Amerncan Museum

ovitates

PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY
CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N. Y. 10024

NUMBER 2300 SEPTEMBER 15, 1967
eee eee ee eee eee ene

A Revision of the Geology and Paleontology
of the Byou Hills, South Dakota

1 1

By Morris F. SKINNER! AND BERYL E. TAYLOR

INTRODUCTION

The two Bijou Hills adjacent to the Missouri River in south-central
South Dakota were the site of Hayden’s early fossil vertebrate collec-
tions and are the type locality for four of Leidy’s generic taxa: Leptarctus
primus, Merycodus necatus, Hippodon speciosus, and Merychippus insignis. The
present report is the result of an investigation of these hills and the fauna
obtained from them. In 1934 the senior author relocated the two hills
from Hayden’s original description and on six subsequent visits prospected
and studied in detail the fossiliferous deposits. The exact geographic loca-
tion of these hills and the stratigraphic description of the Tertiary sedi-
ments are presented here. A new formation, the Fort Randall, has been
named. A type locality has been designated for the sediments called the
“Rosebud Beds” by Gidley (Matthew and Gidley, 1904).

Literature on fossil mammals is replete with inadequate type specimens
and undefined type localities, but seldom has one locality and one
“fauna” shared in so many vicissitudes. The type specimens nearly defy
either definitions or diagnosis; the type of Merycodus necatus, a partial
ramus, is lost; the type of Leptarctus primus is one premolar; the type
of Hippodon speciosus is a lower cheek tooth; the type of Merychippus in-
signis is a maxillary fragment with two milk premolars; Hayden (1857,

? Assistant Curators, Frick Collection, the American Museum of Natural History.

2 AMERICAN MUSEUM NOVITATES NO. 2300

p. 157) mistakenly included the name “Hipparion occidentalis” in the fauna.
The type of H. occidentale consisted of five upper molars and is not from
the Bijou Hills. Moreover, the exact locality from which it came forever
will be uncertain (see below, p. 18).

The types and referred examples of the Bijou Hills genera have been
thoroughly studied as befits their long tenure in North American pale-
ontological literature. We have established a neotype for the lost type
of Merycodus necatus. The types of Leptarctus primus and Hippodon speciosus
are figured and discussed. Hippodon speciosus is here considered a nomen
dubium. The two milk teeth representing the type specimen of Merychippus
insignis have been matched closely to an immature dentition and com-
plete skulls from the Lower Snake Creek deposits of western Nebraska.

Two sets of deposits are well exposed at the Bijou Hills (see fig. 1B,
C). A lower set (Fort Randall, new formation) is the deposit from which
the neotype and paratype of Merycodus necatus were collected. It is probable
that the lost type of M. necatus and the types of Leptarctus primus and
Merychippus insignis were also collected from the Fort Randall Forma-
tion. The higher set, or the undifferentiated Valentine-Ash Hollow equi-
valents, have yielded fragments of horse teeth and a proboscidean tusk.
The type of Hippodon speciosus might have been derived from either the
Fort Randall or the undifferentiated Valentine-Ash Hollow deposits.

The Addendum by Mr. Stout is presented in its entirety, except for
the legends and references which have been incorporated in the main
body of the paper. Whereas Stout’s rodent identifications are accepted
unreservedly, the geologic usages and temporal correlations he employs
do not, in all cases, agree with ours.

ACKNOWLEDGMENTS

The late Mr. Childs Frick not only provided the means for the col-
lection and documentation of the specimens mentioned in this paper,
but he also guided the basic research on them. The many trips to the
Bijou Hills were made by the senior author at the request of Mr. Frick
who long had realized the need for better documentation of Leidy’s
early types. Some of the trips were fruitless, and none was really success-
ful in terms of quantity or quality of collections. The few small exposures
on the Bijou Hills have produced only partial or fragmentary specimens
numbering fewer than 40 in the Frick Collection. Our indebtedness to
Mr. Childs Frick for this support and for much more is here acknowledged.

We are grateful to the members in the Frick Laboratory who have,
by their interest and assistance, helped to bring this paper to publica-
tion. Dr. Malcolm C. McKenna, Dr. Richard H. Tedford, and Mr. Ted

1967 SKINNER AND TAYLOR: BIJOU HILLS 3

Galusha have critically read and given valuable suggestions. Mr. Ray-
mond J. Gooris has prepared the figures. Dr. Richard Estes has furnished
the identification of the turtle fragments from the Bijou Hills. We thank
Dr. Horace G. Richards of the Academy of Natural Sciences of Phila-
delphia for his assistance and generosity in making the Bijou Hills type
specimens in his collection available for study and casting.
The following abbreviations are used:
A.M.N.H., the American Museum of Natural History, Department of Vertebrate
Paleontology
A.N.S.P., Academy of Natural Sciences of Philadelphia

