 1818 (locality

unknown).
SYNONYM: Artibaeus undatus Gervais, in Exped. Amer. du Sud de Castelnau, Zool.,

Mamm., p. 35, pi. 9, fig. 3 (teeth), 1855.
FIGS.: Peters, 1876, pi. 1, figs. 1-3; Anthony, H. E., 1918, fig. 10 (skull and teeth).

This small fruit-eating bat was described 125 years ago from
a specimen in the Paris Museum, but unfortunately there was
no record of its place of origin. Since that time no other living
specimen has been taken, and its habitat has remained un-
known. In 1918, however, Dr. H. E. Anthony recorded the
discovery of over two dozen "good-sized fragments of skulls,
some nearly complete, " during his excavations in the Cathedral
Cave near Morovis, Puerto Rico. The presumption is thus
strengthened that the real home of the species was the Greater
Antilles, perhaps Puerto Rico.

In life this bat doubtless resembled somewhat the larger
Artibeus, found in the same region, but the skull differs mark-
edly in its short, broad rostrum, which, with its parallel and
short tooth rows, is nearly square as seen from below. The
median notch of the palate extends far forward to the level of
the middle of the first molar. The incisors and upper canines
form together a nearly transverse row, while the last molar,
both above and below, is very small. The first two upper
molars are broad for crushing and show little trace of the four
primitive cusps. Total length of skull, exclusive of incisors,
23.1 mm.; zygomatic width, 15.7.

Since no other living examples of this small fruit bat have
been found since it was first made known by Geoffroy, it is

26 EXTINCT AND VANISHING MAMMALS

believed now to have become extinct during the last century.
Anthony's interesting discovery of its remains in the Cathedral
Cave in Puerto Rico indicates not only that this island may
have been its home but also that it was fairly common there.
The curious shape of the rostrum, so short and square, with
heavy anterior molars, may indicate that it was adapted for
feeding on some special kind of fruit, the abundance of which
has been seriously affected through the clearing and burning
of the original forest cover with cultivation on the island.
Coincident with such restriction of food, the bat may have been
much reduced in numbers until its final extinction. Anthony
mentions the possibility that it may still be found on other
islands but has been overlooked on account of a certain resem-
blance to the common Artibeus.

PUERTO RICAN LONG-TONGUED BAT
PHYLLONYCTERIS MAJOR Anthony

Phyllonyderis major Anthony, Bull. Amer. Mus. Nat. Hist., vol. 37, p. 567, Sept. 7,

1917 (cave near Morovis, Puerto Rico).
FIGS.: Anthony, 1917c, pi. 56, figs. 1, 2; 1918, p. 356, figs. 12, a-d (skull).

HAITIAN LONG-TONGUED BAT
PHYLLONYCTERIS OBTUSA Miller

Phyllonyderis obtusa Miller, Smithsonian Misc. Coll., vol. 81, no. 9, p. 10, Mar. 30,
1929 ("Cave near the Atalaye plantation, about 4 miles east of St. Michel,
Haiti").

The bats of this genus are known only from Cuba, Puerto
Rico, and Hispaniola of the Greater Antilles, where slightly
differing local forms have developed on each island. On Cuba
the living species, P. poeyi, though apparently not very com-
mon, nevertheless occurs in numbers in suitable caves, as at
Guana jay, where Palmer secured a large series in 1900 (Miller,
1904), Baracoa, or in the Sierra de Hato Nuevo, where in 1917
Dr. Thomas Barbour found a large colony. On the other hand,
in Puerto Rico it is unknown in the living state, though found
in the superficial cave deposits, and the same is true of the
Hispaniolan form; both are therefore included here among
species nearly or quite gone.

These are small bats with long narrow skulls and long pro-
trusible tongues, which are useful in licking up fruit pulp and

NORTH AMERICA AND THE WEST INDIES 27

juices on which they largely feed. Probably pollen and flower
nectar are also eaten. A small hastate nose-leaf is present at
the tip of the snout, and the interfemoral membrane is narrow,
extending only to the middle of the tibia. The calcar is absent
altogether. In each jaw, both upper and lower, there are two
incisors, a canine, two premolars, and three molars. The last
are rather weak and rounded, the second and third lower
molars without cusps, a character contrasting with the condi-
tion in the related genus Erophylla, occurring in the same
regions. The Puerto Rican Phyllonycteris major is decribed as
closely related to the Cuban P. poeyi, but "noticeably larger,
with wider brain case and heavier dentition"; greatest length
of skull, 27-28 mm.; breadth of brain case, 11.4 mm. The
form from Haiti, P. obtusa, is described as very similar to the
Cuban species "but incisive foramina smaller and anterior
border of premaxillaries as viewed in palatine aspect less
narrowly curved." Greatest length of skull, 22.2 mm.; breadth
of brain case, 10.2 mm.

During the course of his investigations in Puerto Rico caves,
Dr. H. E. Anthony found no evidence that P. major is still
living in that island. In only one cave, that near Morovis, did
he find its remains, including about 60 skulls and many jaws.
This may indicate not only that it was a cave-dweller, like its
Cuban relative, but also that its choice of caves and perhaps
even its numbers were limited. "It was contemporaneous
with Nesophontes as well as with Stenoderma ruficm and other
bats to be found living today." Concerning the Hispaniolan
P. obtusa, Miller (1929a) writes that remains were found in
three distinct places: "The crooked cave near the Atalaye
plantation, St. Michel, Haiti," a skull and mandible; a skull
from a cave near Port-de-Paix; and a skull from owl pellets
found in a cave at Diquini. The last implies that the skulls
were fairly recent, and very likely the species may still be
found there living, although none of the various collections of
bats made in Hispaniola has hitherto yielded specimens. It is,
as in other cases, difficult to assign an exact reason for the
decline of such a species. Nevertheless, since it depends on
a constant supply the year around of various fruits and berries
in their season, it is obvious that any material changes in the
forest or shrub covering, if this eliminated or reduced any one
species of fruit even for a short critical season, might have a
disastrous effect.

28 EXTINCT AND VANISHING MAMMALS

Concerning the Cuban species, P. poeyi, Palmer (in Miller,
1904) found it "very abundant in a wet, ill-ventilated cavern
on Guanajay Mountain. On entering this cave, the vertical
opening of which, about 12 feet across, was concealed by
bushes, we descended about 25 feet, and were then standing
some 20 feet above the lowest level. The slight noise which we
made disturbed the bats in the inner chambers, and we could
distinctly hear the rumbling made by their wings. As we
proceeded this sound increased, until, when we reached the
inner and thickly populated chambers, it became a grand,
rushing roar of thousands on thousands of wildly flying animals.
To reach the inner chamber it was necessary for us to descend
from the first landing to the real floor of the cavern, and there
light our candles, for not a ray of light and very little fresh air
penetrated so far. From the floor we worked our way over
many guano-covered, damp bowlders and through arches and
narrow passages up to a sloping shelf, where, owing to the low
roof, a man could not stand upright. By this time the bad
air and excessive warmth was [were] telling on us, and we were
in a most profuse perspiration. The bats were now thoroughly
aroused, and the noise of their wings was astounding. Many
were darting out through the passage by which we had entered
. . . We began swinging a dip net in every direction,
trusting to chance to secure specimens. About fifteen minutes
of such work usually resulted in the capture of 20 to 30 bats,
nearly all of this species . . . Before June 7, all the females
were big with a single young, but after this date we found the
pink, almost hairless little ones of different sizes hanging to the
roof and scattered over much of its surface. On our last visit,
late in June, the cave was so hot as to be unbearable .
Among the specimens captured at the mouth of the damp cave
near Baracoa . . . were many of this species . . . On
one side of the vertical opening of this cave grew a large tree
whose roots descended like a stream into the cavity. The
people of the neighborhood assured me that the majas (the
Cuban boa, Epicrates angulifer) coil themselves among these
roots and grab at the bats as they fly out. I was told that a
snake frequently secures a bat in this manner."

NORTH AMERICA AND THE WEST INDIES 29

Family NATALIDAE: Long-legged Bats
CUBAN YELLOW BAT

NATALUS PRIMUS Anthony

Natahis primus Anthony, Bull. Amer. Mus. Nat. Hist., vol. 41, p. 642, Dec. 30, 1919
("Daiquiri, Cuba," in the Cueva de los Indies).

The bats of the family Natalidae are small and delicately
formed and have noticeably long and slender legs and distinctly
funnel-shaped ears. The ample interfemoral membrane is
supported in part by the tail, which extends to its border. In
distribution the members of this family are confined to tropical
America and occur on some of the West Indies, to which three
of the genera are confined: Chilonatalus (on the Bahamas,
Greater Antilles, and Old Providence Island), Nyctiellus (the
Bahamas and Cuba), and Phodotes (island of Curagao). The
genus Natalus is the only one occurring on the mainland from
tropical South America north to central Mexico, as well as in
both the Greater and the Lesser Antilles. In the latter it is
known living from the island of Dominica (N. dominicensis
Shamel) and Antigua; in the former, from the Dominican
Republic (Natalus major Miller). It is thus especially inter-
esting that in the course of his cave investigations in eastern
Cuba, Dr. H. E. Anthony should have found the present species
represented by mandibles associated with Nesophontes and
Boromys in the Cueva de los Indios at Daiquiri, "buried but a
short distance under the surface. The bone is stained a very
dark brown and probably represents an extinct form." From
N. major, of the Dominican Republic, hitherto the largest known
member of the genus, "it may be known by its even greater
length of mandible and noticeably heavier teeth. The first
lower molar is especially 'plump' in contour and the tooth
extends externally considerably beyond the alveolar border."
The mandible on which the description is based measured 14.4
mm. in greatest length or about a millimeter more than in the
large form of the Dominican Republic, of which it may be
regarded as the Cuban representative. The color in life was
perhaps yellowish, as in some of its near relatives.

Nothing further is known of the species, but since it has not
been found living by any of the various collectors who have
searched for bats in Cuba, it is probably, as Anthony says,
extinct.

30 EXTINCT AND VANISHING MAMMALS

In this connection it may be surmised that the present
sporadic distribution of this genus in the West Indies, with one
living form on Dominica, one in Hispaniola, and a third prob-
ably recently extinct in Cuba, implies that it may once have
been more widespread over this area and has already died out
on other islands. Of this, however, there is at present no
evidence. The other related genera, too, may be thought of as
in a restricted situation. The little Nyctiellus, one of the
smallest of bats, and formerly believed to be confined to Cuba
and the Isle of Pines, has in late years been found in numbers
in some of the Bahamas (G. M. Allen and Sanborn, 1937) and
is probably in no present danger. The somewhat larger
Chilonatalus is found, in several closely allied races, on Cuba,
Jamaica, the Bahamas, and on Old Providence Island, dwelling
in caves. Wherever its small colonies exist it does not seem
especially difficult to find and may be fairly safe for the present,
since it is insectivorous like the others of the family, and there
would seem to be no lack of sustenance for it. Of the genus
Phodotes, a close relative of Natalus, practically nothing is
known beyond Miller's original description based on a speci-
men from the island of Curagao, off the coast of Venezuela.
If, as supposed, it is peculiar to that island, its future status
may well be somewhat precarious on account of changes taking
place with intensive cultivation.

Family VESPERTILIONIDAE : Simple-nosed Bats
SPOTTED BAT; "DEATH'S-HEAD BAT"

EUDERMA MACULATUM (J. A. Allen)

Histiotus maculaius J. A. Allen, Bull. Amer. Mus. Nat. Hist., vol. 3, p. 195, 1891
(near Piru, Ventura County, California, probably at mouth of Castac Creek).

FIGS.: Miller, 1897a, pi. 1, fig. 11 (ear); pi. 3, fig. 3 (wing); 1907, p. 226, fig. 37 (skull);
Bailey, V., 1931, pi. 21, B; Grinnell, H. W., 1918, pi. 16, fig. 9 (photograph).

For 12 years after its first discovery in Ventura County,
Calif., this bat remained unique; then a second one was found
in the biological laboratory of the College of Agriculture and
Mechanic Arts at Mesilla Park, N. Mex. Since that time only
six additional specimens have been discovered, all in the South-
western United States. This apparent rarity has led to the
belief that the species is actually a waning one, and for this
reason it is here included.

NORTH AMERICA AND THE WEST INDIES 31

This is a fairly large bat with a total length of about 110 min. ;
tail, 50; forearm, 50. In color the fur is deep black, with three
large white spots: one at each shoulder and one on the lower
part of the back, giving a striking contrast. The lower side is
washed with white. The very long ears measured 34 mm. in
one specimen. The two upper and two lower premolars, the
nostrils simple instead of opening upward or in connection with
nasal swellings, and the distinctive color and long ears will
serve to identify it easily.

An account of the skeleton with figures of the important
bones has been given by E. R. Hall (1934), who in a later
note (1939) adds an eighth to the list of known individuals
taken. The localities are: Piru, Ventura County, Calif.;
Mesilla Park, N. Mex.; Yuma, Ariz.; Mecca, Kern County, and
Yosemite Valley, Calif.; Reno, Nev.; and Salt Lake City,
Utah (see Durrant, 1935, p. 226). In every instance only a
solitary individual has appeared, and in such unexpected
places as clinging to a fence rail, or the under side of a rock,
under the eaves of a schoolhouse, in a biological laboratory (!),
or on the side of a building. One was found dead in the overflow
of a railway water tank. From these few instances it has not
been possible to deduce what is the natural resting place of
the species by day. Possibly the one found clinging to the
lower side of a rock indicates more nearly the sort of place they
normally seek for shelter. The area outlined by these eight
known instances takes in roughly the arid region of the Great
Basin of the Southwestern United States, to which probably
the species is confined. Other than the scanty information
supplied by the eight specimens hitherto discovered, nothing
whatever is known of its habits. It is, however, not at all
impossible that it may somewhere be found in greater numbers.
Its evidently solitary disposition and its liking for outside
resting places make it seem unlikely that it is a colonial and
cave-dwelling bat.

Whether this is a species in danger of eventual extermination
can not now be told with certainty, nor does it appear that
anything can at present be done to foster it.

32 EXTINCT AND VANISHING MAMMALS

GALAPAGOS ISLANDS RED BAT
LASIURUS BRACHYOTIS (J. A. Allen)

Atalapha brachyotis J. A. Allen, Bull. Amer. Mus. Nat. Hist., vol. 4, p. 47, Mar. 25,
1892 ("Chatham Island," Galapagos Islands).

This, the only bat at present known from the Galapagos
Islands, is one of the red-bat group, found on the American
continents from north-temperate to south-temperate latitudes.
It is a smaller relative of the hoary bats of the American
continents and Hawaii and like them has a hairy tail-membrane,
short rounded ears, and only a single upper incisor.

The Galapagos bat is said to resemble closely its mainland
representative in its rusty-red coloring and minutely frosted
tipping to the hairs, but the ears are smaller and the wings
slenderer, the dentition lighter, with shorter canines. Forearm
length, 39 mm.

Of this bat almost nothing is known. Dr. George Baur,
who collected the type specimen about 50 years ago, wrote
at the time of its capture (Allen, 1892): "On Chatham Island,
at an elevation of about 1700 feet, where the hacienda is placed,
we observed bats nearly every evening." The specimen was the
only one he succeeded in capturing, and it is at present in the
Museum of Comparative Zoology (its skull, however, unfor-
tunately lost). Dr. Baur adds: "Bats have been observed on
Indefatigable Island by Dr. Habel, and I observed one on
South Albemarle." It is, of course, uncertain whether all
these represented the same species of bat. Red bats are
known to reach Bermuda in their migrations in eastern North
America, so that it is not surprising that this strong-flying
species should have reached the Galapagos at some distant
period. Whether it will continue to survive the arid conditions
and restriction of forest growth on the islands can not at present
be told.

HAWAIIAN HOARY BAT; "OLEPE"

LASIURUS SEMOTUS (H. Allen)

Atalapha semota H. Allen, Proo. U. S. Nat. Mus., vol. 13, p. 173, Sept. , 1890 (Kauai,

Hawaiian Islands).
SYNONYM: Lasiurus grayii in part, J. E. Gray, Proc. Zool. Soc. London, 1862, p. 143.

This is the Hawaiian representative of the continental
hoary bat, which as a species ranges from north-temperate and
sub-boreal latitudes to south-temperate climates, thus covering

NORTH AMERICA AND THE WEST INDIES 33

a wide latitudinal extent. It is migratory in northern United
States and southern Canada; hence perhaps one may assume
that the ancestor of the Hawaiian form reached that group of
islands by flight over sea at some time in the distant past,
coming from the American continent, Although at one time
believed to be the same as the continental L. grayii of Chile,
the Hawaiian form is darker, and in the red phase lacks to a
large extent the hoary tips to the hairs of the upper surface
found in that race and in the North American L. cinereus.

In the Hawaiian hoary bat the color of the fur above may
be either gray as in L. cinereus or dark red, almost blackish
brown, with the middle portions of the hairs dull whitish,
showing through if the fur is disturbed. On the lower flanks,
legs, and interfemoral membrane above, the color of the back
in the red phase passes gradually into a bright chestnut or
mahogany red, with practically no white tipping to the hairs.
Below, the color is paler, with chin and upper throat buffy
white, and an indistinct half-collar of darker, set off by white
tips to the hairs, from the dull-brown belly. Fur along the
under side of forearm and at base of fingers clear buffy. A
minute tuft of short white hairs #t base of thumb and of digit
5, and a third white tuft on the inner side of the forearm just
beyond the elbow. Forearm measures about 50 mm. (in the
original description it is given as 40, but G. S. Miller, Jr.
(1939) shows that this is a misprint for 50 as shown by an
examination of the lectotype in the U. S. National Museum).

This bat was first reported from the Hawaiian Islands by
J. E. Gray in 1862, who, however, believed it identical with
the Chilean race, which Tomes had earlier named Lasiurus
grayii. Harrison Allen in naming it as a distinct form, had
eight specimens from the "Sandwich Islands," only two of
which had the more precise locality of Kauai. Perkins (1903,
p. 465) writes that it chiefly frequents the mountains where,
presumably, more forested areas remain. He adds that it is
locally common in the uplands of Hawaii, and he has seen it,
though rarely, on the islands of Kauai and Oahu. Bryan,
writing in 1915, says: "They have always been rare, but are
apparently still to be seen in the uplands of Hawaii." There
are four specimens in the museum at Tring, England, from
Hawaii taken in 1891, and the Museum of Comparative
Zoology has a specimen taken in 1937 at Waimea, island of

34 EXTINCT AND VANISHING MAMMALS

Hawaii, by Miss Barbara Lawrence, who says that the boys
sometimes find them hanging among the fronds of tree ferns
in the lowland valleys.

While at present the species may be in no particular danger,
it is likely that any considerable changes, such as reduction in
the amount of sheltering tree growth, will affect the species
adversely. As the only bat at present known to inhabit this
group of islands, this species is of particular interest, and its
obvious derivation from the American Continent makes it
further noteworthy from the distributional viewpoint.

Order EDENTATA: Edentates

The New World edentates belong to the suborder Xenarthra.
Two groups are recognized:

(1) Cingulata, the armadillos and their relatives. A race of
the widely distributed nine-banded armadillo is restricted in
its range to a small West Indian island and may be easily
exterminated.

(2) Pilosa, sloths, ground sloths, and American anteaters.
Of these the ground sloths have become entirely extinct; five
species are thought to have persisted into historic times in
Puerto Rico, Haiti, and Cuba. J. E. H.

Family DASYPODIDAE: Armadillos

GRENADA ARMADILLO
DASYPUS NOVEMCINCTUS HOPLITES G. M. Allen

Dasypus novemcinctus hoplites G. M. Allen, Bull. Mus. Comp.. Zool., vol. 54, p. 195,
July, 1911 ("Hills back of Gouyave, island of Grenada," West Indies).

This is a small race of the continental nine-banded armadillo,
but it may be included in this account since any island race
represented by a relatively small population is liable to further
reduction if in the course of years the conditions change
greatly through fire, settlement, or other causes that may
further restrict the available habitat.

Externally this is a somewhat smaller replica of the ordinary
nine-banded armadillo, with the usual external bony cuirass
consisting of a shoulder shield, a rump shield, and nine inter-
mediate half-rings on the body. The tail is covered with a

NORTH AMERICA AND THE WEST INDIES 35

series of 12 complete bony rings, and the color in life is light
flesh-color. From the posterior free edges of the transverse
body-rings project three or four short bristles from each scale
and similar but smaller bristles are present at the posterior
margins of the scales on the shields of body and tail. On the
under side of the body there are transverse rows of small round
scutes, each in the center of a cluster of yellowish bristles. The
skull in addition to being smaller than that of the mainland
representative has one or two less teeth in each series, usually
six or seven instead of eight. Total length, up to 678 mm.;
tail, 82; hind foot, 37; greatest length of skull, 85.5.

At the present day this armadillo is confined to the rough
country covered with rain forest on the hills in the central
part of the island of Grenada. Probably, too, the armadillo
found on the island of Tobago is very similar if not identical.
It is more or less hunted by the local Negroes who esteem its
flesh. They not only hunt the armadillos with dogs but catch
them in deadfalls set in the runways. Whether the armadillo
naturally occurred in these islands or whether it was introduced
there long ago is uncertain, but in either case it has been native
for a long enough time to allow its development into a distinct
dwarf race. De Rochefort, writing in 1658, mentions the
"Tatou" as a native of Tobago, especially calling attention
to the small size. Soon after, Du Tertre in the 1667 edition of
his work on the history of the French Islands, was the first to
record it from Grenada, where he says it was then common.
It was unknown on any of the other French islands, and
apparently all attempts to introduce it in Martinique from
Grenada had ended in failure. Labat in 1742, however,
mentions having eaten the flesh of armadillos that had been
brought to Martinique from Grenada, and in 1700 had several
times eaten it on the latter island.

At the present time I know of no reason to suppose that
this armadillo is in any danger of extinction. When I visited
the island in 1910, I found it well known to the people and a
Mr. John Branch assured me that often the Negroes might
capture several in a single night, hunting with dogs, in the
forest back of Victoria. The animal is fairly prolific and should
survive without difficulty if conditions remain substantially
unchanged.

36 EXTINCT AND VANISHING MAMMALS

Family MEGALOCHNIDAE : Ground Sloths
SMALLER PUERTO RICAN GROUND SLOTH

ACRATOCNUS ODONTRIGONUS Anthony

Acratocnus odontrigonus Anthony, Ann. New York Acad. Sci., vol. 27, p. 195, Aug. 9,
1916 ("Cueva de la Ceiba, near Utuado, Porto Rico").

FIGS.: Anthony, H. E., 1916, pis. 7-10, pi. 11, fig. 1; 1918, figs. 43-48, 50-54 (except
those of A. major'); pi. 69, figs, la-c; pi. 70, figs. 2-6; pi. 71, figs. 1, 3-6; pi. 72,
figs. 1, 3, 4; pi. 73, figs. 1-3, 5, 7; pi. 74, figs. 1-4 (skeletal parts).

LARGER PUERTO RICAN GROUND SLOTH
ACRATOCNUS MAJOR Anthony

Acratocnus major Anthony, Mem. Amer. Mus. Nat. Hist., ser. 2, vol. 2, p. 412, 1918

("Cave . . . near Utuado, Porto Rico").
FIGS.: Anthony, H. E., 1918, figs. 42, 43, 48, 49, 51, 53 (parts pertaining to No. 17169);

pi. 69, figs. 2a^c; pi. 70, figs. 1, 7, 8; pi. 71, fig. 2; pi. 72, figs. 2a-b, 5; pi. 73, figs.

4,6.

HISPANIOLAN GROUND SLOTH
ACRATOCNUS (?) COMES Miller

Acratocnus (?) comes Miller, Smithsonian Misc. Coll., vol. 81, no. 9, p. 26, Mar. 30,

1929 ("Large cave near St. Michel, Haiti").
FIGS.: Miller, 1929a, pi. 5, fig. 2; pi. 6, fig. 2; pi. 8, fig. 1; pi. 10, fig. 1 (limb bones and

atlas).

Among the most interesting finds of recent years is the dis-
covery of bones in the caves of Puerto Rico and Haiti repre-
senting a small ground sloth, which was evidently contem-
porary with aboriginal culture. It was approximately the
size of a two-toed sloth but represented the ground-living
group, known from the mainland of North America by the
gigantic mylodons of the Pleistocene era. The nearest rela-
tives seem to be the small ground sloths of the genus Hapalops
from Miocene formations of Patagonia, although there are
various differences. Their presence on both Puerto Rico and
Hispaniola is additional evidence of the faunal affinities of
these two islands, and since the three forms hitherto described
seem closely allied, they may be considered under a single head.

The smaller ground sloth of Puerto Rico was the first
described and is the best known, from abundant skeletal
remains found in the cave earth during excavations originally
undertaken in the course of archeological research. As in the
living tree sloths, the premaxillae were toothless, and the

NORTH AMERICA AND THE WEST INDIES 37

incisors and canines of the lower series are also lacking. The
skull is relatively long and narrow instead of globular as it is in
the tree-living types, the rostrum is short and broad, and there
is a well-developed median crest extending forward from the
wide occipital ridge to the frontal region, where a lower branch
goes to each of the short postorbital projections. The large
tusklike upper canines are strongly triangular in section,
followed by a diastema of nearly the same length as the diam-
eter of the tooth or somewhat longer, then four cheek teeth in
two parallel rows, each tooth approximately oval in section,
with the broader end on the lingual side. In the lower jaw,
there is a large tusklike first premolar corresponding to the
upper canine, then a space, followed by three cheek teeth
much like those of the upper series. The humerus has a
slender shaft, a wide distal expansion, and a large entepicondy-
lar foramen. The femur is wide in front view, but not thick,
and shows a low third or outer trochanteric ridge. The tibia is
short and stout and the fibula is free. The calcaneum or heel
bone is broadly hatchet-shaped, and the terminal phalanges
were evidently provided with stout claws. The tail is believed
to have been short but rather stout. Dr. Anthony gives the
following dimensions of the major skeletal elements: Skull
length, tip of rostrum to occipital condyles, 136.2 mm.; breadth
of brain case, 28.5; greatest breadth across postorbital proc-
esses, 47.5; alveolar length of upper molar series, 35.5; longest
humerus, 145; ulna, 160; radius, 128.8; longest femur, 163.3;
longest tibia, 128.

Among the remains of this species recovered in caves in
Puerto Rico were a few indicating a similar sloth of slightly
larger size, which Dr. Anthony in a later communication
decided must represent a second and larger species of the same
genus. This he named Acratocnus major, and defined it as
being very similar to A. odontrigonus but "larger and heavier
and with different skull characters"; it has "proportionally a
much broader muzzle and an elevated basioccipital region,"
and a larger upper canine. No complete measurements of the
skull are available, but the breadth across postorbital processes
is 66 mm.; length of ulna, 171; length of tibia, 133. By com-
parison with the corresponding measurements given above for
the smaller species, the obvious size difference may be appre-
ciated.

38 EXTINCT AND VANISHING MAMMALS

The Haitian species, A. (?) comes, was based on a femur from
a large cave and represents an animal much like A. odontrigo-
nus, weighing perhaps 50 pounds. The bone differs from the
corresponding one of the latter in having the neck of the
articular condyle shorter and less bent outward and forward
so that it diverges less noticeably from the general contour of
the shaft. In describing the species, Miller refers to it a few
other fragments, including several caniniform teeth from
Gonave Island, off the coast of Port-au-Prince, Haiti, but the
complete skull is as yet unknown.

In Puerto Rico, remains of this type of ground sloth occur
in dry caves in the limestone of the mountainous parts of the
island. " Caves up on the sides of small hills yielded the most
bones and the size of the cave was immaterial. The cave that
contained the greatest amount of Ground Sloth material was a
small one with the entrance on a rather steep hillside and
opening out on a sheer front of limestone. Inside, the cave
did not open out very wide but had a deep fissure at the left
toward which the floor sloped abruptly. This fissure was richly
packed with bones of the sloth and the large rodent [Elasmo-
dontomys], the bones beginning at near the surface and con-
tinuing down some nine feet when excavations had to be given
up because of the impossibility of reaching any deeper. This
cave had the appearance of a trap for any animal that wandered
into it and certainly would have proved so for any old or sick
animal that had strength enough to crawl up through the
cave entrance."

Concerning the remains of this genus and a larger sloth
found in Hispaniola by Miller, the latter states (1929a) :
"That one or both of these sloths continued to exist on the
island until after the advent of man I have no doubt. The
facts that have led me to this conclusion are as follows : (a) In
the two caves near St. Michel most of the sloth remains were
found within two feet of the surface; and human bones and
pottery occurred to the same depth without any indication
that they had been dug in. (b) Near the entrance to the
smaller of the two main caves bones of ground sloths (certainly
two and perhaps more individuals) were inextricably mixed
with bones of man (adult and infant) and domestic pig. The
remains were scattered among the small fragments of limestone
which made up the greater part of the floor material, and I was

NORTH AMERICA AND THE WEST INDIES 39

unable to determine any definite level-relationship among
them, (c) Near the entrance to the large cave I unearthed
with a trowel, in fine, soft, undisturbed material at the bottom
of a trench two feet deep, the femur of a ground sloth, and,
about 18 inches from it, a fragment of coarse dark pottery . .
the bone and pottery had every appearance of having been
deposited on the former surface of the cave floor and subse-
quently covered by the gradual accumulation of detritus . .
In general the ground sloth bones were associated with the
human remains in exactly the same manner as the bones of
Isolobodon and Plagiodontia, rodents which are positively
known to have been contemporary with man." The external
appearance of these animals may only be conjectured, for no
fragments of hide or hair are known to be preserved.

LARGER HISPANIOLAN GROUND SLOTH
PAROCNUS SERTJS Miller

Parocnus serus Miller, Smithsonian Misc. Coll., vol. 81, no. 9, p. 29, Mar. 30, 1929

("In a large cave near St. Michel, Haiti").
FIGS.: Miller, 1929a, pi. 7; pi. 8, fig. 2; pi. 9; yl. 10, figs. 2, 3.

LARGE CUBAN GROUND SLOTH

MEGALOCNUS RODENS Leidy
Megalocnus rodens Leidy, Proc. Acad. Nat. Sci. Philadelphia, 1868, p. 180 (Cuba).

Miller, in describing the larger ground sloth from Haiti, be-
lieved there was good evidence that it lived to be the contem-
porary of man, since its remains were found in close association
with pottery and in layers of cave earth at no great depth.
Judging from the size of the few bones found, he surmised it to
have been an animal weighing upward of 150 pounds and more
heavily built that Acratocnus. The femur, on which the de-
scription mainly rests, is at once distinguishable by the absence
of the lesser trochanter, "as well as by its greater size and the
much more noticeable antero-posterior flattening of the upper
portion of the shaft." No skulls have yet been found suffi-
ciently well preserved to give much idea of the cranial characters.
An important feature of the humerus seems to be the lack of an
entepicondylar foramen, of which no trace is visible in the
figure published by Miller. The humerus measures 200 mm.

40 EXTINCT AND VANISHING MAMMALS

in greatest length. Three calcanea are essentially similar to
the calcaneum of Mylodon.

While Acratocnus and Parocnus did not perhaps become ex-
tinct till fairly recent times, possibly not until after the dis-
covery of the West Indies by Europeans, the Cuban ground
sloths may have died out at an earlier time. Their remains are
known in some abundance in deposits around a warm spring
at Ciego Montero, associated with bones of a crocodile and a
giant tortoise. Another important locality is the Casimba
in the Sierra de Jatibonico, in central Cuba, a fissure spring
at the bottom of a ravine, where many bones have been
found. Although a full account of these remains was in
preparation by the late Dr. W. D. Matthew and Prof. Carlos
de la Torre, this seems to have been indefinitely postponed
through the death of the former. Nevertheless, in pre-
liminary papers (Torre and Matthew, 1915; Matthew, 1918,
1931) they distinguish no less that four genera, to which names
are given. Matthew (1918, p. 660) restricts Megalocnus to the
largest of these, "about the size of a black bear"; the smallest
is Microcnus, "about the size of a cat, and there are two of
intermediate size, Mesocnus, with a rather long, narrow muzzle,
and Miocnus, with a broad, square muzzle." To this last
genus Acratocnus of Puerto Rico and Hispaniola is said to be
related. Possibly, too, Parocnus may be found to be closely
related to one of the two intermediate genera or even identical
with it, when better material is assembled and restudied. The
three other genera, Megalocnus, Microcnus, and Mesocnus,
have large tusks of a "peculiar dished shape, with a tendency
to approach toward each other like the incisors of rodents. "
Matthew regards the Cuban ground sloths and the Puerto
Rican Acratocnus odontrigonus as descendants of a common
ancestral type of Upper Miocene or Lower Pliocene age, related
to that of Megalonyx. Matthew adds that "there is so much
individual and age variation in ground sloths that it is difficult
to say how many species of these genera are present. Mega-
locnus occurs abundantly both at Ciego Montero and the
Casimba, but apparently only one species at Ciego Montero,
while at the Casimba there may be two or three. The species
found at the eastern end of the island agree better with the
Casimba forms than with the Ciego Montero species. Of the
smaller forms there are clearly two species of Mesocnus, but I

NORTH AMERICA AND THE WEST INDIES 41

see no proof of more than one of Miocnus or Microcnus." So
far nothing seems to have been discovered that would indicate
that ground sloths in Cuba were contemporaneous with early
man.

Order RODENTIA: Gnawing Mammals

The rodents, here considered to include the rabbits and
pikas, are the largest order of mammals, both in number of
species and of individuals. They are characterized by chisel-
like incisor teeth, usually a single pair above and below, but
with a second pair of upper incisors in the rabbits. They are
mostly small mammals, with large populations and rapid rates
of reproduction. There are four suborders:

(1) Lagomorpha (Duplicidentata) : The rabbits and hares,
found in all zoogeographical regions, except the Australian,
and pikas of Asia and western North America. Rabbits have
been introduced into Australia and elsewhere.

(2) Sciuromorpha : This includes the squirrel family, which
is world wide, except for the oceanic islands and the Australian
Region; the scaly -tailed African "flying squirrels," the African
springhaas; the beaver, found in Europe, northern Asia, and
North America; and the sewellel of the Pacific coast of North
America.

(3) Myomorpha: The rats, mice, jerboas, and their relatives;
representatives of this group occur in all regions, although
human agency is doubtless responsible for their presence on
some oceanic islands.

(4) Hy stricomorpha : The porcupines, cavies, chinchillas,
and their relatives; representatives of this group are found in
all regions, except the Australian and the oceanic islands.

The vanishing or extinct North American rodents are the
following :

(1) Sciuridae, squirrels, four forms of a single genus.

(2) Castoridae, beavers, eighteen races of the single Ameri-
can species.

(3) Cricetidae, native rats and mice, eight species of three
genera.

(4) Echimyidae, hystricoid spiny rats, twenty species and
subspecies of nine genera.

(5) Heptaxodontidae, the unique species, thought to have
become extinct about the time of Columbus.

42 EXTINCT AND VANISHING MAMMALS

(6) Dinomyidae, giant hystricoid rats, two species of two
genera.

(7) Dasyproctidae, agoutis, three insular species of the
typical genus.

The last four families belong to the New World Hystrico-
morpha and have their relatives in South America, where they
probably originated. J. E. H.

Family SCIURIDAE: Squirrels

KAIBAB, OR WHITE-TAILED, SQUIRREL

SCITJRUS KAIBABENSIS Merriam

Sciurus kaibabensis Merriam, Proc. Biol. Soc. Washington, vol. 17, p. 129, 1904 ("Head
of Bright Angel Creek, top of Kaibab Plateau, north side of Grand Canyon of
Colorado, Arizona").

FIGS.: Nelson, 1918, p. 448, lower fig. (col.); Goldman, 1928, p. 127, pi. 16 (photo-
graph).

This is perhaps the handsomest of the North American
squirrels, a close relative of Abert's squirrel, with long-tufted
ears, which ranges from Colorado into Mexico. The white-
tailed squirrel has become isolated on the north side of the
Grand Canyon in the high Kaibab Plateau and has developed
striking color characters, such especially as its large white tail.
Because of its restricted habitat and conspicuousness it may be
included here.

In its general appearance it resembles Abert's squirrel, but
the under parts are mainly black instead of white, and the tail
is practically all white instead of on the under side only. The
back is dark grizzled gray, with a wash of ferruginous from
shoulders to rump. The ears are blackish in summer but
rimmed anteriorly with gray in winter; the ear tufts are black,
the face, fore feet, and toes mixed gray and black; hind feet in
summer mainly gray but in winter mainly black. The tail is
ample and bushy, white, with an indistinct buffy-gray stripe
down the middle of the upper side. Size about that of a gray
squirrel.

The Kaibab squirrel is found in a restricted area about 40
miles in length and 20 miles in greatest width, at the broadest
(southern) part of the Kaibab Plateau. Isolated here from its
nearest allies of the south side of the Grand Canyon, it is con-
fined to the portions of the plateau where the yellow pine

NORTH AMERICA AND THE WEST INDIES 43

(Pinus ponder osa) is found. It is largely dependent on this
tree for food and shelter as are its relatives of the Abert's
squirrel type, which, wherever they occur, "are rarely or never
seen beyond the local range of this tree. On the Kaibab
Plateau the altitudinal range is mainly from 7,000 to 8,500
feet" (Goldman, 1928). Major Goldman (1928) writes:
"Yellow pines are rarely hollow and I have never seen one of
these squirrels enter a hole. Dome-shaped nests, about 2 feet
in diameter, made of cut pine branches, are usually well con-
cealed among crowded limbs in the upper part of a tree. Like
most squirrels, the white-tail is most active during the early
morning and late afternoon, but may be found abroad at any
time of day. Practically nothing seems to be known of its
breeding habits." Although it depends partly for food upon
the seeds of the yellow pine, which it extracts by gnawing away
the scales, this supply is more or less precarious in years when
the crop is poor. Its main reliance is therefore the cambium
beneath the bark of the newer twigs of the yellow pine, which
provides an unfailing food supply. "In feeding, the leaf-
bearing branch tip is commonly severed and allowed to drop
to the ground. A subterminal section of tender stem 2 to 4
inches in length is cut off, scaled, and when the cambium sought
as food has been neatly removed, the peeled wood remaining
is also dropped ... In summer, however, they evidently
indulge in a more varied diet. Much time is spent in foraging
upon the ground and I have at various times observed them
feeding upon mushrooms and other funguses, and they doubt-
less eat many other things. "

Although it is known that these squirrels are occasionally
taken by hawks, Major Goldman believes that their numbers
are no more seriously affected by these enemies than are those
of other forms of Abert's squirrel. "The destruction of much
ground cover through overbrowsing by deer may, however,
expose the squirrels to some additional danger from hawks.
Squirrels everywhere vary in numbers from year to year owing
to causes imperfectly understood, but at present there seems
to be no great danger of the extinction of the white-tail. These
beautiful squirrels are objects of great interest to tourists along
the Grand Canyon highway that bisects the Kaibab plateau,
and efforts should be made to establish them in other yellow
pine forested areas. " He adds the caution that care should be

44 EXTINCT AND VANISHING MAMMALS

taken in such an enterprise, not to introduce them into other
areas already occupied by races of Abert's squirrel, since hy-
bridization would inevitably tend to the loss of the distinctive
and conspicuous characters of this particular form. At the
present time the Kaibab squirrel is completely protected by
law and should be fairly well safeguarded, but any influences
that might tend to impair the growths of yellow pine to which
it is restricted would react unfavorably on its welfare in the
limited area in which it is found.

BLACK MANGROVE SQUIRREL

SCIURUS NIGER AVICENNIA A. H. Howell

Sciurus niger avicennia A. H. Howell, Journ. Mammalogy, vol. 1, p. 37, Nov. 1919
("Everglade, Lee County, Florida").

There are few less-inviting places for mammals than a man-
grove swamp. The clean trunks and branches offer little in the
way of food or shelter; the soft ooze from which the trees grow
is daily washed by the tides and thus is unsuitable for ground-
livers ; while the suckerlike stems and rooting tips form a tangle
almost impossible for the larger species to penetrate. It is
therefore the more remarkable that a depauperate form of the
southern fox squirrel has made the mangrove belt of the south-
west coast of Florida its home and become adapted to a life in
this apparently unfavorable environment.

The mangrove squirrel is described as smaller and darker
(more tawny) both above and below than the typical race,
with the feet clearer white and less tinged with buff. The type
is in the so-called buff phase or at least is not of the entirely
black sort, with white face and ears. The head and back are
black sprinkled with cinnamon but more abundantly on lower
back; sides shading to orange-cinnamon; nose, lips, front of
face, and ears white; fore legs blackish washed with orange-
cinnamon, the feet and toes white; hind legs with more orange-
cinnamon than black, the feet edged with white. Tail orange-
cinnamon above, mixed with black, and shading on the sides
to hazel; under side rich tawny, with a submarginal band of
black. Under parts dull orange-cinnamon washed on throat
and breast with black and white. A black phase also occurs,
in which the nose, ears, paws and sometimes the end of the
tail are white, while elsewhere the pelage is deep shining black,

NORTH AMERICA AND THE WEST INDIES 45

with a few pale hairs on the under side. Total length of an
adult male (the type), 535 mm. (21.25 inches); tail vertebrae,
260 (10.25 inches); hind foot, 75 (about 3 inches); skull, occip-
itonasal length, 65.5.

This squirrel is confined to "the damp, dark forests of black
and red mangrove which extend practically without a break
from Marco Pass to Cape Sable and around the southern end
of the peninsula" of Florida, "to the shores of Biscay ne Bay
on the east coast." Here it is apparently not common, for
Ho well writes that " several days spent in hunting through these
mosquito-infested forests resulted" in only a brief glimpse of
one, while the type specimen he secured through an Indian boy
who knew where its home tree was located. A small series of
specimens, however, was secured by W. S. Brooks in the spring
of 1920, for the Museum of Comparative Zoology, mostly in the
black phase. I am indebted to Dr. Thomas Barbour for a few
notes on this squirrel. He tells me that it is well known to the
residents of the region, though not regarded as common, and
that it may be found on even some of the isolated keys grown
up to mangroves. Its feeding habits are remarkable for, in-
stead of depending on nuts and other seeds, it lives on the buds
and bark of the mangroves, gnawing the latter from twigs
which it cuts. Dr. Barbour reports that after the hurricane
that passed over this region a few years ago its numbers seem
to have been lessened. Evidently it is very difficult to secure
satisfactory estimates of the size and extent of this squirrel
population, since the swamps can hardly be penetrated except
by canoes along small waterways. While this race of fox
squirrel may be in no very immediate danger at the present
time, it might easily be affected by any large-scale attempts to
alter the local ecologic conditions.

BRYANT'S Fox SQUIRREL; PENINSULA Fox SQUIRREL
SCIURUS NIGER BRYANTI H. H. Bailey

Sciurus niger bryanti H. H. Bailey, Bull. Bailey Mus. and Library of Nat. Hist., no.
1, p. 1, Aug. 1, 1920 ("Cambridge, Dorchester County, Maryland").

Dorchester County, Md., lies on the southwestern side of
the large peninsula, nearly cut off from the mainland on the
east by Delaware Bay and Delaware River and on the west by
the long estuary formed by Chesapeake Bay. Isolated in the

46 EXTINCT AND VANISHING MAMMALS

area near the tip of this peninsula is a small population of fox
squirrels, which differ from the eastern fox squirrels in colora-
tion by lacking the buff and russet tints. They form a local
race, named by H. H. Bailey in honor of the ornithologist
Walter E. Bryant.

In a small series of specimens of this squirrel the coloration is
remarkably uniform. The entire dorsal surface of the body is a
hoary gray, the result of abundant white-tipped hairs mixed
with black hairs. On the forehead the black hairs slightly pre-
dominate. The under surfaces from chin to root of tail are
white. Feet and ears dull white, sometimes with a faint tinge
of buffy. Tail mixed black and white both above and below,
the black forming a narrow and continuous submarginal
border on the under side.

Although, according to Bailey, the local gunners secure a
few each season in Dorchester County, this race is so localized,
with little or no likelihood of its increasing its range, that it
must be thought of as like an island form threatened with
gradual encroachment on its available habitat. For this reason
it is likely in time to become more and more restricted, with
little possibility of long survival unless given special protection.

Concerning the present status of this squirrel, I am indebted
for a few details to David V. Black, manager of the Blackwater
National Wildlife Refuge, Cambridge, Md., who in response to
inquiries writes, under date of March 21, 1941: "I have talked
with several persons in Dorchester County about the fox
squirrels found here, and they all seem to be of the opinion
that they are fairly abundant and are found in all parts of the
county. There is no particular protection given to Bryant's
fox squirrel. It is hunted indiscriminately with other squirrels
during the open seasons in the State. I do not think that it is
threatened with extermination as yet. Most of the timber
has been cut out in the county, and what little remains seems
to be in the process of cutting with small portable sawmills, so
that the natural habitat for squirrels and other woods dwellers
is continually diminishing. The squirrels do occur on the
refuge, where, of course, no hunting is allowed."

NORTH AMERICA AND THE WEST INDIES 47

EASTERN Fox SQUIRREL; NORTHERN Fox SQUIRREL

SCIURUS NIGER NEGLECTUS (Gray)

Macroxus neglectus Gray, Ann. Mag. Nat. Hist., scr. 3, vol. 20, p. 425, 1867 (Wilming-
ton, Delaware).

SYNONYMS: Sciurus vulpinus Schreber, Saugthiere, vol. 4, p. 772, 1792 (preoccupied
by Sciurus vulpinus Gmelin) (Baltimore, Maryland); Sciurus ludovicianus vicinus
Bangs, Proc. Biol. Soc. Washington, vol. 10, p. 150, 1896 (White Sulphur Springs,
West Virginia).

FIGS.: Audubon and Bachman, 1849, Quadrupeds of North America, vol. 1, pi. 17
(col.) (as Sciurus cinereus); Bangs, 1896, pi. 8, fig. 3 (skull).

The fox squirrels are typical of open heavy forests in the
eastern half of the United States and extend in various sub-
species from New England west to the plains region in South
Dakota, and south to southern Florida, and to Texas and
northeastern Mexico. In the more thickly settled areas of the
East their numbers have been greatly reduced locally, and
especially in the Northeast they have reached nearly the point
of complete extirpation. The typical form, Sciurus niger niger,
is still fairly common in the southeastern States, from southern
South Carolina to Florida, frequenting chiefly the open "piney
woods." The northeastern race*, S. niger neglectus, however,
is gone from its more northern outposts and is in danger of
complete wiping out east of the Alleghenies.

The eastern fox squirrel is slightly larger than that of the
Southeast and somewhat resembles a large gray squirrel. The
upper parts of the head and body are a grizzled buffy and black,
slightly darker on the forehead; cheeks, ears, forearms and
feet, the lower hind legs and hind feet pale rufous; under side
of body white; tail about as long as body, mixed black and
buffy above, rufous below and on the sides. Total length,
about 2 feet; tail, 11 inches; hind foot, 2.75 inches (73 mm.).
Distinguished from the more western race, rufiventer, by its
white instead of rufous belly, and from the southern typical
form by its rufous ears instead of white, and the usual lack of
black color, though rarely entirely black individuals occur.

The eastern fox squirrel probably in former times just
reached southwestern Connecticut but has long since gone.
Linsley (1842), in his list of the mammals of that State, knew
of it only from Northford, a century ago. In New York State
it was formerly found in the southern parts in some numbers,
as attested by Bachman in 1839. He mentions several lately

48 EXTINCT AND VANISHING MAMMALS

received from Orange County but adds that in northern New
York it is exceedingly rare, as he "only saw two pair during
fifteen years of close observation." Merriam (1884) called it a
rare or accidental straggler in the Adirondacks region and
mentions a specimen or two at Lake George in 1872 or 1873 as
the only recent instance he knew of, but there seems even
there to be some doubt whether or not it had been brought in
and escaped. In Rensselaer County he records two killed in
1854. At about the same time Dr. A. K. Fisher knew of its
having been killed in Westchester County, but adds that in
1896 none had been known "of late even in that wild region."
At the present time the fox squirrel is practically gone from
the State. In New Jersey it was present locally at least up to
about 1865 but was apparently gone by the nineties. In
Pennsylvania the species was formerly common in the lower
altitudes but is now much reduced. Rhoads (1903) says of its
status in the early years of this century that it was "probably
always rarer in Chester and Delaware counties and in southern
N. J. than in south central Pa. and northern N. J. Now
exterminated in N. J. but found occasionally in the Pa. counties
bordering the lower Susquehanna, also yet recorded from the
northwestern part of Pa." He quotes Todd that it was shot
"at rare intervals in some of the northern counties" of the
western border of Pennsylvania; and believed that it was
destined to extermination within the entire limits of the State
unless large areas of country in middle Pennsylvania "revert
to a wilderness condition or become game reservations under
state protection." Even in 1896 Bangs wrote that Dr. B. H.
Warren informed him that "the northern fox squirrel is
practically extinct in Pennsylvania except in the counties of
Dauphin and Cumberland" near the south-central part of the
State. In West Virginia, Kellogg (1937) reports that "it now
survives in the heavily wooded and sparsely settled higher
altitudes of the Allegheny Mountains" and that for many
years numbers were shipped to Center Market in Washington,
"from points in western Virginia and from eastern West
Virginia." Apparently in all these regions the numbers have
constantly and slowly dwindled in recent decades, for not only
is the species constantly in demand for food, but it is less
adaptable to the changing conditions of clearing and settle-
ment than its congener, the gray squirrel. It prefers primeval

NORTH AMERICA AND THE WEST INDIES 49

stands of old timber and is much given to foraging and traveling
on the ground. Hunters often employ dogs to tree the squirrels,
after which they may be shot with little difficulty. At the
present time it may be called locally uncommon in less settled
areas from south-central Pennsylvania to the mountains of
central Virginia and West Virginia.

Family CASTORIDAE: Beavers

Beavers of the genus Castor are the only living members of
this family of rodents; they are of circumboreal distribution
in the watered and forested regions of the north-temperate
zone. Since prehistoric times they have been of great impor-
tance in the economy of the human race, chiefly for their fur
and their meat. The Old World beavers (Castor fiber and
races) are regarded as a species distinct from the North Ameri-
can species (C. canadensis), but the actual differences are slight,
as in the proportionate depth of the rostrum (less in the Old
World animal) and the length of the nasal bones (extending
back of the lacrimal level more in the Old World animal).
Of the New World beaver, a nurfber of slightly marked races
have been described, depending on minor differences in color
or in the size or the relative shapes of some of the cranial bones.
Everywhere as the country has been opened and settled the
beaver has been so reduced or extirpated that its present num-
bers and distribution are but a fraction of their original extent.
In various eastern localities reintroductions have taken place,
with varying success. The literature on the American beaver
is extensive, for probably few of our native mammals have been
more often written about. In the following account of the
various races only an outline of the main points in the history
can be given, and in many cases it is hardly possible to obtain
accurate details of recent date. The various races may be
taken up in alphabetical sequence.

CANADIAN BEAVER; NORTHEASTERN BEAVER

CASTOR CANADENSIS CANADENSIS Kuhl

Castor canadensis Kuhl, Beitrage zur Zoologie, 1820, p. 64 (Hudson Bay).
FIGS.: Radford, 1908, 10 unnumbered pis. (col. exterior, photos, skull, works); Nelson,
1916, p. 443 (col.); Elliot, 1901, p. 115, fig. 27 (skull).

50 EXTINCT AND VANISHING MAMMALS

Beavers are stoutly built, the northeastern form weighing up
to 60 pounds. The short, rounded ears and broadly webbed
hind feet are adaptations to aquatic life; the coarse, brown
overfur sheds water and protects the short, dense, and plush-
like underfur. The broad, flattened tail is practically bare and
covered with scales. Swimming is done with the powerful hind
feet, while the small fore feet are held against the breast or
used in transporting material. The skull is stoutly built,
rather triangular in dorsal aspect, with broad nasals that end
slightly behind the level of the upward branch of the premaxil-
lary. In general color the eastern beaver in fresh pelage is
dark chestnut-brown above, the base of the tail and thighs
clearer tawny; below, the throat is buff, the rest of the under
side drab. In this, the typical race, the tail is over twice as
long as broad, a character distinguishing it from the Carolina
beaver, which has a relatively broader tail. Total length, about
1,100 mm. (about 35 inches); tail, 410; hind foot, 175.

The range of this race formerly extended from about New
Jersey northward to southern Labrador and Hudson Bay
(Churchill region) and westward across Canada and Alaska,
as far northward as the limit of trees, in well- watered country.
Over all this area its numbers are now greatly reduced, or in
places it is quite gone. Of the progress of this destruction only
the merest outline can be given here. At the time when the
first white settlers arrived in Massachusetts Bay and the Hud-
son River region, beavers were plentiful along the streams, even
near the coast. The earlier accounts contain many references
to them and their works. Thus, Governor John Winthrop, in
his "History of New England," tells how on January 27, 1631,
he set out with a small company from Boston and "went up by
Charles River about eight miles above Watertown, and named
the first brook, on the north side of the river, (being a fair
stream, and coming from a pond a mile from the river,) Beaver
Brook, because the beavers had shorn down divers great
trees there, and made divers dams across the brook." The
name still persists, though the beavers have long since gone.
Remains of beavers are common in the Indian kitchen middens
all along our eastern coast, as in Connecticut, on Block Island,
and the coast of Maine. In the earlier years of the Colonies
beaver skins were one of the most important trade products
of the country. These were obtained largely from the Indians

NORTH AMERICA AND THE WEST INDIES 51

who exchanged them for various trifles. Traders gathered
together large quantities of beaver pelts, which were sent back
to England. The magnitude of this industry may be gathered
when we read in Winthrop's account that in April 1633 a
vessel from Massachusetts, bound out for London, was wrecked
on the Virginia coast with the loss among other things of four
hogsheads of beaver weighing 900 pounds; and again, in the
same year, he tells of a Mr. Graves sailing with between five
and six thousand weight of beaver. Bradford's "History of
the Plimmoth Plantation" lists seven sailings between 1631
and 1636, in which a total of over 12,500 pounds of beaver was
exported from the colony, with a total value of about 10,000
sterling. Soon after this time, John Pynchon, of what is now
Springfield, Mass., was the chief one to handle the furs brought
in by the Indians from the surrounding region of the Connecti-
cut Valley. His old account books show that in the six years
between 1652 and 1657 he packed for England 47 hogsheads
containing 8,992 beaver skins, weighing over 13,000 pounds,
with an additional 663 pounds sent in bundles. From 1658 to
1674 he packed 6,480 beaver skins. Many of these skins were
from the country to the north add west of Springfield. Other
furs also appear in the inventories, but in lesser quantities.
At that time a beaver skin was sold at a standard value per
pound, about 8 shillings; and in lieu of currency beaver skins
were accepted at standard rates, which varied somewhat from
time to time. The traditional "beaver hats" were made from
the felted fur of this animal, and although these were mostly
manufactured in England there were in later years some among
the colonists who were skilled at this trade. For example,
Deacon Ebenezer Hunt, of Northampton, Mass., manu-
factured hats extensively for 40 years following 1734, buying
his furs chiefly in Boston or Albany. After 1750, beaver hats
sold at from 20 to 42 shillings each (Judd, S., "History of
Hadley, Mass.," p. 355, 1863). At this time beavers were
already becoming scarce in the southern part of New England.
Albany, N. Y., then known as Beaverwyck, and New York
City itself, then New Amsterdam, were the centers from which
the Dutch East India Company's posts gathered in rich
harvests of furs, while in the north the Hudson's Bay Com-
pany, chartered in 1670, formed the outlet for the fur catch in
that region.

52 EXTINCT AND VANISHING MAMMALS

After a century and a half of constant trapping and with the
spread of settlements into the eastern part of the country, the
beaver was extirpated from the coastal regions of New England
and greatly reduced in numbers elsewhere in the eastern
United States. It is said that the Indians were careful to leave
a certain proportion of the beavers to continue the stock, or to
kill adult animals and leave the younger ones to mature, but
the insatiate white trappers killed old and young indiscrimi-
nately and took as many as possible for their immediate gain,
with the result that the animals, bit by bit, were killed out from
the more accessible parts of their eastern range. Yet the re-
mains of their dams lasted for many years after the beavers
had gone and often were mentioned as monuments in land
conveyances, while the old beaver ponds silted up and became
grassy meadows, where frequently the earlier settlers, pushing
their way up the streams into the interior, found the only
available pasturage for their cattle. In southern New Hamp-
shire the last beaver in the town of Rindge is said to have been
killed about 1780 (Stearns, E. S., "History of the Town of
Rindge, N. H.," 1875); and the last one in the town of Peter-
borough about 1790 (Smith, A., "History of the Town of
Peterborough, N. H.," 1876). In the town of Pawlet, southern
Vermont, the "last" beaver was killed about 1800 by one
Ansel Whedon, who came upon it in his cornfield and slew it
with a hoe (Hollister, H., Vermont Hist. Mag., vol. 3, p. 890,
1877). On the Maine coast near Wells, the local, historian
records that a trapper took 17 beavers there in 1755, but they
seem to have quite gone before many years after. In northern
New England, the beaver, though greatly reduced, has never
been quite exterminated. Zadock Thompson, writing in 1842
("History of Vermont," p. 39, 1842), believed that they were
then nearly if not quite gone and mentions that the last one
that he knew of was killed in Essex County, about 1830.
Probably a few lingered in this northeastern corner of the
State until much later, for at that time this region was nearly
a primeval wilderness. Indeed, F. S. Hoag in 1909, after
having made many inquiries among the trappers of Vermont,
was credibly informed by a W. E. Balch that the last beaver he
knew of in the State lived at Neale's Pond, north of Lunenburg,
Essex County, about 20 years before (about 1890). He be-
lieved that it was completely extinct in Vermont by 1909.

NORTH AMERICA AND THE WEST INDIES 53

In the adjacent part of New Hampshire there were beavers in
the Connecticut lakes at least down to 1884, according to Ned
Norton (Forest and Stream, vol. 24, p. 457, 1885). Beavers in
small numbers continued in the remoter parts of Maine till the
middle decades of the last century. At about that time the
fashion for beaver hats abated in favor of silk hats, and the
demand for fur was also lessened by the substitution of nutria
fur from South America. Thoreau, in 1853, on his memorable
journey to the Maine woods, relates that they were getting to
be numerous again in the Lake Chesuncook region, but their
skins brought so little that it hardly paid then to hunt them.
J. G. Rich, an old trapper of those days and a correspondent of
Agassiz, wrote ten years later that the fur of beavers com-
manded only $2.50 a skin, whereas formerly as much as a
dollar an ounce was paid. In 1866, legal protection was first
given the beaver in Maine. Its numbers slowly increased
until during the first decade of the present century it seems to
have become locally common in the remoter townships, where
it was not allowed to be killed at any time. Increasing com-
plaints of timber owners that beavers were damaging their
trees by cutting and flooding resulted in 1910 in the establish-
ment of a short open season for two years in Somerset and
Franklin Counties. In 1911 the area on which beavers could
be trapped was extended to other counties. At the present
time there seems to be no reason why a fair number of beavers
can not be maintained in the wilder areas of the State.

The history of the beaver in New York State has been re-
viewed by Radford (1908), particularly its status in the
Adirondacks. "As early as 1623 the importance of the beaver
to the Dutch colony was so well recognized as to lead to its
incorporation in the seal of New Netherlands." In those times
Indians from the St. Lawrence Valley trapped beavers in
northern New York and traded them to the French companies
at Quebec. Merriam adduces evidence to show that in the
1830's beavers in the Adirondacks were at a very low ebb, and
DeKay, who traversed the height of land between the sources
of the Hudson and the St. Lawrence Rivers in 1840, believed
them nearly extirpated, though reported in 1841 "on Indian
and Cedar Rivers, and at Paskungameh or Tupper's lake"
and in scattered families in parts of Hamilton, St. Lawrence,
and Essex Counties. Radford believes that by 1860 there

54 EXTINCT AND VANISHING MAMMALS

could have been hardly more than 50 beavers remaining in New
York State, chiefly along the Raquette and St. Regis Rivers
and in the well-watered region northwest of Upper Saranac
Lake in Franklin County. Here a slowly dwindling remnant
remained for the next 30 years, their numbers gradually re-
duced by occasional trappers in spite of efforts to protect them
by a local inn-keeper on or near whose lands they lived. In
the winter of 1894-95 probably not more than ten remained.
In the latter year a law was passed removing an open season
for beavers in the State, and in 1904 this was given added force
by increasing the fine for taking beavers to $100 and making it
an offense to set traps for them or to molest their dams or
houses. From the beginning of this century for a number of
years beavers showed a slight increase, and at the same time
small numbers were introduced at favorable localities. Some
of this stock came apparently from Canada; 30 or more others
were obtained in 1906 from Yellowstone Park. Under protec-
tion, and through this addition of new breeding stock, beavers
have so greatly increased in New York State that during the
season of 1924, when trapping was allowed, no less than 2,478
pelts were taken, valued at $39,548, and the following year
yielded 3,573 pelts, which brought in all $71,460 (Couch, 1937).
Under the present law, the Conservation Department of New
York may declare an open season in the month of March only.
Of late years a two-week period has been given in counties that
seem to have a good supply of the animals. The population,
however, is so unevenly distributed that in some regions more
beavers are desirable, while in others, which they are invading
rapidly, it is doubtful if a beaver population can be maintained.
Since 1924 to 1939, seven open seasons of two weeks each have
been permitted. Each trapper is limited to six pelts to be
taken in the two weeks, and these must be tagged by a game
protector. In its practical application it is found that the
protective law can be best enforced when blocks of counties
are governed by like seasons, so that granting open seasons
becomes in good part a problem of law enforcement. In 1939
"there are not more than a half-dozen counties in the State,
aside from those comprising Greater New York City, in which
beaver colonies were not established. Even in Rockland
County, whose border is but a half-hour's motor trip from
Broadway, an open trapping season has been permitted for

NORTH AMERICA AND THE WEST INDIES 55

several years" (Dr. G. Bump, in Hit.). The following sta-
tistics have been obligingly furnished by Dr. Gardiner Bump,
Superintendent of the Bureau of Game, concerning recent
beaver catches in New York State. In 1935, 2,498 pelts were
taken in eleven counties; in 1936 there was no open season; in
1937, a total of 2,014 pelts was taken in thirteen counties; and
in 1938, in twelve counties, 2,629. It is thus clear that under
wise management and intelligent supervision the bearver may
become a considerable asset in regions where conditions are
suitable for the maintenance of a breeding population.

The story in Pennsylvania is much like that in New York
State. Rhoads (1903), after giving in detail such older records
as he could find concerning their former abundance and gradual
extirpation in the State, summarizes these as follows: "It is
evident that this interesting animal was practically extermi-
nated in the eastern half of its Canadian habitat in Pa. about
1830; that some remained in the headwaters of the west branch
of the Susquehanna till about 1840, and that almost the last
stragglers of their race were killed in Elk, Clarion, and Centre
Cos., between the years 1850 and 1865." Beavers killed in
Clinton County in 1884 and on* seen in Cambria County in
1899 may have been escapes. It seems safe to say that beavers
had been practically exterminated in Pennsylvania for 60 years
when in 1917 a pair was imported from Wisconsin. From then
until 1924, 94 beavers "were imported and set free in certain
sections of the State at a cost of about $50 each. The animals
increased so rapidly that it soon became necessary to transfer
some of them to other sections. A survey by the State Board
of Game Commissioners of the streams in Pennsylvania in
1931 revealed 899 beaver dams with an estimated beaver
population of 4,377, which by 1934 had increased to 15,000.
During the 1934 trapping season, 6,455 beavers were legally
taken, which, at the average price of $15 a pelt, brought the
trappers a total of $96,825. In 1936, under increased trapping
restrictions, 2,261 pelts were taken which brought a total
return to the trappers of $22,610, an average price of $10 each
. . . These financial returns demonstrate that the beaver
constitutes a natural resource of great importance" (Couch,
1937).

Profiting in part by the experience of New York and Penn-
sylvania, beavers have in recent years been introduced into

56 EXTINCT AND VANISHING MAMMALS

Vermont, central New Hampshire, and Mount Desert, Maine,
and at the present time they seem well established in several
localities.

From west of the Alleghenies to the edge of the Plains,
beavers seem at the present time to be quite gone over most of
their former range. For Ohio, Bray ton (1882) mentions their
former abundance but implies that they had practically gone
by the middle of the last century. In Indiana, Hahn (1909)
gives 1840 as about the last authentic record, and both he and
Lyon (1936) believe that the beaver of this State may have
represented the southeastern race, while the latter adduces
what seems a reliable recent record of the animal from Wells
County in the northeastern part of the State and implies that
it may have come in from adjacent parts of Michigan, where a
few beavers still exist. Possibly, however, it was introduced,
for in 1935 small "plants" of beavers were made in three
counties, Jasper, Pulaski, and Starke, on game preserves, and
these have so multiplied that in 1939 they had spread into ten
other counties of northern Indiana, and additional "plants"
have been made in Jennings and Clark Counties in the southern
portion, where, however, the conditions of sufficient food and
water are less favorable for them. Strict protective laws are in
force, but in general land owners are interested in the encour-
agement of the beaver, in part from the benefit to water con-
trol and in part on account of the interesting habits of the
animals (from Outdoor Indiana, Nov. 1939).

Still farther west, Cory (1912) writes: "Beavers were for-
merly common throughout Illinois and Wisconsin but at the
present time they are practically exterminated in the former
State," although "it is probable that a very few individuals
may exist in the extreme southern portion," in Alexander
County, whence fresh cuttings were obtained in 1900. In 1854,
Kennicott wrote of beaver dams still existing at various points
in Illinois, as if the animals had practically gone even by that
day. In northern Wisconsin, however, they are now common.

North of the United States, beavers still occur plentifully
in the more remote parts of western Canada but have been
largely reduced in the more accessible regions and, of course,
extirpated in the settled areas. G. G. Goodwin (1924), writing
of the mammals of the Gaspe Peninsula, Quebec, tells of finding
a skull at the forks of the Ste. Anne River and adds that

NORTH AMERICA AND THE WEST INDIES 57

according to his guide "there were a few beavers left above the
forks of the river and that they live in holes in the river bank."
Dr. R. M. Anderson (1939a) states that though formerly
common in the wooded parts of the Province of Quebec it is
trapped out in most localities. However, "some projects for
restoration of the beaver by setting aside preserves have been
gratify ingly successful, particularly in the regions south and
east of James Bay." Although beavers have been trapped in
this region for nearly three centuries, since the establishment
of the posts of the Hudson's Bay Company, they appear to
exist in reduced numbers to this day. Preble (1902) writes,
however, that owing to persistent trapping they were "becom-
ing scarce throughout the region" over 30 years ago, although
skins were annually traded at the Company's posts on the
coasts of Hudson Bay. He noted "the remains of a beaver
house between Pine and Windy lakes and a comparatively
recent dam on a small stream which empties into Hayes River
about 15 miles above York Factory. A number of skins were
seen at Fort Churchill. These had been taken on the Lower
Churchill River. Several black pelts were among the furs at
Norway House." He mentions further that "Dr. Bell reports
that a family of beavers was found by Indians on North
River, a stream that flows into the Bay about 15 miles above
Fort Churchill"; and that according to Hearne's account of
his explorations in this region, published in 1795, the Indians
accompanying him killed beavers on Seal River, the mouth of
which is about 40 miles north of Churchill. This marks perhaps
nearly the northern limit of the beaver on the west side of Hud-
son Bay; of its northward limits on the east side of the Bay no
recent information is at hand. Low, writing in 1895, after
traversing parts of northern Labrador, says that it is common
in the wooded regions and extends into the semibarrens where
food is available.

As to the recent status of beavers in the Athabaska-Macken-
zie region still farther west, Preble (1908) has given a brief
review. He quotes J. Alden Loring that in 1896 evidence was
obtained of the former abundance of the beaver in suitable
localities in western Alberta, but at that time it was nearly
exterminated; tracks were seen on a small stream between
Jasper House and Smoky River, but no other recent traces of
the animals were found. Farther north it was formerly abun-

58 EXTINCT AND VANISHING MAMMALS

dant nearly to the limit of trees; but on his journey in 1903-4,
to the Great Slave Lake and Mackenzie region, he found it
largely gone in many parts, and nowhere common, "though
skins are received annually by all the posts throughout the
region . . . The vast region which stretches from Great
Slave Lake to the Rocky Mountains at present seems to be the
best beaver country in the north. Many skins are brought
from the upper reaches of Hay River by the Beaver Indians,
and from Trout Lake by the Indians who frequent that
locality. The Horn Mountain country also furnishes many
skins. Along my route between Great Slave and Great Bear
Lakes, the beaver has now become scarce, owing to constant
hunting, but my guide intimated that in certain localities off
the main route which we were following he knew of small
colonies of beavers. About Great Bear Lake the best beaver
ground seemed to be to the northward of Fort Franklin, and
I saw several skins, some quite dark, just brought from the
hunting grounds about two days' travel to the northward.
During the winter of 1903-4 several beavers were killed by
Indians in the region about Fort Simpson. In the spring the
animals often descend the smaller streams to the main river
and follow it to the mouth of the next tributary. A young one
was shot near the mouth of the Liard in May, and several
adults and young ones have been killed in recent years near
the mouth of Bluefish Creek, opposite Fort Simpson, as a re-
sult of this habit.

"While descending the Mackenzie in the summer of 1904 I
saw no beavers, but obtained information regarding the traffic
in skins. About 700 skins were said to have been traded during
the preceding winter at Fort Norman, which receives the fur of
a very large extent of country. Skins from the country toward
the Barren Grounds, according to the testimony of C. P.
Gaudet, of that place, are smaller and average darker than
those from the vicinity of the post. Fort Anderson, according
to the fur returns, never received more than five skins annually
during the few years of its existence."

In the Yukon River Valley beavers are nearly extirpated.
Dr. W. H. Osgood, who traversed the region in 1899, wrote
(1900): "It hardly seems possible that half a million or more
beaver skins have been secured in the Territory of Alaska.
The animal is now almost as rare as it is in the United States,

NORTH AMERICA AND THE WEST INDIES 59

the inevitable result of continued pursuit by both whites and
natives, which has so many parallels that it is useless to
emphasize it here. At Fort Selkirk I saw several beaver skins
taken on a small tributary of Stewart River and at St. Michael
I found a very few in the warehouses of the trading companies.
Beyond this I saw or heard nothing of them."

Dr. Francis Harper states that under recent protective laws
beavers are increasing in Alberta at the present time, and it
seems likely that proper restrictions, if they can be enforced in
all this region, will result in an increase of beavers, so that they
may in time prove a regular and profitable source of revenue.

Concerning the habits of the beaver a great deal has been
written. Their general activities center around, first, the dams
constructed solidly of sticks, branches, and larger pieces of
wood, together with stones, mud, and other material from the
pond bottom; then with the resulting formation of a pond, the
construction of a house which serves as a shelter for young and
adults, and has entrances under water into the pond. In the
pond, aspen logs are sunk for a winter food supply, for the
bark of this tree is a favorite food. An older colony will in
time connect various pools with? canals and the dams may be
extended to works of great magnitude. A single litter of from
two to six young is produced in the course of a year, so that
the increase of a colony may be fairly rapid and the young
when of adult size move off to found new colonies. Extensive
accounts of the habits may be found in the following works:

DUGMORE, A. RADCLYFFE.

ca. 1913. The romance of the beaver: Being the history of the beaver in the
western hemisphere, xvi -f- 225 pp., illustr. Philadelphia and London.
BAILEY, VERNON.

1922. Beaver habits, beaver control, and possibilities in beaver farming. U. S.

Dept. Agric. Dept. Bull. 1078, 29 pp., 7 figs., 7 pis.
1927. How beavers build their houses. Smithsonian Inst. Kept, for 1926, pp.

357-360, pis. 1-6.
CHAPMAN, WENDELL and LTJCIE.

1937. Beaver pioneers, xiv -f- 153 pp., illustr. New York and London.
MARTIN, HORACE T.

1892. Castorologia, or the history and traditions of the Canadian beaver, xvi +

238 pp., illustr. Montreal and London.
MILLS, ENDS A.

1913. In beaver world, xiv + 228 pp., illustr. Boston and New York.
WARREN, E. R.

1922. The life of the Yellowstone beaver. Roosevelt Wild Life Bull., vol. 1,
pp. 187-221.

60 EXTINCT AND VANISHING MAMMALS

WARREN, E. R.

1926. A study of the beaver in the Yancey region of Yellowstone National
Park. Roosevelt Wild Life Annals, vol. 1, pp. 13-191, illustr., maps.

1926. Notes on the beaver colonies in the Longs Peak region of Estes Park,
Colorado. Roosevelt Wild Life Annals, vol. 1, pp. 193-234, illustr.

1927. The beaver: Its work and its ways. Amer. Soc. Mammal. Monogr. 2,
177 pp., illustr. Baltimore.

BAILEY'S BEAVER
CASTOR CANADENSIS BAILEYI Nelson

Castor canadensis baileyi Nelson, Proc. Biol. Soc. Washington, vol. 40, p. 125, Sept. 26,
1927 (Humboldt River, 4 miles above Winnemucca, Nevada).

This race of the beaver is darker than frondator and has a
slenderer skull. From C. c. pacificus from the Columbia River
drainage in eastern Washington and Oregon it differs in its
distinctly paler color and slenderer skull. The narrow skull
and especially the rostrum of the Humboldt River beaver
contrast strongly with the massive skull and broad heavy
rostrum of subauratus (Nelson, loc. cit.). The upper parts
are described as "dull rusty chestnut, brightest on crown
with a dull yellowish shade on the cheeks; ears dark brown;
base of tail all around uniform with adjacent parts of body;
tops of hind feet dark chestnut; underparts of body dull drab
brown." Total length of type, 1,064 mm.; tail, 254 by 135;
hind foot, 183.

Concerning this race of the beaver little seems to be in
print. Dr. Nelson had specimens from Winnemucca, Iron
Point, Golconda, and Deeth in Nevada. It is presumably con-
fined to the region of the Humboldt River Basin of Nevada.
Borell and Ellis (1934) wrote of it a few years ago: "Beavers
are found in limited numbers along the main branches of the
Humboldt River, but have been almost completely extermi-
nated along the smaller streams which flow out of the Ruby
Mountains. Mr. August Rohwer and Mr. William Toyn re-
ported that there were still a few beaver in one or two of the
canyons on the west slope of the Ruby Mountains. Character-
istic beaver gnawings found by us on some partly decayed aspen
stumps at about 7,000 feet altitude along Toyn Creek, near the
summit of Harrison Pass, indicated the former presence of
beaver there."

Still more recently, Vernon Bailey (1936) has reported that
"these desert-valley beavers are still found in the Great Basin

NORTH AMERICA AND THE WEST INDIES 61

drainage of northern Nevada along the Humboldt River, and
its tributaries and in the Malheur Lake and Steens Mountains
drainage of southeastern Oregon . . . There are a few
beaver all along the Blitzen from its headwaters down to near
Malheur Lake, and George Benson reports one shot at the edge
of Malheur Lake in 1909. In 1916 they were found in Big
Fish Creek, McCoy Creek, and Kiger Creek. It is fair to
assume in the absence of specimens that the beaver in the
Silvies River and its branches on the north of Malheur Lake
are also of this subspecies. In 1920 they were still common in
many places along the Silvies and Blitzen Rivers, and while in
places they were doing some damage by flooding the meadows,
more often they were merely holding up the water to a better
depth and improving the meadows by subirrigation." Con-
cerning their former abundance in this last region, Bailey
quotes from older records to show that a little over a hundred
years ago, in 1826 and 1828, beavers were so numerous that a
party of trappers took over 300 in a month.

On occasion, Bailey adds, they may dam up irrigation
ditches or locate in the banks of streams near fields and or-
chards where they do serious mischief and have to be removed,
driven away, or destroyed. With intelligent handling, "how-
ever, they could become a valuable asset on many of the
eastern Oregon ranches."

COOK INLET BEAVER
CASTOR CANADENSIS BELUGAE Taylor

Castor canadensis belugae Taylor, Univ. California Publ. Zool., vol. 12, p. 429, Mar.
20, 1916 ("Beluga River, Cook Inlet region, Alaska").

According to its describer, this race of beaver is the one
found from central British Columbia northward along the
coast of the mainland to the mountains of Alaska. Specimens
referred to the race are mentioned from Stuart Lake, British
Columbia, and from various localities on the shores of Cook
Inlet, Alaska. In color the single skin available to the describer
was said to be slightly paler than in the neighboring races,
phaeus and leucodonta. The skull, in comparison with that of
the latter, is "immediately distinguishable through the nar-
rower blades of the hamular processes of the pterygoids," with
a tendency for the maxillary tooth row and ratio of this to the
basilar length to be greater.

62 EXTINCT AND VANISHING MAMMALS

In the region of the type locality, beavers were becoming
scarce 40 years ago. Osgood (1901) mentions three secured by
trappers near the mouth of Turnagain Arm in 1899 and adds
that a few were taken every season along streams in the moun-
tains about 60 miles inland from Tyonek. "A trading station
on the lower Sushitna River also obtains a small quota annually.
Compared with former receipts, however, the number now
obtained is lamentably small." In a later report Osgood (1904)
adds that a little farther to the westward on the Chulitna River
he found evidence of a few scattered small colonies of beavers.
"The extensive area of low land about the sources of the
Chulitna River is covered with hundreds of small lakes and
ponds connected in most cases by small, sluggish streams
eminently suitable for beavers, and no doubt a great many are
still scattered throughout this area." A small number of skins
were annually brought in to the trader at Nushagak at the
mouth of the river of the same name. Farther south, in the
Atlin region of northwestern British Columbia, Swarth (1936)
writes that "as everywhere, beavers hereabout are trapped to
the verge of extinction. They in all likelihood were originally
abundant and generally distributed throughout the lowland
lakes."

NEWFOUNDLAND BEAVER

CASTOR CANADENSIS CAECATOR Bangs

Castor caecator Bangs, Bull. Mus. Comp. Zool., vol. 54, p. 513, July, 1913 ("Near

Bay St. George, Newfoundland").
FIGS.: Bangs, 1913, figure on p. 514 (skull).

Although the Newfoundland beaver was first described as a
species distinct from typical C. canadensis of the neighboring
mainland, it can scarcely be considered as more than the local
representative of that animal, which through a long period of
isolation has become slightly different in certain cranial char-
acters. The color has not been accurately described on account
of the lack of specimens in museum collections, but Bangs, in
naming it, mentions that he had seen two in the flesh, which,
though showing nothing distinctive in external appearance,
nevertheless appeared "rather small" and were of an "excep-
tionally fine rich color." The distinctive characters of the
skull are: the slightly smaller size as compared with the main-
land beaver, typical canadensis; the wider, "more roundish"

NORTH AMERICA AND THE WEST INDIES 63

interparietal; and the lighter, less flaring zygomata, the outer
side of which is more nearly straight, thus giving a much more
triangular outline to the skull as seen from above. The nasals
are also said to be shorter and wider.

Concerning the present status of the beaver in Newfound-
land, little accurate information is at hand. Bangs, in 1913,
wrote that "though much trapped for its fur, it still occurs in
fair numbers in the remoter parts of the island," and probably
this still holds true. It doubtless finds the more barren portions
of the country less suitable to its needs and hence can be
expected to thrive best in the western and central parts of the
island. Dugmore, in his "Romance of the Beaver," published
about 1913, writes that when he first visited Newfoundland
about 1900 and in annual visits during the three succeeding
years, he never saw but one beaver colony, and that a very
small one in a remote and rather inaccessible part of the
country. Under governmental protection, however, beavers
multiplied in following years so that on visiting the same
region in 1912 he counted no less than 27 lodges within a short
day of walking and canoeing. As elsewhere, this illustrates
how readily beavers will breed up to the carrying capacity of
the land if free from persecution, for their natural enemies
are few.

For a comment on the recent status of beavers in Newfound-
land I am indebted to Frank Strickland, who is familiar with
conditions there. He says that previous to the declaration of
a close time on beavers, about 1917, they were plentiful in
Grand Lake and Red Indian Lake, but today they are very
scarce in that section. What beavers are left in Newfoundland
are mostly to be found between Port Saunders north to White
Bay and between Gander River east to the bottom of Bay
d'Espoir, including thus the central districts of the island.

CAROLINA BEAVER; SOUTHEASTERN BEAVER

CASTOR CANADENSIS CAROLINENSIS Rhoads

Castor canadensis carolinensis Rhoads, Trans. Amer. Phil. Soc., new ser., vol. 19, p.
420, Sept. 1898 ("Dan River, near Danbury, Stokes County, North Carolina").
FIGS.: Rhoads, 1898, pi. 23, figs. 1, 2 (skull).

The beaver of the southeastern United States is broader-
tailed than typical canadensis., and in this respect it resembles

64 EXTINCT AND VANISHING MAMMALS

the southwestern race frondator, which, however, is paler in
color. Rhoads describes the type as bright hazel above, with
seal-brown underfur, the hinder back shading into bright hazel
to cinnamon-rufous and tawny-olive; ears blackish, feet bister.
Below, dark broccoli brown, the tips of the longer hairs wood
brown. Total length, 1,130 mm.; tail, 279 by 158; hind foot,
184. The skull is characterized by its short, broad nasals and
short rostrum, with the former scarcely or not extending back
of the level of the orbits.

The limits of the range of this race, though in general from
northern Florida to the Mississippi and northward to the low-
lands of New Jersey and Ohio, Indiana, and Illinois, will
probably never be exactly definable since it is now extinct
over the greater part of the area and but few complete speci-
mens are extant. Rhoads believed that it intergraded with
typical C. canadensis in southern New Jersey and the lower
parts of Pennsylvania, and Cory (1912), in his map of the
range, included southern Ohio, Indiana, and Illinois in the
area of presumed intergradation. Lyon (1936), who figures an
imperfect skull as the only one extant from Indiana, assumes
that it was probably the form occurring in that State.

Rhoads believed that by 1700 beavers had been practically
exterminated from the lowlands of the Delaware, Schuylkill,
and Susquehanna Valleys, which were the regions first settled ;
while by the time of the American Revolution, they were
practically wiped out from all the lowland valleys of Pennsyl-
vania except the northern tributaries of the Ohio; but of this
there is almost no definite record. In New Jersey, owing to
the inaccessible and unproductive character of the lands in
the southern part of the State, beavers persisted longer, in the
most retired swamps of Atlantic and Cape May Counties, but
were probably quite gone by 1830, or earlier. Traces of their
old dams persisted for many years later, and Dr. Witmer
Stone, writing in 1908, mentions remains of a very large one
that he had visited on the Nescochaque branch of Mullica
River. In 1900, according to Dr. Stone (1908), there was a
colony of beavers in Sussex County, N. J., due apparently to
individuals that had escaped from a private preserve where
they had been introduced.

Atwater's " History of Ohio," published in 1838, speaks of
beavers as formerly common on the headwaters of the larger

NORTH AMERICA AND THE WEST INDIES 65

streams in that State but adds that "they have long since
disappeared." For Indiana much the same story is true; Dr.
Lyon (1936) quotes the Prince of Wied's statement, in 1832
and 1833, that beavers were then abundant along the lower
Wabash, but with the opening up of the country to settlement
in succeeding decades, they must have disappeared rapidly.
Nevertheless, he adduces some evidence for the belief that in
Wells County, Ind., beavers were present only a few years
since, possibly introduced; while similarly, Cory (1912), al-
though he writes that beavers are practically extinct in Illinois,
believes it "probable that a very few individuals may exist in
the extreme southern portion of the state."

To the southward of this tier of States that may have marked
the northern limits of this beaver's range, information is
meager, but it indicates that the animals still hold on in a few
widely separated localities, where with real protection there is
some hope that they may continue and even increase con-
siderably.

Within the memory of men now living beavers were present
in small numbers in the counties a short distance south of
Richmond, Va. A writer in Forest and Stream (vol. 7, p. 197)
for November 2, 1876, states that for the two years previous
to that time trappers had been taking beavers in Dinwiddie,
Nottoway, Brunswick, Cumberland, and other counties, some
making good returns on these and other furs. At that time
the pelts of beavers sold for a dollar apiece. Here, too, the
correspondent mentions having seen in some of these localities
places where beavers had totally destroyed acres of corn, caus-
ing serious loss. At the present time, I have no information of
any beavers remaining in the State. Possibly, however, a few
may have persisted in the mountainous parts of this State
and of West Virginia. F. E. Brooks (1911) mentions what he
believes to be a well-authenticated case of a beaver having
been killed in Pocahontas County, W. Va., about 1907, al-
though this was possibly brought in from elsewhere. Still
later, A. B. Brooks (1923) published a note on the reappearance
of beavers in West Virginia. He writes: "A few of the older
trappers and hunters still living state that beaver work was
observed by them many years ago on the headwaters of the
Williams and Greenbrier Rivers," but they had been "counted
as extinct in West Virginia for fifty years or more" until the

66 EXTINCT AND VANISHING MAMMALS

early fall of 1922, when a farmer in Hampshire County re-
ported finding fresh beaver work on a creek tributary to North
River. Subsequent investigation disclosed a colony of beavers
here. They had "built a dam and done much cutting along
both banks of the stream." All efforts to learn of the origin
of this colony were unsuccessful.

Beavers persisted in North Carolina until recent years.
The series of specimens from Dan River, near Danbury, Stokes
County, on which the race carolinensis was based, was col-
lected between 1897 and 1899. C. S. Brimley, writing in
1905, says that it occurs sparingly there and is also reported
from Bertie County. A specimen taken 15 or 20 years pre-
viously at Weldon is in the State Museum at Raleigh. These
localities are all on the Roanoke River or its tributaries. He
adds that "Mr. J. H. Armfield reports a few occurring in
Beaver Swamp in the northern part of Guilford County and
southern part of Rockingham, and Mr. K. E. Shore reports
them from the Yadkin River, between Yadkin and Forsyth
Counties." At the present time, 1939, the beavers seem to
have quite disappeared.

No recent record of the beaver in South Carolina is known
to me. In Georgia, however, a few still remain and are care-
fully protected, in the Flint River section in the southwestern
part of the State and at the headwaters of the Chattahoochee.
Dr. Francis Harper (1927) in his account of the mammals of
Okefenokee Swamp, Ga., adduces evidence that they were
formerly present there. He was shown an incisor tooth taken
from a beaver killed about 1890 by an old Negro woman a
mile or so east of Lloyd's Island. "The story goes that the
woman came upon the animal in the road and clubbed it to
death." Dr. Harper suggests that the southward range of the
beaver may have been to a large extent limited by the presence
of the alligator. He doubts if it ever regularly inhabited the
interior of Okefenokee Swamp, else it would probably have
survived there or left some tradition of its earlier presence.
In former times there were beavers in extreme northern
Florida (Bartram mentions them), but little or nothing is
known of the time of their disappearance. An unmistakable
molar tooth was lately found in an Indian mound on Indian
River by J. H. Rowe.

To the westward of the Alleghenies, the Carolina beaver is

NORTH AMERICA AND THE WEST INDIES 67

believed to have extended to the Mississippi, perhaps as far
north as Iowa. In this State, according to J. A. Allen (1870a),
they had been nearly or quite exterminated in most of the
eastern and southern portions by 1870, although at that time
a few were reported still to exist on the South Raccoon River.
Kellogg (1939) has lately summarized its history in Tennessee,
where it was originally well distributed in former times and
apparently did not begin to show much depletion until after
the middle of the eighteenth century. About 1788 the salaries
of the county clerks and others were paid in beaver skins, 300
for the county clerk, 200 for the clerk of the House of Com-
mons, and three for members of the Assembly. The latest
records for beavers seem to be those of Rhoads (1896) for the
extreme western part of the State. He examined a beaver
house in the cypress swamp bordering Reelfoot Lake, Obion
County, about 1895 and was told that there were then about
20 beavers in that district. Another colony was reported to
him from Hay wood County. Beavers still exist in Alabama,
and according to A. H. Howell (1921) have held their own in
some localities remarkably well, even in settled farming regions.
He says that "apparently they are more numerous at present
[1920] in the central part of the State, in Montgomery and
Lowndes Counties, than in either the wild hill country of the
northern part or the big swamps of the south." Old residents
of Montgomery County assured him that although abundant
there in early times, they had entirely disappeared from the
region, but about 1908 reappeared, coming up from the lower
stretches of Catoma Creek. In Lowndes County, L. J. Gold-
man found beavers in numbers along Big Swamp Creek and in
Jones Lakes, where they do considerable damage to the corn
patches. They are little trapped for fur, although their meat
is much sought after. They now have protection either locally
by the planters or in general by State law, from March to
November. "In the northern part of the State beavers seem
almost to have been exterminated; they formerly occurred in
small numbers in the Tennessee River at Muscle Shoals, but
disappeared about 1895; they lived about the 'towheads'
(small islands) in the river and burrowed into the banks, but
did not build dams. A few were reported in 1916 in Big Wills
Creek, near Collinsville. They disappeared from Talladega
Creek, near Dean, about 1896."

68 EXTINCT. AND VANISHING MAMMALS

These southern beavers attain a large size; Howell speaks of
one killed on Pintlala Creek that weighed 65 pounds. Another
from Catoma Creek weighed 38.5 pounds and measured 1,035
mm. in total length; tail, 290 by 163; hind foot, 170. Compared
with a specimen of the typical race from Canada, its skull is
shorter and relatively broader, with wider brain case and
interorbital region.

In Kentucky, which should come within the western range of
this race, Funkhouser (1925) writes: "We can find no record
of its having been seen . . . within the past twenty
years." In earlier times it was occasionally mentioned by the
pioneers, but "was probably never very abundant in Ken-
tucky."

COLORADO BEAVER

CASTOR CANADENSIS CONCISOR Warren and Hall

Castor canadensis concisor Warren and Hall, Journ. Mamm., vol. 20, p. 358, Aug. 14,
1939 ("Monument Creek, southwest of Monument, El Paso County, Colorado").
FIGS.: Warren, 1910, p. 141, figs. 46-48 (photographs; skull).

Until recently the beaver of Colorado was believed to be the
same as the race frondator, typical on the Mexican boundary of
Sonora. Now, however, the above authors, after a minute
study of a series of skulls, believe that it represents a recog-
nizable subspecies, of dark color, but with a skull resembling
that of mexicanus, yet on the average differing in the size and
shape of the angular process of the mandible, which is longer
and more produced posteriorly and sharp-pointed rather than
short and posteriorly rounded. Other slight differences seem
to separate the Colorado beavers from adjacent races. Greatest
occipitonasal length of skull, to 145.2 mm.; length of nasals,
50-54.4.

Warren (1910) writes that although the beaver "was no
doubt at one time found in every county in Colorado which
contained streams with sufficient water for its needs, and to-
day, in spite of the persecution to which it was at one time
subjected from the trappers, nearly resulting in its extinction,
it is found over a large area of the State, and thanks to the
protection accorded it by law, is on the increase. We have
records of it from Larimer, Weld, Morgan, Grand, Routt,
Arapahoe, Gunnison, Delta, Garfield, Eagle, El Paso, Teller,
and Mineral counties," thus covering much of the west-central

NORTH AMERICA AND THE WEST INDIES 69

mountainous region of the State. The same author (Warren,
1926) has given a detailed account of beaver colonies in Estes
Park, Colo., with descriptions and photographs of their haunts
and works, as well as maps illustrating their engineering enter-
prises. Under protection the beaver seems for the present
fairly safe in Colorado.

SOUTHWESTERN BEAVER; BROAD-TAILED BEAVER;
SONORAN BEAVER

CASTOR CANADENSIS FRONDATOR Mearns

Castor canadensis frondator Mearns, Preliminary Diagnoses of New Mammals of the
Genera Sciurus, Castor, Neotoma and Sigmodon, p. 2, 1897; reprinted in Proc. U. S.
Nat. Mus., vol. 20, p. 502, Jan. 19, 1898 (San Pedro River, Sonora, Mexico, near
monument No. 98 of the Mexican boundary line).

FIG.: Mearns, 1907, p. 351, fig. 57 (skull).

Like the southeastern beaver (C. c. carolinensis) , this is a
broad-tailed race, but its color is paler and brighter. Vernon
Bailey (1931) describes it as light chestnut above; middle of
the belly reddish chestnut; hind feet dark chestnut; colors
slightly darker in winter. Total length, 1,070 mm.; tail from
anus, 360; hind foot, 185; skull length, 133, its width 99.
Weight of type (a female), 62 pounds. Old males are slightly
larger than females and may measure up to 1,130 mm. in length.

This race of beaver extends slightly to the south of the
Mexico-United States boundary in northern Sonora and is
found in Arizona along the Colorado River and in the extreme
western part of New Mexico on branches of the Gila and San
Juan Rivers. According to Bailey (1931) it is the form found
from all points whence specimens were examined along the
Colorado River system, but although formerly numerous, it is
now everywhere much reduced in numbers. "There are still a
few beavers in the headwaters of the Gila . . . also some
in the San Juan River and its tributaries" in western New
Mexico. Mearns (1907) has also given many details of its
presence during the latter part of the last century along the
Mexican boundary of Arizona. Within the rather restricted
range at present assigned to this race, it is evidently rare and
likely to be further reduced in the future.

70 EXTINCT AND VANISHING MAMMALS

VANCOUVER ISLAND BEAVER

CASTOR CANADENSIS LEUCODONTA Gray

Castor canadensis leucodonta Gray, Ann. Mag. Nat. Hist., ser. 4, vol. 4, p. 293, Oct.

1869 (Vancouver Island, British Columbia).
FIGS.: Taylor, 1916, p. 450, fig. J, a; p. 451, fig. K, a; p. 452, fig. L, a (views of skull).

Dr. W. P. Taylor (1916) has briefly summarized some of the
characters that he believes distinguish the Vancouver beaver.
Externally it closely resembles typical C. canadensis from New
Brunswick in color above, but the under side is "about hair
brown in leucodonta, near bone brown, dark grayish brown,
dark vinaceous-drab or natal brown in canadensis. " Cranially,
the external outline of the nasals is different, "tending to be
more parallel in leucodonta than in canadensis, in which there
is a dilation in the outline anteriorly; foramen magnum slightly
broader . . . ; hamular processes . . . definitely
broader bladed. " There are other slight cranial differences
between this and neighboring races.

In the middle of the last century beavers were common on
Vancouver Island, to which this race is now regarded as re-
stricted. Robert Brown (in Green, A. H., 1869), who collected
the original specimens for the British Museum, spoke of them
near Victoria, and especially in the interior, where "they are
almost everywhere abundant and on the increase. In a swampy
lake near the mouth of the Cowichan Lake we found many;
and an extensive swamp near the entrance of the Puntledge
Lake was a great stronghold ... In the spring of 1866,
when crossing the island from Fort Rupert to the head of
Quatseeno Sound with some Indians, a great portion of our
route lay among these beaver ponds and dams. All through
this district beavers swarm. "

Swarth (1912) has presented some measurements and cranial
dimensions from a series of this beaver trapped in 1910 on
Vancouver Island. The external appearance, he says, differs
from that of beavers from southeastern Alaska principally in
slightly paler color. He found that "as a result of a number
of years of protection beavers have multiplied on Vancouver
Island so as to be really abundant in many places . . .
There was a small colony at a point near Errington, but we
found them in far greater numbers at Beaver Creek, near
Alberni . . . There was old beaver work along the streams

NORTH AMERICA AND THE WEST INDIES 71

throughout this valley, and from the size of some of the old
dams they must have existed here in immense numbers at one
time, but at present, though there are a good many left still,
they are more scattered, apparently not more than one or two
families at any one point, and such groups separated by inter-
vals, sometimes of several miles. Nearly all the smaller streams
were obstructed by their dams, and most of the adjoining low
lying land flooded as a consequence . . . We saw no
beavers at Nootka Sound, but were told that there were many
at Vernon Lake, some eight or ten miles inland from the head
of Tahsis Canal" and a few at Central Lake.

In response to inquiry, Maj. Allan Brooks writes me under
date of July 8, 1939: "Beavers still exist in some numbers on
Vancouver Island. There are few localities where the trapping
is not in the hands of licensed trappers, who pay $10 a year for
a denned area where no other trappers are allowed. By this
method they can conserve their furbearers. There are locally
closed areas also, as at Comox, Vancouver Island, where no
beavers are allowed to be trapped and they are numerous.
Over the whole northern third of this island the beavers of late
years have been decimated by co\igars; it will take them many
years to recover. "

Rio GRANDE BEAVER

CASTOR CANADENSIS MEXICANUS Bailey

Castor canadensis mexicanus Bailey, Proc. Biol. Soc. Washington, vol. 26, p. 191,
Oct. 23, 1913 ("Ruidoso Creek, 6 miles below Ruidoso, New Mexico").

The Rio Grande beaver is closely related to the race frondator
with which it had been associated until Bailey distinguished it
in 1913 as slightly different in its duller and paler colors, with
very little chestnut at any season. He gives the weight of an
adult as 47 pounds and the total length of an adult female as
1,070 mm.

The geographic range of this race is believed to be only the
Pecos and Rio Grande drainage in New Mexico and Texas. In
northern New Mexico it shows a tendency to intergrade with
the darker frondator. The most comprehensive account of it is
that of Bailey (1931), who has gathered what information is
available as to its habits and status. In earlier times the beaver
was a source of food and fur for the Indians in the region and
its remains may be found in their kitchen middens. Bailey

72 EXTINCT AND VANISHING MAMMALS

mentions that in 1826 Pattie and a party of trappers took
beavers all along the Pecos River from 20 miles above its
mouth to the first Spanish settlement (probably Anton Chico).
From 1881 to 1884' L. L. Dyche collected on the headwaters of
the Pecos and found at that time "a considerable number of
beavers about 15 miles from the head of the Pecos River" ; there
were many beaver dams and, he adds, "in the pools which
were caused by these dams I found the best trout fishing of any
locality I have ever visited in the Rocky Mountains," an
interesting comment on the effect of such dams in making
pools where trout may live. In 1898, C. M. Barber found a
large colony on Ruidoso Creek, below the town of the same
name, and there he secured the series of specimens on which
Bailey later founded the name mexicanus. At that time the
beavers were feeding on corn, which they cut and dragged to
the stream and floated to their dens. Very few were being
killed by the Mexicans on the ranches, since they lacked the
technique for trapping. In 1902 Bailey visited Ruidoso Creek
" and found that there were still a few beavers along this stream,
which, with its sections of deep water and steep banks, is
peculiarly adapted to the habits of beavers. " In the following
year he also found some of the streams inhabited by them on
the headwaters of the Pecos. "There were old cuttings along
many of the other streams, but in most cases the beavers had
been entirely trapped out ... In the deep water of the
larger valley streams they are not so easily caught, and fortu-
nately enough have escaped in spite of persistent trapping to
prevent the complete extermination of the species .

"In many places along the canyons of the Rio Grande above
Santa Fe there were still some of the animals in 1903-4, and
trappers were then catching them in considerable numbers
. . . and in 1909, E. A. Goldman saw some old beaver
cuttings near Socorro and was told that there were still a few
along the river. He also saw signs of them at Garfield and
found them common in the Rio Grande near Las Palomas, and
they were reported near Las Cruces ... In 1915, J. S.
Ligon reported them as becoming abundant in places along the
Rio Grande above and below San Marcial, where there were
some complaints of their felling trees across the fences.

"In such localities beavers live entirely in the banks of the
rivers and select the deepest water for their operations. They

NORTH AMERICA AND THE WEST INDIES 73

are not easily trapped and usually remain the longest where they
do real damage, while in the higher mountains where they can
do no damage they are easily caught and quickly destroyed. "
While from an economic standpoint beavers may do some
damage to forest growth and even to crops, as in the case just
noted, by feeding on cornstalks, nevertheless this is easily con-
trolled. Bailey mentions that in 1910 the water company of
Santa Fe was offering $50 a pair for live beavers to be placed
in the upper part of the Santa Fe Canyon to aid in conserving
the city's water supply. He adds : "On almost all the mountain
streams they should be protected and encouraged. A series of
beaver ponds and dams along the headwaters of a mountain
stream would hold back large quantities of mountain water
during the dangerous flood season and equalize the flow of the
streams so that during the driest seasons the water supply
would be greatly increased in the valleys. Beaver ponds not
only hold water but distribute it through the surrounding soil
for long distances, acting as enormous sponges as well as
reservoirs. A series of ponds also increases the fishing capacity
and furnishes a safe retreat for the smaller trout and protec-
tion from their enemies . . * . A legitimate amount of
trapping should eventually yield large annual returns over
extensive areas of the country from which they have been
almost exterminated."

MICHIGAN BEAVER; WOODS BEAVER
CASTOR CANADENSIS MICHIGANENSIS Bailey

Castor canadensis michiganensis Bailey, Proc. Biol. Soc. Washington, vol. 26, p. 192,
Oct. 23, 1913 (Tahquamenaw River, 5 miles above falls, Luce County, Michigan).

This is a very dark race of beaver, with a short, broad skull.
According to its describer, it is dark umber-brown above,
brighter on head and cheeks; ears, feet, and nose black. Under
side darker than back, with blackish on the breast and flanks.
The skull has the rostrum shorter and the nasals more quadrate
than in typical canadensis, the zygomata abruptly spreading,
the occiput with an upright crest, giving a short, "sawed-off "
appearance. Though said to be of small or medium size, the
measurements given do not show this: Total length, 1,170 mm.;
tail, 470; hind foot, 185; skull, condylobasal length, 129;
zygomatic width, 96.4; nasals, 46 by 24 (type skull).

74 EXTINCT AND VANISHING MAMMALS

Although described from the northeastern (or "upper")
peninsula of Michigan, between Lake Superior and Lake
Michigan, the range of this race is believed to include much of
the "Lower Peninsula," now or in the recent past. While at
the present time beavers are regarded as extinct in southern
Michigan, they still occur locally in northern parts of the State.
According to Dice, they were found up to about 1920 along
Boyne River in Charlevoix County but are now rare or extinct
there except that beavers have been reintroduced on Spring
Branch. In the neighboring counties, Cheboygan and Mont-
morency of the Lower Peninsula, they are scarce but seem to
be increasing. Presumably intergradation of this race with
neighboring races takes place in Wisconsin or to the northward,
but the exact limits of the range of this form remain to be
worked out.

Because of the rich dark color of its fur, this race of beaver is
especially recommended by Vernon Bailey for introduction
into other and now depleted localities. It has been much used
in restocking areas in Michigan that seemed suitable, and Ruhl
and Lovejoy (1930) have published a summary of such intro-
ductions carried out in various counties in southwestern,
central, and northern parts of the Lower Peninsula in recent
years. In most cases only one or two animals at a time were
available for this purpose, yet these authors conclude that
"there seems to be no difficulty in getting beaver to settle and
winter even if planted as late as November 1," but there is no
assurance that they will remain in any particular locality nor
was the number necessary for efficient colonization determined.
Where large streams are available, these introduced beavers
were found usually to make burrows in banks rather than to
construct lodges. According to Vernon Bailey (1922), choice
skins of this dark form from northern Wisconsin have in recent
years brought as much as $50 each. He writes further: "So
far as at present known, the darkest, richest-colored, and
handsomest beaver fur is found native along the south shore of
Lake Superior, in northern Michigan and Wisconsin [and is of
the race here under discussion] . In this region of heavy forest
and deep snows the outer hairs of the animals are very dark
brown and the under fur is almost black. When tanned and
plucked the skins are very beautiful, and when made up into
wearing apparel they almost equal sea otter in depth of fur

NORTH AMERICA AND THE WEST INDIES 75

and richness of color. They are decidedly superior to the
Canadian and Alaskan skins . . . For many years the
beavers of the region south of Lake Superior have been care-
fully protected . . . They are now fairly abundant in the
region. "

MISSOURI BASIN BEAVER

CASTOR CANADENSIS MISSOURIENSIS Bailey

Castor canadensis missouriensis Bailey, Journ. Mammalogy, vol. 1, p. 32, Nov. 28,
1919 ("Apple Creek, 7 miles east of Bismarck, North Dakota").

This race is said to be "slightly smaller than canadensis and
much paler and duller brown. Skull more triangular in outline,
not so massive and heavy ; much like that of mexicanus, shorter
and heavier than that offrondator. From mexicanus the colors
differ in being noticeably duller and darker; from frondator,
duller, and not so rusty. " The total length of the type (prob-
ably not fully grown) was 900 mm. "Specimens show closer
affinity with those of the Rio Grande drainage, than with those
in the same State [North Dakota] in the streams flowing into
Hudson Bay."

The range of this form of beaver is believed to be the drain-
age system of the Missouri River, typically in western North
Dakota, but further study is needed to settle this. Provision-
ally, South Dakota, Nebraska, and possibly Kansas may be
included in its present or former range. Vernon Bailey (1926)
has given an excellent account of the recent status of the beaver
in North Dakota. He writes that in 1804-5, Lewis and Clark
found beavers abundant along the Missouri, even near long-
established Indian settlements. At that time trappers were
just beginning to find the region a source of pelts, and in the
three decades following they reaped a large harvest. In 1833
Maximilian reported that 25,000 beaver skins had been bought
during the year at Fort Union (now Buford); by the next
decade, however, Audubon found them already scarce. Never-
theless, Bailey writes, where "the beavers had the protection
of the deep water and high banks of the larger rivers" they
persisted and "with characteristic tenacity they still cling to
their old haunts or merely scatter out to establish new colonies
in tributary streams." As late as 1913, he and others found
small numbers as at Buford, Fort Clark, and near Williston.
At this time also they were present on the Little Missouri and

76 EXTINCT AND VANISHING MAMMALS

its tributaries. In 1919, after two years of open season on
beavers, many of the small colonies were reduced or wiped out;
"a few traces of their work were found along the Missouri
River at Sanish and Bismarck, and there were said to be a few
beavers still in Apple Creek and Burnt Creek. Near the mouth
of the Cannonball River they were very scarce, although they
had been fairly common up to 1916." Where cutting of trees
seems too evident, the local residents often become exasperated
and demand the killing-off of the animals, although the loss of
the trees and at times that of the grain growing near at hand
is less than the value of the beavers.

