Comments on

the Taxonomy and Geographic Distribution

of Some North American Marsupials, Insectivores

and Carnivores

BY

E. RAYMOND HALL and KEITH R. KELSON

University of Kansas Publications

Museum of Natural History

Volume 5, No. 25, pp. 319-341

December 5, 1952

UNIVERSITY OF KANSAS

LAWRENCE

1952

University of Kansas Publications, Museum of Natural History

Editors: E. Raymond Hall, Chairman, Henry S. Fitch,


Robert W. Wilson

Volume 11, No. 25, pp. 319-341


Published December 5, 1952

University of Kansas


Lawrence, Kansas

PRINTED IN


THE STATE PRINTING PLANT


TOPEKA, KANSAS


1959

Comments on
the Taxonomy and Geographic Distribution
of Some North
American Marsupials, Insectivores
and Carnivores



BY



E. RAYMOND HALL and KEITH R. KELSON

In preparing maps showing the geographic distribution of North American
mammals we have found in the literature conflicting statements and
questionable identifications, which have led us to examine the
specimens concerned with results as set forth below. Our studies have
been aided by a contract (NR 161-791) between the Office of Naval
Research, Department of the Navy, and the University of Kansas.
Grateful acknowledgment is made to the persons in charge of the several
collections of mammals consulted for permission to examine and study
the specimens therein.

Didelphis marsupialis californica Bennett

From Cuernavaca, Morelos, Hooper (Jour. Mamm., 28:43, February 1, 1947)
lists a specimen, as he says, on purely geographic grounds, as of the
subspecies Didelphis mesamericana tabascensis. We have examined
this specimen, an unsexed skull-only, which falls within the range of
individual variation of Didelphis marsupialis californica and
refer the specimen to that subspecies.

Didelphis marsupialis etensis J. A. Allen

From El Muñeco, Costa Rica, Harris (Occas. Papers, Mus. Zool. Univ.
Michigan, no. 476:7, October 8, 1943) lists as Didelphis
richmondi a specimen (, No. 67550 U.M.). Our examination
of the specimen shows it to be within the range of individual variation
of populations that have been referred to D. m. etensis from
adjoining areas. We identify the specimen as Didelphis marsupialis
etensis.

Didelphis marsupialis tabascensis J. A. Allen

From Minatitlán, Veracruz, J. A. Allen (Bull. Amer. Mus. Nat. Hist.,
14:168, June 15) listed a specimen under the name Didelphis
marsupialis [in the trinomial sense] instead of under the name
Didelphis marsupialis tabascensis, which would be expected, on
geographic grounds, to apply. The specimen is No. 78123, U.S. Nat.
Mus., Biol. Surv. Coll. Our examination of the specimen reveals that it
is within the range of individual variation of Didelphis marsupialis
tabascensis and we identify the specimen as of that subspecies.
From Yaruca, Honduras, Bangs (Bull. Mus. Comp. Zool., 39:157, July,
1903) doubtfully listed as Didelphis yucatanensis a specimen,
No. 10611, M.C.Z. Our examination of the specimen indicates that it
is within the range of variation expectable in Didelphis marsupialis
tabascensis, known from surrounding areas, and we identify the
specimen as Didelphis marsupialis tabascensis.

Didelphis marsupialis virginiana Kerr

J. A. Allen (Bull. Amer. Mus. Nat. Hist., 14:166, May 28, 1901) and A.
H. Howell (N. Amer. Fauna, 45:20, October 28, 1921) have identified
four skulls from Sylacuga, Alabama, as Didelphis virginiana
pigra. The two subspecies virginiana and pigra are
not known to differ cranially. We have, however, examined the skulls
which are Nos. 44057-44060 in the U.S. Nat. Mus., Biol. Surv. Coll.
Because they are from a place north of other localities (Auburn and
Greensboro, Alabama) from which the subspecies virginiana has
been recorded, and within the geographic range of virginiana, we
identify the specimens as Didelphis marsupialis virginiana.

Sycamore Creek (synonymous with Fort Worth), Texas, is a place from
which J. A. Allen (op. cit.:173) recorded a specimen as
Didelphis marsupialis texensis. This specimen (No. 24359/31765
U. S. Nat. Mus., Biol. Surv. Coll.) is in the black color-phase. There
are only a few white hairs on the hind feet, and the basal fourth of
the tail is black. The black phase occurs all through the range of the
species D. marsupialis and our examination of the specimen
reveals no characters by which it can be distinguished from D. m.
virginiana of the surrounding region and we accordingly identify
the specimen as Didelphis marsupialis virginiana.

Didelphis marsupialis pigra Bangs

Davis (Jour. Mamm., 25:375, December 12, 1944) was one writer who
presented evidence that Didelphis virginiana (through its
subspecies virginiana or pigra or both) was only
subspecifically distinct from the species Didelphis mesembrinus
(= D. marsupialis) through the subspecies texensis.
Davis, however, did not actually employ a name combination that would
enforce his conclusion and he remarked that he had not seen specimens
which showed actual intergradation in the color of the toes. As the
remarks below will show, Davis (loc. cit.) was correct in his
supposition that J. A. Allen had seen such specimens.