F.A.M., Frick American Mammals, the American Museum of Natural History
U.S.N.M., United States National Museum of the Smithsonian Institution

History oF THE Byou HILts

The exact location of the two Bijou Hills of Hayden has been for-
gotten in the twentieth century because of their inaccessibility by auto-
mobile, but in Hayden’s time the hills were prominent landmarks for
all voyagers who traveled up the Missouri River by boat or along the
river. The outcrops on the hills are clearly visible from the river (figs.
1A, 2). In a grass-covered country such outcrops would have attracted
the attention of Hayden, and he did make several trips to them. The
first was with F. B. Meek in 1853. The second was in the autumn of
1855. In 1856 Hayden accompanied Lieutenant G. K. Warren, a United
States Army Topographical Engineer, who commanded a reconnoitering
expedition through Nebraska and the Dakota territory. The trip, by
steamboat up the Missouri, was halted when a heavy spring flood
grounded the boat on a sand bar at Cedar Island.1 Some members of
the expedition set out on foot for the 130-mile march to Fort Pierre.
This line of march, which passed between the two main Bijou Hills and
their eastern extension, is shown on Warren’s military map of Nebraska
and Dakota in 1855-1856 (Warren, 1858). The hills were relocated in
1934 from this map and Hayden’s (1857b, p. 156) description which
was as follows: “These Hills are of considerable interest, as forming an
intermediate link uniting Bijou Hills to the main body of the Bad
Lands. The two upper beds [G and H of section, p. 153] of the vertical
section are represented at this locality. The last outlier of this deposit is
seen at Bijou Hills on the opposite side of the Missouri River in lat.

1 Cedar Island is not shown on a present-day map, for it lies under the waters of
Fort Randall Reservoir. Cedar Island was about 15 miles west of the town of Lake
Andes, South Dakota.

4 AMERICAN MUSEUM NOVITATES NO. 2300

43'2°. These are a group of isolated hills towering high above the sur-
rounding country and forming prominent land marks for the voyager.
The two highest hills border upon the river and are from four to six
hundred feet in height. Farther into the interior are two other hills, the
first about two miles long, and the second about eight miles, ranging
in a nearly east and west course, sloping gently down toward the Coteau
de Prairie. In the summer of 1853 I ascended one of the hills nearest
the river in company with my friend Mr. Meek, and, from a denuded
portion near the summit, we obtained several fragments of jaws and
teeth belonging to two new species of mammals, which have been de-
scribed by Dr. Leidy as Hipparion speciosum and Merycadus [sic] necatus.
In the autumn of 1856 I discovered on the denuded summits of the
same hills Hipparion occidentalis } and two new genera, Leptarctus primus,
an animal allied to the raccoon, and Merychippus insignis, a remarkable
new genus of ruminant horse.”

The town of Bijou Hills, founded in 1877, is situated inland about 8
miles east of Hayden’s two Bijou Hills bordering the river. This eastern
extension of the hills has been considered by some authors as the locality
where Hayden obtained the fossils later described by Leidy (Stevenson,
1958, p. 137; Green, 1958, p. 140). A narrow range called the Iona
Hills extends west of the Missouri and a little south for about 50 miles
to the Red Hill Buttes near Carter, South Dakota. Hayden was also
aware of this westward extension (1857b, p. 156).

There were no legal land surveys in Hayden’s time, and even now the
two hills present a difficult geographic description. The difficulty arises
because of a peculiar combination of three different survey corrections
(see fig. 1A): (1) The hills lie astride a township line; (2) there is an
exceptionally large jog correction in the ranges; (3) the hills are in two
different counties, yet border the Missouri River. Although these differ-
ences seem to separate the two hills widely, the bench marks on their
summits are only 7200 feet apart.

“THe Bijou QuarTzITE””
Hayden (1857a, map and profile; 1857b, p. 157) was the first to recog-

’ This was neither the type of, nor was it referable to, Hipparion occidentale. Leidy (1869,
p- 282) corrected the error. Hayden actually collected the type of Hipparion occidentale
in 1855 at some unknown point along the White River.

* So far as we can determine Agnew (1958, p. 129) was the first to use the term
“Bijou quartzite.” Apparently Agnew did not intend this as a formal name, for he used

the term interchangeably with “quartzitic sandstone,” “green quartzite,” and “greenish
siliceous quartzite.”