It is perhaps this form of beaver that was formerly common
on some of the rivers of northern and western Kansas. Hibbard
(1933, p. 241) writes of its recent status, that although once
common throughout the State, along the large streams, it was
exterminated from many parts by early trappers; a few, re-
ferred to this race, are still "found in small scattered colonies
along the Republican, and the Kaw east into Douglas County, "
of eastern Kansas, while a few colonies are reported along the
Arkansas River in the western part of the State.

NORTHWEST COAST BEAVER
CASTOR CANADENSIS PACIFICUS Rhoads

Castor canadensis pacificus Rhoads, Trans. Amer. Philos. Soc., ser. 2, vol. 19, p. 422,
Sept. 1898 (Lake Keechelus, Cascade Mountains, Kittitas County, Washington).

CASTOR CANADENSIS IDONEUS Jewett and Hall

Castor canadensis idoneus Jewett and Hall, Journ. Mammalogy, vol. 21, p. 87, fig. la,
Feb. 1940 (Foley Creek, Nehalem River, Tillamook County, Oregon).

The beaver of the northwest coast from most of Oregon
northward into southern British Columbia was described by
Rhoads as a distinct race in 1898, but for many years thereafter
this name was considered a synonym of leucodonta, the Van-
couver Island race. In 1916, however, Taylor in his review of
the western beavers, reestablished it as a distinct subspecies.
Its characters are slight and relate principally to the form of
the nasal bones, which "in pacificus have their outlines some-
what more invaded laterally by the backward-extending
tongues of the premaxillaries, " beyond which their lateral out-
lines again become somewhat more parallel. This posterior

NORTH AMERICA AND THE WEST INDIES 77

parallel part is longer than in leucodonta. Externally, the color
of the fur is "remarkably close" to that of the Californian
race, subauratus.

Bailey (1936) includes in the range of this beaver in Oregon
the coastal area west of the Cascade divide and the whole
drainage of the lower Columbia River basin and (in eastern
Oregon) the basin of the Snake River. Specimens from the
headwaters of the Deschutes, La Grande, Pine Creek near Pine
in Baker County, Oreg., and from 5 miles south of Walla Walla,
Wash., and from Boise River west of Boise, Idaho, are, ac-
cording to this author, "typical pacificus, " but to the southeast
intergradation takes place with the Nevada race, bailey i; while
in south-central Oregon, in the Klamath section, it is believed
that the beaver is identical with the form shastensis of north-
eastern California.

Very recently Jewett and Hall have further divided this
coastal beaver by describing as a separate race Castor c.
idoneus, a smaller darker subspecies than its immediate
neighbors, with a relatively broad skull and short nasals that
do not extend much, if any, back of the premaxillae. Although
definitely known only from the* type locality (Foley Creek)
and Blaine, both in Tillamook County, Oreg., the range is
believed to include probably the humid coastal area in Oregon,
west of the Willamette drainage. For convenience the two
may be treated together here.

Quoting again from Bailey's account, "a century ago beavers
were abundant in almost every lake and stream in Oregon, so
abundant that the first trapping expeditions brought back rich
returns in fur from the least-inhabited parts of the State. In
the vicinity of extensive Indian settlements beavers were less
numerous, or even scarce, in those days, and in a comparatively
few years of vigorous trapping they became scarce over the
whole State, and later while unrestricted trapping was allowed,
they were reduced to the verge of extermination." As an
example of their abundance, Ogden's trappers from October
1826 till the end of the following March took over 2,200
beavers in the State. West of the Cascades, the Willamette
Valley was considered at one time the finest hunting ground
for beavers in the United States west of the Rocky Mountains,
but by 1824 they had already become scarce, and Ogden in
1827, when he reached the Rogue River, was informed that

78 EXTINCT AND VANISHING MAMMALS

Indian trappers from the Willamette had already visited the
river and taken all the beavers. Even down to 1860, Lord,
traveling along the headwaters of the Deschutes River, re-
ported beavers abundant there. Beavers were gradually re-
duced in number over all this region. Even "at their lowest
ebb a few remained in the larger rivers, however, and during
the past quarter century under the awakened interest in wild-
life and the most rigid legal protection that could be given
them in an area of extensive wilderness, they have come back
to some of their old haunts and increased locally until they
now may be found in many of the streams and lakes of the
State. In fact, they are apparently more common now in the
Grand Ronde and Walla Walla Valleys than they were a
hundred years ago when these valleys were occupied by settle-
ments of Indians who depended largely on the native animals
for food and clothing." In recent years they are reported as
more or less common in a number of counties both east and
west of the Cascade Mountains.

The district forester for Oregon reported in 1930 that the
law providing an open season on beavers had proved to be a
mistake. The eagerness of trappers to be first in the field had
resulted in the greater part of the catch being taken before
the fur was prime, while at the same time the number of
beavers had been very greatly reduced. The effect of their
dams in storing and holding back water during dry seasons was
gone, with much resulting loss. Another observer in 1931
states that "the removal of beaver has been a large factor in
the shortage of water during the drought through which we
are passing. Streams have dried up below former beaver dams
to an alarming extent and water for stock has been reduced."
On a stream where previous to the advent of beavers in 1920
even a bridge was unnecessary, their dams had since increased
so that in the 1931 drought, the driest season on record, "water
was plentiful for a distance of a quarter of a mile below the
beaver dams . . . and at least 20 acres of land that were
dry in the very wet season of 1914 are kept fairly moist."
Such observations indicate the value of beavers in water
storage at the heads of streams during critical seasons.

In southern British Columbia beavers, though "fast disap-
pearing" 40 years ago, are still found in small numbers in
remoter districts.

NORTH AMERICA AND THE WEST INDIES 79

ADMIRALTY BEAVER

CASTOR CANADENSIS PHAEUS Heller

Castor canadensis phaeus Heller, Univ. California Publ. Zool., vol. 5, p. 250, Feb. 18,

1909 (Pleasant Bay, Admiralty Island, Alaska).
FIG.: Taylor, 1916, p. 431, fig. c (tail outline).

The beaver of Admiralty Island was at first believed to be a
very dark race, but Taylor (1916) shows that of six specimens
from Admiralty Island two are hardly different in color from
two New Brunswick examples of typical canadensis, although
others are very dark. The type is described as having the long
hair of the upper parts almost black, with a slight chestnut
tint at the tips; shoulders deep russet, lightening to chestnut-
brown on the head ; underfur very dark seal brown or blackish ;
ears black. On the under side, the longer hairs are seal brown,
becoming slightly reddish at the base of the tail. The ratio of
length to width of tail is even greater than in the southern
broad-tailed beaver, frondator, but the widest part is located
more proximally. The skull is chiefly notable for its narrower
interorbital region, longer and narrower nasals, broader fora-
men magnum, and longer tooth row as compared with typical
canadensis.

This is an island race of beaver, hitherto known only from
Admiralty Island, Alaska, although probably it was present
also at one time on some of the adjacent islands. Heller re-
garded a specimen from Prince of Wales Island as intermediate
toward the race pacificus of the mainland to the south. He
quotes Dixon that there were said to be no beavers on any of
the islands between Sitka and Juneau, so that their discovery
on Admiralty Island was of particular interest. Here they
lived in the lakes, which with their irregular shore line and
quiet little bays made an ideal home. They were seldom
disturbed, although at intervals Indians from the mainland
would come over and take a few. Their main food supply was
found to be spruce bark, although that of willow where obtain-
able was preferred. Four miles to the south of Wide Harbor
it was reported that beavers had been lately exterminated.
On the adjacent Chichagof Island, some old and dilapidated
beaver dams were found, but the beavers seemed to have been
cleared out. Presumably they were of this race.

Concerning the status of the beaver on Admiralty Island in

80 EXTINCT AND VANISHING MAMMALS

the years since its description by Heller in 1909, no recent
information is at hand.

COLORADO RIVER BEAVER
CASTOR CANADENSIS REPENTINUS Goldman

Castor canadensis repentinus Goldman, Journ. Mammalogy, vol. 13, p. 266, Aug., 1932
("Bright Angel Creek, Grand Canyon of the Colorado River, Arizona").

This is said to be a "light-colored subspecies closely allied
to Castor canadensis frondator, of southeastern Arizona and
northeastern Sonora, but upper parts paler, yellowish cinna-
mon, instead of near pecan brown (Ridgway, 1912) as in the
type; cranial characters, especially the long nasals, distinctive.
Similar to C. c. baileyi, of Nevada, but slightly paler, upper
parts more yellowish, less rufescent; skull decidedly broader.
Differing from C. c. mexicanus, of New Mexico, in slightly
yellower coloration (duller brownish in mexicanus) and in
cranial details, especially the longer, less expanded nasals."

Goldman lists specimens from Bright Angel Creek and Yuma,
Ariz.; those from the latter locality with shorter nasals "prob-
ably grade toward frondator" but in the massive jugal and
other respects seem "nearer to repentinus" He adds: "At the
type locality the beavers inhabit Bright Angel Creek which in
short reaches descends the terraced north side, below the outer
rim, of the Grand Canyon. None are known from the Colorado
River in that vicinity, and a measure of isolation would seem
to be due to the rapids and heavy current through rock-bound
gorges extending for many miles, and affording very few places
suitable for beavers to establish homes." According to Dr.
Joseph Grinnell (1933) it was "originally and recurrently
numerous along those parts of this river's course where willow
and cotton wood grow abundantly . . . Since 1911, bea-
vers have invaded Imperial Valley north of Mexican line
following Alamo River and larger distributary canals."

CASTOR CANADENSIS SAGITTATUS Benson

Castor canadensis sagitiatus Benson, Journ. Mammalogy, vol. 14, p. 320, Nov., 1933
("Indianpoint Creek, 3200 feet, 16 miles northeast of Barkerville, British Colum-
bia")-

FIG.: Benson, 1938, fig. c, p. 321 (skulD.

This local race of beaver is described as dark colored, with
the ventral underfur dark grayish; "skull with narrow, pointed

NORTH AMERICA AND THE WEST INDIES 81

rostrum; angle between side of rostrum and zygomatic arch
obtuse, so much so as closely to approach a straight line;
nasals anteriorly narrow and pointed." The slight cranial
peculiarities on which this form is mainly based are not very
striking, but are shown in comparison with skulls of five other
races in the figures accompanying the paper by Benson cited
above; he also makes brief comparisons of its characters with
those of the surrounding western races, and presents tables of
cranial measurements.

This race is "known only from the Cariboo Range, in east-
central British Columbia." It is supposed, however, that it
meets the range of typical C. canadensis in the mountains,
while in south-central British Columbia it probably intergrades
with the race pacificus, and Davis has lately identified as of
this race a skull from northern Idaho. Presumably, too, its
range meets that of being ae "in the northern portion of the
Eraser River drainage system." Concerning its general status
in the east-central part of British Columbia little information
is at hand, but according to Benson it apparently exists in
considerable numbers in the mountains of this region. Cowan
(1939) in his recent account of the vertebrates of the Peace
River district, British Columbia, writes of the beaver, which
evidently represents this race, that it is "rare over most of the
district but according to reports of trappers increasing on
certain areas thanks to the combined efforts of the game
wardens and certain of the more intelligent trappers." He
mentions that a large female from South Pine, British Colum-
bia, and a yearling male from Pine River are typical of cana-
densis and exhibit no trace of intergradation with sagittatus,
the race to the west of the mountains.

SHASTA BEAVER
CASTOR CANADENSIS SHASTENSIS Taylor

Castor subauratus shastensis Taylor, Univ. California Publ. Zool., vol. 12, p. 433,
Mar. 20, 1916 ("Cassel [Hat Creek], Pit River, Shasta County, California").

This race is based on cranial characters of relatively slight
significance. It is closely similar to the race subauratus but
although occurring in the same drainage basin, is found on the
eastern or Great Basin side of the Sierra Nevada. The distinc-
tion is based mainly on the outline of the nasal bones, which

82 EXTINCT. AND VANISHING MAMMALS

show an indentation in the outer margin posteriorly, due to the
premaxillary bone; the interorbital constriction is less than in
subauratus, while the temporal ridges tend to unite to form a
distinct sagittal crest posteriorly and show a "higher degree of
approximation anteriorly." The interparietal also is somewhat
broader.

The type locality "is situated on Hat Creek, a tributary of
the Pit River, which is in turn a tributary of the Sacramento
River." Thus, although living in the same hydrographic basin
as the form subauratus, its range on the eastern side of the
Sierra Nevada is probably bounded by the Pit River Narrows,
"a barrier not regularly crossed by beavers." It is possible,
according to Benson, that the beavers of the Great Basin may
be identical with this form, but the eastern range is not defi-
nitely worked out. Grinnell (1933) gives its area of distribu-
tion as the Pit River basin in the northeastern corner of
California and adds that it is recorded from "Pit River, above
Narrows, north to Willow Creek and Steele Meadow, near
Clear Lake, and east to Lassen Creek, east of Goose Lake, and
to North Fork of Pit River above Alturas these localities
being in Modoc County." Its present status is not altogether
clear, but Grinnell, Dixon, and Linsdale (1937) have sum-
marized all the information they could obtain as to the status
of this beaver in northeastern California and the adjacent
region up to late 1931, as follows: In September of that year
three colonies were known on Fletcher Creek, one consisting of
6 animals at Willow Creek Ranch, another of 10 or 12 indi-
viduals somewhat farther up Fletcher Creek, and the third
and largest in the entire section comprising between 20 and 25
beavers near the "Mulkey place." Another colony, consisting
of about 15 beavers, lived on South Willow Creek, but a colony
formerly living on the North Fork of Willow Creek had ap-
parently been destroyed by trapping. About 8 miles north of
Canby lived a colony of half a dozen beavers. The largest
colony in the county was on Lassen Creek and comprised about
50 animals. It extended from a point about 2 miles above
where the creek enters Goose Lake for a distance of 2J/2 miles.
Four or five individuals lived on Davis Creek where it runs
into Goose Lake, and another group of the same size was
known on the north fork of the Pit River 16 miles north of
Alturas. They formerly (1898) occurred in the Shasta River

NORTH AMERICA AND THE WEST INDIES 83

and Scott River, Siskiyou County, flowing into the Klamath
River, and there is evidence that beavers "were present in the
Klamath River above Requa at least during the years 1915,
1916, and 1917" and were probably of this race. From the
figures given, the approximate population of this beaver in
California in 1931 must have been about 120 individuals.

The range extends into extreme south-central Oregon along
the Klamath River drainage to Goose Lake, a region to the
east of the Cascade Mountains, and Bailey (1936) believes
that the beavers of the Klamath section, Lost River, Sprague
River, and the Yamsay Mountains, are also of this form. As
late as 1860 beavers were very abundant in this district,
There are still a few here, where they seem to be holding their
own or even increasing in recent years (Bailey, 1936).

GOLDEN BEAVER
CASTOR CANADENSIS SUBAURATUS Tajlor

Castor subauratus Taylor, Univ. California Publ. Zool., vol. 10, p. 167, May 21, 1912
("Grayson, Stanislaus County, San Joaquin River, California").

FIGS.: Taylor, 1916, fig. H, a, p. 431 (tail outline); fig. J, b, p. 450; fig. K, 6, p. 451;
fig. L, 6, p. 452 (skull); Grinnell, Dixon, and Linsdale, 1937, vol. 2, pi. 8 (colored
fig. of exterior) .

Fur with a golden sheen dorsally and ventrally; size largest
of the western beavers. Skull with the nasal bones more
expanded and foramen magnum wider in proportion to its
height than in any other western beaver. The average propor-
tion of width of tail to length is about 42 percent, in the race
leucodonta about the same, but in frondator greater, nearly half
(49.1 percent). The type measured: Total length, 1,171 mm.;
scaled part of tail, 320; hind foot, 196.

This is the beaver of the "lower courses of San Joaquin and
Sacramento Rivers and lower portions of larger tributaries of
these main rivers" below 1,000 feet altitude in western Cali-
fornia. Grinnell (1933) gives its former range as "from Tulare
Lake (formerly), Kings County, and from Kings River, near
Sanger (formerly), and at Mendota in Fresno County, north
to Sacramento River and Butte Creek, north of Marysville
Buttes, and at one time to McCloud and upper Sacramento
rivers, in Shasta County." In 1912 Taylor wrote that this
beaver was then "approaching extinction, although the en-
forcement of the present protective law may enable it to

84 EXTINCT AND VANISHING MAMMALS

regain a foothold." Twelve years later Grinnell and Storer
(1924) reported that the protection thus accorded was effec-
tive "and the prospects for their perpetuation are now bright."
Indeed, by 1920 they had so increased as to become troublesome
at Snelling, through interfering with irrigation works. In the
river basins of western California, these beavers feed princi-
pally on the bark of willow and cotton wood. At the present
time they are found in some numbers in the extreme western
end of the Yosemite section, along the lower courses of the
Merced and Tuolumne Rivers west of the foothills. With
proper management it is expected that this population may be
maintained in localities where it does not interfere with human
activities. This seems especially desirable, since the golden
beaver is one of the better-marked races of North American
beavers.

A very thorough account of this beaver in California has
been published by Grinnell, Dixon, and Linsdale (1937, vol. 2,
pp. 629-721) in their recent work on the "Fur-bearing Mam-
mals of California." They find that it is more a "bank"
beaver than most of its relatives, adapting its habits to life
along sluggish lowland streams. The largest specimen of
which there is authentic record weighed no less than 82 pounds.
They indicate that because of individual variation the shape
of the tail is of less diagnostic value in the discrimination of
races than had hitherto been supposed, and further it appears
from measurements of a series of fresh specimens that in males
its width is proportionally greater than in females. The
period of gestation is about three months, and two to four
young is the usual number per litter. Beavers in California
are now "just about free from any natural enmy," with the
reduction of bears, mountain lions, and river otters. Although
the number of beavers in a given locality is often overestimated,
these authors believe that in 1920 there were about 250 inhab-
iting 15 square miles of river bottom in the vicinity of Snelling.
Trapping on the San Joaquin River near Mendota in the same
year indicated about 12 animals to the square mile. On this
basis it was calculated that in 1921 there were about a thousand
golden beavers in California. Much of the area formerly
inhabited by beavers is now reclaimed and in use for agricul-
tural purposes in the delta region, so that the beaver's range
has become more and more restricted, and by 1911 very few

NORTH AMERICA AND THE WEST INDIES 85

were left in that district. With total legal protection given
them in 1911, however, "they had bred up and had become
numerous enough to endanger the levees along Cache Slough,
and in that year seventeen beavers were trapped there, under
special permit." Ten years later beavers had greatly increased
in the Merced River bottom in the region about Snelling and
became something of a nuisance to ranchers by stopping up
irrigation canals. In 1925 legal protection of the golden
beaver was removed, and during the two years following the
population was severely reduced. From that time until 1931
the "decrease has continued until it seems possible that
eventually this species may be eliminated entirely from the
Great Valley of California ... if protection is not
restored."

In agricultural areas in California the damage done by
beavers is summed up by the authors quoted under six heads:
Burrowing into levees or dikes of reclamation work; burrowing
into banks of irrigation canals; obstruction of drainage canals
by dams; flooding and waterlogging of land; gnawing at head
gates of irrigation canals; and cutting down of fence posts and
fruit trees. They also mention the complaints of certain
ranchers that beavers damage alfalfa fields and fruit trees in
their use of these products as food. On the other hand they
may, in regions not under immediate use, be of much value in
the activities of storing water, and instances are mentioned in
which the water held in beaver ponds has saved crops from
drought, or their ponds may be utilized by installing pumps to
distribute water for irrigation. "One rancher states that
beavers have saved him several hundred dollars each season in
this way ... In several known instances, beavers have
lived in streams and ponds of pasture lands among cattle and
hogs . . . with little or no competition with" the stock.
"There appears to be no good reason why beavers, through
their fur value, could not be made to yield a regular income on
such ranches." The authors of the "Fur-bearing Mammals
of California" "believe that much of the waste, dredged-over
land along the Merced, Tuolomne, and American rivers,
wherever these animals are not already reestablished, could be
used to support beavers and thus be made an asset to the
entire State."

8G EXTINCT ND VANISHING MAMMALS

SNAKE RIVER BEAVER

CASTOR CANADENSIS TAYLORI Davis

Castor canadensis taylori Davis, The Recent Mammals of Idaho, p. 273, Apr. 5, 1939
("Big Wood River, near Belle vue, Blaine County, Idaho").

The beaver of the Snake River drainage basin in southern
Idaho and northern Nevada has lately been distinguished on
the ground of its "averaging darker" than the race bailey i;
"nasals long and narrow (breadth averaging 46 percent of
their length); anterolateral rim of orbit narrow (near 7.0 mm.);
occiput nearly vertical."

No details are at hand concerning the past or present status
of this beaver. Davis based his description on 11 specimens
from Ada, Lemhi, Blaine, -and Bannock Counties in Idaho and
four from Goose Creek, Elko County, Nev.

TEXAS BEAVER

CASTOR CANADENSIS TEXENSIS Bailey

Castor canadensis texensis Bailey, North Amer. Fauna, no. 25, p. 122, Oct. 24, 1905

("Cummings Creek, Colorado County, Texas").
FIGS.: Davis, 1940, pp. 85-86, figs. 1, A, and 2, D.

This seems to be a slightly marked form, described as re-
sembling frondator in its pallid coloring, but distinguished by
the skull, in which the sagittal crest is short and the lateral
ridges lyrate or spreading even in old age; supraoccipital crest
doubly curved, nasals long, spatulate, and tapering to a
narrow point posteriorly. These characters shown in the three
specimens studied by the describer "are so well marked and
uniform as to justify describing the subspecies, even on so
scanty material." The localities given are on the Colorado
River of east-central Texas, but "whether the beaver of other
streams north and south" of this valley are the same "can be
settled only by specimens." More recently the matter was
reviewed by Dr. W. B. Davis (1940) with additional specimens,
and he concludes that the distinctive characters of texensis
are its "relatively darker color; posteriorly pointed nasals,
with an expanded area near middle of rostrum; dorsal outline
of skull nearly flat, except for the distal portion of the rostrum
which is abruptly depressed; tail relatively long and narrow,
ratio of width to length, 48 (39-55)" as compared with the

NORTH AMERICA AND THE WEST INDIES 87

race mexicanus. His specimens were from Blanco, Kimble,
and Edwards Counties.

To the eastward, intergradation with the race carolinensis
may be expected, if specimens become available; while to the
westward the beavers of the Rio Grande and Pecos Rivers
probably represent the race mexicanus.

Writing in 1905, Bailey states that "beaver are still found
in many of the streams of eastern Texas, especially in the
larger rivers, where deep water and steep banks afford protec-
tion against relentless trapping. In 1892, at Arthur, in north-
eastern Texas, I was informed that they were fairly common
along the Red River and that trappers caught a few each year.
In 1902, at Texarkana, Oberholser was told that a few were
still found in the Red River, and in 1901, at Mobeetie, was
informed that they were common in Sweetwater Creek, a
branch of the North Fork of the Red River." A few were
found at this time in scattered colonies along the Colorado
River and in the Brazos and Trinity Rivers. In 1904, Bailey
was told that in Polk County, in eastern Texas, beavers were
abundant only a few years before. In northeastern Texas, on
the upper Sabine River, trappers annually obtained them.

A. H. Cook (1940), in a study of live Texas beavers captured
in the summer of 1939, describes an infestation, of three out of
40 animals examined, by the screw- worm fly (Cochliomyia
americana), which had deposited eggs in wounds on sides,
back, or hind limbs. He suggests that such wounds were due
to fighting as a result of overcrowding. One of the infected
animals seemed unable to control movements of the hind legs
and died; while another, treated with "bone oil" (a fly repel-
lent), recovered.

Family CRICETIDAE: Hamsterlike Rats and Mice

JAMAICAN RICE RAT
ORYZOMYS ANTILLARUM Thomas

Oryzomys aniillarum Thomas, Ann. Mag. Nat. Hist., ser. 7, vol. 1, p. 177, 1898
("Jamaica").

The Jamaican rice rat is believed now to be extinct, although
it lived up till about the 1880's. Probably the introduced
mongoose had a large share in its disappearance, for the last

88 EXTINCT AND VANISHING MAMMALS

known examples were taken about 1877, only five years after
the introduction of that carnivore into the island. These
specimens are in the United States National Museum. An
earlier example preserved in the British Museum was made, in
1898, the type of the present species by Thomas, who describes
its color briefly as dull rufous above, rather richer on the rump
and grayer on the head, with a slight lining of black on the
back; under side dull yellowish, not sharply defined, the hairs
slaty gray basally; tail about as long as head and body, nearly
naked, hands and feet dull whitish. Total length about 10
inches (252 mm.) of which the tail is slightly more than half:
hind foot, 29.2; length of skull, 30.5 (Goldman, 1918).

This rice rat, as Thomas pointed out in his original descrip-
tion, is very similar to the mainland 0. couesi or some of its
Central American races. Indeed, it seems doubtful if it
should not be regarded as identical with some one of them, for,
as the only cricetine rodent known from the Greater Antillean
islands, it is obviously a recent intrusion in a rodent fauna
otherwise of a hystricoid type. In his review of the North
American rice rats, Goldman (1918) compares the skull with
that of typical couesi of Yucatan and Honduras but finds
relatively few points of difference, namely, that the nasals
reach slightly farther back beyond the premaxillaries, the
maxillary arm of the zygoma is heavier, and the anterior
palatine foramina are shorter than usual in couesi. Since,
however, only the two specimens in the U. S. National Museum
(from Metcalfe Parish and Spanish Town, respectively) were
available for comparison, it seems likely that these differences
may not be very significant. Without further evidence to the
contrary the name may for the present be retained, but at the
same time the possibility is recognized that the animal may
have been a rather recent introduction through human agency.
That it nevertheless thrived and multiplied is evident from the
fact that Dr. H. E. Anthony (1920a) in exploring caves along
the seacoast of Jamaica found none of the subfossil types of
hystricoid rodents in them, though they often did yield "re-
mains of the extinct rice rat, thus showing that at no very
remote period this small rodent had a widespread distribution
and was so common that it formed an important part of the
diet of the barn owl," by which evidently they were brought
in to the caves. "The reason for its disappearance obviously

NORTH AMERICA AND THE WEST INDIES 89

lies in the advent of the ubiquitous Norway rat and the blood-
thirsty mongoose."

Doubtless this is the field mouse described by P. H. Gosse
in his "Naturalist's Sojourn in Jamaica" (Gosse and Hill, 1851)
as of a "beautiful reddish colour, with a milk-white belly."
"It takes up its habitation chiefly about the hollow roots of
large trees, and the rocky acclivities of gullies and river banks.
It is far from numerous." The type specimen in the British
Museum was collected by Gosse in 1845.

ST. VINCENT RICE RAT
ORYZOMYS VICTUS Thomas

Oryzomys victus Thomas, Ann. Mag. Nat. Hist., ser. 7, vol. 1, p. 178, Feb., 1898 ("St.
Vincent, Lesser Antilles").

Thomas, in describing this species, wrote: "It is difficult to
say to what species this mouse is most nearly allied." At that
time he compared it with the South American Oryzomys longi-
caudatus, giving it the following characters: Size and propor-
tions about as in 0. longicaudatus; color dark rufous, but this
probably affected by the alcohol in which the type is preserved ;
below, buffy white, the hairs with slate-colored bases. Length
of head and body, 96 mm. ; tail, 121 ; hind foot without claw, 25 ;
basal length of skull, 23.8.

The type and only known specimen was presented to the
British Museum by F. DuCane Godman in 1897. Since that
time no further attempt to determine the nearer affinities of
the specimen has been made, nor have any additional examples
been taken. It therefore still remains problematical whether
it is a native or an introduced animal, what its continental
relatives are, and whether it still exists on the island. Since
no representative of the genus is known from any other of the
Lesser Antilles, it may prove to have been introduced in St.
Vincent through local trade from some adjacent part of South
America; on the other hand, the type specimen collected by
H. H. Smith was marked by him "forest rat," implying its
existence in wooded areas on the island. Goldman (1918, p. 87)
in speaking of it, says: "This rice rat seems likely to be en-
dangered by the presence of the mongoose, if it has not already
been exterminated since the introduction of that indiscrimi-
nately destructive animal."

90 EXTINCT AND VANISHING MAMMALS

MARTINIQUE MUSK-RAT; "PILORIE"
MEGALOMYS DESMARESTII (J. B. Fischer)

Miis desmarestii J. B. Fischer, Synopsis Mammalium, p. 316, 1829 ("In Insula Mar-

tinica").
SYNONYMS: Mus pilorides Desmarest, Diet. Sci. Nat., vol. 44, p. 483, 1826 (not of

Pallas, 1778); Hesperomys (Megalomys) pilorides Trouessart, Le Naturaliste, no.

45, p. 5, 1881; Oryzomys piloris Forsyth Major, Ann. Mag. Nat. Hist., ser. 7, vol.

7, p. 205, 1901 (based on Castor Piloris of Zimmermann, not a technical name).
FIGS.: Geoffrey and Cuvier, 1830, vol. 4, pi. 258 ("Le Pilori"); Trouessart, 1885.

ST. LUCIA MUSK-RAT
MEGALOMYS LUCIAE (Forsyth Major)

Oryzomys luciae Forsyth Major, Ann. Mag. Nat. Hist., ser. 7, vol. 7, p. 206, Feb., 1901
("Santa Lucia," Lesser Antilles).

BARBUDA MUSK-RAT
MEGALOMYS AUDREYAE Hopwood

Megalomys audreyae Hopwood, Ann. Mag. Nat. Hist., ser. 9, vol. 17, pp. 328-330, Mar.,

1926 (Barbuda, Lesser Antilles).
SYNONYM: Megalomys majori Trouessart, Cat. Mamm. Viv. Foss., ed. 2, pt. 2, p. 415,

1904 (nomen nudum).
FIG.: Hopwood, 1926, pi. 12.

The so-called "Musk-rat of the Antilles" may formerly
have occurred on most of the islands of the Lesser Antilles,
but it has by now probably been quite exterminated. Although
its range may once have been more extensive, it is certainly
known only from the islands of Martinique (the typical form,
M. desmarestii), St. Lucia (the smaller form, M. luciae), and
from the little island of Barbuda, where a sub fossil jaw was
found. These were rather large rodents, with a head and body
length up to 360 mm. (14.5 inches) and tail only a little shorter,
and belong to the group of sigmodont species, abundantly
represented in South and Central America.

Th Martinique form was the one first named and is the
larger of the two known from complete specimens. Trouessart
describes a specimen in the Paris Museum as having the head
and body 360 mm. long, the tail 330 mm. The head was
rather short in appearance and rounded, the muzzle more
obtuse than in a house rat, and the upper lip with a deep
vertical furrow. The ears were well developed and nearly

NORTH AMERICA AND THE WEST INDIES 91

naked. The fur was long and harsh but not spiny and of a
dark reddish brown, both above and below. The tail was
black at the base, with a white intermediate area and a black
tip. Desmarest, however, speaks of the color as lustrous black
except for the chin, throat, and base of the tail, which are
white. The skull as figured by Trouessart (1885) is provided
with a conspicuous bony ridge on each side from the upper
anterior edge of the orbit back over the brain case to the lateral
corner of the cranium behind. The skull is rather squarely
truncate posteriorly; 70 mm. long.

The St. Lucia musk-rat was slightly smaller, with the greatest
length of the skull fron tip of nasals to foramen magnum about
49 mm. The color was apparently much the same as in the St.
Vincent form, but the belly was nearly wholly brown instead of
largely white.

That the musk-rat occurred on Barbuda to the northward is
interesting, for in this extension to the northern Lesser Antilles,
there is a certain parallelism in distribution with the agoutis.
The Barbuda species, M . audreyae, is based on a lower jaw and
an upper incisor tooth found in a cave breccia, which, though
presumed to be of Pleistocene age, may not be of any great
antiquity. The jaw seems slightly smaller still than that of
the St. Lucia musk-rat, and in certain slight details of the last
two molars that are preserved shows a less development of the
small accessory cusps, or paraconids.

The Martinique musk-rat was first mentioned in literature
by Du Tertre in 1654, in his "Histoire Generale des Isles de S.
Christophe, de la Guadeloupe, de la Martinique, et Autres
dans I'Amerique." He did not know of it from any of the
French islands except Martinique, where, he relates, it was
commonly eaten by the people, who, after first singeing off the
hair, exposed the body overnight to the air, and then boiled it,
throwing off the first water in order to get rid of the strong
musky odor. It was said to live in burrows in the ground and
against it the colonists waged war on account of its destructive
habits in their plantations. In addition to human enemies,
the large serpents of Martinique also attacked it. Du Tertre
mentions killing a large snake in the stomach of which was one
of these rats "almost as big as a cat." Of preserved specimens,
there appear to be scarcely above half a dozen. The specimen
figured by Geoffrey and Cuvier is a mounted one in the Paris

92 EXTINCT AND VANISHING MAMMALS

Museum, one of two or three collected by M. Plee, who lived
at Martinique from 1821-26. Another from the same collector
is said to be in the Leiden Museum. The Paris Museum has
another mounted one brought back by a M. Le Prieur (Pdate)
with no other locality but "des Antilles." The teeth of this
last specimen, but little worn, and the skull and worn dentition
of another are figured by Trouessart (1885). Exactly when the
Martinique musk-rat became extinct is difficult to say, but
that it may have existed up to the end of the nineteenth
century is indicated by the following statement communicated
to me by the late Dr. G. Kinsley Noble, who in 1914 visited
Guadeloupe and was told by a Mr. Delphin Duchamp, a
former resident of Martinique, that "about five years before
the eruption of Mount Pelee [1902] there used to exist in great
numbers among the cocoanut plantations along the Riviere
Blanche, close to St. Pierre, a species of rat which was black
as coal on the back and white as milk below. When adult
this creature was some 40 cm. long without the tail. I killed
many of them, for their flesh is very delicate. The negroes
call this rat the pilorie. It lives almost entirely in [Pamong]
the cocoanut trees but will take to water when driven from
shelter. I never heard of it occurring in any other part of the
island except on these plantations, and since this whole region
was destroyed by the great eruption of 1902, it is very probable
that the rat is now extinct." Inquiries among the hunters on
Guadeloupe showed that it was unknown to any of them as
occurring there, though they knew of its former presence on
Martinique. No doubt the devastating eruption of Pelee
accompanied by clouds of poisonous gas completed its destruc-
tion.

Of the St. Lucia musk-rat, Trouessart (1885) records a speci-
men preserved in the Paris Museum and calls attention to
its smaller size and less amount of white below as compared
with the "pilorie" of Martinique. This individual was
brought back by Bonnecour some time previous to 1881. The
only other specimen extant seems to be the one now in the
British Museum, which served as the type of Forsyth Major's
"Oryzomys" luciae. This in turn was doubtless the one men-
tioned as received from St. Lucia by the Zoological Society
of London in 1849 (see Proc. Zool. Soc. London, 1849, p. 105).
It was sent by Lieutenant Tyler and may have lived in London

NORTH AMERICA AND THE WEST INDIES 93

for three years, since the accession date as shown by its number
in the register of the British Museum was 1852. Possibly this
species was exterminated prior to the final destruction of the
Martinique species.

There is nothing to indicate when the Barbuda musk-rat
was living, but probably it disappeared soon after the occupa-
tion of the island by Europeans and consequent destruction of
cover. Even in Du Tertre's time the musk-rat was not known
from any of the other islands, for De Rochefort's assertion
that it was one of the five species of mammals native to Tobago
may be received with caution.

MUSKEGET ISLAND BEACH MOUSE

MICROTUS BREWERI (Baird)

Arvicola breweri Baird, Mammals North Amer., Pacific Railway Repts., vol. 8, p. 525,

1857 (Muskeget Island, Massachusetts).
FIGS.: Miller, 1896, pi. 1, figs. 1, la (skull).

The Brewer's beach mouse of Muskeget Island, lying be-
tween Marthas Vineyard and Nantucket Island off southeast-
ern Massachusetts, is of special interest on account of the
characters it has developed of large size and pallid coloration,
differing in the latter respect particularly from all the other
island meadow mice of our coast. It is generally supposed that
these traits have resulted from isolation and inbreeding on the
confines of this small sandy island. On the other hand, it may
be that the beach mouse is a relict species that was formerly
an inhabitant of the now- vanished sandplain, which extended
coastwise from New Jersey to Newfoundland, and forms a
case parallel with that of the Ipswich sparrow of Sable Island,
Nova Scotia, a pallid form of that sandy spot, not found
elsewhere at the present day.

In addition to its somewhat larger size when adult, as com-
pared with the common Microtus pennsylvanicus of the neigh-
boring mainland and the island of Nantucket, the beach mouse
is strikingly pale, a "light gray throughout, purest (almost
white) on the belly and tinged with wood-brown on the sides
and back, the latter somewhat darkened by a sprinkling of
longer blackish hairs . . . Tail indistinctly bicolor, brown-
ish dorsally, whitish ventrally" (Miller, 1896). A large per-
centage of these mice have a white mark in the forehead. The

94 EXTINCT AND VANISHING MAMMALS

skull in adults is larger than that of typical M. pennsylvanicus,
with somewhat longer and narrower brain case, more abruptly
flaring zygomatic arches, and the outline of the interparietal
is longer and narrower. Average adult length, about 200 mm.
(8 inches).

Muskeget is distant only 6 miles from the western end of
Nantucket. The intervening waters are shallow and on the
south side it is partly protected by some small islands and long
beaches. In his account of this mouse and its island, Miller
(1896) writes that in 1887 the United States Coast Survey
found the southern point extending out much farther to the
southwest than in later years, during which the sea washed
away all but the tip of the point, which remained as South
Point Island. He found about a dozen species of birds breeding
there and the white-footed mouse occurred in the clumps of
beach plum, or about the numerous small houses along the
south shore of the island. The beach mice seem to have had
various ups and downs. In 1869 they were recorded as "exces-
sively abundant"; in 1890, however, only a few were found by
two naturalists visiting the island, while in 1891, William
Brewster was unable to find any trace of them. Miller records
that the warden, Mr. Sandsbury, who was stationed there in
summer to protect the tern colony and who had been familiar
with the mice for many years, believes that they had actually
been exterminated by cats that ran wild on the island after the
burning of the Life Saving Station a few years before. Since
the mice can burrow but little in the loose sand, they may
have been the more easily exterminated in this way. When
Miller visited the locality in 1892, he found, however, that
there were large and thriving colonies on South Point Island
and the adjacent Adams Island, which had been cut off from
Muskeget previous to the burning of the Life Saving Station.
This accident had therefore saved the mice from total extinc-
tion. Of the further history of this interesting population,
Miller (1896) writes: "The 28th of December, 1892, I spent
on Muskeget and the neighboring islands. The mice were as
numerous as before on South Point Island, but the colony on
Adams Island had greatly diminished. In June, 1893, I again
visited Muskeget, this time in company with Mr. Outram
Bangs and Mr. Chas. F. Batchelder. We found that the
Microtus colony on Adams Island had entirely disappeared.

NORTH AMERICA AND THE WEST INDIES 95

On South Point Island, however, the mice were so abundant
that in less than two hours we caught forty -three. After
selecting as many as we wanted for specimens, we turned out
twenty -six on Muskeget. Although I have not been at Muske-
get since 1893, I have heard that the mice on the main island
are increasing rapidly. Mr. Sandsbury has written several
times to this effect, and Mr. W. K. Fisher collected two dozen
specimens for the U. S. Department of Agriculture during
July, 1895. Mr. Fisher found two colonies both in clumps
of beach plum bushes one at the east end of the island, near
the place where we liberated the mice in 1893, the other about a
mile farther west. He also found that the species was extinct
on South Point Island." The Muskeget colonies have ap-
parently continued to flourish since their reintroduction, for as
lately as the summer of 1937 Donald R. Griffin obtained a few
specimens there. The island is now a reservation for nesting
terns and other birds and is provided with warden service
through part of the year. To this end, also, the chance intro-
duction of cats is especially guarded against, so that for the
present the outlook is bright that the mice will continue to
survive. But it is obvious that any important change in the
living conditions, the introduction of parasites or disease, or
the lessening of human guardianship may again imperil them.
Concerning the special habits of these mice, Miller (1896)
writes further that they have been "modified to meet the needs
of a life among coarse, loose sand, in which no extensive bur-
rows can be made except during the brief and irregular periods
in winter, when the surface is frozen. Throughout the greater
part of the year the animals are exposed to the full force of the
elements, their only natural protection being that furnished
by the scant beach vegetation or by fragments of drift wood
and wreckage. Where the mice are abundant, a labyrinth of
well beaten paths crosses the sand in every direction, and
when one of the animals is chased, he follows these runways
aimlessly and helplessly, until exhausted or until he finds some
place of refuge, perhaps in a tuft of Ammophila, or beneath the
stalks of a beach golden rod. Occasionally a mouse tries to
escape by entering a burrow among the roots of the beach
grass, but these tunnels always afford a very insecure protec-
tion as they are close to the surface and seldom extend more
than a few inches, or at most a few feet. The mice apparently

96 EXTINCT AND VANISHING MAMMALS

make these short tunnels with the sole object of reaching the
soft parts of the beach grass stems. They are in no way con-
nected with the nesting burrows. Another means of shelter is,
so far as I know, peculiar to the Muskeget mice. They con-
struct frail nests or forms which may be seen scattered about
everywhere. The forms are usually open at the top and made
of the finer shreds torn off from the beach grass. Each is
large enough to contain one animal only. The walls are much
thinner than those of the breeding nests, being a mere film of
grass fiber, through which the occupant can be distinctly seen.
The forms may be built on the bare sand or under some shelter
indifferently. Often one partly arches a beaten path.

"The breeding nests, which resemble those of Microtus
pennsylvanicus, are sometimes made under the protecting
stalks of the luxuriant Solidago sempervirens or under cover
of a fragment of wreckage. When no such convenient shelter
can be found the mice construct short nesting burrows. These
are from one to two feet in length and penetrate the sand at an
angle of about 45. In this way the tunnel descends quickly
through the dry crumbling sand at the surface to the more
compact layers beneath, where the walls are less likely to cave
in and smother the helpless young. At the end of the tunnel is
a hollow, completely filled by the bulky nest . . . Four or
five young is the usual number in a litter . . . The food
of Microtus breweri consists chiefly of the tender bases of the
beach grass stalks ... As the supply of Ammophila on
Muskeget is practically unlimited the beach mice never suffer
from hunger, but the struggle for existence must be fierce,
notwithstanding, for few small mammals live in as exposed
situations."

DESERT VOLE; AMARGOSA MEADOW MOUSE

MICROTUS CALIFORNICUS SCIRPENSIS Bailey

Microtus scirpensis Bailey, North Amer. Fauna, no. 17, p. 38, 1900 ("Amargosa
River, near Nevada line, Inyo County, California").

This is a desert-living race of the widespread M. calif ornicus,
from which it differs in being slightly brighter in color, a pallid
neutral gray above, but not so black as in the neighboring race
mariposa of the western foothill belt of the Sierra Nevada;
belly smoky gray; tail indistinctly bicolor, brown above, gray-

NORTH AMERICA AND THE WEST INDIES 97

ish below; feet brownish gray, not dusky. Six adults averaged
in total length, 203 mm., about 8 inches; tail, 65; hind foot,
25.1. Skull, basal length, 31 mm.; zygomatic breadth, 19;
upper molar series, 8.7.

When first described by Bailey this race was known only
from the type locality, where it lived in wet ground under tall
tules (Scirpus olneyi), in a little marsh surrounding a warm
spring. Here the mice had made extensive runways through
the mud and water. Kellogg (1918), in reviewing the meadow
mice of the calif ornicus group, points out that it is an outlying,
isolated form, a relict of what was doubtless in former times
a more or less continuously distributed species. "It is a re-
markable race because of its occurrence away from the main
mountain axes and in an area of extremely high summer
temperature" in southeastern California. Since assiduous
trapping at the original locality in May, 1917, failed to yield a
single specimen, Kellogg believed that it had "been extermi-
nated within recent years, as the type locality, a small tule
marsh near Shoshone, Inyo County, has been burnt over for
several consecutive years and is now being used as a hog
pasture." Nevertheless, it seems that a remnant persists, for
Dr. Joseph Grinnell writes in response to Dr. Francis Harper's
inquiries, in 1937, that it has been "rediscovered within the
last few years, not to be sure, at the type locality, but within a
few miles' radius, as a result of intensive search. This is to be
credited to Miss Annie M. Alexander, who found them in
exceedingly small numbers at two desert seepages, where,
however, the favoring conditions might at any time be wiped
out." On account of its living in such very localized swamps,
where a series of dry years or drainage and fire might at any
time destroy its habitat, the race is in a precarious state much
like that of an animal living on a few small islands where the
environment may become unsuitable through slight changes.

GULL ISLAND VOLE

MlCROTUS PENNSYLVANICUS NESOPHILUS Bailey

Microtus nesophilus Bailey, Science, new ser., vol. 8, p. 782, Dec. 2, 1898 (Great Gull

Island, at entrance to Long Island Sound, New York).
SYNONYM: Microtus insularis Bailey, Proc. Biol. Soc. Washington, vol. 12, p. 86, Apr.

30, 1898 [not Lemmus insularis Nilsson = Microtus agrestis (Linnaeus)].

98 EXTINCT AND VANISHING MAMMALS

This meadow mouse is an island race of the common Micro-
tus pennsylvanicus of eastern United States, from which it
differs only in minor characters. At the present time it is
believed to be extinct.

Bailey describes this as darker than typical M . pennsylvani-
cus, the upper parts dark yellowish bister, heavily mixed with
black hairs, darkest on nose and face; belly dusky, washed
with cinnamon; feet blackish; tail blackish above, dark brown
below. The skull is characterized by its shorter, wider brain
case and its wider and more abruptly spreading zygomatic
arches. In the last upper molar the anterior inner and outer
triangles are confluent and nearly opposite instead of being
separated by enamel walls. Basal length of skull, 26 mm.;
zygomatic breadth, 16.2.

Beyond the original and supplementary descriptions of this
meadow mouse by Bailey (1900, p. 26), nothing seems to be
known of it. That author remarks: "Microtus nesophilus
needs no comparison with breweri or terraenovae, the other two
insular forms from the Atlantic coast. In general appearance
it more nearly resembles pennsylvanicus, but in cranial char-
acters it is as distinctly different as either of the other island
species. During the month of August, 1898, Mr. A. H. Howell
visited Great Gull Island for the purpose of getting specimens
of Microtus, but he found their old haunts covered by the
earth moved in grading the island for fortifications, while no
trace of the animals remained. He thinks they are completely
exterminated." Fifteen specimens are listed in the collection
of the United States Fish and Wildlife Service, but apart from
these and Bailey's account there seems to be no further record
of the race. It is a good example of the precariousness of the
status of an island form, which with some unusual modification
of its habitat is exposed to new dangers,

Since Bailey's account of this race (for such it may best be
considered), the meadow-mouse of Block Island, R. I., has
been described as an insular race by Bangs (1908, p. 20) under
the name Microtus provectus. The differential characters seem
slight but perhaps sufficient to warrant this action. There is
nothing to indicate that it is in any present danger, though
conceivably changing conditions on Block Island might alter
its habitat sufficiently to imperil it. Of the supposed race M.
pennsylvanicus shattucki Howe, of Tumbledown Dick Island,

NORTH AMERICA AND THE WEST INDIES 99

Penobscot Bay, Maine, it has been shown by Wyman (1922, p.
163) that it is identical with the mainland species. Meadow
mice are excellent swimmers and may be expected to occur on
most of the small coastwise islands where conditions are
suitable and where the distances between such islands and the
nearest land are not too great to be crossed at intervals by them.

Family ECHIMYIDAE: Spiny Rats and Their Relatives

HISPANIOLAN SPINY RAT; "MOHUY"

BROTOMYS VORATUS Miller

Brotomys voratus Miller, Smithsonian Misc. Coll., vol. 66, no. 12, p. 7, 1916 ("Kitchen

midden at San Pedro de Macoris, Santo Domingo").
FIGS. : Miller, 1916b, pi., fig. 1 (four views of type rostrum) ; 1929c, pi. 2, fig. 1 (rostrum) ;

1930, pi. 1, figs. 2, 2a, 2b; pi. 2, fig. 5 (femur).

BROTOMYS CONTRACTUS Miller

Brotomys (?) contractus Miller, Smithsonian Misc. Coll., vol. 81, no. 9, p. 13, Mar.

30, 1929 ("Small cave near St. Michel, Haiti").
FIG.: Miller, 1929a, pi. 2, fig. 2 (palate).

These two species of Hispaniola, the second with peculiarly
narrowed palate, may, as Miller remarks, represent even dis-
tinct genera, but with the fragmentary material at present
known, this is still uncertain, though they evidently are related
rather closely.

Brotomys represents a spiny rat, about the size of the South
American Proechimys canicollis, remarkable for its "robust
skull with weak teeth." The antorbital opening is very large
and lacks a secondary canal at the lower corner for the passage
of the facial branch of the fifth nerve, a character readily dis-
tinguishing it from Boromys. The upper cheek teeth are four
on each side, with three short roots each. There is a deep
enamel infolding from about the middle of each side of each
cheek tooth, meeting or slightly overlapping at the center.
Probably a shallower infold is present in the crown of a fresh
tooth, which soon wears down to leave a small island of enamel
in at least the posterior half of each upper molar. Length of
four upper alveoli, 10.8 mm. The palate is slightly emarginate.
In the toothless palate representing B. contractus the very
narrow space between the molar rows is a most striking
character.

100 EXTINCT AND VANISHING MAMMALS

This spiny-rat genus is known only from Hispaniola, where
Miller (1916b, 1929a, 1929c, 1930) has recorded it as relatively
common not only in the cave deposits, doubtless made in large
part by owls, but also in kitchen-middens of the aborigines.
Thus he secured two imperfect skulls and more than 50 mandi-
bles from caves near St. Michel, in central Haiti, while in the
region about Samana Bay in northeastern Dominican Republic
it occurred in every one of the various Indian sites investigated.
He writes (1929c) : "The frequency with which the bones of this
animal occur in the Indian deposits indicates that Brotomys
must have been abundant and generally distributed in pre-
Columbian days. It was probably much like the living South
American spiny-rats in size and general form, but with heavier,
less elongated head. I have little doubt that this animal was
the Mohuy described by Oviedo as ... somewhat
smaller than the hutia [Plagiodontia], its color is paler and
likewise gray. This was the food most valued and esteemed
by the caciques and chief of this island; and the character of
the animal was much like the hutia except that the hair was
denser and coarser (or more stiff), and very pointed and
standing erect or straight above."

Nothing is known of B. contractus except the type palate.
No doubt the two were contemporaneous and were early ex-
terminated after the discovery of the island. It seems unlikely,
however, that hunting for food was the main cause of extinc-
tion. Possibly the introduced rats became a menacing factor,
or the actual extinction may have been the result of several
causes.

LARGER CUBAN SPINY RAT

BOROMYS OFFELLA Miller

Boromys o/ella Miller, Smithsonian Misc. Coll., vol. 66, no. 12, p. 8, 1916 ("Village

site at Maisi, Baracoa, Cuba").
FIG.: Allen, G. M., 1918, pi. 1, fig. 6 (mandible).

LESSER CUBAN SPINY RAT

BOROMYS TORREI G. M. Allen

Boromys torrei G. M. Allen, Bull. Mus. Comp. Zool., vol. 61, no. 1, p. 6, Jan., 1917

("Cavern in the Sierra of Hato-Nuevo, Province of Matanzas, Cuba").
FIGS.: Allen, G. M., 1917, pi., figs. 10-13; 1918, pi. 1, figs. 11-13 (skull and teeth).

NORTH AMERICA AND THE WEST INDIES 101

Remains of these two supposed spiny rats differ chiefly in
size, and both occur together in cave deposits; hence they may
be considered under a common grouping. The genus is related
to Brotomys of Hispaniola, but the skulls differ in that the
latter shows no external swelling of the bone over the posterior
termination of the root of the upper incisor, nor is there a
definitely marked channel for the passage of the fifth nerve on
the floor of the antorbital foramen; in Boromys, on the other
hand, the posterior termination of the incisor root is marked
by an obvious external swelling, and the floor of the antorbital
canal is provided with a groove for the nerve.

So far as known from parts of the skeleton, the two species
were alike in structure. There were four cheek teeth in each of
the tooth rows. Each tooth had a deep inner and outer re-
entrant of enamel extending to the center of the tooth, the
inner slightly in advance of the outer in each jaw. A smaller
reentrant of less vertical depth was present in both the anterior
and the posterior halves of the premolar and in the posterior
half only of the three molars, penetrating from the outer side.
With wear these smaller inlets are soon cut off and appear as
little circular islands of enamel in the center of their respective
sections and with still further wearing down, disappear. The
enamel wall at this stage has nearly the outline of a figure 8.
The alveoli of the molars are squarish in outline with shallow
cavities for short roots. The incisors were orange-yellow in
color. In the larger species, B. offella, the alveoli of the upper
teeth have a combined length of 11.5 mm., against 7.6 mm. in
the smaller animal, B. torrei; while the crown area of a single
molar in the former is about three times that in the latter.

In former times both these species seem to have been well
distributed not only in Cuba, but in the Isle of Pines as well.
The larger animal was first discovered in excavations of an old
village site at Maisi, Baracoa; hence it was probably used as a
food animal by the Indians. Shortly after this discovery, a
number of maxillae and mandibles were found in a cave at
Limones, Cuba, and in similar situations in the Sierra de
Casas, Isle of Pines, by Dr. Thomas Barbour and his associates.
These remains may have been brought in by the now extinct
giant barn owl. Other but "scanty material" representing the
same animal was found at about the same time by Dr. H. E.
Anthony (1919) in the course of excavations in caves at
Daiquiri, in eastern Cuba, brought in probably by owls.

102 EXTINCT AND VANISHING MAMMALS

Of the lesser Cuban spiny rat, remains seem more numerous
in the caves, perhaps because the animals were more easily
taken by the owls. The species was first found in a breccia
solidly cemented by infiltrated lime deposits in a cave in the
Sierra de Hato Nuevo, Matanzas Province. Anthony found
it abundantly represented in the cave deposits at Daiquiri,
while on the Isle of Pines additional remains were unearthed
in caves in the Sierra de Casas by Messrs. Barbour, Brooks,
and Warner and at another cave deposit in the same island by
the Messrs. Link (Peterson, 1917). Nothing further is known
of either species, which must both have become extinct within
the past century from causes not altogether clear. Neither
Gundlach nor Poey seems to have known of the animals, yet
some of the bones recovered, such as the nearly perfect skull
figured in my paper of 1918, are so fresh that they can be of no
very great age. Anthony (1919) writes concerning the bones
he collected in the Cueva de los Indios, at Daiquiri: "Some
idea of the abundance of the members of this genus is shown
by the fact that I have cleaned and examined over five hundred
mandibular rami, while nearly as many fragments have been
ignored as too badly broken up. The small torrei seems to have
been the dominant form, as only about five percent of this
series represents the larger" one. That the smaller species was
ignored as a food animal by the natives may be indicated by
the absence of its remains in kitchen middens.

Although in the general structure of the teeth the genus is
strongly reminiscent of the continental spiny rats of South
America, it has relatively "broader nasals. Miller points out a
close resemblance of the enamel pattern to that of the extinct
Stichomys of the Santa Cruz formation of the South American
Miocene. The fact that the related Brotomys of Hispaniola is
nevertheless unknown elsewhere, and the lack of similar forms
on the mainland, point to long isolation. Miller regards the
two genera as relatives of the Puerto Rican Heteropsomys,
which, if this surmise proves to be substantiated, would imply
some common origin of the three types now regarded as
generically distinct. Anthony, however, believes that Hetero-
psomys is a near ally of the agoutis.

True (1885) records a large spiny rat taken on Martinique
by Ober in 1878. It was sent to the United States National
Museum and there identified as "Loncheres" (= Echimys)

NORTH AMERICA AND THE WEST INDIES 103

armatus of South America, whence it may have been intro-
duced. There seems to be no recent evidence of its continued
presence on the island, although Austin H. Clark in 1904 was
told by the natives that it still occurred.

BLACK-TAILED HUTIA; "ANDARAZ"
CAPROMYS MELANURUS Poey

Caprornys melanura Poey, Monatsber. preuss. Akad. Wiss. Berlin, 1864, p. 384

(Manzanillo, southeastern Cuba).
FIGS.: Dobson, 1884, pis. 18-21 (colored figure of exterior; anatomy); Bucher, 1937,

pis. 10-12 (photographs of captives).

So far as known, the black-tailed hutia is confined to ex-
treme eastern Cuba in the wooded mountains. Its circum-
scribed range places it in a somewhat dangerous position, and
although it may continue to exist there for years to come,
changes incident to human influence, the clearing of forest,
the introduction of the mongoose, and hunting for food are
likely to reduce its numbers more and more.

This is a slightly smaller animal than the common hutia of
Cuba, Capromys pilorides, less clumsily built, and of more
arboreal habits. With a body slightly larger than that of a
muskrat, and with a bushy tail about as long as the body
without the head, the general color is a mixed blackish or
brownish and buffy, due to the presence of scattered all-black
hairs and more abundant hairs having smoky-brown bases
and pale-buff y or ochraceous tips. The forehead and cheeks
are nearly clear dark brown or clearer brown, as are the hands
and the central area of the hind feet. The under side is similar
but paler, mixed grayish and brown, clearer brown on the
throat. There is a varying amount of clear white, which may
include only the axillae and inside of the thighs, or, as in one
specimen examined, the white may include the entire upper
throat, the under side of the arms, and continue over the
entire mid ventral surface to the inner side of the thighs, the
ankles, and the outer margins of the hind feet. The tail is
thickly haired, tapering, with abundant somewhat-stiff black-
ish to maroon hairs from one-half to three-quarters of an inch
long. Total length about 26 inches, of which the tail consti-
tutes about 11 or 12 inches. The claws are slender and sharply
curved for climbing, and the soles are granular for better
holding.

104 EXTINCT AND VANISHING MAMMALS

Although somewhat resembling Capromys prehensilis, the
longer-tailed and prehensile-tailed species of western Cuba,
this seems to be a quite distinct species. Although its tail is
prehensile, the hair is not worn away through use of the tail in
holding as it is in the former. Moreover, as is well known to
those accustomed to hunt it, the tail readily breaks away near
the base if strain is placed upon it. For this reason the hunter
does not attempt to drag the animal out from a hole by the
tail. As with spiny rats, the point of weakness is close to the
base of the tail and is marked externally by a sudden change in
the character of the hair from that of the body to the more
shaggy type of the rest of that member. It comes away easily
without bleeding, leaving a rosette of muscle fibers. The de-
tails of the anatomy and external appearance are illustrated in
the paper by Dobson (1884).

First clearly described by Poey in 1864, the species continues
to be little known and rare in collections. Those in the Museum
of Comparative Zoology were secured in the foothills of the
Sierra Maestra of extreme southeastern Cuba in 1913, and again
others were obtained in the mountains north of Imias in 1936.
This species replaces C. prehensilis in the eastern tip of Cuba
and is closely related to it. The only good account of its habits
is the recent one of Bucher (1937), who kept several of this and
the other larger Cuban species in captivity and found them
interesting pets. Of C. melanurus he writes that they are
active and feed late in the evening and at night. In a wild
state they apparently live in pairs together, and the females
produce one or two young at a birth. The males fight one
another for possession of a female but otherwise seem good-
natured. Unlike C. pilorides, they do not enjoy being petted
and can not be trusted not to snap at one's finger at times.
They chatter their teeth when approached and at mating time
use a short, low birdlike whistle. In captivity the gestation
period seems to be about 17 or 18 weeks. In actions they are
said by Bucher to be very squirrellike, running along branches
or leaping from tree to tree, using the tail as a balance or in a
prehensile manner. Bucher fed his C. melanurus at first on
fruit of "yagruma" (Cecropia), and later they became adapted
to a diet of sweetpotato vine, prunings of guava, any citrus
fruit and grapevine leaves, buds, and bark. Unlike C. pilorides,
they will not eat bread or cracker. Probably their arboreal

NORTH AMERICA AND THE WEST INDIES 105

life and nocturnal habits will help to keep them from exter-
mination for a considerable time to come.

So far as known the genus Capromys is confined to Cuba and
the Isle of Pines. No fossil species are to be found on any of
the other West Indian Islands, which seems a significant fact
in the distribution. The two other large species, C. prehensilis
and C. pilorides, are represented by only slightly differing
races on the Isle of Pines, relictus and gundlachi, respectively.
Bucher (1937) has brought out some interesting details in
which their habits differ. Since, however, neither species
seems in immediate danger of extermination, they may be
omitted here.

DWARF CUBAN HUTIA

CAPROMYS NANA G. M. Allen

Capromys nana G. M. Allen, Proc. New England Zool. Club, vol. 6, p. 54, 1917 ("Cave

deposit in the Sierra de Hato Nuevo, Province of Matanzas, Cuba").
FIGS. : Allen, G. M., 1918, pi., figs. 1-5; Morrison-Scott, 1939, pis. 5-7 (skull and teeth).

This small Capromys was first made known from jaws found
in recent cave deposits at two localities in Cuba. Subsequently
it was found living in the great Zapata Swamp, on the southern
coast of middle-western Cuba, south of the original locality.

About the size of a Norway rat, this is typical Capromys,
with a rather coarse pelage of all-black hairs mixed with others
having a buffy or ochraceous tip, except that on the muzzle
and in front of the ear the particolored hairs have whitish tips
instead. The fingers and sides of the hind feet are whitish,
but the latter have the central area darkened. The tail,
which is slightly shorter than the combined length of head and
body, has the long hair of the back continued to its basal
three-quarters of an inch, beyond which the hairs become
suddenly shorter and somewhat sparse, so that the scaling of
the tail shows through. The upper side of the tail and the
terminal inch of its under side are black; the lower side is
sharply marked off and ochraceous-buff. Total length, about
375 mm.; tail, 176; hind foot, 45.

When first described this dwarf hutia was supposedly
extinct. It was described on the basis of ten small jaws from
cave deposits probably made by barn owls, in the Sierra de
Hato Nuevo and near Limones in western-central Cuba.
Soon after, Dr. H. E. Anthony (1919) procured no less than

106 EXTINCT AND VANISHING MAMMALS

300 mandibles and 60 cranial fragments from caves at Daiquiri,
on the eastern tip of Cuba, indicating that in former times it
inhabited suitable areas in the length and breadth of the island
and probably was regularly preyed upon by barn owls, which
had brought these remains into the caverns where the birds
rested by day. None of these fragments appeared to be "at all
recent and since this mammal formed such a large part of the
owl diet in times gone by, it would almost certainly appear in
recent owl pellets" were it still living in eastern Cuba. In the
meantime, through the efforts of Dr. Thomas Barbour, living
specimens had been secured in 1917 and sent to the Museum of
Comparative Zoology by Don Jose Garcia. These were ob-
tained in the great Zapata Swamp, on the southern coast of
west-central Cuba, an area that is ordinarily nearly impene-
trable but can be entered at times of unusual drought. So
far as known, no other examples have since been obtained,
until November 1937, when Dr. Hans Boker collected a live
pair in the same swamp, south of the town of Jaguey. As pre-
viously stated by Senor Garcia, these had been found living
on small, dry, bush-covered islands in the swamp, where they
are probably at most times fairly well protected against invad-
ing mongooses or other enemies. Senor Garcia told Dr.
Barbour that the dwarf hutias are sometimes routed out when
the sawgrass of these marshes is burned. Morrison -Scott
(1939) records that the pair secured by Dr. Boker were
brought alive to the Berlin zoo, where on December 1, 1937,
the female gave birth to a single young one, but it as well as
its mother died within the next two weeks. These two are
now preserved in the British Museum, while the adult male,
which survived until the following February, is now in the
Berlin Museum. The dentition of the young one, which
Morrison-Scott figures, already consisted of a premolar and
the first molar.

While it seems likely that the species may survive for a
good many years in the Zapata refuge, it must elsewhere in
Cuba have succumbed before the introduced mongoose and
probably the roof rats.

NORTH AMERICA AND THE WEST INDIES 107

JAMAICAN HUTIA; "INDIAN CONY"; BROWNE'S HUTIA

GEOCAPROMYS BROWNII (J. B. Fischer)
Capromys brownii J. B. Fischer, Synopsis Mamm., Addenda, p. 389 [= 589], 1830

("In Jamaica").
SYNONYM: Capromys brachyurus Hill, in Gosse and Hill's Naturalist in Jamaica, p.

471, 1857.
FIG.: Anthony, H. E., 1920a, p. 163 (photograph).

This is the first of the short-tailed hutias to be described,
based by Fischer on the account of the "Large brown Indian
coney" in Patrick Browne's "Civil and natural history of
Jamaica," 1789. Originally placed in the same genus as the
long-tailed hutias of Cuba, it was later included in a separate
subgenus, Geocapromys Chapman, now regarded as a group of
generic rank. The characters of the genus in comparison with
Capromys have been defined by Miller (1929b), of which the
more important are: First lower cheek tooth with a small
additional reentrant of the enamel wall on the inner side of the
first space, making three instead of two indentations; upper
tooth rows convergent anteriorly, so that the bases of the two
premolars of opposite sides are in contact instead of widely
separated; root capsule of the upper incisors passing slightly
outside of and beyond the premolar instead of stopping in
contact with its base; bony bar in front of the eye nearly
vertical or sloping backward instead of forward. Externally,
it is distinguished by the very short tail.

This hutia is about the size of a cottontail rabbit but more
stoutly built, with short legs and a stumpy tail. Above, the
color is a general dark brown, resulting from abundant parti-
colored hairs with dark-gray bases and ochraceous tips, mixed
with all-black hairs, which are most abundant in the middle of
the back. The lower surface is paler, with a buffy-gray throat
and lower belly, while the chest and upper part of the belly are
more thickly clothed with ochraceous-tipped hairs. The tail
and backs of the feet are nearly clear brown. The ears are very
short, rounded, and clothed with minute hairs. Total length,
425 mm.; tail, 50; hind foot, 60.

This was doubtless a common species on Jamaica in former
times and was confined to that island. It formed an important
article of food for the aborigines as attested by the fact that its
bones were commonly found with ashes of their camp sites.
It is said that the name Indian cony is from this association.

108 EXTINCT AND VANISHING MAMMALS

Though evidently now much restricted in habitat and abun-
dance, through hunting and probably through ravages of the
Burmese mongoose, introduced in 1872, it nevertheless still
occurs in some numbers in the John Crow Mountains at the
eastern end of the island where Dr. H. E. Anthony found it in
1919. All his efforts to discover a trace of it in the Blue
Mountains, however, were unsuccessful. In the former region
the native Negro population often hunt them with small dogs,
specially trained for the purpose. These dogs find the dens
of the hutias by scent, under large tree roots, or among loose
rocks. They at once give tongue to their excitement and try
to enlarge the opening and force their way in a few feet to the
chamber where the quarry lives by day. Dr. Anthony (1920a)
recounts a day's hunt of this nature in the course of which a
few specimens were secured. When the dog has penetrated
to the den, the hutia voices its alarm in "faint birdlike chirp-
ings" but is soon dragged out by the dog and at once seized by
the hunters lest the dogs devour it. "The cony possesses
sufficient vitality to withstand a great deal of mauling, and
can put up a good fight." One of them bit a small piece out of
a dog's nose. "Not infrequently several animals are taken
from the one hole," perhaps, as with other hutias, representing
an adult pair and their grown young. In his day's hunt,
however, Dr. Anthony found but a single animal in each den.

Jamaican hutia (Geocapromys brownii)

NORTH AMERICA AND THE WEST INDIES 109

It is said to live on grasses and the leaves of low shrubs,
whence conies its local name of "grazee" in use among the
Negroes. It would be worth while to restrict, if possible, the
hunting of the species in its mountain habitat in order to
protect it from extinction as long as may be. Since 1922 it has
been accorded full legal protection (Journ. Soc. Pres. Fauna
Empire, pt. 2, p. 15, 1922).

Whether any other species of this genus or of the allied
Capromys ever occurred on Jamaica is uncertain. However,
Miller (1916a) reports that, in examining a few bones recovered
from ancient burial sites near Salt River, jaws of two sizes of
hutia were found. Although without teeth, the larger of these
seemed identical with the living Geocapromys brownii, while
the smaller, with obviously reduced last molar, was indis-
tinguishable from the mandible of Geocapromys thoracatus of
Swan Island.

SWAN ISLAND HUTIA

GEOCAPROMYS THORACATUS (True)

Capromys brachyurus thoracatus True, Proc. U. S. Nat. Mus., vol. 11, p. 469, 1888
(" Little Swan Island . . . entrance of the Gulf of Honduras").

This is a slightly smaller and grayer species than the Ja-
maican animal and was first discovered by Dr. C. H. Townsend,
when as naturalist on the U. S. Fish Commission steamer
Albatross, more than 50 years ago he secured specimens on
Little Swan Island, which lies about 110 miles off northeastern
Honduras, and at about a quarter the distance separating that
country from Jamaica.

While resembling in general appearance the Geocapromys
brownii of Jamaica, to which it is probably closest allied of the
known species, this is somewhat smaller and in correlation
with its drier island habitat is less darkened. The general
color of the entire upper side is a very even mixture of blackish
and pale ochraceous, resulting in a rather yellowish-gray
appearance. The backs of the hands and feet are a nearly
uniform dusky, minutely punctate with ochraceous. On the
under side, the most conspicuous feature is a dull whitish half-
collar extending across the lower throat between the arm pits.
Elsewhere the color is like that of the upper side but with very
little darkening, producing a smoky -buff shade. Size less than
in the Jamaican animal, with a total length of some 15.5

110 EXTINCT AND VANISHING MAMMALS

inches (390 mm.) of which the tail is about 2.25 (57 mm.);
hind foot, 2.50 (65 mm.). In both these species, the auditory
bullae do not extend below the level of the basioccipital bones.

There is very little on record as to the habits of this hutia.
Lowe (1911), who visited the Swan Islands in 1908, gives an
excellent picture of the conditions there in his book "A Nat-
uralist on Desert Islands." The two islands comprise a larger
one, which is inhabited and serves as a signal station, and a
smaller lying about a quarter of a mile to the eastward, which
is untenanted except by wildlife and seldom visited on account
of the difficulty of making a landing. The latter island is the
home of many sea birds, which nest on the tops of the scrubby
bush growth or among its rocks, and is the only place where
this hutia is found. It is about a mile and a quarter long and
rises abruptly from the sea with low cliffs on all but a part of
its northwesterly side. "The plateau formed by the top of the
island is densely overgrown with trees and bushes, which form
a closely matted dome of vegetation above its wind-swept
surface." On the occasion of Lowe's visit it was entirely un-
disturbed and could be reached from Big Swan Island only on
very calm days. It appears to be a "big 'stack' of solid coral
limestone which has been thrust bodily up from the sea bot-
tom." In the central portion the surface is somewhat more
even, with vegetable mould and a dense canopy of trees. Here
Lowe and his party captured several of the hutias. They are
diurnal in habits and were seen running about or "bolting into
the big crevasses with which the island is seamed." One
individual ran over some rocks and tried to escape under the
spreading roots of a large undermined tree but allowed itself
to be carefully drawn out and placed alive in a fishing creel.
It was surprisingly mild-mannered and showed not the slightest
fear or suspicion, when later it was taken alive to England and
brought face to face with King Edward's favorite dog, which
it quietly inspected. The two taken captive became very tame
and were allowed the freedom of the yacht's deck, "drank
milk with avidity, and ate any form of vegetable or fruit that
was offered them." In 1912 George Nelson visited Little
Swan Island and secured specimens for the Museum of Com-
parative Zoology.

At the present time the species seems safe enough in the
narrow limits of its island home, but should any important

NORTH AMERICA AND THE WEST INDIES 111

changes take place there, such as clearing, or the introduction
of goats or mongooses, its future would at once be in danger.
How it originally reached this isolated little spot, whether
through introduction by aborigines or by rafting at some pre-
historic time from elsewhere, is still an unanswered question.

CUBAN SHORT-TAILED HUTIA
GEOCAPROMYS COLUMBIANUS (Chapman)

Capromys columbianus Chapman, Bull. Amer. Mus. Nat. Hist., vol. 4, p. 314, 1892

("Cave near Trinidad," Cuba).
SYNONYMS: Synodontomys columbianus G. M. Allen, Bull. Mus. Comp. Zool., vol. 61,

p. 5, 1917; Geocapromys cubanus G. M. Allen, ibid., p. 9.
FIGS.: Chapman, 1892, fig. 3 (type, maxillary bone, and tooth); Allen, G. M., 1917,

pi., figs. 7-9 (teeth).

This short-tailed hutia is not known from living specimens,
yet it must have been common in Cuba up till relatively
recent times. It was first made known by Dr. Frank M. Chap-
man from a half-palate lacking the teeth and from a few other
fragments discovered in a cave near Trinidad, Cuba, "asso-
ciated with the remains of birds^and bats, and also fragmentary
pieces of the bones of Capromys." " The cave is situated in the
southern slope of the coral limestone coast range at an altitude
of about seven hundred feet, and within two hundred feet of
the summit of this part of the range . . . The floor of
the cave was covered to the depth of several feet with a red,
ferruginous earth, and on this was a layer four or five inches in
depth of a dark earth in which the bones" were found.

On account of the fragmentary nature of the original speci-
mens and of other later-discovered remains, satisfactory com-
parisons of this Cuban hutia with other forms of the genus are
yet to be made. It would be important to know the relations
of the audital bullae to the level of the basioccipital bone,
which might indicate a relationship with the Bahaman Geo-
capromys on the one hand or to the Jamaican and Swan
Island species on the other. Dr. Chapman notices the strongly
converging upper tooth rows, which, however, as later finds
showed, do not actually touch in front on the palatal aspect.
The Cuban animal was evidently as large as the Jamaican G.
brownii. A median bony ridge on the roof of the palate ends
abruptly at the palatal margin in the latter and in G. thoracatus
is continued as a short median projection at that point. In

112 EXTINCT AND VANISHING MAMMALS

both G. columbianus and G. ingrahami, however, the ridge
fades out before reaching the palatal margin, while the opening
of the posterior nares seems more narrowed. In the former,
too, the enamel folds are more nearly at right angles to the
long axis of the tooth than in Capromys. Toothless jaws may
often be identified by the indication of the extra small reentrant
on the inner side of the tip of the anteriormost cheek tooth.
It is obvious that until better-preserved skulls of this species
are found, its distinctive characters in comparison with other
forms of the genus are not very well ascertained.

This was evidently a common and well-distributed species
not only throughout Cuba in former days but also on the
adjacent Isle of Pines. On Cuba, in addition to the original
locality, Dr. Thomas Barbour secured numerous remains from
caves in Matanzas Province, as well as from the Isle of Pines,
and Dr. H. E. Anthony obtained an abundance of jaws and
palates from owl deposits in a cave at Daiquiri, in eastern
Cuba. In the latter instance the fragments seemed to be
mostly of immature and hence small animals, indicating that
the adults were too large to be managed by the barn owls that
evidently were the cause of the deposits. Why a species ap-
parently so common should have died out before any modern
specimens were obtained, and in so well populated an island
as Cuba, is not at all clear, for apparently they disappeared
before mongooses were introduced. Possibly the larger size
and arboreal habits of Capromys were factors in their survival,
whereas Geocapromys was ground-living and less able to escape.

BAHAMAN HUTIA
GEOCAPROMYS INGRAHAMI (J. A. Allen)

Capromys ingrahami J. A. Allen, Bull. Amer. Mus. Nat. Hist., vol. 3, p. 329, Dec.,

1891 ("Plana [or Frenchman's] Keys, Bahamas").
FIGS.: Allen, J. A., 1891, p. 335, figs. 1, 3, 5-9 (skull and teeth).

Except for the Jamaican and Swan Island hutias, no living
member of this short-tailed genus is known other than this
small species now confined to the Plana Keys between the
island of Mariguana and Acklins Island in the southern part
of the Bahama Archipelago.

The general color is a mixed "yellowish brown, gray and
black, giving the general effect of grayish brown .

NORTH AMERICA AND THE WEST INDIES 113

Below, nearly uniform pale yellowish brown, barely a little
more yellowish posteriorly . . . some are quite strongly
suffused with yellowish and have less black; others are faintly
suffused, giving the general effect above of pale yellowish gray
mixed with black." A notable feature of the skull is the great
development of the audital bullae, which in this and the other
Bahaman races project conspicuously beyond the level of the
basioccipital bone instead of ending at the same level with
this bone. In the lower molars, the two reentrant loops of
enamel from the inner side are relatively short and straight,
barely reaching the center of the tooth. In size this is slightly
smaller than the other living species.

The discovery of this hutia still living in one of the Plana
Keys is due to D. P. Ingraham, who in mid-February, 1891,
landed there on several occasions while weatherbound under
the lee of the island. This key he describes as a small rocky
islet, the highest point probably not more than 50 feet above
sea level, "with crevices and caves worn in the rocks by the
action of water, and in many places broken strata of rocks,
piled upon each other." It is about half a mile wide and 4 or 5
miles long, quite without permanent fresh water. In the course
of two weeks he secured a small series of specimens for the
American Museum of Natural History. In none of these were
fetuses found. There appeared to be a concentration of
individuals at places where conditions were favorable in afford-
ing many hiding places. He once saw a number of them to-
gether at a distance from the rocks among palmettos, but later
concluded that this was a result of the beginning of the rutting
season. They fed mostly at night, though occasionally foraging
by day, were not shy, and with caution could be approached
within 25 or 30 feet before dashing to the shelter of the rocks.
They fed on the leaves, twigs, and bark of the bushes, "espe-
cially the black button wood, and the succulent growth of the
cactus plants," which doubtless supplied the necessary mois-
ture. They seemed " very fond of the fruit of the paw-paw, and
even of the body of the tree itself," for in some instances
trunks were found nearly as large as a man's body "so nearly
eaten off that they would not sustain their own weight. A
sweet potato left on the shore was eaten up, while the body of a
bird, left to tempt them, was untouched."

So far as the writer knows, no other naturalist visited East

1 14 EXTINCT AND VANISHING MAMMALS

Plana Key after 1891 until James C. Green way, Jr., landed
there in February 1933. He found the hutias still present in
some numbers and brought back a few specimens for the
Museum of Comparative Zoology. They frequented especially
small clumps of bushes and would dash out if disturbed, seek-
ing the shelter of an adjacent clump.

It seems evident that at some former time these or similar
hutias were more widely distributed in the Bahama group.
Catesby mentions a Cuniculus bahamensis and presents a
figure of what may have been one of the Cuban species of
Capromys, though no locality is given. According to J. A.
Allen (1891), the narrators of Columbus's voyage "make
frequent mention of their abundance not only in the Bahamas
and at Jamaica, but also in Cuba and Hispaniola." In recent
years remains of this genus have been brought to light in the
investigation of caves and aboriginal sites in some of the
Bahamas, and these prove to be allied but slightly distinct
forms, a brief account of which may be included here.

CROOKED ISLAND HUTIA
GEOCAPROMYS INGRAHAMI IRRECTUS Lawrence

Geocapromys ingrahami irrectus Lawrence, Occas. Papers Boston Soc. Nat. Hist., vol.

8, p. 190, Nov. 7, 1934 ("Gordon Hill Caves, Crooked Island, Bahamas").
FIG.: Lawrence, 1934, fig. 2 (rostrum).

GREAT ABACO HUTIA
GEOCAPROMYS INGRAHAMI ABACONIS Lawrence

Geocapromys ingrahami abaconis Lawrence, op. cit., p. 191, Nov. 7, 1934 ("Imperial

Lighthouse Caves, Hole in the Wall, Great Abaco Island, Bahamas").
FIG.: Lawrence, 1934, fig. 3 (rostrum).

These two forms are very nearly related and may be treated
together. Both are known from skeletal remains, including
rostra and some other parts of the skull, secured in the course
of archeological investigations in the Bahamas in 1933-34, by
Froelich Rainey. The points of distinction thus far made out
lie mainly in the details of the rostral bones and in the teeth.
G. ingrahami irrectus seems to have been slightly larger than
the Plana Keys animal, still living, with longer tooth row,
wider individual teeth, and longer zygomatic arch. The two
enamel loops on the inner side of the three lower molars are

NORTH AMERICA AND THE WEST INDIES 115

much longer than in the latter animal, extending as long finger-
like folds nearly to the enamel wall of the opposite side of the
tooth instead of about to the center of the tooth. The audit al
bullae, however, were large, as in typical G. ingrahami, ex-
tending well below the level of the basioccipital. The lower
tooth row attained a maximum length of up to 17.5 mm.
(alveoli); upper tooth row, 16.5; humerus, 50; femur, 67; tibia,
64. Remains of this form are known from four of the central
islands of the Bahama group: Crooked Island, Eleuthera, Long
Island, and Exuma. Those from the first three islands men-
tioned were in more or less close association with aboriginal
sites in caves and hence may have been contemporaneous
with early Indian occupation. Those from Exuma Island,
however, though undoubtedly of the same form, may have
been earlier, for the deposit in which they occurred abundantly
(see G. M. Allen, 1937) showed no sign of human influence,
and furthermore contained remains of three genera of extinct
hawks and owls. To what extent the aborigines may have
utilized these animals as food, or transported them from one
island to another, is not now possible to make out.

The Abaco hutia, described' from palates and other cranial
fragments, was closely similar to that on the Crooked Island
group, but with the anterior lower premolar "slightly more
tapering anteriorly with the anterior lobe somewhat smaller
and more slender" and with longer frontal bones, which ended
in a pointed projection "reaching anteriorly between the nasals
and the premaxillaries." The postorbital processes in available
specimens are more reduced than in the two other forms of the
species.

Nothing further is known about these Bahaman hutias.
Possibly they were still living in Catesby's day, but the species
he figures as a Bahaman cony, about 1750, appears to be a
Cuban Capromys. Their extinction can not be wholly due to
the extinct eagle and giant barn owl, which must have preyed
upon them, but more likely was a result of clearing and burning
the cover during the earlier centuries of white occupation, of
which unfortunately there is practically no record.

HEXOLOBODON PHENAX Miller
Hexolobodon phenax Miller, Smithsonian Misc. Coll., vol. 81, no. 9, p. 19, Mar. 30,

1929 ("Small cave near St. Michel," central plain of Haiti).
FIGS.: Miller, 1929a, pi. 3, figs. 1, la, Ib (palate and jaw).

116 EXTINCT AND VANISHING MAMMALS

This is an interesting genus, closely related to Geocapromys
and apparently taking its place on Hispaniola, where it was
contemporary with other species of the recent cave fauna, such
as Isolobodon, Aphaetreus, and the still-living Plagiodontia.
At first glance the tooth pattern in the lower jaw is very
similar to that in Geocapromys ingrahami irrectus in having the
long enamel reentrants of the two last molars extend well be-
yond the center of the tooth. The anteriormost tooth, the
premolar, however, differs from that of Geocapromys and
agrees with Capromys in lacking the additional small inner
reentrant to the first section. In the palate the convergence
of the tooth rows anteriorly is similar to their appearance in
Geocapromys, and the relation of the incisor roots to the first
premolar is likewise the same, but, as Miller points out, the
root of the premolar is "thrown conspicuously forward away
from that of the first molar." Other characters are the less
specialized roots of the cheek teeth, which appear to close with
age instead of continually growing, and "the extension of the
lower incisor root to the outer side of the mandibular tooth
row." Size probably about as in the Cuban Capromys pilorides
but perhaps with a shorter rostrum. Maxillary tooth row
(alveoli), 22 mm.

This animal is known from a palate, six mandibles, and four
isolated cheek teeth, collected by Miller from caves in central
Haiti near St. Michel and 1'Atalaye. At first sight it might
easily be taken for a Geocapromys, which is as yet unknown
from Hispaniola. The characters pointed out by Miller,
however, appear to be distinctive. It was presumably ex-
terminated soon after white occupation.

HISPANIOLAN HUTIA; "RAT-CAYES"

PLAGIODONTIA AEDIUM F. Cuvier

Plagiodontia aedium F. Cuvier, Ann. Sci. Nat., zool., ser. 2, vol. 6, p. 347, 1836 ("Saint-

Domingue," taken as Haiti by Miller).
FIGS.: Cuvier, 1836, pi. 17 (exterior and skull); Miller, 1927, pi. 1, fig. 2.

PLAGIODONTIA HYLAEUM Miller

Plagiodontia hylaeum Miller, Proc. U. S. Nat. Mus., vol. 72, art. 16, p. 4, 1927 ("Gua-

rabo, 10 miles east of Jovero, Samana Province, Dominican Republic").
FIGS.: Miller, 1927, pi. 1, figs. 1, la-lc (skull).

NORTH AMERICA AND THE WEST INDIES 117

Described and figured by Cuvier over a century ago, from a
specimen brought by Alexander Ricord from Hispaniola in
about 1826, no additional specimens of the genus had since
been found living until Dr. W. L. Abbott secured a series of
ten adults and three young at Guarabo, Dominican Republic,
about 10 miles east of Jovero, Samana Bay. These he sent to
the IT. S. National Museum, where Miller (1927), on subse-
quent study, decided that they represented a species distinct
from aedium and proposed to call it P. hylaeum.

Both these species were of similar appearance in life. Cuvier
describes P. aedium as of rather stout build, the head and body
about a foot in length, the tail 5 inches and naked. In a general
way these animals resemble a medium-sized Capromys pilorides,
but the teeth in both species are very different, with the enamel
folds extending diagonally instead of transversely across each
molar. In this respect also they recall the teeth of Isolobodon,
the Puerto Rican hutia, but the enamel reentrants are much
more nearly longitudinal in direction, and the upper tooth rows
are more nearly parallel instead of converging strongly in
front. There were five claws qn all four feet, although that of
the thumb was a short, blunt nail. The pelage in P. aedium
is described as thick, of fine silky hairs above, which were gray
in the basal three-quarters, tipped with tawny and intermixed
with longer all-black hairs. The under surfaces were paler
and clearer, probably nearly buff. The mustachial hairs, as
usual in this group, were long and abundant. Miller, who ex-
amined the type in Paris, adds that the tail was naked and
smooth, its small scales not overlapping but tending to be
rounded-pentagonal and scarcely a millimeter in diameter at
30 mm. from the base of the tail. "Ears naked internally,
thickly furred along edge and apparently on outer side also
. . . hind foot with claws, 74 mm."

The distinguishing features of P. hylaeum are given by
Miller (1927, p. 4) as follows: Anteroposterior diameter of the
first and second true molars in the lower jaw, each less than the
transverse diameter instead of equaling or exceeding it, as in
P. aedium; reentrant enamel folds relatively narrower and
longer; and the size smaller with the length of the mandibular
tooth row (alveoli) in adults less than 21 mm., instead of about
24 mm. as in P. aedium. The color throughout is "nearly
wood-brown" darkening to buffy brown on chest and belly.

118 EXTINCT AND VANISHING MAMMALS

On the under parts the light brown is uniform, but, above, it
is "finely intermingled with dark brown." It is thus probably
a somewhat darker-colored animal than P. aedium, the original
specimen of which is, however, much faded at the present time.
The largest specimen of Dr. Abbott's series measured 405 mm.
in length of head and body; the tail 130; foot, 74.

It seems likely that at the present time P. aedium, the older-
known species of Hispaniola, is quite gone and that the smaller
P. hylaeum is now confined to forested areas in the Samana Bay
region of northeastern Dominican Republic. Dr. W. L.
Abbott, who discovered it here in 1923, wrote that his speci-
mens were obtained for him by an old man who "caught them
with dogs in hollow trees down near a lagoon near sea shore."
The four adult females were all pregnant, each containing, in
early December, a single young, indicating a slow rate of
breeding, which he observes, "will probably help their extinc-
tion." He supposes that they may still have been abundant
in the region and at that time were little hunted for food. They
can climb to some extent, but their short tail and general
stout appearance indicate that terrestrial habits are more
typical. For this reason they are the more liable to attack by
the introduced Burmese mongoose. Since Dr. Abbott's visit
in 1923, William J. Clench, of the Museum of Comparative
Zoology, secured two other specimens for that institution, in
the same general region, in August 1937. These were an
adult male and a half-grown animal of the same sex, doubtless
born earlier in the same year.

Nothing is known of the habits and haunts of P. aedium
beyond the few statements of Cuvier, to whom Ricord reported
that in "Santo Domingo" the animal was known a century
ago as " Rat-Cay es," or house rat, because it frequented settled
areas. It was active only by night, hiding during the day.
The male and female of a pair remained usually together. Its
principal food consisted of roots and fruits; hence its flesh
was good to eat, and the Haitians hunted it with such zeal
that even at that time it had become very rare. From the fact
that Miller unearthed six lower jaws from a cave in the interior
of Haiti in 1925 and de Booy had previously secured a jaw
from old kitchen middens at San Pedro de Macoris on the
south coast of the island about 60 kilometers east of San
Domingo City, it may have been that P. aedium was the

NORTH AMERICA AND THE WEST INDIES 119

animal of western and southern Hispaniola, and P. hylaeum
that of the northeastern parts of the island. Even so, the
ranges could hardly have been entirely separate, for in a later
paper, Miller (1930), reporting on mammal remains from
Indian kitchen middens on the valley floor near Constanza in
the mountainous interior of Dominican Republic, identifies
both species as present. Evidently both were used as food by
the aborigines.

It is significant that this well-distinguished genus is as yet
unknown in either fossil or recent state from any other Antillean
islands or from the mainland.

LEAST HISPANIOLAN HUTIA
PLAGIODONTIA SPELAETJM Miller

Plagiodontia spelaeum Miller, Smithsonian Misc. Coll., vol. 81, no. 9, p. 18, Mar. 10,
1929 ("The crooked cave near the Atalaye plantation, Haiti").

Nothing is known of this supposed third species of Hispanio-
lan Plagiodontia beyond the account given by Miller, who
found 15 mandibles, four in the group of caves near St. Michel
in north-central Haiti, and the>others in the "crooked cave"
near the Atalaye plantation in the same region. Its main
point of difference from P. aedium, remains of which occurred
in the same caves, lies in its smaller size, as indicated by the
mandibular tooth row of 15.6-16.2 mm., as against 23.6 in the
former. Compared with P. hylaeum, the discrepancy is less,
for the latter has a tooth row of 18.6 mm.

PUERTO RICAN ISOLOBODON

ISOLOBODON PORTORICPJNSIS J. A. Alleil

Isolobodon portoricensis J. A. Allen, Ann. New York Acad. Sci., vol. 27, p. 19, 1916

("Cueva de la Ceiba, near Utuado, Puerto Rico").
FIGS.: Allen, J. A., 1916a, pis. 1-5; Anthony, H. E., 1918, pi. 62, figs. la-6b; pi. 63,

figs, la-b, 4-12c; pi. 74, figs. 6a-c; text-fig. 39, A-H (skeletal parts).

This well-marked genus is as yet unknown in the living
state and is undoubtedly extinct, although abundantly repre-
sented in kitchen middens of the Indians in Puerto Rico,
the Dominican Republic, and the Virgin Islands.

The skeletal characters, which alone are available, show that
it was an animal of about the size of the Hispaniolan hutias, to
which it is somewhat closely related. Its teeth, however, are

120 EXTINCT AND VANISHING MAMMALS

very different. The upper tooth rows converge strongly
forward instead of being nearly parallel, while the pattern of
the enamel folding is simpler. Each of the upper molars has
a shallow outer and a much deeper inner reentrant loop, the
latter of which extends inward and forward, diagonally, to
touch the smaller outer one. The first tooth of the four, the
premolar, has in addition a second small outer reentrant in
advance of the one that meets the inner loop. In the lower
tooth row there are two shallow reentrants of enamel on the
inner side of each molar, the posterior of which meets the
single outer reentrant nearly at the center of the tooth. In
Plagiodontia the inner reentrant is much deeper, runs more
diagonally forward, and does not meet the smaller fold of the
outer side. The teeth of young animals have the same pattern
as do those of adults, for they are continuously growing; but
the size of the individual teeth increases with age and wear.

This must have been a common species on Puerto Rico up
till the period of white occupation. Both Dr. H. E. Anthony
and I found its bones on the surface in undisturbed caves on
the island, and they are abundant in kitchen middens, mingled
with broken pottery and ashes, indicating that they formed a
staple article of diet for the Indian aborigines. Shortly after
the discovery of its bones in cave deposits probably of human
origin in Puerto Rico, Miller (1916b) announced the finding of
additional jaws and other parts of the skeleton in the course of
excavating Indian sites at Macoris and San Lorenzo, in the
Dominican Republic. He was unable, however, to distinguish
these remains from those found on Puerto Rico. Again, in a
later paper he (Miller, 1918a), reports the discovery of many
other bones of this hutia, secured by Theodoor de Booy in the
course of excavating two Indian sites in the Virgin Islands, one
at Magens Bay, St. Thomas, the other at Salt River, St. Croix.
The wide distribution of the species Puerto Rico, Hispaniola,
and the Virgin Islands the complete lack of differentiation on
these three areas, and finally the frequent occurrence of the
remains in deposits of human origin, all point to the fact that
it was a regular source of food for the aborigines and that they
carried it about by canoe from island to island, to serve their
needs. Whether its original home was on Puerto Rico or on
Hispaniola must remain uncertain, but Anthony (1918) in-
clines to the view that judged from its apparently wide distribu-

NORTH AMERICA AND THE WEST INDIES 121

tion on the former it was indigenous there and that "it is un-
likely that its primitive range included both Santo Domingo
and Porto Rico because no differentiation between specimens
from the two places is seen." The fact that its bones may here
be found near or at the surface in some deposits is evidence for
his further belief that it was the last native mammal to become
extinct on Puerto Rico. This must have been soon after the
discovery by whites.

If we assume that the Indians used this as a staple food
animal, the implication of its presence on these three different
island habitats is that they kept it in a state of semidomestica-
tion and perhaps bred it in captivity for food. This they
apparently did with the guineapig (the cori), remains of which
also occur in shell heaps in the Dominican Republic and Cuba.
Possibly, too, this was a particularly docile and tractable
species, as some of the hystricomorphs are known to be.

HAITIAN ISOLOBODON

ISOLOBODON LEVIR (Miller)

Ithydontia levir Miller, Smithsonian MisA Coll., vol. 74, no. 3, p. 5, Oct. 16, 1922
(Cave "northeast of St. Michel de 1'Atalaye, northwest end of the central plain of
Haiti").

Isolobodon levir Miller, Smithsonian Misc, Coll., vol. 81, no. 9, p. 14, Mar. 30, 1929.

FIGS.: Miller, 1929a, pi. 2, figs. 3, 3a.

Originally based on a single molar tooth from cave deposits
in Haiti, further excavations in the same region disclosed that
this species was after all an Isolobodon and that the original
tooth was an upper premolar and not a lower molar as was at
first supposed.

From the abundant remains secured in the later work in the
St. Michel area, which lies at the northwest end of the central
plain of Haiti, Miller (1929a) was able to show that this is the
most abundantly represented of the vertebrates found in the
bone-bearing deposits. The fragments of palates and jaws
are constantly smaller than are those of the Puerto Rican
isolobodon, the eight largest among 600 mandibles having
tooth rows 16.2-17.2 mm. in length, while in the latter animal
the range is from 17.6 to 19.2 mm. Such other measurements
as can be obtained from crania of the Haitian animal confirm
its slightly smaller size.

The evidence now available, as Miller brings out (1929a,

EXTINCT AND VANISHING MAMMALS

1929c), suggests that the remains of this genus identified as
I. portoricensis, occurring abundantly in the Indian cultural
deposits in the region of Samana Bay, Dominican Republic,
were probably brought in originally by the Indians, if not
directly from Puerto Rico, at least as domesticated stock from
that island, for food. The smaller form native to Haiti and
doubtless extending to other parts of Hispaniola was found
sparingly in kitchen middens but is the sole one represented in
the great number of remains from the caves near St. Michel in
the central part of Haiti. Imperfect skulls also were found in
a small deposit at San Gabriel near Samana Bay, a deposit
which, like that at St. Michel, was probably due to the giant
barn owl, Tyto ostolaga, now extinct, which must have sub-
sisted largely on these animals and died out when they be-
came too few to sustain it. At what time the animal vanished
is, of course, indeterminate, but the implication is that the
aborigines about Samana Bay, at least, found it easier to
subsist on /. portoricensis, which they may have domesticated,
than to hunt the smaller Haitian animal, which may already
have become scarce or restricted in distribution at about the

time of the discovery.

NARROW-TOOTHED HUTIA

APHAETREUS MONTANUS Miller

Aphaetreus montanus Miller, Smithsonian Misc. Coll., vol. 74, no. 3, pp. 3, 4, Oct. 1C,
1922; also vol. 81, no. 9, p. 16, 1929 ("Larger of the two caves northeast of St.
Michel tie 1'Atalaye, northwest end of Central Plain," Haiti).

Fios.: Miller, 1929a, pi. 2, figs. 4, 4a, 4b.

This is another remarkable octodont rodent, of about the
size of Isolobodon and closely related to it, but so far as known
confined to the island of Hispaniola and now apparently too
exterminated.

It was first described by Miller from mandibles from the
cave deposit northeast of St. Michel, Haiti, among the moun-
tains. The teeth differ conspicuously from those of Isolobodon
and Plagiodontia, which they somewhat resemble in their
diagonally placed enamel plates, through a further develop-
ment of these folds, so that the posterior of the two inner
folds extends across and fuses with the outer, thus cutting off
a posterior transverse space. In addition the teeth are re-

NORTH AMERICA AND THE WEST INDIES 123

markably compressed anteroposteriorly, so that they are
much narrower than wide. In these characters it represents
a more advanced stage of development than Plagiodontia and
Isolobodon, in which the reentrant folds are still separate.

In the excavations at the type locality carried on after the
preliminary work had disclosed the two jaws on which the
original account was based, a more extended search brought
to light abundant remains of this animal, permitting the more
detailed description in Miller's (1929a) second paper. Judged
from the material recovered this was the second most common
species brought in to the caves supposedly by the now extinct
giant barn owl. In a later report Miller (1929c) remarks that
no remains of this animal have as yet been discovered in the
various kitchen middens and cultural deposits in the Samana
Bay region; hence there is no real evidence that the species was
utilized by the aborigines as a food animal, though they may
have hunted it. In addition to the collection of 17 skulls
and more than 200 mandibles recovered from the cave deposit
at St. Michel, other fragments were secured at San Gabriel
near Samana Bay, from a deposit evidently made by owls, as
in the case of the former. Equally interesting is the discovery
of six mandibles, all of immature individuals, in cave deposits
on Gonave Island off southern Haiti, indicating its wide
dispersal in Hispaniola in older times (Miller, 1930). The
animal was apparently unknown to the early Spanish explorers,
nor is it possible to tell whether it survived later than the early
years of white occcupation.

HETEROPSOMYS INSULANS Anthony

Hcteropsomys insulans Anthony, Ann. New York Acad. Sci., vol. 27, p. 202, Aug. 9,

1916 ("Cueva de la Ceiba, near Utuado, Porto Rico").

FIGS.: Anthony, 1916, pi. 11, figs. 2-3; pi. 12, figs. 1-5 (skull); 1918, p. 407, fig. 40,
A-D (skull), p. 410, fig. 41 (vertebrae).

HOMOPSOMYS ANTILLENSIS Anthony

Homopsomys antillensis Anthony, Bull. Amer. Mus. Nat. Hist., vol. 37, p. 187, Jan. 29,

1917 ("Cave . . . near Utuado, Porto Rico").

FIGS.: Anthony, H. E., 1917a, pi. 5, fig. 8 (rostrum); 1918, p. 407, fig. 40, E (skull).

These two species are known from cranial parts chiefly
excavated, with remains of other animals, in caves of Puerto
Rico. In all likelihood they were contemporaneous with

124 EXTINCT AND VANISHING MAMMALS

these latter and may have survived up till relatively recent
times. They are believed to represent genera closely akin to
the agoutis of South and Central America but present certain
differences in cranial and dental characters. They are in-
cluded here for the sake of completeness.

Both were perhaps slightly smaller than the typical agouti
of Trinidad or Brazil, representing the genus Dasyprocta.
Anthony (1916, 1918) describes Heteropsomys as having a
skull smaller than that of an agouti, with a total length of about
70 mm. There are four molars in each jaw, of nearly equal
size, each with a prominent infolding of enamel on the middle
of the inner side in the upper teeth and on the outer side in
the lower teeth. In the few known specimens there are three
small transverse islands of enamel in the first upper tooth,
and two in the second (and probably two succeeding true
molars). An unworn tooth would doubtless show that these
are the result of wearing away the summits of shallow trans-
verse depressions in the enamel covering. The molars are each
provided with two short conical roots. The posterior narial
passage or interpterygoid fossa extends forward in a V-shape
to the level of the second molar. In general the skull is less
elongate and relatively wider than in typical agoutis. Anthony
writes that judged from the position of the remains discovered
"this rodent is doubtless of a later age" than either the small
ground sloth Acratocnus or the large rodent Elasmodontomys
found in the same cave deposits.

Concerning Homopsomys, only a few broken crania are
known from which to determine its relationships. The denti-
tion is missing from these, except for a few fragments, but the
skull shows a similar broad, flattened, frontal area, with
bluntly pointed postorbital projections. In his later paper
Anthony (1918) writes: "Homopsomys is very closely related
to Heteropsomys. Indeed it now appears that the two forms
are probably congeneric, and I would make them so but for
the fact that the material at hand is rather inadequate." The
differences between the two, such as the length of palate,
smaller size of Heteropsomys, its lighter incisors, and more
convex frontals, may be characters due to age rather than
specific or generic characters, so that future specimens, if
obtained, may throw additional light on the relationship of
the supposed species. Although Miller compares his Brotomys

NORTH AMERICA AND THE WEST INDIES 125

with these, Anthony insists that they resemble agoutis rather
than spiny rats and would even make a separate subfamily,
Heteropsomyinae, for their inclusion.

Hitherto these genera have not been found in any of the
West Indies except Puerto Rico.

Family HEPTAXODONTIDAE

HEPTAXODON BIDENS Anthony

Heptaxodon bidens Anthony, Bull. Amer. Mus. Nat. Hist., vol. 37, p. 183, Jan. 29,

1917 ("Cave . . . near Utuado, Porto Rico").
FIGS.: Anthony, 1917a, pi. 5, figs. 1-6 (palate, jaw, teeth).

This must have been a remarkable hystricomorph, and it
adds another to the peculiar mammals of this type known from
Puerto Rico. It is known only from fragments of the jaws
secured in cave excavation at Utuado and near Morovis and
Ciales and hence was probably well distributed in the forested
parts of the island at no very distant period. Its few remains
seem to have occurred in much the same manner as those of
other mammals in these deposits; hence it may be assumed
that it lived up till about the time of discovery by whites.

Of about the size of a woodchuck, this animal was notable
not only for the laminated pattern of its upper and lower
grinding teeth, but also for the fact that these seem to have
been reduced to but two in each series. The upper premolar
was of seven "distinct and separate parallel laminae of strong
enamel . . . oblique to the main axis of the toothrow,"
while the lower premolar was similar but reversed. In the form-
er, however, the last enamel space is not quite completely cut
off from the one in front. In the type fragment a space is
present for a molar, and the palatal notch extends forward to
the level of the hind margin of the premolar. In a lower
mandible there are apparently only the premolar and first
molar present. The length of the premolar is 11.5 mm.

According to Dr. Anthony, its discoverer, the affinities of
Heptaxodon are with the Puerto Rican Elasmodontomys, but
on the other hand it is so different that both may be regarded
as representing distinct subfamilies of the Chinchillidae, or it
may be relegated to a distinct family of its own (Miller and
Gidley). What relationship if any it may have to the genera
discovered by the same investigator in Jamaica, and named

126 EXTINCT AND VANISHING MAMMALS

by him Clidomys, Speoxenus, and Spirodontomys, is still to be
made out. These three are represented by teeth and a few
other fragments almost all of which were in a hard breccia and
may be of greater age than the cave material of Puerto Rico.
The teeth indicate stages of development that may be ancestral
to such a condition as found in Heptaxodon, but as yet only a
preliminary account has been published (Anthony, 1920b).

Family DINOMYIDAE: Giant Rodents

"QlJEMl" OF OVIEDO
QUEMISIA GRAVIS Miller

Quemisia gravis Miller, Smithsonian Misc. Coll., vol. 81, no. 9, p. 23, Mar. 30, 1929

("The crooked cave near the Atalaye plantation," St. Michel, Haiti).
FIGS.: Miller, 1929a, pi. 4, figs. 2, 3.