Deming Station, Matagordo, and Velasco, Texas, are three places from
which J. A. Allen (Bull. Amer. Mus. Nat. Hist., 14:162, May 28, 1901)
listed specimens as Didelphis virginiana. The specimens
concerned are in the Biological Surveys Collection of the U.S. Nat.
Museum and bear catalogue numbers as follows: Deming Station,
32430/44266, 32432/44268, 32433/44269; Matagordo, 32431/44267; Velasco,
32812/44833. In each specimen the tail is shorter than the head and
body. The specimen from Velasco is semi-black, has the basal tenth of
the tail black and there is no white on the ears or tail. The specimen
from Matagordo is grayish, has the basal fifth of the tail black, ears
black, the right hind foot black, but there is some white on the toes
of the left hind foot and on each of the forefeet. Of the three
specimens from Deming Station, all are in the gray color-phase. The
first has the tail black only as far from the base as there is hair and
there is considerable whitish on the hind toes. The second specimen has
the basal fifth of the tail black and a slight amount of whitish on the
hind toes. The third specimen has the basal third of the tail black and
the toes are all black. In the sum total of their characters the
specimens mentioned above are referable to Didelphis marsupialis
pigra. These five specimens, and indeed the three from Deming
Station alone, show intergradation in coloration of the feet between
Didelphis marsupialis texensis and Didelphis virginiana
pigra. Probably there is three-way intergradation here at Deming
Station in that D. v. virginiana immediately to the north is
involved. The specimens mentioned above, along with the information
recorded by Davis (loc. cit.) and other authors (for example, J.
A. Allen, loc. cit., and Bull. Amer. Mus. Nat. Hist.,
16:249-279, August 18, 1902), give basis for arranging the North
American Didelphis as follows:

In listing the subspecific names given immediately above we are aware
of the possibility that a thorough study of the geographic variation in
Didelphis marsupialis may contract or expand the list of
recognizable subspecies. We are aware also that Hershkovitz (Fieldiana:
Zoology, 31 (No. 47):548, July 10, 1951) has arranged several of the
subspecific names listed immediately above as synonyms of Didelphis
marsupialis californica Bennett. We have not employed his
arrangement because he has not given proof that the currently
recognized subspecies are indistinguishable.

Caluromys derbianus canus (Matschie)

Matschie (Sitzungsberichte der Gesellschaft Naturforschender Freunde zu
Berlin, Jahrgang 1917, p. 284 (for April), September, 1917) applied the
name Micoureus canus to a specimen on which the locality was no
more precise than Nicaragua. Comparison of Matschie's description with
specimens in the United States National Museum (including the holotype
of Philander centralis Hollister and referred specimens of
Philander laniger pallidus Thomas) reveals that Matschie's
specimen was intermediate in coloration between the other two kinds of
woolly opossums named above and that there is nothing distinctive, in
the specific sense, in the cranial measurements which Matschie
published (op. cit.). M. canus, therefore, may be merely
an intergrade between the two previously named woolly opossums (C.
d. centralis and C. d. pallidus), an individual variant of a
previously named kind, say, C. d. pallidus, or a valid
subspecies. If it is a recognizable subspecies, it probably comes from
somewhere in the eastern half of Nicaragua. As a means of handling the
name, Micoureus canus Matschie, we tentatively place it as a
subspecies of the species Caluromys derbianus. The name may,
therefore, stand as Caluromys derbianus canus (Matschie), with
type locality in Guatemala.

Caluromys derbianus fervidus (Thomas)

Elliott (Field Columb. Mus. Nat. Hist., Publ. No. 115, Zool. Ser., 8:5,
1907) lists as Caluromys laniger pallidus a specimen from
Honduras that was acquired for the Field Columbian Museum (= Chicago
Natural History Museum) by purchase from Ward's Natural Science
Establishment of Rochester, New York. On August 4, 1951, in the Chicago
Natural History Museum, we found in the catalogue of the collection of
Recent mammals an entry for a male Caluromys bearing catalogue
number 6 and listed as from "San Pedro Sula [Honduras]. From Wards.
Mounted". In the collection of study specimens there is no specimen
from Honduras that was purchased from Ward's, mounted or unmounted. In
the sealed, glass-fronted, exhibit cases of mammals on display there is
one, and only one, Caluromys. It is presumed to be specimen No.
6. This specimen is not C. d. pallidus because it is too dark.
It could be Caluromys derbianus fervidus and we tentatively
refer it to that subspecies.

Caluromys derbianus pallidus (Thomas)

From Puntarenas, Costa Rica, Harris (Occas. Papers Mus. Zool. Univ.
Michigan, 476:7, October 8, 1943) listed as Caluromys laniger
centralis a female, skull and skin, No. 62702 in the Museum of
Zoology of the University of Michigan. We have examined this specimen,
the color of which is darker than in some other specimens of C. d.
pallidus but lighter than that of specimens of C. d.
centralis (for example, specimens from Turrialba, Costa Rica) and
on basis of color we refer No. 62702 to Caluromys derbianus
pallidus.

Scalopus aquaticus aereus (Bangs)

Bangs' (Proc. Biol. Soc. Washington, 10:138, December 28, 1896) name
S. a. aereus was based on a single specimen that shows more than
an average amount of coppery color. Jackson (N. Amer. Fauna, 38:52,
September 30, 1915) and subsequent authors accord full specific rank to
the specimen under the name Scalopus aereus. Blair (Amer.
Midland Nat., 22:98, July, 1939) recorded, from the type locality of
Scalopus aereus, normally colored individuals of Scalopus
aquaticus pulcher Jackson. Previously, Scheffer (Kansas State
Agric. College, Exp. Bull., 168:4, August 1, 1910) reported that in his
examination of 100 individuals of Scalops [= Scalopus]
aquaticus from Manhattan, Kansas, there were two individuals
"that were suffused all over with rich golden brown." Because our
examination of the type specimen of Scalops texanus aereus Bangs
reveals no features additional to coppery color that differentiate
aereus from other individuals of Scalopus aquaticus
pulcher Jackson (Proc. Biol. Soc. Washington, 27:19, February 2,
1914) we conclude that Jackson's name and Bangs' name (Scalops
texanus aereus) apply to the same subspecies. Bangs' name has
priority and the correct name, therefore, for the populations of moles
that in recent years have been designated as Scalopus aereus
Bangs and Scalopus aquaticus pulcher Jackson will be Scalopus
aquaticus aereus (Bangs). This name combination was previously used
by Miller (U.S. Nat. Mus. Bull., 79:8, December 31, 1912).