1967 SKINNER AND TAYLOR: BIJOU HILLS 5

nize and describe the lithology of the two Tertiary beds present on the
Bijou Hills, thus: “. . . The two upper beds of the vertical section are
represented at this locality. The last outlier of this deposit is seen at
Bijou Hills on the opposite side of the Missouri River in lat. 434°...
The two highest hills border upon the river and are from four to six
hundred feet in height. . . The summits . . . are capped with a bed of
bluish-gray compact rock, quite variable in its character. Sometimes it
is very fine, not unlike a metamorphic rock; again it is composed of an
aggregation of particles of granular quartz, interspersed with a few small
water-worn pebbles; then a coarse grained somewhat friable sandstone.”
On Hayden’s geologic chart (1857b, p. 153) Bed H (upper) contains
what was later named the “Bijou quartzite.”! The generalized section
Hayden gave described the deposits on the hills as we have figured them
in detail (fig. 1B, C).

The hard orthoquartzite layers hold up and form a ledge on a prom-
inent range of hills only a few miles wide which extend in an east-west
direction for at least 50 miles. The north and south Bijou Hills are only
a portion of the east end of the range. Most of the range is capped by
the characteristic Pliocene quartzitic layers, the so-called “Bijou Quartzite.”
The quartzitic layers, or zones, commonly vary in thickness from 1 foot
to 4 feet on the range called the “Iona Hills” extending westward to-
ward Carter, South Dakota.

Beds containing the “Bijou Quartzite” have been mapped as “Bijou
Formation” on several areal geology quadrangles of the South Dakota
Geological Survey (table 1). Stevenson and Carlson (1950; see table 1
of the present paper) were the first of several geologists to map the
“Bijou Formation.” In 1954 Stevenson (p. 86) formally named the Bijou
Formation and stated, “Because of the changing lithologic character of
this formation, no type section has been chosen, instead a typical section
is presented.” Stevenson made no mention of the Bijou facies in this
article. By 1958, as table 1 shows, the consensus among some geologists
was that Stevenson’s formation was only a facies.

In 1958 Stevenson revised the “Bijou Formation” and selected a type
locality, but, through an error of typography, omitted the real-estate
section number as follows, “. .. 5 miles north of Academy, South Dakota,
in the Bijou Hills (SW'4, SE'% Sec. [?] T. 101 N., R. 69 W.)” The critical

square mile represented in the omitted section number left the type lo-

1 Hayden’s geologic and faunal charts of 1869 are not in complete agreement with
his earlier ones, for, as his knowledge of western deposits enlarged, he altered his early
concepts.

‘deur oy} jo yoerq ay? uo 3x9) ASojoas ay) ur suvadde , soe nolig,, wa491 oy], :
‘deur ay} jo puasay oy} uo sivodde ..uonewsog nofig,, w0} ay], ,
‘GO6T “FIOFeC YING ‘uoT]IULIEA ‘eioyeq YINog
Jo Ayssoatuy) “sang poordozoan vjoyoq yimog ay) fo suowwoygny Ul UIAIS 219M QCET PUL OCI seep ey, ‘sdew oy) uo uoTeoNqnd jo alep ON g
"ejoyed YINog ‘uorpIwsaA ‘Asaing [eoiojoary vjoyxeq yINog ay) UO pouteigo a19m sdeur asoyy], ,

x x x = 0961 LS6I surT[OD JOUULM
x x x = 6G61 8S6I uoAag ‘UOAIZg 42S
x x px 9X 6561 8S6I1-L661 wostaAaig seed
x x x = gl 8S6T] LS6I wooyos WHIM
x x x = 8661 LG6I woaag ‘uooYysg eye qeAoy
x x x = 861 LS6I sul[[oD BPaMoM,
x x x = 8S6l LS61-9G61 WOsUsArgg A103219
— = <= x ZS61 OS6I-6F61 sisy10 pure sioyxeg OUI
— = = x eS61 8F6I Joyeg “uImpreg Buc]
— -_ = é 1G61 OS6I UIMpTeg o1qeyy
Ba — — x IS6I OS6I uos[IerT “uosuaAc1g sapuy 24e]
— — = x G61 6r6I Joyeg uoxIq
— = = x TS61 661 uos[Ier) “1yeg seon'y
— — _ x OS6I Oc6I uOos[IeD ‘UOsUudAI1¢ JPe}sou0g
= = as é qlosét] 6¥61 Sse) “UIMpleg epedye
UOT} eULIO J uoI}eULIO J sole] UuOneULIOY peysiyqng AdAIMG JsIBOTOI ajSuespend,
MOT[OH, YsSy sUT]UaIe A nolig nolig 21eq jo 21%q Jo 10yIny

pATAUNG TVOIOOTOND VLOAV(] HLNOG AHL dO SdvJJ ADOTOSD Ivauy
NO Gasp) SV ,SaIovy nolig,, GNV .NOLLVWAOY NoOfig,, SWaa], AHL AO ADOTONOAHD

tT WIavi

1967 SKINNER AND TAYLOR: BIJOU HILLS 7

cality still unfixed within the township. Stevenson’s redefinition of the
“Bijou Formation” was slanted heavily toward a facies concept: “One
of the basic ideas therefore, regarding facies is that they are strata char-
acterized by definite and easily distinguished lithology. Formations are
also characterized by diagnostic lithologies, therefore, the Bijou facies
can also be considered as a formation.