ELASMODONTOMYS OBLIQUUS Anthony

Elasmodontomys obliquus Anthony, Ann. New York Acad. Sci., vol. 27, p. 199, Aug. 9,

1916 ("Cueva de la Ceiba, near Utuado, Porto Rico").
FIGS. : Anthony, 1916, pis. 13, 14 (skull and teeth); 1918, figs. 25-37 (skull and other

skeletal parts).

Since almost nothing beyond skeletal remains of these two
species is known, and because they stand evidently in close
relationship, they may be considered together. In general
size they were about alike and probably were nearly as big as a
paca, Quemisia, inhabiting Hispaniola, and Elasmodontomys,
the adjacent island of Puerto Rico. Miller, in describing the
former, expresses his conviction that it is the animal called
"Quemi" by Oviedo in his account of the animals of Hispaniola,
published slightly more than a quarter of a century after the
discovery.

These two mammals, the largest of the extinct rodents known
from their respective islands, are hystricomorphs with rela-
tionships to the South American Dinomyidae and Chinchilli-
dae, but since there is no evidence that they had developed
leaping modifications, it may be better to regard them as
members of the former, as proposed by Miller and Gidley.
In both, the enamel pattern of the cheek teeth consists of
diagonally transverse folds running nearly across each tooth,
so as to present a laminate appearance, with five enamel
ridges to each. In Quemisia, according to Miller (1929a), the

NORTH AMERICA AND THE WEST INDIES 127

main points of difference in contrast to Elasmodontomys are:
Reentrant enamel folds less nearly transverse, slanting for-
ward at an angle of about 21 instead of about 50; the man-
dibular symphysis longer, extending back beyond the middle
of the first molar instead of barely to the middle of the pre-
molar; shaft of the lower incisor not extending behind the
symphysis instead of far beyond it to terminate beneath the
middle of the second molar. The incisors show a shallow
depression on their front face, slightly more marked in the
lower than in the upper pair. The skull of Elasmodontomys
was about 5 inches long and proportionately broader than in
the Cuban Capromys pilorides. Anthony (1918) has well
illustrated the principal bones of the skeleton. He shows that
the sacrum consists of four well-fused vertebrae and that the
tail was probably short. It seems to have been a heavy-
bodied animal, and its short phalanges indicate terrestrial
rather than climbing habits.

According to the brief description of Oviedo, the "Quemi"
resembled the hutia in color but was larger. That it was util-
ized by the natives of Hispaniola as a food animal is indicated
by the presence of its limb bones at a depth of about 4 feet in
a kitchen midden near the entrance of a cave at Boca del
Infierno, in the Samana Bay region, Dominican Republic
(Miller, 1929c). It must have become extinct soon after the
coming of the Spaniards, probably about the first half of the
sixteenth century. As to its representative on Puerto Rico,
Elasmodontomys, nothing is known. Anthony found its bones
"well bedded in the red stalactite formation, the deeper
layers of the cave deposit," so that it may have died out at an
earlier time.

Here may be mentioned appropriately a third but much
larger member of the same group, a rodent almost as large as
a black bear, to which Cope (1868) gave the name Amblyrhiza
inundata and later synonyms based on single teeth of the same
animal. These teeth were discovered in a cargo of cave earth
sent to Philadelphia from the island of Anguilla at the extreme
northeastern point of the Antillean chain. Additional frag-
ments were later discovered in a cave on the adjacent island
of St. Martin. More recently several fragments, including a
large part of a skull and portions of the palate and incisors,
have come to light in the museums at Leiden and Delft,

128 EXTINCT AND VANISHING MAMMALS

collected over 50 years ago. These supply additional facts
concerning the structure and indicate that the animal was
closely allied to the "Quemi" of Santo Domingo, with essen-
tially the same enamel pattern of the molars but with a rela-
tively longer rostrum. A full account of these has been pub-
lished by Schreuder (1933) with figures. The complete skull
is estimated to have been about 400 mm. (16 inches) in length,
and hence the animal must have been a giant in comparison
with the "Quemi". How long ago this species lived, how it
reached the two neighboring small islands on which its remains
are found, and why it should have perished there are questions
still impossible to answer. That it lived well after glacial
times is probable, but whether it was a contemporary of the
aboriginal Indians there is uncertain.

If Miller is correct in identifying his Quemisia grams with
the "Quemi" of Oviedo, the name should stand as Quemisia
quemi (J. B. Fischer), since this name was given in 1830 by
Fischer [Synopsis Mamm., Addenda, p. 389 ( = 589)], who
quotes Oviedo and Latinizes his description. He includes as
synonymous, however, the account of the "Spanish Racoon"
from Browne's "Civil and Natural History of Jamaica," an
unidentified animal that was probably the larger Cuban hutia,
Capromys pilorides.

Family DASYPROCTIDAE : Agoutis
ST. VINCENT AGOUTI

DASYPROCTA ALBIDA Gray

Dasyprocta albida Gray, Ann. Nat. Hist., ser. 1, vol. 10, p. 264, 1842 ("St. Vincent's,"
West Indies).

In former times agoutis were doubtless present in most of
the Lesser Antilles, and they still occur or did so until modern
times on at least the following (named in order from south to
north): Tobago, Grenada, St. Vincent, Barbados, St. Lucia,
Dominica, Guadeloupe, Montserrat, and St. Kitts. This
carries them all the way, practically, from Trinidad and South
America, north throughout the chain to the northeast corner
of the Antilles, a point where the distribution of some other
Lesser Antillean animals stops, and beyond which to the
westward the distribution of certain Greater Antillean species

NORTH AMERICA AND THE WEST INDIES 129

begins (such as that of Amblyrhiza on Anguilla and its relatives
on Puerto Rico) . It is difficult to say whether this is a natural
distribution, owing to the species having succeeded in crossing
from island to island northward, or whether it is due to human
agency whereby they were introduced and successfully colon-
ized. Their flesh is considered excellent food, and they are not
difficult to keep in captivity; hence it is reasonable to believe
that the aborigines may have been responsible for their intro-
duction from the continent or Trinidad to the various islands
to the northward. Du Tertre, writing in 1654, says that
among the French islands implying St. Kitts, Guadeloupe,
and Martinique they were well known at that time and were
much hunted for their flesh, dogs being trained to run them.
When pursued, they seek shelter in a hollow log or tree whence
the hunters smoke them out. The female was said to bring
forth two young at a time, in a nest made on the ground under a
bush. In early days, too, it was said that their sharp incisor
teeth were used as blades, probably set in some kind of a
handle, by the Caribs for cutting their skin all over their
bodies to draw blood in their religious ceremonies. A later
French missionary, Labat, writing in 1742, said that he be-
lieved it was found on all the islands, although he acknowledged
that he did not know of its presence on Martinique. It was
then common on Dominica, Guadeloupe, and St. Kitts. Its
absence, even at the present day, on Martinique is interesting
and requires explanation. Possibly the introduction of so
poisonous a snake as the bushmaster in early times may have
exterminated it there if it previously had been present. Per-
haps, too, volcanic eruptions, with their discharge of poison
gas, may have had an occasional disastrous effect in the past,
as when Mount Pelee went into action a generation ago.
Agoutis are more or less diurnal in habit, feeding on vegetable
matter, and are especially fond of certain forest fruits, so that
hunters sometimes resort to the expedient of sitting in such a
tree and attracting the agoutis by throwing a stone from time
to time through the leafy branches. The agoutis evidently
mistaking the noise of the stone crashing through the leaves
to the ground for the fall of one of the fruits may sometimes be
seen cautiously coming from the thicket where they have hid-
den, to seek the fruit, and thus expose themselves to a shot.
In external appearance this agouti does not differ essentially

130 EXTINCT AND VANISHING MAMMALS

from the Dasyprocta rubatra of Trinidad. The forehead, back,
flanks, chest, and belly are a mixture of blackish with small
tickings, which vary in color from ochraceous or buff on fore-
head and shoulders to orange-rufous on the lower back,
ochraceous-buff on the flanks and belly, and whitish on the
lower throat. The chin is nearly clear whitish to buffy; the
inner sides of the thighs are bright buff; the nape, limbs, and
feet are uniform brownish black, and the central part of the
back may vary to a darker blackish tone. The hind feet have
but three toes, the front feet four, with stout short claws for
digging. There is a short, naked, stumplike tail. Ears short
and rounded. Length of head and body up to 18 inches.

It is perhaps uncertain how far the agoutis on the different
Leeward Islands have become differentiated into island races.
No comparisons have yet been made with sufficient series of
adults from them to be very significant. Nevertheless, as long
ago as 1842, J. E. Gray named as a new species an agouti
from St. Vincent, calling it Dasyprocta albida, on account of
its being "whitish grey, nearly uniform, the hair of the back
elongated, white at the base." This specimen must have been
somewhat albinistic, and probably young, for it is said to have
been the size of a guineapig. On this basis alone the St.
Vincent agouti is not distinguishable, for other specimens from
there prove to be of normal coloring. Nevertheless, scanty
material seems to indicate a smaller size as compared with D.
rubatra of Trinidad, with a shorter hind foot (about 95 mm.
instead of 110), and a smaller skull, having a shorter and
more sharply tapering rostrum and shorter nasals. In view of
these apparent differences the St. Vincent agouti may be re-
garded as distinct and Gray's name will apply to it.

As to the present status of this race, for such it seems best
regarded, little is known. Sclater, in 1874, recorded two
specimens sent to the Zoological Society of London in August,
1868. In February, 1904, Austin H. Clark procured an adult
female there in the forest behind Barrouallie, where probably
it may still be found in small numbers. A similar dearth of
recent information exists in regard to the status of agoutis on
the other Leeward Islands. However, where large stretches of
forest still remain, there the animals may be expected to survive
until further clearing or the introduction of such predacious
animals as mongooses takes place. That on some of these

NORTH AMERICA AND THE WEST INDIES

131

islands apparently local races differing in slight characters of
size or proportion are recognizable, points to an introduction
of agoutis at a long-distant period or else they reached their
present localities by occasional accident at equally remote
times. On a visit to Grenada in 1910, I inquired of native
hunters, who agreed that an agouti still occurred in the moun-
tain forests of the interior, but I was unable to secure a speci-
men. In the nearby group of small islands known as the
Grenadines, Austin H. Clark tells me that an agouti has been
introduced at the southeastern end of the island of Bequia, but
apparently it has not thrived, perhaps on account of the lack
of water. The Museum of Comparative Zoology has a skin of
the Antillean agouti from Grenada that was received about
1879. Another specimen, collected on Barbados about 1870,
came from Governor Rawson and presumably was of local
origin, though the agouti is not mentioned in Hughes 's account
of the natural history of Barbados in 1750. With the gradual
clearing of wooded areas on any of these islands and the intro-
duction of the mongoose (already common for a number of
years past on Grenada) it is tp be expected that agoutis will
eventually be extirpated from all the Lesser Antilles, perhaps
before specimens can be had sufficient to determine the extent
of their local variation.

St. Vincent agouti (Dasyprocta albida)

132 EXTINCT AND VANISHING MAMMALS

ST. LUCIA AGOUTI

DASYPROCTA ANTILLENSIS Sclater

Dasyprocta antillcnsis Sclater, Proc. Zool. Soc. London, 1874, p. 666 (St. Lucia, West

Indies).
FIG.: Sclater, 1874, pi. 82 (colored).

In November 1874, two agoutis from the island of St. Lucia,
West Indies, were presented alive to the Zoological Society of
London and were shortly afterward described as a new species
by Sclater, who, however, compared them only with the very
different Dasyprocta punctata of Central America. His colored
plate shows an animal of closely similar appearance to the D.
rubatra of Trinidad or the D. albida of St. Vincent, so that
there can be no doubt of the near relationship of these two.
That the name may stand for the island animal as a valid race,
however, is indicated by a comparison of two skulls from skins
in the Museum of Comparative Zoology, taken on St. Lucia
probably about 1880.

This agouti is nearly the same in size and appearance as D.
rubatra, but in the skulls available it has shorter nasals, which
seem more narrow and tapering distally. The rostrum is
shorter, and the incisive foramina are prolonged posteriorly
to the meeting of the premaxillary with the maxillary, instead
of ending in advance of that point. How far these characters
will hold with a larger series remains to be seen. Nor is it
certain whether agoutis from other islands of the Lesser
Antilles are locally different.

No recent information is at hand as to the status of the
agouti in St. Lucia at the present time. On the neighboring
island of Dominica Allen and Chapman (1897) recorded a speci-
men from the forested part, and it was said to be still common
in the interior. On Montserrat it occurred also until at least
very recent years, for the United States National Museum has
a specimen received in 1902 from that island. The skull of this
individual shows a number of peculiarities, which if borne out
by additional skulls might warrant naming the agouti of this
island as a distinct local race. Agoutis of this type must once
have been common on St. Kitts, for Du Tertre (1654) and
Labat (1742) both mention them two or three centuries ago.
Two skins from this island are recorded in the catalogue of
mammals in the Museum of Comparative Zoology as having

NORTH AMERICA AND THE WEST INDIES 133

been received about 1881, but they have since been lost sight
of. There is another in the United States National Museum.
Presumably the species may still be present there. St. Kitts
is thus the most northern of the Lesser Antilles from which it
has been reported. It may be added that the South American
Dasyprocta aguti, characterized by its bright ochraceous or
rufous rump, is the one introduced on St. Thomas. Miller
(1911) reported the identity of this agouti through a specimen
received by the National Museum, while in 1917 the Museum
of Comparative Zoology received one from the same island.
In the island of Guadeloupe an agouti has long been known,
and in 1914 I named it on the basis of two specimens secured
by Dr. G. K. Noble.

GUADELOUPE AGOUTI

DASYPROCTA NOBLEI G. M. Allen

Dasyprocta noblei G. M. Allen, Proc. New England Zool. Club, vol. 5, p. 69, 1914
("Goyave, Guadeloupe Island, West Indies").

Two specimens secured in ?914 by Dr. G. K. Noble differ
notably from available skins of the neighboring islands in
their darker appearance, due to extensive blackening of the
head, neck, shoulders, and back, with a suppression of the
reddish or ochraceous ticking, which becomes confined to the
lower flanks. In a third specimen, later obtained, however,
the opposite effect is seen, through the concentration of rusty
tips to the hairs of the entire back, perhaps an individual and
erythristic condition. If the agouti of Guadeloupe thus proves
to be distinct from those on the other islands, it may eventually
be better to regard it as merely an insular race of Dasyprocta
rubatra.

As to the status of this animal on Guadeloupe, Dr. Noble
reported that at the time of his visit in 1914 it was difficult to
secure and was confined to a limited area of uncleared country.
The animals make runways and are occasionally obtained by
waiting nearby for a shot as one may chance to appear.

134 EXTINCT AND VANISHING MAMMALS

Order CARNIVORA: Dogs, Cats, and Their Relatives

The larger carnivorous mammals that compete with man for
food, destroy domestic animals, or threaten man's security,
have been exterminated or greatly reduced in settled areas.
Seven families of true carnivores are recognized :

(1) Canidae, the dog family, includes wolves, jackals, foxes,
and some less familiar types. Representatives of this family
are found throughout the world, except in New Zealand and
some of the oceanic islands.

(2) Ursidae, the bear family, formerly occurred throughout
Europe, Asia, and North America and in the East Indies, the
Atlas Mountains of North Africa, and in the South American
Andes.

(3) Procyonidae, the raccoon family, includes the pandas,
the coati, the kinkajou, and the cacomistle or ring-tailed cat.
Except for the Asiatic pandas this group is confined to the
New World.

(4) Mustelidae, the weasel family, includes the small fur
bearers weasels, minks, otters forms that suffer persecution
chiefly for the sake of their valuable pelts. The weasels and
their relatives occur in all regions except the Australian and
the oceanic islands.

(5) Hyaenidae, the hyaenas and the aard-wolf, are restricted
in Recent times to Africa and southern Asia. The African
aard-wolf is a termite-eating mammal with degenerate teeth
and jaws; it is rare everywhere in its range.

(6) Viverridae, the genets, civets, and their allies, are catlike
or weasellike carnivores with partially retractile claws and with
numerous cheek teeth. Representatives are found in the Ethi-
opian, Oriental, and parts of the Palearctic regions.

(7) Felidae, the cat family, is world wide in distribution,
except for the Australian and polar Regions and the oceanic
islands. Except for the fossane (Cryptoprocta) of Madagascar,
which is genetlike, the cats are a very uniform group, differing
in little except size, length of tail, and color pattern.

Four families are considered in the present volume :

(1) Ursidae: Three races of the black bear, all the grizzlies,
and Alaskan brown bears.

(2) Mustelidae, weasels, martens, mink, and wolverene: 20
species and subspecies, belonging to three genera, are discussed
here; all are valuable fur bearers.

NORTH AMERICA AND THE WEST INDIES 135

(3) Canidae, foxes and wolves: 22 forms, representing two
genera, have been exterminated or endangered.

(4) Felidae, the cat family: 10 races of the puma and two of
the jaguar are treated here. J. E. H.

Family URSIDAE: Bears

THE BLACK BEARS

The black bears are much smaller than the grizzlies, with
shorter claws and usually with black coloration, although in
western North America several local color phases occur. The
skull is characterized by its relatively short snout and short
nasals in comparison with the grizzly bears; the length of
nasals is equal to, instead of being more than, the width across
the front of the first upper molars, and the two posterior cusps
of the first lower molar instead of being in approximately the
same transverse plane are arranged so that the outer is slightly
in advance of the inner. The black bears were formerly found
all over the wooded parts of North America south into northern
Mexico, but at the present tmie this range has been locally
restricted through killing them out in the more settled areas.
Nevertheless so wary and intelligent are these animals that
they have in many regions been able to survive in spite of per-
secution, and hence only a brief consideration is here needed.
Hitherto, in addition to the typical form of the eastern parts of
North America, no less than 13 local forms have been described.
The value of these can not be fully estimated until a general
study is made of the entire group with adequate material.
However, in a recent paper, Dr. E. R. Hall (1928) has reviewed
the bears of this group inhabiting the Northwest coast and
concludes that the seven forms described from the region are
all subspecies of the typical black bear. In the western United
States and on the Pacific coast, a cinnamon phase occurs as
far north as Taku River, Alaska, and young in the same litter
may be either black or cinnamon in color. A gray or bluish
phase is found in the region between Lynn Canal and Cape St.
Elias, and this has been described as Ursus emmonsii, the
glacier bear. Hall (1928) points out cranial characters that
may serve to distinguish these bears as a local race apart from
color, for black individuals occur within the same range.

136 EXTINCT AND VANISHING MAMMALS

Finally, on Gribble and Princess Royal Islands, British
Columbia, three color phases occur, the lightest of which is
nearly white and has been named Ursus kermodei. But "the
light phase is not always pure white and may include yellowish
rufous and bright golden rufous bands and spots." Since the
color characters of Kermode's bear offer but slight grounds for
distinction, Hall recognizes this as a valid race chiefly on the
basis of minor cranial characters.

In the following account only a bare outline of the present
status of some of the forms seems necessary.

EASTERN BLACK BEAR

EUARCTOS AMERICANTJS AMERICANUS (Pallas)

Ursus americanus Pallas, Spicilegia Zoologica, pt. 14, p. 5, 1780 (eastern North

America) .
FIGS.: Elliot, 1901, pi. 34 (three views of skull); Nelson, 1916, p. 440 (col. figs.).

Since the settlement of eastern North America by whites,
the black bear has retreated from the more populous or cleared
sections, but, given a reasonable area of forest, it will readily
adapt itself to contact and persist in wild areas. In eastern
United States a few bears still exist in western Massachusetts,
they are common in the less settled parts of Maine, and are
present in small numbers in the White Mountains of New
Hampshire and the mountainous parts of Vermont. They no
longer are found in Rhode Island or Connecticut but remain in
some numbers in the Adirondacks of New York State and in the
Pennsylvania mountains. According to recent census esti-
mates of the U. S. Fish and Wildlife Service, the approximate
bear population of these States is: Maine, 1,400; New Hamp-
shire, 650; Vermont, 400; Massachusetts, 10; New York 1,000.
Rhoads in 1903 said that at that time it was almost extermi-
nated in New Jersey; however, a few seem to have persisted in
the cedar swamps of the southern part of the State and in
the adjacent parts of Delaware. A bear was reported to have
been killed in Johnsonburg, N. J., in 1920. Pennsylvania,
with its large areas of mountain forests, is estimated to have
a bear population of 3,300; Maryland, 150; and West Virginia,
2,100, in the mountainous regions. Virginia has fewer, an
estimated 600; North Carolina nearly twice that number,
South Carolina 230, and Georgia 400. West of the Alleghenies

NORTH AMERICA AND THE WEST INDIES 137

bears are still present in numbers in the wilder parts of Michi-
gan and Wisconsin, but they were exterminated from Illinois
by 1860 and from Indiana by the middle eighties. Tennessee
still harbors a good number in the mountains, but Kentucky,
which never was very good bear country, holds only a few at
the present time. There are a few in Mississippi, and formerly
they were no doubt present in Missouri, but the census quoted
includes none from the latter State. Hibbard (1933) writes that
though once common in eastern Kansas, the black bear is now
extinct in that State.

Northward of the United States, black bears range through
New Brunswick into southern Canada and as far northward as
Ungava. They occur also in Newfoundland but are said now
to be scarce. In the wilder regions they extend across to Alaska
in timbered country. Bears of this species avoided the open
plains country but formerly extended into the valleys of
North Dakota. At present they are probably very scarce or
nearly gone in that State.

Whether the black bear of Florida and Alabama, known as
Euarctos floridanus, and the so-called yellow bear of Louisiana,
Euarctos luteolus, are to be Regarded as races of Euarctos
americanus or as separate species is still uncertain. At all
events, the former is now rare in Florida, though in some
sections it still occurs and the Wildlife Census credits the State
with 350 and Georgia to the north with 400. In Alabama,
according to A. H. Ho well (1921), bears are at present ex-
terminated everywhere except in the swamps of the southern
part. Skulls from this region are intermediate between the
two forms noted above. The census mentioned gives the
State a population of 300 bears. Westward bears are present
over suitable country to central California. In this State they
seem to have increased in abundance with the disappearance
of the grizzly. They formerly extended to northern Mexico
and no doubt still occur there, but no estimates of status are at
hand.

GLACIER BEAR; BLUE BEAR

EUARCTOS AMERICANUS EMMONSII (Dall)

[Ursus americanus] var. emmonsii Dall, Science, new ser., vol. 2, p. 87, 1895 (St. Elias

Alps, near Yakutat Bay, Alaska).
FIG.: Nelson, 1916, p. 440 (left col. fig.).

138

EXTINCT AND VANISHING MAMMALS

This is a race of the black bear, distinguished from the
typical form by slight cranial characters. As might be ex-
pected it is in these characters most like the race perniger, of
the Kenai Peninsula, but "has the rostrum decidedly longer
and inflated anteriorly over the canines. Also the upper
molars are smaller and the mastoid and zygomatic breadths
are greater" (Hall, 1928). The name will apply to the black
bears of the mainland of southern Alaska. In this race a dilute
color phase occurs, which is blue-gray in appearance, but not
all the bears of the region are thus colored, so that this peculi-
arity, which first led to its recognition in nomenclature, is not
characteristic of all the individuals.

The blue bear, or perhaps in particular those individuals
showing the "blue" phase, is supposed to be "threatened with
extinction" because of over-hunting and because this phase has
a circumscribed distribution, in the vicinity of the Glacier Bay
National Monument. However, since this latter is now a
reserve, adequate patrolling of the region to prevent poaching
should insure its perpetuation.

KERMODE'S BEAR; WHITE BEAR

EUARCTOS AMERICANUS KERMODEI (Horiiaday)

Ursus kermodei Hornaday, 9th Ann. Kept. New York Zool. Soc., for 1904, pp. 81-86,

Jan., 1905 (Gribble Island, British Columbia).
FIGS.: Hornaday, 1905, 2 pis.; Allen, J. A., 1909, p. 234, fig. 1 (skin); p. 236, figs. 2, 3

(skull); Anthony, H. E., 1928, pi. 4, opp. p. 76 (colored).

Kermode's bear (Euarctos americamis kermodei)

NORTH AMERICA AND THE WEST INDIES 139

Originally believed to be a species distinct from the black
bear and characterized as clear creamy white to the roots of
the hairs, further specimens of this bear from the type locality
show that this coloring is not constant but skins may have the
top of the head "yellowish rufous, with the back . . . con-
spicuously varied with bands and irregular patches of bright
golden rufous" (J. A. Allen, 1909). The skull presents dis-
tinctive characters that warrant its recognition as "a strongly
marked form"; however, according to Hall (1928), the latest
writer to consider its status, it is best regarded as a local race
of the black bear. The known range extends "from the lower
part of South Bentinck Arm, Bella Coola River (lat. 52), and
from Aristazable, Princess Royal, Gribble, and Pitt Islands, on
the coast, to a considerable distance into the interior" of
British Columbia. A list of specimens and localities up to
1909 is given in the paper of J. A. Allen (1909).

Because of its circumscribed range, and its relative con-
spicuousness, it might be thought that this type of bear would
be a special object of pursuit and thus be in danger of early
extermination. However, I. M. Cowan, in a verbal communi-
cation to Dr. Francis Harper in 1937, regards it for the present
as "safe enough." It may be thought of as a geographical
race of the black bear, with in some cases a white color phase.

THE GRIZZLY BEARS

In the early settlement of western North America, the
grizzly bear was about the only really dangerous carnivore to
be found; its size and strength as well as its rather truculent
nature combined to make it a formidable beast against which
the rifles of earlier days were often ineffective, and many a
hunter has owed his life to the fact that the grizzly is too large
and heavy to climb a tree that afforded the hunter a refuge.
Because of these traits and its uncertain temper, as well as on
account of its depredations among the rancher's cattle and
sheep, the grizzly bear has been persistently hunted and killed
wherever possible for upward of a century, with the result
that at the present day it has been largely exterminated over a
great part of the former range.

Merriam (1896) in his preliminary review of the North
American bears wrote: "The Grizzly bears (including the

140 EXTINCT AND VANISHING MAMMALS

Barren Ground bear) may be separated into 4 more or less well-
marked forms, as follows: (a) the true Grizzly, Ursus horribilis
Ord, from the northern Rocky Mountains; (b) the Sonoran
Grizzly, 'var. horriaeus' Baird, probably only a subspecies; (c)
the Norton Sound, Alaska, grizzly, probably another sub-
species; (d) the very distinct Barren Ground bear, Ursus
richardsoni Mayne Reid. Whether or not the large Grizzly
from southern California deserves subspecific separation from
the Sonoran animal (horriaeus) has not been determined."
Following this preliminary statement of the case, Dr. Merriam
in succeeding years brought together an enormous series of
specimens from the larger part of the range of the grizzly, and
after intensive study of this material he came to the conclusion
that no less than 86 species of grizzly and big brown bears
could be distinguished as inhabiting western North America!
The main points of difference lie in cranial peculiarities, and
it is uncertain how far the species may be distinguished by
external characters. To many, Dr. Merriam writes, this num-
ber of species of grizzly bears will appear "preposterous," yet
the differences in cranial details are evident on close study.
It becomes then a matter of interpretation as to their signifi-
cance. Thus we do not yet know whether the five species of
grizzly supposed to live together on Admiralty Island will
actually interbreed, or whether the 15 that Dr. Merriam has
described as occurring in British Columbia maintain pure lines
of descent or hybridize. He adds: "Cranial and dental char-
acters among the big bears are very subtle. As a rule com-
parison of any two skulls of essentially the same size brings to
light so many resemblances that one is likely to infer a far
closer relationship than actually exists. This is because the
big bears of the genus Ursus are such a closely interrelated
group that the resemblances far outnumber the differences.
Hence the greatest caution is necessary to avoid misleading
conclusions." Under these circumstances a conservative
course of treatment may be adopted and Dr. Merriam's
original inclusive groups of grizzly bears be considered first in
a very general sense; then the various forms that he since dis-
tinguished may be listed with a brief statement of supposed
range and status.

In a general way the more typical grizzlies inhabited the
West from the edge of the Plains to the Pacific, south into

NORTH AMERICA AND THE WEST INDIES 141

northern Mexico, and north into Alaska and Canada. The
barren-ground grizzly is the type of subarctic western Canada;
the big brown bears are found along the coast from northern
British Columbia to and including the entire Alaskan Penin-
sula, but not to the exclusion of the grizzly -bear forms. From
an economic point of view the grizzly bear is regarded as a
destroyer of game and more especially is disliked on account
of its depredations against livestock, or, recently in northern
Alaska, against the reindeer herds. On the other hand, it
helps keep down the increase of certain burrowing ground
squirrels, and in parts of the North its intestines are used in
making clothing. Its present value to human beings may well
be largely an esthetic and recreational one. To an increasing
number of persons who enjoy the sight of large wild animals or
the thrill of photographing them, these bears can not fail to
give much pleasure, while to the naturalist and to the sports-
man they are a source of interest. The money brought in to
the parts of the country where they are pursued is a substan-
tial help to the local population.

GRIZZLY BEAR

URSUS HORRIBILIS HORRIBIUS Ord

Ursus horribilis Ord, Guthrie's Geography, 2d Amer. ed., pp. 291, 300, 1815 ("Mis-
souri River, a little above mouth of Poplar River, northeastern Montana, " fide
Merriam).

SYNONYMS: For a full list of the names applied to grizzly bears see Merriam, 1918.

FIGS.: Audubon and Bachman, 1854, Quadrupeds of North America, vol. 3, pi. 131;
Baird, 1857, pis. 41, 42 (skull and teeth) ; Nelson, 1916, p. 442 (col. fig.) ; Merriam,
1918, pis. 1-16 (living animal and skulls).

Merriam (1918) writes that "the differences between the
grizzlies on the one hand and the big brown bears on the other
are neither so great nor so constant as at one time believed
. . . The typical brown bears differ from the typical
grizzlies in peculiarities of color, claws, skull, and teeth. The
color of the former is more uniform, with less of the surface
grizzling due to admixture of pale-tipped hairs; the claws are
shorter, more curved, darker, and scurfy instead of smooth;
the skull is more massive; the fourth lower premolar is conical,
lacking the sulca^e heel of the true grizzlies. But these are
average differences, not one of which holds true throughout the
group."

142 EXTINCT AND VANISHING MAMMALS

Baird describes the general color of grizzly bears as varying
from a nearly uniform dark brown with the tips of the hairs
lighter, or brownish yellow, to a condition in which there is a
dark blackish dorsal stripe, and the sides of the body are pale
brownish yellow, bordered by a dark stripe along the flanks.
The limbs are generally black or dark brown; head and belly
yellowish, and a tinge of hoary over the shoulders and sides.
Claws long and nearly straight, those on the fore feet about 4
inches or more long. Skull long, the length of nasals exceeding
the width across last premolars. First lower molar with the two
posterior cusps on about the same transverse plane, and with
a small accessory cusp or two on the inner side between the
anterior and posterior cusps. Weight upward of 1,200 pounds,
but males are larger than females; "in some the disparity in
size is very remarkable ... in a few cases the difference
is slight" (Merriam). Bear skulls undergo a series of changes
from early life to old age and in most species do not attain
their mature form until seven or more years of age (Merriam) .
Baird states that the largest specimens are "from six to seven
feet" in total length. The total length of skull runs to 14 or
15 inches.

In a general way the former range of the grizzly extended
from the eastern edge of the Great Plains in about the longi-
tude of the one-hundredth meridian, westward to the Pacific
coast in northern Lower California, thence south to Durango,
Mexico, and northward to the southern and central parts of
the Alaskan Peninsula and the Arctic coast. In the barren-
ground region west of Hudson Bay it is represented by the
so-called barren-ground grizzly. Baird, in 1857, wrote that
"it appears first to occur on the Missouri, above Fort Pierre
[in South Dakota], and becomes more and more abundant
higher up on the Missouri, and especially on the Yellowstone;
thence to the Rocky Mountains, which it inhabits throughout
its entire extent in the United States . . . To the north
it extends far into the British possessions, and southward into
Mexico." It is safe to say that at the present time the grizzly
bear is extinct over much of this area and survives only in
some of the national parks or in wilderness regions.

In their general habits grizzly bears like somewhat open,
rugged country with scattering thick growth where they may
take shelter. Their food includes a wide range of both animal

NORTH AMERICA AND THE WEST INDIES 143

and vegetable matter. In early times they preyed partly upon
the larger game and animals, such as deer, wapiti, or, on the
plains country, the bison, and later found an easier prey in the
introduced livestock of the settlers. In the Southwest, Bailey
mentions their fondness for acorns, pinyons, cedar berries,
various succulent roots, the sweet mesquite berries, bulbs, and
tubers of various kinds. They also turn over logs and stones
for the insects to be found underneath. In more northern
parts of the range they dig out the burrows of small rodents,
especially of ground squirrels, for their occupants. Fish too are
very much relished, and when salmon are running in the
Alaskan waters the grizzlies and brown bears feast upon them,
wading in and scooping out the fish.

The grizzly bear hibernates in the northern part of its range
from late September to April, more or less, seeking a shelter
among rock crevices or in the hollow base of a large tree or
under its upturned roots. In the southern latitudes, as in New
Mexico, however, Bailey says that hibernation is less regular,
and the bears may be found active at various times in winter.
The young are born in these winter dens and may be two or
three in number, more commonly twins.

In various papers Vernon Bailey has much to say concerning
the grizzly's former range and abundance. In his "Mammals
of North Dakota" (1926) he presents a number of early records
of its occurrence along the eastern border of its range here.
"At the coming of the white man these large grizzlies were
apparently common over practically all of North Dakota."
Alexander Henry while in the Red River Valley in 1800 found
them not so numerous as in the Hair Hills and Devils Lake
region, where they were said to be "very malicious." He also
reported one skin in the catch from the Sandhill River, Minn.,
in 1807, and one from Portage la Prairie, Manitoba, localities
from near the limit of its eastward range. In 1833, Maxi-
milian found them more common the farther up the Missouri
he went. Generally bears were found feeding here on buffalo
carcasses, which were often plentifully distributed in the quick-
sands or along the river banks by floods and breaking ice. "Ap-
parently the Missouri River Valley with its great abundance
and variety of large game, wild fruit, and berries, bulbs, tubers,
roots, and underground beans, was a paradise for these bears
before the days of the rifle." Here, too, Audubon, in 1843,

144 EXTINCT AND VANISHING MAMMALS

found them common, and in 1856 F. V. Hay den of the trans-
continental surveys collected specimens. Bailey quotes an old
resident that in 1867 there were many grizzlies in the river
bottoms about Fort Buford and farther west in Montana.
Old hunters told of killing grizzlies in the Killdeer Mountains
between 1864 and 1870. In the late eighties they were still
to be found in smaller numbers along the Little Missouri, but
they were constantly hunted down, so that "at the present
time there is certainly not a grizzly bear left in the State of
North Dakota, and it is doubtful if there is anywhere a living
representative of tnis original species of the grizzly group that
was first given a scientific name and status in literature . . .
Like some of the savage tribes with which it was associated, it
has in passing left behind a thrilling record of savage bravery
of surpassing interest to red-blooded Americans."

In the middle of the last century grizzly bears were found in
some numbers in Nebraska and Kansas. Baird (1857) lists
specimens in the LTnited States National Museum from several
localities in the former State, and Hibbard (1933) notes that
in earlier years it was common in Kansas west of the Flint Hills.
It has long been extinct in these States and probably in Texas
as well, where, Vernon Bailey (1905) writes, "The only speci-
men of grizzly bear that I have seen or heard of from Texas
was killed in the Davis Mountains in October, 1890" and was
found to agree in all essential characters of the skull with the
so-called Sonoran grizzly. Merriam, however, made it the
type of a special subspecies, Ursus horriaeus texensis, in 1914,
later according it specific rank; he believed it identical with
bears from the adjacent mountains of southern Colorado,
a region where Gary (1911) found grizzlies rare in the wilder
parts over 30 years ago, though occasionally seen in the San
Juan, San Miguel, and La Plata Ranges. In northern Colo-
rado, according to the best information he could obtain, in
1905-6 it was already rare. Along the eastern slopes of the
Front Range, however, where grizzlies were still found in the
early seventies, they were by 1907 apparently quite gone.
He gives a number of recent records of the grizzly in the
southern mountains of Colorado up to 1907, and no doubt a
few still remain in the more inaccessible mountainous regions
of the central and western parts of the State, for a 1939 census
issued by the U. S. Fish and Wildlife Service placed the

NORTH AMERICA AND THE WEST INDIES 145

estimated number of grizzlies in Colorado at only 10 indi-
viduals.

In recent decades grizzlies have become all but extinct in
New Mexico and Arizona and probably in the adjacent parts of
northern Mexico as well, where a century ago they were
common. A skull collected about 1852 near Coppermines
(now the site of Santa Rita, Grant County, N. Mex.) was
regarded by Baird (1857) as representing a smaller brown race,
which he named horriaeus, but Bailey (1931) says that speci-
mens collected in that region in later years are not identical
with it, so that an immigration of grizzlies into this area from
elsewhere is assumed. He quotes the journals of Pattie who
hunted in this region in 1824 and 1825 and wrote that it
"abounds with these fierce and terrible animals." Bailey
(1931) writes that in 1894 Dr. A. K. Fisher was told that
grizzly bears were then common in the mountains about 80
miles north of Silver City, N. Mex. In 1905 a large female was
killed in the Tularosa Mountains and a large male was said to
be still at large in that range. In 1908, there were said to be a
few still in the Mimbres, Mogollon, and San Francisco Ranges.
Since then the numbers seem to have decreased still further
for in 1939 the U. S. Biological Survey estimated that only
three grizzlies remained in New Mexico, so that the species
soon no doubt will no longer exist there. Much the same story
is probably true of Arizona, for which the same report fails to
include any bears at all. It may be doubted, too, if any remain
in the adjacent parts of Mexico, although Merriam lists speci-
mens in the National collection as follows: Arizona, 1856, 1913;
Chihuahua, 1899. A grizzly killed on Mount Taylor, northern
New Mexico, in 1916, may well be among the last from that
State.

A century and more ago grizzly bears were common in Cali-
fornia and extended a short distance farther south into the
mountains of northern Lower California. From the latter
region, however, it must soon have been exterminated, for Dr.
E. W. Nelson (1921) states that the early account by Pattie
of his journey in these regions gives the only Lower Californian
record, namely, that there were grizzlies in the Sierra Juarez
near Santa Catarina Mission. In their work on the "Fur-
bearing Mammals of California," Grinnell, Dixon, and Lins-
dale (1937) have given a very full account of the grizzly bears

146 EXTINCT >AND VANISHING MAMMALS

in that State. Dr. Merriam from a study of available skulls
believes that seven specific types are represented amqng them.
In the coastal counties, Humboldt and Mendocino, grizzlies
were commoner in the river valleys where settlements were
first made, so that on account of their depredations upon
stock and the threat to the lives of children and others passing
along roads, they were soon exterminated, within a short time
after the first settlers arrived, about 1850. According to one
informant the last grizzly taken in Humboldt County was
killed in the Mattole section in 1868. The same person tells
that his father in the early days had once counted 40 bears in
sight at one time from a high point in the Mattole section
that afforded a wide outlook over open country. In Mendocino
County the authors quoted found on inquiry among old resi-
dents that the last grizzlies there were killed in the fall of 1875,
when an old female, a yearling, and a large male were shot.
Various fragments concerning the killing of these bears or
incidents of narrow escapes from them are brought together
in the work mentioned, from which one may gather an impres-
sion of the abundance of grizzlies in the third quarter of the
last century, their boldness and occasional depredations, as
well as the continual warfare waged against them by the
settlers. As a result of constant hunting, bears became
scarcer by the beginning of the last quarter of the century,
but they still persisted in some localities. In the nineties there
were still grizzlies "in the near ranges of the San Gabriel
Mountains due north of Pasadena," even as late as 1897.
"In the heavily chaparral-covered Santa Ana (or Trabuco)
Mountains of Orange and extreme northwestern San Diego
counties" they persisted "until well along in the 1900's," and
the latest record of one being killed there is given as January 5,
1908. One of the last grizzlies killed in the Yosemite region
was in 1887. "The last known southern California grizzly"
was killed October 28, 1916, near Sunland, Los Angeles County,
while so far as the writers above quoted are able to ascertain,
the last certain record of a grizzly killed in California was of
one shot in Tulare County in August, 1922. A few later
reports were un verifiable.

In the State of Oregon these bears are perhaps already gone.
Bailey (1936) has brought together many notes on their earlier
abundance and gradual diminution. In the days of muzzle-

NORTH AMERICA AND THE WEST INDIES 147

loading rifles, grizzlies were bolder, and hunters might well
hesitate to risk coming to close quarters with them. In the
Lake of the Woods region the last grizzly is said to have been
shot in 1868. In the last quarter century these bears have
slowly diminished in numbers even in national forests. Bailey
writes that in 1910 the supervisor of the Siskiyou National
Forest reported grizzlies as very scarce; one was said to have
been killed in the Yamsay Mountains about 1911, and there
were possibly a few in the Mount Pitt section at about that
time. In 1924 and 1925, the United States Forest Service
reported one grizzly on each of the Cascade and Siskiyou
National Forests, and in 1931 two and in 1932 one on the
Wallowa Forest, while in 1933 it reported one on the Willam-
ette. The last (1939) report of the U. S. Fish and Wildlife
Service does not include it in a game census of the State ; in the
State of Washington, however, it gives a surviving remnant
of nine.

At the present time the main population of grizzly bears in
the United States outside of Alaska centers in the Rocky
Mountain region in the States of Wyoming and Montana.
This is partly a result of their protection from usual hunting
in Yellowstone and Glacier National Parks, though others
apparently frequent the wilder parts of these States. In the
Yellowstone Park some grizzlies even come to the garbage
piles near the hotel, but in Glacier Park, writes Vernon Bailey
(1918), few have learned to do this, but most of them are shy
and live in the less frequented portions of the park. A few
are killed around the edges of the park annually. At times,
too, the forest ranger in these parks finds it necessary to elimi-
nate a particular grizzly that has developed troublesome ways.
According to the 1939 census issued by the U. S. Fish and
Wildlife Service (in December, 1940) the following estimates
of the grizzly-bear population by States give at least an ap-
proximation of its present status: Arizona, 7; Colorado, 10;
Idaho, 44; Montana, 620; New Mexico, 3; Washington, 9;
Wyoming, 480. There is thus an estimated total of approxi-
mately 1,100 grizzlies in the United States, excluding Alaska.

Northward of the United States, grizzly bears are not un-
common in the Rocky Mountain region of British Columbia
and Alaska. Cowan (1939) writes that they are now restricted
to the mountainous parts of the country and are fairly com-

148 EXTINCT AND VANISHING MAMMALS

mon in the mountains bordering the Pine River in British
Columbia and in the Rockies north and south of the Peace
River. Still farther north, grizzly bears are well known to
occur in some numbers in the wilder parts of these regions to
central Alaska. Preble (1908), writing of the Athabasca-
Mackenzie area, says that they are found "throughout the
Rocky Mountain range and its eastern spurs west of the Mac-
kenzie, north to the Arctic .coast." Although annually some
are killed, it seems likely they will long hold out here. A report
of the U. S. Biological Survey in December 1938 states that
"the large brown and grizzly bears in Alaska are holding their
own as an outstanding wildlife resource of the Territory."
In parts of Alaska they appear to be more plentiful than for
many years, owing, it is believed, to the favorable attitude of
Territorial residents, to protective regulations under the
Alaska Game Law of 1925, and to sanctuary areas that total
more than 8,500,000 acres. Some three-fourths of these
sanctuary areas are comprised in Mount McKinley National
Park and Katmai and Glacier Bay National Monuments.
In addition, certain areas are now closed to hunting and else-
where regulations afford a closed season from June 20 to
September 1. The bag limit for large brown and grizzly bears
is two a year, except on Admiralty Island, where a limit of one
has been made, to induce nonresident hunters to visit other
areas. The sale of bear hides is now prohibited in Alaska, a
regulation considered as an outstanding factor in permitting
the increase of these big bears, while of equal importance is
the development of a better feeling by the residents, who now
"thoroughly appreciate this source of income and have no
desire to jeopardize it by killing off this attraction at no profit
to themselves." Along the Arctic coast bears may still be
regarded as a nuisance by certain interests, as miners and
trappers, but the Survey's agent at Nome reported that in his
district grizzlies are fairly numerous with no definite increase
or decrease. The few animals that are killed are those slain
by natives working with reindeer herds, and he believes their
actual damage is slight. In these regions that may long serve
as hunting areas it seems likely that these interesting and
remarkable carnivores will continue indefinitely under wise
management.

NORTH AMERICA AND THE WEST INDIES 149

BARREN-GROUND BEAR

URSUS RICHARDSONI Swainson

Ursus richardsoni Swainson, Animals in Menageries, p. 54, 1838 (shore of the Arctic
Ocean, on west side of Bathurst Inlet near mouth of Hood River, Mackenzie,
Canada).

FIGS.: Merriam, 1896, pi. 4, fig. 6; pi. 5, fig. 5; pi. 6, fig. 3 (skull).

This grizzly is believed to range over the barren-grounds
from west of Hudson Bay to the Mackenzie River, and north-
ward to the shores of the Arctic Ocean. The color varies from
yellowish to grizzly brown. Size rather small for a grizzly,
and the skull distinguished by its broadly spreading zygomata
and the sagittal crest, which runs far forward and divides
dichotomously with a branch running to each postorbital
process so far anteriorly that the two ridges are nearly trans-
verse. Dr. Merriam (1918) regards the animal from the mouth
of the Mackenzie as a related but distinct species, which he has
called Ursus russelli, and in his paper he gives a series of
cranial measurements of the two.

Preble (1908) has summarized various bits of information
concerning this little-known type of grizzly. Apparently the
first white man to come in contact with it was Samuel Hearne,
who in his famous journey (1770-1772) to the mouth of the
Coppermine River saw a skin of an enormous grizzled bear
at the Eskimo tents there. Franklin also observed it several
times along the Arctic coast from near Fort Enterprise to
Bathurst Inlet and mentions seeing a female with three cubs.
The stomach of a bear killed at Bathurst Inlet contained the
remains of a seal, a marmot, roots, berries, and grass. In
another case a bear had been feeding on a caribou. Bell met
with the species in 1900 "quite often" along the west and north
shores of Great Bear Lake.

As to its present status, Dr. R. M. Anderson (1939b) writes:
"A few grizzlies of the Barren Ground group, a typical form
being Ursus richardsoni, ranging eastward in the Arctic Zone
as far as Bathurst Inlet, and perhaps a little beyond, have
until the past few years escaped hunting by the Eskimos, who
formerly kept off the bear's range in summer. As these inter-
esting bears do not have the shelter of mountains, it is feared
that with the recent spread of trapping operations and mining
developments in the Far North, and particularly the extension

150 EXTINCT AND VANISHING MAMMALS

of domestic reindeer herding into the Arctic portion of the
Northwest Territories with the effect of patrolling the range
throughout the year, some of these rare species will become
extinct within a few years."

The persistent reports of a grizzly bear existing in the barrens
of northern Labrador have from time to time been investigated
as far as such tales may, but no real evidence that such an
animal is found there has been elicited. Even though the tale
were true that a skin of a grizzly had once been brought in,
this might have been traded from still farther west. In that
way one might account for the origin of the reports. Dr. E.
Stirling (1884) sums up the evidence in a brief statement to the
effect that John McLean in his notes of 25 years' service in
Hudson Bay Territory reports that skins from Ungava had
several times been sent to Europe and that a factor named
Mittleberger said he had known of the animal in Labrador.

KODIAK BEAR; BIG BROWN BEAR

URSUS MIDDENDORFFI Merriam

Ursus middendorffi Merriam, Proc. Biol. Soc. Washington, vol. 10, p. 69, Apr. 13, 1896

("Kadiak Island, Alaska").
FIGS.: Merriam, 1896, pi. 4, figs. 2, 3; pi. 5, fig. 2; pi. 6, fig. 2 (skull); 1918, pi. 3

(skull); Holzworth, 1930 (frontispiece in color and other photographic figures).

This animal is rated by Merriam as the "largest of living
bears, though only slightly larger than Ursus beringiana Mid-
dendorff, from Kamschatka; frontal region in male enormously
elevated, highly arched, and relatively narrow; zygomata
bowed outward to an extraordinary degree ; postzygomatic part
of skull very short." Distinctive features are the contact of
the forepart of the jugal with the lachrymal bone and the
small size of the first upper and last lower molars. The claws
are shorter and more curved than in the usual type of grizzly,
measuring about 100 mm. on the curve.

The Kodiak bear is found on Kodiak Island and the adjacent
Afognak and Shuyak Islands, but it may be taken as a type
of the so-called big brown bears, of which closely allied forms
are found on the nearby mainland, such as Ursus gyas from
Cook Inlet to the entire length of the Alaska Peninsula.
Merriam (1918) writes that "the differences between the
grizzlies on the one hand and the big brown bears on the other

NORTH AMERICA AND THE WEST INDIES 151

are neither so great nor so constant as at one time believed.
And there are species which in the present state of knowledge
can not be positively referred to either group." The color of
the brown bears is more uniform with less of the surface
grizzling, the shorter claws are scurfy instead of smooth, the
skull more massive. The brown bears are in general confined
to the southern Alaskan coast and peninsula and the adjacent
islands. (See Dufresne, 1942, for distribution map.)

In 1912 Sheldon wrote that "the grand bear of Kadiak
Island ... is rarely seen." Osgood (1904), writing of
the big brown bears of the base of the Alaska Peninsula, says
that though formerly abundant persistent hunting by the
natives has much reduced their numbers. They still occur in
many localities but have become extremely shy. Fifteen
years previously it was not uncommon to see "from eight to
fifteen bears scattered about on one mountain side." Their
period of activity varies, but they usually come out from hiber-
nation late in March or early in April, depending on the season,
and are not seen after early November. The young are brought
forth usually in January and may number as many as four at
a birth, though, as with the other grizzlies, two is the usual
number and three not uncommon. Ground squirrels are much
sought after by digging up their burrows, but when salmon are
entering the rivers in spring they become a main object of
pursuit. Small fruits, roots, and grasses also form part of
their fare.

Although these bears a few years ago seemed in much danger
of extirpation, present reports indicate that under the game
laws that went into effect in 1925, and with the creation of
sanctuaries, the outlook is bright that they will remain an
asset to Alaska for a long while to come. Dufresne, the execu-
tive officer of the Alaska Game Commission, reported in 1938
that they are holding their own and that "information from
Kodiak Island and the Alaska peninsula . . . leaves no
question as to the plenitude of the huge brown bears." A
favorable attitude on the part of Alaskan residents, who are
beginning to realize the value of the bears to themselves
through bringing in a certain number of hunters and photogra-
phers, is doing much to insure proper protection and regulation
of the numbers killed. For the present it appears that these
splendid bears are no longer in danger of being exterminated.

EXTINCT AND VANISHING MAMMALS

In order to coordinate the account of the grizzly and brown
bears with the nomenclature proposed by Merriam (1918) in
his review of these animals, the following synopsis of the names
used in his paper and their grouping is added, with a few words
on the status of these forms where information is at hand.

HORRIBILIS GROUP

Ursus horribilis horribilis Ord. BIG PLAINS GRIZZLY

Range: Plains of Montana, Wyoming, Colorado, Dakotas.
Nearly or quite extinct.

Ursus horribilis bairdi Merriam. BAIRD'S GRIZZLY

Range: "Southern Rocky Mountain region from San Juan
Mountains, southwestern Colorado, northward through Wyo-
ming to Montana, and perhaps to southeastern British Colum-
bia." Supposed to be a mountain form; nearly extinct.

Ursus horribilis imperator Merriam. YELLOWSTONE PARK

BIG GRIZZLY

Range: "Yellowstone National Park, Wyoming; limits un-
known." A small number live in the Park and perhaps the
nearby country.

Ursus chelidonias Merriam. JERVIS INLET GRIZZLY

Range: "Head of Jervis Inlet, British Columbia." Status
unknown.

Ursus atnarko Merriam. ATNARKO GRIZZLY

Range : Region of Atnarko River, upper Bella Coola, British
Columbia. Status unknown.

Ursus kwakiutl Merriam. KWAKIUTL GRIZZLY

Range: "Coast region of British Columbia from south-
western corner (Burrard Inlet, Howe Sound, Jervis Inlet)
northwesterly to or beyond the lower Bella Coola." Probably
a few left.

Ursus nortoni Merriam. YAKUT AT GRIZZLY

Range: "Limited apparently to coastal plain on southeastern
side of Yakutat Bay." Probably still found; specimens col-
lected in 1910.

Ursus warburtoni Merriam. WARBURTON PIKE GRIZZLY
Range :" Coast region , . . of southeastern Alaska and

NORTH AMERICA AND THE WEST INDIES 153

adjacent parts of British Columbia from Chilkat River south-
easterly to Atnarko River." Probably still occurs; specimen
collected 1915.

Ursus neglectus Merriam. ADMIRALTY ISLAND GRIZZLY

Range: Admiralty Island. Restricted but probably not in
immediate danger.

Ursus californicus Merriam. CALIFORNIA COAST GRIZZLY

Range: "Humid coast region of California from San Fran-
cisco Bay south about to San Luis Obispo." Probably extinct
since 1908.

Ursus tularensis Merriam. TEJON GRIZZLY

Range: "Dry chaparral hills of interior coast ranges between
the San Joaquin Valley and Los Angeles plain," California.
Extinct in 1916.

Ursus colusus Merriam. SACRAMENTO VALLEY GRIZZLY

Range: "Sacramento (and perhaps also San Joaquin) Valley
and adjacent foothills; westerly ... to Dobbins Creek
canyon on the boundary between southeastern Humboldt and
southwestern Trinity Counties." Extinct.

Ursus dusorgus Merriam. RINDSFOOS'S GRIZZLY

Range: "Head of Jack Pine River near Mount Bess, Al-
berta." Status unknown; type collected in 1916.

PLANICEPS GROUP

Ursus nelsoni Merriam. NELSON'S GRIZZLY

Range: "Sierra Madre of Mexico from northwestern Chi-
huahua and northeastern Sonora south to southern Durango."
Probably a few remain.

Ursus texensis texensis Merriam. TEXAS GRIZZLY

Range: Davis Mountains, Texas, and mountains of southern
Colorado. Probably nearly or quite extinct.

Ursus texensis navaho Merriam. NAVAHO GRIZZLY

Range: "Probably restricted to the isolated Chuska Moun-
tains" between northeastern Arizona and northwestern New
Mexico. Extinct.

154 EXTINCT AND VANISHING MAMMALS

Ursus planiceps Merriam. FLAT-HEADED GRIZZLY

Range: Foothills along the western edge of the plains in
Colorado and Wyoming. Nearly if not quite extinct.

Ursus macrodon Merriam. TWIN LAKES GRIZZLY

Range: Twin Lakes region of Colorado. Probably nearly
extinct.

Ursus mirus Merriam. YELLOWSTONE PARK GRIZZLY

Range : Yellowstone Park region. Probably now uncommon ;
specimen (the type) killed in 1915.

Ursus eltonclarki Merriam. SITKA GRIZZLY

Range: "The Sitka Islands, Baranof and Chichagof,"
Alaska. Probably still occurs.

Ursus tahltanicus Merriam. TAHLTAN GRIZZLY

Range: "Middle and upper Stikine-Skeena region, limits
uncertain/' in British Columbia. Probably still present in
small numbers.

Ursus insular is Merriam. ISLAND BEAR

Range: Admiralty Island, Alaska. Probably not now in
immediate danger.

Ursus orgilos Merriam. GLACIER BAY GRIZZLY

Range: Bartlett Bay and east side of Glacier Bay, Alaska.
Probably still occurs; collected by Hasselborg in 1912.

Ursus orgiloides Merriam. ALSEK GRIZZLY

Range: Italic River region, Alaska; in coast strip southeast
of Yakutat Bay. Probably present in small numbers; type
collected in 1916.

Ursus pallasi Merriam. PALLAS'S GRIZZLY

Range: "Southwest corner of Yukon Territory, east of the
St. Elias Range . . . and adjacent eastern border of
Alaska; easterly to McConnell River and Teslin Lake and
south into northern British Columbia." Probably not in
immediate danger.

Ursus rungiusi rungiusi Merriam. RUNGIUS'S GRIZZLY

Range: Region about the head of Athabaska River and
Kootenay Pass, Alberta, and adjacent parts of British Colum-
bia (Indian Point Creek). Probably still in small numbers.

NORTH AMERICA AND THE WEST INDIES 155

Ursus rungiusi sagittalis Merriam. CRESTED GRIZZLY

Range : Southwestern Yukon, Canada. Probably still occurs
in small numbers and is in no immediate danger. Type col-
lected in 1915.

Ursus macfarlani Merriam. MACFARLANE'S BEAR

Range: Region of Anderson River, Mackenzie; limits un-
known. Specimens from Franklin Bay, Stapylton Bay.
Status unknown; probably still to be found in small numbers.

Ursus canadensis Merriam. CANADA GRIZZLY

Range: "Eastern British Columbia; limits unknown (type
from near Mount Robson; and an adult female from Kootenay
Lake)." Not known to be in danger.

AEIZONAE GROUP

Ursus arizonae Merriam. ARIZONA GRIZZLY

Range: Known from Escudilla Mountains, Apache County,
Arizona. Status unknown, perhaps extinct.

Ursus idahoensis Merriam. IDAHO GRIZZLY

Range: Eastern Idaho (North -Fork Teton River) to Wallowa
Mountains, Oreg. Probably still present in small numbers.

Ursus pulchellus pulchellus Merriam. UPPER YUKON GRIZZLY
Range: Ross River, Ross Mountains and McConnell River,
Yukon, Canada. Probably still present in small numbers;
type collected in 1916.

Ursus pulchellus ereunetes Merriam. KOOTENAY GRIZZLY

Range: Kootenay District, British Columbia. Status un-
known; collected in 1916.

Ursus oribasus Merriam. LIARD RIVER GRIZZLY

Range: Upper Liard River, Yukon, near British Columbia
boundary. Collected in 1916; probably present in small
numbers.

Ursus chelan Merriam. CHELAN GRIZZLY

Range: "Cascade and Cassiar Mountains from northern
Washington to upper Stikine River and Dease Lake, British
Columbia." Nearly extinct.

Ursus shoshone Merriam. SHOSHONE GRIZZLY

Range: "Mountains of Colorado and Wyoming." Probably
very few left in Colorado but more in Wyoming.

156 EXTINCT AND VANISHING MAMMALS

Ursus kennerlyi Merriam. SONORA GRIZZLY

Range: "Nothing is known of the range of kennerlyi except
that the type specimen came from mountains near Nogales,
Sonora," Mexico. "Its affinities with utahensis suggest that
formerly it may have had a disconnected distribution north-
ward in the mountains of central Arizona." Extinct. Type
skull dated 1855.

Ursus utahensis Merriam. UTAH GRIZZLY

Range: "Southern Wasatch and Pine Valley Mountains
[Utah]; limits unknown." Must be now nearly gone, though
specimens have been collected as lately as 1911.

Ursus perturbans Merriam. MOUNT TAYLOR GRIZZLY

Range: Mount Taylor and Datil Mountains, New Mexico;
limits unknown. Now practically extinct.

Ursus rogersi rogersi Merriam. ROGERS'S GRIZZLY

Range: Absaroka Mountains, Wyoming; limits unknown.
The considerable grizzly-bear population of Wyoming must
include a number of this form.

Ursus rogersi bisonophagus Merriam. BLACK HILLS GRIZZLY
Range: "Black Hills of South Dakota and adjacent north-
east corner of Wyoming." Probably extinct; type collected in
1887.

Ursus pervagor Merriam. LILLOOET GRIZZLY

Range: "Interior of southwestern British Columbia; known
only from Lillooet Lake and Bridge River." Probably extinct;
type taken in 1883.

Ursus caurinus Merriam. LYNN CANAL GRIZZLY

Range: "Coast of mainland of southeastern Alaska from
Chilkat River Valley and Lynn Canal south an unknown
distance." Probably rare; type collected in 1911.

Ursus eulophus Merriam. ADMIRALTY ISLAND CRESTED BEAR
Range: Admiralty Island, Alaska. Probably not now in
immediate danger.

Ursus klamathensis Merriam. KLAMATH GRIZZLY

Range: "Siskiyou Mountains of northern California and
southern Oregon, ranging north in recent times to Fort Kla-
math region and Rogue River valley; in early days to lower

NORTH AMERICA AND THE WEST INDIES 157

Willamette Valley (presumably same species); south in Sierra
Nevada an unknown distance." Now about extinct; last ones
killed perhaps about 1911. Bailey (1936) states that according
to Forest Service reports, there were in 1924 and 1925 one
grizzly bear on each of the Cascade and Siskiyou National
Forests, in 1931 two and in 1932 one on the Wallowa Forest,
and in 1933 one on the Willamette.

Ursus mendocinensis Merriam. MENDOCINO GRIZZLY

Range: Known only from Mendocino County, California,
along the northwest coast of the State. Extinct. The last
grizzlies were killed in the county in the autumn of 1875, a
female, a yearling, and a large male, all killed at the same
place on Eel River south of Covelo (Grinnell, Dixon, and
Linsdale, 1937, p. 70).

Ursus magister Merriam. SOUTHERN CALIFORNIA GRIZZLY

Range: Santa Ana or Trabuco Mountains, Cuyamaca and
Santa Rosa Mountains, and probably San Jacinto Mountains
of California. Now extinct; last one killed October 28, 1916,
near Sunland, Los Angeles County, California (Grinnell,
Dixon, and Linsdale, 1937).

HYLODROMUS GROUP

Ursus hylodromus Elliot. FOREST GRIZZLY

Range: "Rocky Mountain region of western Alberta and
eastern British Columbia, including Selkirk Range." Prob-
ably a few still remain.

Ursus kluane kluane Merriam. KLUANE GRIZZLY

Range: "Southwest corner of Yukon Territory east of the
St. Elias Range, extending northwesterly in Alaska to Mount
McKinley region (head of Toklat), easterly in Yukon Territory
to McConnell River . . . and probably south into north-
west corner of British Columbia" (Merriam). Not believed
to be in present danger.

Ursus kluane impiger Merriam. INDUSTRIOUS GRIZZLY

Range: Merriam gives the following localities as indicative
of the range: Brisco, Columbia Valley, in British Columbia;
Morley and Jasper, Alberta; and headwaters of North Fork
Blackfoot River, western Montana. Hollister (1912a) wrote
that "the grizzly bear is still a common animal throughout the

158 EXTINCT AND VANISHING MAMMALS

Robson region" of Alberta, a statement which doubtless still
applies to this animal.

Ursus pellyensis Merriam. PELLY GRIZZLY

Range: Pelly Mountains, Yukon. Grizzlies are still found
in small numbers in the Yukon region, but since it is not
possible to identify animals without killing them and examin-
ing their skulls, it is unknown whether this form is represented
or not.

Ursus andersoni Merriam. ANDERSON'S GRIZZLY

Range: The type specimen is from Dease River, near Great
Bear Lake, Mackenzie; limits of range unknown. Grizzly
bears, according to Treble's (1908) summary, are not uncom-
mon in the region.

HORRIAEUS GROUP

Ursus apache Merriam. APACHE GRIZZLY

Range : White Mountains of eastern Arizona. The type was
collected in 1913, but at the present time the predatory-
animal campaigns have resulted in reducing the numbers to so
low an ebb that already this form may have been extirpated.

Ursus horriaeus Baird. NEW MEXICO GRIZZLY

Range: "Parts of New Mexico, south to Casas Grandes,
Chihuahua, Mexico; probably extending into eastern Arizona"
(Merriam). Probably now extinct.

Ursus henshawi Merriam. HENSHAW'S GRIZZLY

Range: "Lower slopes of southern part of Sierra Nevada;
limits unknown" (Merriam). The type specimen was from
near Havilah, Kern County, California, and Grinnell, Dixon,
and Linsdale (1937) mention as representing probably the
same sort of animal, skins from Madera County. They pro-
visionally refer to this form the grizzly population of the
Yosemite region and the central and northern Sierra Nevada
generally. In the Yosemite region, according to Grinnell and
Storer (1924) the grizzly bear was rare already by 1887, when
the capture of one was considered "an unusual event." The
last one known to have been killed in the Yosemite Park was
shot "'about 1895' at Crescent Lake." Possibly a few indi-
viduals persisted longer, for the authors mentioned tell of one
whose tracks were frequently seen on Bullion Mountain

NORTH AMERICA AND THE WEST INDIES 159

between the years 1908 until 1911. Undoubtedly this form
of bear is now extinct.

STIKEENENSIS GROUP

Ursus stikeenensis Merriam. STIKINE GRIZZLY

Range : Region about head of Finlay River and Dease Lake
region, northern British Columbia (Merriam). Grizzlies in
this region seem in no immediate danger.

Ursus crassodon Merriam. BIG-TOOTH GRIZZLY

Range: Merriam refers to this form, skulls from Klappan
Creek (third south fork of Stikine River) ; White River, Yukon;
Wolf Lake, Yukon; Tatletuey Lake, upper Finlay River; Glen-
lyon Mountains and Quiet Lake, southern Yukon. Since other
supposed forms of grizzlies also inhabit this general region,
they may only be identified by killing them and studying the
skull. At present the grizzly bear is in no danger in this region,
for no cattle are raised, human population is small, and the
occasional hunter is their only enemy.

Ursus crassus Merriam. THICK-SET GRIZZLY

Range: Merriam records skulls from upper Macniillan
River, Yukon; and Anderson River, Horton River, and Lang-
ton Bay, on the Barren-grounds. The type specimen was
collected in 1916. No immediate danger for it is seen.

Ursus mirabilis Merriam. STRANGE GRIZZLY

Range: Admiralty Island, Alaska. With the present re-
ported favorable attitude of the local population and the
recent game regulations, it would seem that this grizzly will
not be hunted beyond a reasonable limit.

Ursus absarokus Merriam. ABSAROKA GRIZZLY

Range: "Laramie and Bighorn Mountains, eastern Wyo-
ming, Black Hills region, South Dakota, and northward along
Little Missouri to Missouri and Yellowstone Rivers." Prob-
ably extinct in the Black Hills and eastern Wyoming. The
latter State still has a relatively large population of grizzly
bears but these center in the Yellowstone Park region, and are
regarded by Merriam as a special type. The type specimen
from near head of Little Bighorn River, northern part of Big-
horn Mountains, Montana, was killed in 1893, and the animal
may still occur in that region.

160 EXTINCT AND VANISHING MAMMALS

ALASCENSIS GROUP

Ursus alascensis Merriam. ALASKA GRIZZLY

Range: Norton Sound region, Alaska, south to Chinitna, on
Cook Inlet; limits unknown. Osgood (1901, p. 68) at the
beginning of this century wrote that "large bears are still very
often seen both on the Alaska Peninsula side of Cook Inlet and
on the mountainous Kenai Peninsula. According to report
they were very abundant about ten years ago, but in the short
time since have been so constantly pursued that their numbers
have been greatly reduced." Probably this local form requires
protection if it is to survive, but no recent report is at hand.

Ursus toklat Merriam. TOKLAT GRIZZLY

Range: "So far as known, restricted to Alaska Range"
(Merriam). The type specimen came from the head of
Toklat River, north base of Alaska Range, near Mount Mc-
Kinley, Alaska. Since this region is now a national part,
there seems to be no immediate cause for anxiety for the
continued existence of these bears in their range.

Ursus latifrons Merriam. BROAD-FRONTED GRIZZLY

Range: Rocky Mountains of western Alberta and eastern
British Columbia from Jasper House northwesterly to head-
waters of Stikine River. Probably the same remarks apply to
the status of this bear as to that of Ursus crassodon. Some
numbers may still occur.

RICHARDSONI GROUP

Ursus richardsoni Swainson. BARREN-GROUND BEAR

Range: The type came from the shore of the Arctic Ocean
on the west side of Bathurst Inlet, near the mouth of Hood
River, and probably is representative of the animal of northern
and northeastern Mackenzie. Anderson notes that of late
years with the extension of reindeer herding by the natives,
these regions, which previously were but little hunted, may
now be more often patrolled and these bears are likely to be
in a much more precarious situation than formerly.

Ursus russelli Merriam. MACKENZIE DELTA GRIZZLY

Range: Lower Mackenzie region; limits unknown. Of the
barren-ground bear in general, Preble (1908) states that it is
rare over the greater part of its range. Probably it is not

NORTH AMERICA AND THE WEST INDIES 161

likely to persist about the mouth of the Mackenzie for any
great length of time.

Ursus phaeonyx Merriam. TANANA GRIZZLY

Range: "Tanana Mountains between Tanana and Yukon
Rivers," Alaska. Dr. W. H. Osgood, who obtained the type
specimen in 1903, wrote that "very little accurate information
is obtainable in regard to the grizzly in the Yukon region,"
and this state of affairs seems to continue. However, with the
development of mining interests and settlement, it is not likely
that this form of the middle Yukon will resist encroachment
for many years.

Ursus internationalis Merriam. ALASKA BOUNDARY GRIZZLY

Range: "Region bordering Arctic coast along international
boundary, and doubtless adjacent mountains, between the
coast and the Yukon-Porcupine; limits unknown." Status
not determined.

Ursus ophrus Merriam. HIGH-BROWED GRIZZLY

Range: The type, collected in> 1915, came from an unknown
locality in eastern British Columbia. Nothing further is
known of its status.

Ursus washake Merriam. WASHAKIE GRIZZLY

Range : The type came from the north fork of the Shoshone
River, Absaroka Mountains, in western Wyoming, between
Bighorn Basin and Yellowstone National Park. Since the
characters distinguishing the form are to be found in the skull
and teeth it is not possible to say what portion of the many
grizzlies found living in western Wyoming are like it.

KIDDERI GROUP

Ursus kidderi kidderi Merriam. KJDDER'S BEAR

Range: "Alaska peninsula for its entire length." According
to the report on Alaska bear sanctuaries issued by the U. S.
Department of Agriculture in December 1939, "information
from . . . the Alaska Peninsula . . . leaves no question
as to the plenitude of the huge brown bears," so that with the
continuance of protective legislation no fears are felt for them.

162 EXTINCT AND VANISHING MAMMALS

Ursus kidderi tundrensis Merriam. TUNDRA BEAR

Range: "Tundra region of northwestern Alaska from Shak-
tolik River on Norton Sound, southerly across the lower
Yukon, Kuskokwim, and Nushagak Rivers to Bristol Bay and
north side of base of Alaska Peninsula" (Merriam). The
Government report above quoted states that "the grizzlies of
the interior seemed to be on the increase during the last year"
(1938). The agent at Nome reports that "the Alaska grizzly
is fairly numerous all through my district, and no increase or
decrease is definitely known. Very few animals are killed and
these by natives working with reindeer herds. Bears are
reported by all reindeer men as destructive to deer, but I
believe that the extent of damage is not important."

Ursus eximius Merriam. KNIK BEAR

Range : Nothing seems to be known of the range of this bear
beyond its occurrence at the head of Knik Arm, Cook Inlet,
Alaska. Probably, like Kidder's bear of the Alaska Peninsula,
it is "holding its own."

INNUITUS GROUP

Ursus innuitus Merriam. INNUIT BEAR

Range: "Coastal region of Norton Sound, Alaska, from
Unalaklik northward and westward; limits unknown" (Mer-
riam). The range of the grizzly bear probably does not reach
the sea to the northward of the region inhabited by this form.
The report of the Nome agent, previously quoted, seems to
indicate that grizzly bears are in no immediate danger of
extermination in this area.

Ursus cressonus Merriam. CHITINA BEAR

Range: "Chitina River valley and adjacent slopes of Skolai
and Wrangell Mountains, westerly doubtless through the
Chugach Mountains to the west side of Cook Inlet; occurs as
far south as the Iliamna region" (Merriam). This local type
may be thought of as present in small numbers, but exact
information is not at hand.

Ursus alexandrae Merriam. Miss ALEXANDER'S GRIZZLY

Range: Kenai Peninsula, Alaska. This is notable as the
largest of the grizzly bears. No recent information of a definite
nature is at hand as to its present numbers.

NORTH AMERICA AND THE WEST INDIES 163

TOWNSENDI GROUP

Ursus townsendi Merriam. TOWNSEND'S BEAR

Range: Mainland of southeastern Alaska. The type came
from Sitka in 1889. No information as to status is at hand.

DALLI GROUP

Ursus dalli Merriam. D ALL'S BROWN BEAR

Range : Malaspina Glacier and region northwest of Yakutat
Bay, Alaska. Osgood (1904) writes of the general region about
the base of the Alaska Peninsula that "Brown Bears were
formerly abundant in much of the country . . . but the
persistent hunting by the natives since the introduction of
modern repeating rifles has reduced their numbers greatly.
They still occur in many localities, but have become extremely
shy and are seldom obtained unless a special campaign for
them is conducted."

Ursus hoots Merriam. STIKINE BROWN BEAR

Range: Stikine River region, British Columbia. This is
believed to be a small race of tj^e brown bear, perhaps closely
related to the form sitkensis. Merriam mentions but three
specimens from the Stikine River. Its present status is un-
known but is perhaps little likely to have altered much recently.

Ursus sitkensis Merriam. SITKA BROWN BEAR

Range: Sitka Islands (Baranof and Chichagof), Alaska. In
the press notice of the U. S. Department of Agriculture for
December 18, 1938, Wildlife Agent Douglas Gray of Juneau
reports that the large brown bear continues to be "the greatest
attraction to the trophy hunter that this district offers. Their
numbers seem to be as plentiful as ever, notably on the Baranof-
Chichagof-Admiralty Island group." With proper regulation
of hunting, therefore, this form seems in no danger.

Ursus shirasi Merriam. SHIRAS'S BROWN BEAR

Range: Restricted to Admiralty Island. This is "a very
large member of the brown bear group," but whether it is
always black like the type specimen is unknown. Its "excep-
tionally broad skull with broad short rostrum, excessively
broad and short frontal shield, and huge massive postorbital
processes," though obviously different in comparison with the
"long narrow skull with slender elongate rostrum, long and

104 EXTINCT AND VANISHING MAMMALS

narrow frontal shield, and insignificant postorbital processes"
of the form eulophus living on the same island, may not be as
distinctive in life. Hence its status on Admiralty Island is
not made out. With the closing of two areas totaling 52,000
acres on Admiralty Island, it is presumed that the brown
bears present in some numbers there must include a proportion
of this type.

Ursus nuchek Merriam. NUCHEK BROWN BEAR

Range: Prince William Sound easterly to Mount St. Elias;
limits unknown. The range continues that of the Yakutat
bear on the Alaskan coast. No late information is at hand as
to its status there, but under present regulations the existing
conditions are at least favorable.

GYAS GROUP

Ursus gyas Merriam. PENINSULA GIANT BEAR

Range: "Entire length of Alaska peninsula from Cook Inlet
to Isanotski Strait and adjacent Unimak Island." This is
"either the largest living bear or second only to the great
Kadiak bear (middendorffi)," according to Merriam. Skulls
of adult males show "a surprisingly wide range in size and
form," so that Merriam distinguishes four types among them.
On account of its large size this bear might form an especial
attraction to hunters and thus be in more danger from this

W3^^

Barren-ground bear (Ursus richardsoni)

NORTH AMERICA AND THE WEST INDIES 165

source than other "species" having the same range. Present
information from the Alaska Peninsula, issued by the U. S.
Department of Agriculture, "leaves no question as to the
plenitude of the huge brown bears."

Ursus middendorffi, Merriam. KODIAK BEAR

Range : Confined to Kodiak and the adjacent islands Afognak
and Shuyak. This bear shares with gyas the distinction of
being the largest of living bears. The two are closely allied.
Present information indicates that its numbers are being
maintained.

KENAIENSIS GROUP

Ursus kenaiensis Merriam. KENAI GIANT BEAR

Range: Kenai Peninsula, Alaska. No recent statistics are
at hand. With the passage of the Alaska game law in 1925,
however, the conditions are probably as favorable as may be
expected. According to the 1938 press information of the U. S.
Department of Agriculture, "there has been a perceptible
change for the better in the brown and grizzly bear population."

Ursus sheldoni Merriam. MONTAGUE ISLAND BEAR

Range: Believed to be confined to Montague Island, Prince
William Sound, Alaska. The name is based on specimens of
the brown bear type taken in 1905. No recent information as
to the bear of this island is at hand.

INOPINATUS GROUP
Vetularctos inopinatus Merriam. PATRIARCHAL BEAR

This genus and species were based on a skull from Rendez-
vous Lake, northeast of Fort Anderson, Mackenzie, but the
genus is currently regarded as inseparable from Ursus, while
the species is not further known beyond the original specimen
collected in 1864. Its peculiarities would ordinarily be as-
cribed to individual variation.

Family MUSTELIDAE: Weasels and Martens
THE PINE MARTENS

Among the northern fur bearers the martens, because of
their soft, handsome fur, are much sought after, and their
pelts bring high prices in the markets. They are characteristic
of the evergreen forest belt of the northern United States
northward into Canada and extend their range southward

166 EXTINCT AND VANISHING MAMMALS

along the higher parts of the Alleghenies in the East and the
Rocky and Sierra Nevada Mountains of western North
America. Like many carnivorous animals they occur in small
numbers relative to the abundance of the rodents and other
animals on which they prey and are given to traveling over
rather wide circuits in their search for food.

In general appearance pine martens are about the size of a
large slender squirrel, with a rather short cylindrical and well-
furred tail, short legs, and erect, rather prominent ears.
Males are larger than females and when adult develop more
prominent median crests on the skull. The general coloring is
yellowish to light brown on the head and forepart of the body,
darkening to a warm brown posteriorly and on the tail and
legs. The under side is cinnamon, darkening posteriorly, and
often with a tinge of brighter ochraceous-orange on the chest.
Adult males weigh about 2.5 pounds, females about a quarter
less. Total length, about 25 inches, of which the tail is 7 or 8
inches.

Over their wide range across the northern part of the conti-
nent pine martens show a slight degree of geographic variation,
both in shades of color and in details of the skull and teeth.
These have been made a basis for recognizing at least two
species, the typical eastern marten and the darker\animal of
the Northwest coast, each with sundry races. While at; present
it seems still uncertain whether only a single species is Actually
represented, with more or less well-marked subspecies, these
may be listed with their type localities in accordance with
current usage, as follows :

Maries americana americana (Turton). EASTERN PINE MARTEN

Mustela americanus Turton, Turton's ed. of Linnaeus's Systema Naturae, vol. 1,
p. 60, 1806 (eastern North America).

Maries americana abieticola (Preble). KEEWATIN MARTEN

Mustela americana abieticola Preble, North Amer. Fauna, no. 22, p. 68, Oct. 31,
1902 (Cumberland House, Saskatchewan, Canada).

Maries americana abietinoides Gray. NORTHWESTERN MARTEN

Maries americana var. abietinoides Gray, Proc. Zool. Soc. London, 1865, p. 106
(Rocky Mountains between Kicking Horse Pass and Columbia River).

Maries americana actuosa (Osgood). ALASKAN PINE MARTEN

Mustela americana actuosa Osgood, North Amer. Fauna, no. 19, p. 43, 1900 (Fort
Yukon, Alaska).

Maries americana atrata (Bangs). NEWFOUNDLAND MARTEN

Mustela atrata Bangs, Amer. Nat., vol. 31, p. 162, Feb., 1897 (Bay St. George,
Newfoundland).

NORTH AMERICA AND THE WEST INDIES 167

Maries americana boria (Elliot). MACKENZIE MARTEN

Mustela boria Elliot, Proc. Biol. Soc. Washington, vol. 18, p. 139, Apr. 18, 1905
(Lower Mackenzie River district; probably same as M. a. actuosd).

Maries americana brumalis (Bangs). LABRADOR MARTEN

Mustela brumalis Bangs, Amer. Nat., vol. 32, p. 502, July, 1898 (Okak, Labrador)

Maries caurina caurina (Merriam). PACIFIC MARTEN

Mustela caurina Merriam, North Amer. Fauna, no. 4, p. 27, Oct. 8, 1890 (near
Grays Harbor, Chehalis County, Washington).

Maries caurina humboldtensis Grinnell and Dixon. HUMBOLDT MAR-
TEN

Maries caurina humboldtensis Grinnell and Dixon, Univ. California Publ. Zool.,
vol. 21, p. 411, Mar. 17, 1926 (5 miles northeast of Cuddeback = Carlotta, Hum-
boldt County, California).

Maries caurina origenes (Rhoads). ROCKY MOUNTAIN MARTEN

Mustela caurina origenes Rhoads, Proc. Acad. Nat. Sci. Philadelphia, 1902, p.
458, Sept. 30 (Marvine Mountain, Garfield County, Colorado).

Maries caurina nesophila (Osgood). QUEEN CHARLOTTE ISLANDS
MARTEN

Mustela nesophila Osgood, North Amer. Fauna, no. 21, p. 33, Sept. 26, 1901
(Massett, Graham Island, Queen Charlotte Islands, British Columbia).
Maries caurina sierrae Grinnell and Storer. SIERRA MARTEN

Maries caurina sierrae Grinnell and Storer, Univ. California Publ. Zool., vol. 17,
p. 2, Aug. 23, 1916 (Lyell Canyon, Yosimite National Park, California).
Maries caurina vancouverensis Grinnell and Dixon. VANCOUVER

MARTEN

Maries caurina vancouverensis Grinnell and Dixon, Univ. California Publ. Zool.,
vol. 21, p. 414, Mar. 17, 1926 (20 miles south of Alberni, Vancouver Island,
British Columbia).

FIGS.: Grinnell, Dixon, and Linsdale, 1937, vol. 2, pi. 4 (col.), of M. c. sierrae; Nelson,
1918, p. 453, upper fig. (col.), of M. americana; Elliot, 1901, p. 334, fig. 66 (skull
of M. americana); Elliott, C., 1942, col. pi. facing p. 75, of M. americana.

A brief survey of the status of the pine marten shows that
although once common in Nova Scotia, New Brunswick, the
northern New England States, New York, Pennsylvania, and
even New Jersey, it is now through much trapping greatly
reduced, though in places where there are large tracts of
forested country small numbers may remain. In New Bruns-
wick, Chamberlain reported it as common 60 years ago, and
probably it is still to be found in wilder areas. Anderson
(1939a) writes that it was fairly common in the wooded parts
of Quebec in earlier days but has been overtrapped and is
"now rare, found only in areas remote from settlement." Of
the large dark Labrador race no recent statistics are at hand,
but there are many skulls in the Museum of Comparative

168 EXTINCT AND VANISHING MAMMALS

Zoology secured from trappers by J. D. Sornborger about 1900.
The Labrador wilderness is not much penetrated by outside
trappers, and the supply of fur is likely to hold out for a long
time to come. The Newfoundland race, however, is said now
to be very much reduced. In northern New England martens
were formerly much trapped in the Maine Woods and the
White Mountains and are still found but in smaller numbers.
A century ago Emmons (1840) wrote that the pine marten was
to be found in the mountainous parts of the Berkshires, in
Massachusetts, especially in beech woods, and describes the
method of trapping it in deadfalls. At that time, he states,
the pelts were worth 90 cents to $1.1^. Probably martens
have been extinct in Massachusetts for many years, but they
still are found in the Vermont forests. Merriam regarded them
as common in the Adirondacks in 1882 and said that hundreds
were trapped for fur every winter; they still persist in smaller
numbers. The fur is prime early in November. Rhoads,
writing in 1903, makes the general statement that the marten
was "once abundant over all the mountain regions" of Penn-
sylvania and New Jersey, invading even the Nfoothills of the
Alleghenies, but was early exterminated fro^the latter
regions, was probably extinct in New Jersey aboutl&>0, and
at the beginning of this century was to be found only in the
forested parts of the mountains in Pennsylvania. South of
this area there seems to be no recent report of the marten's
presence. Rhoads (1903) in his account of the species in Penn-
sylvania states that there it was partial to hardwood forests
rather than evergreens. Reports from various of the counties
are nearly unanimous in telling of its former occurrence and its
later decline, and in most cases tell of none having been found
for a number of years. Rhoads points to the effect of forest
fires as bearing on the depletion of the stock. In Ohio, Brayton
(1882) mentions the marten as already extinct, and there is
apparently no evidence of its occurrence in Indiana within
recent years. Dr. H. H. T. Jackson (1908) regarded it as
practically extinct in Wisconsin 30 years ago and mentions one
taken in Vilas County as lately as 1904-5, but Cory (1912)
wrote that although "steadily declining in numbers" and then
"a comparatively rare animal" in Wisconsin it was still to be
found in the northern part of that State. It seems to be gone
from Illinois, though earlier recorded from Cook County by

NORTH AMERICA AND THE WEST INDIES 169

Kennicott. Northern Minnesota was once good marten
country, and Herrick (1892) mentions that many dark pelts,
especially prized, formerly came from this area. As to its
recent status along the northern border of Minnesota and in
the adjacent part of Canada, areas now forming national
reservations, Cahn (1937) writes: "The marten is now prac-
tically if not completely extinct in the Quetico country. It has
been exterminated from the Superior National Forest on the
American side of the boundary. Trappers report seeing an
occasional track in winter in the larger timber north of Saga-
naga and Saganagonse lakes, and I found one skull at a trap-
per's cabin up the Wawiag River in 1930. With the removal
of the larger timber and their constant pursuit by trappers, the
marten has retreated to the north." Still farther west, toward
the edge of the Plains, martens were common in northeastern
North Dakota a century and more ago. Bailey (1926) quotes
Wied's early account of the number of furs brought in from
local and more western points, including 500 or 600 marten
skins. More particular numbers of martens taken in this
region are quoted from records of Alexander Henry in 1801-7,
and range from three to as many as 75 in a season at the Pem-
bina River. "At present," says Bailey, "there are probably
no martens in North Dakota," nor is it probable that they
would return "even with careful protection" to the limited
forest areas of the State.

The vast region north arid west of the Great Lakes in Canada
is probably now the main producing ground for marten fur.
Concerning the species there Preble (1908) writes: "The marten
is rather common throughout the forest belt of the Athabaska
and Mackenzie regions. It varies greatly in color, being more
or less subject to melanism throughout this area. The 'dark'
martens are more highly prized than the lighter ones, and
bring a higher price, sometimes as much as four or five times
the value of ordinary skins . . . Martens are common
along the Athabaska, Slave, and Peace Rivers, and large
numbers are traded at all the posts on their banks. Skins
taken in these valleys average rather dark in color and furnish
a good grade of fur. The valley of the Athabaska below Grand
Rapid is said to be especially good trapping ground. The
number annually taken by each trapper varies from a few to
a hundred or more. James MacKinlay informed me that

170 EXTINCT AND VANISHING MAMMALS

three hunters working in the Caribou Mountains, southwest
of Great Slave Lake, trapped in one season nearly 500 martens,
an unusually large catch. The lower Liard River region and the
Horn Mountains also are good trapping grounds. Upward
of 3,000 skins are said to be usually traded at Fort Norman.
Fort Good Hope also receives a large number, and as many as
6,000 have been collected at Fort McPherson during an unusu-
ally good season . . . Martens vary greatly in abundance
in the same locality in different years, and to some extent this
increase and decrease are periodic. The winter of 1903-4 was
marked by a great scarcity of martens over much of the upper
Mackenzie Valley, though in other sections the catch was
about normal." At Fort McPherson, A. J. Stone was told in
1900 that martens were then as numerous as they had been
50 years before when the post was first established.

In Keewatin and the Hudson Bay region Preble (1902)
found the marten (subspecies abieticdta^l^Tly common"
north to the tree limit but most abundant in the heavy spruce
forests of the southern part. Many skins come in to the posts
at Norway House, Oxford House, and York Factory, and a
few at Fort Churchill that come from the lower Churchill
River."

Osgood (1900, p. 44), in describing the large marten (M. a.
actuosa) of the Yukon region, Alaska, wrote at the time that
it "is still the commonest fur-bearing animal of Alaska, not-
withstanding the hundreds of thousands that have already
been taken. Trappers are always confident of a harvest of
martens whether other animals are abundant or not." To the
westward, however, in the Cook Inlet region at the base of the
Alaskan Peninsula, Dr. Osgood (1901) found it only moderately
common. One trapper had taken but 15 martens in a season
of two and a half months in 1899 near the mouth of Turnagain
Arm. The martens of this area seem to be the typical sub-
species. The latest report (July 16, 1939) of the Biological
Survey credits Alaska with having shipped in 1938 no less than
9,200 pelts of this animal, surely a large number.

In the western United States and along the Pacific coast
the numbers are much less, though it is still fairly plentiful in
British Columbia. Of the insular race, nesophila, of the Queen
Charlotte Islands, British Columbia, Dr. Osgood (1901)
reports that it evidently is not abundant. The Haida Indians

NORTH .AMERICA AND THE WEST INDIES 171

trap more or less every winter for them, but the annual catch
seldom exceeds 40 skins. According to the fur traders these
martens are short-haired and light-colored and do not com-
mand so high a price as do those from the mainland. On
Vancouver Island they seem to be "fairly common in the
higher mountains " (Swarth, 1912).

While martens are absent from the treeless parts of western
United States, they are common in many areas of forest, in
spite of much trapping. Vernon Bailey (1918) wrote two
decades ago that in the Glacier National Park region of north-
western Montana "they are probably as common in the park
as anywhere in the country, but no animal with the price on
its skin that they have long maintained could well be numerous
or very common. For at least half a century the park region
has been famous for the number of martens caught each year
by trappers . . . The animals are reported to be more
common on the west slope of the mountains than on the east,
but this is probably because the timber there is more dense and
extensive and it has not been possible to trap them out so
thoroughly . . . Steel traps are generally employed, but
many deadfalls of the ordinary type are used." He recom-
mends that "special permits to reliable parties for trapping
them in the park during a limited season when they become
too numerous would probably control their numbers here,
while outside the park there is no danger of their ever becoming
too abundant." They were reported common in the nineties
in the Salmon River and Saw Tooth Mountains of Idaho by
Merriam, who says that prospectors even complained that
martens would carry off their meat from camp. Southward,
the Rocky Mountain marten (subspecies origenes) seems to be
still fairly common throughout the forested areas into northern
Colorado and in smaller numbers into the higher mountain
forests of the northern part of New Mexico. Gary (1911)
writes that from 1900 to 1905 the best regions for martens in
Colorado were the mountain ranges surrounding Middle and
North Parks, the Williams River Mountains, and the moun-
tains south of Aspen, Pitkin County. They were reported
common on the Park Range but very rare on the Medicine
Bow Range. "A conservative estimate of the annual catch
in the Middle and North Park region would be 100 skins."
Fur buyers at Aspen estimated about the same for their dis-

172 EXTINCT AND VANISHING MAMMALS

trict. Cary was informed that 30 years before (about 1875)
a trapper had taken a great many martens in the mountains of
Gunnison County, central Colorado. Farther south the marten
is much less common and reaches the southern limits of its
range in northern New Mexico. Bailey (1931) adduces a few
older records for this region and concludes that "the animals
occur, but are by no means common in the Canadian Zone
forests of the Sangre de Cristo and San Juan Mountains. It is
very doubtful if they occur in any of the ranges farther south."
According to this author (Bailey, 1936, p. 296) it is this race
of the marten that extends into the extreme northeastern part
of Oregon in the Blue Mountains, where, however, they seem
to be uncommon and decreasing in numbers. In the Wallowa
Mountains a few are still taken, but recent reports indicate
that the numbers are small. \

In the Coast Ranges of OregoA the resident animal is re-
garded as typical M . caurina, grading in the southern districts
into the race sierrae of California^/ Bailey (1936) quotes re-
turns for the season November 1, 1913, to February 28, 1914,
of 518 martens taken in that State by registered trappers,
indicating that the species Js still well represented there.
These at prices then current were worth in the neighborhood
of $13,000. In California two races occur, sierrae of the Sierra
Nevada, and humboldtensis of the northern Coast Range.
They are confined to the Boreal Zone and are regularly trapped,
especially in the northern parts of the State. Grinnell, Dixon,
and Linsdale (1937) have published an extended account of the
species in California and show that particularly in summer it
haunts slide-rock areas among the crannies of which it pursues
small rodents, especially mouse-hares and white-footed mice.
The southern limit of range seems to be Tulare County, in the
southern Sierra Nevada; on the coast, however, the marten
extends only to the northwestern part of the State in the higher
mountains. The authors quoted say that "at least three-
fourths of the martens trapped in California are taken by a few
professional trappers who specialize in this animal." One of
these men took as many as 96 in a single season. However,
reports from the trappers show "a marked decline, amounting
to fully 75 per cent, in the number of martens trapped in a 4-
year period," from 452 in 1920 to 121 in 1924, a decrease be-
lieved to be independent of any cyclic decline. The authors

NORTH AMERICA AND THE WEST INDIES

173

believe this is due to over-trapping and that unless more ade-
quate protection is given the species it will in a short time be-
come scarce or absent except in such protected areas as the
national parks. A year-round closed season is recommended.
Concerning the marten of the Coast Range (race humboldtensis)
the authors above quoted regard it now as of "rather sparse
occurrence" in the redwood belt, "though in earlier years it
was more generally distributed and fairly numerous." Its
southernmost occurrence was in Sonoma County. In 1926
and 1927 two trappers were said to have worked a somewhat
inaccessible pocket within ten miles of Carlotta and to have
caught in all 25 martens, which is considered a remarkable
record.

In general it may be said that while the pine marten is still
plentiful in the most favorable parts of its range, as in the
northwest of Canada and in parts of the Rocky Mountains and
Alaska, its outpost areas in the east and south have been so
very much depleted in recent decades that it is in need of
careful protection if it is to maintain anything like former
numbers or to survive at all. They require a large stretch of
territory to range over and a sufficient food supply in the way
of small rodents. Factors that seriously affect them are the
burning by forest fires of parts of their forest range, clearing
through extensive lumbering operations, overtrapping, and

Eastern pine marten (Maries americana americana)

174 EXTINCT AND VANISHING MAMMALS

killing of females. The fact that they are more or less subject
to a periodic cycle of decline, probably due to disease or failure
of food supply through dying off of certain rodents that form
a staple diet, or to both factors combined, probably would
have an influence on the numbers of a small population, and
when these factors coincide with other unfavorable circum-
stances a danger point is reached. Martens are relatively easy
to trap, so that this trait adds to the hazard of their lives.
Finally, an important factor in their decline from a peak of
numbers lies in the fact that their gestation period is long, so
that killing of females often involves the death of a potential
litter. At the present time there have been various attempts
made to raise martens in captivity, but with only moderate
success. Young are born in March and April and are independ-
ent by some three months later. Further experiments being
carried on by the U. S. Fish and Wildlife Service will doubtless
provide a better knowledge of the best methods of management.
In 1936 a five-year closed season on martens in the United
States was urged by the Chief of the Biological Survey and en-
dorsed by the American Wildlife Conference. Some idea of the
number annually killed in western Canada may be gained from
the statement of Cameron (1936, p. 625) that the fur yield of
the Northwest Territories for the year 1933-34 included no
less than 5,580 marten pelts.

FISHER; PENNANT'S MARTEN; "PEKAN"; "BLACK CAT"
MARTES PENNANTI (Erxleberi)

Mustela pennanti Erxleben, Syst. Regni Anim., vol. 1, p. 470, 1777 (eastern Canada).

SYNONYMS: Mustela canadensis Schreber, Saugthiere, vol. 3, p. 492, pi. 134, 1778
(eastern Canada); Mustela melanorhyncha Boddaert, Elenchus Anim., p. 88, 1784
(eastern Canada, based on Pennant); Viverra piscator Shaw, General Zool., vol.
1, p. 414, 1800 (eastern Canada, based on Pennant); Mustela nigra Turton,
Tiirton's ed. Linnaeus's Systema Naturae, vol. 1, p. 60, 1806 (eastern Canada,
based on Pennant); Mustela godmani J. B. Fischer, Synopsis Mamm., p. 217, 1829
(Pennsylvania); Mustela canadensis pacifica Rhoads, Trans. Amer. Phil. Soc.,
new ser., vol. 19, p. 435, Sept. 1898 (Lake Keechelus, Kittitas County, Washing-
ton).

A possibly valid race lately distinguished is

Maries pennanti columbiana Goldman, Proc. Biol. Soc. Washington, vol. 48, p. 176,
Nov. 15, 1935 (Stuart Lake, near headwaters of Fraser River, British Columbia).

FIGS.: Nelson, 1916, p. 446. upper fig. (col.); Grinnell, Dixon, and Linsdale, 1937,
vol. 2, pi. 5 (col.); figs. 72-81 (details, skull, habitat).

NORTH AMERICA AND THE WEST INDIES 175

This larger relative of the pine marten has a range in North
America nearly coextensive with that of its smaller cousin and
covers mainly the coniferous forests from eastern Canada and
Maine across to Hudson Bay, southeastern Alaska and British
Columbia, and southward on the higher parts of the mountain
country to Tennessee in the East, the Rocky Mountains to
Wyoming, and the Sierra Nevada in California. Although
Rhoads in 1898 described as a separate race the fisher of the
Pacific coast, Grinnell, Dixon, and Linsdale (1937) have lately
demonstrated that there are no clear distinctions setting off the
animals of that region, but that only a single recognizable form
covers the immense range indicated. Goldman, however,
believes the British Columbian animal is distinguishable. The
fisher is much larger than the pine marten, measuring about
40 inches in length, of which the tail vertebrae comprise about
16 inches. The ears are shorter, hardly projecting above the
head. Females are smaller than males. Males may weigh up
to about 10 pounds, females to about 5.5 pounds. The general
color is ashy brown on the head, gradually darkening posteri-
orly to black on the legs and tail. There may be irregular
white areas on the throat and chest or on the abdomen.

This large marten is highly prized for its handsome and
durable fur and is much trapped and sought after. The pelts
bring a high price. Not many years ago a prime skin might
fetch as much as $100. Fishers, like martens, inhabit the
heavy woods, subsisting on large and small mammals, such as
hares and even the porcupines, which they seem to attack
successfully; squirrels, wood rats, and mice of various kinds
are also taken. Though hunting and traveling much on the
ground, they are also expert climbers and possessed of great
strength and agility; like the marten they have the habit of
passing at intervals over a somewhat definite course, and are
said to go often in pairs. Two or three young are the usual
number, born late in spring (late in May in California).

On account of the value of its fur, the fisher has everywhere
been relentlessly hunted and trapped, yet it is remarkable that
it has persisted in some areas in spite of persecution. In the
East there are still a few fishers left in northern and central
Maine, northern New Hampshire and the White Mountains,
and in northern Vermont and the Green Mountains. Probably
a few are to be found in southwestern New Hampshire in the

176 EXTINCT >AND VANISHING MAMMALS

Mount Monadnock region; at all events, the late Abbott
Thayer obtained one or two specimens there 20 or more years
ago. Formerly there were fishers in western Massachusetts, as
attested by Emmons (1840) a century ago, but it is safe to say
that they are no longer to be found in that State. Nevertheless
they are still taken in some numbers in the Adirondack region
of New York, and according to the 1939 report of the Biological
Survey that State leads in the total number of fishers taken in
1938, no less than 117, exceeding even Alaska! Dr. Francis
Harper (1929) quotes from the reports of the New York State
Conservation Department the numbers of fishers caught, pre-
sumably all in the Adirondack counties, for the years 1920-25,
as follows: for 1920, 132; for 1921, 186; for 1922, 563; for 1923,
112; for 1924, 144; for 1925, 61. The total for 1922 of 563
animals seems very remarkable. Apparently in spite of these
large catches the animals are still holding their own. This
area must be one of the most favorable for them in the country.
In Pennsylvania the fisher was formerly found in some num-
bers, especially in the beech forests of the more mountainous
portions, but it had much declined by the middle of the last
century. Rhoads (1903) has assembled many notes on its
former occurrence, most of which attest its presence in the
fifties and seventies, with a few instances of its more recent
capture, as in Lancaster County in 1896; but most of the re-
ports speak of it as gone by the beginning of the present cen-
tury. In summary he says: "At the present time [1903] about
the only counties where these animals are to be found are
Clearfield, Cameron, Elk and probably Clinton, Potter, and
Sullivan, and in all of these they are reported to be very rare."
Reports of fur dealers, Rhoads further states, indicate that
probably not over half a dozen fishers were annually killed in
Pennsylvania at that time. In a more recent survey of the
mammals of Pennsylvania by Williams (1930) the animal is
not mentioned at all. Southward of Pennsylvania the fisher in
former times extended along the Alleghenies of West Virginia
to North Carolina and Tennessee, but seems to have long ago
been extirpated from these regions. Dr. Remington Kellogg
(1937, 1939) states that the authentic records for West Vir-
ginia are few; according to Surber (1909) it formerly occurred
in the black-spruce region, but F. E. Brooks (1911) gives no
specific records later than 1873. In 1888 Dr. C. Hart Merriam,

NORTH AMERICA AND THE WEST INDIES 177

after a journey of several hundred miles through "the Great
Smoky Mountains of Tennessee and North Carolina, reported
that the pekan was unknown" to local residents, but "reliable
information exists that this animal formerly occurred in that
area" (Kellogg, 1939). Audubon and Bachman (1846) mention
skins from eastern Tennessee and the capture of an individual
near Flat Rock.

West of the Alleghenies the fisher seems to have been rare
even in early days, south of the Great Lakes. Bray ton (1882)
includes it in his list of Ohio mammals but cites no specific
records. For Indiana, Dr. Lyon (1936) states that there is
considerable evidence of its former presence, "either as a regu-
lar resident or as a wanderer from the north," but specific
instances are perhaps three, which he gives as (1) the statement
of Plummer in 1844 that it was not uncommon at an earlier
period, but had not been seen since 1820; (2) the listing of a
skin purchased of local trappers near Noblesville in 1859; and
(3) the testimony of Wied as to its former occurrence at New
Harmony. Evidently it has been gone from this region for
well nigh a century. Cory in ^912 believed it by no means
improbable "that a few individuals may still exist in some of the
extreme northern counties" of Wisconsin, and he was informed
"that it is occasionally taken in the wilder portions of the
Michigan peninsula." He instanced three from the latter
region, taken in Iron County in 1898 and 1900. The latest
record for this region that he gives is of one killed in November
1900, between Iron Mountain and Pembine, Wis. Kennicott,
half a century before, mentioned it as rather frequently seen
in the heavy timber along Lake Michigan. Evidently it is now
practically .gone from the lake region. As to its present status
still farther northwest, along the border of Minnesota and the
adjacent Quetico Park, Canada, Cahn (1937) writes: "The re-
moval of the large timber has profoundly affected the distribu-
tion of the fisher, as it has that of the marten, and it is now to
be regarded as extremely rare in the Quetico, although some
still are to be found there ... I have in my collection
17 skulls of fisher which I collected from about the camps of
trappers up the Wawiag River. " In former days the range of
the fisher extended to northeastern North Dakota, where
judged from the data presented by Bailey (1926) it was fairly
common at the beginning of the last century, for Alexander

178 EXTINCT -AND VANISHING MAMMALS

Henry records a total of 1,118 skins brought in from the sur-
rounding region during the years 1801-7. In 1833 Maximilian
zu Wied stated that the approximate number of skins annually
brought in to Fort Union (now Buford), N. Dak., was 500 or
600, but a large part of these doubtless came from still farther
west. "At the present time, " Bailey says, "there are certainly
no fishers within the State and there seem to be no authentic
records of their occurrence since the early trapping days. "

In eastern Canada the numbers of fishers have been sadly
depleted in recent years. Although formerly found in Nova
Scotia, there seems to be no evidence of its presence there since
Gilpin's time (1868) when he included it among the mammals
of that province. In New Brunswick a few are found in re-
moter parts, as in the mountains of the Gaspe region, whence
Goodwin in 1924 records a few taken by local trappers. In
Quebec it is now "one of the most rare and valuable fur-bearing
mammals of Canada . . . According to Low (1889) the
fisher does not occur east of Mingan nor north of Mistassini
[central Labrador]. It is still found in small numbers around
the southern end of James Bay but is extremely rare near any
white settlements. With a consistently high price paid for its
fur, the fisher is pursued relentlessly and has steadily decreased
in numbers in all parts of its former range" (Anderson, 1939a).
West of Hudson Bay Preble (1902) found it "sparingly through-
out the southern part" of Keewatin and saw many skins at
Norway House and Oxford House. Near York Factory it was
rare, but "farther south more are taken; about thirty or forty
are annually traded at Trout Lake and a few at Severn River. "

The same author (Preble, 1908) states that in the Athabaska
and Mackenzie region it is found throughout the country north
to Great Slave Lake and Liard River. "It seems to be no-
where abundant, and becomes rare toward the northern limits
of its range. Along the Athabaska and Slave rivers a limited
number of skins are collected at all the posts north to Fort Reso-
lution, and I was informed that the animal was rather common
in the region about the mouth of Peace River." Although
according to Osgood (1900), "there is little doubt that the
animal occurs along the upper Yukon" no Alaskan specimens
could be adduced in proof, nor does Coues (1877) mention any
localities in his statement that its range includes Alaska.