Scalopus aquaticus australis (Chapman)

Quay (Jour. Mamm., 30:66, February 14, 1949) recorded Scalopus
aquaticus from Springhill Plantation, 10 miles south-southwest of
Thomasville, Georgia. He stated that the specimens were intermediate
between the subspecies S. a. australis and S. a. howelli,
but did not refer the specimens to either subspecies. The locality
whence the material was obtained is approximately half way between the
geographic ranges, as previously known, of S. a. australis and
S. a. howelli (see Jackson, N. Amer. Fauna, 38, September 30,
1915).

The specimens recorded by Quay probably are two females in the
Cleveland Museum of Natural History bearing Catalogue Nos. 18136 and
18262 and labeled as from Springhill Plantation, Thomas County,
Georgia. We have examined these specimens and find that they resemble
S. a. howelli in narrowness across the upper tooth-rows, but
that they resemble S. a. australis in length of tail (22, 24),
in shortness of maxillary tooth-row (9.5, 9.5), and in convex dorsal
outline of the skull. Accordingly, we refer the specimens to
Scalopus aquaticus australis.

Sorex cinereus cinereus Kerr

In his revision of the American long-tailed shrews, Jackson (N. Amer.
Fauna, 51, vi + 238, 13 pls., 24 figs., July 24, 1928) referred
specimens of Sorex cinereus from Tyonek, Cook Inlet, Alaska, to
the subspecies S. c. cinereus (op. cit.: 46) and one
specimen from Chester Creek, Anchorage, Alaska, to the subspecies S.
c. hollisteri (op. cit.: 56). Thus, the geographic ranges of
the two subspecies would seem to overlap around the northern shores of
Cook Inlet. In an attempt to resolve this seemingly anomalous
distribution, we have examined pertinent materials in the Biological
Surveys Collection, U.S. National Museum. We agree with Jackson (op.
cit.) that the series of specimens from Tyonek is readily referable
to S. c. cinereus. To our eye, however, the specimen, No.
232691, from Anchorage is referable to Sorex cinereus cinereus,
rather than to S. c. hollisteri. The reference is made on the
basis of the darker color, especially of the underparts. In this
specimen, other characters that distinguish the two mentioned
subspecies are not apparent, probably because it is relatively young;
the teeth show only slight wear.

Sorex trowbridgii humboldtensis Jackson

In his account of the long-tailed shrews, Jackson (N. Amer. Fauna,
51:98, July 24, 1928) listed under specimens examined of Sorex
trowbridgii montereyensis four specimens from 7 mi. N Hardy,
Mendocino Co., California. Under his account of the subspecies S. t.
humboldtensis, however, he (op. cit.:97) mentions that
specimens (seemingly the same four) from 7 mi. N Hardy "have shorter
tails than typical representatives of humboldtensis, but in
color and cranial characters they are similar to this
[humboltensis] subspecies." We conclude, therefore, that the
specimens mentioned were inadvertently listed as S. t.
montereyensis and are Sorex trowbridgii humboldtensis. This
conclusion is supported by the fact that the locality concerned, 7 mi.
N Hardy, is within the geographic range assigned to S. t.
humboldtensis by Jackson (op. cit.:97); his southern records
of occurrence of S. t. humboldtensis are Sherwood and Mendocino,
both in Mendocino County, California. Our conclusion is further
supported by Grinnell's (Univ. California Publ. Zool., 40(2):80,
September 26, 1933) statement of the range of S. t.
montereyensis as "from southern Mendocino County south...."

Blarina brevicauda churchi Bole and Moulthrop

Kellogg (Proc. U.S. Nat. Mus., 86:253, February 14, 1939) tentatively
referred specimens of the short-tailed shrew from the mountainous parts
of eastern Tennessee to the subspecies Blarina brevicauda
talpoides, with the remark that they were unlike specimens of that
subspecies obtained in eastern and southern West Virginia.
Subsequently, Bole and Moulthrop (Sci. Publ. Cleveland Mus. Nat. Hist.,
5:109, September 11, 1942) named the subspecies Blarina brevicauda
churchi with type locality at Roan Mountain, North Carolina. We
have examined the specimens in the U.S. National Museum recorded by
Kellogg (loc. cit.) from the following localities: Shady Valley,
2900 ft. (Catalogue No. 267182); Holston Mtn., 4 mi. NE Shady Valley,
3800 ft. (Nos. 267176-267178, 267180, and 267181); Holston Mtn., 3 mi.
NE Shady Valley, 3000 ft. (No. 267179); Roan Mtn., (Nos.
267469-267475); Mt. Guyot, 6300 ft. (No. 267183); 4½ mi. SE Cosby,
3300 and 3400 ft. (Nos. 267184 and 267185); and Snake Den Mtn., 3800
ft. (No. 267186). Among named kinds of Blarina brevicauda, we
find these specimens to resemble most closely Blarina brevicauda
churchi and so refer them. They are readily distinguishable from
specimens of B. b. kirtlandi, that occurs farther north in the
same mountain range, by larger size and longer tail. Incidentally, in
the specimens that we have examined, we do not find that B. b.
churchi is darker colored than other subspecies of Blarina
brevicauda; B. b. churchi, to us, is indistinguishable in
color from B. b. kirtlandi. Bole and Moulthrop (op. cit.)
thought that B. b. churchi was notably darker than other
subspecies from adjoining areas.