“Diagnostic lithologic properties which characterize the Bijou as a
formation, a facies, and an easily mappable unit, is the nearly homoge-
nous [szc] composition, siliceous cement, greenish color and resistance to
erosion. These properties quickly distinguish the Bijou facies from the
other strata of the Mio-Pliocene stratigraphic unit of which it is a part.”

It is our contention that the “Bijou Quartzite,” or facies, is not a
formation but a post-depositional cementation of certain sand beds that
may, and does, occur throughout several different lithic units or forma-
tions of the Tertiary and is not confined to any of them. A similar but
unrelated condition exists in the area of the Spanish Diggings south of
Manville, Niobrara County, Wyoming, but there the rocks are known
locally as the “Spanish Diggings quartzite.”

Agnew (1958, p. 131) stated, “. .. the Bijou quartzite has been mapped
in Brule, Charles Mix, and Gregory Counties, South Dakota . . . in Holt
and Boyd Counties, Nebraska . . . occurs throughout an east-west extent
of at least 150 miles.” Agnew and Tychsen (1965, p. 44) chose to designate
these silicified sands as the “Bijou Quartzite Facies.” This is a good de-
scriptive field term for a secondary rock condition, but it should never
be used as a formal stratigraphic term.

Green (1958, p. 142) advised that, “. . . the term ‘Bijou formation’
either be dropped or its usage be held in abeyance until accurate strati-
graphic collecting of fossil remains can be made.” We are in agreement
with Green, for the distinction of a rock unit is based on sediments
which must be recognized before the fauna contained therein may be
compared on a biostratigraphic basis with faunas of other rock units.

Stevenson (1954, p. 89) gave a faunal list! with his initial description
of the “Bijou Formation” but gave no lithic allocation for his specimens.
Green (1958, p. 142) stated: “A faunal list from scattered localities of
specimens alleged to have come from the ‘Bijou Formation’. . . is now
understood to be a mixed collection of Lower Pliocene specimens and

1For the sake of taxonomic stability the misspelled generic and specific names in
Stevenson’s lists are corrected: Nannihippus should be Nannippus; Hippoclan should be
Hippodon; Procamelus cf. P. calcaneus may have been P. calcaneum; Ustachoerus should be
Ustatochoerus; moclicus should be modicus; Rhinoceras is rhinoceros; Mustels is mustelid.

8 AMERICAN MUSEUM NOVITATES NO. 2300

3
SCALE: 135000

North Bijou Hill
South Bijou Hill

BRULE CO. | A

LES MIX CO.
GREGORY CO.

VERTICAL SCALE 1"=40

Elevation USBM 2028
NORTH BIJOU HILL

2025

fom Sandstone mamas
4 ,2'-5 Sandstone,very hard

Go 4
J 2'-5' Loose sand, coarse pebbles
; ae A en Oe Gk Fossil chips

Greenish =:

* K quartzite ledge,,
Sod cover over most of mints a

4'-5' Sand, fine, greenish

‘Proboscidean tusk
10' Clay, sandy, pink F en. ail 4
"Bijou quartzite

F:A.M. Nos. 42924, 42924-A we ledges?y:

Fic. 1 (THIS PAGE AND OPPOSITE PAGE). A. Map of the north and south Bijou
Hills from segments of the Dixon, Iona, and Lucas quadrangles of the South
Dakota Geological Survey. B. Section of north Bijou Hill. C. Section of south
Bijou Hill, the type locality for the Fort Randall Formation, new name.

1967 SKINNER AND TAYLOR: BIJOU HILLS 9

VERTICAL SCALE | = 40
SOUTH BIJOU HILL

TYPE LOCALITY OF FORT RANDALL FORMATION Elevation
CHARLES MIX COUNTY, SOUTH DAKOTA
NORTH V2 SECTION 8, T.IOON., R. 7 W.

: 2123 2128

Other hard sandstone zones
weheeeees indicated in talussikatis hye.