Fishers apparently are present in some numbers in the forests

NORTH AMERICA AND THE WEST INDIES 179

of British Columbia, and Goldman (1935) has recently distin-
guished as Maries pennanti columbiana the animal from the
headwaters of Fraser River. The species seems to be absent
from Vancouver Island but is found in small numbers in the
States of Washington and Oregon, particularly in the national
forest reserves. Scheffer (1938), in some recent notes on the
species in the State of Washington, says that it occurs in the
Cascade and Olympic Mountains and on the Olympic peninsula
north to Skagit Valley and Cascade River. In 1937, however,
the U. S. Forest Service estimated that the total number in
the national forests of Washington was about 230 individuals.
In Oregon fishers are found in the cool humid coast ranges, and
probably also in the Blue Mountains section (Bailey, 1936),
but the number taken for fur seems to be small. Thus in the
trapping season of 1913-14, nine fishers were reported to the
State Game Commission by registered trappers as taken in
Washington. At that time skins were quoted at $25 each, but
in 1920-25 the price advanced to as much as $100 to $150
apiece! Fewer numbers are found in the northern Rocky
Mountain region. Bailey (191$), writing of the mammals of
Glacier National Park in Montana, speaks of it as rare, but the
few remaining in the park are now, with proper protection,
"likely to hold their own" for some time to come. "In many
years of trapping in the park region in the early days, Walter S.
Gibb has caught three of these animals, but some of the old
trappers have not secured a skin. Donald Stevenson reports
two skins that were taken by trappers on the Upper Swift cur-
rent in 1910, and tracks which he saw on Swan River and
South Fork as late as 1912. " Merriam in 1891 reported fishers
as already rare in Idaho and mentions a large adult male caught
the previous year near Alturas Lake that weighed 10 pounds 2
ounces. Formerly at least it extended its range to Wyoming,
but no recent report of its presence there is available, nor is
there according to Cary (1911) any evidence of its being found
in Colorado, although in 1874 it was said to occur (J. A. Allen).
Grinnell, Dixon, and Linsdale (1937) have given an excellent
account of the fisher, its habits and range, as known at the
present time in California. It is found still "in the north-
western part of the State south from the Oregon line to Lake
and Marin Counties and east to and including Mount Shasta;
not often in the immediate coastal region (redwood belt) nor,

180

EXTINCT AND VANISHING MAMMALS

so far as known at present, in the Warner Mountains, Modoc
County; south from Mount Shasta and Lassen Peak through-
out the main Sierra Nevada to Greenhorn Mountain in north
central Kern County. " It is found mainly at middle altitudes,
from 2,000 to 5,000 feet in northern California and 4,000 to
8,000 feet in the Mount Whitney region to the southward. At
the present time the fisher is found in the Sierra Nevada mainly
on the west side of the range. On the coast range it formerly
reached Marin County as the southward extension of its area
of distribution. Fishers are nowhere abundant in California,
and at best "it is unusual to find more than one or two to the
township" in suitable country. These authors believe "it is
doubtful if there is one fisher to each 100 square miles" of its
range in California, and in 1926 estimated the entire population
in the State at about 300. Forty years before 1909 they were
found all along the ridges of Mendocino County, but owing to
excessive trapping very few were then left. From 1920 to 1924
the annual catch dropped from 102 to 34, or a loss of 67 percent
in five years. This decrease has involved the entire State and
is not merely local. Since 1925 the fisher has received total
protection in the national parks of California. The authors
mentioned believe that a long-term closed season is much need-

Fisher (Martes pennanti]

NORTH AMERICA AND THE WEST INDIES 181

ed if the species is to continue in the State. They put forth as
major causes in the decrease of the species the following facts :
(1) The fisher is by nature a solitary animal; (2) it requires a
large area of forage territory in order to live ; (3) the areas suit-
able for it are limited ; (4) the rate of reproduction is relatively
low; (5) the forests on which it depends are being constantly
reduced. To these factors may be added the hazard of forest
fires, and, finally, the fact that the gestation period is 11
months, so that trapping operations in winter, through de-
stroying pregnant females, inevitably double or triple the
potential loss of individuals.

The beauty and rarity of fisher fur as well as the high price
it brings make it an attractive proposition for breeding under
artificial conditions. Ashbrook (1928) points out, however,
that the breeding of fishers in captivity for fur has not expanded
to any great extent since 1912, when there were but two fisher
"farms " on this continent. The difficulty seems to lie in getting
them to breed under these conditions even with seemingly
proper food. In the case of an animal naturally so shy and
nervous, breeding seems to be accomplished with difficulty
where there is the least disturbance. At present "little authen-
tic information is available on handling fishers in captivity,"
but no doubt the difficulties will in time be solved.

Dr. Francis Harper has contributed notes as to the number
of fisher pelts yielded in recent years. Thus in the London fur
market in 1927 one dealer offered 1,700 skins; at another sale
2,729 were offered (Journ. Soc. Preservation Fauna Emp., pt.
9, p. 75, 1929). According to Cameron (1936) the fur yield of
the Northwest Territories, Canada, included in 1934 the pelts
of 21 fishers.

SHELL-HEAP MINK; "SEA MINK"
MUSTELA MACRODON (Prentiss)

Lutreola macrodon Prentiss, Proc. U. S. Nat. Mus., vol. 26, p. 887, July 6, 1903
("Brooklin, Hancock County, Maine," from Indian shell-heaps).

SYNONYM: Lutreola vison antiquus Loomis, Amer. Journ. Sci., ser. 4, vol. 31, p. 228,
Mar., 1911 ("Flagg Island, Casco Bay, Maine").

FIGS.: Prentiss, 1903, fig. a (rostrum); Loomis, 1911, figs. 1, 2 (rostrum and jaws).

Remains of a very large mink occur commonly in Indian
shell-heaps along the shores of the mainland and islands of the
coast of Maine, as independently described by Prentiss in 1903

182 EXTINCT AND VANISHING MAMMALS

and by Loomis eight years later. The upper tooth row (front
of middle incisor to back of molar) in the type specimen meas-
ured 30 mm., and in Loomis 's type 29.5, which is larger even
than in the large Alaskan mink. Rostra are infrequent among
the shell-heap fragments obtained, but jaws are fairly common
in kitchen middens along the coast and islands of Penobscot
Bay and Mount Desert Island. Two sizes, evidently repre-
senting males and females, occur.

No other type of mink is found in these Indian kitchen
middens, and since there is an absence of European artifacts
it is believed that these shell-heaps date from previous to, or
perhaps up to, the time of European occupation. Remains are
at present known from such sites as far as Casco Bay in the
south, and northeastward to Mount Desert and Frenchmans
Bay, and Roques Island, Washington County, Maine. Soon
after the description of these bony fragments, the late Manly
Hardy published a note (Forest and Stream, vol. 61, p. 125,
1903) recalling that his father, himself, and other fur-buyers
had in former years "recognized as a distinct form a very large
mink, skins of which were received from islands and coast of
Maine until about 1860. By reason of their large size they com-
manded a higher price than did the skins of other minks . . .
To this animal the name of 'sea mink' was given by the fur buy-
ers" of Maine. "According to Hardy, the animal was large,
very fat, and possessed of an odor entirely different from that
of" the smaller inland animal (see Norton, 1930). Furthermore,
the fur was said to be coarser and of a more reddish color than
that of the latter. There seems no reason to doubt that the
sea mink is the one that inhabited the Maine coasts in Indian
days and persisted up to the latter part of the last century.
With a limited range and on account of eager pursuit by the
trappers, it probably became extinct about the sixties or seven-
ties. Norton (1930) tells of a very large mink skin handled
about 1880 by fur buyers at Jonesport, Maine, which without
the tail measured over 26 inches long and may have been one
of the last of this race. It was taken on one of the islands in the
township of Jonesport. Norton mentions also a large mounted
specimen of a mink taken at Campobello Island, New Bruns-
wick, about 1894, and then in the collection of Clarence H.
Clark, of Lubec, Maine. He speaks of its light color (possibly
a result of fading) and regards it as "probably of this species. "
No measurements are given.

NORTH AMERICA AND THE WEST INDIES 183

It seems probable that this big animal was the only form of
mink to be found in the eastern part of the Gulf of Maine in
earlier times. That it also ranged to Nova Scotia seems likely
from the account by J. B. Gilpin (1867) of large skins from that
region that measured as much as 32.5 inches in total length,
though perhaps slightly stretched. At the present day the
mink occurring along the Maine coast represents Mustela
vison mink, the more southern race, of which a large male will
seldom measure over 23 inches. Evidently this latter race,
which has somewhat of a liking for seacoasts, has taken the
place formerly held by the now extinct sea mink. Possibly,
too, circumstances favorable to this eastward spread within
the last century contributed to the driving out of the larger
animal.

BLACK-FOOTED FERRET

MUSTELA NIGRIPES (Audubon and Bachman)

Putorius nigripes Audubon and Bachman, Quadrupeds of North America, vol. 2, p.

297, 1851 (Fort Laramie, Laramie County, Wyoming).
FIGS.: Audubon and Bachman, 1851, pi. 93; Coues, 1877, pi. 7 (skull) ; Nelson, 1918, p.

449, lower fig., col.

This ferret is of special interest on account of its being in
North America the only representative of the Old World group
of black-bellied weasels to which the polecats and their rela-
tives of eastern and northern Asia belong. On account of its
angular heavy skull and robust premolar teeth, as well as the
color pattern, it is placed in a special subgenus with these,
Putorius, sometimes regarded as a distinct genus. The Ameri-
can representative, however, has only the feet and facial mask
and the end of the tail blackish ; the upper parts are elsewhere
creamy, with a wash of brown over the back, the lower surfaces
whitish. An adult male is about 20.5 inches (529 mm.) long;
the tail relatively short, about 5 inches (130 mm.) (Osgood).

The black-footed ferret is an animal of the interior plains of
North America, east of the Rocky Mountain foothills, from
western North Dakota to northern Montana and Alberta (?)
and thence southward to Texas and central New Mexico. Its
range is practically coextensive with that of the prairie-dogs,
in the colonies of which it lives and upon which it preys. Its
relations with these little burrowers is thus a close one and is
like that obtaining in Mongolia between its relative the black-

184 EXTINCT AND VANISHING MAMMALS

bellied weasel, Mustela eversmanni, and the bobac marmot in
the "towns" of which it lives, preying on these rodents. The
black-footed ferret is believed to rely chiefly upon the prairie-
dogs for sustenance as an abundant and dependable source of
food, yet it doubtless captures other small animals, and
Roosevelt has even recorded a case of one killed in the act of
attacking a small antelope fawn. It is surprising that in view
of the abundance and accessibility of its natural prey, the
prairie-dogs, this ferret should be nevertheless relatively rare,
for specimens are rather uncommon in collections, and indi-
viduals are seldom seen in life even by those living in its habitat.
Bailey (1926) has even suggested "that this very abundance
[of the prairie-dogs] has in some way pauperized the species
until reproduction is restricted," but adds that "apparently
nothing is known of the breeding habits or of the number of
young at a birth. " Since there are three pairs of mammae the
size of a litter probably does not exceed this number and may
normally be less. It is obvious that in a case of such specialized
food preference it would be disastrous for the species to become
so common as seriously to deplete the numbers of its host, so
that undoubtedly some sort of adjustment has been evolved,
but its nature remains uncertain. Possibly enemies as yet un-
recognized are a factor, such as rattlesnakes that might devour
the young, or the hibernating habits of the prairie-dogs, al-
though not well marked in the southern parts of the range, may
affect the ferret's abundance. In its habits it is believed to be
in large part nocturnal, but it has been seen active by day.

In view of the damage done to range vegetation or to culti-
vated crops, the hand of the agriculturist is against the prairie-
dog, and of late years poisoning campaigns on a large scale
have been carried on that in some localities have largely
exterminated these rodents. Undoubtedly a further result of
these activities will be to eliminate or at least to reduce con-
siderably the numbers of the weasel so that the reason for its
inclusion among the vanishing mammals is clear.

Writing of North Dakota, the eastward extension of its
range, Bailey (1926) says that a few have been taken in the
western part of the State and adduces four instances in the
years 1910 to 1915; he quotes Jewett, who killed a specimen
near Quinion in 1913, that none of the old residents to whom he
showed it were acquainted with the species. Coues (1877) in

NORTH AMERICA AND THE WEST INDIES 185

his account of the history of this animal mentions single speci-
mens obtained after special enquiries, in Wyoming, Montana,
Colorado, and Kansas. Hibbard (1933) speaks of it as of
former occurrence in the western part of Kansas, from which
probably it is now largely gone. In eastern Colorado it seems
to be best known, though as elsewhere it is rare. Cary (1911)
mentions that it was reported to Streator in 1894 from three
localities in the Arkansas Valley; Lantz heard of a few in 1905
about Hugo, where they are known as prairie-dog ferrets, and
he had himself heard of it in Baca and Prowers Counties,
" where it seems to be more generally known . . . than in
other sections on the plains." Southward of Colorado black-
footed ferrets "are found over the Staked Plains country of
western Texas and undoubtedly occupy the plains region of all
eastern and northern New Mexico, but few specimens have
been taken" (V. Bailey, 1931) ; a number of specific records are
given for this part of the State, to which Bailey (1905) adds a
few more from east and south of the Staked Plains of Texas.
Although the range may be expected to overlap slightly the
borders of southeastern Arizona and the adjacent parts of
Mexico, where prairie-dogs occur, specific records are not at
hand.

Nelson (1918) is inclined to attribute to its activities the
occasional deserted prairie-dog "towns," which "are not infre-
quently observed on the plains with no apparent reason for the
absence of the habitants." He mentions one case, however,
where a black-footed ferret lived for several days under a
wooden sidewalk in the border town of Hays, Kansas, "where

Black-footed ferret (Mustela nigripes)

186 EXTINCT AND VANISHING MAMMALS

it killed the rats harboring there. " As to its prospects for the
future he writes: "With the occupation of the country and the
inevitable extinction of the prairie-dog over nearly or quite all
of its range, the black-footed ferret is practically certain to
disappear with its host species," a sad prophecy for this re-
markable and useful animal!

WOLVERENE; GLUTTON; "CARCAJOU"
GULO LUSCTJS LUSCUS (Linnaeus)

Ursus luscus Linnaeus, Systema Naturae, ed. 12, vol. 1, p. 71, 1766 (Hudson Bay).
Gulo auduboni Matschie, Sitzb. Ges. Naturf. Freunde Berlin, 1918, p. 153 (Rensselaer

County, New York).

Gulo bairdi Matschie, op. cit., p. 153, 1918 (Fort Union, North Dakota).
FIGS.: Elliot, 1901, pi. 36 (photographs of skull); Nelson, 1916, p. 428 (col.);

Anderson, 1935, p. 87 (photograph).

The wolverene is the largest of the weasel family and an
animal of boreal evergreen forests and barrens. It is repre-
sented by closely similar races in the northern parts of the Old
World, and although several forms have been named from
North America it is very doubtful if their validity can be main-
tained, or if at best some of them are more than slight sub-
species of the Hudson Bay animal.

Audubon and Bachman describe a specimen from New York
State as blackish brown, with a pale reddish brown band ex-
tending from behind the shoulder to the rump on each side.
These two bands join across the rump, and form a contrasting
though not sharply defined light stripe. There is another pale
brownish- white band from the eye to the ear on each side.
The fur is long and rather shaggy. The head and body measure
some two feet nine inches in length, the short tail eight inches.
The animal stands about a foot high at the shoulders, and
weighs 25 to 36 pounds.

Much has been written of the wolverene in the North, of its
great strength in proportion to its size, and of its habit of
breaking into the caches of hunters or trappers and spoiling
what part of the provisions it can not eat. Vernon Bailey
(1926) writes: "Wolverenes are found mainly within timbered
sections of the country, but are great wanderers and at times
may strike out over open country in search of new hunting
grounds. They are omnivorous hunters and scavengers and
have the reputation of being gorging gluttons, a fact which

NORTH AMERICA AND THE WEST INDIES 187

has given them one of their common names . . . They are
said to have from two to four young."

The fur of the wolverene is much in demand among the
Eskimo and other northern people for trimming their parkas,
especially the sides of the hood, for it has the peculiar charac-
teristic that the moisture from one's breath does not readily
freeze and cake on it. Bailey (1936) writes that the fur has
never brought high prices; the quotation for prime skins in
1923 was but $6 to $8, yet at this same time California trappers
received an average of $25 each for wolverene pelts.

The wolverene is an animal of solitudes, usually found singly
and shunning the vicinity of human occupation, so that it
quickly vanishes with the spread of "civilization. " It is there-
fore a species that has already gone from a large part of its
former range in the United States and may be expected to sur-
vive only in the more remote wilderness areas of the North.
Of its occurrence in the northeastern States, very little record
remains. Audubon and Bachman (1846) nearly a century ago,
wrote that they had heard of its existence, "although very
sparingly," in Maine, and abou^t 1810 "we" (probably Bach-
man) had examined three skins in the possession of a merchant
in Lansingburg, N. Y., that were said to have come from the
Green Mountains of Vermont. While these do not constitute
certain evidence of the occurrence of the wolverene in New
England, it is possible that it was formerly found, as these
authors say, "very sparingly. " It should be recalled, however,
that the Canada lynx was sometimes called "wolverene" in
this region, so that hearsay reports have questionable value.
Nor can much credence be placed in a recent (1918) report of
wolverenes in northern New Hampshire (C. F. Jackson, 1922).
Even in New York State the wolverene must have been rare
over a century ago. The only records seem to be those of
Audubon and Bachman (based probably on the latter's obser-
vations) , who give a circumstantial account of the pursuit and
capture of one in Rensselaer County in 1810, the description
and measurements of which are detailed and form the basis of
Matschie's (1918) Gulo " auduboni"', and of a second examined
in 1827 that had been killed near Sacketts Harbor, Jefferson
County, near Lake Ontario. The most southerly record in the
East is that given by Rhoads, of a wolverene killed about 1858
near Great Salt Lick in Portage Township, Pennsylvania.

188 EXTINCT AND VANISHING MAMMALS

Surprising as this seems, Rhoads (1903, p. 165), from his
knowledge of the men who told it to him, had faith in the truth
of this occurrence. The animal was said to have been caught
in a wolf trap. Wolverenes formerly were found in the forests
of New Brunswick but seem to have disappeared quickly.
Chamberlain (1884) mentions in his "Mammals of New
Brunswick" that about 1850 "it was occasionally met with;
but no recent instance of its occurrence is known." In the
Province of Quebec the wolverene "has become extinct or
very rare in the more southern districts, but is found in small
numbers in northern Quebec. It is more commonly found in
wooded districts, but frequently wanders far into the Barren
Grounds of the north and even occasionally to the Arctic
islands" where, Dr. Anderson (1935, 1939) writes, it has been re-
corded from Victoria, King William, and Melville Islands, and
it occurs fairly commonly in the region of Repulse Bay.
Several times it has been taken within 30 miles of Pond Inlet.
"The comparative scarcity of the Wolverene in the southern
half of Baffin Island, and the relative frequency at the northern
end of the island, is good evidence that they come to Baffin
island by way of Melville peninsula. " It still is found in small
numbers in the Labrador Peninsula but probably seldom now
in the southern part where Cartwright found it a century and
a half ago. Audubon and Bachman (1846) mention having
seen specimens "procured at Newfoundland," but it may well
be doubted if they were trapped there. At all events there
seems to be no evidence that wolverenes occurred on that
island within the memory of living men.

Westward along the northern border of the United States
wolverenes were formerly present in small numbers. Cory
(1912) writes that "old trappers living in the vicinity of
Champion, Michigan, claim that wolverines were occasionally
killed in that locality, 30 or 35 years ago" but quotes Edward
G. Kingsford, of Iron Mountain, Mich., who was much in the
woods of northern Michigan, Wisconsin, and Minnesota be-
tween 1880 and 1900, to the effect that the Rainey Lake dis-
trict in the last-named State was the nearest to Michigan he
had ever known of one being killed. Schorger (1939) has
lately unearthed, however, what seems a "satisfactory record"
of one trapped on February 15, 1860, by a man living at
Marquette, Lake Superior, in the northern peninsula of Michi-

NORTH AMERICA AND THE WEST INDIES 189

gan. In earlier times there were a few in Wisconsin, and Hoy
mentions one killed in La Crosse County in 1870. Rather
surprising, however, is the evidence adduced by Hahn (1909)
of a wolverene having been killed in 1840 in Noble County,
Ind., and a second in 1852 in Knox County of the same State.
The records probably mark nearly the southern limit of range
in the east in historic times. In Minnesota, Herrick, writing in
1892, says he has no recent knowledge of it, so that it doubtless
was rare there at that time, but Kingsford, quoted by Cory
(1912), says that "about 1895 to 1897 they were quite plentiful
in northern Minnesota. " More recent records indicate that a
few individuals were found in Minnesota up to a much later
time, for Dr. H. H. T. Jackson (1922) has recorded a skull in
the collection of the U. S. Biological Survey from an animal
killed in Itasca County, January 11, 1899; and C. E. Johnson
(1923), who made special efforts to learn of the wolverene's
former status, records that a fur trader of long experience in
northern Minnesota told him in 1922 that in 30 years of buying
furs in the region he would not average more than one wolver-
ene pelt a year and had bought none during the previous three
years along the Canadian border. He specifically mentions a
pelt bought by this trader that was taken in 1918 in the
northern part of St. Louis County, Minn. It seems to be the
last definite record for that State.

In the early years of the last century these animals were
found at least in the northeastern part of North Dakota and
were taken in small numbers with the fur catch of those days
(see V. Bailey, 1926). It seems to have been present in the
State till about 1850, but there are no later records. According
to Baird (1857) it occurred in "the Black Hills of the Missouri"
at about the same time.

While in the eastern United States and southeastern Canada
the wolverene is doubtless now to be regarded as extinct, it
still is found in some numbers in the wilder and more moun-
tainous parts of the western United States and more frequently
in the Canadian Northwest, although on account of its solitary
habits it can not be accounted a "common" species at any
time. Writing of the mammals of the Athabaska-Mackenzie
district Preble (1908) states that it is found throughout that
region "but is nowhere common, though a few skins are
collected at all the posts we visited. During our trip in 1901

190 EXTINCT AND VANISHING MAMMALS

more skins were seen at Fort Rae than at any other post, and
these had been brought mainly from the Barren Grounds,
where the animal occurs more commonly than in the wooded
country. . . In the semibarren country along the southern
shore of Great Bear Lake, tracks of wolverenes were common.
. . . A large proportion of the wolverene skins which are
obtained from the Indians of the Mackenzie are shipped to
Fort McPherson and traded to the Eskimo, who prize this fur
highly for trimming their clothing." Pike and other travelers
mention it as being met with across the Barren Grounds.
Trappers in the Yukon region secure a few annually. A. J.
Stone (1900) in the Canadian Northwest found them in timber
and on the barrens and mentions seeing them "far out on the
ice of the deep bays along the coast." Osgood (1901) speaks of
them as "apparently rather common" in the Yukon and also
in the Cook Inlet region of Alaska, whence a number of skins
annually are sent via St. Michael to trading posts on the
Yukon.

In British Columbia, on the other hand, wolverenes seem to
be diminishing in numbers. Cowan, in 1939, writes that they
are "rare in the Peace River .district, but according to certain
reports are more plentiful in the Rocky Mountains." He
mentions one trapped on Pine River in 1936. On Vancouver
Island a few wolverenes still exist, according to Swarth (1912),
but are confined to the higher mountains and are seldom
trapped, except by the Indians, who bring in "one or two every
year."

In the Rocky Mountains south of the Canadian border a
small number of these mammals remain. Fifty years or more
ago Merriam (1891) wrote that wolverenes were "tolerably
common in the Saw Tooth Mountains" of Idaho, where in the
previous winter a trapper had caught five. Another had been
taken a few years before 1891 in the Blackfeet Mountains.
No doubt their numbers have since diminished, for Vernon
Bailey in 1918, writing of the mammals of Glacier National
Park in northwestern Montana, says that although in 1895 he
was told by trappers in the St. Mary Lake region that a few
existed there and one was occasionally caught, and that a few
were trapped each year before the park was made a reserve,
there were thought to be none left in the park at the time of
writing. In the high forests of the wilder parts of Colorado

NORTH AMERICA AND THE WEST INDIES 191

wolverenes were still to be found in 1911, but although in
former times generally distributed, they were then apparently
restricted "largely to the San Juan and La Plata Mountains,
the mountains of northern Gunnison County, and the ranges
surrounding North and Middle Parks" (Cary, 1911). A num-
ber of specific records are given by Cary, which prove that the
species was much commoner in the latter part of the last
century. To the southward there is some evidence that
wolverenes extended into the higher mountain forests of
northern New Mexico. Of this Coues (1877) was assured by
hunters there in 1864, but no definite records exist (Bailey,
1931), and it must have soon been extirpated from this southern
outpost of its range. Confirmatory evidence, however, is
given by Seton (1931), who was told by an Acoma Indian in
1930 of an animal known formerly to the people of that part
of New Mexico as the Ho-ho-an or Ko-ko-an, somewhat like a
small bear with a bushy tail, and the Indian then produced from
his medicine bag a small carving of the creature's head, an
excellent likeness of a wolverene's. He said it "was formerly
found in all these mountains but had disappeared. None of
the present generation had seen one." Baird mentions a
specimen from the basin of the Great Salt Lake in Utah
previous to 1857, but it has long been gone from this region.

The status of wolverenes on the Pacific coast still requires
investigation as to numbers and racial relationships. Elliot
in 1905 distinguished specimens from Mount McKinley,
Alaska (type from Sushitna River), as Gulo hylaeus on the
basis of their darker instead of gray heads and larger audital
bullae as compared with specimens from eastern Canada. It
is perhaps a recognizable race. In 1918, Matschie named as a
distinct species on very slight basis, Gulo katschemakensis, from
the Kenai Peninsula, Alaska, but it is doubtful if this is really
different. In the same paper he described Gulo niediecki from
a skull from Dease Lake, British Columbia, but here again the
validity of the race is questionable. Very likely, however, the
animal of the coastal ranges is a recognizable form, described
by Elliot as Gulo luteus, and is accorded subspecific rank by
Dr. Grinnell (see postea).

Recent information gathered by Dr. Francis Harper indicates
that wolverenes are still taken in some numbers in northwestern
Canada and a few in northeastern Canada. In 1931-32, 536

192 EXTINCT AND VANISHING MAMMALS

wolverene skins were produced, according to returns from
Canada (Journ. Soc. Preserv. Fauna Empire, p. 75, pi. 21,
1934). In 1934 the fur yield of the Northwest Territories in-
cluded 98 wolverenes (Cameron, 1936, p. 625).

In 1936, the Chief of the U. S. Biological Survey urged a
five-year closed season on wolverenes in the United States and
a resolution to the same effect was adopted by the North
American Wildlife Conference at Washington, D. C., shortly
after, in the hope of saving the species from complete extermi-
nation within the United States. That Alaska is nearly the
center of the wolverene's abundance at the present time is
indicated by the fact that according to the report of the
Biological Survey, 248 skins were taken in Alaska in 1938.
No reports of it are given for other States.

WEST COAST WOLVERENE; SOUTHERN WOLVERENE

GULO LTJSCUS LUTEUS Elliot

Gulo luteus Elliot, Field Columbian Mus. Publ., zool. ser., vol. 3, p. 260, Dec., 1903

(27 miles south of Mount Whitney, Tulare County, California).
FIGS.: Grinnell, Dixon, and Linsdale, 1937, vol. 1, pi. 6 (col.), text-figs. 91-97.

An excellent account of the characters and habits of this race
is that given by Grinnell, Dixon, and Linsdale (1937), summing
up present knowledge of the wolverene in California. It was
named by Elliot in 1903 as a separate species differing in its
supposedly paler coloration as compared with eastern animals.
The authors quoted show, however, that this was "merely a
feature of a young stage of pelage" and that in the original
description not a single character given is diagnostic of the
race, which may only be distinguished by skull characters,
namely, "skulls from California are slightly smaller than
skulls of corresponding ages from Hudson Bay, Alberta,
British Columbia, and Alaska, and the dentition is noticeably
and uniformly lighter in particular the carnassials are de-
cidedly less massive. Three examples from Idaho and one
from Utah, in the National Museum, are small-toothed like
the California skulls." A Montana and a Minnesota skull are
said to be of intermediate size, while one from Chelan, Wash.,
is "nearer the average of Sierran luteus"

The range of this race may thus be taken to include the
Sierras of California northward to Washington, merging by

NORTH AMERICA AND THE WEST INDIES 193

imperceptible degrees with that of the typical form in Utah
and Idaho. Although formerly the range was more extensive,
over the higher parts of the California mountains, it has, since
at least about 1900, become restricted to the central and
southern Sierra Nevada above 8,000 feet "from the vicinity of
Lake Tahoe south through the Mount Whitney region" to
Kern County. Though the altitudinal range varies from 5,000
to 13,000 feet, the animals are most frequent at about timber-
line. Probably this restricted range has resulted in making
the population of Californian wolverenes never very large.
The first to mention the species in this State was Cooper,
who in 1868 wrote that at that time a few were killed every
winter in the northern Sierras. There is some evidence, ad-
duced by the above authors, that in the middle of the last
century wolverenes occurred in the coastal mountains of Marin
and Sonoma Counties, but no specimens are extant. Slightly
farther north, however, these animals seem to have occurred
down to about 1873, as on the higher peaks of Mount San-
hedrin, Mount Linn, and Yolla Bolly. Merriam is quoted
concerning a specimen killed ^Jbout 1893 between Mounts
Shasta and Lassen.

At the present time there are probably few wolverenes left in
California. "From 1920 through 1924, the average number
reported captured by the trappers of California was less than
three a year. After considering all the available data, the
authors [Grinnell, Dixon, and Linsdale, 1937] believe that at
the present time (1933) there are at most not more than 15
pairs of wolverenes left in the State. This animal is thus one
of the rarest furbearers now living in California." "At present
the main hope" of protecting the species from extinction
within the State is its preservation within the guarded areas of
Yosemite and Sequoia National Parks, but the animals are
likely at times to wander outside the limits of these areas and
are caught or killed by trappers who may legally take fur out-
side the park boundaries.

How far north the characters of the southern wolverene may
hold is tentatively intimated by the above authors as perhaps
extending to the State of Washington. Bailey, in 1936, be-
lieved that they are not yet extinct in the Cascades and
Sierra Nevada and mentions one taken in the upper McKenzie
Valley west of the Three Sisters Peaks, Oregon, in 1912; how-

194 EXTINCT AND VANISHING MAMMALS

ever, at the present time their numbers must be few indeed.
Still later, Scheffer (1938) notes that the U. S. Forest Service
estimates that in the national-forest areas of the State of
Washington, the wolverene population is about 20.

Family CANIDAE: Wolves, Dogs, Foxes
THE KIT FOXES

These small foxes are about half the size of a red fox and
live in the open plains country or sandy deserts of western
North America. The upper side of head and body is some
shade of buff overlain with gray; the sides of neck, shoulders,
and body are clear buff, the ears and legs usually brighter buff,
the tail gray with black-tipped hairs, which terminally con-
centrate to form a black tip. Two species are found, the more
eastern, ranging over the Great Plains, Vulpes velox, and the
more western, centering in the Great Basin and southern
California, V. macrotis. The former is a slightly larger animal
with shorter ears than the latter, and comprises two races; the
latter is more an animal of sandy desert regions and is at
present divided into eight races.

SWIFT; KIT Fox

VULPES VELOX VELOX (Say)

Cam's velox Say, Long's Exped. to Rocky Mts., vol. 1, p. 487, 1823 (South Platte River

(? Logan County), Colorado).
SYNONYM: Canis microtus Reichenbach, Regnum animale . . . , vol. 1, pt. 1, p. 10,

1836, as quoted by Wiegmann in Wiegmann's Archiv fUr Naturg., Jahrg. 3, Bd.

2, p. 162, 1837.
FIGS.: Audubon and Bachman, 1851, The Viviparous Quadrupeds of North America,

vol. 2, pi. 13.

NORTHERN KIT Fox
VULPES VELOX HEBES Merriam

Vulpes velox hebes Merriam, Proc. Biol. Soc. Washington, vol. 15, p. 73, Mar. 22, 1902
(Calgary, Alberta, Canada).

The range of the typical race is from the Staked Plains of
northwestern Texas northward over the Great Plains to prob-
ably South Dakota. The northern race, V. v. hebes, is larger
and slightly grayer, "dark patches on sides of nose darker,
skull larger and heavier; palate much longer; under jaw longer,
heavier, more bellied under sectorial tooth" (Merriam).

NORTH AMERICA AND THE WEST INDIES 195

Average of four males: Total length, 844 mm., tail, 280; foot,
131; ear, about 50. Weight, 4.5 pounds.

"These little foxes," Bailey (1931) writes, "live in burrows
in the open plains country, usually selecting side hills or the
sunny slope of a bank in which to make their dens. They are
mainly nocturnal, are wonderfully swift and graceful in their
motions, but, unlike the red foxes, are so unsuspecting in their
natures as to be readily caught in traps . . . and are so
unable to cope with the advanced civilization that they are
rapidly disappearing from the face of the earth." Gary in 1902
was told that in the Staked Plains of Texas they were at that
time scarce in comparison with former years, for in the winter
they come readily to poisoned bait put out for coyotes and
wolves. This bait the kit foxes will take the first night it is
set out, but coyotes wait till later (V. Bailey, 1905). Formerly
this was a common species over the plains of eastern Colorado,
but by 1911 had "become rare in most sections" (Gary, 1911).
It is found still in the western part of Kansas and has been
taken as far east as Douglas County (Hibbard, 1933). In
1870 Dr. J. A. Allen reported its occurrence "more or less
frequently" in western Iowa, but at the present time it is
probably gone from this eastern outpost of its range.

It is probable that the range of the northern race may be con-
sidered as from about North Dakota north to the Saskatche-
wan River. Bailey (1926) refers those of North Dakota to this
race and states that "apparently the kit foxes of the northern
Plains at one time covered the whole of the prairies of North
Dakota, but at present they are restricted to the western part
of the State, and even there they have become very scarce,
although a century ago they were one of the common fur
animals of the Red River Valley." With the settling of the
country they have easily succumbed to trapping, poisoning,
and capture by dogs, and reports from various localities in that
State indicate that of recent years they have been reduced
nearly to the point of extermination. In northwestern Mon-
tana, Bailey (1918) speaks of this fox as common over the
plains along the eastern edge of Glacier National Park, which
it probably enters in the open area at the lower end of St.
Mary Lake. There is thus apparently no immediate danger
of its extermination in this region, yet on account of its less
wary nature it is likely to suffer gradual reduction with the

196 EXTINCT AND VANISHING MAMMALS

encroachment of settlements and especially with trapping and
poisoning campaigns carried on against coyotes or wolves.
Baird records the usual stomach contents as "fragments of skin,
leather with hair, berries, remnants of mice, and grasshoppers"
so that economically the kit fox is an asset through its destruc-
tion of small rodents and grasshoppers and deserves whatever
protection may be accorded it.

This race of the kit fox formerly extended in range as far
north as the Saskatchewan River in western Canada but is
now gone from many parts of this area. Specimens in the
Museum of Comparative Zoology from Calgary and Buffalo
Lake, Alberta, were received in 1897, and others from Assini-
boia were obtained in the following year. Fowler (1937),
however, writing of changes in the fauna of the High River
district, Alberta, states that it became extinct there shortly
after those dates; "in fact, only those people who were here
before 1900 can recall having seen a kit-fox. My father states
that the last bunch of kit-foxes he saw was in the summer of
1897. These had their den in a sandy knoll about a mile
north of High River."

LONG-EARED KIT Fox; DESERT KIT Fox

VULPES MACROTIS MACROTis Merriam

Vulpes macrotis Merriam, Proc. Biol. Soc. Washington, vol. 4, p. 136, Feb. 18, 1888
("Riverside, Riverside County, California," vicinity of Box Springs, western
margin of San Jacinto Plain, about 10 miles southeast of Riverside).

FIGS. : Grinnell, Dixon, and Linsdale, 1937, pi. 9, bottom fig. in color; and (other races)
figs. 154-162 (skull, live animal, habitat).

The long-eared kit fox resembles V. velox but has longer ears
(about 80 mm.), smaller hind foot, and slightly brighter color-
ing, but otherwise it has a similar appearance. As a species its
range extends from Riverside County in southern California
eastward across the Great Basin into extreme western Texas
and south into Lower California and northern Mexico. Over
this range it divides into a number of slightly marked local
races, of which no less than nine, including the typical form,
are described. These with their type localities are: arizonensis,
from 2 miles south of Tule Tanks near the Mexican border of
Yuma County, Ariz.; arsipus, from Daggett, San Bernardino
County, Calif.; devia, from Llano de Yrais, opposite Magdalena
Island, Lower California; mutica, from Tracy, San Joaquin

NORTH AMERICA AND THE WEST INDIES 197

County, Calif.; neomexicana, from San Andreas Range, Dona
Ana County, N. Mex., about 50 miles north of El Paso;
nevadensis, from Willow Creek Ranch, near Jungo, Humboldt
County, Nev.; tenuirostris, from Trinidad Valley, northwest
base of San Pedro Martir Mountains, Lower California,
Mexico; and zinseri, from San Antonio de Jaral, southeastern
Coahuila, Mexico. In a recent study of the group, Benson
(1938) shows that arizonensis is synonymous with the older-
described race arsipus, so that the number of recognized races
is now eight. This author writes that "kit foxes are now as a
rule few in number even in the areas which seem suitable to
them. They are so unsuspicious that they are easily trapped
and even more easily poisoned. Consequently, wherever
trappers are active, and especially wherever control campaigns
involving the use of poison have been carried out against
predatory animals on areas inhabited by kit foxes, the foxes
have been greatly reduced in number or entirely eliminated.
Unfortunately, there are few kit fox habitats in the United
States that escape visitation by these agencies." Benson
comments that the race mutica pf the San Joaquin Valley and
Walker Basin in California is a well-marked form, while the
race arsipus, ranging east of the Pacific slope drainage, though
also easily distinguishable, clearly extends into southern
Nevada and may have to include nevadensis as a synonym.

The numbers of all these various races seem to be rapidly
decreasing, but the only one which is rather certainly now
extinct is the typical one, V. macrotis macrotis. This form
once ranged over parts of California in the "San Diegan sub-
faunal district, from Riverside County, northwest ... to
Los Angeles County" in the Lower Sonoran Life Zone. It
formerly occurred in the Allessandro, Perris, and San Jacinto
Valleys, westward along the plain at the southwest of the San
Gabriel Range to San Fernando Valley, where in the early
nineties local residents trapped them. The last one known to
have been taken in the San Jacinto Plain was trapped near
Moreno in 1903 (Grinnell, Dixon, and Linsdale, 1937). At the
present time the race mutica is still to be found in the San
Joaquin Valley of California, but the authors just quoted
indicate that its range has shrunk to about half of what it was
not so many years ago, so that it is now restricted to the driest
plains of the southern and western parts of the valley.

198 EXTINCT AND VANISHING MAMMALS

"Open level sandy ground is the preferred habitat of the
kit fox," for here it can easily excavate its burrows, and this
type of ground is likewise preferred by numerous small rodents,
such as kangaroo rats and pocket mice, which form the chief
food of this little fox. " These special adaptations may account
for the rather sharp limitation observable in the general dis-
tribution of this type of fox, which man's presence further
restricts." Their shallow burrows are dug in open ground and
may have one to four entrances. Grinnell, Dixon, and Lins-
dale (1937), from whose account these details are taken, found
that the breeding season is early. Young are found newly born
late in February, and the number to the litter may average four
or five but in one case was as many as seven. It seems to have
few natural enemies. There is a close correlation between the
areas inhabited by the kit fox and the range of the large kanga-
roo rats Dipodomys deserti and D. spectabilis, on which with
other smaller rodents the fox largely feeds, although insects,
occasionally rabbits and snakes, and rarely poultry are also
items in its diet.

Of the race mutica of the San Joaquin Valley, Calif., the
authors quoted say that of recent years large numbers have
been killed for fur. Thus, in 1919, one trapper caught 100
foxes in a single week on an area 20 miles long by 2 miles wide
on the west side of the valley. But "probably the greatest
number of kit foxes has been destroyed by poison campaigns
directed against coyotes . . . Hundreds of kit foxes were

Long-eared kit fox (Vulpes macrotis macrotis)

NORTH AMERICA AND THE WEST INDIES 199

doubtless thus destroyed in a single season." Some unusual
condition added to these efforts at "control" might conceivably
wipe out the remnant of these foxes in the San Joaquin Valley.
However, "in the Mohave and Colorado deserts, which are now
remote from human activity and where the conditions of the
soil and the lack of water will long prevent future settlement,
it is probable that" the race arsipus will endure indefinitely.

The most northerly ranging race of the desert fox is neva-
densis, concerning which Vernon Bailey (1936) writes: "These
graceful little foxes have been found in the valleys of northern
Nevada, southwestern Idaho, and southeastern Oregon, not
far beyond the limits of Lower Sonoran Zone valleys, where
the species generally ranges." Specimens are mentioned from
the Owyhee Valley as the only ones available from Oregon,
but it "may be looked for in the Alvord and adjoining valleys
that open out into northern Nevada." He says nothing of
their present status there but surmises that they will soon be
gone. He further testifies to their adaptability as pets, on
account of their gentleness, their intelligence and affectionate
behavior, and their ready confidence and alertness. "Why
not," he asks, "keep such animals instead of cats and dogs and
save a few from extermination?"

THE AMERICAN WOLVES

Wolves are of boreal distribution in both the Old World and
the New, originally ranging without interruption from western
Europe eastward to the Pacific Ocean in the former and from
Alaska to the Atlantic coast of eastern North America in the
latter area. Pocock (1935) in a recent paper regards them as
of a single species with local races, which appears to be a
logical course, since all are much alike. Even the narrow
Bering Straits, closed by ice during part of the year, probably
form no insuperable barrier to an occasional crossing. In
North America wolves are found from the Arctic regions
southward to the Gulf of Mexico and the Mexican tableland,
but in a recent review Goldman (1937) regards the most
southern or red wolf as a separate species. Exclusive of this
and its two subspecies, there are at present recognized no less
than 20 local races in this area, distinguished for the most part
by minor characters of the skull and teeth; all these are treated
as subspecies of Canis lupus of Europe. It is often said that

200 EXTINCT AND VANISHING MAMMALS

the European wolf is the ancestor of the domestic dog, but the
evidence for this is merely inferential. Usually wolves of any
race may be distinguished from even the largest dogs by their
proportionally larger teeth, shorter ears, and less elevated
forehead due to the smaller development of the frontal sinuses.
In dogs the length of the lower first molar or carnassial tooth
is usually less (and in medium-sized and smaller breeds much
less) than 22 mm., whereas in wolves it is in excess of this
measurement. In their general proportions and a certain
"plumpness," however, the teeth of wolves are exceedingly
like those of domestic dogs and differ from the narrower more
bladelike teeth of coyotes, which some have thought ancestral
to dogs.

Everywhere in North America that wolves have come into
competition with white men, in settlements or in agricultural
communities, they have been regarded as a common enemy
and killed whenever possible. The inevitable result has been
their greater or less reduction in the thinly settled regions and
complete extermination from larger areas where white popula-
tion is denser or agricultural and pastoral use of the land makes
it impossible to tolerate their presence on account of their
depredations against livestock. While this is more or less to be
deplored, it is unavoidable, for the rancher can not be expected
to view with complacency the resulting loss. The outlook for
the preservation of a remnant of this interesting species is
therefore bright only in the wilder and more inaccessible
regions or in large public reservations where the numbers can
be controlled and where, conceivably, a small number of
wolves might serve a useful purpose not only in preventing
too great an increase in wild deer or mountain sheep but also
in actually improving or keeping up a standard of quality by
eliminating sickly or stupid individuals. It is remarkable that
even in parts of Europe to this day as in Spain and France a
few wolves have remained in spite of long occupation by
white races. Furthermore, if one may believe the mass of
published accounts, wolves under the stress of hunger have
been known to pursue and kill human beings in Europe and
Asia, whereas at least in eastern North America there seem to
be few or no authentic cases of such bold actions on the part of
wolves here in early days, although there are instances of
wolves having followed and frightened many of our early
settlers.

NORTH AMERICA AND THE WEST INDIES 201

Like other large carnivores, wolves require food in larger
amounts, so that in addition to what smaller game they may
pick up, it is the larger ungulates that they specially hunt.
In the North the caribou, and on the plains the mule deer, and
even the young or sickly bison were objects of pursuit, while in
the wooded regions of eastern North America the white-tailed
deer was a staple food species. In the case of the last-named
the wolf in areas of deep winter snows was doubtless a main
factor in preventing the spread of this deer to the northward
into regions where now it is abundant.

In recent years a more intensive campaign has been carried
on under the U. S. Biological Survey to eliminate the wolf
entirely from stock-raising regions of the West. This has
resulted in the accumulation by the Survey of sufficient speci-
mens to "afford a fairly satisfactory basis for determining
specific and subspecific relationships" of the American races,
as worked out in Goldman's (1937) paper. Some of these are
already extinct, others nearly so. In the following brief survey,
the races are taken up alphabetically, giving a condensed
statement of the general status so far as at present known. A
final account of the genus and its natural history is to be
desired.

In a general way, the various races of wolves are much alike,
about the size of a modern "police dog" but shaggier, with
short ears and bushy tail, carried usually with a slight hump
near the root. The usual color is a mixed black, white, and
gray, giving a grizzled appearance. The long hairs of the coat
are usually white for the greater part of their length, with
more or less black tipping or white, producing this mixed
appearance. Often the white portion of the hairs is greater
than the black tipping, resulting in a pale coat, and in the
Arctic parts of the range the black may be eliminated almost
altogether, so that the pelage is white. On the other hand,
black may predominate and the pelage in that case is blackish
to entirely black. This wide variation may occur within the
same family group of animals, so that color characters are
often of little avail for the separation of geographic races.
These are therefore in large part based on cranial characters
and on size differences where they are significant. While
wolves must often travel over considerable areas in their
hunting, and thus tend to segregate less than more sedentary

202 EXTINCT AND VANISHING MAMMALS

species, they must nevertheless keep to more or less definite
regions where under varying conditions of environment slight
differences have developed. Intergradation through individual
variation often makes the exact definition of races difficult, and
series are needed in working out the status of animals from
different parts of the range.

According to a useful pamphlet by Vernon Bailey (1907)
wolves enter very little into the national forests of the western
United States but favor foothill regions for breeding and
hunting. In some areas there is evidence of a certain amount of
seasonal migration, in the course of which the wolves of a dis-
trict follow the cattle herds into the mountains in spring, and
again in autumn accompany them into lowland areas, much as
in former times they followed the bison. The same paper has
some account of the abundance of wolves in the West and
South up to the early years of this century, with instructions
for trapping and poisoning.

AMERICAN ARCTIC WOLF

CANIS LUPUS ARCTOS Pocock

Canis lupus arctos Pocock, Proc. Zool. Soc. London, 1935, p. 682, Sept. 12 ("Melville

Island, Arctic America").
FIG.: Nelson, 1916, p. 421.

According to Pocock the skull of a specimen of Arctic wolf
from Melville Island "may be distinguished at a glance by the
great elevation of the frontal region and the resulting much
more strongly convex curvature of the frontorostral line. It is
also broader . . . in the muzzle and palate . . . The
bulla also is noticeably more inflated" as compared with a
skull regarded as representing C. I. tundrarum. This is a
white wolf.

In addition to the skull of the type from Melville Island,
Pocock lists a second in the British Museum from Discovery
Bay, Ellesmere Land. The former was secured in 1853-54, the
second about 1878. Parry seems to have been the first to men-
tion this wolf, which he noticed was smaller than the white
wolf of the mainland (tundrarum) .

The present status of this wolf in Ellesmere Land and ad-
joining islands to the west is uncertain. There are few human
inhabitants, either Eskimo or white. However, since it would

NORTH AMERICA AND THE WEST INDIES 203

be expected that the numbers of caribou in the region would
attract a small permanent population of wolves, it seems
strange that they should be so rarely observed. Peary, who
was familiar with the region from his various polar expeditions,
wrote ("The North Pole," p. 61, 1910) that in the northern
part of the island there might be "perhaps once in a generation
a stray wolf," implying that it was almost unknown a genera-
tion ago. Had it been of regular occurrence following the
caribou herds, the Eskimo would surely have reported it.
Possibly these stray Arctic wolves may have come from farther
south on rare occasions. However, the more recent testimony
of Macmillan is rather at variance with Peary's statements.
In his book "Four Years in the White North" (1918) he has a
good deal to say about the Arctic wolves met with in Ellesmere
Land and Axel Heiberg Land. Late in March, 1916, two were
started at Bay Fiord, near Etah, but they dashed away and
finally disappeared over the rough ice of Smith Sound; later
one was killed near Eureka Sound, and another was started
from the ice foot and also ran off over the ice. "Axel Heiberg
Land is infested with them," he^wrote, after seeing a pack of 12,
all white. In April of that year two others appeared and when
pursued by the sledge dogs ran off over the sea ice toward
North Cornwall. In October, 1916, the local Eskimos returned
from their annual caribou hunt, which covers the region from
Etah to the Humboldt Glacier on the Greenland coast. They
reported no young caribou, but tracks of wolves everywhere,
implying that the young caribou find it difficult to survive on
account of the wolves. Macmillan believes that a large band of
white wolves had crossed Smith Sound and was following the
herds on the North Greenland coast. They also probably
capture a few seals along the ice foot, which would account for
their presence in that place, while muskoxen too must fre-
quently fall a prey. If this explorer is correct in his belief that
wolves cross Smith Sound, from Ellesmere Land to North
Greenland, as seems likely, there appears no reason to suppose
that their wanderings in search of game might not take them
across the northern part of that country or even to the east
coast, so that the supposed distinction of the wolves of the
two regions may prove to be unfounded.

204 EXTINCT AND VANISHING MAMMALS

MEXICAN WOLF
CANIS LUPUS BAILEYI Nelson and Goldman

Canis nubilus baileyi Nelson and Goldman, Journ. Mamm., vol. 10, p. 165, May, 1929
("Colonia Garcia (about 60 miles southwest of Casas Grandes), Chihuahua,
Mexico").

FIG.: Bailey, 1931, pi. 18 (photograph).

This is the wolf of "southern and western Arizona, southern
New Mexico, and the Sierra Madre and adjoining tableland of
Mexico as far south, at least, as southern Durango," and ac-
cording to its describers it is most nearly allied to the plains
wolf (C. I. nubilus) but is smaller and darker, more brownish or
rufescent, with a more slender and depressed rostrum. Its
greater size and other cranial characters indicate "no close
relationship" to C. rufus. Total length, 1,570 mm.; tail, 410;
skull, condylobasal length, 262.

The authors quoted list various specimens from Chihuahua,
one from Sonora, and one from Durango, Mexico, as well as a
number from Arizona and New Mexico. "According to old
residents wolves that formerly inhabited the southern end of
the tableland, near the valley of Mexico, became extinct many
years ago. Wolves are still numerous, however, in the Sierra
Madre as far south at least as Durango." Specimens from
southwestern New Mexico show indications of intergradation
toward the plains wolf (Bailey, 1931, p. 303). "Their greatest
abundance has long been in the open grazing country of the
Gila National Forest" in the open yellow-pine forests and the
orchardlike growth of juniper, nut pine, and oak. "Apparently
they are not known in the Lower Sonoran valleys of this
region." Bailey (1931) has given an excellent account of the
habits of wolves in this area. He states that they persistently
cling to certain localities in their range for raising their young.
"Dens are always placed in the most out-of-the-way places,
with seemingly little regard for convenience as to water or
food." In rough country the dens are generally in a crevice or
natural cavity "usually in the rim of a mesa"; badger holes
are often enlarged, or the burrow may be worked out beneath
a mesquite or other bush that serves as a blind.

Bailey (1931) tells further of trapping wolves in this region
and records that J. S. Ligon with a corps of trappers began a
trap line in 1916 for the purpose of reducing the number of

NORTH AMERICA AND THE WEST INDIES 205

wolves. He estimated that in May, 1917, there were 103
adult wolves in New Mexico and 45 in July, 1918. "In 1927
he said that they were practically eliminated from New Mexico
and that he was then concerned only with reinfestation from
Mexico." Thus it appears that continued and skillful effort
will at length succeed in exterminating even these wily animals,
but that with their tendency to range wide and far, there is a
constant opportunity for new individuals to come in from
outside.

NEWFOUNDLAND WOLF

CANIS LUPUS BEOTHUCUS G. M. Allen and T. Barbour

Cam's lupus beothucus G. M. Allen and T. Barbour, Journ. Mamm., vol. 18, p. 230,

May, 1937 ("Newfoundland").
FIG.: Allen and Barbour, 1937, p. 231, fig. (upper carnassial).

In the Newfoundland wolf the skull is relatively slender,
and the nasal bones are long in contrast to the short ascending
branches of the premaxillaries ; the upper carnassial is char-
acteristic in its shortness and in the size and sharp inward turn
of the anterointernal cusp in contrast to the condition in other
wolves. The color is white, but occasional black individuals
are said to have been known. The largest of four skulls
measured in greatest length, 276 mm.; outer length of nasals,
103; tip of nasal to point of premaxillary, 47; length of upper
carnassial, 25.5; length of first lower molar, 28.7.

In the large size and slender proportions of the skull, as well
as in its white coat, this wolf resembled the white Arctic wolf.
At the present time it is apparently extinct in Newfoundland,
where in former days it followed the herds of Newfoundland
caribou. Bonnycastle in 1842 spoke of it as large in size and
as frequently seen in the neighborhood of St. John's. On
account of its depredations against young cattle, a bounty was
declared on wolves. A writer in Forest and Stream (vol. 4, p.
390) in 1875 spoke of these wolves as then common in the
island, often prowling near the houses of settlers or seen near
settlements in chase. In winter when the caribou were feeding
in the bogs, the wolves cooperated in hunting them. While
some lay in wait along the border of coniferous trees, others
went around to windward, when the caribou, scenting them,
turned and ran directly into the ambush. The same writer
mentions an instance where a Micmac Indian in April, 1866,

206 EXTINCT AND VANISHING MAMMALS

was standing by the side of a small lake formed by the River
Exploits, when he saw an old wolf coming toward him on the
ice, to be followed presently by five or six more. The Indian
ran for his "tilt" and his gun, but the wolves, following him,
gained so rapidly that he sought safety by climbing into a tree
beyond their reach; otherwise, he felt certain, they would have
attacked and killed him. But after remaining about for nearly
an hour they departed. The skin of a white wolf killed north
of Grand Lake, Newfoundland, about 1896, by Dr. El wood
Worcester, was presented by him to the Museum of Compara-
tive Zoology. It was shot as it came over the side of a hill
where Dr. Worcester was waiting, gun in hand, for a chance at
a caribou. Two others, which he saw at about the same time,
were running along the shore of a lake above River of Ponds
in pursuit of a caribou. One of these looked very dark in
color. Bangs (1913) wrote that when Doane was collecting for
him about 1894 one was killed, but neither Millais, writing
about 1906, nor Dugmore in his book on the Newfoundland
caribou in 1913, mentions seeing anything of the wolf, which
by that time must have been greatly reduced in numbers.
The last record known to me is of one reported to have been
killed about 1911. The late Dr. John C. Phillips, who con-
tributed this note, endeavored to obtain the specimen but was
unsuccessful.

While Newfoundland at its nearest point is hardly 10 miles
distant from the mainland of Labrador, it seems that this wolf,
like so many other animals of the island, must have been
isolated here for a long period and gradually developed the
slight cranial differences that distinguish it. At the present
time the Newfoundland wolf is believed to be extinct.

VANCOUVER ISLAND WOLF
CANIS LUPUS CRASSODON Hall

Canis occidentalis crassodon Hall, Univ. California Publ. Zool., vol. 38, p. 420, Nov.
8, 1932 ("Tahsis Canal, Nootka Sound, Vancouver Island, British Columbia").

Skulls of the Vancouver Island wolf differ from those of the
adjacent mainland and southern Alaska in their smaller size,
shortened preorbital region, less elevated tip of rostrum, and
greater shallowness of the lower jaw through the coronoid
process. The color is gray to black. Condylobasal length of
skull, 240 mm.

NORTH AMERICA AND THE WEST INDIES 207

Five specimens are mentioned by Hall in his original descrip-
tion of this wolf. Concerning the status of the race, Swarth
(1912) writes that a quarter of a century ago wolves were still
fairly common "in the wilder parts of at least the northern
two-thirds of Vancouver Island, sufficiently so as to be a serious
menace to the deer in many places." At Beaver Creek he
heard one howl and was told that they were occasionally seen
in the valley during the winter but very seldom in summer.
At Nootka they were said to be abundant. "On the Tahsis
Canal we stayed at the camp of a trapper who made it his
principal occupation during the winter to hunt wolves and
panthers." With the bounty of $15 a head added to the value
of the fur, the trapper evidently secured enough to make it pay.
This trapper used poison exclusively and seemed to have had
no trouble in killing the animals, but in several years' hunting
he had actually seen but two or three alive, so cunning are they
in keeping out of sight. Swarth stated that wolves had been
so common at Friendly Cove during the years immediately
preceding 1912 that deer had been almost completely driven
out of the neighborhood. Of the status of this wolf at the
present day no further information is at hand.

NORTHWEST COAST WOLF; "BROWN WOLF"
CANIS LUPUS FUSCUS Richardson

Canis lupus var. fusca Richardson, Zool. Beechey's Voyage of the Blossom, Mammals,
p. 5, 1839 ("California and the Banks of the Columbia River").

SYNONYM: Lupus gigas Townsend, Journ. Acad. Nat. ScS., Philadelphia, ser. 1, vol. 2,
p. 75, Nov., 1850 (near Vancouver, Columbia River, Washington).

The wolf formerly occurring from northern California to the
Puget Sound region, and probably southwestern British
Columbia west of the Cascades, was regarded as a valid form
by Miller (1912) in notes on the names of the large wolves of
North America. He states that a skull from the Puget Sound
region "indicates an animal differing from the timber- wolf of
the interior region in less great size and in less enlarged teeth."
Hall (1932) contrasts specimens from Oregon with the race
crassodon of Vancouver Island and says that they differ
noticeably in "the markedly less inflated tympanic bullae,
much smaller teeth," in the more convex lower profile of the
jaw and other minor skull characters. The color is "much
darker" than in the Plains wolf (Bailey).

208 EXTINCT AND VANISHING MAMMALS

This race, which was perhaps characteristic of the saturate
region of forest along the Northwest coast, is doubtless now
extinct in California (Grinnell, 1933). In Oregon, Bailey
(1936) writes, "a few of these large, dark gray wolves are still
found in the timbered country west of the Cascades .
and locally northward to British Columbia and Alaska. In
recent years they have been found mainly along the west
slope of the Cascade Range, but before extensive white settle-
ments were made in the State they seem to have been common
in the Willamette Valley and west to the coast ... In
1897 Captain Applegate reported them as formerly common,
but at that time extremely rare in the southern Cascade
region." In 1913-14 bounties were paid on 30 wolves taken in
Oregon, but in later years they seem to have been largely
exterminated within the State. Bailey mentions one taken in
1930 on the Umpqua National Forest, where it had killed
several sheep. It was very old with much worn teeth. A
second old male was killed the same winter in Klamath County,
while in Douglas County and Lane County one each was killed
in 1930 and 1931. At the present time, he writes, their num-
bers are so nearly under control that their damage is negligible.
This probably applies also to Washington, but no recent
statistics are available. In the remoter regions of western
British Columbia they are probably still present in some
numbers.

NORTHERN ROCKY MOUNTAIN WOLF

CANIS LUPUS IRREMOTUS Goldman

Canis lupus irremotus Goldman, Journ. Mamm., vol. 18, p. 41, Feb., 1937 ("Red
Lodge, Carbon County, southwestern Montana").

This is the wolf of the northern Rocky Mountain region and
high adjoining plains, from northwestern Wyoming northward
through western Montana and eastern Idaho at least to Leth-
bridge in southern Alberta. Intergradation with neighboring
races is presumed and makes the precise definition of the
range difficult. It is a light-colored race with narrow but
flattened frontal region, in these characters differing from the
race youngi of the southern Rocky Mountains. It is larger
and paler than nubilus of Nebraska or fuscus of Washington,
and smaller than the form occidentalis to the northward of its
range.

NORTH AMERICA AND THE WEST INDIES 209

The race is based on more than 30 specimens from Montana
and Idaho; hence its characters seem well established. Prob-
ably this is the form of eastern Oregon concerning which
Bailey (1936) has a few interesting notes (under Canis lycaon
nubilus). In this region wolves were numerous in the middle
of the last century and were plentiful even up till the nineties.
At the present time they are largely gone from all this region,
but doubtless still remain in small numbers in Montana and
Wyoming.

LABRADOR WOLF

CANIS LUPUS LABRADORIUS Goldman

Canis lupus labradorius Goldman, Journ. Mamm., vol. 18, p. 38, Feb., 1937 ("Vicinity
of Fort Chimo, Quebec" northern Labrador).

In color this wolf of northern Labrador varies from "dark
somewhat grizzly gray to almost white." It is larger than the
eastern wolf, with heavier dentition and minor cranial differ-
ences, and seems to be only slightly differentiated. Its geo-
graphic range is not at present determined but is presumed to
be the Labrador Peninsula north of the forest region.

Concerning the status of these wolves in Labrador but little
information is available. They are, however, not uncommon
and follow the herds of barren-ground caribou. William B.
Cabot, in his account of travels into the eastern interior of the
peninsula, speaks of them as occasionally to be heard, and in
1905 he saw tracks of perhaps as many as 200. They are
occasionally killed by hunters near the coastal settlements in
winter, but since the interior of the country is rarely pene-
trated by white men it seems unlikely that they will be much
disturbed there for a long period to come.

ALEXANDER ARCHIPELAGO WOLF
CANIS LUPUS LIGONI Goldman

Canis lupus ligoni Goldman, Journ. Mamm., vol. 18, p. 39, Feb., 1937 ("Head of
Duncan Canal, Kupreanof Island, Alexander Archipelago, Alaska").

This wolf is believed to be the race characteristic of the
Alexander Archipelago and adjacent mainland of southeastern
Alaska. It is a race of medium size, "much smaller than Canis
lupus pambasileus of the Mount McKinley region, Alaska, or
Canis lupus occidentalis of the Fort Simpson region, . . .

210 EXTINCT AND VANISHING MAMMALS

Mackenzie." It is similar to the race C. I. fuscus of the coast
districts to the south, but a larger percentage of black or
blackish individuals is normal, while it is less suffused with
buffy in the gray phase. The small auditory bullae distinguish
it from the race crassodon of Vancouver Island, while its longer
rostrum and palate and the longer narrower nasals separate it
from fuscus. None of these characteristics seems very trench-
ant, taken alone, but Goldman considers it a "well-marked
form," closest related to the last.

Swarth (1936), commenting on the distribution of this wolf,
says that it is "found upon the same islands as the Red Squirrel
[i. e., Mitkof, Kupreanof, and Kuiu], and also upon the large
Prince of Wales and Dall Islands with some others of this
southern group that the squirrel has never reached. Evidently
the wolf did not come directly from the north" but reached
those of the inner islands that formed a series of stepping-
stones most easily crossed from the adjacent southeastern
coast. I have no recent information concerning its present
status, but apparently it still occurs on these islands in small
numbers.

EASTERN .TIMBER WOLF

CANIS LUPUS LYCAON Schreber

Canis lycaon Schreber, Saugthiere, vol. 3, p. 353, pi. 89, 1776 ("Vicinity of Quebec,

Canada," as selected by Goldman, 1937, p. 37).
FIGS.: Stone and Cram, 1902, pi. opposite p. 278; Nelson, 1916, p. 423, upper figs.

(color phases).

The eastern timber wolf is slightly smaller than the wolf of
northern Labrador, and usually of a grizzled black, white, and
gray. As long ago pointed out by Baird, it differs from the
Plains wolf and the neighboring northern races in its weaker
rostrum. Weight upwards of 100 pounds.

Three hundred years ago this wolf ranged over the wooded
and open areas of eastern North America from southern Que-
bec to probably the southern States and westward to the
Plains. With the coming of white men and the occupation of
the country, it has been gradually extirpated from the settled
regions until at the present time wolves are gone from practi-
cally all the eastern United States and probably none are now
to be found south of the St. Lawrence and east of Michigan.

When the first English colonists landed on the shores of New
England they found wolves plentiful and troublesome and were

NORTH AMERICA AND THE WEST INDIES

obliged at once to take active measures to prevent them from
seriously depleting their precious cattle and sheep brought
with so much labor from England. No doubt too they feared
the attack of wolves in packs against themselves, particularly
if traveling alone or at night, yet there seems to be no authentic
instance of anyone actually being killed by wolves, though
many a person going after dark to a neighbor's hurried his
footsteps on hearing a wolf howl. Apparently the Indians
along these coasts were little afraid of wolves (for they had no
livestock), but occasionally they caught them in traps, and
wolf bones in small amount occur in the Indian shell heaps.
The early settlers often kept their sheep on the smaller outer
islands where they were comparatively safe and unable to
stray. But flocks or cattle in pasture were in constant danger
even when in pens near the houses. The early records of
Plymouth and Massachusetts Bay colonies show that a de-
termined campaign was waged against wolves. Bounties were
offered for adult wolves and lesser ones for whelps, and the
heads of wolves brought in for the reward were publicly ex-
posed on hooks at the meetinghouse. The Indians were active
in this campaign and secureVl their share of the bounties.
Yet for nearly a century wolves continued to be a menace to
stock. In the "History of Cape Cod" Freeman relates that
in 1717, the early settlers discussed a project for erecting a
high fence of palisades or boards between Sandwich and Ware-
ham to shut off the outer Cape completely as a safe pasture
for the livestock against the wolves. But the scheme finally
was abandoned, since the several towns involved could not
agree on their proportionate share of the expense, while those
outside the Cape did not see why the wolves should be retained
on their side of the fence! Many of the towns maintained
" wolf pits," dug deep enough so that a wolf could not jump out,
the inwardly sloping sides sometimes smoothly faced with
stone, and the whole baited with fresh meat. Wolves jumping
in for the bait were unable to leap out and were later killed.
The remains of such a trap are said still to exist in the Lynn
Woods park in Massachusetts. The idea was doubtless taken
from the Indians.

Even down to the time of the American Revolution wolves
were fairly frequent in or near the settled parts of New Eng-
land, but soon after, the unremitting fight against them had so

212 EXTINCT AND VANISHING MAMMALS

reduced their numbers that they were no longer a serious
menace. Occasional small packs or single cunning survivors
were systematically hunted down by companies of men who
would track them to deep woods or swamps which were sur-
rounded and the animals driven out and often shot. In 1774,
Connecticut repealed the law that formerly offered a bounty
on wolves. But if wolves were then no longer dangerous to
stock in Connecticut, they continued to be so in the less settled
regions of southern New Hampshire and Maine. Bounties
were increased and hunting wolves became a regular occupa-
tion of some of the Indians or white hunters. The constant
war of extermination at length began to tell, and by the early
years of the last century few wolves were left in southern New
Hampshire and Vermont. In his history of the town of Gilsum,
near the present Keene, N. H., Hay ward in 1879 tells that in
March or April, 1828, took place what was then still remem-
bered as "the wolf hunt," when a lone individual that had
killed a number of sheep was finally hunted down, twice sur-
rounded by a party of determined farmers, and shot as it slunk
out from an old spruce top. Many similar tales may be found
in the local histories. The "last wolf" in Lancaster, N. H.,
was trapped about 1840 according to local account, but I have
reason to believe that an occasional one or two came into the
White Mountain region till a much later date. Zadock Thomp-
son mentions a specimen in the University of Vermont that
was killed in Addison County about 1830, but probably they
occasionally were found well after that time. Merriam in his
"Mammals of the Adirondacks" (in 1882) lists wolves on
which bounties had been paid that were taken in St. Lawrence
and Washington Counties, N. Y., even up to the latter date.
When the last wolf was taken in Maine is perhaps uncertain.
The only specimen known to me from that State is a skin now
in the Boston Museum of Natural History, killed near Moose-
head Lake in 1863. The late Manly Hardy, who traveled all
over Maine as a trapper and fur buyer, and whose authority
is unquestioned, wrote in 1884 (Forest and Stream, vol. 22, p.
141) that "in 1853 wolves were very plenty, and for the next
five years were not scarce; plenty could be found within sixteen
miles of Bangor in 1857 and 1858. They seemed to leave
quite suddenly. The last I know of positively being taken was
. . . in 1860, at Munsengun . . . There were rumors

NORTH AMERICA AND THE WEST INDIES 213

of occasional wolves being seen from 1875 to '80, but the first
real proof I can give is that in 1880 I bought the skin of a
freshly killed wolf, taken at Union River. The one who
brought it said it had a mate, and they had been heard at
times for several years in that vicinity." A few later reports
of wolves or their tracks seen in northern Maine bordering
New Brunswick may have been authentic, but evidently they
were practically gone by that time, not only from Maine but
from New Brunswick as well. Indeed, Chamberlain (1884)
writes that although common until about 1860 in the latter
region they had since "entirely disappeared."

In the wilder regions of the Adirondacks of New York,
wolves in small numbers continued till a much later time.
Miller (1899) has published a table showing by counties the
bounties paid on wolves by New York State, incorporating
and continuing the earlier one of Merriam (1882). From this
it appears that between 1871, when a bounty was declared, and
1897, no less than 98 wolves were killed on which a total of
$2,910 in bounties was paid. Further, in the decade preceding
1880, six was the most killed in any year, but 12 were brought
in in 1881 and no less than $1 in 1882! In 1883 nine were
killed, in the next three years two in each, one in 1887, and
two in 1888. Thereafter appears a blank in the record, and
some have believed the animal was then extinct in New York,
but in each of the years 1895, 1896, 1897, bounties were paid
on six wolves! One can not help the thought that something
is strange about this. Of the 98 wolves on which the State paid
bounties between 1871 and 1897, 45 or nearly half came from
St. Lawrence County, which borders the river of that name,
suggesting that there may have been occasional immigrants
from the Ontario side. For the Toronto region, however,
Fleming, in 1913, writes as if they had long since been extinct
and states that "according to the late Dr. Brodie, wolves
occasionally drove deer into Markham as late as 1840."

The status of the wolf in Pennsylvania and New Jersey has
been carefully investigated by S. N. Rhoads, and the results
of his search for records are summarized in his book on the
"Mammals of Pennsylvania and New Jersey" (1903). From
the latter State they were so early extirpated that little record
remains to indicate for how long they were present there, but
perhaps till early in the nineteenth century. Pennsylvania,

214 EXTINCT AND VANISHING MAMMALS

however, with its mountains and extensive wilderness areas,
continued to harbor them up to comparatively recent years.
The last record for New Jersey is given as of one killed in
Wayne County in 1887, which was driven in from New York
State by dogs, but its ultimate source may be questionable.
Wolves seem to have remained common in the less settled and
wilder parts of Pennsylvania until about the middle of the last
century. Thus, Rhoads quotes Warren as authority for the
statement that "about the year 1845 wolves were abundant in
Tomhickon Valley" while between 1808 and 1820 Lucerne
County paid $2,872 in bounties for wolf scalps at about $5
each. As many as 273 wolves were killed in one of these years.
Another correspondent informed him that he saw many as late
as 1857 on the head waters of Pine Creek and Sinnemahonig
in Potter County, where the last one he knew of was killed
about 1875. Tomkins (1931) records a "last" wolf near
Williamsport in 1867 and one near Ralston, in 1872, in the
north-central part of the State. Many records are quoted for
the middle part of the nineteenth century, and Rhoads makes
the summary statement that they finally became "apparently
exterminated in Pennsylvania, within the last 10 or 15 years,"
that is, in the middle or late eighties. More recent reports
lack authenticity, while of various cases investigated in which
bounties had been paid in later years some proved to be
deliberate frauds. In another it turned out that a wolf had
been brought from the West and had later escaped. In Elk
County, Pa., the last wolf is said to have been shot in 1891.
Probably the actual extinction of native eastern wolves in the
State may be set at about that date. A few seem to have re-
mained still in West Virginia wildernesses, where, according to
Fred E. Brooks, "what is supposed to have been the last gray
wolf in West Virginia was killed in Randolph county by
Stofer Hamrick in January 1900."

Apparently wolves were found in North Carolina up to
about the beginning of the present century. At all events,
C. S. Brimley, writing in 1905, states that it was then "found
sparingly throughout the mountains. Dr. Donald Wilson re-
ports it from Graham County, Mr. R. W. Collett says it is
growing very scarce in Cherokee, Mr. Fain says there are a
very few in Buncombe [County]. Mr. H. H. Brimley, Curator
of the State Museum, informs me it has been taken in recent

NORTH AMERICA AND THE WEST INDIES 215

years in Yancey, Caldwell and Watauga. " Dr. C. Hart Mer-
riam states that during one of his visits to Roan Mountain, in
1887 or 1892, a den of wolves was discovered and the young
were captured. In the last quarter of the eighteenth century
wolves were spoken of as abundant in South Carolina, but by
Audubon's day they were much fewer. The last one known to
have been killed in the State was shot in Berkeley County,
near St. Stephens, between 1856 and 1860. E. B. Chamberlain,
of the Charleston Museum, who kindly supplied this record,
had reports of other wolves in the Santee section at about the
same time.

How far to the south the gray or timber wolf extended its
range may never be certainly made out, for the black or
Florida wolf of extreme southern United States is now regarded
as a race of the species Canis rufus by Goldman in his recent
review. Dr. Francis Harper (1927) refers to this form the
wolves formerly found in Okefenokee Swamp, southern Georgia,
where according to his investigations the last one killed was
about 1908.

West of New York State and the Allegheny Mountains, the
timber wolf was formerly common and is believed to have
intergraded with the Plains wolf in the more open region west
of the Mississippi. The story of its original abundance and
gradual extermination is much the same as elsewhere in the
East, except that in the more remote and wilder regions, as in
northern Wisconsin, some still remain. In northern Minnesota
wolves were common till at least the late years of the last
century. Herrick (1892) has an interesting note in his "Mam-
mals of Minnesota" indicating, as in some other instances, an
occasional influx of numbers in certain districts: "During the
winter of 1884-85, wolves became very abundant and insolent
in Wright county, and were seen about the outskirts of Monti-
cello in broad daylight almost daily, though they were suffi-
ciently wary to escape capture." Cory (1912) writes that this
wolf was still "common in northern Wisconsin and possibly
occurs occasionally in other parts of the State," but although
stragglers may at times appear in Illinois, all efforts to secure a
specimen have failed, and most of the reports of wolves prove
on investigation to be based on coyotes. In Wisconsin, recent
records are available from as far south as Buffalo County.

In the region of the Ohio Valley wolves were formerly plenti-

216 EXTINCT AND VANISHING MAMMALS

ful but with the settling of the country a century ago were
soon reduced to innocuous numbers, and in Ohio had become
very rare by 1838 and "nearly extinct" a decade later. Audu-
bon, while living at Henderson, Ky., early in the second decade
of that period found black wolves abundant. He recounts that
one morning he found a black wolf in one of his wild-turkey
pens and shot it in the act of devouring one of the birds of
which it had already killed several. Whether these wolves
were of the timber wolf type or the red wolf type is indetermi-
nable. In Tennessee wolves were exterminated many years
ago (Kellogg, 1939), and but few records exist. Merriam in
1887 found that a few still were present in the Smoky Moun-
tains and in middle Tennessee they were reported as occasion-
ally found in Van Buren County. Late records of wolves
killed, Kellogg gives as: A female and her pups at Waynesboro,
about 1917; and one in 1919 on North Fork River, Cumberland
County. In western Tennessee the last available record is
given as about December 10, 1895, when two were killed near
Brownsville, Haywood County.

In the region to the southward of South Carolina and Ten-
nessee, the black race of the red wolf is believed to be the form
found.

At the present time the timber wolf is still to be found in
limited numbers in eastern Canada, as in the less settled por-
tions of Quebec along the north shore of the Gulf of St. Law-
rence, where near Matamek I was shown in 1929 the skin of a
large one killed the winter before. Doubtless there are many
more between the Gulf and James Bay. At least until recent
years, wolves were present in the Algonquin National Park,
while farther to the west, in Quetico National Park, along the
international boundary from Lake of the Woods to the western
end of Lake Superior, and in the adjacent Superior National
Forest contiguous with it on the United States side in north-
eastern Minnesota, Cahn (1937) writes that "wolves are still
common" and cross back and forth freely between the two areas.
Here numerous and spasmodic drives have been made to rid
the area of these predators but have resulted mainly "in the
killing of about everything except wolves, which usually elude
most cleverly the clumsy efforts to trap them. Late in fall
their call is to be heard frequently all over the Quetico, but in
all my travels I have seen but 2 live wolves Their

NORTH AMERICA AND THE WEST INDIES 217

persecution is to be regretted, for they form a perfectly natural
and normal check on other species." In this last statement is
to be found the key to the preservation of the species in regions
where it comes in contact with white men. Under the nearly
natural conditions of a large national preserve for game, wolves
in these northern regions will be useful in preventing too great
an increase of deer, and to a less extent of moose, which in the
absence of such natural predators may easily become unduly
abundant to the impairment of their range as in the case of the
Kaibab deer or those of the Murderers Creek Game Refuge in
eastern Oregon. Even in such situations, however, the game
ranger has a special problem in determining how many wolves
may safely be allowed on refuge areas, and how best this num-
ber may be maintained through occasional destruction of
young or of adults that become too bold or range outside the
preserve to kill cattle or sheep in adjacent districts.

MOGOLLON MOUNTAIN WOLF
CANIS LUPUS MOGOLLONENSIS Goldman

Canis lupus mogollonensis Goldman, Jouri?. Mamm., vol. 18, p. 43, Feb., 1937 ("S. A.
Creek, 10 miles southwest of Luna, Catron County, New Mexico").

This is a "rather small, usually dark-colored subspecies.
Similar in general to Canis lupus youngi of the Rocky Moun-
tain region . . . but smaller and usually darker in color.
Decidedly larger and usually lighter in color than Canis lupus
bailey i of the Sierra Madre of Mexico. Closely allied to Canis
lupus monslrabilis of Texas, but smaller" and the skull with
"zygomata more widely spreading; frontal region less elevated
and inflated; auditory bullae smaller."

The distribution of this form is said by Goldman to include
the Mogollon Plateau region of central Arizona, extending
eastward through the Mogollon Mountains to the Sacramento
Mountains of New Mexico, and though rather circumscribed
this seems nevertheless to be well borne out by a series of 120
specimens studied by its describer. Vernon Bailey (1931)
writes that "in 1908 he was again over the Gila National Forest
and Mogollon Mountain region and found the wolves still
common." They were feeding on nothing but fresh meat of
their own killing, for "stock was abundant, and they had no
trouble in finding cattle of any age or condition they preferred,

218 EXTINCT AND VANISHING MAMMALS

but they seemed to show little choice in their selection of beef.
Cows, steers, or calves seemed to be killed indiscriminately as
the wolves happened to come upon them when hungry. Cattle
are killed throughout the year and seem to be preferred to
deer, which are more nimble and not so easily caught. In the
wolf droppings along the trails cattle hair was almost the only
recognizable constituent, but occasionally some jack-rabbit fur
could be detected. "

On account of their constant depredations against stock, a
determined warfare has been carried on in this region during
recent years with the result that in 1937, Goldman writes that
this form is now "nearly if not quite extinct." There is some
evidence, however, that if the campaign of extermination is
relaxed for a time, other wolves keep coming across the border
from the adjacent parts of Mexico into southern New Mexico
and Arizona, so that if in later years specimens are taken in the
former range of the Mogollon Mountain wolf, they may be of
some other race.

TEXAS GRAY WOLF; "LOAFER"; "LOBO"
CANIS LUPUS MOKSTRABILIS Goldman

Canis lupus monstrabilis Goldman, Journ. Mamm., vol. 18, p. 42, Feb., 1937 (" 10 miles
south of Rankin, Upton County, Texas").

This is a slightly marked race averaging darker than the
wolves to the north and with a more highly arched frontal
region. Its range formerly included southern and western
Texas and northeastern Mexico, but because of systematic
persecution its numbers are at present greatly reduced.

Vernon Bailey's (1905) account of this wolf furnishes much
valuable information. At that time they were abundant in
and about the Davis and Guadalupe Mountains and over the
Staked Plains and open country east of the Pecos River. He
writes: "When opportunity offers, the 'loafer' not only kills
sheep but often kills a large number, apparently for the
pleasure of killing. His regular and most serious depredations,
however, are on the scattered and unguarded cattle of the
range. Two or three wolves usually hunt together and some-
times pull down a steer, but most of their meat is procured
from yearlings or cows. Occasionally a colt is killed but not
often. Where two or three wolves take up their residence on

NORTH AMERICA AND THE WEST INDIES 219

a ranch and kill one or more head of cattle almost every day,
the ranchmen become so seriously alarmed that they frequently
offer a reward of $50 or $100 apiece for scalps. " Bounties have
been tried, and in some cases several smaller ranches combine
to offer a large reward in addition to that paid by the county,
but the danger is that the hunters are tempted "to save the
breeding females and dig out the young each year for the
bounty, thus making their business not only profitable but
permanent. " A better method seems to be to employ profes-
sional hunters who are well paid by the month to keep down
wolves or other noxious animals.

PLAINS WOLF; BUFFALO WOLF

CANIS LUPUS NUBILUS Say

Canis nubilus Say, Long's Exped. Rocky Mountains, vol. 1, p. 160, 1823 (Engineer

Cantonment, near present town of Blair, Nebraska).
SYNONYM: Canis variabilis Wied, Reise in das Innere Nord-Amer., vol. 2, p. 95, 1841

(Fort Clark, near Stanton, Mercer County, North Dakota).

This was the wolf found on the Great Plains of the United
States from the Dakotas, Nebraska, and Kansas west to the
Rocky Mountains and eastward possibly to the western parts
of Iowa and Missouri. Though not so pale as the races to the
north, it was less dark than C. I. lycaon to the east. The skull
was similar in size to that of the adjacent races monstrabilis
and youngi, but the frontal region was more flattened than in
the former, with wider rostrum.

In the middle of the last century and earlier these wolves
were common on the Central Plains, following the herds of
bison and picking off young or disabled individuals, but at the
present time they are extinct over the greater part of the area
once occupied. With the pushing of the railroads out across the
country, the coming of settlers, and the development of stock
grazing, the buffalo were reduced or exterminated and the
wolves for a time devoted their attention to cattle, with the
result that the ranchers and hunters pursued them relentlessly
until now only a few remain in the remoter regions, where the
roughness or barrenness of the country makes agriculture im-
practicable.

Vernon Bailey (1926) has gathered together many interesting
notes on their former presence in the Dakotas, where in the

220 EXTINCT AND VANISHING MAMMALS

days of the early explorers they were abundant. With the
passing of the buffalo, however, they were poisoned and
trapped in such numbers that they rapidly disappeared from
their old haunts. They remained abundant up to the time of
Wied (1833) and of Audubon (1843), but by 1870 J. A. Allen
wrote that in Iowa "although rather common twenty years
since, they are now scarce, especially in the more settled dis-
tricts." Coues says that in 1873 they did not appear to be
numerous in the Dakotas, while by 1887 they were practically
gone from most of the country across North Dakota. Bailey
states that in 1913 they were numerous enough to be trouble-
some in the Badlands along the Little Missouri, though rare
elsewhere. A few were reported in 1916 west of Cannon Ball,
and in 1922 one was killed near Fargo. At the time of writing,
in 1926, Bailey believed there were still a few wolves left in the
least-settled parts of the rough Badlands region west of the
Missouri.

Dr. M. W. Lyon, Jr. (1936), records the former abundance of
this wolf in Indiana. "About the middle of the last century
they were practically exterminated although a few fairly au-
thentic records date back as recently as 25 years ago," i. e.,
about 1911. The pioneer's method of hunting them in the
Wabash Valley about 1830 was a sort of round-up, with lines
of hunters converging on a central point. Often two to ten
wolves would be taken in a day at these hunts, while the bounty
provided an added stimulus for their destruction.

In Kentucky this type of wolf may once have existed, but
Funkhauser (1925) believes it doubtful if at the present time
any are to be found in the State. On the other hand, it is
surprising that in Missouri it still occurs in some numbers.
In their recent survey of the game and fur animals of that
State, Bennitt and Nagel (1937) believe that on the basis of
careful investigation lately made "the maximum density of
wolves is about one per 10 square miles and the minimum about
one per township (36 square miles)." They estimate the total
wolf population of Missouri as in the neighborhood of 3,500
wolves, with the maximum concentration in the ten west-
central counties south of the Missouri River (a portion may
be C. rufus gregoryi). The State paid bounties on 249 wolves
between July 1, 1923, and July 1, 1925, but it is likely that a
good part of these were for large feral dogs of which there are

NORTH AMERICA AND THE WEST INDIES 221

many, perhaps exceeding the wolf population. Wolves and
coyotes probably furnish a most effective check on the undue
increase of woodchucks and ground squirrels, but the part
played by each is uncertain. Wolves also kill some deer.
With the diminution of good wolf cover in recent years, there
has probably been a concentration of the animals proportion-
ately where cover is good, and it is even believed that their
numbers have shown an actual increase in later times. If they
become really numerous, these authors observe, they must be
controlled, but under usual wild conditions they are a natural
element of the fauna and their presence may even have a bene-
ficial effect on the species that they prey upon in improving
the quality of the race through their elimination of the less
able animals.

Farther west, the Plains wolf is apparently quite gone from
Kansas (Hibbard, 1933), though it was common in that State
in the sixties or later, and the same is probably true for Ne-
braska where they remained in some numbers till about 1890.
In Montana, Bailey (1918) writing of the fauna of Glacier
National Park, says: "Along the eastern edge of the park, in
the open country, are still founS the large, light-colored plains
wolves . . . and a few occasionally range up through the
valleys and over the high parts of the mountains . . .
Along the trails in the Belly River valley in August, 1917, I
saw wolf signs that were not very old. The wolves had evi-
dently been down in the cattle country as the sign was com-
posed mostly of cattle hair. Don Stevenson reports them in
the country about Chief Mountain for the past 20 years, where
they have ranged up and down the edge of the Plains killing
cattle and some horses, and in 1914 he saw their tracks on St.
Mary Ridge at the park line. There are said to be some in the
North Fork valley, where it is probable they are attracted by
the abundance of deer, as they are on the eastern border by the
abundance of stock. In 1895 they seemed to be no more
common than at the present time, as I saw then only a few
tracks on the prairie below St. Mary Lake, and some fine skins
among the Indians on the Blackfeet Reservation. As the
valleys settle up, more vigorous hunting and trapping is likely
to crowd the wolves back into the park at any time and make
them more numerous than they are at present. If so, their
destruction of game will be correspondingly increased * . .

EXTINCT AND VANISHING MAMMALS

If unmolested they seem to prefer the domestic stock where it
is abundant and easily accessible. " With the demands of the
cattle interests for their destruction, their only possibility of
long surviving is in large preserved areas such as the national
parks.

NORTHERN PLAINS WOLF

CANIS LUPUS OCCIDENTALS Richardson

Canis lupus occidentalis Richardson, Fauna Boreali-Americana, vol. 1, p. 60, 1829

(Simpson, near mouth of Liard River, Mackenzie).
SYNONYM: Canis lupus griseus Sabine, Franklin's Narr. Journ. Polar Sea, p. 654, 1823

(Cumberland House, Keewatin, Canada).

This is the wolf of the northern interior plains east of the
Rocky Mountains and north of the United States, northward
to the Arctic tundra. It is still paler than the Plains wolf,
C. I. nubilus, and larger.

Writing of the Athabaska-Mackenzie region in 1908, Preble
says: "Gray or Timber wolves are found throughout the
wooded parts of the region, and are fairly abundant and ap-
parently increasing in some sections. In 1901 we saw numerous
skins at nearly all the posts visited, and found a skull at a
trapper's cabin on Slave River, 10 miles below the mouth of
the Peace. Among a number of skins seen at Fort Rae, most
of which were in the normal or gray phase, was one the color of
which was mainly dark bluish gray, the throat and back were
nearly black, the latter flecked with a few white hairs; the chest
had a white patch; the belly and tail were bluish gray, the
latter blackish toward the tip.

"During the season of 1903 we heard that wolves had been
rather abundant for several years past in the region west of
Smith Landing, in the Birch Mountains, and in the vicinity of
Athabasca Landing." In the same winter a black one was
killed at Fort Simpson. Tyrrell says that this wolf occurs in
the country between Athabaska Lake and Churchill River,
but not plentifully. Preble quotes the record of skins of this
wolf received by the Hudson's Bay Company as follows : From
1858 to 1884, from the Athabaska district, 2,119; from 1885 to
1889, a further 339 skins. For the district of Mackenzie,
between 1863 and 1884, a total of 1,880 skins; in 1889 only 49
skins; Fort Resolution and Great Slave Lake, between 1862

NORTH AMERICA AND THE WEST INDIES 223

and 1887, 193 skins. Dr. Francis Harper (1932) traversing
this region in 1914 noted seeing a number of skins and other
evidences of wolves. No doubt they will continue in the general
region in slowly decreasing numbers for a long time, though
annually trapped and poisoned.

GREENLAND WOLF

CANIS LUPUS ORION Pocock

Canis lupus orion Pocock, Proc. Zool. Soc. London, 1935, p. 683, Sept. 12 ("Cape

York, on Baffin Bay, Greenland").
FIGS.: Manniche, 1910, figs. 18, 19 (animal), fig. 20 (skull) ; Jensen, 1928, p. 322, fig. 2

(mounted specimen).

The Greenland wolf is described as of "a uniform whitish
grey, except for the tolerably extensive marbling and streaking
of the black-tipped contour hairs of the upper side, but the
ears are pale buff, turning to drab towards the point; the top
of the head as far as the angle of the eye, and of the muzzle,
apart from its white tip, is very pale brownish grey . .
the tail has a black tip, but the rest of its upper side is not so
conspicuously black and white as the back." The skull is said
to be small and with smaller teeth than those of the races arctos
of Ellesmere Land or tundrarum of the Arctic coast of Canada.
"In the fore feet, the digital pads are greatly reduced in size
and there is no trace of the pollical or carpal pads. "

Professor Jensen (1928) states that this wolf "has at the
present time the same distribution as the musk-ox, in that it
occurs on the north coast and the northern part of the east
coast, as far as the region round Scoresby Sound . . .
Even in the regions where the wolf now occurs, it is by no
means frequent; it is only found scattered and is comparatively
rarely seen. Before 1899, when Nathorst, on his expedition to
northern East Greenland, observed several polar wolves, this
animal had not been seen in those parts, although they had
been visited by various travellers; this caused Nathorst to set
forth the view that the wolf had only quite recently immigrated
to the east coast from the north. " The caribou or reindeer on
which these wolves must have preyed are now gone from the
east coast, but the muskox still remains in small numbers, and
presumably furnishes a major part of the wolfs sustenance.
Manniche (1910) could find no evidence that the Arctic hare

224 EXTINCT AND VANISHING MAMMALS

or the fox formed any part of the prey. In his account of the
mammals observed by the Denmark Expedition to East
Greenland, during which the ship of the expedition wintered
near Hvalrossodden, he states that the party found evidence
of wolves between latitudes 75 and 83 10' N. He believes
that they have been derived from the North American regions
by way of northern Greenland. At all events they are few in
number and, curiously, seemed very wary, though one would
have expected the contrary from their very slight contact with
man. In the last of August, 1906, he saw a large white wolf
and at length succeeded in shooting it by lying in ambush
wrapped in the skin of a muskox he had killed, the meat of
which served as bait. This female was in excellent condition,
"almost fat," and weighed 35 kilograms. Its total length was
about 1500 mm.; tail, 420; ear, 100 long by 100 wide. Two
others were seen and tracks were found among the mountains
inland from Hvalrossodden and slightly to the west and north.
During the winter of 1906-7, while the expedition was frozen
in near this place, a group of three Arctic wolves came about
the ship from time to time, but were very wary, and in the dim
polar night offered no good chance for a shot. Several sledge
dogs were once set upon them, but they were driven back by
the wolves, which eventually succeeded in killing four of their
animals and wounding others. If a sledge journey were made
to the meteorological station or elsewhere in the neighborhood
one or more of the wolves would follow at a distance, but even-
tually they became bolder, so that by the end of the winter all
three had been shot or trapped. Of these the largest was a
male weighing 29 kilograms, very thin and emaciated.

It thus appears that the Greenland wolf at the present time
is confined to the narrow strip of country between the inland
ice and the sea from northern Greenland to about the region of
Scoresby Sound, following the muskox. That in former times
it ranged much farther south can hardly be doubted, and it
probably followed the caribou now largely extinct in Greenland.
It seems probable that the wolf, like the lemming and the
muskox, spread across northern Greenland and down the east
coast rather than the west coast, on account of the great ice
barriers in the region of Inglefield Gulf. The wolf seems to
have been gone from southern and western Greenland long ago,
though mentioned by early historians. Even Egede's "De-

NORTH AMERICA AND THE WEST INDIES

scription of Greenland" in 1745 mentions them as a thing of
the past. Hence it was a notable circumstance that in the
winter of 1868-69 one was killed near Unamak and is now
mounted in the Museum at Copenhagen. It was one of a pair
seen together, but where they could have come from is diffi-
cult to tell. At the present time the wolf population of north-
eastern Greenland must be limited, and whether there is oc-
casional immigration from Ellesmere Land over the ice is also
uncertain. The slight basis on which orion was described
makes necessary a further comparison between wolves of the
two areas when a sufficient number of skulls can be assembled.

ALASKAN WOLF; "AUTOCRAT TIMBER WOLF"
CANIS LUPUS PAMBASILEUS Elliot

Canis pambasileus Elliot, Proc. Biol. Soc. Washington, vol. 18, p. 79, Feb. 21, 1905
("Sushitna River, region of Mt. McKinley, Alaska").

Skull and teeth larger than in the neighboring races; "ridge
of sagittal and occipital crest nearly on a level with frontal and
with only a very slight descent at occiput and very deep at
that point; maxillae very b?oad . . . premaxillae ex-
tending considerably over one-half the length of the nasals
Color from nearly uniform black to white and black in
various mixtures. " The total length of the type skull is 263
mm.; length of carnassial tooth, 27.

Elliot describes these wolves from the upper waters of the
Sushitna River in the region of Mount McKinley as "remark-
able for their large size and black color." He believes their
skull characters are distinctive, and Goldman in his paper of
1937 lists this as a valid race, without attempting to define its
range. Evidently, however, it is larger than the race ligoni of
the Alexander Archipelago and somewhere intergrades to the
south with the northern Plains wolf, C. I. occidentalis. Its very
dark color is in contrast with the usual paleness of the latter.

The late Charles Sheldon (1930) has left a vivid account of
his hunting experiences during a year spent in the "Wilderness
of Denali" (the Indian name for Mount McKinley). He saw
little of these wolves, though on nine occasions he found in the
snow the tracks of wolves that had attempted to catch moun-
tain sheep by dashing upon them from above from a distance
of a hundred yards or so. In each case the wolf was unsuccess-

226 EXTINCT AND VANISHING MAMMALS

ful; nevertheless this danger to the sheep caused them to be
very wary and to keep to the high ridges. The wolves seemed
to find the caribou at lower levels a more dependable source of
food. Just to the north of Mount McKinley, Sheldon found
wolves along the upper Toklat River very abundant, always
hovering about the feeding herds of caribou and following them
as they roamed, usually in a fairly well-defined circuit. Late
in September the caribou frequent ridges more than the low
country and would be fairly conspicuous to a wolf. He re-
marks also on the innate caution of the wolf, which, though
rarely pursued in this region, is so wary that he never saw or
trapped one. Living as they do, in McKinley National Park,
it seems likely that the wolves of this area will continue for a
long time to harry the caribou of the outer ranges and less
often penetrate the upper slopes for sheep.

BARREN-GROUND WOLF
CANIS LUPUS TUNDRARUM Miller

Canis tundrarum Miller, Smithsonian Misc. Coll., vol. 59, no. 15, p. 1, June 8, 1912

("Point Barrow, Alaska").
SYNONYM: Canis lupus albus Sabine, Franklin's Narr. Journ. Polar Sea, p. 655, 1823

(Fort Enterprise, Mackenzie, Canada). (Not Canis lupus albus Kerr, from

Siberia.)

The tundra wolf has a narrower palate and slender rostrum
in comparison with the Plains wolves and the northern timber
wolves. It is decidedly larger than the eastern timber wolf
and in color is "said to be frequently white or whitish." It is
the form found in the barren-grounds of the Arctic coast region
from near Point Barrow eastward toward Hudson Bay, and
perhaps too in the Arctic Archipelago north of those shores.

Little recent information concerning this wolf is at hand.
In the middle of the last century and earlier, explorers found
them in some numbers about the Isthmus of Boothia, and
Back's expedition saw white wolves near Artillery Lake.
Preble (1908) has summarized some of these later observations
as follows: "Armstrong states that a wolf was seen near Prin-
cess Royal Islands in February, 1851 ; and that many were seen
at Mercy Bay, Banks Land, during the winter of 1851-52.
McDougall states that a pack of wolves was seen on Melville
Island near Cape Russell in June, 1853; one was seen on May
27, 1854, near Cape Hotham, Cornwallis Island; McClintock

NORTH AMERICA AND THE WEST INDIES

reports that wolves were observed in October, 1858, at Port
Kennedy, and in May, 1859, by Lieutenant Hobson on King
William Land. Kennedy records one seen in the central
part of Prince of Wales Land in April, 1852. While in the
Barren Grounds to the northeast of Fort Rae in the early
spring of 1894, Russell found wolves rather common. Of a
band of six, two were snow white, the others a light gray.
During his exploring trip between Great Slave Lake and Hud-
son Bay in 1900, J. W. Tyrrell found large wolves on the east
side of Artillery Lake. J. M. Bell informs me that during the
same season he occasionally saw wolves near . . . the eastern
end of Great Bear Lake. Hanbury, while traveling overland
between Baker Lake and the Arctic coast in the early spring of
1902, noted an occasional wolf. On April 30, when the party
was near latitude 67 between Lake Garry and Ogden Bay,
Darrell, his companion, encountered a band of 16 large wolves.
Darrell writes me that of this band 13 were of the ordinary
dirty white color, 2 were nearly black, and 1 pied. He states
that though these wolves live largely on caribou they are not
very successful in killing these animals unless they can separate
one from the herd, and that they always seem to be starving
. . . The band of 16 was the largest pack seen, the animals
usually being found singly or in pairs, though occasionally half
a dozen were observed together."

It seems clear that this race of wolf is nowhere very common
but is found in small companies or more often singly or in
pairs scattered over a large area and is not likely to become
exterminated for a long time. How great an area the animals
may cover in their search for food may not be possible to find
out, but it must be considerable, extending as it does to the
Arctic Archipelago. With the accumulation of specimens in
later years, a critical comparison is needed to demonstrate the
validity of the Ellesmere Land and Greenland races in com-
parison with that of this area.

SOUTHERN ROCKY MOUNTAIN WOLF

CANIS LUPUS YOUNGI Goldman

Canis lupus youngi Goldman, Journ. Mamm., vol. 18, p. 40, Feb., 1937 ("Harts Draw,
north slope of Blue Mountains, 20 miles northwest of Monticello, San Juan
County, Utah").

FIGS.: Grinnell, Dixon, and Linsdale, 1937, vol. 2, figs. 208-214 (as C. I. nubilus).

228 EXTINCT AND VANISHING MAMMALS

The wolf of the southern Rocky Mountain region is described
as a rather light-colored race of medium size, resembling the
Plains wolf, C. I. nubilus, but larger, its upper side more black-
ish and suffused with buff and its frontal region rising less
steeply. It is larger and paler than C. I. baileyi of Chihuahua,
with a broader, less depressed rostrum.

This wolf takes the place of C. I. nubilus, the Plains wolf,
west of the prairie region of Nebraska and Kansas and was
formerly found over "southeastern Idaho, southwestern Wyo-
ming, northeastern Nevada, Utah, western and central Colo-
rado, northwestern Arizona (north of Grand Canyon), and
northwestern New Mexico. " Over most of this range it has
been exterminated within recent decades by persecution of
hunters and of ranchers with whose cattle interests it interfered.
Bailey (1931) writes that through such efforts wolves were
practically eliminated from New Mexico by 1927. Goldman
(1937) states that at the present time it is mainly restricted to
northwestern Colorado. Warren (1910), in his account of the
mammals of the latter State, tells of one wolf, whose track was
identifiable through its having lost two toes in a trap, that be-
came so bold that special measures had to be taken for its
destruction. Its tracks with those of others were in a short
period found about the carcasses of 75 head of cattle and horses.
He quotes the figures of wolves killed in 1907 (from U. S.
Biological Survey Circular no. 63) as follows: In Colorado, 69;
in Wyoming, 1,009; in Idaho, 14. In 1910, Warren believed
there was not a county in the State of Colorado that did not
harbor some wolves. Cary (1911) wrote at about the same
time that wolves were still found in considerable numbers in
North Park and in Routt and Rio Blanco Counties, where they
kill a great many range cattle, and they were similarly destruc-
tive in Baca and eastern Las Animas Counties, in spite of much
trapping. Cary was told by professional trappers that so per-
sistently have wolves been hunted, trapped, and poisoned that
they would rarely come to a scent of any kind and seldom to a
baited trap, while poisoning was no longer successful. Traps
set blind in trails or near water holes were most successful. In
the Lily Park region on the lower Bear River wolves were
numerous until 1902, but in the two years following a trapper
killed 61, which nearly put a stop to their local depredations.
By 1907 they had become uncommon over most of southern

NORTH AMERICA AND THE WEST INDIES 229

Colorado. Such constant persecution is gradually reducing the
wolf population in all this region.

Whether this form of wolf originally occurred in the extreme
eastern parts of California may never be known. Grinnell,
Dixon, and Linsdale (1937) report the wolf as having been
practically exterminated from California for many years. No
specimen of wolf from the State had been preserved in any
museum up to 1922, while the two since taken on the eastern
border these authors identify as "clearly of the plains wolf
type" referring them to the race nubilus.

RED WOLF
CANIS RTJFUS RUFUS Audubon and Bachman

Canis lupus var. rufus Audubon and Bachman, Quadrupeds of North America, vol. 2,

p. 240, 1851 (Texas; fixed as 15 miles west of Austin).
FIG.: Audubon and Bachman, 1851, pi. 82.

FLORIDA WOLF

CANIS RUFUS FLORIDANUS Miller

1
Canis floridanus Miller, Proc. Biol. Soc. Washington, vol. 25, p. 95, May 4, 1912

("Horse Landing, St. Johns River, Florida").

SYNONYMS: Canis lycaon $ americana Hamilton Smith, Griffith's Cuvier, Anim. King-
dom, vol. 5, p. 144, 1827 (not Canis alopex americanus Kerr, 1795) ; Canis ater
Richardson, Fauna Bor.-Amer., pp. 70-72, 1829 (in part).

MISSISSIPPI VALLEY WOLF
CANIS RUFUS GREGORYI Goldman

Canis rufus gregoryi Goldman, Journ. Mamm., vol. 18, p. 44, Feb., 1937 ("Macks
Bayou, 3 miles east of Tensas River, 18 miles southwest of Tallulah, Madison
County, Louisiana").

FIG.: Arthur, 1931, fig., p. 147 (photograph of captive animal).

In his review of the American wolves, Goldman (1937) states
that the red wolf, Canis rufns, "usually regarded as a wolf
. . . , exhibits a departure from the true wolves, and in
cranial and dental characters approaches the coyotes." He
therefore gives it full specific rank and relegates to it as sub-
species the Florida black wolf and the wolf of the lower Mis-
sissippi Valley which he describes as C. r. gregoryi. This course
may be tentatively accepted and the three treated together.

The typical form of red wolf was slightly the smallest of the

230 EXTINCT AND VANISHING MAMMALS

three races and the most southwestern, with a range extending
from central Texas south west ward to the Mexican tableland.
In central Texas its range meets that of the Texas gray wolf,
Canis lupus monstrabilis, "but the two are not closely allied."
The latter has a much more massive and highly arched skull,
its dentition is simpler, the upper carnassial with a smaller
(often obsolete) anterointernal cusp, and the large upper molar
with less prominent cusp development on the internal lobe;
the posterior upper molars are relatively smaller than in rufus.
The color is reddish mixed with gray.

Bailey, writing in 1905, believed that the ranges of the red
and the gray wolves, though contiguous, did not overlap, but
whether there is overlapping or intermingling future studies
will perhaps show. As a result of special attempts to secure
specimens of C. rufus from Texas, Bailey reported that at the
time mentioned there were 14 skulls and 4 skins in the U. S.
Biological Survey collection on the basis of which he outlined
the range as "the whole of southern Texas north to the mouth
of the Pecos and the mouth of the Colorado and still farther
north along the strip of mesquite country east of the plains,
approximately covering the semiarid part of the Lower Sonoran
zone." He mentions a specimen from Matamoras, Mexico.
In this area these wolves are destroyed by the ranchmen, since
they kill young cattle, goats, and colts. The result is that at
the present time this, the typical race, may now be extinct
(Goldman) .

In Audubon's day "wolves" were common in Kentucky and
were usually black. They may or may not have represented a
race of the red wolf, but in his paper of 1937 Goldman definitely
names the wolf of the lower Mississippi region a race of it,
Canis rufus gregoryi. He characterizes it as a "large but slender
form of a small species . . . decidedly larger and grayer,
less tawny" than C. r. rufus, with a more slender skull and
lighter dentition than the Florida race. The middle and sides
of the face are mixed black and gray, changing to black and
cinnamon -buff on top of head ; upper parts from nape to rump
light buff, heavily mixed or overlain with black; outer surface
of legs between cinnamon and cinnamon-buff, paling on the
feet. Black individuals occasionally occur.