Blarina brevicauda carolinensis (Bachman)

Blair (Amer. Midland Nat., 22(1):99, July, 1939) referred specimens of
the short-tailed shrew from the Arbuckle Mountain area of Oklahoma to
Blarina brevicauda hulophaga and specimens from Mohawk Park,
Tulsa County, Oklahoma, to B. b. carolinensis. Later Bole and
Moulthrop (Sci. Publs. Cleveland Mus. Nat. Hist., 5:108, September 11,
1942) saw two of the specimens from Mohawk Park and assigned them to
B. b. hulophaga. According to the most recent published account,
therefore, B. b. hulophaga would seem to have a peculiarly
discontinuous geographic range. We have examined the material seen by
Blair and by Bole and Moulthrop (Nos. 75946, 75947, 75643, Mus. Zool.
Univ. Michigan) in an attempt to form our own judgment as to their
subspecific identity. The teeth of No. 75946 are well worn, whereas the
teeth of the other two are scarcely worn. We are unable to distinguish
No. 75946 from topotypes of B. b. carolinensis by size, color,
or cranial features. The two younger specimens are smaller and paler,
but do not agree with the description of B. b. hulophaga. The
nearly-complete narrow, white girdle of No. 75947 is clearly an
individual variation. We assign the animals to Blarina brevicauda
carolinensis (Bachman) as did Blair (loc. cit.).

Blarina brevicauda minima Lowery

Bailey (N. Amer. Fauna, 25:207, October 24, 1905) identified as
Blarina brevicauda carolinensis one specimen from Joaquin and
two specimens from Big Thicket, 8 mi. NE Sour Lake, both localities in
eastern Texas. Strecker and Williams (Jour. Mamm., 10:259, August 10,
1929) later recorded the specimens again under the same name. The
subsequent naming of B. b. plumbea from Aransas National
Wildlife Refuge, Aransas County, Texas (Davis, Jour. Mamm., 22(3):317,
August 14, 1941) and B. b. minima from Louisiana (Lowery, Occas.
Papers Mus. Zool., Louisiana St. Univ., 13:218, November 22, 1943)
leaves the identity of the specimens from eastern Texas in doubt. We
have examined the following specimens in the Biological Surveys
Collection, U.S. National Museum: No. 117372, from Joaquin; No.
136407, from 7 mi. NE Sour Lake; and No. 136788, from 8 mi. NE Sour
Lake. We judge these to be the specimens referred to by Bailey (loc.
cit.). We find that they are indistinguishable from specimens of
Blarina brevicauda minima and they seem to differ from B. b.
plumbea in being chestnut rather than plumbeous in color and in
lacking the highly-arched posterior border of the palate. They are
easily distinguished from B. b. carolinensis by their chestnut,
rather than slaty-black, color and small size. They are distinguishable
from B. b. hulophaga, to which they might conceivably be
referred on geographic grounds, by their color and small size. We refer
them to Blarina brevicauda minima Lowery.

Spilogale angustifrons angustifrons A. H. Howell

In his "Revision of the skunks of the genus Spilogale" (N. Amer. Fauna,
26, November 24, 1906) A. H. Howell identified certain specimens in the
United States National Museum as follows:

Hall and Villa (Univ. Kansas Publ., Mus. Nat. Hist, 1:448, December 27,
1949) inferred that No. 47177 from Pátzcuaro was instead referable to
Spilogale angustifrons angustifrons. Our examination of No.
47177 and of each of the other specimens mentioned by catalogue number
immediately above leads us to conclude that they all are of one
species, and that, among named kinds of Spilogale, they should
be referred to the subspecies Spilogale angustifrons
angustifrons Howell.

Our examination of all of the specimens that Howell (op. cit.)
identified as Spilogale [angustifrons] angustifrons reveals that
none of the specimens from the type locality had attained full adult
stature; the holotype is a subadult and the other specimens from the
type locality are even younger. The small size of these specimens from
the type locality seems to have misled Howell into thinking that they
were taxonomically distinct from the larger specimens—those from
Jalisco, Michoacán and Hidalgo—that he identified as other kinds.

Spilogale gracilis gracilis Merriam

In the genus Spilogale four specific names, concerning the
status of which we have been uncertain, are listed below in the order
of their appearance in the literature.

In 1906 (N. Amer. Fauna, 26:1-55, 10 pls., November 24) A. H. Howell's
"Revision of the skunks of the genus Spilogale" was published and the
four names listed above were retained by him as applying to four
species (not subspecies). His map (op. cit., pl. 1) showing the
geographic distribution of the four kinds looks reasonable enough at
first inspection and does not indicate any overlapping of the
geographic ranges of the species in question, but if a map be made by
plotting the localities of occurrence recorded by Howell (op.
cit.), for specimens examined by him, a notably different
geographic distribution is shown. For one thing the geographic ranges
of gracilis, leucoparia, arizonae and
ambigua coincide over a considerable part of Arizona. Also,
specimens collected in recent years from Arizona and adjoining areas do
not readily fit into the "species" recognized by Howell; some specimens
are structurally intermediate between two or more of these species and
other specimens combine the diagnostic characters ascribed to two or
more of the alleged species. For these and other reasons a re-appraisal
of the application of the names mentioned above long has been
indicated.