Sod covered

a “Bijou quartzite

" \'25' Hard sondstone, greenish

210s

6 coverad

* 5§'Clay, massive, sandy, greenish brown 2085

UNDIFFERENTIATED

-'15' Sand, loose, well sorted
channel type deposit

VALENTINE - ASH HOLLOW

Sod covered

== 4'Cla sandy, greenish- Concretions, white 2065

18'Sand,fine, silty, light greenish
2 zones of concretions, white

Rae ees 2045

i aes 6' Clay, sandy, greenish buff

2025
: (5' Sand, very fine, massive, gray to pink
F:A.M. No. 42925

17'Clay, massive, pink at base to reddish at top 2008

F:A.M. Nos. 87011,51790, ——~ Le sate 5 Clay as below ~ Fossils and
= barite rock rose zone

S127, 42922 A, 42923 (1 Clay, massive, brownish pink,
weathers crumbly

1985

z
2
te
<
=
a
&
4
a
<
ra)
Zz
<f
«
e—
a
&

II Clay, reddish ot base to buff and
sandy at top

Hee 5' Clay,sand ra
: Yy, y, gray, weathers crumbl aes

ov 21' Clay, sondy, massive, pink,
weathers to bluffs and steep slopes

lreaviar

1945
Approximate contact bottom of CRETACEOUS
draw on N.W. face of hill

Mean woter level of Missouri river before construction of Fort Randall dam i RO

‘mole ue Aq PdJESIPUT SI 9UOZ ,,9SO1 YOOI,, OY, “W2] Ie} ay) uO st [[tFT nolig you sy Ts
‘Opls JOATY WMossipy 94} Wo premjseo Suryoor ‘TITY noltg yinos ay} jo so"j ysamyisoU ay} UO saimsodxs jo ydesZoI0Ug °z “OI d

TH NOPIG HLYON

8u0Z 8804 ysoy a

UOIjDWIO4 ||DPUDY 4JO4 yo doy

f2lz

TMH Nora HiNnos

1967 SKINNER AND TAYLOR: BIJOU HILLS 11

some Upper Miocene forms... That some of these, in South Dakota at
least, are clearly Lower Pliocene deposits is indicated by the presence of
Ustatochoerus medius (Leidy) from one of the quartzite beds. The specimen
referred to was collected by Stevenson. He informs the writer that it
was found in place.”

The locality 5 miles north of Academy, South Dakota, cited by Steven-
son (1958, p. 137) carries the usual sequence of rocks and lithology ob-
served in the Upper Valentine and Lower Ash Hollow formations of
north-central Nebraska and south-central South Dakota. This locality is
also the eastern extension of the Bijou Hills, as pointed out above in
the present paper. The sediments exposed north of Academy, however,
are not to be confused with the older Fort Randall Formation exposed
on the more westerly south and north Bijou Hills bordering the Missouri
River.

STRATIGRAPHY

TERTIARY SYSTEM

FORT RANDALL FORMATION,! NEW

Type Section: The type section is on the south Bijou Hill in the
northwest corner of Charles Mix County, South Dakota, in the N. 1/2,
sect. 8, T. 100 N., R. 71 W. (fig. LA, C). When viewed from the river,
the slowly eroding exposures near the summit on the northwest face of
the hill make a prominent landmark (see fig. 2). The major outcrops are
in one small, cedar-covered wash which has a 5-foot zone of scattered
barite crystals, or “rock roses,” at an elevation of 1995 feet. The details
of the lithology of the Fort Randall Formation are described on the type
section (fig. 1C).

Discussion: Several writers? have observed two sets of Tertiary beds
in the Bijou Hills. We are naming the lower part (fig. 1C, 1945 to 2067
feet) the Fort Randall Formation, here considered a new and previously
unnamed set of Tertiary sediments.

The known thickness of the Fort Randall Formation varies from 60
to 120 feet and represents a cyclic type of deposition of siltstones and
clays and a secondary deposition of barite crystals in one distinctive
zone (“rock rose” zone on the section). The Fort Randall is uncon-

‘Named for old Fort Randall and the Fort Randall Reservoir on the Missouri River.

? The following have indicated that either two beds or two faunal units were present:
Hayden (1857b, p. 153), Agnew (1958, p. 129), Stevenson (1958, pp. 135, 137), Green
(1958, p. 142).

12 AMERICAN MUSEUM NOVITATES NO. 2300

formably overlain by undifferentiated sediments approximately equiva-
lent to the Valentine-Ash Hollow formations and rests unconformably
on the Cretaceous below. The Fort Randall Formation is, in reality, a
limited remnant of a former set of deposits of unknown areal extent.
The erosional hiatus between the Fort Randall and the Cretaceous indi-
cates an immense gap in time. The fauna from the Fort Randall sug-
gests that the erosional hiatus between the overlying undifferentiated beds
and the Fort Randall Formation represents a span of time partially
equivalent to that between the Lower Snake Creek beds and the Lower
Valentine Formation of Nebraska.

The lower part of the Valentine Formation as exposed in the drain-
ages of northern Brown and Cherry counties in north-central Nebraska
is composed of massive, loose, sandy deposits of well-sorted channel
sand with very little clay content, very different in grain size from the
fine sandy silts and clays of the Fort, Randall Formation. The 15-foot
bed of sand at the 2067-2082 level (fig. 1C), above the Fort Randall
Formation, is more like that of the upper part of the Valentine Forma-
tion in texture and appearance.