Of this race, Goldman has examined over 150 individuals
from the lower part of the Mississippi Valley. He states that

NORTH AMERICA AND THE WEST INDIES 231

it is found mainly on the western side of the river in south-
eastern Missouri, Arkansas, southeastern Oklahoma, eastern
Texas, and Louisiana. It grades into the typical form in Texas
and apparently into the eastern race floridanus in northern
Alabama. Arthur (1931) publishes a photograph of a captive
Louisiana individual which was black and shows the relatively
long coyotelike ears. He says that it is found in the prairie
and marsh sections of Louisiana, preying on rabbits, native
rats and mice, squirrels, ground-nesting birds, and fawns, as
well as on domestic calves, sheep, and hogs. The young are
born in January and February and number from three to a
dozen to a litter, averaging about half a dozen, so that the rate
of increase may be fairly rapid. "While the wolf has been
persistently hunted by man in Louisiana ... it appears
to be on the increase, and, slowly but surely, extending its
range in the state ... In La Salle parish . . . the
wolves have become very obnoxious because of their depreda-
tions on live stock, and cattle men in West Feliciana parish now
fear an increase of their number as the cattle and sheep raising
business in this former cottonraising territory is growing in im-
portance . . . It is not unlikely, therefore, that in a few
years stringent and systematic campaigns against wolves in
this state will have to be planned. " Similar reports of increased
numbers are credited by Bellinger and Black (1940), who
write that " wolves are becoming rather common in the Ozarks "
with records of several, including two black wolves, taken in
late years in the northwestern and western part of Arkansas.
Stanley P. Young (1940) finds that in eastern Texas the red
wolf is abundant; more than 800 were caught during the pre-
vious year (1939) on account of livestock depredations. He
believes that on account of its habit of living in thickets it will
not easily be exterminated.

The Florida wolf is now regarded as a race of the red rather
than of the gray wolf, from which it differs presumably in much
the same characters as does C. rufus gregoryi, but it has ap-
parently a much stouter skull and dentition. Black was the
usual color, but no doubt a mixed phase also occurred. At the
present time this wolf is probably extinct in Florida, as well as
in southern Georgia, which formerly perhaps marked its north-
ward range. Dr. Francis Harper (1927) has reviewed its status
in the region of the Okefenokee Swamp, where the last one actu-

EXTINCT AND VANISHING MAMMALS

ally killed was about 1908, some 10 miles north of Fargo.
Later accounts of what may very likely have been wolves in
this region point to its survival there perhaps as late as 1918.
In peninsular Florida, wolves were present in the Everglades,
according to Cory (1896), up to the middle nineties. Miller in
his original description of the race makes as the type a specimen
obtained by the U. S. National Museum on the St. Johns
River, August 12, 1890. This and a skull taken by Dr. Henry
Bryant many years before, and now in the Museum of Com-
parative Zoology, seem to be the only extant specimens from
the State. In Bradford County they were reported up to 1895
(V. Bailey, 1907).

A. H. Ho well, writing in 1921, says that in former times
wolves probably ranged over the entire State of Alabama, but
are now on the verge of extinction. In Bartram's time, late in
the eighteenth century, they were common, roaming the moun-
tainous areas in small packs and preying considerably on the
smaller domestic animals, sheep, goats, pigs, and sometimes
calves. On the Gulf coast, "strangely enough, in the big
swamp country in Baldwin and Mobile counties, where deer are
still numerous, wolves were apparently exterminated many
years ago; the last one of which there is record was reported
killed near Carlton about 1894." During the second decade
of the present century wolves were still present in small num-
bers in northwestern Alabama in the "rough, hilly country
stretching from Walker County northwestward to Colbert
County." Here three or four were killed in 1912 near South
Lowell, while in 1915 wolves were destructive to stock in west-
ern Cullman County, afterward moving westward into Winston
and Marion Counties, killing "thousands of dollars' worth of
sheep and goats" and some calves. In 1917, a wolf was killed
in Colbert County and is described by Ho well in some detail.
It is the latest record mentioned by him. Goldman says of
this specimen that it is somewhat intermediate between typical
floridanus and gregoryi, but in its heavy dentition is nearer the
former.

There is some evidence that this small type of wolf formerly
ranged northward into Tennessee. Kellogg (1939) refers to the
race floridanus a right mandible excavated from an Indian
mound near Citico Creek, Hamilton County, and thinks it
quite likely that this form ranged over southeastern Tennessee

NORTH AMERICA AND THE WEST INDIES

at the time of the first white traders. Possibly, too, the form
gregoryi was the wolf formerly found in the extreme western
part of the State. At all events, he quotes Benjamin C. Miles
that the small black wolf was exterminated about 1870 in
Haywood and Lauderdale Counties, while Major Shaw is au-
thority for the statement that it had been replaced in later
times by the larger and fiercer gray wolf.

Family FELIDAE: Cats

THE PUMAS

The pumas, or mountain lions, are typically American and
constitute a single species having a wide latitudinal range from
the northern United States and southern Canada southward
in forested country to Patagonia. In South America they are
spoken of by people of Spanish descent as "leon" in contra-
distinction to the "tigre," or jaguar. Other names are
"puma," used in Peru, and "couguar, " said to be a corruption
of a Tupi word signifying an animal of the same color as a deer.
For pumas are associated with Seer, on several species of which
they prey in different parts of their vast range. In disposition,
these cats are not especially fierce in spite of many tales told by
frightened persons, and very seldom indeed do they attack
human beings. In South America they are not feared, the
jaguar being held in high respect. Slightly smaller than the
latter, the pumas are characterized by a uniform coloration of
tawny shades above, whitish lips and eye-stripes, and a black
spot on each side of the muzzle. With a total length of about
8 feet for an adult male, and a weight ranging up to some 200
pounds, there is a wide local variation in size, shade of color,
and in cranial details, which has led to the distinction by Nelson
and Goldman (1929), the last to review the matter, of no less
than 19 subspecies, of which 9 are South American, including
the typical race of southern Brazil, and 10 are found in North
and Central America. The tropical races tend to be smaller
and brighter in color, becoming a distinct reddish in tone.
The southernmost races are paler, the race pearsoni of Pata-
gonia being a silvery gray. In the United States, at the more
northerly part of their range, pumas may show two color
phases, a redder and a grayish brown, sometimes spoken of as

234 EXTINCT AND VANISHING MAMMALS

red and blue, corresponding to the color phases of deer, though
not seasonal as in the latter.

While in South America there seems as yet little to indicate
a widespread reduction in numbers of any of these races, except
in the more settled regions, in North America pumas have been
exterminated over most of their range in the eastern United
States; in the West, where ranching interests are paramount,
they are locally gone, and will doubtless ere many years become
few and restricted to large areas of forest country or to national
parks, where with proper measures their numbers can be kept
under control. Nevertheless, in other areas a good many yet
remain and probably will continue for a long time to come.
The North American races are here considered in order.

EASTERN MOUNTAIN LION; "CATAMOUNT"; "PANTHER"
FELIS CONCOLOR COUGUAR Kerr

Felis couguar Kerr, Linnaeus's Anim. Kingdom, p. 151, 1792 ("Pennsylvania").
FIG.: Nelson, 1916, p. 413, upper fig. (colored).

Originally the range of the eastern mountain lion extended
from central New England westward to the edge of the Plains
and southward through the forested country to Georgia and
northern Alabama, where presumably it merged with that of
the Florida puma. Over this region it is at the present time
probably extinct, with the possibility, however, that a very few
may remain in the southern Alleghenies. Its range coincided
more or less with that of the Virginia deer, which formed its
main object of prey.

The eastern mountain lion is described as dark reddish
brown, darker along the middle of the back, the tail similar,
becoming blackish at the tip, and the feet dark also. The
muzzle is whitish, with a black marking just back of the white
and in advance of the eye ; a small white line in front of the eye ;
under side paler, yellowish white. A grayer phase probably
occurred as well. Length of body when full grown, about 6
feet, tail about 3 feet; weight about 175 pounds for a Vermont
specimen. Smaller animals, perhaps females, measure less,
about 7 feet over all, with a weight of 118 pounds (Vermont
specimen) .

The catamount, or panther, as these big cats were usually
called in New England, was not especially common even in

NORTH AMERICA AND THE WEST INDIES 235

early days but was nevertheless evidently troublesome to the
first white settlers on our shores, for bounties were offered and
many were killed though with little record. At first they were
thought to be young lions, and one is so reported by John
Josselyn at Cape Ann in earlier times. As early as 1694 Con-
necticut offered a bounty of 20 shillings apiece for catamounts
and as late as 1769 paid for four or five. Massachusetts first
offered a reward of 40 shillings for killing these animals in 1742.
In 1753 this was increased to 4. Probably by the time of the
Revolution panthers were fairly well gone from the three
southern States of New England. Wood, in his early account
of New England, tells that "Plymouth men have traded for
Lyons' skins in former times, " but they seem to have been few
in eastern Massachusetts and were early driven back. In the
Boston Gazette of April 20, 1741, is a notice of a strange animal
exhibited at the Greyhound Tavern in Roxbury that had been
"caught in the woods about 80 miles to the westward," hence
in western Massachusetts. "It has a tail like a Lyon, its legs
are like Bears, its claws like an Eagle, its Eyes like a tyger."
It was "called a Cattamount." Judd, in his "History of
Hadley, Massachusetts," 186^?, mentions that one was killed
by some Northampton hunters in 1764, and others were said
to have been shot in later years in Hampshire. Emmons in
1840 regarded the species as then extirpated in Massachusetts,
yet one was killed in 1847 or 1848 in West Greenwich, R. I.,
was mounted and for many years was preserved in the museum
of the Providence Franklin Society, and then was secured by
the Boston Society of Natural History, in whose museum it
still is. This seems to be the last record for the State. In
Connecticut it disappeared at about the same time, for Linsley
(1842), writing of the mammals of Connecticut a century ago,
had no later record of it than of one he saw some years previ-
ously that had been killed in the northern part of the State.
Panthers were occasional in the southern half of Maine during
the early days of settlement but were practically gone by 1815,
although what may have been the last one killed in the State
was shot about 1845 at Sebago (Norton, 1930). From time to
time there have been rumors of panthers in the wilder parts of
Maine even down to the present day, but it seems safe to say
that these reports deserve little credence. In southern New
Hampshire panthers were occasionally killed, and there are

236 EXTINCT AND VANISHING MAMMALS

several accounts in local histories of such events in the latter
part of the eighteenth century, but the last one actually cap-
tured and still preserved seems to be the specimen now in the
possession of the Woodman Institute, Dover, N. H., which was
shot on November 1, 1853, in the town of Lee. I have one or
two later accounts of panthers seen at short range in the
White Mountains as lately as about 1880, the details of which
are so circumstantial that they seem quite credible. Vermont,
however, was the best country for these animals, and in the
days of settlement previous to the American Revolution they
were often hunted and killed. At the present time there are
extant perhaps three specimens, possibly four, that were killed
in the State. The last one was taken in the town of Barnard
in 1881 and is in the State Museum at Montpelier. Since then
there have been a few other reports, some of which seem
credible, but probably the species has ceased to exist in Ver-
mont for nearly 50 years

The wilder parts of New York State were formerly good
country for panthers. In DeKay's time, a century ago, there
were still a few in the Catskills, and in his book of the mammals
of New York he mentions that as a boy he recalled the appear-
ance of one in Westchester County within 25 miles of the city
of New York and was also informed of one that had been killed
in Warren County. Merriam has given an account of the
species in the Adirondacks in the latter part of the last century.
Here they were then present in some numbers, and when in
1871 a bounty was offered on them no less than 46 were re-
ported killed in the succeeding decade, about half of them in
St. Lawrence County. Merriam (1882) comments that in
1882, owing to the incentive of the bounty, they had been
killed down to nearly the point of extermination, yet Miller
(1899) continues the bounty tabulation down to 1890, with an
additional 107 panthers on which $20 each had been paid.
Since the last date no further bounties were claimed up to
1897, when a panther, now mounted at Albany, was killed on
Sumner Stream. The last bounty was paid on one killed
December 27, 1899. Since that date, however, there were oc-
casional reports of panthers from the Adirondacks, the last in
1903, said to have been seen at Big Moose, though it seems
safe to say that they were virtually extinct in the State by
about the close of the century.

NORTH AMERICA AND THE WEST INDIES 237

To the westward there were formerly a few panthers in
southern Ontario and to the northward they extended into ex-
treme southern Quebec. But the Quebec records, according to
Seton, were long ago: Sherbrooke, about 1840; and near Sorel,
October 3, 1863; and there is one old record for the Toronto
region of Ontario. Ohio never seems to have had many of
these animals, or else they were early killed out. Kirtland, in
the Ohio Geological Survey for 1838, wrote that they had al-
ready disappeared but that there were Ohio specimens in the
museum of a Mr. Dorfeuille at Cincinnati. The bounty of $4
for a grown panther killed in Ohio was discontinued in 1818.
The last record for the State is perhaps of one killed in 1805
near Newark (Brayton, 1882). For Indiana, Lyon (1936)
writes that they were gone from the southern part of the State
at the time of the Prince of Wied's sojourn there in 1832-33
but that they persisted a little longer in the northern part.
A few final records take the panther to about 1838, with
additional vague reports up to about 1850, when Hahn be-
lieved it had become extinct in Indiana. Cory (1912) lists
several later occurrences in Illinois: the last one killed in Ma-
coupin County about 1840; one killed in Alexander County
about 1862; one in Daviess County in 1840. In Wisconsin one
was shot on the headwaters of Black River in December 1863,
and one was killed in Vernon County, near Westly, about 1870.
Possibly a few remained into the next decade. What is perhaps
the only extant specimen from this State is one recorded by
Schorger (1938) that was killed at Appleton, Wis., November
22, 1857, and is preserved at Lawrence College. The last
record of panther in Minnesota is of one killed at Sunrise,
Chisago County, in 1875 (Herrick).

According to Bailey (1926) this species must once have
ranged over much of the Dakotas, but it is not mentioned by
Alexander Henry and his trappers in the early years of the last
century. After detailing a few records for western North
Dakota up to the late years of the nineteenth century, Bailey
mentions a report of a pair seen at Sully s Lake in 1907, and
comments: "It is not improbable that a few may still lurk in
the very rough Badlands country in the western part of the
State, but it is more probable that the last record for the State
has been made." He refers the Dakota panther to "Felis
hippolestes," the type locality of which is Wyoming in the

238 EXTINCT AND VANISHING MAMMALS

Wind River Mountains, but it seems likely that those formerly
believed to occur in the eastern parts of the Dakotas were of
the race couguar.

Summarizing the former presence of the mountain lion in
New Jersey and Pennsylvania, Rhoads (1903) says that though
originally found in every part of both States it was always
more plentiful in the Allegheny Mountains. As early as 1697,
on account of its destructiveness to stock, a bounty was placed
on this animal in New Jersey, amounting to 20 shillings for
"whatsoever Christian shall kill and bring" in the head, or
half that sum to a Negro or an Indian. In 1730, this bounty
was reduced to 15 shillings, and evidently it had been effective,
for the species seems to have been extirpated by about 1830
or 1840. In the wilder parts of Pennsylvania, however, the
mountain lion persisted till much later. In Audubon and
Bachman's time, about the middle of the last century, it was
so common "among the mountains of the headwaters of the
Juniata River . . . that one man has killed for some years
from 2 to 5, and one very hard winter, 7." The latest date
when one was killed in Pennsylvania was, according to Rhoads,
about 1871, although later reports of two panthers killed in
Treaster Valley take the date down to 1893, and may be
trustworthy. Three subsequent reports of tracks and even an
animal seen are as recent as 1913 (Shoemaker, 1914). It is not
at all impossible that panthers may still survive in the wilder
parts of the West Virginia and Tennessee Alleghenies. Kellogg
(1937, 1939) has summarized the available information on this
matter and finds that although numerous enough at the time
of settlement to give the pioneers some trouble, their numbers
had been much reduced by the middle of the last century. He
cites statistics showing that in West Virginia 11 were killed in
1853, 14 in 1856, 11 in 1858, and 6 in 1859. The last record
for Lewis County was in 1855, and there is a skeleton in the
U. S. National Museum from Hampshire County, taken in
1850. Although no record of any having since been killed is
given, Kellogg includes various reports of late date, such as
"tracks of a panther in the snow on Black Mountain during
the winter of 1935 and also in 1936"; but all such reports
should be received with caution, even though made by persons
who might be capable of correctly identifying the species.
Shoemaker (1914), however, states that one was killed in No-

NORTH AMERICA AND THE WEST INDIES 239

vember, 1913, several miles north of Washington, D. C. The
panther records assembled for Tennessee relate mostly to
earlier days, with some evidence that they persisted up till
about 1875, or slightly after. Merriam, who traveled through
the Great Smoky Mountains in 1888, concluded as a result of
his inquiries that it was at that time unknown to the local
inhabitants. Nevertheless, Kellogg accepts a report of one
said to have been killed in 1929 in the Holston Mountains,
Johnson County, and another of one "seen crossing the trail
on Roan Mountain on September 18, 1937," without adducing
further evidence. C. S. Brimley (1905) records that the last
authentic report for North Carolina is of one killed near Rose
Bay, Hyde County, some years before the Civil War. Prob-
ably the range extended slightly farther south into the northern
parts of Georgia and Alabama and then merged with that of
the lowland Florida race. Ho well (1921) writes that in the
latter State (Alabama), although it is now "nearly, if not
quite, exterminated . . . recent reports, although rather
indefinite, indicate that a very few may still remain in the big
swamps of the southern counties." If these reports may be
credited, it is likely that they Vefer to the Florida form, and
Dr. Francis Harper regards the animal of southern Georgia as
the same, giving records of its presence up till at least the early
years of the present century.

From the foregoing brief survey, it is clear that the eastern
panther, though fairly common for so large a beast of prey,
was exterminated from the more settled parts along the
Atlantic coast in the colonial days, while in the wilder and more
mountainous regions, as in Vermont and New Hampshire, the
Adirondacks, and the Alleghenies, it persisted in some numbers
till about the middle of the last century, after which the last
surviving scattered individuals were gradually shot or trapped
until they were exterminated, or practically so, by the last
years of the nineteenth century. Later reports, though in
some cases perhaps credible, must nevertheless be received
with skepticism for, as shown by Cory in his (1912) account of
the animal, even a trained eye and a person familiar with the
appearance of the animal may sometimes be deceived. The
testimony of those who report "tracks" requires careful sub-
stantiation, since few hunters or trappers in the East have
ever had opportunity to examine these, much less to recognize
the animal in the wild.

240 EXTINCT AND VANISHING MAMMALS

FLORIDA COUGAR
FELIS CONCOLOR CORYI Bangs

Felis coryi Bangs, Proc. Biol. Soc. Washington, vol. 13, p. 15, Jan. 31, 1899 ("Wilder-
ness back of Sebastian, Brevard County, Florida").

SYNONYMS: Felis concolor floridana Cory, Hunting and Fishing in Florida, p. 109, 1896
(not Felis floridana Desmarest, 1820); Felis arundivaga Hollister, Proc. Biol. Soc.
Washington, vol. 24, p. 176, June 16, 1911 ("12 miles south of Vidalia, Concordia
Parish, Louisiana").

The puma of the Florida Peninsula and the adjacent parts of
southern Georgia, Alabama, and Louisiana is a brighter-
colored animal and smaller than the eastern puma, of a "rich
ferruginous or intense rusty red" above, with smaller feet.
Hollister, in describing the Louisiana puma as distinct, sepa-
rated it partly on the basis of its color, which is a "grayish
fawn . . . with a decided cast of ecru drab," especially
on flanks and legs. Since, however, Nelson and Goldman
(1929) have relegated this name to the synonymy of the
Florida puma, it seems evident that his description applies to
the grayer color phase of the same animal. Hollister describes
the skull as "much larger than the skulls of F. couguar, with
well developed crest and much larger nasals" and ear bullae.
The condylobasal length of the skull in an adult male described
by this author was 193.5 mm.; greatest length of nasals, 63.

The Florida puma, though fairly common in many parts of
Florida and the adjacent States at the end of the last century,
has apparently become much reduced in numbers of late years.
Cory (1896), in his book "Hunting and Fishing in Florida,"
has much to say of his experiences with this animal. He speaks
of it as "still not uncommon in the more unsettled portions of
the State" on both the east and the west coasts. He himself
had found tracks of seven in one week within 30 miles of Lake
Worth in southern Florida and mentions that one John Davis
had killed six in that region in the season of 1895. The Indians
reported to him that they were also "numerous" in the vicinity
of the Big Cypress south of Fort Myers on the western side of
the peninsula. Farther north they were even then less common
and were "scarce" on the peninsula east of the Indian River,
"but were common there a few years ago. " He recounts that
in the eighties a hunter named Quarterman killed several in
the vicinity of Canaveral, once making a double shot at two
old males that he discovered fighting. Outram Bangs, in 1898,

NORTH AMERICA AND THE WEST INDIES 241

secured several through a hunter who obtained them in the
wilderness back of Sebastian and in his paper on Florida mam-
mals writes (1898) that though the exact range could not then
be given, "the puma is extinct in all the region directly north-
east of Florida, and I believe in northern Florida as well."
At the present time there are still a number of pumas in the
Everglades, where they live largely on deer, but probably they
have been extirpated from most other parts of the State.

Dr. Francis Harper (1927) has gathered a mass of notes as
to the presence of the puma in the region of the Okefenokee
Swamp in southern Georgia and refers the animal to this form.
Here he says "it is now very nearly if not entirely extinct. Yet
it lingered well into the present century, and it is perhaps not
beyond the bounds of possibility that some solitary survivor
may yet be taken. " At all events, he records one killed about
1883 and one seen alive about 1903. Older residents of the
region supplied other instances of animals killed about 1885.
Most of the later reports are of tracks seen or other signs,
with a final record of one said to have been killed about 1925
on the southern edge of the Swamp. The residents in that
region universally speak of it*as the "Tiger." Hunting is
usually done in Florida with dogs. When started, the puma
usually runs a short distance and then takes to a tree. The
dogs keep it at bay until the hunters come up and shoot it.

While probably now quite gone from northern Florida and
Georgia, this race doubtless still occurs in parts of southern
Alabama and the canebrakes of Louisiana. Writing of the
former State, Howell (1921) speaks of it as now "nearly, if not
quite, exterminated"; recent reports, however, "although
rather indefinite, indicate that a very few may still remain in
the big swamps of the southern counties. " Tracks were seen
near Scale about 1912 by an old trapper, and another was re-
ported as actually seen about 1905 in Baldwin County, but
these are rather questionable bits of evidence. In 1911 Hoi-
lister wrote that pumas "are still fairly common in the wilder
parts of the cane brake region of eastern Louisiana," and he
himself while hunting in the Bear Lake Cane in February,
1904, several times heard panthers "crying."

Panthers formerly were found in somewhat similar country
in Arkansas, and True (1889) reports a few killed or seen up to
the late eighties, but it is not certain that they were of this race.

242 EXTINCT AND VANISHING MAMMALS

At the present time what few may remain are probably to
be found in the Florida Everglades and in southern Louisiana.

TEXAS COUGAR
FELIS CONCOLOR STANLEYANA Goldman

Felis concolor Stanley ana Goldman, Proc. Biol. Soc. Washington, vol. 51, p. 633, Mar.

18, 1938 ("Bruni Ranch, near Bruni, southeastern Webb County, Texas").
SYNONYM: Felis concolor youngi Goldman, Proc. Biol. Soc. Washington, vol. 49, p. 137,

Aug. 22, 1936. (Preoccupied by Felis youngi Pei, for a fossil Chinese species.)

Goldman has lately distinguished the cougar of southern
and central Texas and northeastern Mexico as larger than
azteca, with lighter upper parts of a more grayish tint, black on
tail usually more restricted, the dentition heavier, the nasals
depressed. It is smaller and paler than hippolestes to the
north. Total length of the type, 2,134 mm.; greatest length of
skull, 220 mm. A series of specimens examined by Goldman
represents 16 localities in Texas, and there were also two from
Matamoros, Tamaulipas, across the border in northeastern
Mexico. On the east it presumably intergrades with the race
coryi of Louisiana and Florida. Its general status may be
considered together with that of its near ally:

MEXICAN COUGAR

FELIS CONCOLOR AZTECA Merriain

Felis hippolestes aztecus Merriam, Proc. Washington Acad. Sci., vol. 3, p. 592, Dec. 11,
1901 ("Colonia Garcia, Chihuahua, Mexico").

The characters claimed for this cougar are its large size,
though slightly smaller than the Montana hippolestes, "general
color dull fulvous [as in the latter] ... but tail darker,
browner, with longer black tip and no white underneath"; ears
almost wholly black. The skull is large and massive, though
not equalling that of hippolestes, but shorter and the teeth
somewhat smaller. The total length of an adult male, the
type, was 2,268 mm.; tail vertebrae, 731; basal length of skull,
171.5. It is thus a somewhat larger, duller animal than browni.

The range of this race is believed to extend from the southern
border of the United States in Arizona and New Mexico, south
into Mexico for an undetermined distance, but probably over
the tableland in at least the northern half. Of its present
status in Mexico and Arizona, little information is at hand.

NORTH AMERICA AND THE WEST INDIES 243

Bailey, writing in 1905, says: "In most of eastern Texas
panthers are reported as formerly common, but now as very
rare or entirely extinct. Individuals have been killed, however,
within a few years in the swamps not far from Jefferson in the
northeastern part and Sour Lake in the southeastern part of
the State. At Tarkington Prairie Mr. A. W. Carter says there
were a few panthers when he was a boy in 1860, but he has not
seen one since. In the Big Thicket of Hardin County a few
panthers have been killed in past years, and Dan Griffin, who
lives 7 miles northeast of Sour Lake, says a very large one
occasionally passes his place. He saw its tracks in the winter
of 1903-4 ... In the rough and sparsely settled western
part of the State mountain lions are still fairly common in
certain sections, where they often lay a heavy tribute on colts,
calves, and sheep." In the region about Langtry they were
reported as common a few years previously, but in 1903 were
become already scarce; and there were a few at that time in the
Franklin Mountains, where they annually killed numbers of
colts. "In the Davis Mountains these cougars have been
hunted with hounds till scarce, but in the Santiago Range, in
the Chisos Mountains, and along the canyons of the Rio
Grande and Pecos they" were still common in 1905 (Bailey).
Even at the present day a few cougars still trouble the cattle
ranchers in this region, but are constantly being hunted down.
Of their future prospects, Bailey writes: "The rough desert
ranges, full of canyons, cliffs, and caves, are the favorite haunts
of the panthers, and will be their last strongholds, not only
because of the advantages they offer for foraging but because
of the protection they afford from hounds and hunters. " In
extreme southern Texas they were, according to Attwater in
1896, "not as scarce as the Jaguar in the country west of San
Antonio, but they are fast becoming killed out" (J. A. Allen,
1896). He knew of but two records in recent times that he
considered trustworthy; these, in Bexar and Kerr Counties,
respectively, were sight records made in 1893 and 1894.

Mountain lions probably referable to the race azteca "are
or have been common over practically all of New Mexico, but
they are rapidly decreasing in numbers and in 1931, writes
Bailey, "are rare or absent from most of the open plains
country, but are still found in many of the rough or timbered
mountain ranges, which afford them cover and game." In a

244 EXTINCT AND VANISHING MAMMALS

report for 1917, J. S. Ligon had killed during the year 84 and
estimates that there were still about 400 left alive in New
Mexico. "At the present time they are probably most com-
mon in the Mogollon Mountain region, and in the Animas,
San Luis, and Sacramento Mountain Ranges, with a few
scattered through some of the small desert ranges over the
southern and western parts of the State." The continual
hunting down of these animals by ranchers and by government-
paid hunters can not fail within a few years more to reduce
their numbers to very small proportions. Bailey (1931) quotes
the following figures for panthers killed by hunters in govern-
ment employ in New Mexico: In 1917, 17; in 1918, 14; in 1919,
41; in 1920, 63; and in 1921, 29, a total of 164 in five years.
With such a determined campaign waged against them, it
seems likely that the species will soon be largely wiped out over
much of the State. When one locality is rid of them, they may
come in from surrounding territory, for they are great travelers.
Bailey mentions a case where a hunter followed a track for two
days and estimated that the animal had covered 30 miles
without making a kill or stopping for any length of time.
They may have as many as six young at a litter but usually
four or even two; the gestation period was 96 days in the case
of a captive individual (J. A. Allen, 1896). Hence animals
killed in early or mid winter have probably not yet bred, but
with females shot late in winter the potential increase is like-
wise destroyed.

YUMA MOUNTAIN LION; COLORADO DESERT PUMA
FELIS CONCOLOR BROWNI Merriam

Felis aztecus browni Merriam, Proc. Biol. Soc. Washington, vol. 16, p. 73, May 29,
1903 ("Lower Colorado River 12 miles south of Yuma, Arizona").

Very little is known concerning this puma. Grinnell, Dixon,
and Linsdale (1937) describe it as "closely similar to the Cali-
fornia Mountain Lion but with shorter pelage, paler tone of
coloration, and smaller skull and teeth." Merriam (loc. cit.)
in describing the original specimen emphasized similar charac-
ters in comparison with the more eastern race, azteca, calling
attention further to the smaller and lower audital bullae, small
canines and carnassials. Merriam suggested that the slender
canines indicated that the animal fed upon smaller game than

NORTH AMERICA AND THE WEST INDIES 245

do the larger neighboring races. The type specimen was said
to have measured in the flesh 7 feet 4 inches from tip of nose to
tip of tail, of which the latter was 28.5 inches. It weighed 170
pounds (perhaps an estimate). Greatest length of skull, 198
mm.

This small race of the puma was probably confined to the
desert regions of the lower Colorado River in southeastern
California and the adjacent parts of Arizona, where, according
to the authors above quoted, it "is now rare, perhaps even near
to extinction." They add that in the "spring of 1910, while
exploring the Colorado River from Needles to Yuma," they
obtained evidence, both direct and through report, of the pres-
ence of mountain lions about Riverside Mountain in the ex-
treme northeastern part of Riverside County, California. The
animals here "appeared to range chiefly through the densely
wooded bottom lands, though one was reliably reported to
have been seen among the rough desert hills which constitute
'Riverside Mountain'." They saw a fresh track on May 1,
1910, four miles below Potholes, on the California side of the
River, and had further reports from Ehrenberg, Cibola, and
Potholes; and secured two hides and skulls from a rancher 18
miles north of Picacho who had shot the animals the previ-
ous autumn. These pumas had appeared in the region where
none had been seen for ten years, and one by one killed off
eight of the rancher's hogs. Finally one was hunted down and
shot. "Meanwhile the hogs had become thoroughly frightened
and had taken to swimming the river twice daily to forage for
mesquite beans on the Arizona side, where they appeared to
feel safer. " A second lion continued the depredations and was
eventually shot. Inquiry of the rancher 12 years later brought
out the fact that he had seen no more pumas since that event.
In 1909, a pair of mountain lions appeared near Calexico and
were occasionally seen by the ditch tender swimming the main
Imperial Canal at that point. These lions lived on pigs that
had gone wild. Since 1910 no information is available from
California as to the status of this race, and it is doubtful if it
still exists.

Outside of the immediate delta region of the Colorado, it is
likely that this subspecies was found in northern Lower Cali-
fornia. Nelson (1921) includes it without comment in his lists
of the fauna of the region. Probably it was closely similar to

246 EXTINCT AND VANISHING MAMMALS

the race improcera of the central and southern parts of the
peninsula. How far it may have ranged into the adjacent part
of Sonora is also unknown, but it may have already become
nearly extirpated.

LOWER CALIFORNIA PUMA

FELIS CONCOLOR IMPROCERA Phillips

Felis improcera Phillips, Proc. Biol. Soc. Washington, vol. 25, p. 85, May 4, 1912

("Calmalli, Lower California").
FIG.: Phillips, 1912, pi. 5 (skull).

Little seems to be known of this puma. The adult male
upon which the subspecies was founded is small, with a skull
of only 150 mm. in basal length, and of a rounder, less elongate
form than in neighboring races. The color is "dark fawn,"
darker along the back, the general hue less reddish than in
F. c. azteca.

This small puma is believed to be characteristic of central
and southern Lower California. Calmalli, the type locality,
lies in the Vizcaino Desert district, which occupies the middle
third of the peninsula and includes the great lava plateau,
with scattered mountains, low ridges, and groups of granite
mountains. Two remarkable plants are largely characteristic
of this region, the "cirio, " which forms polelike forests, and
the " elephant wood, " with fantastic massive trunks and small
branches. There is abundance also of giant and other cactuses.
In this arid region of the Lower Sonoran desert and in the Arid
Tropical area of southern Lower California, this little puma is
found. Practically nothing is recorded of it. Dr. Phillips be-
lieved it "probably a rare animal throughout the entire length
of the peninsula, " for he adds that he is informed by W. W.
Brown, Jr., that it occurs "even south to the vicinity of Cape
St. Lucas. " On account of its restricted habitat and the barren
country in which it lives, its numbers can not be large and are
not likely to increase. Unless changes come in this habitat, it
is likely to remain as it is except as it comes in contact with the
scattered ranches and interferes with stock raising. At the
present time there is no evidence at hand.

NORTH AMERICA AND THE WEST INDIES 247

ROCKY MOUNTAIN PANTHER OR COUGAR
FELIS CONCOLOR HIPPOLESTES Merriam

Felis hippolestes Merriam, Proc. Biol. Soc. Washington, vol. 11, p. 219, July 15, 1897
("Wind River Mountains, Wyoming").

OREGON COUGAR
FELIS CONCOLOR OREGONENSIS Rafinesque

Felix (sic) oregonensis Rafinesque, Atlantic Journ., vol. 1, p. 62, 1832 (Northwest coast

of United States).
SYNONYM: Felis hippolestes olympus Merriam, Proc. Biol. Soc. Washington, vol. 11, p.

220, July 15, 1897 (Lake Cushman, Mason County, Washington).
FIGS.: Elliot, 1904, pt. 2, pis. 44, 45 (skull).

CALIFORNIA COUGAR
FELIS CONCOLOR CALIFORNICA May

Felis calif ornica May, California Game "marked down," p. 22, 1896 (Upper Kern

River, Kern County, California).
FIGS.: Grinnell, Dixon, and Linsdale, 1937, pi. 11, col. (showing red and gray phases),

figs. 215-239 (skull, photographs of old, young, habitat, etc.).

VANCOUVER COUGAR
FELIS CONCOLOR VANCOUVERENSIS Nelson and Goldman

Felis concolor Vancouver -crisis Nelson and Goldman, Proc. Biol. Soc. Washington, vol.
45, p. 105, July 15, 1932 (Campbell Lake, Vancouver Island, British Columbia).

The present status of these races may be considered briefly
together, for although none of them seems now in actual danger
of extermination, nevertheless in the years to come, with the
constant encroachment of human occupation, with hunting for
sport or for bounties, and with forest fires and other causes,
it is likely that their ranks will eventually dwindle, and that
they can survive in any numbers only in large reserves or in
country that remains in a more or less primeval condition,
with shelter of forests and food in the shape of larger game,
meaning chiefly deer. The fallacy of completely killing out
these larger predators in deer reserves has been several times
demonstrated. In the case of the Kaibab Forest, from which
panthers and wolves were eliminated by purposeful hunting
and trapping, and from which deer hunters were excluded, the
mule deer in the course of a few years became so abundant
that not only did the animals destroy all the available food but

248 EXTINCT AND VANISHING MAMMALS

also permanently injured the sparse forest of the area so that
hundreds died from starvation. A nearly similar state of affairs
took place in the Murderers Creek Game Refuge of eastern
Oregon, which was established in 1929 in the belief that the
deer there needed protection (Englis, 1939). In the course of
some years the mule deer so increased that in one section
12,000 deer were trying to exist "where even 6,000 would have
gone hungry. " Bitterbush, mountain mahogany, and juniper,
the chief winter food plants, were so overbrowsed that bitter-
bush became mere stubs, and the junipers were trimmed as
high as a deer could reach. Many deer starved to death and
the others were small and undernourished. In 1935 the area
was reopened to hunting and the refuge given up. In such
cases it is clear that a proper proportion (to be determined) of
panthers would not only keep the deer down to a number
commensurate with the carrying capacity of the range, but also
afford a certain return from fur, besides adding to the human
interest of the situation, in hunting or observation. This view
of the value of predators has only in recent years been brought
out forcibly as a result of the experiments in "control" that
have been more or less intentionally carried out by federal and
other agencies in our West during the past 60 or 70 years. It
now becomes clear that "predatory animals are to be con-
sidered as an integral part of the wild life protected within
national parks, and no widespread campaigns of destruction
are to be countenanced, " but native predators may be allowed
to continue their normal utilization of other park animals, unless
in special cases where the need for some regulation becomes
obvious (Cahalane, 1939b). This logical and scientific policy,
based on an accurate knowledge of the various species con-
cerned, should become increasingly established over all the
larger areas of unsettled country.

Of the four races listed above, Felis c. hippolestes is the largest
member of the species; it is of a dull fulvous color and has a
large massive skull with a highly developed sagittal crest. A
large male measured 8 feet in total length and weighed 227
pounds (Merriam, 1901, p. 586). The skull measured 196 mm.
in basal length. Young are born in January and February in
Colorado. This race ranges from extreme northern New
Mexico (probably) northward through Colorado and the ad-
jacent parts to Wyoming, Montana, and British Columbia.

NORTH AMERICA AND THE WEST INDIES 249

Bailey (1931) refers to this race a large specimen in the collec-
tion of the U. S. Biological Survey from the Jemez Mountains
of northern New Mexico and believes that other dark-brown
skins from Taos and Pecos River Mountains of that region are
the same. In the latter district they were "common" in 1903,
and a few were killed every year in the Taos Mountains. "In
1910 officials of the Forest Service reported mountain lions as
fairly abundant on the Carson National Forest and as still very
common in the Jemez Mountains. In 1914 they reported 3
killed on the Carson, 3 on the Pecos, and 2 in the Jemez Na-
tional Forest; in 1915, 4 on the Carson and 4 on the Santa Fe
National Forests; and in 1916, 4 on the Carson and 7 on the
Santa Fe National Forests." These large cougars could and
did kill even elk, as well as deer, but as these and "other native
game animals become scarce, the mountain lions turn their at-
tention to domestic stock and seem especially to relish colts,
but if these are not to be found, they take horse meat of any
kind. In spite of the bounty usually paid for their destruction
and the efforts of stockmen and hunters, they have until re-
cently held their own in the rougher parts of the country, but
with the present organized effort it will not be long before they
are sufficiently reduced in numbers to prevent any great
losses" (Bailey, 1931).

In Colorado the range of this race, which formerly inhabited
all the "rough parts of the State, and in early times" was oc-
casionally seen "even well out on the plains along the more
heavily brush-fringed streams," has now become much more
circumscribed. In 1911 Cary wrote that it was already rare
east of the Continental Divide, though holding its own fairly
well in the rough canyon and mesa country of the western and
southwestern parts of the State, where locally the animals
were still sufficiently numerous to be very destructive to stock,
especially young colts. Cary (1911) lists eight regions in the
northern mountains of Colorado in all but one of which they
were said to be from uncommon to "common"; in the Snake
River region where formerly they were common, none remain.
Reports from southern Colorado for ten regions where they are
found indicate that they are much less numerous, though a few
still occur. They may be found locally up as high as 10,000
feet in some of the mountain ranges. At least until recent
years the cougar was of local occurrence in the mountainous

250 EXTINCT AND VANISHING MAMMALS

parts of Utah and Idaho, but its numbers can not be very
great, as a result in part of determined efforts of stockmen and
professional hunters.

Bailey (1918) writes that they were still common in Glacier
Park, northwestern Montana, 20 years ago but were confined
for the most part to the western slopes, where dense forest
cover and abundance of deer offered excellent living conditions.
"Apparently they are still almost as numerous as they were in
1895," when he first went through the region. "In this
region," he adds, "their food consists mainly of the white-tail
deer, which abound on the west slope of the park. " Although
Bailey advocates their destruction, as contributing little to the
tourist interest of the park, since they are seldom seen, never-
theless their value as a check on the over-increase of deer might
well justify the maintenance of a proper number.

In southern British Columbia the cougar is uncommon but
extends into the Peace River district. Cowan in a recent report
(1939) writes of a female killed in March 1937 by Ted Strand
of Little Prairie, and quotes Seton ("Lives of Game Animals")
as to a specimen shot in November, 1921, near the junction of
Cypress Creek and Halfway River. This latter "seems to be
the northernmost record for North America."

On Vancouver Island the panther still occurs in some num-
bers in wilder areas and has lately been distinguished as a sepa-
rate race, characterized by its darker, more rufescent upper parts
and the more elevated frontal region of the skull, in comparison
with the Rocky Mountain and Oregon races. Swarth, in his
paper of 1912, wrote that for an animal of this type it was
abundant "throughout the wilder parts of Vancouver Island,
and frequently seen near many of the smaller towns also."
They are shy and secretive but are often hunted successfully
with dogs. One farmer in the Beaver Creek Valley, near
Alberni, had killed as many as 13 during the previous winter.
In a letter dated July 8, 1939, Maj. Allan Brooks writes me
that "over the whole northern third of this island beavers have
been decimated by cougars of late years, and they will take
many years to recover if they ever do. " Cougars have also
been an active agent in the destruction of deer, concentrating
on the more healthy upland population, on account of the
dying off of deer in the lowlands from infection by liver flukes.
At the present time, therefore, the Vancouver cougar seems in

NORTH AMERICA AND THE WEST INDIES 251

no danger, unless more intensive operations for its destruction
are undertaken to preserve the deer.

Farther south, along the coasts of southern British Columbia,
Washington, and Oregon, occurs the race oregonensis, which is
nearly as large as the Rocky Mountain subspecies but darker
and richer in color, with more black on the tail. A large one
was said to have weighed 150 pounds. Concerning its status
in the State of Washington, Taylor and Shaw (1929) write
that it is "probably more common in the Olympic Mountains
than elsewhere, but present also in the Cascades and Blue
Mountains. Predatory-animal control campaigns are markedly
reducing its numbers." In Oregon cougars were common in
the forests west of the Cascade Mountains up till recently but
at the present time have been so reduced by professional
hunters over the State "that they are no longer a serious
menace to livestock industries" (Bailey, 1936). Some idea of
their numbers in recent years may be gained from the figures
given by Bailey (1936), who states that for the period from
October 1, 1913, to December 31, 1914, bounties totalling
$4,035 were paid on 269 mountain lions killed in Oregon. In
Curry County alone no less th&n 60 were killed, and smaller
numbers in 19 other counties. In the fiscal year 1930, Jewett
reported 17 killed by U. S. Biological Survey hunters in Oregon,
"where they had been reported killing stock or game. While
the number is insignificant, it shows a marked decrease in these
big cats during recent years and that their destruction of live-
stock and game is being well curbed. " It seems clear from these
figures that mountain lions will ere long become uncommon in
the State if this policy is continued. In the wilder sections, it
may be that a certain number of these predators will be valu-
able in keeping down too great an increase of deer.

Grinnell, Dixon, and Linsdale (1937) have published an ex-
cellent summary of our knowledge of the race calif ornica as it
occurs in California. The race is less dark and richly colored
than the Oregon cougar. Both red and gray phases occur,
with an intermediate condition, all three in about equal pro-
portions. Adult males weigh upwards of 165 pounds, with an
average of about 140. It occurs in forested and chaparral-
covered areas mainly in the mountains west of the Great Basin
and desert divides from Oregon to the Mexican line. For the
most part it lives at middle altitudes in the mountains between

252 EXTINCT AND VANISHING MAMMALS

2,000 and 5,000 feet. Young seem to be born in almost any
month of the year, but with a peak in April. The usual number
seems to be two young to a litter in California. The chief
prey is deer, on the increase of which it forms a natural check.
Some valuable data on the abundance and feeding habits of
cougars, based on much field experience, give perhaps for the
first time a fairly accurate picture of the economic status of
mountain lions, at least in California. In good lion and deer
country there is on an average about one lion to a township
(36 square miles), and in a few very favored localities about one
to each 10 square miles. Where deer are most plentiful, as in
national parks and game reserves, there also the mountain
lions tend to be most concentrated. A careful study indicates
that at the time of writing, according to the authors quoted,
the mountain-lion population of California was about 600 and
that the annual toll of these animals upon the deer was roughly
21,600. Yet in spite of this, and of a nearly equal number an-
nually taken by hunters in the State, deer are actually increas-
ing. From 1908 to 1921 the annual kill of mountain lions
averaged 258, and the State even employs a professional hunter
for their reduction. Yet it seems that the species is about
holding its numbers in spite of this toll and a natural mortality.
In regard to stock-killing propensities, it appears that usually
those lions that have learned by experience that this sort of
prey is easily secured are the main ones to be feared, and with
their elimination in grazing areas the trouble is much abated.
The present prospect for the California mountain lion is that
although inevitably it must retreat before the spread of settle-
ment and agriculture, nevertheless in wilder areas it will per-
sist for many years to come and continue to be a source of
slight revenue for hunters and perhaps in places a real asset in
the natural control of deer populations in protected areas. Its
situation is perhaps more favorable than is that of the other
three races of the West, the Rocky Mountain, the Vancouver,
and the Oregon cougars.

THE JAGUARS

In a recent paper Nelson and Goldman (1933) have re-
viewed the systematic relationships of the local races of jaguars
in America. These, the largest of American cats, constitute
but a single species, which inhabits the tropical and subtropical

NORTH AMERICA AND THE WEST INDIES 253

regions from south-central Argentina northward to southern
United States. With the general appearance of a leopard,
having a yellowish ground color marked with black spots or
circles of spots, the jaguar is slightly heavier of build than a
leopard, with a more massive head and relatively shorter tail.
The color pattern is variable, but usually at least some of the
rosettes of black spots have a central spot in the jaguar, which
is absent in the leopard. The authors mentioned have given
descriptions of no less than 16 local races of jaguar over the
wide range covered by the species as a whole. The characters
on which these are based are usually not striking and depend
largely on slight peculiarities in the skull. Of the 16 races,
five occur in North America, the others in South America.
Concerning most of these, information is rather scanty and in
general, although jaguars are frequently hunted wherever they
come into proximity of man, there is little to indicate that they
have been effectively depleted in numbers except perhaps in
the regions of thickest settlement. Since the two forms that
occur along the southwestern border of the United States have
probably become much reduced in numbers and in the extent
of the country they occupy, they^may be included together here.

ARIZONA JAGUAR
FELIS ONCA ARIZONENSIS Goldman

Felis onca arizonensis Goldman, Proc. Biol. Soc. Washington, vol. 45, p. 144, Sept. 9,

1932 ("Near Cibecue, Navajo County, Arizona").

NORTHEASTERN JAGUAR
FELIS ONCA VERAECRUCIS Nelson and Goldman

Felis onca veraecrucis Nelson and Goldman, Journ. Mammalogy, vol. 14, p. 236, Aug.,

1933 ("San Andres, Tuxtla, Vera Cruz, Mexico").
FIG.: Nelson, 1916, colored fig. on p. 413.

These two races of the jaguar are characteristic, respectively,
of the subtropical areas of the "mountainous parts of eastern
Arizona north to the Grand Canyon, southern half of western
New Mexico, and northeastern Sonora, " and the "Gulf slope
of eastern and southeastern Mexico from the coast region of
Tabasco north through Vera Cruz and Tamaulipas to central
Texas." Their ranges are therefore separated by the high
tableland of north -central Mexico. The race arizonensis is a

254 EXTINCT AND VANISHING MAMMALS

large northern form distinguished by its flatter, more depressed
nasal bones from other subspecies; it differs additionally from
the race hernandesii to the south by its more massive skull
with broader rostrum, wide anterior nares, and narrower
posterior nares. The race veraecrucis is the largest of the North
American races and has the nasals more arched. A tanned
skin of arizonensis measures in total length, 2,145 mm.; tail,
660; the skull has a condylobasal length of 237 mm. An adult
male of veraecrucis has a total length of 1,993 mm., the skull a
condylobasal length of 247 mm. (Nelson and Goldman, 1933).
The northeastern jaguar just reaches the southern border of
the United States in Texas but apparently has never been
known to be common, and by now it is doubtless nearly extir-
pated in this area. J. A. Allen in 1894 wrote that in Aransas
County, Texas, on the Gulf of Mexico, it was then already gone
from the region. He mentions a skin formerly owned by a
Captain Bailey that was killed in 1858 on Live Oak Peninsula,
but Attwater who contributed the note had heard of none
since in that region. The same collector, quoted by J. A. Allen
(1896), speaks of it as formerly present in Bexar County,
southern Texas, but adds that it was then (1896) "rare east of
the Nueces River, but still taken occasionally in the chaparral
thickets in the counties bordering the Rio Grande. " How far
to the eastward it may once have extended is uncertain. True,
in 1885, in his provisional list of the mammals of North and
Central America and the West India Islands, even stated that
it ranged from Louisiana to Patagonia, but the basis of the
Louisiana report is not given. Seton (1920) has called atten-
tion to an account of what seems to have been the jaguar in
an old book on "Rocky Mountain Life," by Rufus B. Sage,
who while encamped on Soublets Creek, headwaters of the
Platte within 30 or 40 miles of Longs Peak, Colo., mentions a
"strange looking animal" encountered by one of his party.
He believed it to have been "of the Leopard family" and adds
that they (meaning jaguars) "are not infrequently met in some
parts of the Cumanche country, and their skins furnish to the
natives a favorite material for arrow-cases. " If this refers to
the jaguar, as seems likely, it furnishes the most northeasterly
record. Possibly, however, the ocelot was the animal meant.
Bailey (1931) mentions a jaguar killed near Center City, Texas,
as lately as 1903. It was treed by some dogs and after being

NORTH AMERICA AND THE WEST INDIES 255

wounded with a revolver shot was finally driven into some
brush, surrounded, and shot, but not before it had killed a dog
and a horse. Nelson and Goldman mention a specimen ex-
amined from Goldthwaite, Mills County, Texas.

Concerning the Arizona jaguar, the latter authors write that
at the present time its range marks the most northern point
where the species is known. Formerly, according to old rec-
ords, it ranged into extreme southern California, and in recent
years, "while not very abundant, it appears to be a regular
resident of southeastern Arizona. " They examined specimens
from Cibecue, Greaterville, and Nogales in that State. That
jaguars occurred in southern California in the first half of the
last century is attested by Merriam (1919), who quotes several
older accounts: that of Langsdorff, 1814, who mentions it as
among the species of the Monterey region ; Beechey (Narrative,
1831), who in 1826 says it was reported to be found in the
country between San Francisco and Monterey; and Saint-
Amant, who included it in 1854 as a California species. He
further calls attention to a circumstantial account of the finding
of a jaguar family in the Tehachapi Mountains by the famous
James Capen Adams about the> middle of the last century, as
detailed in Hittell's (1860) account of the adventures of this
old hunter. The Indians of a former generation living in
southern California apparently were well acquainted with this
jaguar, and Merriam (1919) was told by an old chief of the
Kammei tribe that in the Cuyamaca Mountain region in San
Diego County it was there known as "big-spotted lion" in their
native tongue. Another early writer, Pattie, mentions one
killed on an island at the mouth of the Colorado River about
1822. Other Indian records, as gathered by William D. Strong
(1926) in this region, indicate that the jaguar formerly occurred
in the mountains bordering the Mohave Desert and that the
"old people made a practice of following jaguar and mountain
lion trails in order to uncover and eat the deer remains the
animals buried. The last jaguar that his informant could re-
member as having been killed in the region was near Palm
Springs, Calif., about 1860. In New Mexico the jaguar is now
rare if not quite extirpated. Bailey (1931) has summarized a
number of later occurrences of jaguars in the southern part of
the State: Grafton, in 1900; Ute Creek, San Miguel County,
1902-3; near Fulton, 1903; Datil Mountains, 1902; Clanton

256 EXTINCT AND VANISHING MAMMALS

Creek ranch, 1903; and Sierra de los Caballos, in about 1904
or 1905. Since these dates, no later instances are mentioned,
and probably the species has been reduced to near the vanishing
point in the United States. It has been found to kill cattle on
the ranches in these regions and hence is regarded as a menace
by the ranchers, to be killed wherever found. Probably it still
occurs in small numbers in northeastern Mexico. According
to Nelson (1916) the jaguar has little of the truculent disposi-
tion of the leopard. In parts of Mexico he made careful in-
quiry without hearing of a single case where one had attacked
human beings, although the natives everywhere fear it on ac-
count of its size and strength. He writes: "Jaguars are very
destructive to the larger game birds and mammals of their
domain and to horses and cattle on ranches. On many large
tropical ranches a 'tigrero,' or tiger hunter, is maintained,
whose duty it is immediately to take up the trail when a 'tigre'
makes its presence known, usually by killing cattle. The
hunter steadily continues the pursuit" in which dogs are ordi-
narily used, until the animal is killed or driven far away. He
mentions that along the Mexican coast in spring, when sea
turtles come ashore to lay eggs, fresh tracks of jaguars may be
found showing where they have traveled along the beach for
miles in search of the eggs. In the province of Guerrero,
Mexico, the hunters have a way of imitating the call of the
jaguar during the mating period and thus enticing it within
shot.

Order ARTIODACTYLA : Even-toed Ungulates

The cloven-hoofed herbivores form a large and complex
group. About a hundred genera are usually recognized from
the Holarctic, Oriental, Ethiopian, and Neotropical Regions.
The artiodactyls do not occur in Australia, but the deer reach
Celebes, and the pig family is found on most of the oceanic
islands, probably carried by man to the islands west of Celebes.
European and American deer have been introduced into New
Zealand and some of the other islands, where they are doing
well.

Since these herbivores are everywhere hunted for food and
sport, many forms have become greatly reduced in numbers.
There are three major groups:

(1) Suina, the hippopotamuses, pigs, and peccaries.

NORTH AMERICA AND THE WEST INDIES 257

(2) Tylopoda, the camels and their South American rela-
tives, the huanaco and vicuna.

(3) Pecora, the ruminants, to which belong the three families
treated herein:

(1) Cervidae, the deer. Six forms of wapiti, 11 forms of the
American deer, and 13 forms of caribou are considered to come
within the scope of this work.

(2) Antilocapridae, the North American pronghorn antelope.
Four races of the single Recent species are here recognized.
The familv is the only endemic group of the Artiodactyla in the
New World.

(3) Bovidae, the cattle and their relatives. All the American
representatives of this group, except the Rocky Mountain
goat (19 forms, 3 genera), are included here as vanishing mam-
mals. J. E. H.

Family CERVIDAE: Deer

THE WAPITI, OR AMERICAN ELK

These large deer, although almost universally called elk in
America, should more properly be termed by the American
Indian name wapiti, since elk in the Old World was originally
applied to the species resembling our moose, found across
northern Europe and Asia. In general appearance this animal
is much like the familiar red deer of Europe but is much larger
and of a darker color, without the whitish eye ring of the latter.
The antlers, too, in addition to their larger size, seldom show
the "cup" of the Old World species, in which the three termi-
nal tines tend to come off together. Originally found over
most of temperate North America, this species has become very
greatly restricted in range within the three centuries since
active settlement by Europeans. At present some half dozen
geographic races are recognized, which may be taken up in
detail, as follows :

EASTERN WAPITI, OR ELK
CERVUS CANADENSIS CANADENSIS Erxleben

Cervus elaphus canadensis Erxleben, Syst. Regni Animal., vol. 1, p. 305, 1777 (Eastern

Canada; as later restricted, "near Montreal," Quebec).
FIGS.: Audubon and Bachman, 1851, vol. 2, pi. 62 (colored figure of animals caught

when young in western Pennsylvania).

258 EXTINCT AND VANISHING MAMMALS

The general appearance of the eastern wapiti is described by
Audubon and Bachman (1851), the latter of whom kept on his
grounds near New York City a pair that were obtained as
young animals in western Pennsylvania and that later served
as the originals from which their plate was drawn by Audubon.
Head dark brown, neck darker, blackish; a white patch on
each side of the under jaw, with a black stripe between, passing
down on to the throat. No pale eye-ring. General color dark
gray all over except for the prominent white rump-patch.
The male in winter develops a heavy fringe of long hair on the
throat and back of the neck. Under surface of body, brown.
The winter pelage is somewhat grayer than that of summer
which is redder, and the young are spotted with white. The
large antlers have a rounded beam and extend back, giving off
a brow tine and two others, the bez and trez, before the longest
is reached, beyond which is another fork, its tines in the same
anteroposterior plane. Although there are a number of antlers
in existence, there is apparently but a single skin of the eastern
wapiti preserved, namely, one in the possession of the Academy
of Natural Sciences of Philadelphia. Bailey (1937), on the
basis of the slender evidence available, believes that the
eastern animal was not so large and heavy as the Rocky Moun-
tain race and was brighter brown and more richly colored than
any of the western forms.

Three hundred years ago the eastern wapiti ranged from
southern Quebec to the edge of the Plains and southward
through extreme western New England and western New
York, perhaps as far south as North Carolina on the Atlantic
seaboard, and through the Allegheny Mountains to northern
Alabama. Such a large animal could not fail to have been
useful to the early settlers on account of its meat and hide;
hence it is not surprising that localities the elk haunted or
where they were killed became commemorated and distin-
guished from other wilderness spots by appropriate place-
names. A fairly accurate map of the former range of this
animal in the eastern United States might be made by plotting
the cities, counties, creeks, and rivers named after it. Thus in
western New York we have Elkcreek (town) and Elk Creek
(stream), Elkdale; in Pennsylvania, Elk City, Elk County,
Elk Creek, Elk Grove, Elk Hill, Elk Horn, Elk Lake, Elkrun;
in Maryland, Elk Neck and Elkton; in Virginia, Elk Creek,

NORTH AMERICA AND THE WEST INDIES 259

Elk Garden, Elk Hill, Elkton, Elkwood; in West Virginia, Elk
Garden, Elk Horn; in North Carolina, Elkville; in Tennessee,
Elk Horn, Elk Mills, Elk River, Elkton; in Kentucky, Elkfork,
Elk Horn, Elkton; in Alabama, Elk and Elk River. Other
place names of a similar sort are on the maps of Ohio, Indiana,
Michigan, and Iowa. But the native elk has long since gone,
although in a few places as in Corbin Park, N. H., western
stock has been imported and maintained under fence.

In the northeastern part of its range the animal was known
to the French explorers of the St. Lawrence River, but Dr.
R. M. Anderson (1939a) writes that although found in the prov-
ince of Quebec south of the St. Lawrence in early days it has
been gone from there for at least a century. He quotes W. P.
Lett that about a hundred years prior to 1884 elk were present
in small numbers in Carleton County, Ontario, and the antlers
are frequently turned up by the plow in the vicinity of Ottawa.
Presumably elk were present at that time on the Quebec side
of the Ottawa River. In New England there is no record of
the wapiti in historic times, yet it must occasionally have
reached western Vermont, for antlers have been found in bogs
in that region. Western New York was a part of the wapiti's
range at one time. DeKay, in 1842, after correcting a few
erroneous conceptions concerning this species, writes that they
were at that time found in "the northwestern counties of
Pennsylvania, and the adjoining counties of New- York. In
1834, I am informed by Mr. Philip Church, a stag was killed
at Bolivar, Allegany county. My informant saw the animal,
and his description corresponds exactly with this species."
He also quotes a certain "Mr. Beach, an intelligent hunter,"
that he had shot at one in 1836 on the north branch of the
Saranac and another hunter, Vaughan, who had actually
killed one at nearly the same place. Merriam (1884), however,
is inclined to doubt these last records on the ground that no old
hunter with whom he had talked in the Adirondack region had
ever seen or heard of one; nevertheless he admits that they
must once have been common in the Adirondacks, since a
number of antlers have been discovered, the best preserved of
which that he had seen was found in a bog on Third Lake of
Fulton Chain, in Herkimer County. The largest of several
discovered at Steels Corners in St. Lawrence County measured
12 1/2 inches in circumference at the burr and 10 inches immedi-

260 EXTINCT AND VANISHING MAMMALS

ately above it. Thus, in spite of Merriam's doubts, it is clear
that the eastern wapiti formerly occurred in the northwestern
part of the State and probably became extinct a century or
more ago.

For Pennsylvania the record is much fuller, and Rhoads
(1903) in particular has gathered what evidence he could find
of their former presence. Early accounts tell of their abundance
in that State, especially in the vicinity of salt licks, to which
they had, with other large mammals, worn broad trails. Peter
Kalm related how in the winter of 1705 great numbers came off
the mountains because of the deep snows and were killed.
Barton in 1806, a century later, implies the reduction of the
herds: "In the memory of many persons now living, the droves
of elks which used to frequent the salines near the river Sus-
quehanna in Pennsylvania were so great that for 5 or 6 miles
leading to the licks the paths of these animals were as large as
many of the great public roads of our country. Eighty elks
have sometimes been seen in one herd upon their march to the
salines." In the vicinity of these congregating places they
were apparently much hunted and killed, which may in part
explain their early extermination. Rhoads says : " In the north-
eastern Alleghenies of Sullivarr, Luzerne and Wyoming counties
they seem to have totally disappeared in the second decade of
the 19th century, although a few remained in a favorite haunt
called 'Elk Forest' in the Pocono range of Wayne Co. until
exterminated between 1830 and '40. In Tioga, Lycoming and
Potter counties they haunted the headwaters of Pine Creek
and the Black Forest until 1862, when the last was killed. In
Somerset and Bedford counties, where the mountain glades
and saline or sulphur springs were sought out by numerous
bands of wapiti and buffalo in early colonial times, their ex-
termination must have been of very early date . . . Even
more obscure is the evidence of their former occurrence in the
southwestern counties of Pennsylvania, and in the parts of
New Jersey pertaining to the valley of the upper Delaware. "

Stragglers, however, driven from the Pennsylvania moun-
tains are believed occasionally to have been taken in northern
New Jersey in former days. A writer in 1835 (R. C. Taylor,
in London's Mag. Nat. Hist., vol. 8, p. 536) says that at that
time elk were almost extinct in Pennsylvania; nevertheless a
few continued for a number of years and were shot by hunters

NORTH AMERICA AND THE WEST INDIES 261

at the salt licks, or were tracked down with the help of dogs
and shot. It was not thought unusual for a hunter to follow a
stag 40 miles before finally bringing it to bay. Shoemaker
(1915) writes of these final captures: "Taken by sections, the
last elk in the Blue Mountains was killed about 1800; in the
Pocono Mountains in 1845; in Lacka wanna County five or ten
years earlier. Caleb Mitchell killed the last elk of the Seven
Mountains at the head of Treaster Valley, Mifflin County, in
1857; James David killed the last elk in Clearfield County in
May, 1865. It was brought to Lock Haven on a raft from the
mouth of Medix Run, where it was killed . . . Jim Jacob -
son, a half-breed Indian, killed an elk in Elk County in 1867,"
which according to Rhoads is the last to have been shot in the
State. Shoemaker (1915), however, says that this same
Jacobson killed others "annually until Nov. 19, 1875, when he
killed his last near Roulette, Potter County," and adds that
the very last Pennsylvania elk was shot on September 1, 1877.
Up to the middle of the last century "there was quite a thriving
business of catching elks alive in northern Pennsylvania. "

In the adjacent State of West Virginia wapiti were common
till about the middle of the eighteenth century and thereafter
dwindled in numbers up to about the time of the Civil War.
Kellogg (1937), in summing up their brief history, notes that
"between 1830 and 1835, elk were killed at a deer lick near
'The Sinks' on Gandy Creek . . . Three elk were killed
on the Black Fork of Cheat River near Davis, Tucker County,
in 1843 . . . During 1845, seven elk were seen near
Durbin, Pocahontas County." McWhorter in a historical
account of the region states that an elk was killed in 1867 at
Elk Lick on Middle River, in the same county, and tracks were
said to have been seen even in 1873. At the present time a few
elk, escaped from an enclosure in Marlinton, are said to be at
large on the ridges of the eastern part of the State, but these
were imported.

Still farther south, C. S. Brimley (1905) states that this
animal doubtless occurred in North Carolina a century and a
half previously, and it seems also to have been present in
Virginia in the years after early settlement, up to about 1847
(Audubon and Bachman, 1846-54). For Tennessee, Kellogg
(1939) has assembled evidence to show that "elk at one
time were plentiful . . . occurring not only in the high

EXTINCT AND VANISHING MAMMALS

passes and narrow valleys of the mountainous sections but
also in association with the buffalo visited the licks of middle
Tennessee, browsed along the rivers and creeks in the southern
counties, and wandered through the canebrakes of the Missis-
sippi bottomlands." Crockett refers to them repeatedly be-
tween the years 1820 and 1830 as then present in the bottom-
lands of Ob ion and Dyer Counties. The last records as given
by Rhoads (1896) seem to be: About 1849 one killed at Reelfoot
Lake by David Merri wether and one reported killed in 1865 in
Obion County. Doubtless a few must have reached the
mountains of northern Alabama as evidenced by the place-
names, previously mentioned, and may have wandered into
the adjacent part of Georgia, but no contemporary record is
known to me.

West of the Alleghenies to the edge of the Plains this race of
elk was formerly common. Historical accounts of Ohio tell of
its presence in the pioneer days of settlement, but it seems to
have been much reduced by the end of the eighteenth century.
Kirtland, however, writing in 1838, says it was "frequently to
be met with in Ashtabula county until within the last six years.
I learn from Col. Harper, of that county, that one was killed
there as recently as October "of the present season." There
were elk in Kentucky in pioneer days, affording the early
settlers a desirable source of meat and leather but they were
early reduced in numbers. Audubon wrote that when he first
settled in that State (in 1808) there were still some to be found
and a few in Illinois across the Ohio River, but by 1847 they
seem to have been pretty well gone. Cory has assembled many
older accounts for Wisconsin and Illinois. In 1818 they were
no longer to be found east of the Illinois River, in Illinois,
though a few seem to have remained in the northern part of
the State. In southern Illinois they are said to have been com-
mon up to about 1820; the last records in Indiana are of one
killed in Knox County in 1829 and another in the following
year. In Wisconsin they apparently held on till much later.
Hoy wrote in 1882 that they were found on Hay River in that
State in 1863, but Bray ton, in 1882, said they were still found
in the vicinity of Green Bay, Wis. They seem to have been
"numerous" in eastern Michigan as late as 1860 in Huron and
Sanilac Counties about the headwaters of Cass River, but so
relentlessly were they being hunted with rifle and trap pens

NORTH AMERICA AND THE WEST INDIES 263

that their early extermination was foreseen by Miles. In 1856
they were not uncommonly seen at Saginaw Bay. Elk were
formerly abundant in Minnesota, and according to Herrick
(1892) they were occasionally killed by the Indians to the north
of Lake Superior as late as 1885, and he was informed of their
presence in that year at Red Lake. Although Cory (1912)
stated that a few individuals were said still to remain in the
extreme northern part of Minnesota, no evidence of this is
given. Cahn (1937) says that in Quetico Park, adjoining the
Superior National Forest in that region, elk have long been
extinct, but that in 1914 or 1915 they were introduced into the
latter reserve, and on at least one occasion "some members of
this herd were seen in the Quetico Park" on the Canadian side
of the common boundary. At the present time, according to
a Department of Agriculture press statement dated January 29,
1939, there are in Minnesota some 45 elk, and in Michigan 5,
but what part of these, if any, represents the native stock is
uncertain. From the same source it appears that estimates of
elk populations show the following figures : In New Hampshire,
250; in New York, 100; in Virginia, 140. Presumably these are
all from stock introduced fronrthe West.

How far to the westward the typical eastern elk ranged may
perhaps never be demonstrated. From analogy with climatic
conditions and from what is known concerning other animals
of wide distribution, it may be presumed that somewhere
along the eastern edge of the Great Plains the characters
showed intergradation with those of the Rocky Mountain elk,
so that available notes from the States of the eastern Plains
may be grouped under that race. In North Dakota, where
this transition may have been found, Bailey (1926) writes of
the former abundance of elk all over the State, where they were
equally at home in the timber and on the open prairie. Ex-
plorers and others tell of the numbers found at the beginning
of the last century, but they rapidly disappeared before on-
coming settlement. A few remained into the early eighties,
one of the last reports being that of six killed in 1883 near
Elkton in Cavalier County. In former times the valleys seemed
to be their wintering grounds, where they found a dense cover
of timber and abundant food. The antlers were shed by the
males mainly during March and April, and quantities of these
might be picked up. "Next to the buffalo," says Bailey, "the

264

EXTINCT AND VANISHING MAMMALS

elk at the height of their abundance were the easiest to hunt
and hence the most rapidly killed of the large game, but when
much hunted they became very wild, and it is probable that
besides the vast numbers killed in the State, many were driven
out of its borders." Running down elk on horseback was a
favorite form of plains sport. The meat was in great demand as
about the best of game animals, while the hides were much
used as leather not only by the Indians for moccasins but by
the whites for various purposes. The upper canine teeth were
a prized decoration of Indian women, and in later times were in
great demand as watch charms among the whites, a demand
that for a time led to much ruthless slaughter.

The possibility of raising elk commercially under fence has
been somewhat developed and is the subject of more recent

Eastern wapiti (Cervus canadensis canadensis}

NORTH AMERICA AND THE WEST INDIES 265

experiment. Bailey (1926) says that "in domestication elk
have proved more hardy and prolific than other stock and al-
most as easily handled under well-fenced range. If in the future
the production of elk meat proves as profitable an industry as
it promises, there will be found ideal conditions for elk pastures
in many parts of western North Dakota, where rough and
steep slopes lie close to brushy bottomlands, and winter browse
and summer grass can be inclosed in single or adjoining areas.
The severe winter weather which means suffering and loss to
domestic stock without shelter is a joy to these native born and
bred deer if a suitable and adequate food supply be available.
Along many of the stream valleys with Badlands borders,
which now lie idle or are of little use for stock, elk would find
an abundance of their favorite food and choice living condi-
tions. The time seems ripe for adding this industry to the
many resources of the State."

ROCKY MOUNTAIN WAPITI; NELSON'S WAPITI
CERVUS CANADENSIS NELSONI Bailey

Cervus canadensis nelsoni Bailey, Proc. Biol.^Soc. Washington, vol. 48, p. 188, Nov. 15,

1935 ("Yellowstone National Park," Montana).
FIGS.: Elliot, 1901, pi. 12 (skull and antlers); Nelson, 1916, p. 454 (col. fig.).

The wapiti of the Rocky Mountain region, from northern
New Mexico northward into Canada, and including formerly
perhaps the Great Plains area, is regarded as a race larger in
size of body and in antlers than that of the eastern woodlands,
and in color it is the palest of the races except the dwarf elk
(C. c. nannodes) . Bailey describes the type as light buffy fawn
in summer, fading to creamy buff; rump-patch creamy buff or
whitish; head, neck, legs, and belly dull rusty brown to dark
umber and blackish; eye-rings buffy. In winter, the body
color is buffy gray over lavender, with dusky tips that wear
off leaving a creamy or soiled-whitish appearance; large rump-
patch, including the tail, whitish or clear white; head and neck
dull rusty brown with dark-brown manes, darkest on lower
throat; ears dull light brown, lined with pale buff. The basal
length of the skull of the type (male) was 430 mm.; length
of antler over the beam, 1,250-1,260 mm.

It may now be impossible to ascertain where this race
intergraded with the eastern form or whether the range of the

266 EXTINCT AND VANISHING MAMMALS

Rocky Mountain wapiti should be regarded as extending east-
ward to the edge of the Great Plains. At all events, the species
formerly was found over much of this region. Thus, in 1804-5,
Lewis and Clark recorded elk along the Missouri River all the
way through North Dakota, and 40 years later Audubon
found them abundant along the Missouri and Little Missouri
Rivers. Bailey (1926) mentions being told of "thousands"
seen along the Lower Yellowstone in 1864 and they remained
common in the Dakotas till about 1880, but "as the country
filled up with settlers they rapidly disappeared," and were
practically gone from the State by the middle eighties. The
late J. A. Allen (1870) wrote of their occurrence in Iowa, that
though formerly numerous, they were by 1879 "extinct in
most of the region." An old resident with whom he talked in
1867, at New Jefferson, Greene County, told him that only
seven years before "the elk were abundant in some parts of
that county. Prior to this date he used to see herds nearly
every day, and sometimes several in a day, some of them of
very large size. During the early settlement of this part of
Iowa they were of great value to the settlers, furnishing them
with an abundance of excellent food when there was a scarcity
of swine and other meat-yielding domestic animals. But, as
has been the case too often in the history of the noblest game
animals of this continent, they were frequently most ruthlessly
and improvidently destroyed. In the severer weather of winter
they were often driven to seek shelter and food in the vicinity
of the settlements. At such times the people, not satisfied with
killing enough for their present need, mercilessly engaged in an
exterminating butchery. Rendered bold by their extremity,
the elk were easily dispatched with such implements as axes
and corn-knives. For years they were so numerous that the
settlers could kill them whenever they desired to, but several
severe winters and indiscriminate slaughter soon greatly re-
duced their numbers, and now only a few linger where formerly
thousands lived, and these are rapidly disappearing. Their
home here being chiefly the open country, they much sooner
fall a prey to the Vest ward march of civilization', through the
most merciless treatment they receive at the hands of the emi-
grant, than does the deer."

On the plains of Kansas the elk was formerly abundant but
has long been extinct. In these open regions elk were con-

NORTH AMERICA AND THE WEST INDIES 267

spicuous because of their large size, while the excellence of their
meat made them an object of pursuit by the pioneers. Add to
these points their gregarious habits, the dangers to which
they were exposed through severe winters, prairie fires, and
diseases, and it becomes obvious that they would be among
the first to go in the face of these perils. Shortly after the
Civil War, with the increase of settlers over the prairie country,
elk became reduced in numbers on the Plains until they were
left only in the more rugged country in the foothills of the
Rocky Mountains.

This race of elk reached its southern limits in the mountains
of extreme northern New Mexico. In the San Juan Mountains
they were reported as abundant by Pike in 1807 and by Cope
as not uncommon in 1874, but in succeeding years these num-
bers seem to have quickly dwindled, although they were still
to be found in 1892. In September, 1909, forest rangers
reported two elk in these mountains. Bailey (1931), who re-
cords these facts, was unable to learn of elk in the Jemez
Mountains when he was there in 1906. Nelson, who in 1883
was staying near the head of Pecos River, had a few reports
of elk in the nearby mountains at that time, but the last
record that Bailey could obtain was of one killed in 1902.
While probably the native stock was by then wiped out,
several attempts at restocking have since taken place, by im-
porting animals from Colorado and Wyoming, as in 1911,
when small numbers were "planted" in Potato Canyon,
Cimarron Canyon, Gallinas Canyon, and later in San Miguel
County. These introductions, in part under fence, were re-
ported as doing well and increasing slowly. With the setting
aside of several game reserves, it seems likely that a stock may
eventually be built up.

In Colorado, elk were early exterminated in the eastern parts
and in 1871 were already rare even in Park County in the
eastern hills. Cary, writing in 1911, says that they were then
exterminated over much of their former range in Colorado,
and "the few bands which remain in the wildest parts of the
western plateaus and mountains are small and widely scat-
tered . . . 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 . . .

268 EXTINCT AND VANISHING MAMMALS

Careful estimates made by the Forest Service officers in the
spring of 1911 show a total of about 2,100." Since Gary's
report in 1911, the herds seem to have been built up very
greatly, for the U. S. Biological Survey in its census published
in 1939 gave an estimated total of 23,000 elk for Colorado.

At the present time, owing to adequate protection in recent
decades, the elk herds of Wyoming, Idaho, and Montana
have increased to such large proportions that they are no
longer in danger. The chief centers of abundance are in such
national parks as the Yellowstone, where formerly hundreds
were slaughtered by hunters; another famous place is the
Jackson Hole region, Wyoming, a regular wintering area for
elk coming down off the surrounding mountains where they
spend the summers. In Glacier Park, Montana, there are
smaller numbers, but these are apparently increasing. One
difficulty in maintaining elk in large preserves is, as Bailey
points out, the problem of providing upland mountainous
areas for their summer range and suitable lowland places
where food is available for wintering them. The 1939 wildlife
census provides the following figures for elk populations in the
Rocky Mountain States, from south to north: New Mexico,
5,000; Colorado, 23,000; Utah, 3,800; Wyoming, 40,700;
Idaho, 24,400; Montana, 26,700; Nevada, about 200.

In addition to these there are a good many elk in other
States that have been introduced from the Rocky Mountain
area. Some of these are living under fence, and others may be
free. Thus the report above quoted gives the following:
Arizona, 4,400; California (in several places in the northern
part); Michigan, 5; Nebraska, 31; New Hampshire, 250; New
York, 100; Oklahoma, 230; South Dakota, 3,400; Texas, 350;
Virginia, 140. The Rocky Mountain elk is obviously in no
present danger. The feasibility of restoring elk to regions
from which they have been extirpated is evident from the
following quotations (gathered by Dr. Francis Harper). Ac-
cording to Barker (1936, p. 177), in New Mexico, "elk, com-
pletely exterminated about 1900, have been restored so that
we now have about 5,000 animals." "Prior to 1912, elk had
disappeared from the ranges of the State of Utah. Since 1912
there have been planted within our State, a few at a time,
193 head in seven different areas. These plantings were made
during the period from 1912 to 1925, and the elk were imported

NORTH AMERICA AND THE WEST INDIES 269

from Jackson Hole and Yellowstone Park . . . We now
have approximately 3,500 head on the ranges" (Cook, 1936,
p. 187). In Wyoming, Montana, and Idaho they have been
completely saved from extinction through adequate protective
measures. In 1932 there were, in national parks of the United
States, an estimated total of 15,420 and in national forests
over 96,800 (Phillips, 1935).

In a recent survey of the elk situation, it is shown by Skinner
(1928) that in the Yellowstone region, for example, any ade-
quate plan for maintaining the elk population should consider
providing a sufficient food supply for winter as well as summer.
This seems to consist under natural conditions of "browse"
as well as grass, so that the hay supplied in recent years for the
Jackson Hole herd in winter is probably not an adequate
diet. He gives a partial list of food plants used from month
to month. Protection against unusually heavy snows and cold
is another problem, for when elk are forced out from Yellow-
stone Park into the surrounding low country in winter, thou-
sands are killed by hunters, without supervision or restraint.
There is the further problem of competition with bison herds
of the region for food, the depredations of coyotes, wolves,
and mountain lions, though these are less important. Finally
diseases, some of which may be introduced through domestic
stock, and parasites are as yet insufficiently known. Murie
(1930) has made a special study of winter losses among the
elk at Jackson Hole, Wyoming, and attributes a large portion
of them to necrobacillosis, the symptoms and general etiology
of which he describes and illustrates. The sharp awns of the
squirreltail grass, which produce lesions in the skin of mouth,
tongue, or throat and so afford entrance for the disease organ-
isms, seem to be a contributory factor in this. The elimination
of this grass, if that is possible, might aid in preventing the
trouble. All these problems are matters of importance in
maintaining the elk herds and increasingly demand careful
consideration. A beginning in this had already been made by
the U. S. Biological Survey, in concert with the Forest Service
and a program of management laid 30 years ago (see Graves
and Nelson, 1919). The present excellent condition of the
herds in Yellowstone Park reflects the wisdom of the policies
pursued.

A decade ago, during prohibition days, a number of elk

270 EXTINCT AND VANISHING MAMMALS

were imported into eastern Massachusetts and ostensibly
maintained for breeding and experimental purposes, until it
developed that the pine grove in their spacious and well-
guarded paddock concealed a large illicit still!

North of the United States- Canadian Boundary this race
of elk extended its range in former times to about latitude 60
nearly to Lake Athabaska but was driven back in the last
century. Nevertheless there are still "some numbers in East
Kootenay district, British Columbia, where they are hunted
to some extent . . . Some small herds of typical Rocky
Mountain elk have within the past few years been demon-
strated to occur on some of the southern tributaries of Nelson
River in northeastern British Columbia" (Anderson, 1939b).
They are, however, apparently now gone from the Peace River
region where they formerly occurred. In a manuscript report
of Harry Snyder to the Canadian Secretary of the Interior,
as communicated by Dr. Francis Harper, the former tells of the
discovery of a herd of 150 to 200 wapiti on the Prairie and
Henry River watersheds of northern British Columbia, an area
comprising 750 to 800 [Psquare] miles, about 500 miles north of
the nearest known area now inhabited by elk.

MANITOBA WAPITI
CERVUS CANADENSIS MANITOBENSIS J. G. Millais

C\ervus\ c\anadensis\ manitobensis J. G. Millais, The Gun at Home and Abroad, vol. 4,
p. 281, 1915 ("Manitoba and eastern Saskatchewan," Canada).

The elk in the most northerly part of its range in North
America are "apparently larger and duller colored than typical
canadensis; not so large or light colored as nelsoni of the Rocky
Mountains. In summer rich chestnut brown, darker on head
and neck and belly; in winter somewhat lighter. The name
manitobensis, applied to this northern animal by Millais in
1915, is regarded as valid by Bailey (1935) for the race found
in Manitoba and Saskatchewan and presumably still farther
to the westward. Its precise area of distribution, however,
is somewhat uncertain. In former times the wapiti, whether of
this race or nelsoni, ranged to the Rocky Mountains on the
west and northwestward, to the plains of Peace River, where
Alexander Mackenzie found it in abundance near the trading
post that he established at the mouth of Smoky River in the

NORTH AMERICA AND THE WEST INDIES 271

autumn of 1792 (see Preble, 1908). In that day it ranged as
far north as Fort Nelson River, according to Richardson, but
the nineteenth century saw its numbers so depleted that in
1894 Caspar Whitney placed its northern outpost as on the
Saskatchewan, between Edmonton and Lac la Biche. Loring,
in 1896, reported that a few were said still to exist near the
head of Pembina River, where, however, they had been nearly
exterminated by Indians through "crusting." Preble (1908),
whose account summarizes these facts, adds that a few were
seen on the south side of the Athabaska River in 1897. The
most recent account of the status of this race is that of Dr.
R. M. Anderson (1939b), who writes that in 1923 there were
estimated to be 400 or 500 elk at Riding Mountain, Manitoba,
and other bands on Duck Mountain Forest Reserve, but none
elsewhere in Manitoba. The Riding Mountain herd in seven
years under careful protection had by 1930 increased to about
8,000 head, but in the severe winter of 1934 the heavy mortality
cut this down to about 6,000, with an estimated 300 in Duck
Mountain Reserve, 250 in Porcupine Reserve, and 600 in the
Inter-Lake district. In 1923 the Provincial Game Guardian of
Saskatchewan estimated that in* the Moose Mountain Reserve
in that province there were about 100 elk, and in the region
north and east of Prince Albert about 1,200 more. In 1927 a
short open season was allowed here in the second half of Novem-
ber. The Prince Albert National Park, established that same
year, "now shelters numbers of elk, but the exact figures are
not known, and there are several bands outside the Park."
It thus appears that in these two provinces, to which this race
may be considered restricted at present, there are something
over 8,000 head, largely in government reserves under careful
supervision, so that in these areas there is as yet no fear for
the elk's extinction.

MERRIAM'S ELK, OR WAPITI

CANADENSIS MERRIAMI Nelson

Cervus merriami Nelson, Bull. Amer. Mus. Nat. Hist., vol. 16, p. 7, Jan. 16, 1902

("Head of Black River, White Mountains, Arizona").
FIGS. : Nelson, 1902, figs. 1, 3, 5 (skull); Bailey, 1931, pi. 3 (antlers).

Equaling in size the Rocky Mountain elk, this race is de-
scribed as paler and more reddish in color, with more massive

272 EXTINCT AND VANISHING MAMMALS

skull and more erect antlers; the nasal bones are remarkably
broad and flattened, with a well-marked constriction in the
middle.

Fifty years ago this elk was common in the Sacramento,
White, and Guadalupe Mountains east of the Rio Grande
River and in the Mogollon group of mountains west of it, in
southern New Mexico, as well as in the White Mountains of
Arizona. There are old records for the Datil and Gallina
Mountains of Socorro County, New Mexico, and a doubtful
record for the Manzano Mountains. At the present time this
race, with a restricted range in the mountains of southern
New Mexico and Arizona, is believed to be extinct (Bailey,
1931).

Bailey has summarized the history of this deer. He surmises
that it was antlers of this race that Montezuma showed to the
followers of Cortez. That it was once common in southern
New Mexico is inferred from the statement published by J. A.
Allen in 1874 that a trapper and guide in the Rocky Mountain
region had met with droves of 2,000 individuals as far south
as the Mexican boundary. In the White Mountains and the
Blue Range of Arizona, "more particularly on the head of
Black River," according to D. B. Rudd, a forest ranger, they
were present in large bands in 1876 when his father moved to
that part of the country. "As late as 1890 elk could be found
but not so plentifully. Since the year 1895 I cannot find that
any have been seen. " Nevertheless, another ranger had seen
"bedding signs" of a small band between Black River and the
higher plateau of the Blue Range, about 1904, which seems to
be the latest report of them for that region. In the Mogollon
Mountains H. W. Henshaw found tracks of these elk in 1873,
and in 1882 Nelson heard reports of their presence near the
extreme headwaters of the Gila River. In 1886 he collected
specimens in the White Mountains of Arizona. Reports of the
forest rangers indicate that these elk became scarce in the
Mogollon Mountains about 1888, although up to that time
they seem not to have shown any alarming decline. Bailey
adduces a few instances of animals shot in the Mogollon Moun-
tains up to 1890, when a cow and a bull were killed. The last
record, however, seems to have been in 1894, when a fine male
was reported seen on Lily Mountain, north of the main peaks
of the Mogollons and tracks of three later in the same year.

NORTH AMERICA AND THE WEST INDIES 273

In the Mimbres Range Bailey was unable to find any tradition
of elk, nor in the Datil Mountains could Hollister in 1905 find
any ranchers who remembered seeing them, although old
antlers were occasionally discovered. A similar story of traces
of their former presence was told in the Bear Spring and Indian
Spring Mountains, which probably once formed the northern
boundary of this elk's range. In the Sacramento Mountains
elk were said to have been killed up to 1898, but Bailey him-
self, in 1902, could get no record of elk killed or even seen
there later than 1893.

Concerning these elk in the White Mountain region of
Arizona, Nelson (1902) wrote that their main range covered an
area about 30 by 50 miles in extent, at an elevation 8,000 to
10,000 feet above sea level. At that time, between 1885 and
1888, "elk were far from numerous" but seemed to be most
often found "about some beautiful damp meadows in the
midst of the dense fir forest on the rolling summit of the
Prieto Plateau, between the Blue and the Black Rivers."
W. W. Price, who collected in this region in 1894, saw several
and shot a male at 9,000 feet elevation. The report of tracks
seen here in 1901 by a local httnter, who followed them un-
successfully for two days, indicates that a few survived until
that time, perhaps till about 1904, as previously noted. The
fact that this survivor was trailed for two days shows that elk
were still hunted in spite of legal prohibition.

In the Chiricahua Mountains of southeastern Arizona
these elk in early times were numerous, but with the extension
of grazing interests in the latter part of the last century there
came "a long slow reduction in numbers due chiefly to the
elks' inability to compete with cattle on the over-grazed range."
Hunting accomplished their final extermination in the region,
for the last small band was killed in the vicinity of Fly and
Chiricahua Peaks about 1906 (Cahalane, 1939a).

DWARF ELK; CALIFORNIA WAPITI; TULE ELK

CERVUS CANADENSIS NANNODES Merriam

Cervus nannodes Merriam, Proc. Biol. Soc. Washington, vol. 18, p. 23, Feb. 2, 1905

(" Buttonwillow, Kern County, California").
FIG.: Merriam, 1905, p. 23, fig. (from photograph of a male).

Although called "dwarf elk," this race of southern Cali-

274 EXTINCT AND VANISHING MAMMALS

fornia is after all only slightly smaller than the form of the
Rocky Mountains. Compared with the latter it is paler, with
a smaller and narrower rump-patch. The legs are shorter,
and the skull is slightly smaller but has longer palatal bones
in proportion. Premolar and molar teeth are of the same length
in both, hence relatively larger in nannodes. Antlers in general
similar to those of the Rocky Mountain form but smaller and
with the posterior terminal prong less developed. "Front of
legs and feet bright golden fulvous; back and flanks varying
from buffy gray, slightly washed with fulvous, to grizzled
buffy whitish"; inner side of the ears "buffy white, the white,
particularly at posterior base, much more extensive than in the
other" races. The type specimen, a two-year old male, had a
total length of 2,030 mm.; tail, 140; hind foot, 620; basilar
length of skull, 358 mm. as against 388 in a male of same age
representing the Rocky Mountain form.

Prior to 1860 this form of elk was common in nearly the
entire San Joaquin and Sacramento Valleys, California,
especially in their lower parts. It was found at least as far
north as Butte Creek, in Butte County, and south to near
Bakersfield, Kern County, and westward through the southern
inner coast ranges as far as the plains of the Cuyama Valley,
San Luis Obispo County, and extreme northern Santa Barbara
County, and to the south end of San Francisco Bay, Santa
Clara County (Grinnell, 1933). "The encroachments of civili-
zation have resulted in the gradual extermination of this elk
over the greater part of its former range, until" by the end of
the last century it had become "restricted to a small area
between Tulare and Buena Vista lakes, where at present
[i. e. in 1905] the survivors are confined almost exclusively to
lands included in an extensive cattle ranch (Buttonwillow
Ranch) owned by Miller and Lux. " About 1903 these owners
presented the herd to the United States Government and a
park for its reception was constructed on Kaweah River in
the Sequoia National Park. In November, 1904, an attempt
was made to round up and corral this herd and move the
animals to the new area, but it was not a success, for "the elk
refused to be driven and escaped to the adjacent foothills of the
Temploa Mountains. " During the process several were roped,
with fatal results, but the skins and skulls were preserved for the
U. S. Biological Survey collection (Merriam, 1905). Since

NORTH AMERICA AND THE WEST INDIES 275

then, apparently, the remnant of this herd has continued, wild
or in part under fence, in "western Kern County between
Tulare and Buena Vista lake basins and at times in adjacent
hills to westward. In 1932, about 170 individuals remained"
(Grinnell, 1933). "Many transplan tings of Dwarf Elk from
the remaining 'herd' in the Button willow district to other parts
of the State have been made, first to Sequoia National Park,
Tulare County, in 1904 and 1905, and more recently to Yosem-
ite Valley, to Monterey County, etc. None of these has re-
sulted in establishment under conditions of really wild freedom.
The Sequoia animals had entirely disappeared by 1926"
(Grinnell, 1933). With the protection now accorded them,
unless local requirements change, it seems likely that this
remnant in the Buttonwillow district may continue for a long
time.

ROOSEVELT'S ELK; OLYMPIC ELK OR WAPITI

CERVUS CANADENSIS ROOSEVELTI Merriam

Cervus roosevelti Merriam, Proc. Biol. Soc. Washington, vol. 11, p. 271, Dec. 17, 1897
("Mt. Elaine (on ridge between heads of Hoh, Elwah, and Soleduc rivers) near
Mt. Olympus, Olympic Mts., State of Washington").

SYNONYM: ? Cervus occidentalis Hamilton Smith, Griffith's Cuvier, Animal Kingdom,
vol. 4, p. 101, 1827. [Not certainly applicable to any American deer (Bailey)].

FIGS.: Bailey, 1936, pi. 20 (photographs of wild individual and a herd).

This elk of the Northwest coast is perhaps the largest of the
races, the males with massive skull and antlers that are often
"cupped," that is, the three terminal tines tend to come off
together. The beam of the antler is relatively short and
straight, with the terminal prong reduced. The body color in
summer is a rich cinnamon-buff; head, neck, and belly dark
brown with much black; in winter the body is dark gray with
a dusky dorsal stripe and with dusky on the face, mane, and
belly. In the skull, the frontals are broad and much flattened,
and the preorbital cavity is small, as compared with the Rocky
Mountain race. Total length of an adult male, 2,490 mm.;
tail, 80; spread of antlers, 990 (3 feet 3 inches); length of left
beam (in type), 1,050 (41.25 inches). The length of antler
is about 500 mm. (20 inches) shorter than in a Rocky Mountain
elk of comparable development.

The range of this race is confined to the humid forest belt
along the Pacific coast from northern Vancouver Island south
through the coast ranges of Washington, Oregon, and Cali-

276 EXTINCT AND VANISHING MAMMALS

fornia to (formerly) about San Francisco Bay. Eastward in
California the range included Mount Shasta. Altitudinally
the area inhabited extended in California to about 7,000 feet
(Grinnell). On Vancouver Island, its northward limit, it was,
according to Swarth (1912), formerly abundant everywhere
in the forests, decreasing in numbers southward to within 30
miles of Victoria. "They keep well back from the settled
districts and are quite scarce near the east coast and the ad-
jacent woods. They are most abundant in the north end of the
island, particularly in the northwestern section and the vicinity
of Kuyuquot Sound . . . The wapiti are seldom hunted
by the Indians, who prefer the more easy task of killing deer on
the beaches." In the interior of Vancouver Island the wapiti are
still fairly safe, since the forests are relatively unexplored and
the animals are little disturbed (Sheldon, 1912). The Govern-
ment took active steps to stop the killing of these wapiti for
their canine teeth when, about 30 years ago, they were in
active demand. Dr. I. M. Cowan, in 1937, told Dr. Francis
Harper that there were supposed to be only about 700 elk left
on Vancouver, this decimation being due in part to the ravages
of wolves and cougars, but possibly also to disease that may
have come among them, as in the Rocky Mountains.

This elk is not found on the mainland of British Columbia
opposite Vancouver but appears again across the Strait of
San Juan de Fuca in the mountains of the Olympic Peninsula
and in scattered bands in the coast ranges of Washington and
southward. Not many years ago its numbers seemed to have
reached a low level in this State. Taylor and Shaw, in 1929,
reported that in addition to the small groups found south to
Pacific County, a small herd was said to be living on the head-
waters of the Cispus River, south of Mount Rainier. It was
at one time feared that this remnant might be even further
reduced, but recent reports indicate that it is at present
thriving under protection and is in no present danger. It
should be borne in mind, however, that according to these
authors Rocky Mountain elk have been introduced into several
counties of Washington: Stevens, Garfield, and Walla Walla
in the eastern part and Skagit, Snohomish, King, and Yakima
in the central part. It is believed that in former times the
ranges of these two forms did not overlap in the Northwest.

In Oregon the status of this deer has been summarized by

NORTH AMERICA AND THE WEST INDIES 277

Vernon Bailey (1936), who reports that there are still con-
siderable numbers of them in the coast ranges to which they
are confined, from the Columbia River to the California border.
Originally they inhabited the western slope of the Cascade
Mountains and thence to the coast over all western Oregon.
In 1841 Wilkes found them plentiful in the Willamette Valley,
and Peale reported them from the mountains south of the
Columbia River. Suckley and Gibbs about 1854 met with them
abundantly in the mountains west of Astoria. "Later as the
country was settled, elk were reported from almost every
valley and mountain range of western Oregon, including the
west slope of the Cascade Mountains, but there seems to be
no record from the east [drier] slope of the range. In recent
years, under rigid protection, apparently they have been
holding their own, while in some localities actually increasing.
The continual spread of settlement, though, is restricting their
range, and the greater number of hunters each year makes it
more difficult to prevent poaching in out-of-the-way places"
(Bailey, 1936). In 1926 there was an estimated total popu-
lation of 436 elk on the national forests of western Oregon, a
figure that, says Bailey, covers the greater part of the Roose-
velt's elk in the State. This is a slight increase (of about 40)
over the estimate for 1933, but less than those for 1930 and
1931. It may be pointed out that the 17,000 elk estimated in
1939 as inhabiting the State of Oregon (Federal census), are
doubtless mostly the Rocky Mountain elk, native and im-
ported, in eastern parts of the State.

Still farther south this elk formerly occurred in abundance
all the way to San Francisco Bay in the humid coast belt of
northwestern California and eastward to the nearer inner
coast ranges and to Mount Shasta. Of recent years, however,
its numbers have dwindled, until by 1933, according to Grinnell
(1933), it was to be found only in Del Norte and northern
Humboldt Counties, while the total population was believed
not to exceed 400 head. Its vertical range extended originally
from sea level up to at least 7,000 feet, as formerly in the Scott
Mountains, Trinity County. Its usual habitat was in and
about openings in forest.

While the population of this elk is small, with a few hundred
in Vancouver Island, and in each of these three northwestern
States, nevertheless with the protection now given it there

278 EXTINCT AND VANISHING MAMMALS

seems no reason to be apprehensive for its survival for a long
time to come. Nevertheless, in order to preserve the race in
its purity, care should be exercised that introductions of the
Rocky Mountain elk are avoided within the range of this
coastal form.

EASTERN MULE DEER; PLAINS MULE DEER
ODOCOILEUS HEMIONUS HEMIONUS (Rafinesque)

Cervus hemionus Rafinesque, Amer. Monthly Mag., vol. 1, p. 436, Oct., 1817 (Sioux

River, South Dakota).
SYNONYM: Cariacus virgultus Hallock, Forest and Stream, vol. 52, p. 404, May 27,

1899 (Near Hallock, Kittson County, Minnesota).
FIG.: Nelson, 1916, col. fig. on p. 455.

Unlike the white-tailed deer, this is more a species of open
or rough brushy country, so that where its range meets that
of the former it is found in a different type of habitat. The
antlers, while often having a basal prong, are usually charac-
terized by the dichotomous forking of the main beam. The
ears are longer than in the white-tailed group; the tail is short
and either black-tipped or largely black above (black-tailed
deer of the Pacific coast) . The metatarsal gland is much longer
and higher up on the foot. Both these types of deer are charac-
terized by the extension of the vomer backward to form a
complete longitudinal division of the posterior nasal passage;
both also belong to the group Telemetacarpalia, in which the
lateral digits of the forefeet are represented by the outer in-
stead of the inner ends of the metacarpals.

As a species the mule deer and its west coast forms, the black-
tailed deer, are or originally were found from the eastern edge
of the Plains to the Pacific Ocean. While most of the western
races are still abundant locally, the eastern form is largely
extinct and may therefore be mentioned here. Baird describes
its color as "ashy brown, pointed or varied with gray, and
without any rusty tinge whatever. There is a distinctly marked
stripe from the crown of the head to the root of the tail, the
hairs in which are much darker throughout their extent than
elsewhere. The under parts generally appear to be ashy
brown, like the back, but without annulation of the hairs.
The only whitish portion of the inferior surface is seen beneath
the head, around the axillae, and in the groin. On each side
of the tail, on the end of the rump, is a dull white patch,

NORTH AMERICA AND THE WEST INDIES 279

crossing above the tail and involving its entire basal half or
two-thirds. The tail itself is quite slender and cylindrical on
the basal two-thirds; the hair there being compact and close
and then expanding into a dense tuft, which is entirely black"
(Baird, 1857). Total length from tip of nose to tip of tail
vertebrae, about 6 feet 6 inches.

Formerly the eastern mule deer was found in the Plains
country from northern Oklahoma, Kansas, Nebraska, the
Dakotas, and extreme northwestern Minnesota northward
into southern Manitoba and westward to the foothills of the
Rocky Mountains, where it is believed to have intergraded
with the Rocky Mountain mule deer, 0. h. macrotis, which is
still common in the region from northern New Mexico north-
ward into southern British Columbia. In the Dakotas this
was a common species in the early days, and Bailey (1926)
has brought together many notes of their presence from the
accounts of early travelers in the region. In the seventies of
the last century they were more abundant than the white-
tailed deer on the upper Missouri, but in the succeeding decade
the increasing influx of settlers and hunters had greatly reduced
their numbers. J. S. Weiser, emoted by Bailey, reported them
so common in the vicinity of Valley City in 1878 "that one
could not travel 5 miles without seeing them," and JohnHailand
stated that though common at that time the last one was
shot in 1885 or 1886. The latter adds: "There was so much
venison in camp during the first years that visitors' ponies
were usually loaded down with it before they returned. There
was no sale for venison nor for skins, they were so plentiful.
Skins were used for mattresses; they would get damp and
deteriorate during summer and a new supply was provided
each fall for the winter's sleeping." In 1896, Seton (1909) in
a 15-mile ride across the Badlands of the Little Missouri,
counted only three where ten years before he and his companion
had counted 160 over the same ground. The numbers seem to
have declined rapidly in the early years of the present century.
Bailey states further that in 1912 they were reported rare in
the Turtle Mountains of North Dakota, but just across the
line in Manitoba they were more common and a number were
killed each year. In 1913 Jewett found them "still fairly
common" in the Badlands along the Little Missouri, but in
the Killdeer Mountains he found that all had been killed off

280 EXTINCT AND VANISHING MAMMALS

near settlements. He was told that at that time they were
found only in the rougher parts of the Badlands, where a few
may still persist, "but, if not already extinct, this finest of all
native species of the smaller deer will soon have vanished from
the State" (Bailey, 1926). They are apparently rare also in
South Dakota.

There were mule deer in Nebraska in the seventies, but their
numbers have been much reduced, and from Kansas they are
quite gone. At the end of the last century a few existed in the
open northwest corner of Minnesota, from Red Lake to Lake
of the Woods, but they were exterminated soon after. In 1911
Gary wrote that "apparently none remain on the plains east
of the mountains" in Colorado, "where they were common in
early times." The mule deer of the mountainous areas here
are perhaps of the race macrotis, although, as Bailey says, so
few specimens of the typical plains form are in existence that
we may never know its precise characters.

In the 1939 estimate of game animals by the U. S. Biological
Survey, the following figures are given for the mule deer in the
States east of the Rockies where it formerly was found:
Oklahoma, none; Kansas, none; [eastern Colorado, gone;]
Nebraska, 350; South Dakota, 4,900; North Dakota, 100;
Minnesota, none. Although the basis of the figures for South
Dakota is not given, it is likely that their center is in the
Black Hills region, now a national park. Here Dr. Walter
Granger reported them "numerous" in 1895 although "about
extinct in the Bad Lands" (see J. A. Allen, 1895). With
strict protection it seems likely that they might be permanently
preserved in such areas as parts of the Badlands afford, with
due consideration of the fact that they are wont to move down
from the higher hills in autumn to winter in the more sheltered
places of the lowlands.

WHITE-TAILED DEER; VIRGINIA DEER

ODOCOILEUS VIRGINIANUS VIRGINIANUS (Zimmermann)

Dama virginiana Zimmermann, Specimen Zoologiae Geogr., p. 351, 1777 ("America,"

assumed to be Virginia).
FIGS.: Elliot, 1901, pi. 15 (skull), as 0. americanus; Nelson, 1916, p. 458, upper fig.

(col.).

The white-tailed deer is found in woodlands of the temperate

NORTH AMERICA AND THE WEST INDIES 281

zone quite across the United States, barely entering southern
Canada; to the southward it extends to Central America and
northern South America. Over this wide range it develops
several local races, some of which are in danger of being un-
duly reduced. A brief survey of these is here given, since this
animal is one of our most important game species.

The total length from nose to tail is given by Baird as 59
inches for a doe from Virginia; tail about 9 inches. Males are
slightly larger and heavier, weighing up to 200 pounds. In
winter coat, "pale grayish chestnut, faintly annulated"; in
summer "bright uniform rufous" the "blue" and the "red"
coats of hunters. Under parts of body and tail, inside of the
limbs, the area between the jaws, and lining of ears, white;
a pale eye-ring. Tips of ears and a spot at the angle of the
mouth black. Antlers with a basal prong, the beam curving
up, out, and then slightly inward and forward, usually with
three erect prongs, but often with additional small points.

The typical race of Virginia deer is believed to range from
central or southern Pennsylvania south to the region of Palm
Beach in eastern Florida; westward it extends at least to the
Mississippi basin in Ohio and Kentucky, and formerly to the
edge of the Plains. Over most of this range it still is found in
woodland regions and survives in spite of persistent hunting,
even in fairly well-inhabited areas.

In New Jersey and southern Pennsylvania the white-tailed
deer formerly abounded, but Rhoads (1903) in the early years
of this century writes of it as confined to limited areas in
southern New Jersey, while from the lowlands of the Susque-
hanna, Allegheny, Monongahela, and Delaware River Valleys
it was exterminated. In New Jersey it was found then in
Burlington, Atlantic, Cape May, and Ocean Counties in small
numbers. At the same time in West Virginia deer were "still
rather plentiful" though much less abundant than in earlier
times (F. E. Brooks, 1911), and this is generally true for other
nearby States. Thus C. S. Brimley (1905) reports it rare or
absent in the more thickly settled parts of North Carolina
but "not uncommon" in the eastern section of the State and
in the mountains. Dr. Francis Harper (1937) notes that in
the Okefinokee Swamp region of southern Georgia it was
"abundant in former times" but "had become greatly reduced
in numbers through excessive hunting in late years, even before

282 EXTINCT AND VANISHING MAMMALS

the cutting of the timber and the great increase of the human
population" of the area. On account of its ready adaptability
it can maintain itself in varied cover and persist even in prox-
imity to settled districts if given a reasonable chance. With
two young at a birth its numbers readily come up with pro-
tection, so that even where much hunted it will continue in
the face of considerable persecution. In the Great Smoky
Mountains National Park it is said that there were believed
to be "not more than a score of deer" remaining in this region
when it was established in 1931, but already they show some
evidence of coming back (Campbell, 1939). Kellogg (1939)
summarizes records showing the "incredible number" of deer
formerly found in parts of Tennessee and the importance of
their meat and hides to the early settlers. By the latter part
of the last century they had been so reduced in many areas of
the State that the "Cumberland Mountain range has been
almost entirely depleted of its stock of deer. " Action of the
legislature in 1895 prohibiting their killing for five years in
five of the counties of Tennessee and the later prohibition of
their hunting in the Great Smoky park will doubtless save the
remnant. In 1882, according to Brayton, the deer was then
"rarely met with in Ohio," and the same story holds elsewhere
of its great depletion in the eastern States. Even in Alabama,
Ho well (1921) tells that they "once ranged in large numbers
over all" parts of the State but are "now exterminated in all
but the wilder and more inaccessible" districts.