Before re-appraising the names it is pertinent to recall that Howell's
paper in 1906 on Spilogale was only the second revisionary paper
that he prepared. It was prepared by a man who at that time lacked much
taxonomic experience, and who held to a morphotype concept. Howell
worked under the guidance, in the literal sense, of Dr. C. Hart
Merriam. The concept of species and subspecies held by Merriam
fortunately was recorded by him (Jour. Mamm., 1:6-9, November 28,
1919). Merriam's reliance on degree of difference and his disregard of
intergradation were naturally (and necessarily, we think, in Howell's
work in 1906) adopted by Howell. For example, of six specimens from
Point Reyes in west-central California, a place less than ten miles
from the type locality of Spilogale phenax phenax, Howell
(op. cit.:33) assigned one specimen to the subspecies
Spilogale phenax latifrons! S. p. latifrons occurs in
Oregon and in northern California—no nearer than 200 miles to Point
Reyes. Howell's assignment of this specimen to S. p. latifrons
was not a lapsus, as persons with the modern (geographic)
concept of a subspecies would be likely to suppose. Howell's assignment
of the one specimen to S. p. latifrons and the other five
specimens to S. p. phenax was intentional, as he told one of us
(Hall). He explained that he relied upon the morphological characters
of the individual animal instead of upon the morphological characters
of a population of animals. To him, therefore, there was nothing
inconsistent in his procedure in 1906. Also, variation that was the
result of difference in age and variation that was the result of
individual deviation were not understood, or at least not taken into
account, by Howell in 1906, nor by Merriam in 1890. For example,
Merriam selected the most extensively white specimen available to him
for the holotype of Spilogale leucoparia. He, and Howell in
1906, used the extensiveness of the white areas of that particular
specimen (see fig. 3, pl. 2, N. Amer. Fauna, 26, 1906) as a character
diagnostic of the "species" S. leucoparia although each of the
authors had available two other specimens of S. leucoparia from
the type locality, and all of the other referred specimens in the
United States National Museum, that were less extensively white than
the holotype. The individual specimen was the primary basis for
the species or subspecies and one selected specimen alone often was
used in making comparisons between a given named kind and some other
species or subspecies. Also, be it remembered, degree of difference,
and not presence or absence of intergradation, was the basis on which
subspecific versus specific rank was accorded to a named kind of
animal. Howell wrote on the labels of some specimens of
Spilogale "not typical" when the individuals differed from the
type specimen in features that owe their existence to individual
variation, and he wrote the same words on the labels of other specimens
that had not yet developed mastoidal crests because the animals were
not yet adult.

Anyone who examines the specimens that Howell used will do well to bear
in mind the circumstances noted above concerning Howell's paper of
1906; otherwise the reasons for Howell's identifications of certain
specimens can not be understood.

We have examined and compared the holotypes, and other specimens used
by Howell. While doing so we have borne in mind the degree of
individual variation well shown by each of several series of specimens
(for example, that in six adult males, from the Animas Mountains of New
Mexico, recorded by V. Bailey, N. Amer. Fauna, 53:339, 1932) and age
variation (for example, that shown in specimens of S. interrupta
from Douglas County, Kansas). The degree of each of these kinds of
variation, although considerable, is not extraordinary. That is to say,
the variations are of approximately the same degree as we previously
have ascertained to exist in Mephitis mephitis and in Mustela
frenata, two species that are in the same family, Mustelidae, as
Spilogale. As a result of our comparisons, we conclude, first
that the four names mentioned at the beginning of this account all
pertain to one species, and second that the three names S.
gracilis, S. p. arizonae and S. ambigua, and probably
also S. leucoparia, were based on individual variations in one
subspecies. S. gracilis has priority and will apply; the other
names are properly to be arranged as synonyms of it, as follows:

Some information in support of the above arrangement, along with some
other observations on Spilogale, are as follows: The type
specimen of Spilogale gracilis bears on the original skin-label
in the handwriting of Vernon Bailey, the collector, the statement that
the tail was imperfect. The recorded measurements of 400 for total
length and 142 for length of tail, therefore, are presumed to be
subject to correction. This presumption and the further circumstance
that other specimens from Arizona and New Mexico are as large as
specimens of comparable age and sex that we have examined from Nevada
and Utah of Spilogale gracilis saxatilis Merriam, indicate that
S. g. saxatilis differs less from the allegedly smaller S. g.
gracilis than was previously thought. Nevertheless, from north to
south (for example, from northern Nevada to southern Arizona) there is
an increase in extent of white areas at the expense of black areas of
the pelage. As a result, the lateralmost white stripe in S. g.
saxatilis averages narrower (and often is wanting) than in S. g.
gracilis. The absence, or narrowness, of the lateralmost white
stripe seems to be the principal basis for recognizing S. g.
saxatilis, just as the tendency to narrow rostrum in Coloradan
specimens seems to be the principal basis for recognizing Spilogale
gracilis tenuis A. H. Howell. Both S. g. saxatilis and S.
g. tenuis are "poorly" differentiated from S. g. gracilis
and from each other.