The fauna of the south Bijou Hill is extremely limited. From the Fort
Randall Formation, shell fragments of cf. Chrysemys (painted turtle) and
cf. Geochelone were obtained. From the “rock rose” zone of the Fort
Randall, the following were collected: lagomorph calcaneum; Eucastor cf.
curtus teeth; Peromyscus (Copemys) sp. ramus; Merychippus sp. maxilla frag-
ment with P*-M?; Merycodus necatus neotype and paratype, a partial ramus
and associated horn core; and from the talus of the “rock rose” zone,
a Skull condyle of a rhinoceros and a peccary-like canine.

A superficial examination of the color and texture of the Fort Randall
sediments suggests that they might be correlated with the Rosebud
Formation as defined by Gidley (in Matthew and Gidley, 1904, pp. 245-
246). Gidley did not designate a type section but stated, “The lower
formation above mentioned, for which we propose the local term Rosebud
Beds, is best exposed along the Little White River and in the vicinity of
the Rosebud Agency.” [Italics ours.] In the interest of nomenclatural sta-
bility, we select as the type section of the Rosebud Formation the only
set of exposures mentioned by Gidley that can be locally identified.
These exposures are in the vicinity of the Rosebud Agency buildings
on both sides of Rosebud Creek in the E. 1/2, sect. 34, and the W. 1/2,
sect. 35, T. 38 N., R. 30 W., Todd County, South Dakota, where some
100 feet or more of pink sandy sediments are exposed below overlying
Pliocene deposits.

The north Bijou Hill is in the center of the 8S. 1/2, sect. 30, T. 101
N., R. 70 W., in the southwest corner of Brule County, South Dakota.

1967 SKINNER AND TAYLOR: BIJOU HILLS 13

The bench mark on the top of this hill has an elevation stamped on it
of 2028! feet, to which we have tied our stratigraphic section (fig. 1B).

The fauna from the north Bijou Hill is even smaller than that from
the south hill. In the Fort Randall Formation, a partial skull of Myla-
gaulus sp. was collected 27 feet above the base (fig. 1B). Thirty-five feet
above the base of the Fort Randall a worn upper molar of Merychippus
sp. was collected; from the talus a calcaneum of Ustatochoerus sp., near
U. medius in size, was obtained.

UNDIFFERENTIATED VALENTINE-ASH HOLLOW FORMATION

The upper deposits on the south Bijou Hill (above 2067 feet, fig. 1C)
are approximately equivalent to the Valentine-Ash Hollow formations
undifferentiated. A broken P?, referable to “Hipparion” inasmuch as it
has an isolated protocone, was collected from the upper deposits 31 feet
above the contact with the Fort Randall Formation.

The exposures on the north hill are fewer and are not so large as
those on the south hill. We discovered on the north hill a loose sandy
channel deposit with coarse pebbles and fossil chips that can be allocated
to the distinctly Pliocene phase of deposition. A prominent hard out-
crop of the silicified quartz sandstone (orthoquartzite) is exposed on the
southeast side of the hill. One of these outcrops contained portions of a
proboscidean tusk which were left in the rock.” The quartzitic sand is
part of the channel deposit that extends westward to Carter, South
Dakota. The capping deposits are unconformable upon the earlier Fort
Randall Formation and have at least 50 feet of erosional relief, as il-
lustrated on the section (fig. 1B). This relief may also be interpreted as
large blocks of quartzite which have moved down the hill as a result of
gravity slump or soil creep. There are no distinct bedding planes at this
place to allow for an exact interpretation.

In the upper deposits from a channel of loose sand above the erosional
contact of the Fort Randall Formation, we collected a first phalanx of
Longirostromeryx or of a merycodont. From the so-called “Bijou Quartzite”
in the upper deposits we obtained a fragment of a proboscidean tusk.?

1See United States Geological Survey Iona Quadrangle, South Dakota, 15-minute
series (topographic), scale 1:62,500, edition of 1940, which gives 2037 feet. We do not
know the reason for this inconsistency. We simply note it.

2To date we have not found proboscidean remains in the Fort Randall Formation.
We have never found proboscidean bones in the Lower Snake Creek deposits of Sioux
County, Nebraska, though they are very common in the Valentine Formation.

3 Green (1958, p. 140) gave the location of the Bijou Hills as “Sec. 27, T. 101 N., R.
70 W., Brule County, South Dakota.” This is the eastern extension of the Bijou Hills
described by Hayden (1857b, p. 157) and not the source of Hayden’s fauna (see p. 4).

14 AMERICAN MUSEUM NOVITATES NO. 2300

SYSTEMATICS

CLASS MAMMALIA
ORDER CARNIVORA
SUBORDER FISSIPEDA
FAMILY MUSTELIDAE

SUBFAMILY LEPTARCTINAE
LEPTARCTUS LEIDY, 1857

Tyrer Species: Leptarctus primus Leidy, 1857.