As a good example of what may be done, however, is the
result reported by F. B. Chapman (mimeographed report of
the Ohio Wildlife Research Station, release 105, 1939). Deer,
almost if not quite extirpated from Ohio by 1904, were re-
introduced in 1922-30, and under protection they have so
increased that by 1939 they numbered more than 2,000, dis-
tributed in 31 counties, in spite of a certain amount of poaching
and occasional accidents and disease. The sex ratio is found
to be about four or five does to one buck. As might be ex-
pected, the abundance of deer is roughly proportionate to the
amount of forest cover.

Estimates of the deer population of the States in which this
race occurs were given by the U. S. Biological Survey in Janu-
ary, 1939, as follows: New Jersey, 8,000; Indiana, 400; Ohio,
2,000; Iowa, 450; Virginia, 15,000; Kentucky, about 700;

NORTH AMERICA AND THE WEST INDIES 283

Missouri, 700; West Virginia, 17,500; Tennessee, 1,700;
North Carolina, 51,000; South Carolina, 23,000; Georgia,
16,400; Alabama, 20,200; Arkansas, 8,400; and Mississippi,
4,600. A comparison of these figures with the statements
given above indicates that under recent legal protection the
numbers have probably increased very greatly over what
they were a few decades ago, so that a certain amount of hunt-
ing will become warranted.

At the present time the white-tailed deer is the largest and
most valuable of our game animals in the eastern United States.
As an object of interest to the increasing numbers of persons
who delight in seeing wildlife, as an important item of food for
those living in rural or wilderness areas, and as a chief object
of pursuit by the many who enjoy hunting in the invigorating
autumnal season, this deer may be reckoned as one of our
most important species in the regions where it is found. The
readiness with which it adapts itself to the proximity of man
and the rapidity with which its ranks may build up if given
proper protection should make its continuance in reasonable
abundance a matter easily managed by any efficient game
department.

NORTHERN VIRGINIA DEER

ODOCOILETJS VIRGINIANUS BOREALIS Miller

Odocoileus americanus borealis Miller, Bull. New York State Mus., vol. 8, p. 83, Nov.

21, 1900 ("Bucksport, Hancock County, Maine").
FIGS.: Stone and Cram, 1902, pi. opposite p. 42; Barbour and Allen, 1922, pi. 4, fig. 2

(antlers).

The northern Virginia deer is a slightly larger animal than
the more southern race, weighing when full grown up to 250
pounds or even more. In winter coat the pelage is slightly
longer and thicker, and the antlers of the males tend to be
wider, more spreading, and less bent in at the tips.

The northern Virginia deer probably did not range farther
north in early times than the coastal regions of Maine and
adjacent parts of New Brunswick and Nova Scotia, avoiding
the central parts of Maine, northern New Hampshire, and
northern Vermont and southern Canada. The northern and
central parts of New England and New Brunswick are regions
of heavy and lasting snowfall, in which the movements of deer
become much restricted in winter, while in times of crust forma-

284 EXTINCT AND VANISHING MAMMALS

tion, wolves would probably get most of the deer in such
regions. A hundred years ago, says Dr. R. M. Anderson
(1939a), "deer were said to be very seldom seen north of the
Ottawa River." Within the past 50 or 75 years, however,
considerable changes have taken place. Wolves are gone
south of the St. Lawrence River, and even to the northward
are much reduced in Quebec, so that they hardly form a
deterrent to the spread of deer. Furthermore, with the in-
tensive lumbering industry in New Brunswick, northern
Maine, Michigan, and Wisconsin, areas of primeval spruce
and white-pine forest have been cleared, with a consequent
springing up of a great sprout growth of deciduous trees, which
afforded cover and abundant food. These two factors have
probably contributed in great part to the spread northward of
this deer over Nova Scotia, New Brunswick, northern Maine,
and southern Quebec, until "in 1930 a deer was taken in
Abitibi River district within one hundred miles of James Bay"
(R. M. Anderson, 1939a). Deer have been introduced and are
thriving on Anticosti Island. They are common in northern
Michigan, northern Wisconsin, northern Minnesota and the
adjacent parts of Canada as in Quetico Park (Cahn, 1937),
where, though some are killed by wolves during winter, "such
depredations are not at all serious."

In his map of the distribution of eastern races of the Vir-
ginia deer, Cory (1912, p. 66) shows as blank areas from which
the deer is "now practically extinct," western New York,
northwestern Pennsylvania, practically all of Ohio, Indiana,
and Illinois, the southern third of Michigan and Wisconsin,
all of Iowa, the southern third of Minnesota, and a part of
northern Missouri. Over much of this area it has in recent
years come back. According to Dr. M. W. Lyon, Jr. (1936),
deer at the present time are extinct in Indiana, where in pioneer
days they were abundant. "The last stand of the deer in the
state was in the northwest in the Kankakee region and in
Knox County," where the last wild deer were seen near Red
Cloud in 1893. Probably the estimate of 400 deer in the
State given by the Government census in 1939 requires con-
firmation, but indicates a very recent return. Hollister, writing
in 1908, says that the last wild deer were exterminated in
southeastern Wisconsin nearly 60 years before. In Walworth
County, where they were abundant a century ago, they rapidly

NORTH AMERICA AND THE WEST INDIES 285

decreased after 1842, and the last one was seen near Delavan
in 1852. "Probably the most southern limit of their range in
Wisconsin at the present time is Sauk County" (Cory, 1912).
Although formerly abundant in Illinois, deer seem to have
been greatly reduced by the last decade or so of the last century,
and by 1900 were regarded as practically exterminated within
the State, although even as late as 1910 Cory had reliable
reports of a very few in Alexander County in the southern
portion of the State. Possibly this remnant has increased in
the succeeding quarter century enough to justify the figure
given in the 1939 Federal estimate of 250 animals.

The way in which this deer will "come back" if depleted
numbers are allowed to breed up for a period of years is well
illustrated by the experience of the three southern States of
New England. Before the settlement of the country deer were
a staple source of food for the Indians. From deer skins they
prepared clothing, and from the straight metapodial bones
they skilfully cut out portions for use as implements. The
first settlers also depended in part on these animals for sus-
tenance and buckskin, and there was a large trade in the hides.
As early as 1646 Rhode Island had a closed season on deer
from May 1 till November 1. In 1698 Connecticut ordered
that deer should not be killed between January 15 and July 15,
and this period was soon after extended to include the re-
mainder of these two months. Massachusetts soon followed
suit, and deer reeves were appointed to enforce the law. For
nearly a century the numbers continued without notable de-
pletion, but by the close of the eighteenth century deer had
become scarce or nearly exterminated over much of southern
New England, until in 1842 Linsley, in his account of the mam-
mals of Connecticut, makes mention only of one killed the
previous year in Waterbury. In 1869 J. A. Allen wrote that
in Massachusetts they were gone except for a few in Plymouth,
Barnstable, and Berkshire Counties, where they were "strin-
gently protected by law. " Much of southern New Hampshire
and parts of Massachusetts were in the middle of the nineteenth
century cleared and used as grazing country, supplying the
nearby markets until the opening of the more extensive areas
of the Middle West, after which these regions gradually re-
verted to woodlands. Toward the end of the century the
remnant of deer that had hung on in the Berkshires of the

286 EXTINCT AND VANISHING MAMMALS

western part of Massachusetts and in the Plymouth region of
the eastern section gradually built up a small population,
which overflowed into the surrounding regions and, helped by
a few introductions, began at length to attract notice. Under
protective laws the numbers have increased so greatly that by
1910 Massachusetts declared an open season of six days late
in November. Since then there has been an annual open sea-
son with few exceptions, and of about the same length. The
number killed each year has varied from as low as 832 to over
2,000 (in 1924). There seems no reason to believe that with
proper regulation the deer may not continue to be a game
animal not only in Massachusetts but also in the other New
England States where a somewhat similar increase also took
place, in Connecticut, Rhode Island, and Vermont (in the
last-named as a result in part of introduction). A similar wise
administration has brought about a like result in Pennsylvania,
where in early days it abounded, but by the end of the last
century, Rhoads (1903) said, it had become "sparsely scattered
or locally exterminated" though most often found in the
Pocono and South Mountain regions. At the present time,
the estimate published (in 1939) by the U. S. Biological Survey,
places the white-tailed deei population of Pennsylvania at
793,000 ! The success of these eastern States in the replenish-
ment of their forest covers with deer, and the asset which this
game may be to the commonwealths, should prove an example
to other States, in which the deer have been greatly depleted,
to encourage an attempt at restocking.

FLORIDA DEER
ODOCOILEUS VIRGINIANUS OSCEOLA (Bangs)

Cariacus osceola Bangs, Proc. Biol. Soc. Washington, vol. 10, p. 26, Feb. 25, 1896

(" Citronelle, Citrus County, Florida").
SYNONYM: Odocoileus virginianus mcilhennyi F. W. Miller, Journ. Mamm., vol. 9, p.

57, Feb. 9, 1928 ("Near Avery Island, Louisiana").
FIGS.: Barbour and Allen, 1922, pi. 5, figs. 5-7 (antlered skulls).

In the southern part of the Florida Peninsula, and extending
up on the western side and along the Gulf coast to Louisiana,
the representative of the Virginia deer is a small animal, with
notably smaller antlers than the typical race. Although the
type locality is in the intergrading area of the two races, the
name may for the present be retained. Of these small Florida

NORTH AMERICA AND THE WEST INDIES 287

deer, Cory (1896) states that he has killed full-grown bucks
that did not weigh over 110 pounds, although they average
larger. Apparently the recently described small deer from the
coast marshes of Louisiana may be considered the same and
is here included as a synonym.

The exact status of the small Florida deer is not well ascer-
tained. It is apparently common in southern Florida and was
considered to be "of very general distribution" over much of
the peninsula in 1898 by Bangs, although "in the more thickly
settled and accessible parts of the State it has been much re-
duced in numbers. " This reduction has of late been acceler-
ated, because of the killing of deer on account of their being
hosts to ticks that spread disease. On the Gulf coast they are
said to be no longer found in western Florida, but in extreme
southwestern Alabama what may possibly be this same race
still exists. In this State, according to A. H. Howell (1921),
deer are now practically exterminated except in a few areas in
the northern part (presumably typical mrginianus) and in
"the big wooded swamps of the lower Tensaw and Mobile
Rivers " (presumably osceola) . Here a number are killed during
the open season every fall. Ttey are hunted with dogs that
drive the deer past the waiting hunter at some favorable
"stand." "The deer take readily to the water and swim
easily from one island to another in this great swamp; in this
way they are able to keep ahead of the dogs, but are often
shot while swimming a creek or river or when crossing an
opening in the timber. Deer are still found in moderate num-
bers in the sandhills and swamps of southern Baldwin County
. . . Twenty years ago or more they were common in the
sandhills and small swamps of Mobile County, but now ap-
parently all have been exterminated from that region."
Both these counties border the Gulf of Mexico in extreme
southwestern Alabama. No data are at hand concerning the
deer in coastal Mississippi, but it seems probable that the
small deer with large toothrows of south-central Louisiana
described as a race, mcilhennyi, are the same and are of local
distribution in the coastal region.

Probably some restriction on the killing of deer in these
areas is desirable.

288 EXTINCT AND VANISHING MAMMALS

KEY DEER; "SPANISH DEER"

ODOCOILEUS VIRGINIANUS CLAVIUM Barbour and Allen

Odocoileus virginianus clavium Barbour and Allen, Journ. Mammalogy, vol. 3, p. 73,

May, 1922 ("Big Pine Key, Florida").
FIG.: Barbour and Allen, 1922, pi. 5, fig. 4 (an tiered skull).

This is the smallest of the eastern races of this deer, charac-
terized by its pale colors, very small antlers, and small teeth
(upper cheek teeth 67 mm. in combined length). It is short-
coated at all seasons and somewhat buffier in color than the
mainland form.

This small deer is confined to a few of the outer keys off the
southeast coast of the Florida Peninsula, at the present time
being found only on the chain of islands from Big Pine Key
southwestward to Boca Chica, a small island 7 or 8 miles from
Key West; on the latter there were deer in the early nineties.
"Deer were killed on Stock Island, a small key adjoining Key
West," about 1910, but none has since been known there.
There were deer on Boca Chica until about the same time,
when they disappeared and were not again noticed there until
late in 1920, when two were seen. "Deer have always been
found from time to time on Ramrod Key, all three of the Torch
Keys, and probably Newfound Harbor. They swim readily
from key to key and if hunted on the smaller islands they leave
and go back to Big Pine Key," the largest of the "lower
keys. " On this island and some of the others the dense thorny
thickets of pricklypear and other growth offer them a safe
refuge, and in 1922 they were said to be more plentiful than
they had been a few years previously.

On account of its restricted range and the assiduity with
which it is hunted, even by parties coming over from Cuba,
this small deer may be considered to maintain a rather pre-
carious existence, but unless its stronghold of thorny growth
is burned or destroyed in some other way it is likely to hold
out in small numbers for a long time.

TEXAS WHITE-TAILED DEER

ODOCOILEUS VIRGINIANUS TEXANUS (Mearns)

Dorcelaphus texanus Mearns, Proc. Biol. Soc. Washington, vol. 12, p. 23, Jan. 27, 1898
(Fort Clark, Kinney County, Texas).

NORTH AMERICA AND THE WEST INDIES 289

This is the race of the semiarid part of southern and middle
Texas, but the exact limits of its range remain to be defined.
It is said to be of small size, with relatively short legs, small
ears, and with small and strongly incurved antlers. The molar
and premolar teeth are relatively very large, the colors pale.
Total length, 1,585 mm. Weight of bucks up to 100 pounds^
females 75 pounds (Mearns, 1907).

The range of this race extends into Mexico from southern
Texas at least as far as San Luis Potosi. Mearns (1907)
wrote of it that in the early years of this century it was abun-
dant in the bottomlands and low mountain ranges of southern
Texas, notwithstanding the fact that the Seminoles kill num-
bers of them every year. He had seen them in great bands in
Devils River Valley. Bailey (1905) says that "on many of the
large ranches between Corpus Christi and Brownsville, where
the oak and mesquite thickets are interspersed with prairie and
grassy openings, deer find ideal conditions, with abundant food
and cover. The nature of the ground is such as to protect them
from wolves and other natural enemies; but it is well suited to
either hunting on horseback or still hunting, which, if freely
allowed, would soon exterminate them. With protection,
however, they increase rapidly, and in many places are abun-
dant . . . On certain large ranches they are still numerous,
while on others they have become extremely scarce and would
be entirely exterminated but for the recruits from surrounding
and better protected ranches ... In spite of the pro-
tection of State laws and ranch owners there are still remote
sections of rough, uncontrolled range where every year hunters
kill wagonloads of deer for the market, or worse, kill the deer
for the hides only, leaving the carcasses to rot. I was told
that in the winter of 1901-2 hundreds of deer skins were
brought out of the country west of Kerrville." Bailey at-
tributes much of the present abundance of this deer to the en-
lightened attitude of the ranch owners who give it reasonable
protection.

LOUISIANA DEER

ODOCOILEUS VIRGINIANUS LOUISIANAE G. M. Allen

Odocoelus [sic] virginianus louisianae G. M. Allen, Amer. Nat., vol. 35, p. 449, June,

1901 (Mer Rouge, Morehouse Parish, Louisiana).
FIG.: Allen, 1901, fig. on p. 453 (antlered skull).

290 EXTINCT AND VANISHING MAMMALS

The status of this form is imperfectly known. It is a large,
rather pale animal, not very different apparently from mr-
ginianus. Its range is believed to be the northeastern part of
Louisiana, northeastward into the adjacent region. No
statistics of its numbers or notes on its status are at hand.
Probably, however, the 2,000 deer credited to Louisiana by
the 1939 Federal census of big game are largely of this race.

PLAINS WHITE-TAILED DEER

ODOCOILEUS VIRGINIANUS MACROURUS (Rafinesque)

Corvus (i. e. Cervus) macrourus Rafinesque, Amer. Monthly Mag., vol. 1, p. 436, 1817

(Plains of Kansas River, eastern Kansas) .
FIG.: Bailey, 1931, pi. 2, fig. D (antlered skull).

This race of the mountains and plains of the western United
States is described by Bailey as large for a white-tailed deer,
color in winter pale gray, yellowish red in summer; tail long
and bushy, ears small. He describes a doe from Texas as
lacking the black tips on the ears, and in general brighter
colored than the race texanus. It is larger and paler than
typical virginianus.

Originally well distributed in wooded and brush-grown
areas in valley bottoms and along streams from the Dakotas
west to the Rocky Mountains and south to eastern New
Mexico, it is naturally restricted in these regions to the local
habitats it favors, avoiding open country, but seeking the pro-
tection of thickets along the streams. In North Dakota,
Bailey (1926) writes that though much reduced in numbers
locally, this deer, because of its secretive habits has hung on
and is even in places common in spite of settlement. It may
even be farmed with profit, like domestic stock, so readily
does it respond to careful handling. Elsewhere its numbers
vary. Thus in Kansas, where it was originally described, it is
now extinct (Hibbard, 1933). In early days, wrote Gary
(1911), it was found generally over the Plains region of Colo-
rado, but it is at present uncommon in the State and largely
restricted to the foothills and eastern slopes of the front
ranges, where it occurs sparingly across the entire width of the
State. In recent years deer are reported as locally common in
the mountains of eastern New Mexico, although much hunted.
They especially frequent the "willow thickets along stream bot-

NORTH AMERICA AND THE WEST INDIES 291

toms. Bailey (1931) writes that "this white-tailed deer will
soon become exterminated from the open country in New
Mexico, and the only possible hope of keeping it from entirely
disappearing from the State will be to give it a permanently
protected breeding ground where conditions are suitable for
food and shelter. Within its present range these conditions
could be obtained in great perfection on the eastern slope of
the Sacramento Mountains and along the southern and eastern
slopes of the Sangre de Cristo Range. "

In recent years, according to Seton, this deer has extended
its range into parts of Utah.

NORTHWESTERN WHITE-TAILED DEER; YELLOW-TAILED
DEER; "PEND OREILLE DEER"

ODOCOILEUS VIRGINIANUS OCHROURUS Bailey

Odocoileus virginianus ochrourus Bailey, Proc. Biol. Soc. Washington, vol. 45, p. 43,

Apr. 2, 1932 ("Coolin, south end of Priest Lake, Idaho").
FIGS.: Bailey, V., 1918, pi. 7, fig. 2 (photograph of doe), fig. 3 (head).

This race is found from Idaho and Montana to the eastern
slopes of the Cascade Mountains, Oregon, and a short distance
into northern Nevada and the extreme northeastern corner of
California, and extends northward into the southeastern corner
of British Columbia (Cowan).

Equaling the Plains white-tailed deer in size, it differs in
being darker with less black and more ochraceous on the upper
side of the tail. In winter coat it is dark buffy gray above,
becoming bright ochraceous on top of tail, on the legs, and on
the edges of the belly. Forehead and top of head dark brown,
brisket dusky ; sides of nose and eye-ring light gray. In summer
the upper parts are bright tawny or light bay, legs but little
paler, not yellowish as in the race macrourus. Antlers large
and heavy when well developed, and frequently having the
brow tine and the proximal long tine slightly forked.

Concerning this race Bailey (1936) writes: "There is very
little on record of the habits of this deer in Oregon, except
that they are found mainly in thickets and willow bottoms
along the streams and valleys. In recent years they are some-
times found back in the hills, crowded back probably by
settlements. Like most of their group they are secretive and
would rather hide than run. In the more extensive thickets

EXTINCT AND VANISHING MAMMALS

and forests of Idaho and Montana they are still abundant in
favorable locations but over much of their ranges they are
doomed to be crowded out by settlements, or killed by preda-
tory animals as they gather on winter ranges and are easily
pulled down by coyotes and dogs in the deep snows." Bailey
(1918) has given a further account of this deer in Glacier
National Park (under 0. v. macrourus) where it is abundant
and protected. In extreme northeastern California, it was
evidently at one time common in Modoc County but apparently
now is no longer so. Grinnell (1933) gives as "a late definite
record" one taken in extreme eastern Lassen County in
January, 1922. On the eastern slopes of the Cascade Moun-
tains in Oregon and Washington this race seems still to be
common.

Estimates of white-tailed deer populations for the States in
which this race occurs are given by the U. S. Biological Survey
for 1939 as follows: Idaho, 10,000; Montana, 24,000; Oregon,
not given; Washington, 7,500.

By Executive Order in 1939 a Federal refuge for this deer
and other forms of wildlife was established in Stevens and
Pend Oreille Counties, Washington. It will contain eventu-
ally about 65,000 acres and is well adapted to the management
of these deer, since they spend the spring and summer in the
mountains and the autumn and winter in the lowlands of the
area. The need for such a refuge has for some time been urged
because the deer have become locally much limited through
the operations of farming and logging.

PACIFIC WHITE-TAILED DEER; COASTAL WHITE-TAILED DEER;
"TIDELAND DEER"; "COTTONTAIL DEER"

ODOCOILEUS VIRGINIANUS LEUCURUS (Douglas)

Cervus leucurus Douglas, Zool. Journ., vol. 4, p. 330, 1829 (Lower Columbia River,
Oregon; description based on examples from Falls of the Willamette and mouth
of the Columbia).

FIGS.: Bailey, V., 1936, pi. 23, fig. A (antlered skull); Scheffer, 1940, p. 275 (photo-
graphs of a doe).

This is the only race of white-tailed deer to reach the Pacific
coast of the United States.

Small in size, standing about 3 feet 5 inches at the shoulder,
and characterized by its small delicate antlers, this is likewise
a darker race than ochrourus, its nearest neighbor, an appear-

NORTH AMERICA AND THE WEST INDIES 293

ance due to the narrowness of the pale subterminal rings of
the hairs. There is a poorly defined median dark dorsal line
on neck and shoulders ; upper side of tail light ochraceous-buff
at base, more dusky toward tip; brow patch, with prominent
suffusion of chestnut; white of under parts not reaching the
axillae. Summer coat redder.

Formerly this deer was found "at least in the Cowlitz and
lower Columbia River Valleys of western Washington, possibly
also eastward for some distance along the Columbia, there
being one report from Pasco" but at present is much restricted
and has even been supposed to be extinct in that State (Taylor
and Shaw, 1929). South of the Columbia River it formerly
ranged along the coast west of the Cascades nearly to the Cali-
fornia line. It was typically an animal of the river bottoms.
In 1826 Douglas reported this as the commonest deer in the
districts adjoining the Columbia River and in the fertile
prairies of the Cowlitz and Willamette Rivers. He found it
also near the head of tidewater on the Umpqua River. There
are no positive records of its occurrence farther south in the
Coast Range valleys but northward these deer extend to the
Puget Sound country of Washington. In 1872-75 they were
still of regular occurrence in the Willamette Valley, though
less common than 20 years before, but after that time they
rapidly disappeared. According to the report of a reliable
hunter in that valley, the last whitetail he knew of was killed
about 1898 near Sweet Home, Linn County (Bailey, 1936).
These deer were said to be common in the foothills about
Beaverton, Washington County, Oreg., from 1860 to 1875.

Until very recently the only remaining deer of this race were
supposedly to be found in Douglas County, Oregon, where,
according to Vernon Bailey (1936), Stanley G. Jewett in 1915
reported a few still inhabiting the Long Tom Swamp west of
Eugene and a few on the oak-covered hills along the Umpqua
River, some miles northeast of Roseburg. In this latter region,
according to Scheffer (1940), the "herd" is now believed to
number 200 to 300 animals, upon which the State Game
Commission enforces a strict closed season. Curiously, how-
ever, Scheffer reports that this deer is still to be found in some
numbers on both sides of the mouth of the Columbia River,
a fact that had previously quite escaped notice. According to
his detailed account (Scheffer, 1940), the present range of these

294 EXTINCT AND VANISHING MAMMALS

deer includes chiefly low brushy islands and tidelands, below
the 100-foot contour; above this the hills rise sharply from the
river valley and the whitetails give place to the black-tailed
deer. In the State of Washington the area inhabited by the
former includes chiefly Skamokawa, Elokomin, and Cathlamet
Islands, Puget Island, Tenasillahe Island, and Price Island, a
total area of over 12,000 acres, with an estimated whitetail-
deer population of between 400 and 500. On the Oregon side
of the river's mouth the area inhabited is slightly greater
(nearly 14,800 acres), but the deer population is less than half
that in Washington, since the bulk of it is confined to the less
cleared Webb and Westland districts. "As far as the main-
land is concerned, the four corners of the range are fixed by
steep bluffs that rise from the water's edge on both the Oregon
and Washington shores." With practically no competition
from blacktail deer, and with nearly complete freedom from
large predatory mammals, the danger to this group seems small.
Nevertheless in some of the islands as on Puget Island, the
deer "have increased to the point where their depredations
were distinctly annoying to the farmers." A few are annually
killed in spite of legal protection, but with the gradual clearing
and settlement of their restricted range, this remnant, Scheffer
believes, will be in danger of extermination. "The obvious
solution," he writes, "is to move one or more breeding stocks
of the deer to nonagricultural areas," if suitable ones can be
found, or to make a sanctuary of some tideland area already
occupied by them and as yet only slightly devoted to farming,
as on Skamokawa Island. This deer is well known to the
farmers and fishermen of this region as "tideland deer" or
"cottontail deer."

SONORA DEER; FANTAIL
ODOCOILEUS COTJESI (Coues and Yarrow)

Cariacus virginianus var. couesi Coues and Yarrow, Rept. Geogr. and Geol. Explor.

and Surv. west of 100th Meridian (Wheeler), vol. 5, p. 72, 1875 (Camp Crittenden,

Pima County, Arizona).
FIG.: Nelson, 1916, p. 458, lower fig. (colored).

Though possibly to be considered a race of the Virginia deer,
Bailey (1931) believes the Sonora deer are best regarded as a
distinct though closely allied species. About half the size
and weight of a Virginia deer, they are extremely pale in color,

NORTH AMERICA AND THE WEST INDIES 295

a light gray, somewhat more rusty in summer, but in pattern
much the same as their larger relative. Old bucks reach a
maximum weight of about 100 pounds, does 75 pounds. Total
length of a large buck, 1,530 mm.; tail, 270; ear, 203.

This small deer is still common on many of the wooded
mountains "of middle and southern Arizona, southern New
Mexico, western Texas, and in the Sierra Madre of Chihuahua
and Sonora, Mexico," which constitute its range, but "with
the growing occupation of their territory by cattle and sheep
and the increase in the number of hunters, these once abundant
deer are rapidly diminishing" (Nelson, 1916). In winter, if
snows became deep on the upper levels of their mountain
haunts, these little deer would move down into the valleys,
and it was not uncommon to see bands of 20 to 100 in the
White Mountains of Arizona in the nineties of the last century
(Nelson). Concerning their status in New Mexico, Bailey
(1931) writes that they occupy the mountain ranges "west of
the Rio Grande as far north as the Datils and possibly to the
Zuni Mountains ... In 1908 Goldman reported them as
formerly abundant in the Burro Mountains, but at that time
apparently all gone. In 1909 he 5 was told by residents that there
were a very few of them in the Zuni Mountains, but in previous
years" others had failed to find them. In the Mimbres Range,
Bailey reports that Goldman in 1909 found them in limited
numbers and that same year the Forest Service reported them
in the Datil, Gallinas, Magdalena, and San Mateo Mountains.
In the Animas Mountains they were found common by Bailey
in 1906, but he believes that their numbers are a result of the
protective natural features of their haunts, and that "a few
persistent hunters could easily exterminate them." This deer
probably reaches its eastern limit in the Chisos Mountains of
western Texas, where according to Bailey (1905) it ranges from
5,000 to 9,000 feet in the oak, juniper, and nut-pine cover.
Where undisturbed they may often be seen feeding by day.

As to the future outlook for this deer, Bailey writes of it in
New Mexico: "In September, 1915, J. S. Ligon reported Sonora
deer still common in parts of the Mogollon Mountains, but
much less so than formerly, owing to unrestricted hunting in
season and out. On December 31, he wrote: 'The number of
deer killed in New Mexico during the season just closed far
outnumbers the increase for the year, I am quite sure.' These

296 EXTINCT AND VANISHING MAMMALS

mountains now form the principal range of the species in New
Mexico, but as the country fills up with settlers the deer will
entirely disappear unless given better protection than they
have received in the past . . . The many game refuges
in these mountains will now doubtless insure the perpetuity of
this interesting little deer." In the Chiricahua Mountains of
southeastern Arizona, Cahalane (1939a) has lately reported
them as numerous "from the tops of the highest peaks to the
lower limit of the upper Sonoran Zone." Owing to their
aversion to leaving the shelter of forest cover, they are prac-
tically "marooned" on these isolated ranges and in 1933 were
found to be increasing so rapidly that already they "were too
abundant for the carrying capacity of the range . . . The
depletion of food was becoming a serious problem at this
time" and reduction of their numbers by hunting was insuffi-
cient on account of the rugged nature of the country they
inhabit. "A check on this undesirable increase would be
effected by dropping control of mountain lions," the reduction
of which during the previous decade by the "predatory -animal
control" "has been followed by a too rapid growth of the deer
population. "

THE CARIBOU

Two types of caribou are found in the northern parts of the
New World, the barren-ground and the woodland. Both are
divided into several geographic races, not very sharply defined.
The barren-ground caribou, as its name implies, inhabits the
Arctic tundra. It is shorter of limb but often longer in its
antlers than the woodland caribou, which is found slightly to
the southward of the barrens, in regions where open bogs
alternate with evergreen forest. The term "caribou," almost
universally applied to these animals in America, is from a
Micmac word meaning "a shoveler," in allusion to the way in
which these animals paw away the snow in winter to get at the
reindeer moss on which they feed. In contrast to other mem-
bers of the deer family, the females as well as the males may
have antlers, although less often in the woodland than in the
barren-ground species ; when present, however, they are smaller
in the females and in the woodland species are generally spike-
like. The hoofs are broad and spreading as an adaptation to
living on boggy or snow-covered ground, and the body is stocky

NORTH AMERICA AND THE WEST INDIES 297

for a deer. Because of their gregarious habits, their manner of
living in more or less open country, and a sort of stupid curi-
osity they may often be closely approached or attracted and
so in former days were easily brought within gunshot and
several killed before the herd could escape. In many parts of
their range they have become greatly reduced in numbers and
merit protection wherever found. As a source of food and
clothing they are a mainstay of Eskimo and Indians in the
parts of their range where these primitive people live.

BARREN-GROUND CARIBOU
RANGIFER ARCTICUS ARCTICUS (Richardson)

Cervus tarandus var. arctica Richardson, Fauna Boreali-Americana, vol. 1, p. 241, 1829

(Fort Enterprise, Mackenzie, Canada).
FIGS.: Grant, 1902, pis. [18, 19] (antlers and skull).

As Murie (1935) points out, the characters distinguishing
the barren-ground type from the woodland type of caribou
are sometimes "confusing." In general, however, the former
is shorter of limb and lighter in color, the antlers tend to have
the main beam more nearly cylindrical rather than flattened,
and the antlers themselves are less compact, with long, backward
then forward curved beam, branching dichotomously at the
summit. The two brow tines tend to show a wide triangular
expansion or palm. So great is the variation in these structures,
however, that scarcely any two are alike, and the opposite
antlers of the same individual may be considerably different.
Hollister (1912a) states that in the barren-ground type the
lower incisors (which in both are small and weak) decrease in
size from the middle to the outer pair by conspicuous steps,
with the outer pairs very small, whereas in the woodland type
the gradation from the middle to the outer pair is more uniform.

General color in summer "clove-brown, mingled with reddish
and yellowish brown, under-parts white; in winter entire coat
dirty white " (Lydekker, 1915) . In size this is one of the smaller
American races, standing about 46 inches in height at the
shoulder. The basal length of skull averages about 365 mm.,
the maxillary tooth row about 89, but there is a slight over-
lapping with the larger Alaskan race (Murie, 1935). The
antlers of the male are "very long, slender, and rounded, with
few points on the expanded portion of the beam, which is

298 EXTINCT AND VANISHING MAMMALS

separated by a long interval from the third tine, . . . back
tine usually, if not always, wanting; female antlers much
smaller, simpler, and scarcely curved at all" (Lydekker). In
the few specimens available the antlers seem to be less spreading
than in other races, but, as Murie has shown, there is great
individual variation in their form.

The range of this subspecies is believed to be continental
North America from the west side of Hudson Bay to the Mac-
kenzie River and northeastward to include Boothia, South-
ampton Island, and Baffin Island. The migratory habit so
characteristic of caribou leads to more or less irregular wander-
ing, apparently in search of better feeding areas, and the ani-
mals often congregate and pass through certain sections in
great numbers, while in other seasons they may hardly be
found at all. This uncertainty leads at times to much hardship
on the part of Indians and Eskimos who depend upon them as
a source of food. Preble (1902) states that the southern limit
of the barren-ground caribou on the west coast of Hudson Bay
is Churchill River. "Even in former years these caribou were
seldom known to cross that river, and they are still killed within
a few miles of Fort Churchill. Farther inland they reach the
south end of Reindeer Lake." In his later paper on the
Athabaska-Mackenzie region, Preble (1908) has given a de-
tailed account of the presence of this animal as recorded by
various travelers along the western part of the barren grounds,
the shores of the Arctic Ocean, and the outlying large islands
as Banks Land and Victoria Land. "As winter approaches,
the caribou which have summered on the Barren Grounds
move southward in herds, many of which enter the wooded
country. Their movements are more or less irregular. " Few
persons traverse the inland barrens so that this race probably
remains, as Dr. R. M. Anderson says (1939b), "the most
numerous of all the caribou, although now very rare or missing
from much of its fairly recent former range along the Arctic
coast." Nevertheless, he adds, it "has perhaps not been
greatly reduced in total numbers, as where it has disappeared
many of the natives have not followed the caribou. The
caribou have also retreated farther from the shores of Hudson
Bay, and there have been many reports of large concentrations
of caribou in the interior east of Great Slave Lake and south of
Bathurst Inlet. In some cases shifting of range is due to human

NORTH AMERICA AND THE WEST INDIES 299

crowding and in others the destruction of winter forage of
lichens by fire may have been the prime cause of extended
movement . . . Within the past few years more caribou
than usual have come into parts of northern Manitoba and
Saskatchewan, in winter, and quite recently barren ground
caribou have crossed the Slave River into the Wood Buffalo
Park and still farther south of the Park in northeastern Alberta
where the species was never known before. This southward
trend of the barren ground caribou of course brings them out
of the comparative protection of the large native hunting
reserves and restrictive regulation of white trappers in the
Northwest Territories into the northern parts of the Prairie
Provinces where white trappers abound. Reduced killing of
caribou in parts of the north has perhaps been counterbalanced
by increased slaughter in parts of the winter range. Recent
expert investigators inform the writer that there has not been
much new evidence to change his estimate [of 1930] of about
three million caribou in Canada."

On Southampton Island in northern Hudson Bay there is a
local herd of which Sutton and Hamilton, in 1932, saw a few
animals. They write that this caribou was "once a very abun-
dant animal all over Southampton Island. At the present time
it appears to have disappeared almost altogether from the
southern part and to be restricted principally to the region of
the high country between East Bay and Duke of York Bay,
and to the more or less unknown country inland from the
coast and north of Cape Kendall." The Eskimos evidently
hunt them only occasionally, and there is no evidence that they
are greatly depleted. In Baffin Land both Hantzsch and Soper
report the barren-ground caribou as local in occurrence and
irregular in its appearance about the misson settlements of
Cumberland Sound. Almost nothing is known of their num-
bers or their migrations, but it seems likely that they keep
mainly to their archipelago, which they must have originally
reached by way of Boothia.

300 EXTINCT AND VANISHING MAMMALS

LABRADOR BARREN-GROUND CARIBOU
RANGIFER ARCTICUS CABOTI G. M. Allen

Rangifer arcticus caboti G. M. Allen, Proc. New England Zool. Club, vol. 4, p. 104,
Mar. 24, 1914 (30 miles north of Nachvak, eastern Labrador, Canada).

SYNONYM: Tarandus rangifer labradorensis Millais, The Gun at Home and Abroad,
vol. 4, p. 259, 1915 ("Nain, Davis Inlet, and Fort Chimo").

FIGS.: Grant, 1902, pis. [8, 9] (skull and antlers).

This is the race of the Labrador Peninsula north of its
forested base. It is believed to be characterized by the wide
antlers, with very sweeping backward and forward curve, and
the great palmation of the brow and bez tines. If other dif-
ferences occur, as in color, these have not been made out.

Dr. R. M. Anderson (1934b) writes that this caribou "is
found in more or less scattered bands over the treeless Arctic
Zone area and through parts of the scantily timbered Hud-
sonian Zone of the peninsula. In most of the districts near the
coast they have been greatly reduced in numbers, but con-
siderable numbers are still found in the more isolated inland
districts." These are regions seldom penetrated by white
hunters, and accurate information as to the status of this
animal is difficult to obtain. Low, at the beginning of this
century, was told by the Nascopie Indians that there were
three main herds or stocks: One near the coast on the high-
lands between Nachvak and Nain; one in the hinterland of
Ungava Bay that crosses the lower Koksoak and passes the
summer on the tundra along Ungava Bay and Hudson Strait;
and one on the east coast of Hudson Bay, south to Clearwater
Lake. Before 1886 the last herd was said to have fallen off
greatly. Of the second, W. B. Cabot (1912) and others have
witnessed something of the migrations and mention the hunting
of the migrating animals by the Indians who rely partly on the
caribou for sustenance. Prichard (1910) states that these
animals often work out to the coast by November, and at
that time a good many may be taken, but their appearance is
not to be counted on, and they may come out at different
points in different seasons. There is some evidence that ex-
tensive fires have destroyed the available food in parts of their
range, thus excluding them from return to such districts.

There is every reason to believe that this caribou never
crosses the rough water and ice of Hudson Strait to Baffin
Land but is isolated on the Labrador Peninsula. To the south-

NORTH AMERICA AND THE WEST INDIES 301

ward its range is limited by the forest country of the base of
the peninsula, where the woodland caribou is found.

STONE'S CARIBOU
RANGIFER ARCTICUS STONEI J. A. Allen

Rangifer stonei J. A. Allen, Bull. Amer. Mus. Nat. Hist., vol. 14, p. 143, May 28, 1901
("Kenai Peninsula, Alaska").

SYNONYMS: Rangifer excelsifrons Hollister, Smithsonian Misc. Coll., vol. 56, no. 35, p.
5, Feb. 7, 1912 ("Meade River, near Point Barrow, Alaska"); Tarandus rangifer
ogilvyensis Millais, The Gun at Home and Abroad, vol. 4, p. 263, 1915 (Ogilvie
Mountains, north of Dawson, Yukon, Canada); Rangifer mcguirei Figgins,
Proc. Colorado Mus. Nat. Hist., vol. 3, no. 1, p. 1, Dec. 28, 1919 (Kletson Creek,
a tributary of White River, Yukon, Canada).

FIGS.: Allen, J. A. 1901, figs. 1-4 (head, skull, antlers).

In his recent review of the Alaska- Yukon barren-ground
caribou, Murie (1935) has shown that only one race is recog-
nizable over most of central and northern Alaska (exclusive
of the Alaska Peninsula and Unimak Island), ranging to the
western part of Yukon, Canada. This is recognizable by its
large size, with an upper maxillary tooth row averaging 94 mm.
The large cheek teeth, large size, maximum for the arcticus
type, dark color, with well-developed white fringe on the
throat, and large, heavy, and "rangy" antlers are given as
diagnostic characters.

Murie states that this caribou has now disappeared from the
Kenai Peninsula, whence it was first described, but the form
of the interior of Alaska is the same, and it is probable that the
Kenai animals simply represented those remaining from a
periodic overflow of the herds of the interior.

Stone's caribou, named in honor of Andrew J. Stone, who
collected the type, is at present represented by four herds,
which Murie (1935) briefly characterizes as follows:

(1) The Bering seacoast group, consisting of scattered bands
distributed over a wide area of lowland bordering Bering Sea
from Bristol Bay to Bering Strait, but which even by the
early eighties had begun to decline. Residents agree that at
the beginning of this century caribou were much commoner
in the country north of the lower Yukon and inland from
Norton Sound and may still be found in the hills of that region
particularly at the head of Unalakleet, but at the present time
are "entirely absent or occur only as stragglers on the Bering

302 EXTINCT AND VANISHING MAMMALS

Sea coast, in most of the Kuskokwim region, along the lower
Yukon, and in the vicinity of Norton Sound, and much of this
area is now occupied by domestic reindeer" (Murie, 1935).

(2) The Alaska Range herds, consisting of scattered bands
along the entire length of the Alaska Range. Murie believes
that an approximate estimate of this group would be between
25,000 and 30,000 animals, of which the principal aggregation
is included in Mount McKinley National Park.

(3) Northern herds, including the scattered remnants on
the Arctic slope, comprise the animals "ranging along the
Endicott Mountains, through the upper Koyukuk and Chanda-
lar watershed, across the Porcupine into Yukon Territory."
An estimate of 60,000 for the animals of this area is "probably
conservative." In early times caribou were plentiful along
the Arctic coast, but "in later years they largely disappeared
in this area, owing no doubt partly to the activities of whalers,
fur traders, and natives."

(4) The Yukon-Tanana herds, which comprise the largest
numbers of all, occupy the country west of the Mackenzie
River, on the uplands between the Yukon and Tanana Rivers.
Migrations of these herds, often numbering many thousands
of animals, are described by Murie from accounts gathered.
Except locally, their numbers show no depletion of importance.

GRANT'S CARIBOU
RANGIFER ARCTICUS GRANTI J. A. Allen

Rangifer granti J. A. Allen, Bull. Amer. Mus. Nat. Hist., vol. 16, p. 122, Mar. 31,

1902 ("Western end of Alaska peninsula, opposite Popoff Island").
FIGS.: Allen, J. A., 1902a, figs. 1-6 (skulls and antlers).

The caribou of the Alaska Peninsula and Unimak Island is
distinguished by "averaging somewhat smaller and paler than
R. a. stonei, many antlers diverging widely and with sharply
recurving beams, although these antler characters may not
hold uniformly in a large series." The average length of the
maxillary tooth row is 94.5 mm., or about as in stonei.

"In their present depleted numbers" the caribou of the
peninsula are "islolated from the interior herds" but appear
to have been "sufficiently isolated to form a local race"
(Murie, 1935). The antlers show a "tendency toward wide
divergence of the beams," which "average lighter in weight

NORTH AMERICA AND THE WEST INDIES 303

than those of stonei. " At one time these caribou were abun-
dant throughout the length of the Alaska Peninsula from a
point nearly opposite Kodiak Island west to the island of
Unimak and thence all around the Alaskan shore of Bering
Sea (Nelson, 1887), as well as on Nunivak Island. At present,
writes Murie (1935), "the numbers on the peninsula have
dwindled greatly, but a few herds are still holding their own.
One herd, numbering possibly 2,500, ranges from Morzhovoi
Bay to Herendeen Bay, with the center of abundance in the
vicinity of Pavlof Valley. The other herd, apparently a little
larger, ranges from Moller Bay to the vicinity of Black Lake.
The latter herd appears to be less molested by hunters than the
others, but as a whole the caribou of the peninsula have been
and are rapidly dwindling . . . Caribou are scarce to the
eastward of Port Heiden, but a few have been recorded as far
east as Becharof Lake in comparatively recent years. Several
islands adjacent to the peninsula were inhabited at one time.
Caribou occurred in considerable numbers on Unga Island,
according to A. J. Stone, and Deer Island is also said to have
had caribou, but today none are found on either. In July 1925
a caribou skeleton was found on Amak Island, 12 or 14 miles
off the Bering Sea shore . . . This island cannot be much
over 2 miles square. Unimak Island, roughly 30 by 75 miles
in extent, represents the westernmost point of distribution for
caribou and harbors a group that has been estimated at from
7,000 to 10,000 animals; a visit to the island in 1925 inclines
the writer to favor the lower figure. " According to reports of
Donald H. Stevenson, quoted by Murie, the caribou here were
decreasing rapidly in the eighties and early nineties, when it
was thought many caribou were killed for food by men hunting
sea otter. With the cessation of these activities the caribou
commenced to increase until "by 1905 the island held all that
the range could carry. About 1908 they again began to decline
in numbers, but soon were able to hold their own and later
increased once more. " At intervals these island animals have
been known to cross to the mainland of the peninsula, and
there is some evidence of migratory movements lengthwise up
and down the country.

From the evidence brought together by Murie it seems that
in spite of a depletion that has restricted these caribou to
parts of the peninsula and islands, and cut them off from the

304 EXTINCT AND VANISHING MAMMALS

herds of interior Alaska, they are at present in no danger of
extinction and with proper regulation of hunting should hold
their numbers for a long period.

OSBORN'S CARIBOU
RANGIFER ARCTICTJS OSBORNI J. A. Allen

Rangifer osbarni J. A. Allen, Bull. Amer. Mus. Nat. Hist., vol. 16, p. 149, Apr. 16,
1902 ("Cassiar Mountains (60 miles southeast of Dease Lake), British
Columbia").

FIGS.: Grant, 1902, pis. 18-21 (unnumbered) (head with antlers).

This race of the barren-ground caribou is darker than R. a.
stonei with the dusky tipping of the white throat fringe more
pronounced and the maxillary tooth row longer, averaging
about 100.5 mm., according to Murie. The differences are,
however, actually slight, and it is hard to find any very tangible
distinctions between the two, beyond these average ones of
size and color. The antlers sometimes show the flattened form
of beam characteristic of the woodland caribou.

The geographic range of R. a. osborni includes southern
Yukon and northern British Columbia, where according to
late information it is still -"reasonably common" (R. M.
Anderson, 1939b). This race is believed to be much less mi-
gratory than the others.

PEARY'S CARIBOU; ELLESMERE LAND CARIBOU
RANGIFER ARCTICUS PEARYI J. A. Allen

Rangifer pearyi J. A. Allen, Bull. Amer. Mus. Nat. Hist., vol. 16, p. 409, Oct. 31, 1902

("Ellesmere Land, N. Lat. 79").
FIGS.: Allen, J. A., 1908, figs. 1-6 (antlered skulls); Nelson, 1916, p. 421 (col. fig.).

The caribou of Ellesmere Land, Grinnell Land, and Grant
Land is smaller than the Greenland caribou or the typical
R. a. arcticus, its nearest neighbors, and is nearly pure white
in color, with, however, a slight admixture of gray to gray-
brown in the middle area of the back. The maxillary tooth
row ranges from 83 to 95 mm. The antlers "have a much
greater upward curvature than in arcticus, in proportion to
their length," and the main beam has an average length of
about 1,019 mm.

From Peary's notes on these caribou, which range the Arctic
islands mentioned, it appears that they live in small scattered

NORTH AMERICA AND THE WEST INDIES 305

groups of up to a dozen individuals of both sexes, and penetrate
as far as latitude 83 N. on the northern coasts of these areas.
They are not known to show any distinct mass migrations but
remain the year round thriving on the lichen, moss, and other
dwarf and scanty Arctic vegetation. Their chief enemies are
the Arctic wolf and, of course, man, particularly the native
Eskimo, who hunts them for food and clothing. "In these
northern wilds, amid the most intense cold, the caribou passes
from three to five months of continuous night, its wanderings
lighted only by the moon, stars, and the marvelous displays of
waving northern lights" (Nelson, 1916). According to Ander-
son these caribou are not very numerous, although Sverdrup
found them in abundance on the west coast of Ellesmere
Land. Since there are rather few human inhabitants it is
likely that the caribou will persist for a long time to come.

GREENLAND CARIBOU
RANGIFER ARCTICUS GROENLANDICUS (Gmelin)

[Cervus tarandus] groenlandicus Gmelin, Linnaeus's Systema Naturae, ed. 13, vol. 1,

pt. 1, p. 177, 1788 (Greenland).
FIGS.: Grant, 1902, two unnumbered plates (4, 5) (skull with antlers).

Lydekker (1915) describes the Greenland caribou as "closely
allied to R. t. arcticus, with a broad sharply defined white ring
round each eye, and distinct broad white bands above the
hoofs; skull with an elevated frontal region." Whether or
not these color characters are really distinctive, J. A. Allen
(1908) states that it is darker in color than typical arcticus
and much darker than pearyi, resembling greatly the coloration
of the Woodland caribou in its dark-brown body, with neck and
ventral area much lighter. Three male skulls average in
condylobasal length, 368 mm., which is greater than in arcticus
and considerably more than in the little pearyi. The upper
maxillary tooth row is also greater, 81-99 mm. Antlers of
adult males are large and spreading; the brow tines are not
greatly palmate, the main beam is bent sharply forward at
near a right angle at the most posterior point, at which there
is a small but well-developed backward tine. Although the
Greenland caribou has by some naturalists been associated with
the Old World reindeer as a subspecies, and by others is re-
garded as a derivative of the New World barren-ground cari-

306 EXTINCT AND VANISHING MAMMALS

bou, it seems quite as likely that Jacobi (1931) is correct in
calling the forms of both merely races of a single circumboreal
species, Rangifer tarandus.

The Greenland caribou, or reindeer, is believed to be derived
from North America by way of the islands lying to the west-
ward, and in that way to have reached the Greenland coast.
Since, however, Peary's caribou now occupies the northern
Arctic Archipelago, one must assume either that both have
since differentiated from a common stock or that the Greenland
animal arrived in some other way.

Formerly the Greenland caribou was found over the whole
outer land fringing the ice cap of this country, except, ap-
parently, the northern part of the coast. During the last
century and a half constant hunting on the part of both
Eskimo and white population has greatly reduced their num-
bers, so that at the present time they are found only locally in
the areas most suitable to their needs on the southwest and
west coasts but are gone from the eastern coast. Jensen
(1928) has given some interesting data on the past and present
distribution. These caribou were frequent on the southern-
most part of the west coast even as late as the end of the
eighteenth century, but they disappeared from this southern
tip of the peninsula during the early part of the nineteenth
century. Nevertheless, a good many still remain not far to
the northward, near Narssalik (south of Frederikshaab) .
The best hunting district is a little farther to the northwest,
Sukkertoppen, at latitude 65 N., where "even now" about a
thousand animals are killed annually. The Godthaab district,
slightly to the south, is also a favored habitat, where Jensen
says about 600 are taken yearly. The present distribution of
these caribou extends still farther northwestward to the
southern part of the Upernavik district, in about latitude 73,
so that the range of the animal now is mainly limited to the
stretch of coast covering about 800 miles from the southern
Frederikshaab district to the Upernavik area. They no longer
occur, however, on Disko Island about in the center of this
stretch. Jensen states that in former times these animals
extended their range still farther north to the Thule district,
"at least as far north as Rensselaer Bay (lat. 78 40' N.),"
where now they have "practically disappeared." Neverthe-
less, a remnant still persists here and apparently forms the

NORTH AMERICA AND THE WEST INDIES 307

dependence of the Eskimo of the Etah and Smith Sound area.
Macmillan (1918), in his account of experiences here, tells of
the Eskimo setting forth on their annual hunt for the caribou,
about 50 miles north of Etah, on September 10, 1914, and
returning on the 23rd with "forty -two warm skins, invaluable
for bed-robes, coats, and sleeping-bags for the extreme tem-
peratures to come." At another time the Eskimos reported
killing 19 in two weeks. In February, 1916, Macmillan
noticed the wandering of the caribou herd southward to the
vicinity of Etah, because of deep snows covering their feeding
grounds between there and Humboldt Glacier, which forms a
natural barrier to the northward extension of their range. Of
six killed near Etah, the heaviest weighed but 120 pounds.
It is believed that this northern herd is much preyed upon by
wolves, for in 1916, Macmillan writes, the Eskimos returning
late in October from the annual caribou hunt throughout the
region extending from Etah to Humboldt Glacier, reported "no
young caribou whatever and tracks of wolves everywhere."
It was his belief that the wolves had crossed over Smith Sound
on the ice from Ellesmere Land and had accounted for the
young caribou. The Eskimo ^stated that they had seen a
number of caribou sleeping on the ice in the center of lakes,
probably for security against attacks by wolves.

Returning to the southern tip of Greenland, Jensen notes
that these caribou were in former times found at Angmags-
salik but are exterminated. During recent decades they have
also disappeared from the coast between Cape Dalton and
Cape Brewster. In the northern part of East Greenland the
Ryder expedition in 1891-92 "encountered reindeer with com-
parative frequency in the region round Scoresby Sound; in
1899 Nathorst only saw a few herds." The "Danmark Ex-
pedition" in 1906-8 found "innumerable traces (cast-off
antlers and excrements) of the former occurrence of reindeer
in great numbers, as far north as Holm Land (lat. 80 20' N.),
but now they have entirely disappeared"; this disappearance
from the northeast coast is corroborated by later reports of
explorers and hunters. Coincident with this disappearance
may have been the abandonment of the region by the Eskimo,
whose deserted camps have been investigated in the same
region. Nathorst, who in 1899 saw only a few herds in the
Scoresby Sound region, attributed this disappearance to wolves,

308 EXTINCT AND VANISHING MAMMALS

which are said to have appeared in East Greenland "rather
suddenly" at about the time of the disappearance of the cari-
bou. This, however, may be only part of the reason. Quite
possibly overhunting had much to do with it, or there may
have been other contributory causes.

From an economic standpoint, the Greenland caribou is an
important asset to the Eskimo and white inhabitants of that
country and one that should be exploited with care. For a
great many years the export of caribou hides to Denmark has
been going on, to an extent that was evidently unwarranted
by the size of the herds. Jacobi (1931) quotes figures showing
a decline in the numbers during part of last century. It is
said that 37,000 caribou were killed in Greenland in 1839;
from 1841 to 1850 an average of 13,900 hides were exported;
from 1851 to 1860, an average of 5,667 yearly; from 1860 to
62, the average yearly number was below 1,000; and since
1862, below 100. In 1891, none was shipped. Nevertheless,
there seems to be some evidence of an increase in numbers
locally at the close of the last century.

"Among land mammals the reindeer is the most important
object of hunting to the Eskimos. The meat is eaten, and the
fat is used, for instance, as cfeam for coffee, and the contents
of the paunch are considered a special delicacy; the skins are
used as underlayers on sleeping platforms and for sleeping
bags and garments, the antlers for hunting implements, the
sinews for thread" (Jensen, 1928). Apparently caribou hunt-
ing is a summer occupation in South Greenland, when the
Eskimo and their families seek the great plateaux in the neigh-
borhood of the inland ice, and having selected their camp sites
the men carry on the hunting till about the first of September,
when they return to their dwellings on the outer coast. In the
Etah region the hunting season is apparently in October.
Since 1924 this pursuit has been somewhat regulated by the
government in South Greenland. It has forbidden the hunting
of caribou or molesting of the calves between May 20 and July
20, and during the open season no more animals may be killed
than can be utilized on the spot or taken care of for future use
by the hunters. Obviously these animals are of great impor-
tance in the economy of the Eskimo, to whose use it would
seem they should mainly be relegated. Macmillan writes
(1918, p. 74) that caribou meat "is lamentably lacking in

NORTH AMERICA AND THE WEST INDIES 309

strength and stamina-producing properties, " yet it is neverthe-
less more palatable than walrus or polar bear and forms a
welcome change from seal meat.

QUEEN CHARLOTTE ISLANDS CARIBOU
RANGIFER DAWSONI Seton

Rangifer dawsoni Seton, Ottawa Nat., vol. 13, p. 260, Feb., 1900 (Graham Island,

Queen Charlotte Islands, British Columbia).
FIGS.: Seton, 1900, pis. 4, 5 (type specimen, antler and skull); Sheldon, 1912, pi.

opposite p. 234.

This is supposed to be a small island form, darker in color
than the barren-ground caribou of the mainland. The single
antler of the type is peculiar in that the brow tine does not
project parallel to the forehead but turns upward and is well
separated from it. The terminal forks of the main beam turn
backward instead of extending forward. Length of type
antler following outer curve to highest point, 28% inches.

Very little is known of this caribou. Indeed, its very exist-
ence in the Queen Charlotte Islands was doubted by Osgood
(1901) and it was not until years after its description that
satisfactory evidence of its presence on Graham Island, the
most northerly island of the group, was obtained. Dr. G. M.
Dawson, while making a geological survey of the islands in
1878, was the first to learn of a caribou there, but from the
meager description given him by the natives he at first was
inclined to suppose the animals were wapiti. Nevertheless he
noted the use of pieces of caribou antler in the implements of
the native Haidas. After some effort the portion of cranium
with an antler that afterward served as the type was secured
through the Indians with the aid of a local trader, on Virago
Sound. In 1904, 1905, and 1906, the Rev. Charles Harrison,
a missionary stationed there, found evidence in the shape of
tracks, dung, and hair. In 1906 the late Charles Sheldon
undertook an intensive search for caribou in the country west
and southwest of Virago Sound on Graham Island but found
nothing more than old tracks and dung. In 1908 a bull and two
cows were killed, and a calf was seen by two half-breeds near
Virago Sound, and the specimens were acquired by the Pro-
vincial Museum in Victoria, B. C. (Keen, 1909). In October,
1910, Francis Kermode found two old tracks in the same dis-
trict, and expressed the opinion that the form was nearly

310 EXTINCT AND VANISHING MAMMALS

extinct. Wharton Huber informed Dr. Francis Harper that
in 1930 two prospectors again found tracks on Graham Island
but saw no animals. This was on grass-covered mountains
about 2,000 feet high in the northwestern part of the island.
I. M. Cowan further informed Dr. Harper that in November,
1935, "Ed White reported fresh caribou tracks on Massett
Island." It seems likely, then, that a few still exist in the
interior mountainous parts of these islands. Since 1910 the
British Columbia authorities have prohibited the killing of
these animals, and because the local Haidas are more a fishing
than a hunting race and do not care for caribou meat, it seems
probable that the law will be observed.

There has been much speculation as to the origin of this
caribou group and how it should have survived in such seem-
ingly small numbers with no known natural enemies. The
nearest point of land is about 30 miles away to the coastal
archipelago, and the nearest relatives of the caribou live 150
miles distant at the present time. Dawson has suggested that
the ancestral stock reached the archipelago during the Ice Age,
when the islands were more easily accessible. If the race
proves to be one of the barren-ground group, as seems likely,
the fact will be all the more interesting.

EASTERN WOODLAND CARIBOU
RANGIFER CARIBOU CARIBOU (Gmelin)

[Cervus tarandus] caribou Gmelin, Linnaeus's Systema Naturae, ed. 13, vol. 1, pt. 1,

p. 177, 1788 (eastern Canada).
FIGS.: Grant, 1902, 2 pis. opposite p. 18 (photographs); Nelson, 1916, p. 459 (colored

fig-)-

The woodland caribou, as its name implies, is a more south-
ern animal than the barren-ground species, inhabiting the
boreal evergreen forests and their bogs quite across North
America from Newfoundland to British Columbia. It is larger
of body but with less elongate antlers than the latter type; the
antlers usually tend to have a more flattened beam, and the
lower incisor teeth are more uniformly graded in size from the
middle to the outer ones. The females occasionally have small
antlers, but these are ordinarily much less developed than in
females of the barren-ground caribou. About five closely allied
forms are at present recognized.

The eastern woodland caribou is usually darker than the

NORTH AMERICA AND THE WEST INDIES 311

barren-ground species, with the head in front of the ears, the
body from shoulders to tail, and the legs to just above the
hoofs a dark brown; the lips and neck with its fringe of long
hair below, and the legs just above the hoofs white or hoary.
There is a narrow white area bordering the buttocks, but the
upper surface of the short tail is dark. The antlers of the male
are shorter than in the barren-ground type, less sweeping, and
extend outward and forward, with a short tine at the bend,
palmate brow tine on one or both sides and a varying number
of tines on the terminal part of the main beam. Large antlers
may measure in length of beam 42 inches on the outside curve.
The range of this eastern race originally extended from Nova
Scotia westward across New Brunswick, northern Maine, ex-
treme northern New Hampshire, and Vermont to southern
Ontario, and on the north side of the Gulf of St. Lawrence to
the wooded parts of southern Labrador west to James Bay.
Westward of the latter region it grades into the race sylvestris.
Its southward limit was the north shore of Lake Superior. It
apparently did at one time reach the Adirondack region of
New York. Over this range the woodland caribou is now
largely gone, chiefly as a resulfe of hunting and partly perhaps
as a result of lack of reproduction on account of reduced living
area and the harrying of civilization. The caribou south of
the St. Lawrence have probably long been cut off from their
western and northern neighbors since the settlement of southern
Canada and may be thought of as a separate herd having a
tendency to wander irregularly over the area. Thus during
the nineteenth century at intervals of about 15 years, the
records show a great influx of their numbers into northern
Maine, so that they came down as far as tidewater in the
eastern part of the State and to the large bogs near Bangor in
the interior. For a few years there would be caribou in some
numbers in the Maine woods, then they would drift away again
eastward into New Brunswick and Nova Scotia and might be
nearly if not quite absent from Maine for another period.
New Brunswick with its greater extent of barrens and southern
Quebec south of the St. Lawrence were probably more attrac-
tive feeding grounds. Caribou are easily killed on account of
their gregarious habits and a curious inquisitiveness that leads
them to stop their flight and approach an unfamiliar object.
In former times a hunter on being discovered might lie on his

312 EXTINCT AND VANISHING MAMMALS

back and kick his legs in the air or wave a red cloth, when
presently the whole herd would turn and approach to investi-
gate. The hunter, waiting his chance, would leap up and get
in several shots before the caribou could dash off. This south-
ern herd must have become gradually reduced by hunting and
the effects of settlement without much cognizance being taken
of it. The last great infiltration in Maine was in the early
nineties, when they were reported for a time more numerous
than deer. In 1895 and again in 1896 the Bangor and Aroostook
Railroad alone shipped out about 130 caribou each year that
had been killed by visiting sportsmen, but after that the num-
ber in the State began to decline, owing in part probably to the
eastward movement of the animals once more and in part to
summer and winter killing in both Maine and the adjacent
parts of Canada. There were a few caribou in the bogs of
northern New Hampshire until at least 1885. In 1899 a closed
time on caribou in Maine for six years was established and on
its expiration was renewed for another six years. Meanwhile,
however, the animals had practically gone from the State.
The last caribou in the Mount Katahdin region, formerly their
favorite haunt, was seen in 1905 and a small herd remained on
the northwestern border of the State near the St. Johns River
until 1916, but since then there seem to be no certain records.
The history of this herd is much the same in New Brunswick
and Nova Scotia a gradual diminution and retreat. Dr.
R. M. Anderson (1939b) writes that in Nova Scotia caribou
were so numerous in the sixties that a certain English sportsman
killed 120 in a single day in Lunenberg County. "The last
caribou on the mainland of Nova Scotia was killed in Guys-
borough County about 1912," and they were "almost certainly
gone" from the once famous caribou grounds of Victoria and
Inverness Counties in northern Cape Breton Island by 1925,
although in only the previous year he found recent "traces"
of them. "In New Brunswick a closed season was estab-
lished for caribou in 1919," but it was too late to help them,
for in a few years they seem to have altogether gone. "Fairly
authenticated records in the upper Tobique River in 1924"
are mentioned by Dr. Anderson, and the New Brunswick
Game Department is quoted as saying that "reports from
wardens and woodsmen show that caribou probably remained
in the Province in small quantities until about 1926, and one

NORTH AMERICA AND THE WEST INDIES 313

warden actually reported seeing five caribou east of Bartibogue
Station during the spring of 1927." Since then, however, the
herd south of the St. Lawrence seems to have vanished, except
for a small remnant still surviving in the Gaspe region. This
remnant is chiefly confined to the Shickshock Mountains in
the northern part of the country and though small is never-
theless of sufficient size to allow a few to be shot each year.
There is a considerable barren area on the summits of some of
these mountains that is attractive to caribou, and the sur-
rounding wilderness is at present little disturbed. As the last
remaining woodland caribou south of the St. Lawrence River,
this little group should be given careful protection until it
appears that it is numerous enough to survive. One may
believe, however, that it will not persist for a great many years,
when annually hunted. In southern Labrador "a few" wood-
land caribou are found "here and there in the interior north of
the St. Lawrence River and a small number scattered through
the back districts of northwestern Quebec . . . The situ-
ation in Ontario is even less hopeful. On a map prepared at
the Royal Ontario Museum of Zoology in 1935, caribou are
considered to have disappeared Entirely from the east of a line
drawn from east end of Lake Superior to James Bay, the dates
of extirpation being dotted on the map here and there 1894,
1908, 1912, 1919, 1926, etc. The most southern band at present
is said to occupy Shakespeare Island in Lake Superior, with a
few about Lake Nipigon, Lake of the Woods, and the Rainy
Lake area. Local bands exist north of the Canadian National
Railway lines, but the most recent reports are to the effect
that the numbers of caribou are not large even in the more
northern parts of Ontario." The range in southern Labrador
probably extends to Hamilton Inlet and Sandwich Bay on the
east, and on the Hudson Bay side to Great Whale River or
thereabouts. According to Eidmann (1935) it has been killed
out over most of this range, but small herds are still occa-
sionally met with, especially in the southeastern part of Labra-
dor, where a small herd appeared in the Matamek area in 1930
and several were killed. Its extermination has been especially
hard for the native Indians, who relied on it in part for food
and clothing.

Minnesota is the only one of the eastern United States in
which caribou still exist in a wild state. C. L. Herrick (1892)

314 EXTINCT AND VANISHING MAMMALS

found them 50 years ago "only about the headwaters of the
White-face River and along the St. Louis River near Knife
Fall. There it was in 1884 not rare, though so shy as to be
secured with difficulty. Along the North Shore of Lake
Superior it is less shy and the animals may be seen feeding
quietly in groups along the upland meadows." G. S. Miller
(1897b) in 1896 found these caribou "very abundant on the
north shore of Lake Superior" and saw "heads, antlers, and
jaws of caribou at White River, Peninsula Harbor, Schreiber,
and Nepigon. A wet pasture among the hills a mile or more
northeast of Peninsula Harbor is a favorite feeding ground of
these animals. " At the west end of Lake Superior, in what is
now Quetico Provincial Park of Ontario, caribou are extinct,
"although this region was once the approximate southern
limit of its winter migration. There are no records in recent
years. Nash reported it as very shy but not rare in the White-
face and St. Louis River country of northern Minnesota as
late as 1894 . . . There is an abundance of its favorite
food, the caribou moss, Cladonia rangiferina, throughout the
region" (Cahn, 1937). A century ago there were caribou
in Chippewa County, northern Michigan (Schorger, 1940). A
recent report by BreckenridgeXl935) shows that a last remnant
of the woodland caribou still holds out "in the muskeg country
lying between Upper Red Lake and Lake of the Woods in
northwestern Minnesota. " In company with game officers he
spent from February 28 to March 3, 1935, investigating the
status of these animals. Of three adults seen at close range on
March 1, one retained its antlers. Evidence of at least six ani-
mals in this region was obtained, a number that is reported in the
August, 1939, Biological Survey census to have been increased
to 12, but the basis of this estimate is not given. The district
where these caribou live is now included in the Red Lake
Game Refuge, where they may be considered fairly safe except
from wolves which still are found in small numbers. The
refuge includes over 400,000 acres of land, "which is prac-
tically worthless, not only to the agriculturalist, but to the
forester" (Swanson, 1936).

Finally, in the hope of building up the stock more rapidly,
W. T. Cox (1941), in charge of the refuge, arranged to introduce
new stock from Saskatchewan. With the aid of Indians, seven
calves and an adult bull were captured alive and after pre-

NORTH AMERICA AND THE WEST INDIES 315

liminary conditioning were brought in 1938 to the refuge.
Nearly a year later the bull was seen ranging with the cows.
The calves, however, were kept in a corral and small pasture
and later were transferred to a larger fenced area, to be liberated
later. At last accounts the little group was thriving encourag-
ingly, and every effort is being made to ensure the success of
the attempt at colonization. The danger is that the animals
may wander outside the reserve and fall prey to settlers or the
Indians on a nearby reservation.

To sum up, the woodland caribou, once so common in the
muskeg areas from Nova Scotia to western Minnesota and
north to southern Labrador and James Bay, is now everywhere
so greatly reduced that it is in actual danger of extermination.
South of the St. Lawrence a single small group still exists in
the mountains of northern Gaspe; it is gone from Nova Scotia,
New Brunswick, southern Quebec and northern New England.
A few scattered groups still remain in southern Labrador, but
to the westward it is all but gone from Ontario, and a very
small group still holds out in northern Minnesota. The main
factor in its extermination has no doubt been overshooting;
other factors are doubtless the animal's innate aversion to the
presence of settlements, thus tending to a restriction of its
range and lowered prolificness, its gregarious habits and curi-
osity often making easier the killing of several individuals at
a time by hunters; finally forest fires may affect the food
supply, and wolves, though fewer now than formerly, may be
a factor. Of possible diseases nothing is known, but that deer
of this group are subject to serious attack by bot flies is well
ascertained, although what effect these may have in causing
lowered vitality is not clear.

NEWFOUNDLAND CARIBOU
RANGIFER TERRAENOVAE Bangs

Rangifer terraenovae Bangs, Preliminary Description of the Newfoundland Caribou,
p. 2, Nov. 11, 1896 (privately printed) ("Codroy, Newfoundland").

FIGS.: Grant, 1902, two pis. unnumbered (14, 15) (heads); Prichard, 1910, col. pi.
opposite p. 80; Dugmore, 1913, many photographs.

The caribou of Newfoundland is not very different in general
from the woodland caribou of the neighboring mainland, even
though usually accorded specific status. It is said to be lighter
in color, the back, sides, and legs drab, somewhat mixed with

316 EXTINCT AND VANISHING MAMMALS

yellowish- white hairs paling on flanks to white. Face and inner
surface of ears rich bister brown, nose and chin white; tail
short, drab above, white beneath (Bangs). The antlers are
rather distinctive, being "low, widely spread, much forked and
with the points pointing forward and inward" (Bangs). They
are, in a well-developed head, rather shorter of beam, with
large palmated brow and bez tines. The skull is said also to
be somewhat larger than in the typical woodland caribou.
An adult male stands from 46 to 49 inches at the shoulder.
Does are in about half of the cases provided with small branch-
ing antlers. Weight of a full-grown male up to about 300
pounds.

Both Prichard (1910) and Dugmore (1913) have given
excellent accounts with illustrations of this caribou. According
to the former, the population of this species in the first decade
of the present century was said to consist of three groups:
One north of the railway that crosses the island from east to
west, especially frequenting the Humber River valleys and
Birchy Lake region; the main herd, largely south of the rail-
road, inhabiting the central and southern parts; and a small
"herd" consisting of a few animals that still survive in the
Avalon Peninsula, south of St. John's. The last group is
believed to be nonmigratory, but the other groups perform
annual north and south migrations, with more or less regularity,
although they may not always be found at the same places in
different seasons. The first of these groups is said to be the
best known and most accessible and crosses the railway in its
southward wanderings, especially in the vicinity of Howley.
To this region resort the greater number of sportsmen in
search of trophies and of settlers who to some extent depend
on caribou meat.

In 1913 Dugmore believed that a conservative estimate of
the caribou population of the island was not less than 150,000
animals. Some estimates were much larger. It is difficult to
judge of the general status, since those visiting the hunting
grounds annually for short seasons may not find the larger
groups each time or the caribou may be for various reasons
less in evidence in different years. The interior barrens are
vast and travel is difficult. Nevertheless the consensus seems
to be that the numbers have greatly declined in the last two
decades, owing in large part to overshooting. Jacobi (1931)

NORTH AMERICA AND THE WEST INDIES 317

writes that in February, 1899, an observer saw 550 carcasses
unloaded in the harbor of St. John's and two weeks later about
the same number, so that the price of meat fell to 3 cents per
kilogram. However, as late as 1913 they were still common
and could be seen in large bands, especially during the migra-
tions, but large heads were more difficult to secure.

In a letter to Dr. Francis Harper, dated October 4, 1934,
Greville Haslam writes that he had spent nine of the previous
fourteen summers in the island, and found in 1919, 1920, and
1921 "a few caribou" each year in the region north of the
railroad near the Humber River and Birchy Lake. "There
was no question in anybody's mind that the caribou were dis-
appearing, but they all claimed that this was because the
animals no longer migrated, and that they would be found in
large numbers in some other section, especially the country
lying at the head of the Gander River and extending around
Mount Cormack and west around Meelpaig Lakes. In 1929
we spent three weeks in this country but saw not a single cari-
bou and only a few tracks made some months before. In the
summer of 1934 we spent three weeks on the Serpentine River
but saw no tracks." It seems \hat the numbers have appar-
ently fallen off considerably in late years, but exact statistics
are unavailable. At all events there was a general impression
in 1934 that caribou were again increasing and in that year
the Government permitted an open season, charging $50 for
a license to shoot a caribou. Brooke Dolan reported that
according to a Newfoundland official, 60 licenses (including
two to aliens) were issued in 1936 and that some caribou were
killed. One herd of about 700 was reported seen north of the
railroad, but shooting was permitted to the southward of the
railroad only. In 1938 John K. Howard found small numbers
a little distance north of the head of White Bay. Thus it
seems that there are still a good many caribou left in the
country, but in order that the numbers may be maintained at
an economic level some care in administration is needed.

Dugmore (1913) presents some interesting notes on habits.
Summer in Newfoundland begins late in June. During this
month the young caribou are born, and at this time the does
seek the thick forests of spruce and fir. Usually there is a single
young at a birth but twins are not rare. With the passing of
wolves on the island, the lynx is about the only potential

318 EXTINCT AND VANISHING MAMMALS

enemy. On the other hand, the swarms of black flies and
mosquitoes must be a great source of annoyance to the animals
as they are to human beings at this season. "During the
warmer months the caribou are more or less solitary in habit,
going about singly or in pairs and only rarely in small herds of
half-a-dozen or more. In the day-time they keep very largely
to the woods, coming out to feed at the approach of evening
. . by September their habits have completely changed
and they become almost entirely diurnal," and bands gather.
An attempt is being made to introduce this caribou from
Newfoundland into Nova Scotia, in the hope that it may, if the
plan is successful, in time replace to a certain extent the
eastern woodland caribou, which has for a number of years
been extinct in this province. On April 10, 1939, nine females,
of which five were "with calf," arrived in Halifax from New-
foundland. The five pregnant females were at once taken to the
Liscombe Game Sanctuary in Guysboro County, while the
four others were being held temporarily at Halifax pending the
arrival of the three males, which were to be imported in order
to start this new herd (R. W. Tufts, 1939). The result of this
experiment may be awaited with interest.

WESTERN WOODLAND CARIBOU
RANGFIER CARIBOU SYLVESTRIS (Richardson)

Genus tarandus var. (3 sylvestris Richardson, Fauna Boreali-Americana, vol. 1, p. 250,

1829 (southwestern shores of Hudson Bay).
FIG.: Hollister, 1912b, pi. 2, fig. 2 (skull).

The western woodland caribou is much the same as the
eastern animal, but according to Hollister (1912b) it may be
distinguished by its longer and slenderer skull, with a narrower
rostrum and larger teeth. The tooth rows, especially the lower,
are longer. The neck and head are darker, the ears, back, and
sides of the neck are much darker, the hairs brown to their
roots. Total length of a male skull, 417 mm.; upper tooth
row, 107 mm. These characters, while not very sharply
marked, are supposed to distinguish the animals west of Hud-
son Bay; presumably those of the Lake Superior region and
James Bay are intermediate but may be referable to this
rather than to the typical race.

The precise range of this race has not been well defined but
is believed to have originally extended from the southwestern

NORTH AMERICA AND THE WEST INDIES 319

shores of Hudson Bay westward across northern Manitoba to
the Anderson River, overlapping the winter range of the
barren-ground caribou to some extent and possibly reaching
the Mackenzie River region.

Dr. R. M. Anderson (1939b) writes that it is widely dis-
tributed in the timbered portions of the Northwest Territories,
but never common, ranging as far north as Great Bear Lake.
It is fairly certain that human settlement is the controlling
factor in the present range of the species. "Small numbers of
woodland caribou are found in the wooded parts of northern
Saskatchewan and Alberta, but many reports indicate that
the large number of trappers and prospectors who have
strung trap lines all along the northern parts of the Prairie
Provinces are rapidly causing the woodland caribou to disap-
pear. Bush fires destroy their winter food of lichens which are
slow to recuperate, and the woodland caribou are shy of settle-
ments as well as being easy to kill on the 'barren' openings in
the forest in winter." Writing of this caribou in 1902, Preble
says that he found it throughout the region he traversed be-
tween Norway House and Hudson Bay. Between York
Factory and Fort Churchill 3, few small bands are found
throughout the year on the "barrens." Dr. Milne, who had
resided 14 years at York Factory, told Preble that Cape
Churchill was considered a good place to hunt them at any
time of year and that he believed the small bands occurring
here formed the northern fringe of those that migrate to the
coast in spring, the great majority of which cross to the south
of Nelson River. Their return movement occurs from about
the middle of October to the last of November. "During
these semiannual movements the animals are much pursued,
especially in the fall, when the weather is usually cold enough
to preserve the meat for winter use." A resident at Oxford
House said that the species was much less common than
formerly. In his report on mammals of the Athabaska-
Mackenzie region, Preble (1908) wrote that it is found in the
"country between Lake Winnipeg and Athabaska Lake, and
though nowhere in large numbers is more abundant on the
southern than on the northern shores of this lake. Between
Athabaska and Great Slave lakes . . . the animal is
met with chiefly on the west side of Slave River, and through
all the country lying between Peace River and Great Slave

320 EXTINCT AND VANISHING MAMMALS

Lake." Caribou still exist in small numbers in the Peace
River district, as in the vicinity of Peavine Lake and on
Mount Bickford and Tuscoola Mountain (Cowan, 1939), and
are presumably of this race. According to Anderson, there is
some evidence that in areas visited by game officers in the
winter of 1937-38, the caribou are increasing. With proper
protection, they should hold their own in most of their range
for the present.

MOUNTAIN CARIBOU
RANGIFER MONTANUS Seton

Rangifer montanus Seton, Ottawa Nat., vol. 13, p. 129, Aug., 1899 (Selkirk Range,

near Revelstoke, British Columbia).
FIGS.: Allen, J. A., 1902b, pp. 154, 155, 156, 157, figs. 3-6 (antlered skulls).

RANGIFER FORTIDENS Hollister

Rangifer fortidens Hollister, Smithsonian Misc. Coll., vol. 56, no. 35, p. 3, Feb. 7, 1912
("Head of Moose Pass branch of the Smoky River, Alberta (northeast of
Mount Robson) ").

FIG.: Hollister, 1912a, pi. 10 (antlered skull of type).

Hollister, in 1912, described the large caribou of the woodland
type from Mount Robson region as a species distinct from R.
montanus of the Selkirk and Gold Ranges in southeastern
British Columbia. The two must be closely related (the type
localities are little over a hundred miles apart) and may even
prove to be much the same when further study can be made
with adequate specimens. Hence it is convenient to treat both
together here.

Both agree in being darker in color than the more eastern
forms. Dr. J. A. Allen (1902b) describes R. montanus in
September coat as in general terms "a black caribou, with the
neck and shoulders, especially in the males, much lighter than
the body and limbs." The "whole body and legs blackish
brown, varying (in different specimens) to glossy black over
the middle of the dorsal area from shoulders to rump; lighter,
more brownish black, on the flanks and ventral surface;
inguinal region, sides and under surface of tail, a narrow band
bordering the hoofs, and ventral median line of neck, grayish
white; nose and edge of lips grayish white; sides of neck
grayish brown varied with blackish and, in the males, tinged
more or less with rusty. The females are much darker than
the males, especially on the neck and shoulders, but have the

NORTH AMERICA AND THE WEST INDIES

321

grayish white areas of the males replaced by nearly pure
white." The distinction between the two "species" seems to
be that in R. montanus the size is somewhat less in skull and
teeth, with shorter beams to the antlers and a more exuberant
development of tines ; whereas fortidens is said to be largest of
the caribou with less branching and less shortened antlers.
No doubt both should eventually be regarded as races of R.
caribou.

The range of the more southern montanus was the Selkirk
and Gold (or Columbia) Ranges and doubtless the adjacent
country, southward to the northern border of Washington.

Newfoundland caribou (Rangifer terraenovae)

322 EXTINCT AND VANISHING MAMMALS

Anthony states that it formerly reached the mountains of
Montana and Idaho, but certainly it has not been known in
these States for a long time. In the State of Washington,
Taylor and Shaw (1929) wrote that it was formerly "of
occasional occurrence along the Canadian boundary in the
northeastern part of the State, south to Usk, and west to
Okanogan County" but that it was exterminated a decade
ago (about 1920). Nevertheless the latest wildlife census by
the IT. S. Biological Survey credits the national forests of the
State with an estimated four individuals in 1938. In a letter
to Dr. Francis Harper in April, 1937, I. M. Cowan wrote that
in British Columbia there are still a number of montanus but
none now south of the main line of the Canadian Pacific Rail-
way. In general, caribou are "definitely on the wane through
overhunting," he says, "mainly by Indians."

Cowan's statement probably applies in part to the Mount
Robson form, fortidens, as well, for he says they are "almost
gone" in the Chilcotin region. Hollister in 1912 said they were
not common along the line of the railroad east of the boundary
line of Jasper Park but were formerly common down to the
eastern foothills. Dr. R. M. Anderson wrote in 1938 that
R. montanus "is still fairly common in the Rocky Mountains
of western Alberta, but the animals are extremely scarce on
most of their former range in central and southern British
Columbia." In 1929 he had "fairly conclusive evidence of the
presence of a small band ranging along Summit Creek on both
sides of the British Columbia-Idaho border, and if these still
exist they are probably the most southern representatives.
At that time a few were still supposed to be in the mountains
southwest of Nelson, British Columbia." Carl Rungius (in
Ely et al., 1939) has lately published an account of hunting R.
fortidens in the Mount Robson region, but he gives no dates.
At the time, caribou were fairly numerous there.

Family ANTILOCAPRIDAE : Pronghorns
THE PRONGHORN ANTELOPE

The pronghorns are the last surviving members of a family,
the Antilocapridae, that has developed exclusively in North
America. Though related to the true antelopes of the Old

NORTH AMERICA AND THE WEST INDIES 323

World, they differ in that the horn sheaths are forked instead
of simple and are shed and renewed annually instead of being
permanently retained. In many of the fossil forms, the bony
core of the horns was forked to correspond to the forked sheath,
but in the living species this front branch of the horn core is
represented by a slight bulge only. In some of the extinct
species the horns showed a twisted core recalling that of the
Old World bushbucks. Only a single species is known to have
lived down to the Recent period, but smaller forms, such as
Capromeryx, with a forked horn core, must have persisted to no
very distant period, and have left their remains in the tar pits
of Rancho La Brea, California, and elsewhere in the Southwest.
Four races of the pronghorn have been described, but they
differ in such slight characters that they may be best con-
sidered together.

PRONGHORN ANTELOPE; PLAINS PRONGHORN
ANTILOCAPRA AMERICANA AMERICANA (Ord)

Antilope americana Ord, Guthrie's Geography, 2d Amer. ed., vol. 2, pp. 292, 308, 1815

("Plains and highlands of the Missour^").
FIGS.: Stone and Cram, 1902, pi. facing p. 64 (photographs); Nelson, 1916, p. 451,

upper fig. (col.); 1925, pis. 1-6 (photographs and habitat).

MEXICAN PRONGHORN
ANTILOCAPRA AMERICANA MEXICANA Merriam

Antilocapra americana mexicana Merriam, Proc. Biol. Soc. Washington, vol. 14, p. 31,
Apr. 5, 1901 ("Sierra en Media, Chihuahua, Mexico").

OREGON PRONGHORN
ANTILOCAPRA AMERICANA OREGONA V. Bailey

Antilocapra americana oregona V. Bailey, Proc. Biol. Soc. Washington, vol. 45, p. 45,
Apr. 2, 1932 ("Hart Mountain (Warner Mts.), Oregon").

LOWER CALIFORNIA PRONGHORN
ANTILOCAPRA AMERICANA PENINSULARIS Nelson

Antilocapra americana peninsularis Nelson, Proc. Biol. Soc. Washington, vol. 25, p.
107, 1912 ("Forty-five miles south of Calmalli, Lower California").

Light of body and limb, the pronghorn stands about 34 to
36 inches at the shoulder, with a total length of about 54 inches
in adult males, females slightly less (Anthony). The forehead,

324 EXTINCT AND VANISHING MAMMALS

nape, and upper part of the back and the outside of the limbs
are rich reddish brown, with a mixture of blackish on the
muzzle and in the short mane. On the throat the reddish
brown of the upper parts extends as two collars, separated by
a white space, around the throat, although the lower one is
often incomplete in the middle line below. Under parts and
flanks and a prominent rump-patch, white. Tail and a central
dark line dividing this patch are colored like the back or
darker. Horns erect and diverging, the posterior and longer
point curving backward, the anterior point short and blunt,
projecting forward.

The Mexican race is similar but paler, with a tinge of cinna-
mon. That of the peninsula of Lower California has the ears
darker, with the facial markings dark and strongly contrasting
with the pale areas. The Oregon race is slightly larger than
the typical race of the plains, with relatively larger feet,
longer horns, and slightly paler color. The distinctions between
these races are slight and the distributional areas need more
careful mapping.

Pronghorns are characteristic of open plains country of a
semi-desert nature and seldom enter tree-grown areas except
in winter to find shelter from storms. Their flashing of the
erectile white hair of the rump-patch makes a remarkably
striking semaphore visible at a great distance in sunlight and
affords an automatic danger signal to others of their kind.
They go in bands of varying size according to the circumstances
and when once alarmed can reach a great speed, seeming
fairly to fly over the ground. A curious habit seen in this and
in some other animals, as gazelle in the Gobi or seabirds in open
ocean, is a seeming desire to match speed with pursuers, and,
attaining a sufficient lead, to cross in front. Another char-
acteristic trait is curiosity, leading animals often to approach
unfamiliar appearing objects. Nelson tells of enticing shy
animals up within gunshot by donning a white sheet and ap-
proaching them on all fours.

In an important paper on the status of the pronghorned
antelope, Dr. E. W. Nelson (1925) presents a careful summary
of its former and recent distribution. It originally ranged over
an enormous area, from the present provinces of Manitoba,
Saskatchewan, and Alberta in the north to the southern part
of Texas and the Mexican tableland in the south, and from the

NORTH AMERICA AND THE WEST INDIES 325

eastern edge of the plains in Minnesota, Iowa, Kansas, and
Oklahoma westward to eastern Washington, Oregon, and the
valleys and coast of southern California and most of the penin-
sula of Lower California. "In Mexico it occupied the open
plains country of the tableland south almost to 20 of latitude,
nearly to the valley of Mexico " and the western part of Sonora.
Originally it was abundant and well distributed over this
territory and exceeded the bison in numbers where both
occurred together. It is estimated by Nelson that at the time
of the settlement of this continent by Europeans the prong-
horn population was "not less than thirty to forty millions,
possibly more." A recent census of those still existing in the
United States, Canada and Mexico (1922-24), indicates that
of these vast numbers about 30,000 then remained, but a later
figure (1939) makes this over 180,000.

The pronghorn just reached the prairie country of south-
western Minnesota in its northeastward distribution, but
Herrick, writing of it in 1892, speaks of it as having "long since"
gone. That it may even have penetrated farther east is
possibly indicated by the recent discovery of a horn buried in
a few inches of earth at Moline, ill., although it may have been
brought there by Indians in earlier years (Fryxell, 1926).
Formerly abundant all over the Dakotas, it began to decline
in the late seventies, and with the encroachment of settlement
gradually disappeared. By 1924, Nelson wrote that but five
small herds aggregating 225 animals remained, and these in
the southwestern corner; in South Dakota about three times
that number were found in the western half of the State.
"Of the countless thousands of antelope which once roamed
the plains of Nebraska but 10 small bands remain, containing
a total of about 187 animals." "At one time Kansas was
inhabited by myriads of pronghorns, and for years after the
construction of the transcontinental railroads they were a
familiar sight to passengers on the trains. In 1923, however,
they had become almost exterminated throughout the State,"
and the few remaining were to be found only in the extreme
southwest corner where they wandered at times into the
adjacent parts of Oklahoma. Otherwise the species once so
common in this State was gone, but in 1910 an attempt was
made to establish a herd in the Wichita National Game Pre-
serve in Comanche County. After several unsuccessful im-

326 EXTINCT AND VANISHING MAMMALS

portations of antelope from Yellowstone National Park, the
difficulties were finally overcome, and in 1925 Nelson reported
that the birth of three pairs of young had brought the total
number of this small band up to 17.

Except for Oregon the Rocky Mountain States seem now to
hold the largest antelope populations. Nelson's account shows
Montana with some 44 herds and a total population of over
3,000; Wyoming leads with 27 areas in which a total of nearly
7,000 is carried; Colorado has 28 localities mostly in the eastern
half of the State where over 1,200 pronghorns are found; in
New Mexico the numbers were decreasing, but a total of 31
localities and nearly 1,700 animals was recorded in 1924.
In Arizona the estimates were 18 bands totaling about 650
animals; they appeared to be increasing on cattle ranges but
decreasing on sheep ranges in the State. In Utah the number
of bands was 10, with about 670 animals; in Idaho 14 bands
with about 1,500 animals. In Nevada, where once they were
plentiful, they were found in only 11 limited areas, but since
some of these are areas sparsely occupied by man the numbers
of antelope are large, aggregating over 4,200. Southeastern
Oregon, where the form oregonus is found, "forms part of a
rough, rocky desert covering also northern Nevada and south-
western Idaho, on which natural conditions have been ex-
ceedingly favorable for antelope. This region constitutes one
of the few areas in the United States where large herds of these
animals numbering hundreds still continue to congregate during
the winter season." Here they have increased in recent years
and were in 1924 estimated at over 2,000 animals. In Texas,
antelope formerly abounded on the plains in the western part,
"but with the occupation of the country they have decreased
until it has been possible to obtain definite information of only
42 existing bands, numbering about 2,400 animals, for the
entire State." While most of these are in the western third, a
few hundred are also found in the southwestern tip. A few
antelope remain in California, where once they were abundant;
in 1923 there were six widely separated areas in the State
where small bands were found totaling about 1,057.

In comparison with these estimates of 1924, the following
issued by the U. S. Biological Survey are significant: South
Dakota, 4,508 (large increase) ; Nebraska, 750 (large increase) ;
Kansas, none (decrease); Oklahoma, 36 (increase); Montana,

NORTH AMERICA AND THE WEST INDIES 327

6,740 (doubled); Wyoming, 24,071 (more than tripled); Colo-
rado, 1,770 (slight increase); New Mexico, 26,564 (great in-
crease); Arizona, 9,410 (great increase); Utah, 35,350 (great
increase); Idaho, 12,328 (great increase); Nevada, 12,700
(about tripled) ; Oregon, 28,550 (great increase) ; Texas, 9,075
(great increase); California, 14,212 (great increase). The
estimate published by Nelson in 1925 gave a total of 26,604
pronghorns in the United States; the 1939 estimate by the
U. S. Biological Survey was 186,114, or nearly a sevenfold
increase for the same States. It is clear that the animal is no
longer in danger; in fact, in some areas it is common enough to
cause resentment among the ranchers for fear that the herds
will make the grazing less available for their cattle. In several
of the States antelope refuges have been established, which,
especially in Nevada and Oregon, are well populated by the
animals.

Obviously, with the continued increase of these herds and
their tameness as a result of protection, there will come points
where the numbers are too great, and their competition for
food with domestic stock must result in measures being taken
to reduce the numbers to a reasonable proportion. It remains
to be seen whether this is best done by having brief open
seasons such as have been declared in Wyoming and Nevada
or in some other way. The natural enemies of the pronghorn
are chiefly coyotes, which kill the fawns, and bobcats, which in
winter may kill a few. Wolves formerly were doubtless a
principal enemy. With present measures against these ani-
mals, the menace seems slight. These antelope are said not to
do well under fence, for it has been found that "within such
areas" they seem "to lose their freedom of movement and be-
come extraordinarily helpless. This is particularly the case
during heavy snowstorms, when they remain within more or
less definite areas, in which predatory animals capture them
with surprising ease."

In Canada the pronghorn formerly ranged eastward to
southwestern Manitoba but is now restricted to southwestern
Saskatchewan and southern Alberta. Nelson's (1925) estimate
for 1924 gave a total of 1,327 antelope for Canada, but Ander-
son (1939b) shows a figure nearly double this, about 2,400 in
1932, as a result of protection, and adds that in 1938 the Fish
and Game Commissioner of Alberta estimated the antelope

328 EXTINCT AND VANISHING MAMMALS

population of that province at 15,000, and there have been two
short open seasons. In certain districts they have even ex-
tended their range north of the South Red Deer River, as well
as into the adjacent parts of Montana and Saskatchewan.
In the latter province there was an open season in 1936 during
which 267 antelope were taken.

In Mexico the situation is less clear, but Nelson's (1925)
report gives an estimate of about 2,400 animals, of which 500
are in Lower California, the rest being in the States of Sonora
adjoining and Chihuahua, Durango, and Coahuila to the east-
ward. In Lower California the pronghorn was formerly
found over much of this desert country nearly to the tip of the
peninsula, which is mountainous and unsuitable for them.
At present the range extends, according to Nelson, only to the
basal half, on the plains east of the central mountain range,
and on the Desert of Vizcaino west of it. With the establish-
ment of a close season at the time of Nelson's report, he be-
lieved that their prospect of surviving in these sparsely settled
desert lands was very good.

From the human viewpoint pronghorn antelope may be a
source of food in some areas, and they have an esthetic and
recreational value as well for those who enjoy the sight of wild
game or take pleasure in hunting it with gun or camera.
Large herds such as gather in winter may, on the other hand,
cause some competition with grazing stock of which ranchmen
are jealous. On the whole, however, it seems that the preserva-
tion of the species at least in certain large areas of our West is
well assured at present and that its numbers can be readily
controlled where the necessity exists. Some pertinent sugges-
tions for the management of this antelope in the arid South-
west have been put forth by Dr. W. P. Taylor (1936), who
points out that in these regions herds of pronghorns in severe
winter weather are wont to come down off the mountains to
lower levels where they come into competition with cattle on
the ranches for winter food. Ordinarily they do not like sheep
ranges but prefer those suitable for larger stock. These factors
are often of critical value in the selection of reserves for their
preservation. The ranchmen have often done much to en-
courage these animals and give them protection.

NORTH AMERICA AND THE WEST INDIES 329

Family BOVIDAE: Sheep, Goats, Cattle
THE MUSKOXEN

Much has been written on the relationships of the muskox to
other ruminants. Although many fossil skulls have been found
of this and related animals, there seems to be little evidence
that the type represented by the living species was derived
from the Old World. It is therefore believed that it must have
originated in the New World and in late geological times spread
into northern parts of the Old. While some zoologists have
thought it related to the sheep or the goat-antelopes, the latest
suggestions favor its close affinity to the bison. The single
living species is of high Arctic distribution from (until recently)
Alaska to Hudson Bay and the Arctic Archipelago, and thence
across northern Greenland to the eastern coast of that country.
Over this area it becomes differentiated locally into three
generally recognized races, for a full account of which the
reader is referred to the extensive papers of J. A. Allen (1913)
and Elisabeth Hone (1934).

BARREN-GROUND MUSKOX
OVIBOS MOSCHATUS MOSCHATUS (Zimmermann)

Bos moschatus Zimmermann, Geographische Geschichte, vol. 2, p. 86, 1780 (based on
Pennant's description; type locality "therefore the region adjoining Hudson Bay
between the Seal and Churchill rivers (about latitude 59) ", Keewatin (J. A.
Allen, 1913)).

SYNONYM: Ovibos moschatus mackenzianus Kowarzik, Fauna Arctica, vol. 5, pt. 1, pp.
97, 116-122, 1910.

FIGS.: Allen, J. A., 1913, figs. 28, 30, 32, 39 (exterior and skull).

The typical race of the muskox is "very dark, nearly black
on the head, neck, sides, and underparts, with the feet and
nose white; the back is lighter (browner black), with a still
lighter "saddle" behind the shoulders. There is no white
area on the head in adult males, although individual white
hairs may often be found on the face; in young animals and
females there is sometimes quite a trace of white on the front
of the head" (J. A. Allen, 1913). The long hair of the body
forms a protecting fringe on the sides. The horns are charac-
teristically dark brown and are very broad at the base in pro-
portion to their length. They sweep downward and slightly
forward, then turn upward to form a hook; the record length

330 EXTINCT AND VANISHING MAMMALS

of horn on the outside curve is 29 inches. A large bull killed
at Aylmer Lake was measured by Seton as follows: Total
length, 96 inches; tail, 4 inches; height at shoulder, 59 inches.
The basal length of the skull in adult males is about 466 mm.,
with a maxillary toothrow of 132 mm.

The continental range of the muskox formerly was more or
less coextensive with the barren grounds west of Hudson Bay,
extending to the Arctic coast and the adjacent islands, which
must have been reached by crossing on the ice in winter. On
account of certain characteristic habits this species became
peculiarly liable to extermination when hunted by men with
modern weapons. Lacking the speed to escape such enemies
as wolves, and having the habit of living in small groups, these
animals when attacked would form a close circle, the adults
with heads out to the enemy, the young huddled at the inner
side against their mothers' quarters. When closely approached
by wolves, one of the company would suddenly dash out from
the ring in an endeavor to rip one of the assailants with its
horns, then as quickly dart back into its place. This method
of defense, while probably efficient against wolves, its only
natural enemy, was often its undoing when attacked by men
with modern guns, for the entire group might then be killed
at close quarters. Against Eskimo and Indians, who in former
days often depended in part upon muskox for winter food, this
defense was probably more effective before the use of firearms.
During the summer season the almost impenetrable muskegs
of the north form a fairly safe retreat, and the animals' long
silky coat seems sufficient protection against cold and bliz-
zards. But against Eskimo and Indians with modern arms,
they can make little stand, while the thoughtlessness of human
pursuers in wantonly killing entire groups has in places brought
the species to the verge of extinction.

There is evidence that at no very distant time the range of
the muskox extended westward at least to Point Barrow,
Alaska, where according to Frank Russell, in 1898, the oldest
natives "say that their fathers killed muskox, which were then
abundant." The fact that skulls of muskoxen have been found
on the neighboring tundra confirms this statement. J. A.
Allen (1913), in reviewing this evidence, believes that muskoxen
existed on the tundra of northern Alaska till about the middle
of the nineteenth century. A. J. Stone, after extensive inquiry

NORTH AMERICA AND THE WEST INDIES 331

among the Indians and Eskimo west of the Mackenzie, came
to the conclusion that the species had not inhabited that region
for a very long period. Their western limit, at the close of the
nineteenth century, he writes (1900); is "far to the east of
Anderson River and Liverpool Bay." In the region about
Artillery Lake, where the muskox was common shortly before
1901, it was practically exterminated a few years later. Con-
tinual hunting by the Eskimo from the coasts and by the
Indians from the southern borders of the barren grounds had
reduced their numbers still further in the early years of this
century, so that Dr. R. M. Anderson, in 1917, believed that
they were extinct west of the Tree River and much reduced
elsewhere, especially near the accessible regions along the
Arctic coasts,

In 1913 Dr. J. A. Allen well summed up the situation at that
time as follows: "With all its natural fitness to survive, the
muskox is doomed wherever it can be utilized by man as a
commercial asset. The history of its restriction in range and
decline in numbers over the western part of its former range
during the last quarter of a century . . . indicates clearly
its fate wherever it can be reafched by the white man, either
directly or through his Eskimo or Indian allies . . . But
wherever its range is shared by the Eskimos, as many portions
are, the muskox's fate is sealed, as shown by its extermination
over the greater part of the large expanse of land known as
Victoria Island. It will also rapidly decline over the more
accessible parts of Ellesmere Land and Grinnell Land, through
intrusions of ambitious sportsmen. It is doomed throughout
the mainland of northern Canada unless the Canadian Govern-
ment takes the utmost care in restricting the killing .
Much could be done to preserve a considerable remnant of
these unsuspicious animals if the Danish, Canadian, and other
governments would declare muskox peltries contraband and
suppress all traffic in them, while the Canadian and Danish
governments might set aside reservations within which neither
Eskimos and Indians nor white hunters should be allowed to
kill them." Fortunately the Canadian Government in 1927
established a sanctuary known as the Thelon Game Sanctuary,
with an area of 15,000 square miles, northeast of Great Slave
Lake, which so far as then known included the last important
herd of muskoxen remaining on the mainland of Canada. No

332 EXTINCT AND VANISHING MAMMALS

person may hunt or trap within its borders, and entry is pro-
hibited except under permit from the Minister of the Interior.
In 1929 it was estimated that there were 250 muskoxen in this
sanctuary. In 1935 a representative of the department made
an airplane survey and counted 180 muskoxen north of the
junction of the Thelon and Hanbury Rivers (Cameron, 1936),
and Dr. R. M. Anderson (1937) adds that there are a few
scattered bands and individuals still farther north, although
few if any animals are now known to occur very near the
Arctic coast. With this protection it seems likely that the
remnant of the continental muskox is safe enough at present.

HUDSON BAY MUSKOX

OVIBOS MOSCHATUS NIPHOECUS Elliot

Ovibos moschatus niphoecus Elliot, Proc. Biol. Soc. Washington, vol. 18, p. 135, Apr.

18, 1905 (type locality, head of Wager Inlet, Hudson Bay, not "600 miles north

of Hudson Bay" as originally given).
FIG.: Kowarzik, 1910, vol. 5, fig. 8.

In his key, J. A. Allen (1913) characterizes this race as having
"usually no coronal nor facial white areas in adult males, but
traces of them (often well developed) in young males and
females; horns more slender and longer in proportion to their
basal breadth, and generally light-colored; toothrow relatively
longer (max[illary] series, cf, 130 [mm.]); basal length of
skull in old males, 442 mm." The males are usually more
intensely black than the typical race, and the horns are
lighter colored. While males generally lack the white on face
and crown, the females and young are much like the Green-
land race. In body size it is much as in the typical race.

The range of this race, and indeed its characters, do not
seem to be very well defined. The region whence came the
original specimens was at the head of Wager Inlet, in the
northwestern corner of Hudson Bay. According to Captain
Comer, who collected them, the muskox does not now range
south of Chesterfield Inlet, but it is believed that the distribu-
tion of this race extends northward from the latter point to the
Arctic coast of the mainland and inland for an undetermined
distance. They are not found on Melville Peninsula or in
Baffin Land, nor do the natives there have any tradition of
their former occurrence. North of Baker Lake the natives say
that the muskoxen are larger, perhaps representing the typical

NORTH AMERICA AND THE WEST INDIES 333

race. Captain Comer states that formerly the native Eskimo
hunted these animals but seldom on account of the risks in-
volved in making the inland trips, but are encouraged to under-
take their pursuit by traders who are eager for the pelts. Up
till recent years at least, the remoteness of the region where
these muskoxen occurred was itself a factor of safety, for few
hunters would enter it unless well provisioned (J. A. Allen,
1913).

Though explorers agree that no muskoxen are now known to
occur on Melville Peninsula or on Baffin Island, there is some
evidence that they may formerly have been found there but
have been exterminated by the Eskimos. "The Canadian
commission report (1922, p. 13) notes a tradition of a Muskox
once killed on Baffin island . . . ," while on Melville
Peninsula Freuchen states that it has so recently been exter-
minated there that the Eskimos "still knew the names of men
who have hunted it" (Hone, 1934). Apparently there is
nothing to indicate that muskoxen ever occurred naturally on
Southampton, although, as noted by Miss Hone, Freuchen
"says that some teeth were found in the settlement at Kuk on
the west side of York Bay, and* according to the Eskimos there
was a skull in a house ruin on the south shore."

WHITE-FACED MUSKOX
OVIBOS MOSCHATUS WARDI Lydekker

Ovibos moschatus wardi Lydekker, Nature (London), vol. 63, p. 157, Dec. 13, 1900

(Clavering Island, off East Greenland).
SYNONYM: Ovibos moschatus melmllensis Kowarzik, Fauna Arctica, vol. 5, pp. 113-116,

1910.
FIGS.: Allen, J. A. ( 1913, figs. 29, 31. 33, 34-38, 40-44, pis. 11-17 (exterior, live animals,

skull, teeth).