The holotype of Spilogale ambigua Mearns is slightly smaller
than other adult males of comparable age, and the braincase, relative
to its width, is slightly deeper than in the average adult male. These
variations, nevertheless, are within the range of individual variation,
as also are those characterizing the holotype of Spilogale phenax
arizonae Mearns. The latter specimen is an adult male, with much
inflated mastoidal bullae, nearly straight dorsal profile on the skull,
relatively shallow braincase, and only slightly worn teeth.

The holotype of Spilogale leucoparia Merriam, as pointed out
above, is an extreme example of the extensiveness of the white areas of
the pelage at the expense of the black areas. This feature occurs more
often in the southwestern desert areas of the United States than it
does farther north. In addition to the extensiveness of the white
markings, the other two characters allegedly distinctive of S.
leucoparia are broad and much flattened braincase and great degree
of inflation of the mastoidal bullae. Although these three mentioned
features do distinguish S. leucoparia from S. indianola
to the eastward, they seem not to set S. leucoparia apart from
S. gracilis to the westward. For example, in Arizona some
specimens are extensively white and some others have the braincase
flattened and the mastoidal bullae much inflated. V. Bailey (N. Amer.
Fauna, 53:339, 1932) refers to a specimen (, No. 147252 USBS)
from the head of the Rio Mimbres in New Mexico in which, as our
comparisons show, the inflation of the mastoidal bullae exceeds that of
any Texan specimen of S. leucoparia, the holotype included.
Also, at the type locality of S. leucoparia, subadult male No.
188467 USNM and adult male No. 188468 USNM are narrower across the
mastoidal region than is the holotype. In summary and review, specimens
from the eastern part of the range heretofore ascribed to S.
leucoparia nearly all have much inflated mastoidal bullae whereas
less than half of the specimens of Spilogale from western New
Mexico and Arizona have these bullae as greatly inflated; but, in No.
147252 from the head of the Rio Mimbres of New Mexico the inflation of
the bullae is more extreme than in any specimen that we know of that
has been referred to S. leucoparia.

If intergradation occurs between Spilogale gracilis gracilis and
Spilogale indianola and between one or both of these kinds on
the one hand and Spilogale interrupta on the other hand, central
Texas would be a logical place to collect intergrades. We suppose that
such intergradation will be found to occur and that eventually
Spilogale putorius will be the specific name to apply to all of
the Recent subspecies of spotted skunks. Until proof of such
intergradation is forthcoming we employ current nomenclature.

Spilogale gracilis microdon A. H. Howell

A. H. Howell (N. Amer. Fauna, 26:31, November 24, 1906) listed as
Spilogale arizonae martirensis one specimen (
sad.-yg., 145886 USBS) from Comondú, which is the type locality of
S. microdon. Our examination of No. 145886 convinces
us that it is referable to S. microdon.

Examination of the materials used by Howell (op. cit.) reveals
that there is an increase in size of animal and its skull from within
the geographic range of S. g. martirensis southward to Cape St.
Lucas which is the type locality of S. lucasana. Specimens of
S. microdon, which so far has been recorded only from Comondú,
the type locality, are, as would be expected, intermediate in size
between S. g. martirensis and S. lucasana. The
differential characters of these three named kinds of Spilogale
are principally those of size, and we can see no characters judged to
be of more than subspecific worth. Consequently the named kinds should
stand as:

Spilogale gracilis microrhina Hall

When Hall (Jour. Mamm., 7:53, February 15, 1926) named as new
Spilogale phenax microrhina, he did not mention specimens
previously recorded by A. H. Howell (N. Amer. Fauna, 26:32, November
24, 1906) as Spilogale phenax from San Bernardino Peak (57026
USBS), La Puerta (99580 USBS), Dulzura (55848, 56173, 56873,
33693/45728, 36291/48656 and 36292/48657) in southern California. On
geographic grounds these specimens would be expected to be S. g.
microrhina although geographically slightly outside the area that
could be delimited by Hall's (op. cit.) marginal record-stations
of occurrence. Our examination of the pertinent specimens reveals that
they are Spilogale gracilis microrhina. The localities from
which the specimens came are, respectively, the northeasternmost,
easternmost and southernmost occurrences so far listed for the
subspecies.

Conepatus mesoleucus mearnsi Merriam

Examination of the holotypes of Conepatus filipensis Merriam,
Conepatus pediculus Merriam, Conepatus sonoriensis
Merriam, and Conepatus mesoleucus mearnsi Merriam, and other
specimens of the two kinds last named, convinces us that all are the
same species and that the names should stand as follows: Conepatus
mesoleucus filipensis Merriam (type locality, Cerro San Felipe,
Oaxaca); Conepatus mesoleucus pediculus Merriam (Sierra
Guadalupe, Coahuila); and Conepatus mesoleucus sonoriensis
Merriam (Camoa, Río Mayo, Sonora).

One method of designating the ages of individuals in Conepatus
is to recognize four categories from younger to older, as follows: 1)
juvenile—retaining one or more deciduous teeth; 2) young—sutures open
and clearly to be seen between bones of the facial part of the skull;
3) subadult—skull of adult form, but lacking sagittal and lambdoidal
crests and retaining faint traces of sutures between facial bones; and
4) adult—sutures obliterated, lambdoidal ridge high and temporal
ridges (of females) or sagittal crest (of males) prominent.