Type: Left P*, A.N.S.P. No. 11293 (fig. 3).

Loca ity: It is not known whether the type was collected on the south
or the north Bijou Hill. The exact stratigraphic unit is also unknown.

Description: The type of Leptarctus primus, a P*, has characters which
Leidy (1857a, p. 311), believed were similar to those of P* of a coati-
mundi; therefore Leptarctus was placed in the family Procyonidae which
Bonaparte established in 1850. The P* compared well with the correspond-
ing tooth in a complete fossil skull described by Matthew (1924, p. 146),
who referred Leptarctus to the Mustelidae on the basis of characters re-
vealed in the more complete specimens. Matthew’s revised diagnosis
(1924, pp. 139-140) was based on two “neotypes,”! even though the
type was in the Academy of Natural Sciences of Philadelphia. One of
the “neotypes” (A.M.N.H. No. 18241) was a skull collected by Albert
Thomson in the lower part of the Lower Snake Creek deposits in Sioux
County, Nebraska. A photograph taken by Thomson? of this spot shows
that it is the same as the Frick Laboratory Trojan Quarry. The ramus
(A.M.N.H. No. 18240) is from Thomson’s Quarry “B.” In this case,
Thomson’s photograph shows that Quarry “B” is in a slightly higher
zone, which directly overlies the Trojan Quarry, and is synonymous
with the Frick Laboratory Far Surface Quarry.

The P*, as Matthew observed, has certain characters that are similar
to those of the procyonids, especially the well-developed hypocone. It
should be noted, however, that the protocone is also prominent and

1 Matthew followed Osborn (1918, p. 35) whose usage of neotype was expressed as,
“The specimen chosen in a subsequent paper by the original or another author as refer-
able to the same species and amplifying the characters of the type by affording fuller
material for description.”’ Matthew’s “neotypes” (A.M.N.H. No. 18241, a skull, and A.M.N.H.
No. 18240, a jaw) are here considered referred specimens of Leptarctus primus.

2 Albert Thomson personally gave a collection of labeled photographs of the Sioux
County area to the senior author.

1967 SKINNER AND TAYLOR: BIJOU HILLS 15

about one-fourth larger than the hypocone. The parastyle is a well-
developed production of the anterior buccal cingulum. We have veri-
fied Leidy’s and Matthew’s measurements of the type of Leptarctus primus,
which are as follows: anteroposterior length of crown, 7.4 mm.; width of
crown, 6.3 mm.; crown height, 5.2 mm.

Discussion: In 1869 Leidy supplemented his original description of
the type of Leptarctus primus with figures (Leidy, 1869, pl. 1, figs. 15, 16).
Matthew observed (1924, p. 139) that the figure of the type P* was inac-
curate. In reality the figure of the occlusal view is accurate, but it had
been placed on the plate in such a way that the hypocone appears to be

external

anterior

Fic. 3. Leptarctus primus, type, left P4, A.N.S.P. No.
11293, drawn from cast. Left: Lateral buccal view. X 1.
Right: Occlusal view. X 2.

the protocone. The figure was rotated 90 degrees, completely misorientat-
ing it in relation to all the other figures. The type of Leptarctus primus is
refigured in this paper (fig. 3).

Leidy’s type of Leptarctus primus cannot be assigned to any particular
late Miocene-early Pliocene temporal unit on the basis of stratigraphically
controlled specimens in the Frick Collection. Specimens that are referable
to L. primus are found in deposits ranging from the Lower Snake Creek
into the Upper Valentine Formation, but below the Burge Member in
north-central Nebraska.

ORDER PERISSODACTYLA
SUBORDER HIPPOMORPHA
FAMILY EQUIDAE
SUBFAMILY EQUINAE
HIPPODON LEIDY, 1854

Type Species: Hippodon speciosus Leidy, 1854.

Type: Right M, or M,, A.M.N.H. No. 465 (fig. 4).

Locatiry: Although it cannot be proved, the type may have come
from the south Bijou Hill, only because it has more and larger outcrops
than the r orth hill. Meek and Hayden, in the summer of 1853, made the
collection for Prof. James Hall of Albany, New York. The Hall Collec-

16 AMERICAN MUSEUM NOVITATES NO. 2300

tion was later transferred to the American Museum of Natural History.

STRATIGRAPHIC Unit: Unknown, but the type may have been derived
from the upper beds. The fossilization of the type of H. speciosus does
not resemble that of specimens which we have collected from the Fort
Randall Formation.