This northernmost race of the muskox has "conspicuous
areas of white between and behind the horns, and face and
sides of the head sometimes suffused with white to a greater or
less extent in old males, in which much of the original white
area is obliterated by the development of the horn bases;
horns long and slender in proportion to their basal breadth,
very light creamy white; tooth row relatively longer than in
moschatus (max[illary] series in males, 140 mm.); basal length
of skull in old males, 442 mm." "In general coloration wardi
is not so dark as either moschatus or niphoecus; the saddle

334 EXTINCT AND VANISHING MAMMALS

area and especially the horns are much lighter in color" (J. A.
Allen, 1913). Tables of cranial measurements for this and
other races of the muskox are given by the latter author.

The range of the white-faced muskox occupies "a narrow
coast belt of Greenland from about latitude 70 on the east
side north as far as land extends, and thence southward along
the west coast to about 81, and within historic times as far
south as Westenholme Sound (latitude 78), where its further
progress south appears to have been checked by impassable
glaciers. It formerly occupied practically the whole Arctic
Archipelago from Grant Land and Ellesmere Land westward
to Prince Patrick Island and south to Lancaster Sound and
Coronation Gulf. Thus it must have nearly met the range of
niphoecus on the mainland west of the Gulf of Boothia, and
the range of moschatus thence westward to Coronation Gulf
and Dolphin and Union Strait . . . They have been ex-
terminated from the greater part of Victoria Island, including
Victoria Land, Wollaston Land, and part of Prince Albert
Land. Hundreds have been killed on Melville Island, and
thousands in northern Ellesmere Land, Grinnell Land, and
Grant Land, mainly by explorers for the support of their dogs
and men. They are found in* winter as well as in summer on
the most northern known land, being in no sense migratory"
(J. A. Allen, 1913).

On Banks Island muskoxen were formerly numerous, for
there was much good pasturage there. They were extermi-
nated, however, by the Eskimos, who killed entire bands,
utilizing little of the meat. The latest record given by Miss
Hone (1934) is of a band killed by Eskimos in the spring of
1911. At the present time the muskox is said to be extinct on
this island. On Victoria Island there were formerly "plenty"
in the little- visited northern part, and there are reports of them
having been killed as recently as 1924 (Hone, 1934) ; the stock
is, however, apparently small. Prince of Wales Island, at
least within a few years, held a population of muskoxen that
Anderson estimated at 1,500. They were present on Somerset
Island in the middle of the last century but are now gone.
Melville Island was the first of the Arctic islands where musk-
oxen were found; Parry in 1819-20 reported them "in con-
siderable numbers." Storkersen in 1916 estimated that there
were about four thousand on this island. They occur also on

NORTH AMERICA AND THE WEST INDIES

335

Byam Martin Island and on Bathurst Island, where Anderson
estimates there are about 1,500. The numbers on Devon
Island seem fewer according to recent reports. On Axel
Heiberg they are plentiful, although, according to Anderson,
they are found only on the east side and may number in all
about 1,000. On Ellesmere Island they were common along
the east coast up till the eighties, but since then they have
seldom been seen there, although numbers occur in the western
areas and may in these less-visited parts be still numerous.

On the short stretch of coast opposite Ellesmere Island in
west Greenland muskoxen formerly occurred, as far south as
Cape Alexander. South of this point the obstructing glaciers
prevent further extension. A few animals were to be found
northward of these points till about 1850, when the last living
muskoxen in the region were reported seen by Eskimo near Cape
George Russell. Macmillan (1918) mentions numerous skulls
to be found from Etah north to Humboldt Glacier, but the ani-
mal no longer occurs there and is believed to have become ex-
tinct about 1860. Various expeditions have reported muskoxen
in some numbers at various localities on the north coast of
Greenland and Miss Hone (1$34) has lately assembled many

White-faced inuskox (Ovibos moschatus wardi)

336 EXTINCT AND VANISHING MAMMALS

notes on their occurrence in bands large or small along the east
coast as far as the region of Scoresby Sound. Anderson, a
decade ago, believed there were approximately 1,500 head in
East Greenland, but others believe this figure too low. Jennov
places the numbers between 6,000 and 10,000. There is some
evidence that they are more abundant than they were when
this region was inhabited by Eskimo. Their distribution may
be sporadic, depending on the depth of winter snows and on
whether the summer has been warm enough to produce pasture.

There has been some controversy as to whether muskoxen
are in need of protection in Greenland. While the inhospitable
nature of their habitat prevents much hunting by visitors, it
is obvious that near the Eskimo settlements, as in northwestern
Greenland, they will continue to dwindle, unless perhaps, as
has been hmted, these Eskimo themselves may be decreasing.
After four years in East Greenland, Pedersen is of the opinion
that protection is more than ever needed, partly on account of
the excessive use made of the animals as food by trappers from
Norway, and partly on account of the trade in young animals
for zoological gardens, for in order to secure the calves, the
herd of adults must first be shot. Although regulations have
been proposed for limiting the number of animals that may be
killed as well as prohibiting hunting for sport or for the capture
of calves for commercial purposes, the latest reports at hand
do not indicate great progress in this direction.

An interesting experiment was undertaken a few years ago
in introducing this muskox into Alaska, in order if possible to
reestablish the animal over areas formerly occupied by it.
In 1930, on the suggestion of the legislature of Alaska, the
United States Congress appropriated $40,000 for obtaining the
necessary stock. Thirty-four animals, both calves and adults,
were captured in East Greenland, transported to Norway,
thence to New York, and by rail to Seattle, whence they were
finally brought by boat to Seward, Alaska (Bell, 1931). This
herd has prospered, so that in 1936 Dr. Bell wrote to Dr.
Harper in response to inquiry that the animals were doing
well, had reached breeding age, and already had produced two
crops of calves, five the first year and six in the second year.
In 1935 two pairs of animals were transferred to Nunivak
Island as an experiment and have thrived there, since food is
plentiful. The herd was later placed on Nunivak National

NORTH AMERICA AND THE WEST INDIES 337

Wildlife Refuge. At last accounts the herd was in excellent
condition and had increased to 90 animals (Dufresne, 1942).

THE AMERICAN BISON

Although evidence is accumulating that at least one large
species of extinct bison was contemporaneous with early man
in North America, only one species was found on the continent
at the time of its discovery by Europeans (see Meserve and
Barbour, 1932). This species is closely related to the Old
World bison, Bison bonasus, but differs in numerous details,
as in the larger chest, smaller pelvus and shorter tail. In
early days the "buffalo," as it is almost universally called, was
found in great numbers over a vast range in this continent,
but with the westward expansion of settlements it became an
object of exploitation on a tremendous scale, so that literally
millions were killed, and it was wiped out in the East and later
over much of its western range. The story of this decimation
has been many times told, but more especially by J. A. Allen
(1876a, 1876b, 1877), Hornaday (1889), and recently by
Garretson (1938). Although .attempts have been made to
distinguish several local races, the characters of these are for
the most part imperfectly known, and the respective ranges
undefined. The northern form, or wood bison, however,
constitutes a fairly well marked race.

PLAINS BISON; "BUFFALO"
BISON BISON BISON (Linnaeus)

[Bos] bison Linnaeus, Systema Naturae, ed. 10, vol. 1, p. 72, 1758 (Mexico).

Bos americanus Gmelin, Linnaeus's Systema Naturae, ed. 13, vol. 1, pt. 1, p. 294,
1788.

Bison bison hanningtoni FIGGINS, Proc. Colorado Mus. Nat. Hist., vol. 12, no. 4, p.
30, pis. 8, 9, Dec. 5, 1933 ("Head of Rock Creek, northeast South Park, Park
County, Colorado"). Doubtfully distinct.

Bison bison septentrionalis Figgins, op. cit., p. 28, pi. 7, Dec. 5, 1933 ("Six miles north-
east of Palmer, Nebraska"). Doubtfully distinct.

FIGS.: Allen, J. A., 1876a, pis. 5, 6, 9, 10, 12, fig. 1-6 (skulls and teeth); Garretson,
1938, pis. facing title page and p. 5.

EASTERN BISON

BISON BISON PENNSYLVANICUS Shoemaker

Bison americanus pennsylvanicus H. W. Shoemaker, A Pennsylvania Bison Hunt
(Middleburg, Pa.), p. 9, 1915 ("Pennsylvania").

338 EXTINCT AND VANISHING MAMMALS

Since the first accounts of the American bison were those
brought back by the Spanish explorers of northeastern Mexico,
this is taken as the type locality of the so-called Plains bison,
the range of which is believed to have covered much of interior
North America from the tableland of Mexico and the grass-
lands of the West to the eastern Alleghenies, perhaps even
reaching the coast in the southeastern States. In a recent
history of the bison in Pennsylvania, Shoemaker has named the
animal ranging "between the east and west slopes of the
Alleghenies, migrating between the Great Lakes and the valleys
of Southern Pennsylvania, Maryland and Virginia, to Georgia, "
as a distinct eastern race, pennsylvanicus, but unfortunately
the description is not based on a comparison of specimens, but
upon local tradition that the bison of this region was larger
than the Plains animal, and "very dark, many of the old bulls
being coal black, with grizzly white hairs around the nose and
eyes"; the hump "was notable by its absence" (which seems
strange), while the legs were "long" without the contrast
between the height of the fore and hind quarters seen in more
western animals. It is difficult to know what value to give
such an account, but the probability that these eastern bison
were somewhat different from those of the Mexican tableland
warrants the tentative recognition of the name. The charac-
ters claimed for the bison of Colorado and of Nebraska, named
B. b. hanningtoni and B. b. septentrionalis, respectively, seem
more likely to be merely individual variations in tooth struc-
ture, so that I have for the present considered these names as
synonyms of B. bison bison.

An adult male Plains bison stands about 5% to 6 feet at the
highest point of the shoulder, but only about 4% feet at the
hip, so that the hind quarters are proportionately small and
the back is sloping. The females are somewhat smaller than
males. J. A. Allen (1876a) gives the following measurements:
Muzzle to insertion of tail, male, about 9 feet (2.75 m.) ; female,
about 6.5 feet (2 m.). The horns are short, thick at the base,
curving outward and upward, then somewhat inward. In the
female they are slenderer than in the male. "In winter the
head, neck, legs, tail, and whole under parts, are blackish-
brown; the upper surface of the body lighter. The color above
becomes gradually lighter towards spring; the new short hair
in autumn is soft dark umber or liver-brown. In very old

NORTH AMERICA AND THE WEST INDIES 339

individuals the long wooly hair over the shoulders bleaches to
a light yellowish-brown . . . The chin and throat are also
covered with long hair, which under the chin forms an immense
beard, eight or ten inches to a foot or more in length. Thick
masses of long hair also arise from the inner and posterior
surfaces of the fore legs, where the hair often attains a length
of six or eight inches. A strip of long hair also extends along
the crest of the back nearly to the tail. The tail is covered
with only short soft hair till near the tip, from which arises a
tuft of coarse long hair twelve to eighteen inches in length"
(J. A. Allen, 1876a). Rarely, black or melanistic individuals
occur, and still more rarely an albino. Many cranial measure-
ments are tabulated in the monograph of J. A. Allen (1876a,
1876b).

Much has been written on the history, distribution, decima-
tion, and reestablishment of this species. In the eastern part
of North America the bison (" pennsylvanicus") occurred as
far east as the western parts of New York State, but in inter-
glacial times it probably extended to New England, as proved
by the discovery of a piece of the maxilla with characteristic
milk premolars found in glacia*! till on Cape Cod (G. M. Allen,
1920). From western New York southward small herds were
found in former times in the mountains of Pennsylvania,
West Virginia, and Tennessee, into the upper parts of North
and South Carolina, following the valleys of the New, Holston,
and French Broad Rivers. Probably the extreme southeastern
limit was in Georgia, where in the southeastern part is a creek
still known as Buffalo Creek. There seems to be no certainty,
however, that it was found in the present limits of the State of
Florida, although its occasional presence there in former
times is not unlikely. It is believed that the eastward extension
of the bison's range was taking place at the time of the dis-
covery, aided in part by the clearing of forests by the Indians
and in part by the attraction of salt licks as in the mountains
of Pennsylvania and West Virginia. To the northward the
animals reached the shores of Lake Erie in their annual mi-
grations and thence ranged westward to Lake Winnipeg and
in increasing numbers over the Great Plains, to the edge of
the Great Basin, and north of that to the extreme northeastern
part of California and southeastern Oregon. The bison of the
latter State has recently been described as a distinct race
(see below).

340 EXTINCT AND VANISHING MAMMALS

In its southward extent, the bison seems to have reached
northern Alabama, central Mississippi, and Louisiana but at-
tained the Gulf coast only in extreme southern Texas and
northeastern Mexico (see map in J. A. Allen, 1876a).

There is but little contemporary record of buffalo in early
days east of the Alleghenies, but what evidence remains indi-
cates that they were locally common in western Pennsylvania,
West Virginia, and the Carolinas, but were constantly perse-
cuted by settlers, explorers, and to some extent by the Indians,
for meat and hides, and even wantonly destroyed. In their
spring migration northward and during the autumnal migra-
tion southward over a considerable extent of country, they wore
deep trails following the easiest gradients and natural passes,
going to better feeding grounds. Many of these paths came to
be the trails and routes used later by the settlers pushing west-
ward from the coast. This emigration received impetus after
the Revolution when "thousands of officers and men who had
served in the war received their pay in land script" (Garretson,
1938). Even before this, the settlers had found that numbers
of bison were incompatible with their own safety, for on oc-
casion the herds would eat and trample down their slender
crops and on one occasion an e&rly settler on Toby and Licking
Creeks (now Oil and Clarion Creeks), Pennsylvania, had his
cabin demolished by a herd of these animals that persisted in
rubbing their backs and sides against its timbers. In his first
two seasons this man and his companions killed 600 or 700
bison, the skins of which brought but 2 shillings apiece. Such
continued slaughter together with the destruction of the natural
food of the bison, through firing of the grass and canebrakes by
the settlers, gradually reduced their numbers so that by the
close of the eighteenth century the "buffalo" in Pennsylvania
"had been reduced to one herd, numbering between three and
four hundred animals which had sought refuge in the wilds of
the Seven Mountains, where, surrounded on all sides by settle-
ments, they survived for a short time by hiding on the most
inaccessible parts of the mountains" (Garretson, 1938). Ac-
cording to Garretson, the last buffalo migrations from the Ohio
country into Pennsylvania had ceased prior to 1783, and the
year 1795 marked the disappearance of the last herds in the
northwestern parts of the State. In the very severe winter of
1799-1800, what was probably the last herd in Pennsylvania

NORTH AMERICA AND THE WEST INDIES 341

was slaughtered when, huddled together in the deep snow in
a great hollow known as the "Sink" in the White Mountains
of Union County, they were rendered nearly helpless. In the
following year a bull, a cow, and a calf were seen in the same
county, and the bull was killed the following year; it was be-
lieved to be the last wild buffalo to be shot in the State. The
cow and calf were hunted but eventually disappeared, and
with that the bison became extinct in Pennsylvania. The
story of the buffalo in West Virginia is very similar, but they
persisted a little longer. According to Garretson (1938), "the
last buffalo killed in Kanawha County, West Virginia, was in
1815, on the waters of the Little Sandy Creek of Elk River,
about twelve miles from Charleston. It is also recorded that
as late as 1 825 a buffalo cow and her calf were killed at Valley
Head, near the source of Tygart's River, and these are be-
lieved to be the last buffalo killed in the East. " Thus by 1825,
with the rapid opening up of the Middle West, and the
slaughter of these animals by the settlers, the buffalo had be-
come practically extinct east of the Mississippi, although a few
stragglers were killed in Wisconsin as late as 1833 (Cory, 1912).

With the transcontinental Purveys followed by the trans-
continental railways in the two or three decades after 1830,
added to the expansion of the fur trade in the West, the slaugh-
ter of the vast herds beyond the Mississippi began in earnest.
"As early as 1840 the American Fur Company sent 67,000
robes to St. Louis and in 1848, 110,000 robes were received,
also 25,000 tongues." The skins of cows only were used for
robes, for those of bulls were too heavy. Hayden, of the U. S.
Geological Survey, who visited "the upper Missouri country
in 1850-1860, estimated the number of buffalo killed every
year to be about 250,000 of which 100,000 were for robes"
(Garretson). The railroad companies, advancing their lines
across the western country, employed hunters to keep their
camps supplied with buffalo meat, and the hunters likewise
shipped back incredible quantities of tongues and hides for
sale in the East.

After the Civil War the army posts on the Plains increased
in number, and hunters on contract supplied the camps with
meat. In the seventies bison were recklessly slaughtered by
the hundreds of thousands, and for every one utilized Hornaday
(1889) believes two were wasted. Only the best hides and the

342 EXTINCT AND VANISHING MAMMALS

choicest parts of the meat were saved. South of the main
transcontinental railroad lines, what came to be known as the
"southern herd" centered in Kansas, Oklahoma, and Texas.
In December 1877 and January 1878 the "last great slaughter"
of this group took place when more than 100,000 hides were
taken by an army of hunters. By the end of the latter year,
this herd had been practically wiped out. A few remained in
the southwestern corner of Kansas until 1879, when the last
one in that State was killed west of Dodge City. In Texas,
scattered herds survived later, but it is believed that a little
group of four killed near Buffalo Springs were the last survivors
of the southern herd. It was said that at this period, when
war was still being carried on against some of the tribes of
Plains Indians, the extermination of the buffalo on which they
chiefly depended for food was a factor also in the destruction of
the Indians. This period of the seventies marked the turn of
the tide for the buffalo, and their numbers rapidly decreased
both in the southern and the northern part of their range.
Garretson (1938) has published some interesting pictures of the
vast quantities of bones bleaching on the sites of slaughter,
and for years after persons made a living by collecting these
bones in cartloads and shipping them east to be made into
fertilizer.

In the late eighties naturalists and others interested in wild
life began to realize that the bison was approaching extermi-
nation. Apart from a few privately owned herds and a herd
closely protected by the Government in Yellowstone National
Park very few remained in a wild condition in the United
States. It was at this time that Dr. Hornaday (1889) made
his stirring protest against the extermination of the species.
By 1900 there were but two herds of bison remaining in a wild
state in North America: the small one in Yellowstone Park
and the wood bison in Athabaska. Notwithstanding supposed
protection of the former, there was for a time considerable
poaching, and it was not until May 1894 that an effective
law for the preservation of the bison was passed by Congress.
In 1902 Congress appropriated $15,000 for the purchase of
buffalo from privately owned herds to build up the small stock
then remaining in Yellowstone Park. Soon after, through the
efforts of the New York Zoological Society, the Wichita herd
was established, following the plan that numerous small herds

NORTH AMERICA AND THE WEST INDIES 343

in different areas would be the wisest way to build up a stock
of these animals. In 1905 the American Bison Society was
founded, starting with a group of 16 public-spirited citizens
who saw the need for steps being taken if the bison was not to
become endangered. The story of its work in establishing at
various suitable places small herds as nuclei, and thus building
up a sufficient population so that the species would no longer
be in danger, has been graphically told by Garretson (1938).
The effectiveness of these efforts is evident from the table
published by the latter, showing that in 1889 the total popu-
lation of bison in existence was placed at 1,091 but by 1933
had been built up to well over 21,000, of which the greater part
(17,043) were in Canada. There are now some 121 small herds
in 41 States totaling 4,404 animals. Thus the species seems
no longer in danger of vanishing from the face of the earth,
but there is good assurance that in suitable places and under
supervision sufficient numbers may be maintained under fence
to make certain of its preservation.

From a practical point of view, bison were formerly the main
source of meat for many of the Plains Indians, as well as for
the early settlers in certain regions. Their later exploitation
by hunters was no doubt largely unnecessary, yet one should
not lay too great a blame upon the shoulders of those who, in
the presence of seemingly unbounded stores of animals, made
use of them for their own gain. For as yet the needs of "con-
servation" had not been made obvious.

Experiments have from time to time been made in crossing
the bison with domestic cattle, but although the resulting
"cattaloes" possess certain desirable qualities, it does not
appear that the hybrids have proved popular with agricul-
turalists.

One of the largest of the protected herds has been that
maintained at Buffalo National Park at Wainwright, Alberta.
Its numbers were lately estimated at about 3,000 head, and in
addition on the reserve were nearly half as many wapiti, as
well as numbers of deer and moose, together with some im-
ported yak from Tibet. Now, 1940, as a late development of
the war in Europe, comes word that this national park must
be utilized for other purposes, which according to rumor,
include the training of aviators, and the animals must be
cleared from the area. According to a quotation in Science

344 EXTINCT AND VANISHING MAMMALS

(vol. 91, pp. 12-13, Jan. 5, 1940), the bison are to be killed and
the meat and hides sold.

OREGON BISON

BISON BISON OREGONTJS Bailey

Bison bison oregonus Bailey, Proc. Biol. Soc. Washington, vol. 45, p. 48, Apr. 2, 1932
("Dry bed of Malheur Lake, Oregon").

In the extreme western extension of its range, the local bison
was slightly different from the typical Plains bison of south-
western Texas, being "slightly larger, with relatively longer
and straighter and less abruptly tapering horn cores, indicating
wider and straighter horns of a somewhat larger animal. The
rostrum or arch formed by the upper premaxillary bones is
slightly longer and relatively narrower than in southern
specimens; interpterygoid fossa wider and larger; auditory in-
flations smaller than in typical Texas skulls; molars larger."
The external characters are unknown.

This race of the bison is now extinct, but a century or more
ago it was found over southern Idaho and extreme northern
Nevada to southeastern Oregon and northeastern California,
areas that it reached from the more eastern plains through
broad flat valleys such as those of the Quinn River and Alvord
Valley. Vernon Bailey (1936) has gathered together what is
known of its history. From the lips of older Indians of the
region, he heard that these animals formerly abounded about
the Cow Creek Lakes country in the early part of the last
century, and Townsend in 1834 found them across southern
Idaho to the Malad River. They seem to have vanished from
Oregon before the arrival of the white man, and disappeared
from Idaho soon after the coming of the early explorers. This
disappearance Bailey attributes largely to the fact that by the
beginning of the last century the Indians of this region were
then well supplied with horses and thus were able to make
more serious inroads into the ranks of the bison, finally ex-
terminating them. Merriam (1926) was able to obtain definite
accounts of its former presence in northeastern California
where old Indians of several tribes said that their fathers had
killed these animals, as in Pine Creek Valley west of Eagle
Lake. The Indians believed that they came in small bands
from still farther north, indicating that here as elsewhere the

NORTH AMERICA AND THE WEST INDIES 345

bison made seasonal migrations in search of better feeding
grounds. Bailey learned from an old chief of the Piute Indians
that in the Malheur Lake region of Oregon bison were found
all over the district at a time probably about the middle of the
last century. They went into the mountains in summer and
came down into the valleys in winter. During dry years,
Bailey writes, the waters of Malheur Lake became very low,
and numerous bison skeletons were laid bare, so that a series
of specimens was collected on which the new race was described.
These remains were evidently of animals that "had bogged
down in search of water at some dry period long ago when the
water had receded; or else, in attempting to cross the lake on
the ice in winter, or to get out to open water, they had broken
through and drowned." There is therefore "no question that
only a few generations back buffalo covered in considerable
numbers many of the large valleys of southeastern Oregon,
and that they disappeared after the introduction of horses
among the Indians and before many firearms were obtained. "
According to an old chief, Yakima Jim, said in 1916 to be 110
years old, the "last of the buffalo were killed during a hard
winter when the snow was so 1 deep that they could not get
grass and a good many tumbled over the high bluffs on the
Owyhee River" (Bailey, 1936). Grinnell (1933) comments
that although the bison evidently was sporadically distributed
in the northeastern part of California in Modoc and Lassen
Counties a century ago, there seems to be no good evidence
that it ever reached the Sierra Nevadas. It may be that
future comparisons will show, if material becomes available,
where this race intergraded with the typical Plains bison.

WOOD BISON; WOODLAND BISON
BISON BISON ATHABASCAE Rhoads

Bison bison athabascae Rhoads, Proc. Acad. Nat. Sci. Philadelphia, for 1897, p. 498,
Jan. 18, 1898 ("Within 50 miles southwest of Fort Resolution, Mackenzie,
Canada").

FIGS.: Garretson, 1938, pi. opposite p. 12 (head and refuge map).

The wood bison is a distinct northern race characterized by
darker color, more dense and silky coat, somewhat larger size,
and more particularly by its longer and slenderer horns and
horn cores, as compared with the Plains bison to the south. The
longer and more incurving horns give the animals a distinctly

346 EXTINCT AND VANISHING MAMMALS

different look from their shorter- and stouter-horned relatives,
which is obvious even in photographs of the living animals.

The range of this race is (or was) north of the United States,
in northwestern Canada, east of the Rocky Mountains to
about the 95th meridian, and between latitudes 55 and 63
N., approximately. According to Dr. R. M. Anderson (1937)
there "is evidence that the wood bison formerly was found
some little distance northwest of Great Slave Lake as far as
Horn Mountains and Liard River, and for an indeterminate
distance up the Peace River valley, and southward, but it is
now restricted to the Wood Buffalo Park area, on both sides of
the 60th parallel." Preble (1908) has assembled a large
amount of data gleaned from the accounts of older travelers
and explorers concerning the bison of this northern region.
It appears that the animals "formerly ranged over immense
areas north to Great Slave Lake and Liard River," where in
1772, Samuel Hearne found them "very plentiful." At the
beginning of the last century Mackenzie recorded "numerous
herds" on the plains near Vermilion Falls and in the Peace
River region, which was probably close to its southern limit.
There is some evidence that by 1828 it was already diminishing
in numbers here, and according to Cowan (1939) the last
record of its presence in the district is furnished by Dawson,
who, in his report of his expedition down the Peace River in
1879 and 1880, mentions the many scattered bones and the
saucershaped wallows of the buffalo, adding that "the Beaver
Indians report having seen in the summer of 1879, six woodland
buffaloes of which they killed one in the vicinity of Pouce
Coupe." Probably they did not long persist in this part of
British Columbia after that time. Their destruction in this
region may not have been wholly due to man, for in 1877
J. A. Allen published a letter giving observations of two young
men who had reached the Yukon through British America in
which they state that in making a portage from Peace River to
Hay River, they saw "thousands of buffalo skulls and old
trails in some instances two or three feet deep, leading east and
west. They wintered on Hay River, near its entrance into
Great Slave Lake, and there found the buffalo still common,
occupying a restricted territory along the southern border of
the lake. This was in 1871. They made inquiry concerning
the large number of skulls seen by them on the portage, and

NORTH AMERICA AND THE WEST INDIES 347

learned that about fifty years before snow fell to the estimated
depth of fourteen feet, and so enveloped the animals that they
perished by thousands." Such wholesale decimation must
have been rather rare, yet nevertheless shows what an unusual
season might so. Another danger to which these animals
sometimes must have succumbed is illustrated by an account
given to Preble (1908) of a herd of about 50 bison that were all
drowned in attempting to cross a small lake too early in the
season, before the ice would bear their weight. This was near
where the Petite Riviere Bouffante enters the Athabaska.

Macfarlane, who resided at Fort Chipewyan during the
years 1870-85, said that the fort hunters seldom failed to kill
a few bison each winter, mainly on the north side of lower
Peace River (Preble, 1908). Writing in 1888, William Ogilvie
(quoted by Preble, 1908) reported that the wood bison was
"nearly a thing of the past ... In the winter of 1887-88,
on the headwaters of Hay River, which flows into Great
Slave Lake, and west of Battle River, a tributary of the Peace,
the Indians saw three bands containing 17, 10, and 4, respec-
tively; they killed 5 ... The same winter three bands
were seen between Salt River and Peace Point, on Peace
River, numbering 50, 25, and about 25, respectively . . .
During the winter of 1886-87, between the north end of Birch
and the south end of Thickwood Mountains, distant about
one day or 30 miles from Fort McMurray, on Athabasca River,
one band of about 13 was seen. Since then 5 of this band have
been killed. Below Red River, a tributary of the Athabasca,
and between Birch Mountains and Athabasca River, and
ranging down to Poplar Point, on the Athabasca, another band
said to contain about 20 was seen. Altogether we have only
about 180 head of wood buffalo in this vast extent of territory. "
In 1891 the same explorer estimated the wood-buffalo popula-
tion as not exceeding 300 in all. He speaks of the Indian
method of killing them by stampeding the animals into a bog
where they soon become mired and are easily slaughtered.

In February, 1894, Caspar Whitney "estimated the total
number then living as about 150" (Preble, 1908). That the
numbers were indeed low is indicated by the fact that Preble
on his journey through their region in 1903 and 1904 was able
to obtain but few reports of their presence. In the winter of
1902-3 he found two small bands, aggregating 24 individuals,

348 EXTINCT AND VANISHING MAMMALS

in the thickly forested region about 125 miles southwest of
Fort Smith. There were apparently no young animals with
this herd, and it was believed that wolves had accounted for
any there may have been. In 1907 Inspector A. M. Jarvis and
Ernest Thompson Seton in much the same region found two
herds of 13 and 20.

Fortunately, in December, 1922, the Canadian Government
set aside as Wood Buffalo Park an area of 10,500 square miles,
which included the entire habitat of the known herds, as a
sanctuary. This was subsequently enlarged to 17,300 square
miles and placed under the charge of a dozen skilled rangers.
By 1929 the number of animals had increased to about 1,500,
or nearly three times the number estimated to inhabit the
region in 1914 (Harper, 1932). This satisfactory increase has
continued, but an undesirable element has been that since 1925
there have been transported to Wood Buffalo Park the increase
of the Wainwright herd of Plains buffalo from central Alberta
in at least four annual shipments totaling 6,673 animals. On
account of the close relationship of the two races, it seems in-
evitable that they must interbreed and that the remnant of the
wood buffalo will eventually be represented by a mongrel
stock. From a zoological point of view this result will be
highly undesirable, but on the other hand may, through the

Wood bison (Bison bison athabascae)

NORTH AMERICA AND THE WEST INDIES 349

increase of the herd, tend to add to the living resources of the
human population in the region. The combined herds have so
increased that in 1934 the total number of buffalo in the park
was estimated at about 8,500 (Anderson, 1937, p. 103).

Outside and to the northwest of the park there is said to be
a small band of 20 to 30 buffalo in a place known locally as the
Buffalo Mountains on the South Nahanni River, and it is not
impossible that these may help to perpetuate the woodland
stock, since its requirements are slightly different from those
of the Plains bison (notes of Dr. Harper from manuscript
report of H. M. Snyder) . Thus, while the wood bison may be
considered as no longer in danger of extermination by man, it
remains to be seen whether its type will disappear through
interbreeding with the imported Plains stock.

THE BIGHORN, OR MOUNTAIN SHEEP

Of all American game animals, probably the bighorn offers
the greatest thrill to the sportsman. A dweller in open moun-
tainous country, keenly alert to the least danger, with excep-
tional eyesight and hearing, it is the most difficult to approach
of all American mammals, while its magnificent horns make a
fine trophy for the successful hunter. The American members
of this genus appear to have come in from Asia at a time when
it was possible to cross from the northeastern extremity of
that continent to Alaska, probably during the Pleistocene
period. Thence they have spread southward through the
western mountain chains to northern Mexico and eastward to
the Dakotas. At the present time zoologists recognize two
American species, the white, or DalPs, sheep in the northwest
and the mountain sheep, or bighorn, from British Columbia
south. Several races of each have been named at various
times, but only recently has a comprehensive study of the
group been made (Cowan, 1940), the results of which have
been followed here.

WHITE SHEEP; BALL'S SHEEP

OVIS DALLI DALLI Nelson

Ovis montana dalli Nelson, Proc. U. S. Nat. Mus., vol. 7, p. 13, June 3, 1884 (mountains
west of Fort Reliance, Alaska, on the divide between Tanana and Yukon Rivers).

FIGS.: Hornaday, 1901, figs, on pp. 83, 86, and pis. facing pp. 86, 92, 94; Nelson, 1916,
p. 450, lower fig. (col.); Sheldon, 1930, pis. facing pp. 78, 79 (photographs).

350 EXTINCT AND VANISHING MAMMALS

This is slightly smaller than the other North American sheep
and with its races is considered to represent a species different
from the more southern 0. canadensis and its races. In general
appearance it is similar, but the horns are slenderer and the
coat is white at all seasons, with sometimes a few scattered
black hairs, especially near the base of the tail, where at times
they may form a distinct black line. As a species distinct from
0. canadensis, Ovis dalli, the white sheep, is smaller, with horns
of greater length but of less basal circumference, "with the
anterior surfaces strongly rugose, the orbital angle prominent,
and often developed into a pronounced ridge overhanging the
orbital surface of the horn" (Cowan, 1940). Height at shoul-
ders, 39 inches; total length, 58.5 inches; tail, 4 inches. Record
length of horn, 47.5 inches. Weight of rams about 200 pounds
(Seton) ; of ewes 50 or more pounds less.

Originally white sheep were found in the mountain ranges
of Alaska "from those bordering the Arctic coast south
through the interior to the cliffs on Kenai Peninsula, but are
now scarce or gone from some mountains. " " Coming within a
few miles of the Arctic coast south of Herschel Island," they
follow the Mackenzie Mountains as far south as Nahanni
River, north of the Liard River. Probably at one time their
range was practically continuous over this area. Hornaday
(1901) mentions the presence of this sheep in the mountains
east of Nome, Alaska; it was not uncommon in the region about
Cook Inlet, and extended as far eastward as about fifty miles
from the Mackenzie delta. Southward it extends along the
Rocky Mountains to about latitude 60 N. At the beginning
of this century it was regarded by A. J. Stone as common
throughout most of its range, but "not nearly so abundant as
formerly. Where hundreds roamed eight years previously, we
saw but sixty -four. " Sheldon (1930) found them still abun-
dant in the Mount McKinley region, and Dr. R. M. Anderson
(1939b) writes that it is "holding its range in most parts of
Yukon Territory and limited numbers are found along the
eastern slopes of the Mackenzie Mountains in Northwest Ter-
ritories from the Arctic coast south nearly to Liard River."

Thus, although on the borders of its distributional area it
has undoubtedly been reduced in numbers, it is still common
on the main mountain ranges of Alaska and Yukon. Various
excellent accounts are in print of its habits and hunting, no-

NORTH AMERICA AND THE WEST INDIES 351

tably by A. J. Stone (1900), and in the Mount McKinley region
by Charles Sheldon. The natural enemies are chiefly the wolf,
less often the lynx (Sheldon), and Lee emphasizes particularly
the depredations of grizzly bears, which ambush them and suc-
ceed often in killing them. During winter, when snowstorms
cover their upland pasturage, they may lie up for days among
sheltering rocks without venturing forth to feed. Their range
is usually above timberline on the open slopes where they can
readily detect the approach of danger from below. Two
young are usually produced at a birth, some time between
May 1 and August 1. "The males and females are hardly
ever found together during the summer months. The males
generally inhabit the roughest and highest peaks, while the
ewes and lambs keep along the high plateaus" (Lee, in Horna-
day, 1901).

Except for food and trophies these sheep are little hunted,
for the coat even in winter is so brittle that it is not in demand.
Also since their area of distribution is mainly outside tracts
used for grazing domestic sheep, the bighorns of this type are
relatively safe from infection with scabies, which has proved
so destructive to their more southern relatives in the United
States. Nevertheless, there is always danger of introducing
new diseases where domestic animals are brought in. Thus
Maj. Allan Brooks (1923) writes that "lumpy -jaw was very
prevalent in the range of Ovis dalli and its subspecies (north
side of Stikine River) up to 1908, but is now apparently
stamped out. "

KENAI SHEEP

OVIS DALLI KENAIENSIS J. A. Allen

Ovis dalli kenaiensis J. A. Allen, Bull. Amer. Mus. Nat. Hist., vol. 16, p. 145, 1902
("Head of Sheep Creek, Kenai Peninsula,' 5 Alaska).

This race, confined to the Kenai Peninsula of Alaska, is
similar externally in its white pelage to typical 0. d. dalli. It
is distinguished, however, on the basis of cranial characters:
basilar length less; molar series significantly shorter (70-72
mm.); basioccipital narrower; angle between basioccipital axis
and palatal axis apparently greater (Cowan, 1940).

The sheep on the Kenai Peninsula constitute a population
now quite isolated from their mainland relatives and charac-
terized by slight though significant differences in certain cranial

352 EXTINCT AND VANISHING MAMMALS

proportions, so that Cowan believes they deserve recognition
as a distinct geographical race. This fact, he comments,
"probably indicates a very long period of residence there and
an effective degree of isolation by the lower ground between
the peninsula and the mountains of the contiguous mainland. "
Very little seems to be recorded as to the present status of the
sheep on the Kenai Peninsula. W. T. Hornaday (1901) repro-
duces an account of hunting them in the Kenai Mountains by
Harry Lee, who mentions their great reduction in numbers at
the hands of prospectors and miners in the region forty years
ago. Previous records for this race from the base of the Alaska
Peninsula are regarded by Cowan (1940) as referring to typical
dalli.

Undoubtedly on account of its limited range it should be
protected if it is to be saved from eventual extermination.

STONE'S SHEEP; BLACK SHEEP
Ovis DALLI STONEI J. A. Allen

Ovis stonei J. A. Allen, Bull. Amer. Mus. Nat. Hist., vol. 9, p. Ill, Apr. 8, 1897 (head-
waters of the Stikine River, British Columbia, altitude 6,500 feet).

SYNONYMS: Ovis liardensis Lydekker, Wild Oxen, Sheep, and Goats, p. 215, fig. 41,
1898 (Liard River, Canada); Ovisfannini Hornaday, Fifth Ann. Rept. New York
Zool. Soc., p. 78, June 1, 1901 (Dawson City, Northwest Territories); Ovis cowani
Rothschild, Proc. Zool. Soc. London, 1907, p. 238 (mountain chain near Mount
Logan, British Columbia); Ovis canadensis niger Millais, The Gun at Home and
Abroad, vol. 4, p. 324, 1915 (Skeena River, mountains at the head of, British
Columbia).

FIGS.: Hornaday, 1901, col. pi. opposite p. 78 (0. 'fannini'}, pis. opposite pp. 92, 98,
100; Nelson, 1916, p. 450, upper fig. (col.).

South of the range of the typical form of the white sheep
occurs a dark-colored race, with intermediate conditions shown
by specimens from the mountains west of the Yukon between
Selkirk and Forty-mile River (the 0. "fannini") while vari-
ations from the mountain ranges in northern British Columbia,
though given distinct names, are now regarded as all referable
to Stone's sheep.

Of the same general type as Ball's sheep, with slender, out-
curving horns of triangular cross section, Stone's sheep de-
velops a partly pigmented coat, varying from one with a gray
saddle to a more gray or dark-brown condition, with a black
dorsal stripe from the back of the head to the tip of the tail.

NORTH AMERICA AND THE WEST INDIES 353

Face and mid ventral area white. Height at shoulders in males
said to be as in dalli, in an adult female 32 inches. Record
length of horn on front curve, 51% inches, with a tip-to-tip
spread of 31 inches (Muskwa River, British Columbia) (Ely
et al., 1939).

The range of Stone's sheep approaches that of the Rocky
Mountain bighorn south of the upper Peace River, British
Columbia. It is still common about the "upper Stikine River
and its tributaries; thence it extends easterly to Laurier Pass
in the Rocky Mountains, north of Peace River, and south
perhaps to Babine Lake. Unfortunately it seems to have be-
come extinct in the southern border of its range" (Nelson,
1916). Cowan (1939), however, mentions the occurrence of a
single young ram on the shore of Charlie Lake during the sum-
mer of 1937, in the Peace River district. It is said to be espe-
cially common east of Dease Lake. The intermediate form, the
so-called Fannin's sheep, or saddleback sheep, occurs in the
same groups with white animals, and there is a condition in
which the white sheep show in some individuals a decidedly
black tail, especially in the Nahanni Mountains. All such,
however, are best regarded as connecting steps in the passage
from the two extremes of coloration, rather than as distinct
races.

Nelson (1916) writes that this sheep occurs in "one of the
most notable big-game fields of the continent. Its home above
timberline is shared with the mountain goat and in the lower
open slopes with the caribou, while within the adjacent forests
wander the moose and two or more species of bear. Owing to
its frequenting remote and sparsely inhabited country it con-
tinues to exist in large numbers; but if its range becomes more
accessible, only the most stringent protection can save this
splendid animal from the extermination already accomplished
on the southern border of its range. "

BIGHORN, OR ROCKY MOUNTAIN SHEEP

OVIS CANADENSIS CANADENSIS Shaw

Ovis canadensis Shaw, Naturalist's Miscellany, vol. 15, text to pi. 610, 1804 (mountains

on Bow River, near Calgary, Alberta, Canada).
SYNONYMS: Ovis cervina Desmarest, Nouv. Diet. Hist. Nat., vol. 24, pp. 5, C,

1804 (Alberta); Ovis montana G. Cuvier, Regne Animal, vol. 1, p. 267, 1817 ( =

Dec., 1816) (Alberta).

354 EXTINCT AND VANISHING MAMMALS

FIGS.: Hornaday, 1901, figs, on pis. opposite pp. 102, 106, 110; Nelson, 1916, col. fig.
on p. 447; Chapman, W. and L., 1937, pis. following p. 86.

The bighorn group has the facial part of the skull an inch
or two longer than in the white sheep and its race, Stone's
sheep; the body is larger and the horns stouter, less spreading.
The pelage is coarser and in winter somewhat shorter, crinkled.
In color the Rocky Mountain bighorn is a uniform gray -brown,
varying much in individuals, even in the same group, and on
the front of the hind legs becoming darker. The end of the
muzzle, a conspicuous patch on the rump, and the backs of the
limbs, white to creamy. "This dark-colored abdomen seems . . .
to render this mountain animal less conspicuous from below"
(Hornaday, 1901). The massive horns of the male usually
make one complete turn, sometimes more, in a rather close
spiral. A record head with about one and one-half turns is
figured by Hornaday (1901, opp. p. 106). In females the horns
are shorter, more nearly erect. There is no groove under
either edge such as is seen in horns of male white sheep.
Height at shoulder, 40 inches (1,018 mm.); length from nose to
base of tail, 58 inches (1,476 mm.); tail, 3 inches (77 mm.).
In the record head, the greatest length of horn was 52.5 inches
on outer curve; the longest listed by Ely et al. (1939) is 49.5
inches. Cowan (1940) states that the short ear is additionally
diagnostic of the typical race.

The range of this sheep is from the mountains of western
Alberta and southeastern British Columbia to about 120 miles
south of Peace River, thence southward in the Rocky Moun-
tains to Colorado, Utah, and northern New Mexico. Over
parts of this range, owing to persecution, and to some extent
to other causes, it has become scarce or even extirpated.
"Their wariness, their strength and agility in climbing, and
the rugged mountains which they inhabit combine to render
them difficult to find and difficult to kill," while the impressive
horns of an adult male form a spectacular trophy, so that
more perhaps than any other game mammal of North America
they have a strong appeal to sportsmen. According to Dr.
Hornaday, the young, usually one or more rarely two, are
born late in spring, between May 15 and June 15. The young
lambs are sometimes the prey of eagles, and the adults are
occasionally killed by wolves in winter. The same author
states that although they have been exhibited in zoological

NORTH AMERICA AND THE WEST INDIES 355

gardens in eastern Unites States, they do not long survive
change in altitude, humidity, and food.

In the Rocky Mountains of western Alberta, Dr. R. M.
Anderson writes (1939b) that this sheep is "still fairly common
. . . and many good specimens are taken annually."
Under present conditions and with adequate legal protection,
the prospect for their survival seems excellent. In Canada
ewes are protected at all times, and the rams during about
two-thirds of the year, while in a national park created by the
Canadian Government on the boundary adjacent to Glacier
National Park in northwestern Montana, there is of course
year-round protection. In the Glacier Park, Montana, Bailey
(1918) reported that these sheep are abundant "on practically
all the high, rugged ranges . . . especially on the rocky
slopes above Two Medicine Lakes and around Chief and
Gable Mountains. In summer they scatter out over the high
and more inaccessible ridges above timberline . . . but
during the winter they come down on the lower slopes and,
especially in spring and early summer, are much in evidence."

Bailey estimated the total sheep population in the park as
about 2,000. Their chief ene*my here is the large mountain
coyote, but outside the Park "the animals are not easily
protected from poachers" who hunt for meat. In the adjacent
State of Idaho, sheep were formerly plentiful and widely dis-
tributed. In 1884 "thousands of sheep were in the Lost River
area; Hornaday (1901) reports that in 1887 trappers encoun-
tered 2,000 to 2,500 head on the Middle Fork of the Salmon
River" (Davis, 1939); Merriam a few years later found them
common in the Lemhi and Pahsimeroi Mountains and in
smaller numbers in the Sawtooth Mountains. A 1939 estimate
of the bighorn population in Idaho was but 2,450, of which
the greater part were in national forests. The same source
(U. S. Dept. Interior, Wildlife Leaflet BS-142) placed the
bighorn population of Wyoming at 5,079, or nearly as great as
the number in Montana and Idaho combined. These animals
were said to be about equally divided between grazing areas
and national forests, and hence largely in the western part of
the State. The estimate given for Utah was 252, probably
mostly in the Ashley National Forest. When about 20 years
ago I accompanied Dr. Theodore Lyman to the Uinta Moun-
tains in search of sheep, we could gather nothing but vague

356 EXTINCT AND VANISHING MAMMALS

reports of a few animals said to have been seen within recent
years. However, Cowan (1940) reports 150 head there at
present. Most of the upper parts of the range are grazed by
domestic sheep in summer. In Colorado, Merritt Cary in 1911
reported that bighorns were to be found in small numbers on
nearly all the high mountain ranges, especially in the northern
parts of the State. Since 1885 they had been protected by law,
but this apparently had not always been well enforced. With
better protection and a developing public sentiment against
killing them, an encouraging increase in numbers took place
in the early years of this century. In 1902, Dall De Weese
estimated that there were probably 200 in the State. By 1907,
"sheep were reported on most of the mountain ranges of south-
ern Colorado, and seemed to be on the increase." The 1939
census credited the State with some 2,285 sheep, of which all
but about 150 were in national forests. To the southward the
typical Rocky Mountain bighorn extended formerly into the
northern parts of New Mexico, but at the present day there
are few if any left there. Bailey gives as its former limits of
distribution the Sangre de Cristo Mountains as far south as
the Truchas Peaks, Pecos Baldy, and Santa Fe Baldy on the
east of the Rio Grande Valley, and probably through the San
Juan Mountains to the Jemez on the west of the valley. Sheep
disappeared from the lower and more accessible San Juan
Mountains some time toward the close of the last century for
in 1904 Bailey could obtain no recent record of them from
local ranchmen, while from the Jemez Mountains still farther
to the south they must have gone soon after 1880. Sheep were
common in the Santa Fe region in 1873, and probably a few
lingered among the mountains at the head of the Pecos till
the early years of the present century. Bailey (1931) believes
that given adequate protection bighorns will in course of time
increase in Colorado and the overflow will repopulate northern
New Mexico, following the mountain chains. The future of
the species seems well assured, for summing the censuses of
1939 for the five Rocky Mountain States gave an estimated
total of nearly 11,000 animals, of which about half are in the
State of Wyoming.

Cary (1911) as well as Warren (1910) mentions the coyote
as an occasional enemy, especially in winter when, in deep
snows, it is sometimes impossible for the sheep to escape,

NORTH AMERICA AND THE WEST INDIES 357

whereas with a light crust the coyotes are able to give pursuit
effectively. Another danger that has developed is the intro-
duction of scab by the grazing of domestic sheep on the upper
ranges in summer, in the States where bighorn occur. Warren
was told of an instance in which 75 bighorns were found dead
from this cause in the West Elk Mountains. More recently
the bighorn in Glacier Park have in some instances shown the
development of a fatal pneumonia, to the extent in one case
that 15 out of a band of a hundred had died. The primary
trouble seems due to a lungworm, which opens the way for
secondary bacterial infection. Most of this seemed to occur
among bighorns that in a certain locality were fed with hay,
where local pasturage had been overgrazed. Other cases of
young lambs dying of pneumonia due to infection by Pasteu-
rella, without primary infection by the lungworms, were also
found. No suggestions for eradicating this danger have so far
been made. The lungworms probably pass one stage in a
snail and then, if the small snails are eaten by the sheep while
grazing, the parasite continues its development in the new
host, and invades the lung tissue (see Marsh, 1938). * The late
George Bird Grinnell (1928) wrote that "many years ago Col.
Edward L. Munson expressed the belief that an epidemic of
anthrax communicated to them from domestic sheep feeding
on the plains below had exterminated the wild sheep of the
Bear Paw Mountains in Montana." They were formerly
common there, but in later years none has been found. Thus
it seems likely that bighorn may be susceptible to various
introduced diseases that are brought in by domestic sheep and
against which they may need in some way to be guarded.

As a preliminary to more intelligent management and pro-
tection of this species, H. B. Mills (1937) has made an intensive
study of the bighorn population of the Yellowstone National
Park in the northwestern corner of Wyoming and the adjacent
edge of Idaho. Here, notwithstanding Vernon Bailey's earlier
estimate that the area might easily support ten times the
present population, the number of bighorns has remained
about 200 for the past 20 years or more and even dropped to
about half that number as from 1927-33. Thus under protec-
tion from both man and natural predators, the animals show
a decrease. A careful census of the bighorns in the park in
1934-35 gave a total of 240. They keep more or less in groups,

358 EXTINCT AND VANISHING MAMMALS

each having its range, migrating in autumn to wintering
areas and in spring to the extensive summer ranges. A proper
winter range thus becomes essential to their well-being, but
it appears that at the time when the report was made (1937),
the winter range had "been so depleted by the large bands of
wapiti that the bighorn were feeding on very short grass all
winter . . . cropping grass so closely that plants were up-
rooted and the roots and adhering soil were swallowed." In
addition to this competition for food, "two diseases, scabies
and lung worm infection, have caused reduction of the bighorn
population in this region. " There is also a certain amount of
infection from nematodes, which may be more or less normal.
Mills believes that the conditions may be alleviated by re-
storing the winter range to a better condition (involving some
reduction in the wapiti herd as well as a possible acquisition of
additional park territory) and maintaining a proper proportion
of large predators that will eliminate unhealthy individuals.
The study points the way to a better understanding of the
requirements of the bighorn and of methods for maintaining a
normal population.

BADLANDS BIGHORN"; AUDUBON'S BIGHORN
Ovis CANADENSIS AUDUBONi Merriam

Ovis canadensis auduboni Merriam, Proc. Biol. Soc. Washington, vol. 14, p. 31, Apr. 5,

1901 ("Upper Missouri," believed to be the Badlands of South Dakota).
FIG.: Audubon and Bachman, 1846-54, vol. 2, pi. 73 (col).

This is the most eastern race of the bighorn, but it is now un-
happily extinct. It seems to have been but very slightly
different from the typical Rocky Mountain bighorn, but the
upper tooth row averaged longer, and the tooth rows were less
nearly parallel. Audubon and Bachman (1846-54) described
the color as "light grayish brown," the rump and under parts
grayish white; weight of an adult male 344 pounds, and of a
female 240 pounds.

A century ago this sheep was abundant in the broken country
or Badlands of the upper Missouri and Little Missouri in
western Dakota. Bailey (1926) has summed up its history in
North Dakota; its former range in that State "included all of
the very rough Badlands country along and west of the Mis-
souri River. " It was found in western South Dakota, western

NORTH AMERICA AND THE WEST INDIES 359

Nebraska, and to an undetermined distance westward probably
into eastern Montana and Wyoming, where it must have inter-
graded with the typical form of the Rocky Mountains. At
the present time it is believed to be extinct over all this region.
Audubon in ascending the Missouri River in 1843 first met with
the bighorn in western North Dakota above the mouth of the
Little Knife River, probably, as Bailey points out, at about
the same place where Maximilian ten years before had found
them. Probably this was very near the northeastern limit of
distribution, for Bailey quotes a resident long acquainted with
that region as stating that there "never were any mountain
sheep near the Missouri at Cannon Ball, but that formerly
the Indians went farther west to hunt them." Dr. George
Bird Grinnell (1928) pointed out that in those days sheep were
far less shy than later, when the white hunter with his rifle
proved a deadly enemy; it was then common to find the
bighorn grazing on open prairie near the high buttes to which
they could escape if threatened with danger. The Platte
River in southern Nebraska marked about the southeastward
extent of the range. Grinnell recalls that in earlier days,
about the middle of the last century, "on that stream certain
isolated buttes and pinnacles were their favorite resorts, but
when the country about these high points for example,
Scott sbluff began to settle up, the sheep were cut off from the
mountains and could neither escape to other refuges nor could
their numbers be added to by others of their kind from the
mountains.

"In the early eighties Theodore Roosevelt . . . hunted
mountain sheep in the Badlands along the Little Missouri,"
but their numbers were apparently not large. In 1888 three
were killed near the present town of Oakdale, in the Killdeer
Mountains, North Dakota. Three others were killed from a
little band of five in the Badlands of the Little Missouri in
1898. The last positive record for the bighorn in North
Dakota is said by Bailey to be an old ram killed about 1905 on
Magpie Creek, in the Killdeer Mountains. There are later
reports, however, for the Badlands of South Dakota. Ap-
parently the sheep held out longest in the Black Hills region,
where in 1895 Dr. Walter Granger was told of the presence of
a small herd in the vicinity of Harney Peak. "In the Bad
Lands," he wrote (in J. A. Allen, 1895), "they are quite com-

360

EXTINCT AND VANISHING MAMMALS

mon. Several were seen by our party, and their tracks could
be seen at any time. They live mostly in the high flat-topped
buttes, where there is good grass." Just when the last ones
disappeared from this State and from Nebraska, where they
were earlier exterminated, is difficult to say, but they must
have held out in the Black Hills region till rather recent dates.

Badlands bighorn (Ovis canadensi? auduboni]

NORTH AMERICA AND THE WEST INDIES 361

CALIFORNIA BIGHORN; LAVA BEDS BIGHORN;
RIMROCK SHEEP

OVIS CANADENSIS CALIFORNIANA Douglas

Ovis calif ornianus Douglas, Zool. Journ., vol. 4, p. 332, Jan., 1829 ("Near Mt. Adams,

Yakima County, Washington").
SYNONYMS: Ovis canadensis samilkameenensis Millais, The Gun at Home and Abroad,

vol. 4, p. 324, 1915 (Similkameen Mountains, British Columbia); Ovis cervina

sierrae Grinnell, Univ. California Publ. Zool., vol. 10, p. 144, 1912 ("East slope of

Mount Baker, Sierra Nevada, Inyo County, California").
FIGS.: Bailey, V., 1936, pi. 17 (horns).

This race was believed to have been characterized by its
slightly darker color than the Rocky Mountain bighorn, with
heavier jaws and teeth, and especially by its horns, which
were slightly more spreading and less closely coiled, as well
shown in Bailey's (1936) figure. Cowan (1940), however, be-
lieves the color hardly different from that of canadensis.

Originally described on the basis of a specimen from Mount
Adams in western Yakima County, Washington, sheep referred
to this race extended northward into the mountains of south-
central British Columbia, and southward through the lava-
beds region of extreme northeastern California and perhaps
the adjoining part of western Nevada to Tulare and Inyo
Counties, California. The race at the present time is believed
to be nearly extinct in California, where probably it occurred
west to "include neighborhood of Mount Shasta, and to Sheep
Rock . . , east side of Scott Valley, and to Siskiyou
Mountains, in Siskiyou County . . . ; also south as far
at least as Observation Peak, near Nevada line in eastern
Lassen County" (J. Grinnell, 1933). In their account of the
vertebrate natural history of the Lassen Peak region in ex-
treme northeastern California, Grinnell, Dixon, and Linsdale
(1930) adduce a few last records of sheep in that corner of the
State. A band of about 40 "that had lived on Observation
Peak were thought all to have perished in the severe winter of
1922," and many skeletons were found there the following
summer. In 1927 a small band thought to consist of four
females and two males was located in the extreme southeastern
corner of the region, close to the Nevada line. In Lassen
County the last sheep of which the authors could secure in-
formation was seen in 1872, but on Lassen Peak itself there
seems to be no evidence that the animal ever was found.

362 EXTINCT AND VANISHING MAMMALS

According to Dr. Joseph Grinnell (1933), it ranged in the
high Sierra Nevada of California, from the vicinity of Sonora
Peak, in Alpine County, southward to southeastern Tulare
County; probably also this was the race formerly present on
the upper parts of the White Mountains in Mono County,
eastern California. At the date of writing this account, Dr.
Grinnell said that it was still to be found from the vicinity
of Mammoth Pass, Mono County, south to the vicinity of
Olancha Peak and the Kaweah Peaks, Tulare County. Its
altitudinal range extended from 5,000 feet (in winter) in Tulare
County up to 13,000 feet on the ridge east of Whitney Pass.
There was some evidence of an autumnal movement to lower
altitudes in fall and winter, on the eastern slopes of the Sierra.

In a consideration of the present status of this sheep in
California, Dr. J. Dixon (1936) believes it is now in a very
precarious position. In various ways this is a result of human
influence, the chief factors being deer hunting, in the course of
which bighorns may at times be unlawfully killed though ac-
corded full legal protection, and grazing by domestic sheep on
the lower winter ranges, which results in the eating of food that
should be reserved for the bighorns. Intensive human intru-
sion in the form of summer camps is an added disturbing
factor. In September, 1935, Dr. Dixon knew of but a single
band still in existence, probably numbering seven animals,
near Mount Baxter. Testimony of deer hunters and others in
the region is nearly unanimous that some bighorns are shot
every year as camp meat and not for trophies. The critical
time is only about a month in the hunting season, so that by
bringing in a "roving ranger" at that period and by obtaining
the active cooperation of the Forest Service, which is with-
drawing domestic sheep from all parts of the bighorn's range
where they might compete with the latter for food, there is
still a chance for its survival. Those who have eaten it agree
that no other meat that the game animals of this continent
yield is equal to that of mountain sheep; hence arises the temp-
tation to hunters to kill an occasional animal for the pot, even
at considerable risk of detection.

In Oregon this was a sheep of relatively low country, oc-
curring over most of the State east of the Cascade Mountains.
Bailey (1936) has gathered much detailed information on their
former presence, from which a few notes are here extracted.

NORTH AMERICA AND THE WEST INDIES 363

"Originally mountain sheep inhabited every canyon, cliff, and
lava butte as well as many of the rough lava beds of Oregon
east of the Cascade Mountains. They were common until
recent years in the Steens and Warner Mountains and are
still found in the Wallowa Mountains and along the canyon
walls of the Imhaha River . . . There are also records of
sheep seen within the memory of many now living over most
of the extensive lava beds and buttes of eastern Oregon."
They were formerly abundant along the Deschutes Canyon;
old settlers who came to the Bridge Creek region on the John
Day River in 1873 found these sheep in bands of as many as
50 or more, but they have long been extinct. In the early
nineties they were numerous "on all the rimrock of the sur-
rounding country from Burns to Bend, on the rough rim of
Dry Basin, on Glass Mountain, Rams Rock, Juniper Moun-
tain, in the Warner and Abert Mountains, around Christmas
lake, and even out on the sagebrush plains where they some-
times joined the herds of domestic sheep and fought the rams.
They seem to have been last seen in the Mount Warner area
about 1912. Captain Louis, an old Indian, told Bailey that in
former days they frequently crossed the sagebrush valleys and
then were hunted on horseback witji bows and arrows. His
people reported them in the Wagontire and Juniper Mountains
as lately as the autumn of 1915. In the Steens Mountains the
last one was killed in 1911; at least a thorough search of these
mountains in 1916 revealed no trace of living sheep. To the
east of these mountains, they had gone at about the same time.
In south-central Oregon the bighorn sheep disappeared shortly
after the settlement of the country by white men. "In 1905
James H. Gaut was told by people living west of lower Klamath
Lake that mountain sheep had been numerous on the lava
ridges near there up to 1885 and that the last was killed in
1890." In Lake County, where formerly sheep were numerous,
they had nearly disappeared by 1897. In the 1939 census of big
game by the Fish and Wildlife Service, Oregon is credited with
50 Rocky Mountain bighorns, but though these may include a
few referred to this race that reach the extreme northeastern
part of the State (see Bailey, 1936, p. 65), the figure probably
also includes the remaining few Lava Beds bighorns mentioned
as still found in the Wallowa Mountains. To the southeast
there are probably still small numbers in western Nevada, but
their precise subspecific status is unknown.

364 EXTINCT AND VANISHING MAMMALS

In the State of Washington this sheep is "now nearly extir-
pated, though a few are reported as of irregular occurrence in
the Mount Chopaka and Mount Bonaparte region. " Formerly
they occurred easterly in the Cascade Mountains from the
Canadian boundary south to the Columbia River (Taylor and
Shaw, 1929). In the 1939 census this State is credited with
only ten bighorns. A few still exist also, according to Dr.
R. M. Anderson (1939b), "in the southern interior of British
Columbia (Okanagan, Similkameen, Lillooet, and Chilcotin
districts), but apparently no sheep ever ranged in the Selkirk
Mountains. "

With this race now near extinction from its former range,
one asks again why it should so suddenly have begun to dis-
appear in the late eighties and nineties, and the answer is no
doubt in part that its habit of frequenting lower levels, where
it came in contact with the domestic sheep then on the increase,
resulted in many areas in its contracting scabies and perhaps
other diseases, for there is much testimony to the effect that
hunting by white settlers alone did not account for the rapid
decline. For example, Bailey (1936) quotes W. F. Schnabel
that in the Mahogany Mountains of Oregon, where mountain
sheep were plentiful, the^y practically disappeared in 1885,
during the winter of that year and the one before. "They did
not starve but were killed by some disease. I found their
carcasses everywhere and grass and feed were plentiful in those
days." Nowadays, with the greater care taken by ranchers
to free their domestic stock from the ravages of the scabies
parasite, the danger of infecting the feeding ranges is probably
lessening, but in many instances this precaution may have
come too late to save the bighorns.

According to Maj. Allan Brooks (1923), the "mountain
sheep of the greater part of the dry interior [of British Colum-
bia] were wiped out forty or fifty years ago by the introduction
of rifles to the Indians and the introduction of domestic sheep
to their range. Scab decimated the sheep of the region east of
the Fraser River about 1870. A virulent disease that affects
the heart and liver is now being introduced in [the Chilcoten
and Similkameen River districts] ... by domestic sheep
that are brought over to graze from the state of Washington
by sheep-herders. The government veterinary at Osoyoos
is unable to determine this disease Ticks were

NORTH AMERICA AND THE WEST INDIES 365

very bad in 1897 and 1898. The ears of rams killed were
packed to the drum with larval ticks, pale blue with sulphur-
yellow legs. None were found in the ears of rams killed
in 1902 and 1905." Brooks believes that coyotes and golden
eagles annually kill "at least seventy -five per cent of all
the lambs on two of the ranges on which" he has an op-
portunity to observe, which "prohibits all possibility of any
increase." As corrective measures he recommends the total
prohibition of grazing permits for domestic sheep on any range
inhabited by mountain sheep and the appointment of wardens
whose duty it should be to enforce the game laws and to reduce
the predatory eagles and carnivores.

According to Cowan (1940) the largest remnant of the Cali-
fornia bighorn is now to be found in British Columbia; a few
are still living in the Ashnola and Similkameen district where
they were formerly abundant. Small bands are present in the
mountains near Vasseaux Lake and the northern end of
Okanagan Lake.

MEXICAN BIGHORN; DESERT BIGHORN

Ovis CANADENSIS MEXiCANA Merriam

Ovis mexicanus Merriam, Proc. Biol. Soc. Washington, vol. 14, p. 30, Apr. 5, 1901
("Lake Santa Maria, Chihuahua, Mexico").

SYNONYMS: Ovis canadensis texianus V. Bailey, Proc. Biol. Soc. Washington, vol. 25,
p. 109, June 29, 1912 ("Guadalupe Mountains, Texas"); Ovis canadensis gaillardi
Mearns, Mamm. Mexican Boundary, U. S. Nat. Mus. Bull. 56, p. 240, 1907 ("Gila
Mountains, between Tinajas Altas and the Mexican Boundary Line in Yuma
County, Arizona"); Ovis sheldoni Merriam, Proc. Biol. Soc. Washington, vol. 29,
p. 130, Sept. 6, 1916 ("El Rosario, northern Sonora, Mexico").

FIGS. : Elliot, 1904, pt. 1, pis. 24, 25 (skull of type) ; Mearns, 1907, figs. 35-38 (feet,
skull, horns).

This is a pale desert race of the bighorn, with long ears, a
short broad skull, and relatively short, moderately spreading
horns. Height of ear from notch, 106 (95-120) mm. (Cowan,
1940). The latter author describes the color in September
topotypes as paler than in any of the other races in comparable
pelage with the possible exception of cremnobates; above, pale
vinaceous-fawn and vinaceous-buff, tail wood brown, lower
legs between avellaneous and wood brown; much white on
abdomen, extending forward narrowly on to chest, and as a
narrow white line down fore and hind legs. Horns pale, taper-
ing more rapidly than in nelsoni.

366 EXTINCT AND VANISHING MAMMALS

In his recent review, Cowan (1940) regards the Texas and
the Arizona bighorn as insufficiently distinguished, while 0.
sheldoni of Sonora is based on a 'runt' or dwarfed individual,
as pointed out by Dr. H. H. T. Jackson (in Ely and others,
1939).

The range is from extreme southwestern Texas, southern
New Mexico, and Arizona south across northwestern Chi-
huahua and in Sonora to Seriland opposite Tiburon Island.
Northward it intergrades with neighboring races.

This race of mountain sheep was formerly found in the most
arid desert ranges of western Texas in the San Andreas and
Guadalupe Mountains and on the western slopes of the latter
range in extreme eastern New Mexico as well as in eastern
Chihuahua. Its numbers and distribution are now greatly
reduced. Bailey (1905, 1931) has given a good summary of
notes he obtained as to the status. In 1912-14 the number of
these sheep in eastern New Mexico was estimated at 200 in
the Guadalupe Mountains, but a more thorough census in
1916 resulted in a count of about 100 or less. "All the reports
from the Sacramento, Capitan, and Jicarilla Mountains assert
that no sheep have been known in these ranges in modern
times." To the northwest oJ the Guadalupe Range, in the
San Andreas Mountains of New Mexico, there were a very few
left in 1902, according to Gaut who made a special trip in
search of them and as late as 1914 located about 30. At the
same time he was "informed that sheep formerly were found
along the crest of the Organ Mountains, but that none had
been seen there in recent years." On account of the accessi-
bility of their range, they were much hunted over this region,
and Bailey in 1931 concluded that within the limits of New
Mexico this race is now to be found only on the Guadalupe and
San Andreas Ranges. He believes that "with adequate pro-
tection for a term of years the mountain sheep could doubtless
be brought back to its original range and abundance. The
difficulty of enforcing game laws, however, in these uninhabited
mountains is almost insurmountable, and wholly so without
the full cooperation of the resident population. The fact that
these sheep occupy land that will always remain practically
worthless for stock raising or other agricultural purposes makes
it doubly important that their numbers should be increased. "
Since now the "whole summit and eastern slope of the Guada-

NORTH AMERICA AND THE WEST INDIES 367

lupe Mountains from Guadalupe Peak in Texas north to Dog
Canyon in New Mexico is "a permanent game refuge that
could easily support a thousand bighorns, the desirability of
reestablishing the local race is apparent. "Under wise control
and a definite plan for use of the surplus game, either for
hunting or stocking other ranges, such an area could be made
not only self-supporting but a valuable piece of property"
(Bailey, 1931).

In western Texas, in addition to the Guadalupe Mountains,
a few sheep formerly were found in the Eagle and Corozones
Mountains and on the northwest side of the Chisos Mountains.
"They come into the Grand Canyon of the Rio Grande
mainly from the Mexican side . . . The sheep are by no
means confined to isolated mountain ranges. In several
valleys I saw tracks," writes Bailey (1905) "where they had
crossed from one range to another through open Lower Sonoran
country. In this way they easily wander from range to range
over a wide expanse of country in western Texas, and might be
considered to have an almost or quite continuous distribution
between the Guadalupe Mountains and the desert ranges of
Chihuahua." Here they have more or less held their own for
many years, and in the wildlife census of 1939 were estimated
to number 280 individuals. .

No data are at hand as to the status of this race in eastern
Chihuahua, Mexico, but since the time of Bailey's investiga-
tions a study of the bighorn sheep in Texas was carried out in
1938 by W. B. Davis and W. P. Taylor (1939), which affords
an excellent summary of present numbers and living condi-
tions. This sheep is now confined in Texas to the extreme
western wing of the State west of the Pecos River, an arid
region having a scanty rainfall averaging between 9 and 17
inches annually, depending somewhat on elevation. About 1
percent of this area is under cultivation, and the rest is used
for grazing. "The present range of the Texas bighorn in Texas
is considerably more restricted than it was when Vernon Bailey
worked in trans-Pecos Texas at the turn of the century."
None occur now in the Chisos and Corozones Mountains or in
the Grand Canyon of the Rio Grande, and they seem to have
been absent in these areas for the past ten or fifteen years.
"The heaviest concentration is north of the Texas and Pacific
Railway in the Beach, Baylor, Carrizo, and Diablo mountains,

368 EXTINCT AND VANISHING MAMMALS

where the bighorns doubtless cross freely from one mountain
range to another." The authors quoted state that "probably
not more than 25 mountain sheep occur in all the Guadalupes,
none of them on the top in the heavily wooded sections."
The numbers here seem to be decreasing but were probably
never very high. In other ranges single rams or rarely small
bands appear sporadically from time to time, but in other of
the 16 mountain masses for which data were gathered sheep
have been absent for years. Perhaps not more than 30 occur
south of the Texas and Pacific Railway. Summarizing the
figures given, there are approximately 410 bighorns in western
Texas, of which some 300 are found in the four mountain
ranges Baylor, Beach, Carrizo, and Sierra Diablo, while the
remainder are chiefly in the Apache (40), Guadalupe (25),
Delaware (17), Eagle (8), and Glass (13) Mountains. The
requirements of these sheep include not only craggy regions
but also a particular type of vegetation, of low, shrubby,
xerophytic sorts, low enough to afford an unobstructed view.
In the Baylor Mountains the evidence seems to show that
bighorns move into them seasonally and leave "when condi-
tions are adverse." Young are born in March and April,
usually a single one at a time, and they follow their mothers
throughout at least the first summer and autumn. Davis
found that in the Yellowstone National Park grass constituted
nearly 60 percent of the food for Rocky Mountain bighorns,
but for the Texas bighorns investigation shows that grass con-
tributes about 3 percent only, and that the animals prefer
mountain-mahogany, Mexican tea, yellow trumpetflower,
mock-orange, and wild onion. Davis and Taylor consider
carefully the reasons for the decrease in numbers of this big-
horn and conclude that it is not due to overhunting, for the
season has been closed since 1903; nor is it a result of depre-
dations by mountain lions, for these predators have been re-
duced to very small numbers. Golden eagles may take toll of
a few lambs, but these birds are not a very significant factor,
and besides predators have always lived in the region. They
conclude that "the incursion of domestic sheep possibly has
been the most serious factor affecting bighorn numbers" and
suggest that this is in part because of the introduction of
diseases to which domestic sheep are liable, in part to direct
competition for food and water, and in part perhaps to an

NORTH AMERICA AND THE WEST INDIES 369

intangible factor of psychic incompatibility, for bighorns are
commonest where sheep are fewest. On account of the rather
unfavorable conditions of the area, they doubt if it will ever be
possible to increase the bighorn population enough to warrant
an open season in western Texas.

There is little recent information at hand as to the present
status of the bighorn in northern Chihuahua or the adjacent
parts of New Mexico, beyond the fact of its evident deple-
tion within the past 50 years. In extreme southwestern New
Mexico Mearns in 1892 found them in some abundance in
the Dog, Big Hatchet, and San Luis Mountains. In 1900
sheep were brought in and sold for meat at Deming from the
mountains near the international boundary; but by 1908 there
were only a few remaining in the Big Hatchet Mountains,
"and probably none in the other ranges of southwestern New
Mexico. In the country north of Deming there seem to be no
recent records," nor could Goldman and Bailey at that time
procure any records for the Burro and Carlisle Mountains or
the Mogollon Mountains. In 1905 old horns were reported to
Hollister as found from time to time in the little ranges from
the Magdalenas to the Zunis, But no one could tell when the
animals had last lived there.

In northern Chihuahua and the adjacent sierras of Sonora
these sheep were found up to a few years ago on all the ranges,
but now, owing to relentless hunting, and in spite of the fact
that these sheep are not to be killed legally, they are in danger
of becoming extinct. A reservation for this animal is contem-
plated in the Altar district of Sonora (Zinser, 1936, p. 9).

Bighorn sheep, presumed to be of this race, formerly in-
habited the Chiricahua Mountains in the extreme southeastern
corner of Arizona but are now extinct there. In former years,
according to information supplied to Cahalane (1939a, p. 439),
a life-long resident of the mountains recalled that they were
"fairly numerous" in all the lava hills of the vicinity but were
gradually shot out, though he believed that the last remnant
(probably in the Rincon Mountains) succumbed to the drought
of 1903-5.

During the early nineties mountain sheep were common along
the international boundary and the adjacent parts of Arizona.
Mearns (1907) writes that he saw many horns in the Papago
Indian settlement, and the Indians reported them common in

370 EXTINCT AND VANISHING MAMMALS

the higher ranges visible from Nariz, where they killed many,
in consequence of which the animals were much scarcer than
they had formerly been. The surveyors of the party saw sheep
"in the rugged Tule Mountains in 1892. When my party was
there, in February, 1894, no sheep were seen, but many tracks
and heaps of horns were noted, as also in the neighboring
Granite Mountains. During our stay at Tinajas Altas, at the
foot of the Gila Mountains, from February 14 to 23, 1894,
sheep were seen on four occasions, in flocks of 6, 3, 3, and per-
haps as many as 20. They were feeding largely upon a Cy-
lindropuntia cactus, in valleys at the base of the mountain,
but tracks and beds were seen at all altitudes."

Dr. W. P. Taylor (1936, p. 653) speaks of this region as a
desert of rare beauty, and one of the most arid in the United
States. "There is but one small cow outfit in the entire
region, at least on the American side. A few prospectors
work the territory. Aside from a handful of Indians and an
occasional smuggler or United States custom agent, there are
no other inhabitants of the area, and no permanent resi-
dents at all. Both antelope and bighorns are at present subject
to poaching on both sides of the international boundary
through lack of adequate warden service. "

In 1939, by Executive order, two game ranges were estab-
lished in Arizona, with a view especially to protecting the
sheep, mule deer, antelope, and peccaries of this region. One
in central Yuma County, the Kofa Range, comprises over
660,000 acres and is at one point within 15 miles of the border
of southeastern California. The other, the Cabeza Prieta
Range, lies 35 miles to the south on the international border,
in Yuma and Pima Counties, and includes 866,880 acres.
Negotiations are still pending with the Mexican Government
to set aside a corresponding adjacent area on the Mexican side
of the border. It is believed that these game ranges will prove
of great value in preserving the larger wild mammals of the
State. "A few years ago this section abounded in game, but
indiscriminate hunting and poaching came dangerously near to
eliminating some of the most valuable species. "

As to the present status of this sheep in Arizona, Cowan
(1940) quotes A. A. Nichol, who has made it the subject of
special investigation lately. He gives an approximate census
for 1937 as follows: Granite Mountains, 10; Aquila Moun-

NORTH AMERICA AND THE WEST INDIES 371

tains, 4; Gila Mountains, not including the Tinajas Altas, 15;
Kofas, 25; Trigos, 10-15; Chocolate Mountains, 6 or 7; Buck-
skin Mountains, 10-12; west end of Harcuvar Mountains, 7 or
8; Black Mountains, 60-75; Superstition Mountains, 47; while
C. T. Vorhies adds that there were known to be at least 6 in
the Tucson Mountains and about 71 in the Santa Catalina
Mountains altogether approximately 275 animals.

NELSON'S BIGHORN; DESERT BIGHORN

Ovis CANADENSIS NELSONI Merriam

Ovis nelsoni Merriam, Proc. Biol. Soc. Washington, vol. 11, p. 218, July 15, 1897
("Grapevine Mountains, on boundary between California and Nevada").

This is a "comparatively small Bighorn, with rather slender
horns somewhat triangular in cross-section near base; color
very pale; white rump patch divided lengthwise to base of tail
with dark line" (H. H. T. Jackson, in Ely and others, 1939).
Molar teeth heavy, horns slender. Total length, 1,280 mm.;
tail, 100; height at shoulder, 830. In its pale salmon-gray
pelage, this subspecies is strikingly different from the Rocky
Mountain bighorn ; and its ears' are noticeably longer.

This pale, small race is found in the desert mountain ranges
of southeastern California east of the Sierras and in the ad-
jacent parts of southern Nevada. Although the limits of dis-
tribution have not been carefully worked out, Dr. J. Grinnell
(1933) regarded it as the form now or lately in the mountain
ranges of Inyo and Mohave Desert regions, northeast to Owens
Valley, Mono County, south to the Chocolate Mountains in
Imperial County, and west to San Bernardino, San Gabriel,
and (formerly) the Tejon Mountains and the southern end of
San Joaquin Valley, California.

Recent observations on the status of this desert race are few.
According to Burt (1934, p. 424) they are still common in the
Sheep Mountains of southern Nevada, and he was informed of
a herd of 24 at Corn Creek which watered at a cattle tank.
They were formerly common in the Charleston Mountains,
but few if any were left by 1934, although Burt was told of a
small band seen shortly before, along the southwestern border
of the mountains. In California they are somewhat more re-
stricted than formerly but still remain in some numbers in the
desert regions of the southeastern parts of the State. The

372 EXTINCT AND VANISHING MAMMALS

1939 estimate of their numbers by the Biological Survey
credited California with 1,770 desert bighorns, and Nevada
with 1,485. If these figures are approximately correct, the race
is no longer in immediate danger. In both States they are
protected by law throughout the year. Cowan (1940) has
given a good summary of its recent history with a list of speci-
mens examined, including one from as far west as Caliente
Peak, San Luis Obispo, Calif.

LOWER CALIFORNIA BIGHORN

OVIS CANADENSIS CREMNOBATES Elliot

Ovis cervina cremnobates Elliot, Field Columbian Mus. Publ. 87, zool. ser., vol. 3, p.

239, Dec., 1903 (Matomi, San Pedro Martir Mountains, Lower California,

Mexico).
FIGS.: Elliot, 1904, pt. 1, figs. 26, 27 (drawings of adult ram, immature ram, and ewe).

This race of bighorn is described as resembling 0. c. nelsoni
"but of a much lighter color, the head of a three-year-old ram
being nearly white, with a very small caudal patch not divided
from color of upper parts by any perceptible line; fore part of
legs almost black . . . ; head very broad between orbits,
from 20 to 25 mm. broader in old rams than the head of 0. c.
nelsoni; horns of adult rams very large. " Length along outer
curve of horns in an old ram, 850 mm. (Elliot, 1904) . "Perhaps
the palest" of the races.

The isolated desert mountains of the northern half of Lower
California, Mexico, northward to extreme southern California,
on the eastern and northeastern faces of the mountains along
the west side of the Colorado Desert, and to the lower northern
slopes of the San Jacinto Mountains in San Gorgonio Pass,
form the range of this palest of the desert bighorns. In south-
ern California it is not known ever to have invaded the Pacific
slopes but reached the eastern edges of San Diego and River-
side Counties.

This was the first of the races of the bighorn known to Euro-
peans, for it was briefly described as long ago as 1702, in a
memoir by the Jesuit priest Francis Maria Piccolo, on the
discovery of a passage by land to Lower California. The
animal was called the "Taye" and was figured in 1758 by
Venegas, as detailed by J. A. Allen (1912) in whose historical
paper the figure is reproduced. He writes: "Sheep still exist
where the first Spanish missionaries found them in 1697. Dr.

NORTH AMERICA AND THE WEST INDIES 373

Charles H. Townsend on his recent 'Albatross' expedition to
Lower California . . . obtained some imperfect skulls of
mountain sheep from the natives at Conception Bay, and saw
a living specimen in the low mountains at the head of that bay.
He also informs me that mountain sheep are said still to in-
habit the low mountains near the Gulf coast as far south as
Saltillo del Rey, or to within about one hundred miles of La
Paz, and that they range thence northward in all the high hills
and mountains, especially on the Gulf side, nearly to the
United States boundary." Mearns (1907), during the survey
of the international boundary, was informed by Indians that
sheep were then (1894) abundant in most of the rocky ranges
of northern Lower California. In more recent years, however,
the needs of miners and prospectors have jeopardized the con-
tinued existence of bighorns in the region. Dr. E. W. Nelson
(1921) wrote that they were then "still widely distributed on
the main mountains of the eastern half of the peninsula,"
but "the difficulties of transportation and the scarcity of live
stock through most parts of the peninsula render the securing
of food supplies so difficult that great numbers of game ani-
mals, especially mountain sheep^ have been killed and the meat
dried, or jerked, to supply mining camps and other communi-
ties. This deplorable and wasteful slaughter still continues
and unless checked will ultimately result in the extermination
of the sheep . . . Mountain sheep are peculiarly en-
dangered by this slaughter owing to their habit of going in
bands and drinking at certain watering places, where hunters
lie in wait behind blinds built of loose stones and kill indi-
viduals of all ages and sexes, often in a few minutes destroying
an entire band. I have been informed of one party having
killed more than 100 sheep in this manner to make dried meat
during a single season."

Although in 1917 the Mexican Government prohibited the
hunting or exploitation of mountain sheep in Lower California,
"no effective enforcement of this regulation appears to have
been undertaken" (Nelson, 1921, p. 132).

Cowan (1940) regards specimens from slightly north of
Calmalli, Lower California, as intermediate between this and
the race weemsi of the lower part of the peninsula, while others
from the Chuckawalla Mountains, California, are intergrades
between cremnobates and nelsoni.

374 EXTINCT AND VANISHING MAMMALS

WEEMS'S BIGHORN

Ovis CANADENSIS WEEMSi Goldman

Ovis canadensis weemsi Goldman, Proc. Biol. Soc. Washington, vol. 50, p. 30, Apr. 2,
1937 (Cajon de Tecomaja, Sierra de la Giganta, about 30 miles south of Cerro
de la Giganta, southern Lower California, Mexico).

This recently named race occupies the mountainous country
at the tip of the peninsula of Lower California and is remarka-
bly dark for a desert-frequenting animal. In its "usually
darker" color it seems chiefly to differ from the race cremno-
bates of the upper parts of the peninsula, "varying from very
dark brown more or less mixed with black, " rump patch white,
almost divided by the tail stripe. Horns in the adult female
are said to be remarkably long and gradually tapering; in males
the lateral surface is flatter and the external ridge more promi-
nent than in other races.

Very little information is at hand as to the status of this
sheep. Major Goldman describes the country in which it
lives as supporting a richer vegetation than that to the north
where the race cremnobates is found, indicating a greater rain-
fall. The Sierra de la Giganta is the highest range of moun-
tains in the southern part of the Cape, about 60 or 70 miles
long and isolated by a low gap on the north and the south,
close to the Gulf of Mexico. Its eastern face is much steeper
and more broken than the western. The flora includes a num-
ber of subtropical species not found to the north. Here this
southernmost race of mountain sheep still occurs in small
numbers. Goldman mentions four specimens secured by his
party at the type locality in 1936. There is a mounted head
in the Museum of Comparative Zoology obtained in 1925 near
La Paz, and it is of the dark-brown color characteristic of the
race. Although doubtless the puma occasionally attacks this
sheep, as Goldman intimates, yet this animal is probably not
now common enough in Lower California to be nearly so im-
portant an enemy as man. The Mexican Government nomi-
nally gives the sheep of this area full legal protection, but, as
earlier mentioned by Dr. Nelson, the law is difficult to enforce.
Nevertheless it is encouraging to learn that this small remnant
of mountain sheep still holds out in the isolated ranges of the
tip of Lower California. Cowan (1940) regards specimens
from Sierra San Borjas, 20 miles north of Calmalli, as inter-
mediate between this race and 0. c. cremnobates.

SOUTH AMERICA

Order EDENTATA: Edentates

THE NEW WORLD edentates are discussed elsewhere in
the present volume, p. 34. The only edentate considered
in this section is the Patagonian giant ground sloth, which be-
came extinct after man reached southern South America.

Family MEGATHERIIDAE : Giant Ground Sloths

PATAGONIAN GIANT GROUND SLOTH

GRYPOTHERIUM LISTAI (Ameghino)

Neomylodon listai Ameghino, Premiere Notice sur le Neomylodon listai, un Representant

Vivant des Anciens Edentes Gravigrades Fossiles de 1'Argentina, La Plata, Aug.

1898; transl. in Natural Science, vol. 13, pp. 324-326, Nov., 1898 (Consuelo Cove,

Last Hope Inlet, Patagonia).
SYNONYM: Grypotherium domesticum Hauthal, Roth, and Lehmann-Nitsche, Revista

Mus. La Plata, vol. 9, pp. 409-474, 1899.
FIGS.: Hauthal, Roth, and Lehmann-Nitsche, 1899, pis. 1-5; Woodward, 1900, pis.

5-9 (piece of skin).

Much interest was aroused by the announcement in 1898 of
the discovery of a large piece of skin of a ground sloth in south-
ern Patagonia. In a preliminary notice, F. Ameghino briefly
described this and, believing it to represent a new genus re-
lated to the extinct giant Mylodon, named it Neomylodon listai,
after the ill-fated explorer Lista, who described what he be-
lieved to have been a living example in the region of the eastern
base of the Andes. Although it is probable that he was mis-
taken in identifying the animal he briefly saw, nevertheless the
finding soon afterward of the portion of skin with the hair in-
tact, and with the small bony nodules imbedded in it, as
characteristic of the Pleistocene Mylodons, raised at the time
high hope that a living example might yet be found. With
this object in view, H. Hesketh Prichard (1902a) undertook a
journey to the region, an account of which he gives in his book
"Through the Heart of Patagonia"; no evidence did he find,
however, that this mysterious animal was still extant.

375

376 EXTINCT AND VANISHING MAMMALS

The piece of skin on which the accounts of the animal are
based was so fresh that the hair and serum on it, though
dried, were still intact. The hair was 4 or 5 cm. long, reddish
tinged with gray, while in a large portion of the fragment were
numerous small oblong or rounded dermal ossicles. The skin
was discovered in a large cavern, where at a shortly subsequent
date further investigations were made by Hauthal, geologist
of the La Plata Museum. Here he found not only "another
piece of skin, but also various broken bones of more than one
individual of a large species of ground-sloth in a remarkably
fresh state of preservation. Moreover, he discovered teeth of
an extinct horse and portions of limb bones of a large feline
carnivore, in association with these remains; he likewise met
with traces of fire, which clearly occurred in the same deposits
as the so-called Neomylodon. All these remains were found
beneath the dry earth on the floor of an enormous chamber,
which seemed to have been artificially enclosed by rude walls.
In one spot they were scattered through a thick deposit of
excrement of some gigantic herbivore, evidently the ground-
sloth itself; in another spot they were associated with an ex-
tensive accumulation of cut hay. Dr. Hauthal and his col-
leagues, indeed, concluded that the cavern was an old corral in
which the ground-sloths had been kept and fed by man"
(Woodward, 1900).

Remains of at least three individuals were found in this cave.
So fresh were these that bits of dried tissue still adhered to
some of the bones, and one of the skulls showed evidence of
having been cut away at the occiput by an instrument. In the
lower jaw there were four cheek teeth without any evidence of
a caniniform tooth in front of them. The upper jaw likewise
had four cheek teeth, thus fixing the identity as Grypotherium
instead of Mylodon, in which there are five, as also in
other large ground sloths. Of particular interest are the
masses of excrement, which have been determined to consist of
grasses largely. A few pieces of stems appear sharply cut and
suggest that, like the quantity of cut hay found in association
with the remains, these had been fed to the animals by their
human captors. Complete measurements are not available,
but these ground sloths were doubtless as large as a cow,
though differently and heavily proportioned.

SOUTH AMERICA 377

Order RODENTIA: Gnawing Mammals

The rodents, as an order, are discussed hereinbefore (p. 41).
Representatives of four family groups in South America are
considered vanishing forms:

(1) Cricetidae. The rats and mice of the New World are
separated from most of the Old World types by important
dental characters, although frequently similar in appearance.
Six species of two genera of rice rats, peculiar to the Galapagos
Islands, are in danger of extinction.

(2) Myocastoridae, hystricoid water rats, or coypus. Three
races are in need of protection because of persecution for their
fur.

(3) Dinomyidae, hystricoid giant rats. A single species is
considered here.

(4) Chinchillidae. Three races of the rare chinchilla are in
need of protection. J. E. H.

Family CRICETIDAE: Hamsterlike Rodents
THE GALAPAGOS RICE RATS

The Galapagos group of islands, lying on the Equator nearly
600 miles due west of the western coast of Ecuador, is remark-
able for the various peculiar endemic birds found there, in-
cluding the flightless cormorant, the dwarf penguin, and such
land birds as Geospiza, while among reptiles there are sundry
species of giant land tortoises, large iguanas of a genus peculiar
to the islands, and others. The land mammals, however, until
a few years ago were believed to comprise but two species
related to the continental rice rats of the genus Oryzomys.
Later explorations have now raised the number of indigenous
rodents to six, representing two related genera, Oryzomys and
Nesoryzomys. Commenting on these discoveries, Osgood
(1929) writes that the native rodents now "take on consider-
ably more importance than formerly and will doubtless need
serious consideration in speculation regarding the derivation
of the insular fauna. Until thorough study of mainland forms
is made, however, no satisfactory conclusions are to be ex-
pected. While the oryzomyine rodents are widely distributed
and greatly varied in South America, they are also highly
developed in Central America, and no competent study of the

378 EXTINCT AND VANISHING MAMMALS

whole group has yet been undertaken. When such a study is
made, it may be possible to determine something more than
the general affinity of the insular and continental forms. At
present, it can not be affirmed even that the nominal genus
Nesoryzomys is confined to the Galapagos, for at least in some
of its characters it is closely paralleled by certain forms of the
mainland. However, the number and diversity of the island
rodents may perhaps be taken as indicating that their existence
on the islands is not an accidental matter and the view is
somewhat substantiated that the present land area has been
reduced from former larger proportions. Furthermore, it
seems quite certain that before the introduction of house rats
the native rodents were more generally distributed throughout
the different islands of the group than at present." Since the
various species have much general similarity in appearance, they
may have been frequently confused with house rats by casual
visitors, and so neglected with the result that very little is as
yet known of them. The six species hitherto described are the
following:

CHATHAM ISLAND RICE RAT

ORYZOMYS GALAPAGOENSIS (Water-house)

Mus galapagoensis Waterhouse, Zool. Voyage of the Beagle, Mammalia, pt. 2, p. 65,

1839 ("Chatham Island," Galapagos group).
FIGS.: Waterhouse, 1839, pi. 24 (animal in color), pi. 33, fig. 8, a-c (skull and teeth),

pi. 34, fig. 14, a (lower jaw).

Slightly smaller and more delicately built than a roof rat,
this species has rather large thin ears, a tail about as long as the
body, and in color is brownish above, with a mixture of black
hairs and others having a subterminal yellowish band; the sides
are yellowish, and the under surface of the body is whitish, with
a faint yellowish wash and with dark gray bases to the hairs.
Feet whitish; tail distinctly bicolor, dark above, white below.
Head and body about 6 inches long (150 mm.) ; tail, 4.75 inches
(120 mm.); hind foot, iVe inches (about 30 mm.); length of
skull, 31 mm.

The colors and general appearance of many of the small
rodents of South America are rather nondistinctive, so that
cranial characters give the best clues to the genera. Oryzomys
is characterized by the double series of enamel tubercles on
the anterior molars, with narrow secondary enamel folds
between them.

SOUTH AMERICA 379

When Darwin visited Chatham Island in 1835 he found this
rice rat abundant. "It frequents the bushes," he wrote,
"which sparingly cover the rugged streams of basaltic lava,
near the coast, where there is no fresh water, and where the
land is extremely sterile." He seems to have made search for
it or other native rats on other islands, for he remarks that he
was unable to find it on other parts of the archipelago; nor has
any one since taken specimens. Indeed, Heller (1904), who
visited the island in 1898-99 and searched for it unsuccessfully,
believes "it is now probably extinct or else restricted to the
barren eastern part of the island where Darwin secured his
specimens." If it has really been extirpated, one may suppose
that the introduced brown and black rats have been too
active competitors.

BAUR'S RICE RAT

ORYZOMYS BAURI J. A. Allen

Oryzomys bauri J. A. Allen, Bull. Amer. Mus. Nat. Hist., vol. 4, p. 48, May, 1892
("Barrington Island," Galapagos group).

The rice rat of Barrington Island, in the Galapagos Archi-
pelago, is believed to be very Closely allied to 0. galapagoensis
of Chatham Island but "differs mainly in having a somewhat
shorter tail and less yellow on the upper parts" (Heller, 1904).
The general color is said by its describer to resemble very much
that of the cotton rat, Sigmodon hispidus. A series of average
measurements given by Heller: Total length, 273 mm.; tail,
136; hind foot, 30-31; greatest length of skull, 31.5.

This rodent is known only from Barrington Island, about 30
miles to the east of Chatham Island. Here Dr. George Baur
obtained the three specimens that served for the description of
the species a number of years ago. As Heller points out, the
lack of actual specimens of 0. galapagoensis makes close com-
parison out of the question, but there is every reason to believe
that it is distinct from the latter. Dr. Baur found it rather
common on Barrington, "between the bushes near the shore,
and also high up between grass and the lava rock." When
Heller visited the island ten years later, in 1898-99, he found
it "very abundant," inhabiting crevices among the loose lava
rocks and burrows and runways beneath bushes and brush
piles. "In habits it appears to be somewhat diurnal and was
as often seen at midday as at other times." Nothing further

380 EXTINCT AND VANISHING MAMMALS

seems to be known of the species, but presumably it still is to
be found on its island home.

INDEFATIGABLE RICE RAT
NESORYZOMYS INDEFESSUS (Thomas)

Oryzomys indefessus Thomas, Ann. Mag. Nat. Hist., ser. 7, vol. 4, p. 280, 1899 ("Inde-
fatigable Island," Galapagos group).
FIGS.: Orr, 1938, pi. 25, figs. 3, 3a (skull).

This and the three other remaining rodents of these islands
are placed in a somewhat aberrant genus, Nesoryzomys, which,
though thus far considered endemic, may nevertheless eventu-
ally prove to be represented on the mainland of South America,
as Osgood (1929) has remarked. Compared with typical
Oryzomys, the frontal bones are much narrower, with rounded
orbital edges; the snout is more elongate, the nasals narrower
and less convex in profile, and the tooth rows are shorter.
Externally the fur and the tail are shorter. The present species
is dull mouse gray above, heavily lined with black, and with a
slight tinge of fawn on the posterior back; below, the hairs are
slaty at the base, tipped with white. The tail is about as long
as head and body, well haired, black above and white below.
An average specimen measures: Total length, 269 mm.; tail,
108; hind foot, 29; ear, 22. Skull length, 35.3.

This species was described from a specimen taken on Inde-
fatigable Island by the Rothschild Expedition. Later, in
1898-99, Heller found it abundant on the same island, as well
as on South Seymour Island, and secured a series of specimens.
He did not find it on the adjacent North Seymour Island.
He writes (1904) that it "seems to be more nocturnal in habits
than Oryzomys. It inhabits burrows or rock crevices beneath
bushes." Nothing further is known of its habits. Although
at present believed to be still common, it is probable that if
Old World rats and mice or house cats (the latter recently
introduced) should become established on these islands, as
they have on some others of the group, this native rat would
be in jeopardy.

NARBOROUGH RICE RAT
NESORYZOMYS NARBOROUGHI Heller

Nesoryzomys narboroughi Heller, Proc. California Acad. Sci., ser. 3, zool., vol. 3, p. 242,

Aug. 31, 1904 ("Narborough Island," Galapagos group).
FIGS.: Orr, 1938, pi. 25, figs. 2, 2a (skull).

SOUTH AMERICA 381

The rice rat of Narborough Island is described as similar to
N. indefessus of Indefatigable but larger, the feet and ears
especially larger, the former 31 mm. or greater in length;
coloration much darker above, chiefly blackish mixed with
some rusty brownish; below, darker gray (Heller, 1904). The
skull is wider, with longer nasals and shorter palatal foramina
than in the latter. Total length, 303 mm. ; tail, 131 ; hind foot,
33; ear, 23. An adult male skull measured 41.5 in length.

According to Heller's account, "this species was found in-
habiting the cracks and fissures in barren black lava fields near
the coast of Mangrove Point, Narborough Island. Individuals
were rather scarce at this locality, perhaps owing to the paucity
of the vegetation. The contents of several stomachs were
examined and found to contain a reddish material resembling
pulverized Crustacea." At the time of Heller's visit the Euro-
pean rats and house mouse apparently had not reached Nar-
borough Island. Should they later become established there,
it is likely that this native species will suffer.

SWARTH'S RICE RAT
NESORYZOMYS SWARTHI Orr

Nesoryzomys swarthi Orr, Proc. California Acad. Sci., ser. 4, vol. 23, p. 304, Sept. 1,
1938 ("From the vicinity of Sulivan [ = Sullivan] Bay, James Island, Galapagos
Islands").

FIGS.: Orr, 1938, pi. 25, figs. 1, la (skull).

This island race is evidently closely related to N. indefessus
of Indefatigable Island and is indistinguishable from it in
color, but the skull shows broader nasals, a larger rostrum,
greater diastema and much larger molariform teeth than any
of the other species of the genus yet known. The average
measurements of three males are given as: Total length, 312.3
mm.; tail, 133.7; hind foot, 35.7; length of skull, 40.6.

The four specimens from James Island, from which no
mammals had previously been known, were collected in 1906
by J. S. Hunter for the California Academy of Sciences and
may be regarded as the representative form on that island, of
the indefessus type, which is again present on Narborough
Island in the much darker N. narboroughi. In the 30 years and
more since the species was discovered, no other examples have
been collected by more recent expeditions, so that, as Orr
writes, "the possibility exists that this form may now be

382 EXTINCT AND VANISHING MAMMALS

extinct as a result of the introduction of non-native old world
rats. Examples of Rattus rattus alexandrinus were taken dur-
ing the same visit to James Island on which the specimens of
N. swarthi were secured." Heller (1904) remarks that "leav-
ing out of consideration Albemarle, James, Charles, and
Duncan, the islands which are inhabited by introduced species,
it seems remarkable that large islands like Abingdon, Bindloe,
and Hood should lack indigenous species of Muridae [Criceti-
dae]. A careful search, however, failed to discover any mam-
mals on these islands. It is probable that James, Duncan, and
Albemarle islands, intervening between the ranges of the two
species of Nesoryzomys, were until recently inhabited by
indigenous species of this genus which became extinct upon
the introduction of" the house rats. This suggested distribu-
tion is now partly substantiated through the subsequent dis-
covery of Nesoryzomys of a related type on James Island.
Moreover, the fact that Heller failed to find it in 1898-99,
while Hunter secured four specimens in 1906, merely shows
how elusive small mammals may be, even when sought for by
so skilled a collector as Heller.

DARWIN'S RICE RAT
NESORYZOMYS DARWINI Osgood

Nesoryzomys danvini Osgood, Field Mus. Nat. Hist., zool. ser., vol. 17, p. 23, July 12,

1929 (Academy Bay, Indefatigable Island, Galapagos group).
FIGS.: Orr, 1938, pi. 25, figs. 4, 4a (skull).

From its nearest island relatives, N. indefessus and nar-
boroughi, this species differs in its decidedly smaller size and in
its bright fulvous coloration, which extends to the entire under
parts. The skull is slender and without sharp ridges or angles.
Total length, 222 mm.; tail, 89; hind foot, 27; length of skull,
30.

This small, bright-colored species is remarkable in that it
furnishes the only instance known of two species of the genus
occurring side by side on the same island, "indicating a dis-
tinction of long standing." Three specimens were taken at
Academy Bay and a fourth at Conway Bay, Indefatigable
Island, while on the same dates 12 examples of N. indefessus
were taken at the former and 13 at the latter locality, indicating
that it is probably much the less numerous. Orr (1938) records

SOUTH AMERICA 383

the capture of two additional specimens at Academy Bay by
the California Academy's Expedition of 1905-6, making six
known examples. Though this species at present may be in no
danger, it is likely to disappear first if Old World murids are
introduced or other unfavorable circumstances intervene.

Family MYOCASTORIDAE : Coypus

COYPU; "NUTRIA"; " SWAMP BEAVER"
MYOCASTOR COYPUS COYPUS (Molina)

Mus coypus Molina, Saggio Stor. Natur. Chile, p. 287, 1782 (Chile).

SYNONYM: Guillinomys chilensis Lesson, Nouv. Tabl. Regne Anim., p. 126, 1842.

ARGENTINE COYPU
MYOCASTOR COYPUS BONARIENSIS (E. Geoffroy)

Myopotamus bonariensis E. Geoffroy, Ann. Mus. Hist. Nat. Paris, vol. 6, p. 81, 1805
(Parana River, Paraguay).

MYOCASTOR COYPUS SANTACRUZAE Hollister
%

Myocastor coypus santacruzae Hollister, Proc. Biol. Soc. Washington, vol. 27, p. 57,
Mar. 20, 1914 ("Rio Salado, near Los Palmares, Santa Cruz, Argentina").

FIGS.: Waterhouse, 1848, pi. 15, fig. 1; pi 16, fig. 1 (skull) ; Hudson, 1892, p. 12 (female
and young)

The coypu, or nutria, is so important as a fur animal and has
in some places become so depleted that it may be briefly con-
sidered here. With somewhat the appearance of a large musk-
rat, this is a dark yellowish-brown or reddish-brown animal,
with longer and coarser guard hairs than a muskrat, which
when plucked disclose a soft velvety coat of under fur of a dark
slaty color. The tail is about as long as head and body and is
covered with coarse scales, but unlike the tail of the muskrat
it is round instead of compressed from side to side. The large
hind feet are webbed for swimming. Tip of muzzle and chin
white. Length of head and body, about 25 inches; tail, 17.5
inches; hind foot, 5.5 inches. The large incisor teeth are of a
deep orange-red; the cheek teeth are four in each row, with
three outer and one inner enamel fold in each of the upper and
the reverse in each of the lower series set at an angle directed
forward. Length of skull, 4.8 inches.

As a species this large aquatic rodent is found in the tern-

384 EXTINCT AND VANISHING MAMMALS

perate parts of southern South America, from the valleys of
central Chile in about latitude 33 S., southward to the Straits
of Magellan and eastward across Argentina to extreme southern
Brazil, Paraguay, and Uruguay. Over this wide area there is
some geographic variation, so that at present three living
races are recognized: the typical form of Chile at lower alti-
tudes; the race bonariensis of northern Argentina, Uruguay,
Paraguay, and southern Brazil, in which the notch at the
posterior border of the palate is broadly arched and concave
instead of strongly V-shaped as in the typical coypus; and
lastly the Patagonian race M. c. santacruzae, of southeastern
Argentina, in which the skull is larger with the posterior border
of the palate V-shaped as in the Chilean animal, but the upper
cheek teeth increase conspicuously in size from before back-
ward, the last with nearly twice the crown area of the first,
instead of being of uniform size. All three races are evidently
closely related and to be distinguished only upon comparison
of skins and skulls.

The coypu, like the North American muskrat, is an animal
of lakes, streams, and swamps, but on the south coast of Chile
it is said to frequent the coastal tide waters as well, living in
the bays and channels among the small islands. Although
primarily a vegetarian, it is said also to eat shellfish. According
to Bennett the breeding season is in September and October,
but Gay (1847) states that there are two litters, or even three,
in a year, and young may be up to five or seven to a litter.
Hudson (1892), however, says that the number may be as
many as eight or nine, and he has published a sketch of an
adult female, with her brood, swimming, some of the young
being ferried along on her back, the others paddling behind.
He mentions the peculiar moaning cries the animals make,
but these may be given mostly at mating seasons, as with the
muskrats.

The soft, plushlike underfur of this animal early attracted
the notice of Europeans. At the beginning of the last century,
Azara in his account of the species in Paraguay (where it was
known by the native name of "quouiya") says that the fur
was then beginning to be used for felt hats. Since his day this
use has greatly increased, while the pelts also, after the long
coarse guard hairs have been plucked, are much used in fur.
They dye easily and give a soft plushlike effect. So great has

SOUTH AMERICA 385

become the demand for this fur that in recent decades the
animals have in places become greatly depleted.

Hudson, writing in 1892, said that it was much more abun-
dant in the La Plata region 50 years previously than at that
time. Its skin was largely exported to Europe. "About that
time the Dictator Rosas issued a decree prohibiting the hunting
of the coypu. The result was that the animals increased and
multiplied exceedingly, and abandoning their aquatic habits,
they became terrestrial and migratory, and swarmed every-
where in search of food. Suddenly a mysterious malady fell
on them, from which they quickly perished, and became
almost extinct." This account of a sudden and decimating
epizootic among these rodents recalls the cyclic fluctuations
known among such northern rodents as meadow mice and
snowshoe hares. No other evidence is known to me of such an
epizootic elsewhere among the coypus.