On this basis of designating age, the holotype of C. pediculus
is young and nearer the juvenal than the subadult stage. Its small size
is partly the result of its youth. Other than its small size we find no
characters to distinguish it from C. m. mearnsi. Unfortunately
no young male of C. m. mearnsi of the same age as the holotype
of C. pediculus is available. Also, from the general area of the
Sierra Guadalupe, Coahuila, only the one specimen of Conepatus
mesoleucus (the holotype of C. m. pediculus) is known.
Consequently, we can not yet prove that some young males of C. m.
mearnsi are as small as the holotype of C. pediculus.
Because of this lack of proof we tentatively recognize the subspecies
Conepatus mesoleucus pediculus instead of placing the name
Conepatus pediculus in the synonomy of Conepatus mesoleucus
mearnsi.

The holotype of C. sonoriensis is a young female, older than the
holotype of C. pediculus, and approximately midway between the
juvenal and subadult stages.

The holotype of C. filipensis is an adult male.

We suppose that C. mesoleucus mesoleucus Lichtenstein and C.
mesoleucus mearnsi Merriam on the one hand, and Conepatus
leuconotus leuconotus Lichtenstein and C. l. texensis
Merriam on the other hand will be found to intergrade, in which event
the name Conepatus leuconotus, having page priority over
Conepatus mesoleucus, will apply to the species. Proof of
complete intergradation is not yet available. The one difference
between the two that prevents our uniting them as subspecies of one
species is the larger size of C. l. leuconotus and C. l.
texensis. Measurements of the smallest adult male and female
available to us of C. l. texensis and of the largest adult male
and female of C. m. mearnsi are given below.

Where the geographic ranges of the two species approach one another the
only taxonomically significant difference detected by us is in size,
C. leuconotus being larger than C. mesoleucus. Other
characters that are useful in separating the two alleged species now
are known to vary geographically in a fashion that indicates only
subspecific status for the two kinds. For example, three specimens from
Laredo, Texas (previously recorded by V. Bailey, N. Amer. Fauna,
25:205, October 24, 1905—Nos. 24839/32237, 24840/32238 and 24842/32245
USBS), bridge the gap in color pattern between C. l. texensis to
the east and C. m. mearnsi to the west. C. l. texensis
characteristically has the white stripe terminating anteriorly in an
obtuse angle, and on the hinder back the area of white is restricted to
a narrow line or is wanting. C. m. mearnsi characteristically
has the white stripe truncate anteriorly and approximately as broad on
the hinder back as on the shoulders. In the specimens from Laredo, the
young female, No. 24842, has the white nearly truncate anteriorly
(pointed in the other two specimens, adult females). In No. 24839 the
area of white on the hinder back is only slightly restricted in width
(noticeably restricted but present in the other two specimens).

The proof of intergradation, or the lack of it, between the two alleged
species, Conepatus mearnsi and Conepatus leuconotus,
would seem to be profitably sought by obtaining specimens along the Rio
Grande in Texas between the Blocker Ranch ("50 miles southeast of Eagle
Pass") and Laredo.

Measurements illustrating the size difference between the two alleged
species are as follows:

Table 1. Measurements of Conepatus from Texas

 
 

[1] Measured dry.

Conepatus mesoleucus venaticus Goldman

When Goldman (Jour. Mamm., 3:40, February 10, 1921) named C. m.
venaticus from Arizona he did not mention material which Merriam
(Proc. Biol. Soc. Washington, 15:163, August 6, 1902) had recorded from
Ft. Verde, Arizona, under the name Conepatus mesoleucus mearnsi.
This material seems to be specimens in the American Museum of Natural
History of which the two oldest specimens are as follows: No.
2486/1921, male, adult, from Box Cañon, 20 mi. S Ft. Verde; No.
2487/1922, female, subadult, from Verde River, Arizona. Pertinent
measurements of these specimens are, respectively, as follows:
condylobasal length, 72.4, 68.8; zygomatic breadth, 50.0, 44.2; width
of braincase at constriction behind zygomata, 36.4, 33.8; mastoidal
breadth, 44.3, 38.4. Comparison of these measurements with those given
for C. m. venaticus (Goldman, loc. cit.) reveals that the
specimens concerned agree in narrowness of skull with C. m.
venaticus (C. m. mearnsi is relatively wider) and it is on
this basis that we refer the specimens to Conepatus mesoleucus
venaticus.

Urocyon cinereoargenteus costaricensis Goodwin

J. A. Allen (Bull. Amer. Mus. Nat. Hist., 20:48, February 29, 1904)
listed two specimens of gray fox from Pozo Azul, Costa Rica, as
Urocyon guatemalae. Goodwin, in his "Mammals of Costa Rica"
(Bull. Amer. Mus. Nat. Hist., 87(5):271-474, December 31, 1946) did not
mention any material from Pozo Azul. We have examined the skull of the
adult female (No. 19208 AMNH) taken on July 17, 1902, at Pozo Zul
[sic], by M. A. Carriker and find it to be indistinguishable from other
specimens of Urocyon cinereoargenteus costaricensis to which
subspecies we therefore refer the specimen.