In his account of the mammals of Uruguay, Sanborn (1929)
writes: "Coypus have been so hunted for their fur that they
are now quite scarce in settled districts. I heard of many at
the Laguna Negra in Rocha but did not have a chance to
visit there. At most places tBe people said there were a few
left where many had been." The Argentine Government, in
1931, issued a bulletin on the nutria, remarking on the great
persecution to which it had been subjected on account of the
value of its fur and calling attention to the fact that in conse-
quence it had now disappeared from many parts of Argentina,
while recent droughts had augmented these losses. The export
of the fur is now under governmental control, and already
there are farms for breeding it. The report adds that the
hunters are now their worst enemy. They hunt not only for
the fur but also for the meat, which is palatable and is said
even to have been served by a hotel in Goya.

Numerous breeding farms have of recent years been estab-
lished in Uruguay and have given excellent results. Devin-
cenzi (1935) states, however, that "pitiless persecution by the
professional hunters, without regard to season, sex, or size
has so decimated the numbers in many sections of Uruguay,
that where once it was abundant, it is now rare." Frank G.
Ashbrook, summing up the situation, writes to Dr. Harper
(1935): "The nutria is perhaps not in as great danger of ex-
termination as the chinchilla, but strict conservation laws

386 EXTINCT AND VANISHING MAMMALS

applied to this animal are necessary." There is the added
circumstance that the nutria is much more prolific than the
chinchilla, so that with reasonable opportunity of breeding it
should maintain its numbers better.

Live animals have also been exported to various European
countries for breeding purposes and appear to have flourished.
In France, where it is known as the "ragondin," the nutria
was bred in captivity, at least as early as 1882, by Pays-
Mellier at Champigny on Veude, and in 1888 by Edgard
Roger in the park of the Chateau de Naudy (Seine et Marne) .
In the two decades preceding the World War of 1914-18,
there were apparently a number of persons breeding these
animals in France, but the war seems to have put a stop to
such enterprises until about 1925, when the raising of this
animal was again taken up, especially in southern and south-
western France. The nutria seems to adapt itself readily to
conditions under captivity and reproduces normally. Eco-
nomic conditions during the past ten years in France appear
to have been unfavorable, however, for the hopes entertained
for financial profit from this source and most of these breeding
colonies have been given up. Many animals have escaped and
established the species as an -element of the wild fauna, which
may in time prove a valuable asset or the reverse. At present
the capture of such wild individuals is subject to the hunting
restrictions applied to other game (Bourdelle, 1939).

Family DINOMYIDAE: Giant Rats
"PACARANA"

DlNOMYS BBANICKII Peters

Dinomys branickii Peters, Monatsb. Akad. Wiss. Berlin, 1873, p. 551 (Colonia Amable
Maria, Montana de Vitoc, Andes of central Peru).

SYNONYMS: Dinomys pacarana Ribeiro, Arch. Escola Sup. Agric. Med. Veterin., vol.
2, pp. 13-15, 1919 (" Amazonas, Brazil "); Dinomys branickii occidentalis Lonnberg,
Arkiv Zool., vol. 14, pp. 49-53, 1921 ("Gualea, Ecuador"); Dinomys gigas
Anthony, Amer. Mus. Nov., no. 19, pp. 6-7, 1921 ("La Candela, Huila, Colombia").

FIGS.: Peters, 1873, pis. 1-4; Sanborn, 1931, pi. 5 (skull); Mohr, 1937b, figs. 1-9
(photographs).

It is remarkable that this large rodent should continue to be
so rare in collections that, although first described in 1873 on
the basis of a specimen captured in the Andes of central Peru,

SOUTH AMERICA 387

it remained practically unknown thereafter until 1904, when
Goeldi (1904) published an account of two living individuals
sent to him at Para from the upper Rio Purus, Brazil. These
were the first he had known of in 20 years' acquaintance with
the fauna of the country. Another, perhaps from the same
region, was described by Ribeiro in 1919 as Dinomys pacarana
on the basis of its brown instead of black color; and two years
later Lonnberg named as a new race, occidentalis, a specimen
from near Gualea, Ecuador. The animal had meanwhile been
taken in Colombia, near La Candela, Huila (J. A. Allen, 1916b),
and this specimen was in 1921 made the type of D. gigas
Anthony. In 1922-23, Edmund Heller secured a series of 23
specimens from the natives about Buena Vista and Vista
Alegre on the Rio Chinchao, and at Pozuzo, Peru, and pur-
chased another at Manaos, Brazil. On the basis of the speci-
mens available in North American museums, Sanborn (1931)
reviewed the history of the species and showed that all the
names given doubtless applied to but a single form. In 1937
Dr. Erna Mohr (1937b) published an account, with photo-
graphic illustrations, of a live one in the zoological gardens at
Hamburg, with remarks on jireviously known specimens in
Europe.

From Sanborn's account it appears that this large and
heavily built rodent is found in the Andean region from
"central Colombia through Ecuador to central Peru, and east
to the Rio Purus region of Brazil." Attaining a length of head
and body of about 730-790 mm., with a tail of about 190 mm.,
it is of a black or brown color, with on each side of the midline
two more or less continuous, broad white stripes, and on the
sides two shorter rows of white spots. Apparently older
animals have the stripes broader and more conspicuously
white. The ears are short and rounded, the tail stout and
cylindrical; the skull measures in length some 153 mm. (about
6 inches) in the adult male, but the females are smaller, with a
skull length of about 141 mm., and the bodily proportions
correspondingly less. The incisors are disproportionally
large, the cheek teeth small relatively, each showing three
transverse enamel folds, with in the three last teeth a small
additional posterior fold.

The Tupi name, pacarana, signifies "false paca," since the
size and color pattern recall those of the common agouti-like

388 EXTINCT AND VANISHING MAMMALS

paca (Cuniculus). The present animal, however, is extraor-
dinarily different in its stout, heavy appearance and thick
tail of about half the body length. Goeldi describes the gait
as "waddling" on account of the plantigrade position of the
feet and the shortness of the legs, while the whole appearance
"reminds one of an immense rat well advanced in develop-
ment towards a bear." Of his captive animals Goeldi (1904)
writes that they are "of a peaceful, phlegmatic disposition,"
exhibiting as a predominant trait "a combination of leisurely
movements and supreme good nature . . . It is not easily
irritated, and permits one to stroke and to scratch its head
and back, and only occasionally manifests its displeasure by a
low guttural growl. I have never yet observed a manifest
intention to bite. When let out of the cage it makes no attempt
to escape, and limits its excursions to an exploration of the
immediate neighborhood in search of something to eat . . .
This phlegmatic disposition seems to me to be a very precarious
endowment for the struggle for life ; and considering the evident
advantages which result to the smaller domestic rodents . . .
from their nervously active constitution, it would not be
strange if the species should tend to disappear. The apparent
rarity of Dinomys may possibly find its explanation in the con-
sequences of such a psychological endowment in a more
nervous environment; but it is also possible that this rarity
is because of the circumstance that the real habitat of the
species has not yet been ascertained." Goeldi's captive
female shortly gave birth to a young one and died in unsuccess-
ful parturition of the second fetus, so that two young at a
birth is probably normal, indicating a slow rate of increase
(see also Tate, 1931). Dr. Mohr corroborates Goeldi's account
of the slow and gentle behavior of the animal.

Heller, who secured the series in the Field Museum, is the
first to have gleaned much information as to the habitat of the
species, and owes the skulls he secured to the habit of the
natives of the upper Huallaga River of preserving the skulls
of the game they kill, hanging these in their huts for good luck
in hunting. He states (in Sanborn, 1931) : "It is not a fighter
but merely fights as a last resort to save its life. It is slow in
motion and can not turn about quickly, therefore it has no
rear protection from alert foes like ocelots, tayras, coatis, etc.
It therefore lives in rocky cliffs, or holes in the ground by

SOUTH AMERICA

preference, where it can back up and secure rear protection."
In several months' stay in the region where it occurs he never
met with one of these animals, though he was taken to their
haunts by the natives, who use dogs to trail it.

Because of its large size the pacarana is sought out by
natives for food, while its lethargic manners and inoffensiveness
make it easily vulnerable to its enemies. In habits it is likely
to prove largely nocturnal, but captives seem to be active by
day as well. Unless in some way given protection it is likely
to be exterminated in the limited area where it occurs. A
similar fate at the hands of the natives very likely in former
days overtook its West Indian relatives, Elasmodontomys and
Amblyrhiza.

Family CHINCHILLIDAE : Chinchillas

SMALLER, or CHILEAN, CHINCHILLA; " COAST CHINCHILLA";
" CHINCHILLA BASTARD"

CHINCHILLA LANIGERA LANIGERA Bennett

Chinchilla lanigera Bennett, Gardens and Menagerie Zool. Soc. London, vol. 1, p. 1,
Oct., 1829 (Coquimbo, Chile).

SYNONYM: Eriomys chinchilla Lichtenstein, Darstellungen neue oder wenig bekannt.
Saugeth., pi. 28 and text, 1830 (Chile implied) ; Chinchilla velligera Prell, Zool.
Anz., vol. 108, p. 100, Nov. 1, 1934 (based on Bennett's description; hence Co-
quimbo, Chile).

FIGS.: Lichtenstein, 1830, pi. 28 (colored fig. of animal); Bennett, 1833, vol. 1, pi. 1,
(skeleton).

PERUVIAN CHINCHILLA; "CHINCHILLA REAL"
CHINCHILLA LANIGERA BREVICAUDATA Waterhouse

Chinchilla brevicaudata Waterhouse, Nat. Hist. Mammalia, vol. 2, Rodentia, p. 241,
1848 ("Peru").

BOLIVIAN CHINCHILLA; "CHINCHILLA CORDILLERANA"

CHINCHILLA LANIGERA BOLIVIANA Brass
Chinchilla boliviana Brass, Aus dem Reiche der Pelze, vol. 2, p. 613, 1911 (Bolivia).

The nomenclature of the chinchilla has been much confused
owing to the fact that the naturalists of over a century ago
believed that Molina's description of his Mus laniger applied
to it. In 1830 Lichtenstein pointed out that Molina's animal

390 EXTINCT AND VANISHING MAMMALS

was too small to be regarded as identical with the chinchilla,
and himself proposed a new genus Eriomys and named the
animal Eriomys chinchilla. However, in the year previous
Bennett had already described the animal anew from specimens
and given it the name Chinchilla lanigera, thus creating for it
a special genus. Molina's name, Mus laniger, is now regarded
as pertaining to a small, long-haired mouse, Abrocoma, so that
Bennett's new name is not invalidated and becomes the correct
term for the chinchilla. His account was based, he says, on
two animals brought back to England by Surgeon Collie,
who accompanied Captain Beechey on his exploring voyage
around Cape Horn to the northwest coast of North America in
1825-28. On the voyage out their vessel stopped at Concep-
cion and Valparaiso, Chile, but on the return put in for a few
days at Coquimbo, a short distance to the north of these ports.
Here evidently were obtained two chinchillas, one of which
survived the voyage and reached England alive, while the
other died and was preserved as a skin and skull to serve later
as the basis of Bennett's description. In the later account by
Waterhouse (1848) this specimen is mentioned as in the British
Museum. The living specimen seems to have been presented
by Surgeon Collie to Lady Kriighton, who in turn gave it to the
London Zoological Gardens, where it lived "for some time"
and was "said to be from Coquimbo." It was perhaps the
skeleton and internal anatomy of this individual that were
described by Yarrell and Bennett. It seems clear that what-
ever Molina's M us laniger was it could not have been a true
chinchilla and that the first tenable name for the latter was
applied to the animals brought back by Collie from Coquimbo.

This is a stocky, short-limbed rodent, with a head and body
about 9 inches long, and a well-haired, somewhat tufted tail
about 5 inches long to the end of the vertebrae, beyond which
the tuft projects about two inches. The ears are broad, about
two-thirds the length of the head, prominent and oval. The
fur, about an inch in length and extremely soft, is a beautiful
gray strongly mottled with dusky or black above, passing into
an "impure yello w- white " below. Feet dull white. Tail
along the middle line above and below, as well as its tuft,
black, with the sides dull white. Length of cranium about 2.5
inches.

From time immemorial the thick fur of a delicate buffy-gray

SOUTH AMERICA 391

tint has been used in wearing apparel, first by the Incas and
other native peoples of the southern Andes and later by Euro-
peans. The former likewise utilized the long hair for weaving
into cloth and the flesh was highly regarded as food (Ashbrook) .
When the fur was introduced by dealers to the European
trade, it became highly prized and much sought after. The
price in recent decades has been so high and the animals
themselves have become so reduced that naturalists have
found it difficult to assemble sufficient series to determine the
limits of geographic variation, so that the definition of races
has been to this day unsatisfactory. A recent writer (Prell,
1934) has attempted, however, to distinguish and allot names
to three races, the Chilean, the Bolivian, and the Peruvian,
but it can not be said that the characters of these are as yet
sufficiently defined or their respective ranges traced. The
name Eriomys chinchilla was given by Lichtenstein to a skin of
a chinchilla obtained through fur traders and is said to be still
in the Berlin Museum. A careful perusal of that author's
account reveals nothing of the origin of the specimen beyond
the fact that it was one of others traded through Carthagena
and La Guayra, Venezuela. t)n the other hand, he implies
that Chile is the home of the species, rather than Peru, as
sometimes given, for in mentioning Molina's chinchilla or
Mus laniger from Chile, he adds, "Die Uebereinstimmung des
Namens, sowie der Fundort, machen es sehr wahrscheinlich,
das damit unser Thier gemeint sei" (the correspondence of the
name as well as of the locality makes it very probable that our
animal is the same) . Moreover, the tail length of the specimen
figured of half size, would be 6 inches, of the hind foot about 2.5.
It was the largest of various skins at a fur dealer's. It thus
seems most likely that Lichtenstein's name is a synonym of
C. laniger a, published by Bennett very shortly before. This
leaves the latter's Chinchilla brevicaudata as the first name
applied unequivocally to the Peruvian chinchilla. Prell (1934)
believes that the Bolivian chinchilla is also a distinct race and
adopts for it the name boliviana given by Brass in a German
treatise on fur. The ranges and characters of these races may
be further outlined, as follows :

The Chilean chinchilla is said to occur from about the Rio
Chupa in latitude 32 S. northward along the western base of
the Andes to the region of Copiapo in northern Chile. It is

392 EXTINCT AND VANISHING MAMMALS

said by Gay to prefer the warmer areas along the coastal
foothills and the inner valleys. It is thus somewhat of a low-
land animal. In the fur trade it passes under the names of
small, or coast, chinchilla, Chilean chinchilla, and bastard
chinchilla, and its fur is the least valuable.

The Peruvian chinchilla is described as somewhat shorter-
tailed and larger in body (length of head and body, 14 inches ;
tail vertebrae 3 inches). The general tint is more silvery than
in the Chilean form, the back silvery gray with a clouding
of blackish. Its range is said by Prell to be at higher altitudes
on the western slopes of the coastal Cordillera of Peru at alti-
tudes of 8,000 to 10,000 feet. In the trade this is known as the
big chinchilla or royal chinchilla and is the most prized as fur.

The Bolivian chinchilla has more rounded ears than the
two others and is short-tailed like the Peruvian race, but in
color it is with difficulty distinguished from the coast chinchilla,
even by the fur dealers among whom it passes under the names
Bolivian, La Plata, or Argentine chinchilla, after the ports
whence it is shipped out. The range of this race is the eastern
Andes and upper plateaus of Bolivia and northern Argentina,
especially the provinces of Jujuy, Salta, Catamarca, and La
Rioja.

According to Ashbrook (1928) the chinchillas are very
swift in movement and in the early morning and late forenoon,
when they are abroad, are shy, dashing at once to the shelter
of their holes at the least alarm. Their curiosity, however,
often prompts them to reappear shortly after, to see the cause
of their fright. They generally feed actively early in the
evening, sitting on their haunches and holding the food in their
fore paws. They are fond of grains, seeds, fruits of shrubs, dry
and green herbs, mosses and lichens. "Of all fruits, they seem
to prefer the Algarrobilla, the seed of which is sweet and nutty,
although the pod is astringent. The pods are found stored in
their dens."

Because of the extremely soft quality of the fur and its
delicate tints of gray, the chinchilla is the most valuable and
most sought after of all South American furs. Formerly so
numerous that travelers in their haunts could see "thousands
of them daily," they have now become so rare that in parts of
the range, especially in Peru, whence the royal chinchilla
comes, they are practically exterminated. Ashbrook (1928)

SOUTH AMERICA 393

states that till recently at least they were still common in
Villenar Province, Chile, but are gone from the provinces of
Antofagasta and Tacna. The countries in which these animals
occur have of recent years realized the value of chinchillas as
fur bearers, and not only have passed laws to safeguard them
from extermination if possible but also have made some at-
tempts to breed them in confinement for fur. Dr. Francis
Harper through correspondence has assembled some notes on
this subject, from which some of the following details are
extracted: In the winter of 1923, M. F. Chapman succeeded
in securing five male and seven female coast chinchillas and
bringing them to California for breeding purposes. At the end
of the third breeding season (1926) his stock had increased to
65 animals. He found them easy to handle, for they quickly
become tame, but they do not thrive in a damp atmosphere.
In Chile captive animals have been found to produce two
litters a year and are polygamous. The gestation period is
about 111 days. Females will mate and commence breeding at
four months of age, but it is inadvisable to breed them so
young. Those brought to the United States have, after four
years in captivity, raised even \hree litters a year. The num-
ber of young at a birth averages two but varies from one to
four (Ashbrook, 1928). Following Chapman's success with
these animals, the stock was divided in order to establish
other "ranches." In 1935 M. L. Weaver, who has had a share
in raising this stock, wrote that "we have ranches at Logan,
Utah; Idaho Falls, Idaho; Afton, Wyoming; and Madison,
Wisconsin ; all doing well." He knew of several later shipments
of live animals from South America, but none had been success-
ful. In Chile, the home of this chinchilla, local attempts are
being made to raise the animals and to develop an improved
cross-breed, but efforts to hybridize the coast chinchilla with
other forms have been unsuccessful. The former, though
prolific, is less valuable than the Peruvian chinchilla, has
coarser and less grayish fur, and is worth about half as much in
price. Hitherto most of the efforts of breeders in the United
States have been directed to building up a stock, some of
which has been sold for fabulous prices to prospective breeders
as royal chinchillas, which strictly they are not. There are
several chinchilla farms in Chile, where the coastal form is
still locally common.

394 EXTINCT AND VANISHING MAMMALS

Concerning the Bolivian race, the so-called "Indian" or
"Cordilleran" chinchilla, Carlos Garcia-Mata, of the Argentine
Embassy, wrote Dr. Harper that in April, 1933, the Argentine
Government started a chinchilla farm in Abra Pampa, Jujuy,
at 12,548 feet altitude in the Andes, and has been very success-
ful with it. From a start of 9 animals, in three years the number
had increased by breeding to 61, an average per pair of 3.8 a
year. It was planned to start selling breeding stock to local
breeders by 1937. A decade before, the Argentine Government
had made a similar attempt in cooperation with local hunters
at the same place, but most of the stock was lost through
ignorance of proper methods of care. They prove very sensi-
tive to humidity and quickly succumb if kept in a damp place.
At the present time there are stringent laws establishing close
seasons and regulating the hunting of these animals. In 1926
the hunting, exportation, transportation, and sale of chin-
chilla skins were prohibited by law. In Bolivia, likewise, laws
have been passed for its protection, and the exportation of the
fur was prohibited under laws of 1920 and 1922. The hunting
of chinchillas in Bolivia is said to be in the hands of a monopoly,
but because of the scarcity of the animals very few are taken.
In 1931, according to the American vice-consul at La Paz,
there were no chinchilla farms in Bolivia.

The royal chinchilla ("chinchilla real") of the Andes of
western Peru is the most valuable of the three forms because of
its longer and silkier fur and its pale "bluish" tint. The most
valuable skins formerly came to market at Oruro or Tacna and
Arica. This form has now been brought nearly to the verge of
extinction. Not only were they persistently trapped and
hunted by the natives for the fur trade, but it is said by Water-
house (1848) on the authority of Bridges, who traveled in Peru
a century ago, that they even trained a species of grison (of the
weasel family) to hunt them by entering their burrows and
capturing them as they endeavored to escape. Dogs also were
used in hunting. In a letter to Dr. Harper in 1935, Carlos
Garcia-Mata, of the Argentine Embassy, wrote that a few
still remain in the steep and inaccessible rocks of the lower
Andes, but that attempts to capture a pair for breeding stock
have resulted in failure. Where 30 or more years ago they were
common, and the pelts brought only $6 or $7 a dozen, by 1930
they had become so scarce that pelts brought as much as $200

SOUTH AMERICA

395

apiece. However, as early as 1920, hunting and the sale of
pelts as well as exportation were prohibited by the Peruvian
Government, except under license. There continued to be
nevertheless a small amount of illicit trade, but this is now
apparently reduced to a minimum, for the animals are too
scarce to make the trade remunerative.

In 1931, William C. Burdett wrote to Dr. Harper, in response
to inquiries, that "it is almost impossible to obtain accurate
information in regard to the Peruvian chinchilla, since it is
practically extinct. No one in this office (American consul-
general's) has ever heard of a chinchilla ranch. As far as is
known no live chinchillas have been seen in Lima for over 20
years. The Peruvian center for the traffic in wild animals and
their pelts is Sicuani, a mountain town near Cuzco in southern
Peru, and the Indian dealers there report that they have not
seen any specimens of chinchilla for many years."

Formerly "most of the Chilean pelts were purchased by
buyers in Coquimbo," and some statistics from the customs
office of that city presented by Bidlingmaier (1937) are inter-
esting. It seems that in 1905 the quantity sold amounted to
18,153 dozen valued at from $100 to $110 (? per dozen). In
the following year the number sold was about half that figure,
and in 1907 it was again reduced to half that of 1906. In 1909
it had dropped to 2,328 dozen, and the price had risen to $400
to $500. European markets instruct their agents "to purchase

Peruvian chinchilla (Chinchilla lanigera brevicaudata)

396 EXTINCT AND VANISHING MAMMALS

skins at any price" thus increasing the greed of the "chinchil-
lero," or professional native hunter, who redoubles his efforts.
The relentless and systematic methods of destruction by the
latter are recounted and the difficulties of exercising any effec-
tive control by governmental agencies. Bidlingmaier suggests
that perhaps the most feasible way would be to enact "stringent
laws especially aimed at the traffic in contraband skins."
Some attempts have been made to breed these animals in
Chile, as in the vicinity of Vallemar and Copiapo, where their
natural food plants are available, especially the "algarrobilla,"
but so far "nothing of great value has been revealed . . .
Today there remain only five licensed criaderos (breeding
farms), two of which are now presumed to be working in
conjunction with a syndicate from the United States."

Order CARNIVORA: Dogs, Cats, and Their Relatives

The families of carnivores are discussed on page 134. Four
groups occur in South America, the cats, the dogs, the raccoons
and their relatives, and the bears. The bears occur only in the
Andean Mountains, while the raccoon family developed in this
continent. The bear and the Falkland fox, or wolf, are the
only carnivores considered in this section. J. E. H.

Family URSIDAE: Bears

SPECTACLED BEAR; "HUCAMARI"

TREMARCTOS ORNATUS ORNATUS (F. Cuvier)

[Ursus] ornatus F. Cuvier, in E. Geoffrey and F. Cuvier, Hist. Nat. Mammiferes,

vol. 3, pt. 50, p. 2 and pi, June, 1825 (Cordillera of Chile).
SYNONYM: Ursus frugilegus Tschudi, Fauna Peruana, pp. 11, 90, 1844 (Peru, probably

near Lima).

NORTHERN SPECTACLED BEAR
TREMARCTOS ORNATUS MAJORI Thomas

Tremardos ornatus majori Thomas, Ann. Mag. Nat. Hist., ser. 7, vol. 9, p. 216, 1902
("Southern Ecuador, probably the province of Azuay").

SYNONYMS: Ursus ornatus thomasi Hornaday, Bull. New York Zool. Soc., no. 45, p.
748, 1911 ("Andes of southern Colombia"); Tremarctos lasallei Maria, Bol. Soc.
Colombiana Cien. Nat., vol. 13, no. 76, p. 115, Aug. 1924 ("Arauca, Boyaca
Province, Colombia").

FIGS.: Geoffroy and Cuvier, 1825, pi. 50; Hornaday, 1911, pp. 747-748, 4 figs, (photo-
graphs).

SOUTH AMERICA 397

This is the only member of the bear family living in South
America, where as a species it occurs from the Andes near the
Venezuela-Colombia boundary, across Colombia and Ecuador,
to Peru, northern Chile, and Bolivia. Although several names
have been given to specimens of the species, the supposed
differences appear to rest mainly on individual variations.
Thomas (1902) found that skulls from Ecuador differed from
the typical southern skulls in being larger, longer in proportion,
and more slender, and the teeth "rather larger throughout."
He therefore named these latter specimens as a northern race,
majori, but even yet it is uncertain what are the limits of
variation in the more northern bears, or how the ranges of the
two races should be drawn. As long ago as 1844, Tschudi dis-
tinguished the Peruvian animal as Ursus frugilegus on account
of the supposed shorter head, shorter soles of the feet, and
much slenderer body, but the differences appear again to be
of an individual nature, and the Peruvian animal is currently
believed to be inseparable from that of Chile. In 1911 Horna-
day casually bestowed the name Ursus ornatus thomasi on a
bear of this species from the Andes of southern Colombia
living in the New York Zoological Gardens, but, again, the
supposed diagnostic character, the lack of white markings, is
now known to be individually variable ; and Maria's Tremarctos
lasallei from Arauca, Colombia, based on a mounted skin,
having longer muzzle and claws than typical majori, can
hardly be other than a synonym of the latter.

In coloring and pattern both northern and southern races
are apparently alike. The entire body is uniformly black or
blackish brown, except that usually there is a narrow line of
white beginning on each side about halfway between the ear
and the eye, continuing forward nearly to the bridge of the
muzzle in front of the eye, then turning downward across the
cheek to the midline of the throat, and continuing parallel
with the corresponding mark of the other side, to the chest.
This marking, which from the semicircle about the eyes, gives
the bear its English name, shows, however, considerable varia-
tion. Its upper part may be washed with pale yellowish;
again, the muzzle may be light tan, or more extensively
whitish, to include the forehead, nose, cheeks, and throat.
As an opposite extreme, the white markings may be lacking
altogether in animals from the same region. Cuvier mentions

398 EXTINCT AND VANISHING MAMMALS

the total length as about 3 feet in the original specimen, height
at the shoulder about 20 inches. Thomas's measurements for
the type specimen of the northern race are: Head and body,
1,625 mm.; hind foot with claws, 210; ear, 75; greatest length
of skull, 263 mm. (in one of the typical race, 231, a male in
each case); zygomatic width, 169 mm. (in typical race, 163).

Various writers reiterate that very little is known of this
bear beyond the fact that it seems to be rare throughout its
range and is seldom seen or hunted. This fact and its restricted
range may warrant its inclusion among species in a precarious
situation. The specimen on which Frederic Cuvier founded his
description and of which he published a colored figure, was said
to have been brought from the Cordillera of Chile, and it
doubtless did come from that region, probably in northern
Chile. It is also known in Bolivia, where as well as in Peru, it
is called "hucamari" in the Quichua tongue. Krieg (1931)
mentions that it is reported from the Bolivian (?) Chaco, and
he obtained a skin from near Caraparisito. There it was said
to be exceedingly rare and very shy of men, though occasionally
accused of killing cattle. The individual exhibited in the New
York Zoological Gardens in 1911 was from near Quito, Ecuador,
and it seems likely that the center of abundance now lies in
southern Colombia, Ecuador, and parts of Peru.

Osgood (1912, 1914), who made special efforts to learn
something of the habits of this bear, writes of securing an adult
female and wounding a male accompanying it, in the moun-
tains about 10 miles northwest of Menocucho, east of Truxillo,
northern Peru. The region inhabited by this bear is here
extremely arid and mountainous, with a scanty vegetation
consisting mainly of cacti and small thorny bushes. The
mountains range from 1,000 to 5,000 feet in height and are not
greatly different in character from the desert plain stretching
westward to the sea. The principal food of the bears seemed to
be a pear-shaped fruit with a hard outer shell enclosing numer-
ous seeds, a species of Capparis locally known as "chapote."
"The region is for long periods almost waterless and animal
life is very limited . . . From reports received from local
sources, it is evident that bears are fairly common in numerous
localities in the arid region." The two seen by Osgood were
feeding at midday in the full glare of the tropical sun, but the
natives advised that the early morning was the best time for
hunting them and they use dogs in the pursuit.

SOUTH AMERICA 399

Quite different is the type of country inhabited by these
bears in southwestern Ecuador, where Tate (1931) found them,
and farther northeast in the Paramo de Tama near the Colom-
bia-Venezuela boundary, where, writes Osgood (1912, p. 58),
they are very seldom seen and are decidedly rare. In four
weeks' hunting he found only a fragment of dung and no tracks;
"natives say the bears live almost exclusively in the forest and
it is only on the very rare occasions when they wander out into
the cultivated clearings that they have been killed." Tate, in
company with a native hunter, found traces of an adult pair in
the rain forest on a large plateau in southwestern Ecuador, on
the Andean slope at about 4,000 feet altitude. He followed
the tracks a long distance up the slopes, and came upon six
places where the bears had "bedded down." He saw also
where they had broken down many trees 2 inches in diameter
and "were feeding on the seeds of a palm called "pambili"
trees from 80 to 100 feet high. They evidently climbed the
trees and brought down the whole fruit-stalk, which looks
somewhat like that of the royal palm. Numbers of the trees
had been climbed, some of them several times, since they had
both old and fresh claw marksX Other food was secured by
breaking down young palms, tearing open the green stalk and
eating the unopened leaves in the interior. According to
Olalla [a skilled collector], in the region about Quijos in eastern
Ecuador bears appear at a certain season of the year upon the
mountain sides to feed upon the ripe fruit of a certain tree,"
but the statement that they make large nests of sticks in the
tree tops for sleeping purposes, as Olalla told him, seems open
to question.

Nearly a century ago Tschudi (1844) gathered together many
notes on this bear in Peru, from which the following are culled.
He transcribes a few interesting points from the accounts of
the early Spanish explorers. Thus Ulloa said that the bear was
common in the provinces of Guijos, Macas, and Jaen de Braca-
moros. It was sometimes lassoed from horseback. It was
also found in the forests east of La Paz, Bolivia. According
to the account of Garcilasso de la Vega, it was rare in Peru, a
fact that he attributed to the method of hunting. Great
annual drives were held in which the Incas used as many as
25,000-30,000 Indians, whose lines would cover 20 to 25
leagues, converging toward a funnel into which the game of the

400 EXTINCT AND VANISHING MAMMALS

enclosed area was driven. As the circle became smaller, the
lines of Indians would become deepened by the addition of
their number on the outer edge of the line, to prevent animals
from breaking through. When the enclosed animals were
finally penned, the carnivores were all killed, but of the deer,
vicunas, or other such game, only a certain number of males
and old females were slain, the others freed. Tschudi had seen
similar drives in his time, but on a smaller scale, and agrees
that the number of bears thus captured or killed stands in very
small proportion to the number of other large carnivores, so
that their apparent scarcity is probably an actual one. He
doubts if the bear that served for Cuvier's description and
figure really came from the Chilean Cordillera but suggests
that the only likely harbor on the west coast of South America
from which such an animal might be shipped was Truxillo.
In distinguishing his supposed new species, "frugilegus," from
ornatus, Tschudi stresses the reports that while the latter
preys upon young deer, vicunas, and huanacos, the former is
chiefly a vegetarian and often does much damage in plundering
the maize fields of the natives. The available evidence does not
indicate that this bear is much of a predator but finds abun-
dant food in the way of fruits', leaves, or roots.

Family CANIDAE: Wolves, Dogs, Foxes

ANTARCTIC WOLF; FALKLAND Fox

DUSICYON AUSTRALIS (Kerr)

Canis vulpes australis Kerr, Linnaeus's Animal Kingdom, p. 144, 1792 (West Falkland

Island; see Osgood, Journ. Mamm., vol. 1, p. 35, 1919).
SYNONYMS: Canis antarcticus Bechstein, Uebers. Vierfiiss. Thiere Pennant, vol. 1, p.

271, footnote, 1799; Dusicyon antarcticus Thomas, Ann. Mag. Nat. Hist., ser. 8,

vol. 13, p. 353, 1914.
FIGS.: Mivart, 1890, pi. 8 (col.); Pocock, 1913, figs. 70B, 71 A, 73, 74B, D (skull and

teeth).

The Falkland Island fox, although often called a wolf, is not
a wolf at all and is not even closely related to the North
American coyotes, as Huxley formerly supposed. Instead it
is a near relative of the group of South American foxes, which
are now regarded as distinguishable under the generic title
Dusicyon. How this animal reached the Falkland Islands
will doubtless ever remain a matter for speculation.

SOUTH AMERICA 401

Foxlike in appearance, this animal stands 15 inches at the
shoulder. Its coat is thick and soft, lacking coarse long hair
and having a bushy tail. Pocock (1913) describes a specimen
as having the prevailing color of the body brown "relieved by
fine speckling due to the narrow pale band on the individual
hairs." Below, the color is brownish, except that the posterior
part of the belly and upper end of the throat are white, the
chin and lower jaw white with a fuscous tint. There is a
marked fuscous patch above the hock of the hind leg. The tail
on its basal two-fifths is like the back, the middle part black
and the tip white. The ears, according to Pocock, are unusu-
ally small. The skull is characterized among other things by
the lyrate sagittal area and truncated instead of pointed occip-
ital crest. In these respects it agrees with some other of the
South American "foxes." Mivart (1890) gives the following
measurements : Length of head and body, 970 mm. ; tail, 285 ;
hind foot, 180; ear, 65; skull length, 110. It is said that the
animal formerly occurring on the East Falkland Island was
smaller and redder than the one on West Falkland.

The history of the Falkland fox has been several times writ-
ten, most recently by Renshawtl931). He points out that the
animal was first discovered in January, 1690, on South Falk-
land by Strong's party, who captured one alive and kept it for
several months on their ship. Unfortunately the vessel
finally had occasion to discharge its guns, which so startled the
fox that it leaped overboard and perished. In his voyage of
1763-64, the French navigator Pernetty again found the
species, and the discovery of it had been credited to him by
writers, until Strong's earlier account was brought to notice.
In 1765 Commodore Byron took possession of the islands for
Great Britain, and his landing party records that several of
these foxes came to meet the men, wading out toward them in
curiosity at the strange apparition. The sailors, however,
believing the "wolves" were actuated by ferocity, at first
retreated, but later found that the animals were quite without
fear. The animals were so numerous that the men set fire to
the long grass and presently saw great numbers of them run-
ning to escape the flames. Byron brought a live one back to
England, and this individual was later described by Pennant
and thus became the basis of Kerr's name.

When Darwin during the voyage of the Beagle stopped at

402

EXTINCT AND VANISHING MAMMALS

the Falklands in 1833, the animal was still common there, but
its absurdly tame habits, he foresaw, would lead to its early
extermination. Indeed, he mentions that the Gauchos would
capture them by holding out a bit of meat to a fox in one hand
and stab the animal with a knife held in the other, when the fox
came within reach.

Such an abundant and soft-furred animal, thus easily caught,
attracted the notice of John Jacob Astor, then active in the
fur trade, who in 1839 sent men to the Falklands to collect
pelts, and great numbers of them were taken. Hamilton
Smith mentions seeing quantities of them in Astor's warehouse
at New York. Renshaw even says that the extermination of
the species from East Falkland may date from this exploita-
tion; at all events by 1863 they were already extinct in the
eastern part of East Falkland. When, later, the Scotch set-
tlers started raising sheep on the Falklands, the foxes seem to
have developed a taste for mutton and would kill sheep by
attacking one or two or three together. As a result a poisoning
campaign was undertaken and many were destroyed. In 1870
Byng wrote to the Zoological Society that they were almost
exterminated, and the last one is said to have been killed in
1876 at Shallow Bay, West Falkland.

Antarctic wolf (Dusicyon australis)

SOUTH AMERICA 403

The Antarctic "wolves" were said to feed on various native
birds, harrying the penguin colonies and driving the upland
geese to nest on small islands off the coasts. Seals were eaten
too. Their extreme tameness may have been a result of long
isolation and lack of contact with man, but their failure to
develop any wholesome fear of him may have been in part a
result of the use of such silent weapons as bolos and knives
and probably traps, rather than firearms.

Captain Fitzroy of the Beagle and Darwin in 1836 brought
back four of these animals, two of which are still preserved as
specimens in the British Museum, which has a skeleton in
addition. The Royal College of Surgeons had two skulls
which may now be in the British Museum, for according to
Pocock (1913) there are five crania in that institution. Pocock
adds that the other known material representing this animal
is in Paris, but he does not tell of what it consists. Renshaw,
however, states that the Leiden Museum has three specimens.
It was first exhibited by the Zoological Society of London in
1845. Twenty years later, in 1868, a pair was again sent to
the Society, but one only survived the journey. Again in
1870, a pair was sent by Byng, of which the male died on the
voyage.

Order PERISSODACTYLA: Odd-toed Ungulates

This order is represented by three Recent families:

(1) Equidae, the horses, asses, and zebras. The wild horses
are Eurasian, but all except a few individuals of the Przewalski
horse of Mongolia have become extinct.

(2) Rhinocerotidae, the rhinoceroses of southern Asia and
Africa.

(3) Tapiridae, the tapirs, with nose and upper lip produced
into a short proboscis, are found in Central and South America,
the Malay Peninsula, and part of the Malay Archipelago.

The odd-toed ungulates are represented in South America
only by the tapirs. The mountain tapir is poorly known and
is thought to be endangered by the agricultural development
of the northern Andes. J. E. H.

404 EXTINCT AND VANISHING MAMMALS

Family TAPIRIDAE: Tapirs

ROULIN'S TAPIR; MOUNTAIN TAPIR; "HAIRY TAPIR"
TAPIRUS ROULINII Fischer

Tapirus roulinii Fischer, Synopsis Mammalium, Addenda, p. 602, 1830 ("Summos

monies Andes Americae australis")-
SYNONYMS: Tapirus villosus Wagler, Syst. des Amphib., p. 17, 1830; Tapirus pinchacus

Blainville, Osteograph., Unguligrades, Genus Tapirus, pi. 3, fig., 1845; Tapirus

leucogenys and enigmaticus Gray, Proc. Zool. Soc. London, 1872, pp. 488, 490

(Assuay and Sufiac, Ecuador).
FIGS.: Gray, 1872a, pi. 21, fig. 1; pi. 22, fig.l (col. fig. of young and subadult) ; Sclater,

1878, pi. 39 (col., adult); Hatcher, 1896, pi. 4, figs. 2, 2a; pi. 5, fig. 2 (skull).

Very little seems to be known of the habits and present
status of this tapir of the northern Andes. It was first brought
to the notice of naturalists by M. Roulin, who communicated
to Baron Cuvier an account of the animal, which was published
by Cuvier in the Annales des Sciences Naturelles in 1829 and
further elaborated by Roulin in the following year in the same
journal. Roulin gave the tapir a French name, "Tapir pinch-
aque" from the name used by the native Indians to denote a
fictitious monster, but the Latin form was not used. In 1830,
both Fischer in the supplement to his Synopsis Mammalium
and Wagler in his System des Amphibiens, etc., proposed new
Latin names based on Roulin's description, but since it is not
evident which had precedence, it seems better to follow Thomas
(1880) in adopting Fischer's name, which commemorates the
discoverer of the species.

This tapir does not apparently differ in size from the lowland
species of the Amazon Valley, which is about as large as a pony.
The hair, contrary to the usual belief, is said by Sclater (1878)
to be rather short, the individual hairs about an inch long, and
the color nearly uniform black, shading to brownish; outer
edge of the ears and a spot at the corner of the mouth white.
Iris light bluish hazel, rather than brown as in the lowland
tapir. The form of the nasals is very different from that of
the latter, being long, and tapering, with concave outer borders.
The brain case is shorter and the sagittal crest less high.
Measurements given by Roulin are : Tip of snout to tip of tail,
5.5 feet; height at shoulder 2 feet 9 inches, but these are per-
haps not of a fully adult animal. Greatest length of skull (M.
C. Z.), 380 mm.; median length of nasals, 95.

The home of the mountain tapir seems to include the high-

SOUTH AMERICA 405

level forests from central Colombia to Ecuador and possibly
northern Peru. Most of the specimens seem to come from the
Andes of Ecuador. Its discoverer, Dr. Roulin, obtained his
specimens from the Paramos of Quindiu and Suma-Paz, during
his residence at Bogota, and it was met with by Goudot on the
peak of Tolima between 1,400 and 4,400 meters, in south-
eastern Colombia. P. L. Sclater (1870) quotes a letter from
Robert B. White dated from Popayan, Colombia, June 8, 1869,
in which he speaks of finding this tapir on the central Cordillera,
in the region of the volcano of Purace, adding that "they are
very shy, and I have not been able to get near them, but have
seen them at a distance of half a mile, with a telescope, bathing
themselves in a small lake. I have also seen the skins occa-
sionally brought in by the Indians . . . It is never found
at a lower elevation than 3,500 metres above sea-level
and it exists up to 4,200 metres. These animals are rarely
killed, because the skin only sells for" 3 shillings.

In the Santa Marta region of northern Colombia, there
are tapirs in the somewhat isolated Santa Marta Range, but
judged from a skull from Dibulla, in that region, the species
there is T. terrestris. Farther s\mth, however, in the Depart-
ment of Santander, M. A. Carriker, Jr., informs us that tapirs
are rather common in the southern part of the area at altitudes
from 8,000 to 10,000 feet, and are much hunted by the natives.
Owing to clearing of the forests for cultivation, the tapirs are
likely in time to become driven out or much depleted in
numbers.

Thomas (1880) mentions that a Mr. Buckley secured a series
of 15 specimens of this tapir at Sarayacu, Ecuador, but "un-
fortunately before the skins were prepared, a troop of revo-
lutionary soldiers put in an appearance and cut off the hoofs of
every specimen to make into amulets," which so vexed Mr.
Buckley that he abandoned the entire lot!

Mountain tapirs were first exhibited in the London Zoologi-
cal Gardens in 1878, according to Sclater (1878), who probably
did not then regard as the same the two youngish ones from
Ecuador that Gray in 1872 named enigmaticus and leucogenys.
There is a mounted skin and a skeleton from Ecuador in the
Museum of Comparative Zoology, and there are two speci-
mens in the Academy of Natural Sciences of Philadelphia
taken in the Uanganatis Mountains on the headwaters of the
Curaray River, Ecuador, at 14,000 feet, in 1935-36.

406 EXTINCT AND VANISHING MAMMALS

Whether the range of this tapir extends to Peru, or formerly
did, is uncertain. Tschudi (1844), however, wrote that the
presence of this species seems almost certainly to be indicated
on the testimony of the natives that a tapir is to be found in
the eastern slope of the inner Andes at 7,000-8,000 feet,
especially in the Ceja region. No one seems to have verified
this supposition, nor is there enough available information on
the species at hand to give any idea of its present abundance.

Order ARTIODACTYLA: Even-toed Ungulates

This order is represented in South America by two families,
the camel group and the deer; these are discussed on page 256.
The wild South American camel-like mammals, the guanaco
and the two races of the vicuna, are in need of protection from
hunters, chiefly natives. Four species of deer, representing
three endemic genera, are rare and appear to be decreasing in
numbers.

Family CAMELIDAE: Camels, Llamas

GUANACO; WILD LLAMA
LAMA GLAMA HUANACUS (Molina)

Camelus huanacus Molina, Saggio Storia Nat. Chile, vol. 1, p. 317, 1782 (probably the

Chilean Andes).

SYNONYM: Auchenia llama Water-house, Zool. Voyage Beagle, Mamm., p. 26, 1839.
FIGS.: Prichard, 1902a, pi. opp. p. 160; Cutright, 1940, pi. opp. p. 119 (photograph).

The wild huanaco, or guanaco, is believed to be the original
source of the domesticated llamas and alpacas of the Peruvian
aborigines, and with the smaller vicuna it is the only living
member of the camel family in America. The family seems to
have originated in North America, becoming much diversified
in middle and late Tertiary times and spreading to the Old
World and to South America. It has since completely died out
in North America.

The guanaco has somewhat the slender build and long neck
of the camel, but the ears are proportionately long, the tail is
short and bushy, there is no "hump," and the coat is soft and
woolly. In color the wild species is a dark fawn-brown above,
with a blackish face and white under surfaces. It stands about
3 feet 7 inches high at the shoulder, and the skull is about 11.5

SOUTH AMERICA 407

inches in basicranial length. The upper incisor and canine are
lancet-shaped and capable of inflicting a bad wound, but the
lower incisors are procumbent. There are callosities on the
inner side of the fore limbs.

The range is from the Andes of south-central Peru south-
ward to Patagonia, where it comes to the lower altitudes, and
thence to Tierra del Fuego. Very likely, when sufficient
series of specimens are available, more than the single geo-
graphic form may be distinguished. Indeed, Lonnberg (1913)
has described as Lama huanachus cacsilensis, an animal of
small size from Cacsile, Nunoa, Peru, which seems to be more
like the vicuna and according to Osgood (1916) is "not closely
related" to the common guanaco. Its status may await
further study.

Osgood (1916) states that the herds he saw on the Pampa de
Arrieros, between Arequipa and Puno, at the northern part of
the animal's range, are "almost if not quite the northernmost
now existing." Inhabiting as they do the alpine zone between
14,000 and 18,000 feet above the sea, they are perhaps also
the most lofty-living of the species. Barren though these
heights appear, the animals obtain pasturage sufficient to sus-
tain existence. Though little hunted by white man, they are
persecuted by the Indians, who shoot them from blinds
erected at waterholes. The skins are used in making beautiful
rugs or for saddle cloths, while the meat is of excellent quality.
B. T. Colley, writing to Dr. Francis Harper from Oroya, Peru,
in April, 1934, states that to the south of the Puno region, and
beyond the Atacama Desert, the guanaco is rather plentiful,
and he has seen herds of over a hundred animals on the con-
tinental divide above Santiago, while farther to the south
there are many more. Here, as winter sets in, they move down
from the higher regions to lower altitudes, retiring chiefly to
the Argentine side because of the abruptness of the Chilean
Cordillera. They have a characteristic habit of making zigzag
trails in which the angles tend to be equal and regular.

In Chile, Jose M. B. Toro reports (in litt., 1934), the guanaco
has become so much reduced in numbers in the past 20 years
or more because of persecution that in 1929 a decree was
passed prohibiting the hunting of it for three years, after
which the close time was extended for two years more to
December 1, 1935. At the present time there are still a few

408 EXTINCT AND VANISHING MAMMALS

scattered bands to be found in the steeper parts of the Cordil-
lera and along the Argentine border, but any close estimate of
their numbers is difficult.

On the pampas of southern Argentina and over most of
Patagonia the guanaco seems to be still common from the Rio
Colorado in about latitude 40 S., and is even said to have
increased within recent years.

Darwin has given a brief account of the abundance and
habits of the guanaco as he saw it a century ago in the Santa
Cruz region of Patagonia. They go in small herds of from half
a dozen to thirty, but he mentions one herd of "at least five
hundred." He speaks of their taking readily to water and
swimming to islands near the mainland. They have a certain
curiosity when alone but in herds easily become bewildered and
stampeded, a fact of which the Indians take advantage in
killing them. Both Darwin and Prichard (1902a) mention
coming upon places where great numbers seem to have perished
and left bones bleaching on the ground. Prichard was told
that in the winter previous enormous numbers of guanacos had
sought Lake Argentine and perished there of starvation. "In
the severities of winter they seek drinking-places where there
are large masses of water likely to be unfrozen. The last few
winters " had been so severe that great havoc had been wrought
among the animals. Prichard crossed Patagonia from the Rio
Chubut southwesterly to the Andean foothills, and thence
proceeded southward, coming out at Punta Arenas on the
Straits of Magellan. Over most of this country guanacos
were common, and in regions where they were not much hunted
were not very shy.

Apparently the main danger to which this species may be
exposed, apart from the decimating effect of severe winters, is
the extension of grazing by the cattle and sheep of the settlers,
and the exploitation of the guanacos for their hides and meat.
As to the former, a writer in an Argentine journal (La Chacra,
1936) states that in 1913 ranchers in Santa Cruz pleaded for
the destruction of the guanaco on the ground that it was a
detriment to sheep raising and a national plague. Wire fences
used to enclose ranges prove the ruin of many guanacos,
victims of cold and hunger. On the other hand, the animal
does not interfere with sheep raising in parts farther south that
are unsuitable for sheep. It is much persecuted for its hide

SOUTH AMERICA 409

and meat. In the report of the world's fur production for 1928
(Journ. Soc. Preserv. Fauna Empire, pt. 12, p. 64, 1930) the
number of guanaco skins brought to market is given as 300,000.
While at present it can hardly be said that the species is in
danger of extermination, nevertheless the demand for hides
to be used as robes, and the encroachment of grazing and
hunting, will doubtless much restrict its area and its numbers
in years to come.

SOUTHERN VICUNA
VICUGNA VICUGNA VICUGNA (Molina)

Camelus vicugna Molina, Saggio Storia Nat. Chile, vol. 1, p. 313, 1782 ("Probably
Peru" Lydekker; but Molina says in the Cordillera of the provinces Coquimbo
and Copiapo).

PERUVIAN VICUNA
VICUGNA VICUGNA MENSALIS (Thomas)

Lama vicugna mensalis Thomas, Smithsonian Misc. Coll., vol. 68, no. 4, p. 3, Apr. 10,

1917 ("Incapirra, Junin, Peru").
FIGS.: Tschudi, 1844, pi. 17 (col.); Royal Nat. Hist., vol. 2, p. 412, fig., 1894.

The vicuna resembles the guanaco in general form but is
about a fourth smaller, somewhat slenderer, and paler in color,
a pale fawn, without black on the face. The fore limbs have
no callosities such as are present on the inner side of the
"knees" of the guanaco. Although formerly placed in the
same genus with the latter, Miller (1924c) regards it worthy of
generic distinction, since the lower incisors are peculiar in
being long, slender, and ever-growing from persistent pulps,
like those of rodents, but with the enamel on the inner side.
This condition is unique among living artiodactyls but recalls
that of the dwarf wild goat (Myotragus) , the remains of which
are found in Pleistocene cavern deposits of the Balearic Islands.
Osgood (1916) gives the following measurements of a Peruvian
specimen: Length from between ears to root of tail, 1,250 mm.;
base of ear to point of shoulder, 670. The height at the shoul-
der, according to Lydekker, is about 2 feet 9 inches; Tschudi
(1844) gives 2 feet 6 inches. Length of skull, about 220 mm.

Although Lydekker says that the "typical locality is prob-
ably Peru," Molina's account, on which the name is based,
states that they are found in the parts of the Cordillera belong-
ing to the provinces Coquimbo and Copiapo, but keep to the
steep summits of the mountains, minding neither snow nor ice.

410 EXTINCT AND VANISHING MAMMALS

Whether or not the vicuna did occur in Molina's day, 1782, as
far south as the Cordillera of Coquimbo, in Chile, it apparently
no longer does so, but one may regard the type locality as the
latter region on Molina's authority. Northward the range
extends to northern and central Bolivia, Peru at high altitudes,
and southern Ecuador. Thomas has named, as a distinct race,
mensalis, the northern animal, basing his description on speci-
mens from Incapirra, Junin, Peru. The characters claimed are
the more strongly fulvous color and slightly smaller size and
smaller teeth, as compared with true V. vicugna, the type
locality and size of which are not indicated. The range of this
race is given as "Peru and Bolivia" and is assumed to include
also southern Ecuador at high altitudes. This would leave
typical V. vicugna as the form of Chile, where it is possibly now
extinct. The two races may be considered together. An adult
male measured from "between ears to root of tail," 1,250 mm.;
base of ear to point of shoulder, 670, or more than in the guana-
co. The skull of the type of mensalis, a male, had a greatest
length of 240 mm.; length of molars, 45.

The vicuna is apparently at the present time gone from
Chile, but the British Museum has a mounted skin from Cata-
marca, northwestern Argentina. In Bolivia it is found hi a
restricted area in the north-central part. In Peru, its chief
centers of abundance are said by Maccagno (1932) in his
recent monograph to be: Junin, Huanta, Ayacucho, Puno,
Cuzco, Apurimac, Huancavelica, and Arequipa.

These animals live on the vast plains at altitudes between
11,500 and 18,500 feet in Peru, going in small droves of 10 to
15 females and an adult male. In general they are said to be
easily tamed, and often one or two may be seen about ranch
houses, but in confinement they do not breed readily. On the
other hand, if allowed practical liberty under fence with
sufficient area, they are said to breed freely and may be caught
and sheared for their wool, which is much sought for its fine
quality.

In Peru the vicuna was accorded legal protection as long
ago as 1825, but the law apparently remained a dead letter
until the animals had become much reduced through hunting.
Finally, on October 8, 1920, a decree was passed prohibiting
the making of goods from vicuna wool and forbidding the sale
of skins. A heavy fine was provided for infractions. As a

SOUTH AMERICA

411

result it is said on good authority that in Peru the vicuna is
fast recuperating and is now locally common as in the Junin
area (William C. Burdett, in litt.). Nevertheless it is difficult
to enforce these laws and many of the animals are killed.
In Bolivia the exportation of vicuna wool and hides is pro-
hibited under laws of 1920 and 1922. Vicuna rugs are often
manufactured in Peru, but their export is forbidden except that
a person leaving the country may take one out by securing an
official permit. The robes are made of small strips of the hide
perfectly matched in color and texture. Several grades are
made by using the skin from different parts of the body, the
back and sides for one type, the neck for another, and the legs
and belly for a third. The most valuable are those made from
the back and sides, and some of the best may sell for as high as
$100 (Carriker). It is believed that it may eventually be

-"->

Vicuna (Vicugna vicugna)

412 EXTINCT AND VANISHING MAMMALS

feasible to raise vicunas on ranches and shear them for the
wool, which brings a good price, as much as $5 a pound.
This would require special sanction of the government con-
cerned. Like the llama and alpaca, this animal is subject to
such parasitic infections as lungworm, scab, and flukes, particu-
larly at lower altitudes (B. T. Colley, in Hit.).

In earlier days the native Indians made much use of this
species. Tschudi (1844) describes how they carried out ex-
tensive drives, gradually working the animals into a narrow
funnel-shaped place among the rock walls, where many were
captured, their wool sheared, and the captives then freed.

Family CERVIDAE: Deer
ECUADORIAN PUDU

PUDU (PUDELLA) MEPHISTOPHILES (de Winton)

Pudua mephistopheles de Winton, Proc. Zool. Soc. London, 1896, p. 508.
FIGS.: De Winton, 1896, pi. 19; Lydekker, 1898, pi. 24, fig. 1.

This small deer stands about 14 inches high at the shoulder
and is peculiar in its short metapodials, very short spikelike
antlers in the male, short ears, and practical absence of an
external tail. Very little is known of it, but since its habitat is
restricted and is likely to be further limited by various develop-
ments in the future, it may deserve mention here.

The general color is a rich brown, due to a blackish-brown
ground color, sprinkled with bright rufous. Ears relatively
short, with long white hairs lining them. Face and legs nearly
black.

Originally made known from a specimen taken on the paramo
of Papallacta, Ecuador, very few examples of this deer have
since found their way into collections. The Swedish consul
Soderstrom sent a specimen to the British Museum and one to
the Royal Museum at Stockholm, the latter with antlers about
78 mm. long. All these are from an altitude of about 12,000
feet at Papallacta, outside of which the species is unknown.
It is believed that even here it is uncommon.

SOUTH AMERICA 413

PERUVIAN GUEMAL; "TARUGA"

HIPPOCAMELUS ANTISIENSIS (d'Orbigny)

Cervus antisiensis d'Orbigny, Ann. Mus. Hist. Nat. Paris, vol. 3, p. 91, 1834 (Andes,

near La Paz, Bolivia).
SYNONYMS: Anomalocera huamel Gray, Scientific Opinion, p. 384, 1869; Xenelaphus

chilensis Gray, Ann. Mag. Nat. Hist., ser. 4, vol. 12, p. 61, 1873.
FIG.: Lydekker, 1898, pi. 23 (col.).

This small deer is confined to a rather restricted alpine
habitat in the Andes and may be included here as a game
animal that may need more protection. Somewhat smaller
than a Virginia deer, having a coarse brittle coat and lacking
metatarsal glands, this species is of a nearly uniform speckled
brown and buffy, with a darker line on the forehead. Tail dark
brown on base and most of the upper surface, but its tip and
lower side are white. The antlers are present in males only,
and consist of a short fork commencing close to the burr.
Height at shoulder about 34 inches (Lydekker).

In the Andes of Ecuador, Peru, Bolivia, and northern Chile
this species is found at high altitudes, mainly between 14,000
and 16,000 feet but at times lower. Lydekker in 1898 wrote
that it was abundant in Ecuador on Chimborazo, Pichincha,
and Cotapaxi, but Richardson in 1912-13, collecting for the
American Museum of Natural History, failed to obtain speci-
mens when he was in that region. In Peru, Tschudi gives some
account of its habits and speaks of it as frequenting rocky
areas, often hiding in caves by day, and coming out in the
evening to feed on mosses and other vegetation, and lichens.
Osgood (1914) in his journey to northern Peru in 1912 did not
find even a track of one and writes that "so far as learned from
inquiry, it never has been common in the region and it was only
at rare intervals that we met a man who ever had seen one.
A few doubtless remain in the higher parts of both the western
and the eastern cordillera but at the points we were able to
touch " none was found. The Peruvian Indians call it "taruga,"
and Tschudi mentions that he was to have named it Cervus
taruga but discovered that d'Orbigny had already described it.
It seems likely that the drives formerly held on a large scale by
the Peruvian Indians must often have captured these small
deer. On his later visit to the Arequipa region, Osgood (1916)
secured a single specimen at Pampa de Arrieros, Peru. It was
found up to an altitude of 13,000 feet, ranging somewhat lower

414 EXTINCT AND VANISHING MAMMALS

than the guanaco and vicuna. In Bolivia, Neveu-Lemaire
and Grandidier (1907) have recorded it from the Yuru district,
at 5,000 meters. On account of the limited distribution and
the hunting to which this deer is subject, it is evidently becom-
ing uncommon and requires adequate protection.

PAMPAS DEER; "VEADO BRANCO"; "GuAZim"; "VEADO

CAMPEIRO"
BLASTOCERUS BEZOARTICUS (Linnaeus)

Cervus bezoarticus Linnaeus, Syst. Nat., ed. 10, vol. 1, p. 175, 1758 (Brazil).
SYNONYMS: Cervus campestris F. Cuvier, Diet. Sci. Nat., vol. 7, p. 484, 1817; Cervus

azarae Wiegmann, Isis, col. 954, 1833 (Paraguay).
FIGS.: Lydekker, 1898, pi. 22 (col); 1915, p. 190.

The pampas deer is much smaller than its congener the
marsh deer, about the size of the European roebuck, but more
delicately built. The color is a light reddish brown, the face
darker, and occasionally a black patch on the crown; "tarsal
tuft, a patch at base of backs of ears, a ring round pedicles of
antlers, another round each eye, lips, throat, chest, under-
parts, fronts and inner sides of thighs, and inner sides of but-
tocks and upper part of fore-legs whitish; . . . tail dark
blackish brown above and white below." The antlers of the
male are small, with a large brow tine and a posterior beam
that forks usually once; length of beam up to 14.5 inches.
Males usually have the upper canine present.

The general range of this small deer includes the open
campos of Brazil, Paraguay, and Uruguay to the pampas of
Argentina and northern Patagonia (Lydekker, 1898). Possi-
bly the animals of the more southern part of the range may be
separable as a distinct subspecies for which, as pointed out by
Lydekker, the name azarae is available. It is said to inhabit
dry open plains, avoiding forests and thickets. "Formerly,
when the tussocks of tall pampas-grass were dotted more or
less thickly over all the plains, it had plenty of covert; but in
the more settled districts it now has to live almost completely
in the open, and has consequently become wary in the extreme"
(Lydekker, 1898). It may be found in small groups or pairs,
but the adult males are often solitary for most of the year.
In the evening they are active but during the daytime lie up in
concealment. They are said to have a strong and characteristic
odor. When started, they bound off at a considerable speed,

SOUTH AMERICA 415

so that a good horse is necessary to overtake one. The natives
sometimes capture them with the bolas. In parts of Brazil the
animal goes by the native name of "guazuti," and in Uruguay
it is known as "gama."

According to Lydekker (1901) it has completely disappeared
from many districts of Argentina and Uruguay. As long ago
as 1894, Aplin (1894) wrote that in the neighborhood of Santa
Elena, Uruguay, it had been exterminated except for a small
herd of about a dozen does and seven bucks preserved in a
certain district. He found it rare on the Rio Negro, but in
some other areas it was still common. Sanborn (1929) in
1926-27 found it plentiful in one locality in Rocha, Uruguay,
but in most other places rare, and this is corroborated by
Devincenzi (1935), who speaks of Rocha as the district where
it is now to be found, although 30 years before it was abundant
in the whole country.

Dr. Roberto Dabbene, in a letter to Dr. Francis Harper in
1937, stated that though formerly common in the northern
and central regions of Argentina as far as northern Patagonia,
it is now very scarce as far as the Chaco, and if not protected
adequately is certain to disappear from the Argentine fauna.
He deplores the introduction of exotic deer into the national
parks, rather than the encouragement of the native species.

While further and more precise information on the present
and past status of this species is much needed, it is evident
that it is much reduced in many districts and requires careful
protection. In Argentina hunting this species was prohibited
by presidential decree several years ago.

MARSH DEER; "VEADO GALHEIRO GRANDE"; "GuAzupuco"
BLASTOCERUS DICHOTOMUS (Illiger)

Cervus dichotomies Illiger, Abh. Akad. Wiss. Berlin, for 1811, pp. 108, 117, 1815

(Brazil).
SYNONYMS: Cervus paludosus Desmarest, Mammalogie, pt. 2, p. 443, 1822; Cervus

palustris Desmoulins, Diet. Classique Hist. Nat., vol. 3, p. 379, 1823.
FIGS.: Lydekker, 1898, pp. 284, 285 (antlers, animals).

This is the largest South American deer, attaining about the
size of the British red deer, but more slenderly built, standing
about 46 inches at the shoulder (Azara). "General colour in
summer bright rufous chestnut, in winter brownish red, becom-
ing lighter on flanks, neck, and chest; legs black from knees

416 EXTINCT AND VANISHING MAMMALS

and hocks downwards, and tarsal tuft also black; abdomen,
inside of thighs, chin, and insides and bases of backs of ears
white ... a whitish line above, or a ring round eyes,
most marked in females . . . tail yellowish rusty red
above and black beneath . . . Fine antlers attain a length
of from 21 to 24^ inches" (Lydekker, 1915); they are doubly
forked, each of the two branches having a simple fork.

The range is extensive in the tropics and subtropics of South
America, from probably Guiana southward through Brazil to
Paraguay and Uruguay and the Chaco or wooded districts of
northern Argentina. Unlike its smaller relative, the pampas
deer, it inhabits dense jungle on the borders of streams or
swamps and it is said to go in small parties of three to five
individuals. The skin is much used for leather by the natives,
but the meat is apparently not held in high esteem. Lydekker
remarks on the similarity in color between this deer and the
maned wolf, which is found in the same areas in parts of the
range.

Although little information is at hand as to the details of
distribution and relative abundance of this large deer, it is
evident that in certain of the more settled regions in the south-
ern part of the range it is becoming much reduced in numbers.
Devincenzi (1935) writes that in Uruguay the Departments of
Rocha and Treinta y Tres have been "considered the last
refuge of the species/' while Sanborn comments that it is now
very rare in Uruguay but was reported to be found in small
numbers in Rocha. No effort has been made to preserve them.
Dr. Roberto Dabbene writes, in response to inquiry by Dr.
Harper, that in Argentina it was formerly common and oc-
curred as far as the islands of the Delta of La Plata but is now
(1937) confined to the Territory of Formosa, where it is not
common. By presidential decree the hunting of this deer is
prohibited in Argentina.

Though no data of importance are available for Brazil, it is
apparently to be found in fair numbers.

OCEANIC MAMMALS

Order CARNIVORA: Dogs, Cats, and Their Relatives

E carnivores, considered as an order on p. 134, are repre-
- sen ted by the sea otter (family Mustelidae) of the northern
Pacific Ocean. Two races of this species are recognized; both
were brought to the verge of extinction because of their
valuable fur, but now they are recovering in numbers and
may be an important resource in the future.

Family MUSTELIDAE: Weasels, Martens, Otters

NORTHERN SEA OTTER
ENHYDRA LUTRIS LUTRIS (Linnaeus)

[Mustela] lutris Linnaeus, Systema Naturae,, ed. 10, vol. 1, p. 45, 1758 (Kamchatka).
SYNONYM: Lutra marina Schreber, Saugthiere, vol. 3, p. 465, pi. 128, 1778 (ex Steller) ;
Nov. Comment. Petropol., vol. 2, p. 367, pi. 26, 1751.

SOUTHERN SEA OTTER
ENHYDRA LUTRIS NEREIS (Merriam)

Latax lutris nereis Merriam, Proc. Biol. Soc. Washington, vol. 17, p. 159, Oct. 6, 1904

("San Miguel Island, Santa Barbara Islands, California"),
FIGS.: Royal Nat. Hist., vol. 2, p. 98, fig., 1894; zur Strassen, 1914, pp. 12*. 13*;

Nelson, 1916, col. fig. p. 434.

The distinction between the northern sea otter and the
southern race rests apparently upon slight cranial differences;
hence it is not possible to distinguish the two without recourse
to the skulls. Accordingly they may be treated together as
representing but a single species, while the limits of geographi-
cal and individual variation still remain to be more precisely
defined. It is believed that the southern race is the one
occurring north at least to the coast of Oregon.

The sea otter is of rather heavy, robust form, about 4 feet in
length, of which the tail is about a foot. The fore feet are small,
with naked palms, but the hind feet are long and broad,
webbed and with furry soles, recalling the flippers of a seal.

417

418 EXTINCT AND VANISHING MAMMALS

The color is dark brown, becoming hoary on the head. The
sparse whiskers are stout and short. The posterior teeth are
remarkably enlarged, broadened, and their cusps blunted to
form low rounded knobs suitable for crushing the shellfish that
form their diet. Dr. Merriam states that the skin of the type
of the southern race was 6 feet long, but it may have been
stretched, or perhaps a very large individual. Barabash-
Nikiforov (1935) gives a maximum length of 1,635 mm., of
which the tail was 330, and a maximum weight of 35 kilograms.

Sea otters were abundant formerly from the coasts of south-
ern Kamchatka to the Kurile Islands in the western North
Pacific and in the waters about the islands of the Bering Sea
and Alaskan coasts southward following the cooler currents
even to the coasts of southern California. On the Asiatic side
they ranged at one time as far south as Yezo. On account of
the richness of its fur the sea otter was ruthlessly pursued
since its discovery by the Russians in about the middle of the
eighteenth century, down to more modern times, until by the
first decade of the present century it was very nearly exter-
minated. The few skins that then came on the market sold for
as much as $1,000 apiece. Finally the United States Govern-
ment in 1910 passed a law forbidding its capture in American
waters and negotiated treaties with other interested nations
for giving similar protection. Now, after a lapse of little over
a quarter of a century, there is encouraging evidence that the
species is recovering and it has lately appeared in some num-
bers off the California coasts.

The history of the pursuit and exploitation of the sea otter
has at various times been written. From some of these ac-
counts the following pertinent facts are gleaned. Active
trade in sea-otter skins seems to have begun in 1742, when
Bering, after being wrecked in the sea bearing his name,
returned to Petropaulovsk with about 900 skins stowed in the
small boat built from the wreck of his vessel, the St. Peter.
These at first were chiefly traded with the Chinese, by whom
they were highly valued. Shelikof, a Russian trader, at once
saw great possibilities in further trade in this fur. He founded
a colony on Kodiak Island and made plans for collecting sea-
otter skins on a large scale. He died before his object was
accomplished, but his son-in-law, Nicholas P. Rezanof, carried
on the work and in 1799 obtained from Emperor Paul the

OCEANIC MAMMALS 419

charter of the Russian-American Co. Shortly after, under the
management of Shelikof and Baranof, the work was organized,
and a fleet of bidarkas, manned by native Aleuts, started an
intensive campaign. H. W. Elliott (1875) writes: "During
the first few years the numbers of these animals taken all along
the Aleutian chain, and down the whole northwest coast as
far as Oregon, were very great ; for instance, when

the Prybilov Islands were first discovered, two sailors, Lukan-
non and Kaiekov, killed at St. Paul's Island, in the first year
of occupation, five thousand, the next year they got less than a
thousand, and in six years after not a single sea-otter appeared,
and none have appeared since. When Shellikov's party first
visited Cook's Inlet, they secured three thousand; during the
second year, two thousand; in the third, only eight hundred;
the season following they obtained six hundred; and finally, in
1812, less than a hundred, and since then not a tenth of that
number. The first visit made by the Russians to the Gulf of
Yahkutat, in 1794, two thousand sea-otters were taken, but
they diminished so rapidly that in 1799 less than three hundred
were taken. In 1798 a large party of Russians and Aleuts
captured in Sitka Sound and "^neighborhood twelve hundred
skins, besides those for which they traded with the natives
there, fully as many more; and in the spring of 1800 a few
American and English vessels came into Sitka Sound, and
anchored off the small Russian settlement there, and traded
with the natives for over two thousand skins, getting the
trade of the Indians by giving fire-arms and powder, ball, &c.,
which the Russians did not dare to do, living then, as they were,
in the country. In one of the early years of the Russian-
American Company, 1804, Baranov went to the Okotsk from
Alaska with fifteen thousand sea-otter skins, that were worth
as much then as they are now [1875], viz., fully $1,000,000."
These tremendous inroads very quickly had an alarming
effect. A Russian report, quoted by Elliott, tells that in 1826
the total number secured in the entire Aleutian chain was only
15, where previously more than 1,000 had been regularly taken
each year. In 1835 the number from this district was 70 to
150 annually. The natives employed in collecting these furs
were frequently subjected to great dangers, not only from the
elements but from other native tribes who were unfriendly.
When Alaska was purchased by the United States, writes H.

420 EXTINCT AND VANISHING MAMMALS

W. Elliott (1875), "the Russians were taking between four
and five hundred sea-otters from the Aleutian Islands and
south of the peninsula of Alaska, with perhaps a hundred and
fifty more from Kenai, Yahkutat, and the Sitkan district; the
Hudson's Bay Company and other traders getting about two
hundred more from the coast of Queen Charlotte's and Van-
couver's Islands, and off Gray's Harbor, Washington Territory.
Now, during the last season, 1873, instead of less than seven
hundred skins, as obtained by the Russians, our traders secured
not much less than four thousand skins. This immense differ-
ence is not due to the fact of there being a proportionate in-
crease of sea-otters, but to the organization of hunting parties
in the same spirit and fashion, as in the early days . . .
The keen competition of our traders will ruin the business in
a comparatively short time if some action is not taken by the
Government.

"Over two-thirds of all the sea-otters taken in Alaska are
secured in two small areas of water, little rocky islets and reefs
around the island of Saanach and the Chernobours, which
proves that these animals, in spite of the incessant hunting
all the year round on this ground, seem to have some particular
preference for it to the practical exclusion of nearly all the rest
of the coast in the Territory. This may be due to its better
adaptation as a breeding ground. It is also noteworthy that
all the sea-otters taken below the Straits of Fuca are shot by
the Indians and white hunters off the beach in the surf at
Gray's Harbor, a stretch of less than twenty miles; here some
fifty to a hundred are taken every year, while not half that
number can be obtained from all the rest of the Washington
and Oregon coast-line; there is nothing in the external appear-
ance of this reach to cause its selection by the sea-otters, except
perhaps that it may be a little less rocky.

"As matters are now conducted by the hunting-parties, the
sea-otters at Saanach and Chernobours do not have a day's
rest during the whole year. Parties relieve each other in
succession, and a continuous warfare is maintained. . . So
the bad work goes on rapidly, though a majority of the natives
and traders deprecate it." The optimum region mentioned by
Elliott is described as a chain of small islets, most of them bare
at low tide but with numerous reefs and rocky shoals with beds
of kelp. "As the natives have never caught the mothers

OCEANIC MAMMALS 421

bringing forth their offspring on the rocks, they are disposed to
believe that the birth takes place on kelp-beds, in pleasant or
not over-rough weather. The female has a single pup [rarely
two], born about fifteen inches in length. . . The sea-otter
mother sleeps in the water on her back, with her young clasped
between her fore-paws."

The methods employed in taking sea otters Elliott describes
as four: Shooting them in the surf at long range with a rifle,
and waiting till they drift ashore if the surf is too rough to
permit of launching a boat; surrounding an otter by a party
of spearers in their boats and awaiting its return to the surface
after it becomes exhausted in several dives; clubbing them in
winter when they may at times be stealthily approached among
rocks and kelp; and using nets 16 to 18 feet long and 6 to 10
feet wide, of coarse meshes, spread out on the kelp beds.
Frequently several at a time are thus taken, for when enmeshed
they seem to make little or no attempt to get free. It is said
that this method is preferred by the Japanese, since it permits
the release of inferior animals or breeding females. The young,
according to Scammon (1874), are met with at all times of the
year, so that there appears to 1be no definite breeding season ;
the period of gestation is believed to be eight or nine months.

Writing in 1874, Captain Scammon speaks of the Lower
California coasts as the haunt of sea otters and adds that
Cerros, San Geronimo, Guadalupe, San Nicolas, and San
Miguel Islands were "regarded as choice places to pursue
them." Earlier, when California was part of Mexico, the
pursuit of sea otters was prohibited by that Government under
severe penalty. In recent years bones of the sea otter have
been found in Indian shell heaps on Santa Cruz Island and
near Monterey (E. M. Fisher, 1930).

Previous to the purchase of Alaska from Russia, Wrangell
had already instituted somewhat more far-sighted methods of
making the annual catch, allowing no part of the hunting
grounds to be used for two consecutive years, and thus some-
what restricting the number killed. From 1842 to 1862 the
average catch, including that of the Kurile Islands, was about
1,249, and the total for the 20 years was 25,899. The result
was to restore the sea-otter population to a slightly better con-
dition. But these careful methods were abandoned when
Alaska became part of the United States, and the pursuit was

422 EXTINCT AND VANISHING MAMMALS

followed with the greatest intensity, so that from 1881 to 1890
the take was 47,842, or nearly 5,000 a year (C. L. Andrews,
1937). The immediate reduction of the breeding stock by
thus overexploiting these animals caused the Alaska Commer-
cial Co., which had practically secured control of the Alaskan
fur production, to close four or five of its posts by 1897. In
1900 the company was able to secure but 127 skins, and ten
years later the entire catch of its fleet of 16 schooners was but
31 skins. Even at the price of $1,000 a skin there was no
profit in this undertaking. Fortunately, in 1910, the taking of
sea otters by citizens of the United States was prohibited, and
by treaties of similar import with other interested nations the
species was given respite just in time.

The gradual recuperation of the sea otter since 1910 from
the verge of extinction to appreciable numbers during the last
30 years since protective laws were established has been most
encouraging, although relatively meager data only are at hand
as yet. In the Commander Islands their principal habitat is
the rocky Copper Island, in the northwestern part of the group.
Here from 1930 to 1932 the Russian naturalist I. Barabash-
Nikiforov (1935) spent much time in an extensive study of
their biology, and he concludes that the entire "herd" there
then numbered between 600 and 700 animals and that the
yearly increment was about 7 percent. The animals seem to be
slowly building up a considerable population under careful
protection. At about the same time Eyerdam (1933) published
a brief report on his observations in the Aleutian Islands,
where in a few localities "they seem to be on the increase."
From fishermen and natives he learned that they are now
frequently seen along the coasts of Afognak Island, and he
himself in 1922 saw several during his stay there. In July,
1932, while staying on Atka Island, he learned that they are
now becoming fairly common in the western Aleutians, adding
that "among the dangerous, windswept foggy islets and reefs
between Atka and Adak Islands, more properly known as the
Sitkin Islands, sea otters can nearly always be seen if the
weather permits." A captain of the Bureau of Fisheries
vessel Crane, that summer, reported that "he had counted up-
wards of 40 sea otters between Adak and Atka Islands although
the weather was unusually stormy at that time." Eyerdam
also reported that sea otters seem to be increasing at the Sanakh

OCEANIC MAMMALS 423

Islands, formerly a favorite resort for them; the residents
there see them not infrequently. There is said, however, to be
a small amount of poaching by Japanese vessels, for much of
these coasts are uninhabited and can not readily be patrolled.
With regard to the sea otter on the coast of California
Grinnell, writing in 1933, observed that, though formerly
abundant about the islands and open seashore the whole length
of the State, it was then rare but that nevertheless "individuals
have been reported as seen almost every year off coasts of
Monterey and San Luis Obispo counties. Last actual speci-
men" taken from near Monterey, September 9, 1915. On the
Oregon coast the species was apparently exterminated half a
century ago, for the last report given by Bailey (1936) is for
1876. Very recently a considerable herd appeared off Mon-
terey, Calif., and formed the subject of interesting and valuable
studies by Edna M. Fisher (1939). The animals were first
noticed on March 19, 1938, off the mouth of Bixby Creek and
attracted much attention. After about two weeks they moved
slightly farther northward but remained in the general vicinity
at least into early September. A careful count and estimate
placed the number in this gtoup as slightly less than 100,
about 94 or fewer having been made out on various occasions.
Miss Fisher's studies and sketches add greatly to a knowledge
of the habits of the animal. The food secured by the otters
was found to be red abalones, sea-urchins, and crabs; occa-
sionally they nibble at the kelp. No evidence of their feeding
on mussels was observed. These notes corroborate those of

Sea otter (Enhydra lutris)

424 EXTINCT AND VANISHING MAMMALS

Barabash-Nikiforov, who analyzed many droppings and found
that in the Kurile Islands remains of sea-urchins constituted
about 59 percent of their content, mollusks 23 percent, crabs
about 10 percent, fish about 7 percent, and seaweeds a trace.
Further details may be found in a paper by Murie (1940).
Aside from man, their chief natural enemy is probably the
killer whale, small groups of which haunt these seas and
especially those about the more northern fur-seal rookeries.

While this recovery in numbers is encouraging, it must be
evident that the building up of a considerable population will
be slow. At the present time the Kurile herd must number
nearly 700 or more, while information supplied by the Alaska
Commercial Co. of San Francisco states that reports of sub-
stantial numbers have come to them from Atchitka and Rat
Islands of the Aleutians; other estimates of the sea-otter popu-
lation in the western part of this chain place the numbers at
several hundreds. From southeastern Alaska there have been
as yet no reports. That the numbers in the Kurile Islands are
now sufficient to warrant the taking of a certain proportion for
commercial purposes is indicated by the fact that according to
the Fur Trade Review Weekly of May 4, 1939, 50 sea-otter
skins were sold at recent London fur sales. "At about the
same time The New York Auction Company offered 22 and
the Fouke Fur Company of St. Louis, 10 pelts of this animal."
Investigation has shown that all these were legally killed in the
Kurile Islands by the Japanese Government, and "were certi-
fied in accordance with the Act of August 24, 1912" (Journ.
Mammalogy, vol. 20, p. 407, 1939). If the sea otter continues
to increase it may eventually with proper management be
made a productive source of profit once more.

Order PINNIPEDIA: Seals, Sea-lions, Walruses

The sea-lions, seals, and walruses are related to the carni-
vores, differing from the latter group in the finlike structure of
the feet. There are three families :

(1) Otariidae, the eared seals, sea-lions, and fur seals, or sea-
bears. Representatives of this family occur on the temperate,
subarctic, and austral shores of the Pacific, southern Atlantic,
Indian, and northern Antarctic Oceans. Six species of the
southern fur seal and three forms of the northern genus are dis-
cussed here.

OCEANIC MAMMALS 425

(2) Phocidae, true seals or hair seals. These are more highly
specialized for aquatic life than the sea-lions. The seals occur
along the shores of all seas and oceans, and certain forms are
found in the Caspian Sea and in Lake Baikal in central Asia.
Seven species are threatened with extinction.

(3) Odobenidae, walruses. Two species of the single Recent
genus are recognized; both are endangered by hunting activi-
ties, although they are still represented by large herds in the
Arctic Ocean. J. E. H.

Family OTARIIDAE: Sea-lions, Fur Seals

SOUTHERN FUR SEAL
ARCTOCEPHALUS AUSTRALIS (Zimmermann)

Phoca australis Zimmermann, Geographische Geschichte, vol. 3, p. 276, 1782 (Falk-
land Islands, based on Pennant's account).

SYNONYMS: For full synonymy see J. A. Allen, 1905, who gives among important
synonyms the following: Phoca falklandica Shaw, General Zool., vol. 1, pt. 2, p.
256, 1800; Arctocephalus ursinus Gray, List Spec. Mamm. Brit. Mus., p. 103,
1843; Arctocephalus nigrescens Gray, Zool. Voy. Erebus and Terror, and Proc.
Zool. Soc. London, 1859, pp. 109, 360:

FIGS.: Townsend, in Jordan et al., 1899, vol. 3, pi. 35; Allen, J. A., 1905, pi. 15, fig. 1;
pi. 16, fig. 2; pi. 17, fig. 2 (skull).

The fur seals of the genus Arctocephalus are allied to the
North Pacific species of the genus Callorhinus but are dis-
tinguished by having the facial portion of the skull slender,
narrow, and elongated, with the upper profile sloping instead
of nearly flat, the molars larger. The several isolated groups
are in some cases regarded as separate species, but the differ-
ences are based largely on skull characters. All have been
much hunted for their pelts and are largely depleted or even
on the verge of extinction in parts of their range.

The pelage is of two kinds a long, coarse, blackish overhair,
tipped with gray or yellowish gray, giving a grizzled effect,
except on the lower surface; and a thick, soft, brownish under-
fur, lighter at the tips and darker basally. Skull short and
broad, the brain case rather squarish, the antorbital region
very short, with short nasals; sagittal crest slightly developed.
Unworn teeth tricuspid, with a main cusp and small anterior
and posterior cusps. Teeth relatively much smaller than in
A. philippii. Basal length of skull, 235 mm. ; zygomatic width,

426 EXTINCT AND VANISHING MAMMALS

to 148 mm. Like other eared seals, the neck is relatively long,
the fore limbs longer than the hind, which are capable of
turning forward.

According to J. A. Allen (1905), from whose account the
foregoing details are taken, no adequate description of the
external characters of this fur seal has as yet been published.
Its breeding places "formerly included the Falkland Islands,
New Year's Island, Staten Island, Desolation Islands, and
other islands and coasts off the southern portion of South
America, and probably the more southern South Shetland,
South Georgian, and Sandwich groups. They doubtless still
resort to most of these localities, but only in small numbers in
comparison with their former abundance" (J. A. Allen, 1905).
It ranged northward along the Patagonian coast to the mouth
of the La Plata, frequenting the small islands off Maldonado,
Uruguay, and in 1887 was known to straggle as far north as
Rio de Janeiro, doubtless folio wing the cooler northward-flowing
currents. The Straits of Magellan were a favorite resort,
while on the west coast of Chile it ranged at least to the south-
ern part of the archipelago.

The status of the fur seal about the southern part of South
America is somewhat difficult to make out definitely, but it
still occurs in some numbers locally. Dr. Barnum Brown (in
manuscript notes published by J. A. Allen, 1905), at the be-
ginning of this century saw "considerable numbers on the
south coast of Tierra del Fuego, but they were not observed
off the coast of Patagonia. One herd estimated to contain
1500 head, was seen near Cape Hall, west of the Strait of Le
Maire, and two smaller herds were seen south of Lenox Island,
having less than 200 individuals each. These seals are poached
by a few natives, but owing to the abrupt, rugged rocks they
are seldom found on shore and cannot be driven to a killing
ground. The Argentine Government sends a gunboat to these
waters once a month to keep off poachers."

The rugged shores of the Magellan region probably now
offer a last stronghold for this fur seal, where the very nature
of the seas and shores affords it partial security against human
predation. Elsewhere, however, the species is largely gone.
Dr. Robert Cushman Murphy (1918) has given a summary of
the history of its pursuit. "Sealing on the coast of Patagonia,
the Falklands, and the islands north and east of Cape Horn

OCEANIC MAMMALS 427

began during the third quarter of the eighteenth century.
Alexander Dairy mple, writing in 1775, reports that there was
at the Falkland Islands an abundance of ... fur seals in
'such numbers that they killed eight or nine hundred in a day
with bludgeons on one small Islot.' ' : At present fur seals are
no longer found on the Falklands, but the date of their disap-
pearance can not have been very long after the beginning of
the last century, for already soon after the American Revolu-
tion New England and British sealers were pushing their search
for pelts to South Georgia, and then to the South Orkneys and
the South Shetlands. The Portuguese were also according to
Forster engaged in this pursuit in its earlier years. About
1800, sealing at South Georgia had already about reached its
peak. In that year, the Aspasia, of New York, one of 18
sealing vessels at the island, "secured the season's prize catch
of 57,000 fur seal skins. This record was never again equalled,
although the hunting evidently continued, for when . . .
Bellingshausen sailed along the blustery, uncharted south
coast of the island in December, 1819, he met with two English
three-masters in one of the fjords. These ships had already
been there four months, or through the southern winter, and
had carried on a profitable business." Five years later, in
1824, when Capt. James Weddell visited the island, he found
that seals of all kinds had become almost extinct. He men-
tions that the American sealers traded these skins to China
where they frequently brought $5 or $6 apiece. He estimates
that not less than 1,200,000 of the hides had been gathered
there since the opening of the sealing exploitation. Fur seals
are believed to have been practically exterminated at South
Georgia about 1874, "but rumor has it that a New England
vessel made a small illegal catch there in 1907." In 1874 no
less than 1,450 skins were taken, and in the year following 600.
In 1892, according to Lonnberg (1906), 135 fur seals were
killed at this island "and they may have been the very last
ones," for in 1905 "a Chilenian sailing vessel visited the coasts
of this island," hunting in every cove and corner for fur seals
but found none. However, Dr. Murphy records that about the
middle of February, 1915, some Norwegian whalers discovered
a single fur seal, evidently a straggler, on the beach near the
eastern end of South Georgia, and unfortunately killed it.
With this exception no fur seal has been definitely known from

428 EXTINCT AND VANISHING MAMMALS

the island since 1910, when the first sealing license was issued
by the Falklands.

With the reduction of the fur seals on the Falklands and
South Georgia, the hunt was pushed still farther south to the
South Shetlands. Weddell writes of the success of his and
other crews here: "The quantity of seals taken off these islands,
by vessels from different parts, during the years 1821 and 1822,
may be computed at 320,000 . . . This valuable animal,
the fur seal, might . . . have been spared to render
annually 100,000 furs for many years to come. This would
have followed from not killing the mothers till the young were
able to take the water; and even then, only those which ap-
peared to be old, together with a proportion of the males,
thereby diminishing their total number, but in slow progres-
sion." Shortly after Weddell's time, the pursuit of fur seals in
this region was nearly abandoned, since the seals had been
largely extirpated. Fifty years later, however, the remnant
seems to have somewhat recovered, for between 1871 and 1891
at least 18,000 skins were taken from various parts of the
group. "From this second catastrophe they were never
allowed to recover, and their final extermination is believed to
have taken place in the opening years of the present century.
The last authentic capture occurred in 1902, when the Swedish
expedition found a single fur seal on Nelson Island. Since then
. . . none has been recorded from the group " (Marr, 1935).

The South Orkneys, lying to the northeast of the South
Shetlands, seem to have yielded relatively fewer fur seals.
Only three were reported by Weddell in 1823, but Dallmann
50 years afterward reported 165 seals, and he is the last to
record the seal's presence there, "but it is by no means certain
that it was he who exterminated it there," for these islands
seem a less favorable place since they are surrounded by pack
ice for a much longer period each year than are the South
Shetlands (Marr, 1935, pp. 370-373). On the South Sandwich
group, still more to the northeast, Rudmose Brown (1913)
supposes this seal may still rarely occur for "it certainly was
found there in comparatively recent years."

While thus the southern fur seal is believed to be gone from
the island groups where it formerly occurred within the Falk-
land dependencies, a few probably remain in the Straits of
Magellan region, and it may survive under some protection

OCEANIC MAMMALS 429

for a while longer. Dr. Roberto Dabbene in a letter to Dr.
Francis Harper dated January, 1937, writes that it has almost
vanished from the Argentine coasts because of constant perse-
cution for its fur. Thirty years before, when he visited Punta
Arenas, a single skin brought a pound sterling. He saw but
two individuals during a voyage through the Fuegian canals.
It has disappeared from "La Isla de los Estados" (Staten
Island) and the Patagonian coast.

In striking contrast to this depletion, however, is the wise
policy of Uruguay in protection and "farming" the rookeries
of these fur seals on Lobos Island off its coast near Maldonado.
This island is less than a mile in length, low and brush-covered,
but with rocky shores. A few houses are in its center. The
fur seals here have for many years been carefully protected
and managed under governmental supervision so that they
yield an annual return. "Commercial sealing was carried on
here prior to 1820. The present lessees of the island, operating
under the direction of the Government of Uruguay, placed
upon the London market, from 1873 to 1897, 319,746 salted
skins, or an average of over 13,000 a year" (C. H. Townsend,
in Jordan and others, 1899). Here is an excellent example that
should be followed by other nations of these latitudes, for
under proper management and protection the fur-seal colonies
might be made to yield a good return for years to come.

NEW ZEALAND FUR SEAL
ARCTOCEPHALTJS FORSTERI (Lesson)

Otaria forsteri Lesson, Diet. Classique Hist. Nat., vol. 13, p. 421, 1828 (New Zealand).
SYNONYM: Gypsophoca tropicalis Gray, Proc. Zool. Soc. London, 1872b, p. 659.
FIGS.: Gray, 1872, figs. 5, 6; 1874, pi. 18.

SOUTH AUSTRALIAN FUR SEAL
ARCTOCEPHALUS DORIFERUS Wood Jones

Arctocephalus doriferus Wood Jones, Rec. South Australian Mus., vol. 3, no. 1, 1925
("South Australia").

FIGS.: LeSouef. Burrell, and Troughton, 1926, figs. 12, 13, opposite p. 101 (photo-
graphs); Wood Jones, 1925. figs. 246-248 (skull and feet).

TASMANIAN FUR SEAL
ARCTOCEPHALUS TASMANICUS Scott and Lord

Arctocephalus tasmanicus Scott and Lord, Papers and Proc. Roy. Soc. Tasmania for
1925, p. 189, 1926 ("Tasmania").

430 EXTINCT AND VANISHING MAMMALS

While J. A. Allen (1905) regarded the fur seals of the Aus-
tralian and New Zealand seas as representing but a single form,
A. forsteri, those of South Australia and of Tasmania have
lately been given distinctive names. Whether these are
really separate races or species, or whether the characters
represent merely individual variations, seems as yet to be
uncertain. According to Wood Jones, A. forsteri has the
dorsal surface dark brown, grizzled with white-tipped hairs,
and the ventral surface reddish brown, with a fine red-brown
underfur. A distinguishing feature is said to be that the hind
foot has the prolongation of the middle three digits only
slightly shorter than the marginal digits, and the teeth have an
accessory cusp on only the front edge instead of on both fore
and hind margins as in the South Australian animal, which he
describes as A. doriferus. In the Tasmanian fur seal the large
size is apparently a supposed distinction; condylobasal length
of skull 280-290 mm. against about 250 in A. doriferus for
adult males. The three may be briefly considered as to status,
for although now more or less protected, they are reduced to a
small part of their former numbers; and while supposed to be
distinct, are not certainly identifiable when alive so that the
applicability of notes is uncertain.

Hutton and Drummond (1923) have given a brief history of
the exploitation of the fur seal in New Zealand. When dis-
covered by Capt. James Cook on his first voyage, he noted
them on the south island and on his second voyage found the
fresh meat of seals killed at Dusky Bay a welcome addition to
the larder. It was not until 1792, however, that the first
sealing crew landed there and in the course of nearly a year at
Dusky Bay procured over 4,500 fur-seal skins. In the opening
years of the nineteenth century, with the decline of the sealing
in Bass Straits, a schooner from Sydney, New South Wales,
traded a cargo of 2,000 sealskins at Dusky Bay, Breaksea
Sounds, and Solander Island. Sir Joseph Banks, who had
accompanied Cook on his first voyage, was much impressed by
the numbers of fur seals: "The beach is encumbered with
their quantities, and those who visit their haunts have less
trouble in killing them than the servants of the victualling
office have, who kill hogs in a pen with a mallet." In 1806 he
added that on one vessel bound to London from Sydney there
were 30,000 sealskins! The result of this reckless slaughter

OCEANIC MAMMALS 431

was that as early as 1810 the numbers were seriously depleted,
but in that year the discovery of the Campbell and Macquarie
Islands "gave new life to the trade." Nevertheless the search
for seals was carried on relentlessly, and in 1824 ten vessels
were said to have secured 70,000 to 80,000 on New Zealand and
the adjacent islands. Two years later a vessel spent six months
cruising for new sealing grounds, but obtained only 449 skins.
Stewart Island was a specially favored spot.

In former times fur seals were abundant on the Tasmanian
coasts, in Bass Straits, and along the islands in the bight of
South Australia. Peron saw them at Kangaroo Island in 1802-3.
The breeding season was from about November to April.
Sealing continued in these regions until a comparatively
recent date. Wood Jones, writing in 1925, states that the last
large haul of skins from Kangaroo Island and its outlying rocks
was made "nearly forty years ago [i. e. about 1885], and it is
very doubtful if a fur seal has been seen on the Island since
then." On six visits to the islands of Nuyts Archipelago and
the Investigator group he had not seen a single one, although
he was told of recent sealing there. He adds that "it is prac-
tically certain that the animals^still live on the inner Casuarina
Island, which was one of their great strongholds in the old
days, and from which only 13 years ago [1912] 20 fur seals
were said to have been taken." This island is now a sanctuary.
The fur seals were almost exterminated from Tasmania, "but
protection there has led to the re-establishment of some small
herds." It is said that on the coast of Western Australia it
still breeds in small numbers in the extreme southwest, off
Cape Leeuwin, and perhaps rarely on the Houtmans Abrolhos,
and that stray individuals have been seen as far north as
Shark's Bay (Shortridge, 1936, p. 745). While spending two
weeks on the Abrolhos in November, 1931, however, I saw no
sign of fur seals. It is said that in 1921 an expedition to the
Recherche Archipelago gathered 300 skins of a fur seal but
realized very little on them. They are not protected here and
no permit is required for taking them. "Doubtless on the out-
lying islands, which are very difficult of access, fur seals may
still be found, but unless strict measures are taken for their
protection they will soon disappear" (Hull, in Hanna, 1922, p.
14). It seems a pity that these remnants should not be pro-
tected and allowed to increase until the time when they can
be successfully maintained as a useful asset.

432 EXTINCT AND VANISHING MAMMALS

KERGUELEN FUR SEAL
ARCTOCEPHALUS GAZELLA (Peters)

Arctophoca gazetta Peters, Monatsb. Preuss. Akad. Wiss. Berlin, 1875, p. 396 (Kergue-

len Island, southern Indian Ocean).
SYNONYM: Arctophoca elegans Peters, Monatsb. Preuss. Akad. Wiss. Berlin, 1876, p.

316 (St. Paul and Amsterdam Islands, Indian Ocean).
FIGS.: Turner, 1888, pi. 6, figs. 4, 6 (skull).

The fur seal of Kerguelen in the southern Indian Ocean was
described by Peters on the basis of a specimen brought back
by the exploring ship Gazelle in 1874. The back, sides of head,
neck, and crown are grayish brown, this color extending for-
ward in a triangle between the eyes and on the sides of the
neck to in front of the ears. Below, the fore neck and upper
chest are pale yellow, becoming rusty brown posteriorly. Lips
rusty red and vibrissae white. The underfur is rusty red. The
specially distinctive characters lie in the skull, which has very
small auditory bullae. Turner (1888) speaks of the skeleton
as slenderer than in the South American species, with nasals
on the same plane as the top of the skull, and the cheek teeth
without secondary cusps but with a weak cingulum. The sixth
postcanine tooth is much smaller than the others. The adult
male measured in total length, 1,770 mm.; ear, 40; hind foot,
305. The female showed a total length of 830 mm.; tail, 35;
ear, 30. Greatest length of skull, 212 mm.; zygomatic width,
131; length of palate, 92.

Kerguelen or Desolation Island, lying in latitude 49 S., in
the southern Indian Ocean, has long been known as a resort
especially of the elephant seal, but it maintained also small
numbers of fur seals that apparently differed slightly in
cranial characters from other isolated groups in the southern
hemisphere. Peters emphasizes particularly the small bullae
and lack of accessory cusps on the postcanine teeth. The seal
found on the St. Paul and Amsterdam Islands, not very far to
the north of Kerguelen, Peters later distinguished as Arcto-
phoca elegans, a name to which Trouessart (1904-5) gave sub-
specific status, but Dr. J. A. Allen (1905, p. 122) believed that
the supposed differences were insufficiently defined to recognize
more than the one form among these, and probably on the
Crozet Islands as well as on Kerguelen, which seems a reason-
able conclusion, for the colonies are not so far apart that they
might be expected to develop peculiarities through isolation.

OCEANIC MAMMALS 433

Apparently the fur-seal population of Kerguelen has not
been large during recent times. As early as 1830 vessels in
pursuit of elephant seals visited it and doubtless obtained a
few fur seals as well. In 1874 there were four expeditions
there to observe the transit of Venus, the German with the
ship Gazelle, the English on the Challenger, and a French and an
American expedition. Moseley of the Challenger wrote that
in January, 1874, two of the whaling schooners then at the
island killed over 70 fur seals in one day and upwards of 20 on
another at some small islands off Howes Foreland, and he
deplores the wantonness and shortsightedness of their destruc-
tive methods. At the same time J. H. Kidder (1876) who
spent four months on Kerguelen, saw none of this species but
notes that "sealers speak of a few scattering fur-seals upon
this and Heard's Islands, but they have never been found in
large numbers." The latter island, though a favorite resort of
elephant seals, seems to have been less attractive to fur seals.
J. A. Allen (in Jordan and others, 1899, p. 316) quotes Capt.
George Comer as to the seals at Kerguelen. He spent five
months there in the winter of 1883-84 and obtained only six
skins. He adds, "About 1850* this island was visited by an
American who practically cleaned off the seals. The captain
I shipped with Joseph Fuller visited the island in 1880 and
took 3,000 seals practically all there were and this was the
increase for thirty years from 1850."

The St. Paul and Amsterdam Islands lie about 11 north of
Kerguelen and may be taken as the northward limit of this
fur seal. They were visited by Capt. Henry Cox in 1789, in
May, who on landing "found the shores covered with such a
multitude of seals that we were obliged to disperse them before
we got out of the boat . . . We procured here a thousand
skins of very superior quality, while we remained on the island
of Amsterdam, besides several casks of good oil." Lord
Macartney, who touched at Amsterdam in 1773, found there
five men collecting seal skins for the Canton market. He says
of the seals: "In the summer months they come ashore, some-
times in droves of 800 or 1,000 at a time, out of which 100 are
destroyed, that number being as many as five men can skin
and peg down to dry in the course of a day . . . Most of
those which come ashore are females, on the proportion of
more than thirty to one male." In 1874, Charles Velain of the

434 EXTINCT AND VANISHING MAMMALS

French transit of Venus expedition, reported considerable herds
of fur seals there (J. A. Allen, in Jordan and others, 1899), but
the more recent status of these animals seems difficult to
determine.

In 1924 the French Chamber of Deputies completely pro-
hibited taking or hunting seals in the Crozet Archipelago, St.
Paul and Amsterdam Islands, Howe, McMurdo, and Briant
Islands in the north of the Kerguelen Archipelago, and on
part of the south coast of Kerguelen, islands that together now
constitute a national park.

Presumably the fur seals formerly frequenting the Prince
Edward and the Crozet Islands were the same as the Kerguelen
fur seal, or possibly were nearer to the form of South Africa.
No precise studies of the matter seem to have been made.
The former group lies some 900 miles southeast of the Cape of
Good Hope, and the latter about the same distance to the east
and half way between the Prince Edward and Kerguelen
groups. According to J. A. Allen (in Jordan and others, 1899)
fur seals formerly abounded on the Prince Edward group.
About 1806, Capt. H. Fanning obtained a full cargo of fur
seals there, as did other vessels at the same time, but definite
statistics are unavailable. He was also the first sealer to visit
the Crozet Islands (in 1805), but although he saw an abundance
of seals there he passed on to the Prince Edward Islands.
Later on, however, many seals were taken over a number of
years at the Crozets. At Possession Island, the largest of the
group, Captain Brine in 1876 found "hundreds of seals, which
were resting on the damp grass bordering on the stream which
at this point enters the sea." They must have been greatly
depleted soon after, however, for in 1887, Captain George
Comer visited the islands on the recommendation of people at
Cape Town who had formerly taken great numbers there, but
after five months his party succeeded in obtaining only three
seals (J. A. Allen, in Jordan and others, 1899). What numbers
if any now occur on these islands would be interesting to know.

CAPE FUR SEAL

ARCTOCEPHALUS PUSILLUS (Schreber)

Phoca pusilla Schreber, Saugthiere, vol. 3, p. 314, 1776; p. 584, 1777; pi. 86, 1775
("Im indischen Meere" but assumed to be South Africa).

OCEANIC MAMMALS 435

SYNONYMS: Phoca antarctica Thunberg, Mem. Acad. Sci. St. Petersbourg, vol. 3, p.

222, 1811; Otaria peronii Desmarest, Encyclop. Meth., Mammalogie, p. 250, 1820;

Arctocephalus delalandii Gray, Proc. Zool. Soc. London, 1859, p. 107 (Cape of

Good Hope) ; Arctocephalus nivosus Gray, Ann. Mag. Nat. Hist., ser. 4, vol. 1, p.

219, 1868 (Cape of Good Hope); Arcto-cephalus schist-hyperoes Turner, Journ.

Anat. and Physiol., ser. 2, vol. 2, p. 113, 1868 (Cape of Good Hope).
FIGS.: Gray, 1859, pi. 69; Sclater, W. L., 1900, vol. 1, figs. 34, 35; Shortridge, 1934, vol.

1, 4 pis. opposite p. 204.

While there seems some doubt as to the strict applicability
of the name pusilla to this species, it is generally assumed that
it was given to a South African specimen; the next available
name, and the one sometimes used, is antarctica.

The characters of this fur seal have not been well defined.
According to Gray the palate is concave, "hinder aperture
narrow, with a rather acute, ovate anterior edge; teeth large;
lower jaw rather short, strong." Otherwise no characters are
pointed out that might distinguish it from A. gazella. Length
of adult male, 8 feet from nose to root of tail; an average female
about 4 feet 4 inches ( W. L. Sclater) .

This seal was formerly abundant on the islands off the
coasts of South Africa, but W. L. Sclater (1900) wrote that by
the beginning of this century 4 ^the number killed of late years
has not been very great, as they have been nearly exterminated,
and it is considered very desirable to allow them to increase."
This wise course evidently resulted in their recuperation, for
at the present time there are some flourishing rookeries on
these coasts. The most northern breeding colony is at Cape
Cross, 100 miles north of Walvis Bay, on the mainland of
Southwest Africa, and is the only colony on the mainland. It
is now visited annually "by a far larger breeding colony than
is on any of the islands" (Shortridge, 1934). Other localities
are the various islands between Walvis Bay and Liideritz,
comprising a number of small islands, though from some the
seals seem to have departed. On the east coast the fur seals
breed as far as Rocky and Bird Islands, in Algoa Bay, and
occasionally are seen along the coast as far as East London.

At one time, Lydekker wrote, as many as 70,000 to 80,000
skins were annually imported from the Cape to London, but
the number "is now much reduced." The females come ashore
in November to give birth to their young. Since Sclater in
1900 did not mention the large rookery at Cape Cross, one
may conclude that it has been established in later years as a
result of protection.

436 EXTINCT AND VANISHING MAMMALS

It is believed that it was this fur seal that formerly was
abundant at Tristan da Cunha and the neighboring islands.
This group was first "visited for fur seals in 1790 by Captain
Patten, of the American schooner Industry, of Philadelphia,
who secured 5,000 skins. Large numbers are said to have been
subsequently obtained there, probably from the smaller islands
of the group, Inaccessible and Nightingale Islands. The latter
is apparently still frequented by a few fur seals. Gough
Island, somewhat to the southward of the Tristan group,
formerly abounded with fur seals " (J. A. Allen, in Jordan and
others, 1899); but according to Capt. George Comer there
were practically none left by 1887, when his vessel put six men
ashore there for nine months. They were able to secure only
40 or 50 skins. Nevertheless "odd specimens were killed in the
Tristan group up to 1920, but now appear to have been ex-
terminated" (Shortridge, 1934).

Whether the fur seal of the Crozet Islands was the same as
the South African form or more resembled that of Kerguelen
Island does not seem to have been determined. However, for
the present it is included under the latter.

PHILIPPE'S FUR SEAL
ARCTOCEPHALUS PHILIPPII (Peters)

Otaria (Arctophoca) philippii Peters,