Canis lupus griseoalbus Baird

In 1823 Sabine (No. V, Zoological Appendix, p. 654, In Narrative
of a journey to the shores of the Polar Sea ... xvi + 768, 30 pls., 4
maps, 1823, London, by John Franklin) applied the name Canis
Lupus-Griseus to the gray wolf in the vicinity of Cumberland House,
Saskatchewan. On the following page (p. 655) he employed the name
Canis Lupus-Albus for a white wolf obtained at Fort Enterprise,
Northwest Territories. In 1937 Goldman (Jour. Mamm., 18(1):45, February
14) did not consider the wolves of the Cumberland House region to be
sufficiently different from animals from surrounding areas to warrant
nominal separation for them and he placed the name Canis lupus
griseus Sabine as a synonym of Canis lupus occidentalis
Simpson. Anderson (Jour. Mamm., 24(3):386, August 17, 1943) revived
Sabine's name griseus and assigned to Canis lupus griseus
an extensive geographic range in central Canada. Later, Goldman (Part
II, Classification of wolves, p. 395 and 424, In The Wolves of
North America, American Wildlife Institute, May 29, 1944) by
implication, again arranged griseus of Sabine as a synonym of
Canis lupus occidentalis and pointed out (op. cit.:395)
that, in any event, the name griseus is preoccupied by
[Canis] Griseus Boddaert, 1784 [= Urocyon
cinereoargenteus (Schreber), 1775]. Still later, Anderson (Bull.
102, Nat. Mus. Canada, p. 54, January 27, 1947) again recognized the
subspecies formerly known as Canis lupus griseus Sabine, and,
because of Boddaert's prior usage of [Canis] griseus, renamed
the subspecies Canis lupus knightii. It appears, however, that
there is an earlier name available for this subspecies. Goldman (op.
cit., 1943:395) points out that "apparently combining the names
Canis (Lupus) griseus and Canis (Lupus) albus of Sabine
... as Canis occidentalis var. griseo-albus, Baird
[Mammals, Repts. Explor. and Surv. for R. R. to Pacific Ocean,
Washington, p. 104, vol. 8, (1857) July 14, 1858] seems to have
entertained a somewhat composite concept of a widely ranging race
varying in color from 'pure white to grizzled gray.' No type was
mentioned and the name does not appear to be valid or clearly
assignable to the synonomy of any particular race." We agree with
Goldman that Baird's concept was a composite one, but Baird's name,
Canis occidentalis var. griseo-albus, was clearly based
on the primary names of Sabine (griseus and albus), of De
Kay (occidentalis), of Maxmillian (variabilis, a synonym
of Canis lupus nubilis) and of Townsend (gigas, a synonym
of Canis lupus fuscus). Nevertheless, the name
griseo-albus was applied to, among others, the subspecies of
wolf the type locality of which is at Cumberland House, Saskatchewan,
and, by restriction, the name Canis lupus griseoalbus Baird is
available for the subspecies and, of course, antedates Canis lupus
knightii of Anderson (op. cit., 1947:54). It might be argued
that Baird did not intend to propose a new name, but that he did so is
a fait accompli. Canis lupus albus Sabine, 1823, is not
available since it is preoccupied by C[anis]. Lupus albus Kerr
(Animal Kingdom, Class I, Mammalia, p. 137, 1792), a name applied to
the wolf of the Yenisei region of Siberia.

The name and synonomy of the wolf of central Canada should stand as
follows:

Canis lupus griseoalbus Baird

The name Canis Lupus-Albus Sabine, 1823 (nec C[anis]. Lupus
albus Kerr, Animal Kingdom, p. 137, 1792) should, of course, be
retained as a synonym of Canis lupus mackenzii Anderson as
arranged by Anderson (Bull. 102, Nat. Mus. Canada, p. 55, January 24,
1947).

When Anderson (op. cit.:54) recognized the subspecies Canis
lupus knightii [= C. l. griseoalbus] he made no mention of a
specimen of wolf from Norway House, Manitoba, which Goldman (op.
cit., 1944:427) had referred to C. l. occidentalis, but the
subspecific identity of which was placed in doubt by Anderson's action.
We have examined the specimen, No. 115995, in the Biological Surveys
Collection, U.S. National Museum, and have compared it with specimens,
including topotypes, of C. l. occidentalis and C. l.
hudsonicus. The specimen fits the description of C. l.
griseoalbus and differs from C. l. occidentalis in its long
and narrow incisive foramina, larger skull, more nearly straight
frontal profile (not markedly concave), and slightly higher coronoid
processes. Other differences alleged to obtain between these two
subspecies offer no assistance in the present case. The specimen from
Norway House differs from C. l. hudsonicus in larger size of
skull and stouter, blunter, postorbital processes, the posterior
borders of which turn less abruptly inward. In brief, among currently
recognized subspecies, the specimen from Norway House seems best
referred to Canis lupus griseoalbus Baird.

Canis niger rufus Audubon and Bachman

Goldman (Part II, Classification of wolves, p. 486, In The
wolves of North America, American Wildlife Institute, May 29, 1944)
referred two specimens of the red wolf from Reeds Spring, Missouri, to
the subspecies C. n. gregoryi. Leopold and Hall (Jour. Mamm.,
26(2):143, July 19, 1945) referred wolves from 5 mi. N Gainesville and
from 3 mi. N Thomasville, both localities in Missouri, to C. n.
rufus. The identification of Leopold and Hall was made on the basis
of the small size of their specimens and they did not have the
advantage of comparative material. The locations of these and other
records of occurrence in Missouri and Arkansas suggest that the
specimens from Reeds Spring might be better referred to C. n.
rufus, the more western subspecies. An examination and comparison
of the two specimens from Reeds Spring, Nos. 244127 and 244527,
Biological Surveys Collection, discloses that they are intergrades
between C. n. rufus and C. n. gregoryi. They resemble
C. n. rufus in small size and cranial characters, but are more
nearly C. n. gregoryi in the darker, less brightly rufescent
color of the pelage. Being, in this case, more strongly influenced by
the size and cranial features than by the color, we consider the
animals from Reeds Spring best referred to Canis niger rufus.

Transmitted July 15, 1952.