Description: (See fig. 4). The type of Hippodon speciosus (A.M.N.H.
No. 465) is a partially worn molar, nearly free of cement, of a mesodont
equid. Leidy (1854a, p. 90) observed that, “It appears to have had an
envelope of crusta petrosa as in the horse, though nearly all removed in
the specimen.” It is possible that some earlier worker removed the ce-
ment, for a vestige still remains on the internal part of the postflexid. On
the buccal side the enamel surface is finely crenulate. On the lingual side
the enamel surface is smooth. The anteroflexid is shallow and does not ex-
tend to the base of the tooth. The postflexid is deeper than the antero-
flexid and extends nearly to the base. The ectoflexid (between the proto-

Fic. 4. Hippodon speciosus, type, M, (or M,),
reversed, A.M.N.H. No. 465; from the Bijou
Hills. Left: Occlusal view. X 1. Center: Anterior
view. X%. Right: Buccal view. XA.

conid and the hypoconid) is deep and separates the metaconid-metastylid
column almost to the base. The anterocingulum is transversely wide at
the base and extends obliquely upward, narrowing at the crown, whereas
the hypoconulid is not so transversely wide at the base and extends more
directly up to the crown. A distinct posthypoconid groove extends half-
way down the crown to the base, and a distinct postentoconid groove
extends from the crown half of the length of the tooth. The combination
of the two grooves sets the hypoconulid off as a distinct entity on the
posterior side of the tooth. The following measurements were made of
the type (A.M.N.H. No. 465): length of crown (paralophid to hypo-
conulid), 18.4 mm.; width of crown (protoconid to metaconid), 8.4 mm.;
greatest length at base of tooth, 14.6 mm.; greatest width at base, 10.7
mm.; greatest height of tooth, which is worn (metaconid to base of pro-
toconid), 23.0 mm.

Discussion: The first North American equid genus to be named was
Hippodon, but unfortunately the single lower molar that is the holotype
of the type species, Hippodon speciosus, has characters that are generally
common to several equids. Leidy (1856, p. 311) indicated that he, too,
was puzzled by the relationships of Hippodon when he transferred it to
Hipparion, thus: Hipparion (Hippodon) speciosum. Later Leidy (1869, p. 282)
referred certain specimens to Hipparion speciosum that Hayden had col-

1967 SKINNER AND TAYLOR: BIJOU HILLS 17

lected in the Bijou Hills in the autumn of 1856, and these were the
specimens that Hayden (1857b, p. 157) had misidentified as “Hipparton
occidentalis.” Leidy (1869, pp. 320-321) specifically retained, however, the
identity of the name Hippodon speciosus for the type in text and plate
(pl. 19, fig. 23). |

Gidley (1907, p. 877) could not find the type and assumed that, since
it was a lower molar, “. . . [it] would probably show no distinguishing
characters, even if located.” He cited both the genus and species as in-
determinate. Osborn (1918, p. 129) reproduced Leidy’s 1869 figure of the
type, but followed Gidley by citing Hippodon speciosus as “Gen. et sp.
indet.” and then placed it as “. . . intermediate between a small Mery-
chippus and a small Protohippus.”

The genus Hippodon is based on a nominal species and as such is
nomenclaturally valid. Quinn (1955, pp. 13-17) went one step farther
and treated Hippodon as a diagnostically useful genus. He gave a de-
tailed description of the type of Hippodon speciosus and referred a low-
crowned upper dentition from Texas and Florida to it. In so doing, he
inferred diagnostic characters for the type which could be seen only in
the referred examples. The characters limited to the type of Hzppodon
speciosus may be associated with Parahippus, Merychippus, Desmatippus, and
other groups. The single molar may be matched in several disparate
groups of equids.

We contend that it is not in the interest of nomenclatural stability to
resurrect Hippodon. We prefer to leave it as a monotypic genus, nomen
dubium, with a single species standing for its type [Stoll and others, 1961,
Articles 42 (b) and 45 (b)]. Because the type species of Hippodon occurred
in the same area as that of Merychippus, there is the possibility, which
we cannot prove, that the two genera are synonymous, and that Mery-
chippus is a junior synonym.

Hipparion occidentale Leidy, 1856

Discussion: The species Hipparion occidentale was based on a syntypic
series of “five superior and one inferior molar teeth” (Leidy, 1856, p. 59).
Leidy (1869, p. 281) acted as his own reviser and fixed the type when he
stated: “The teeth consist of four upper molars of the right and one of
the left side, and are represented in figures 1-5, plate XVIII. They were
accompanied by a fragment of a last upper molar and a lower molar,
apparently of another equine animal.” We interpret this as a valid re-
vision not subject to change by Gidley (1907, p. 877) who designated a
right P?, P3, M?, and a left P? as the type.

18 AMERICAN MUSEUM NOVITATES NO. 2300

There would never have been a question but that the Little White
River was the type locality for Hipparion occidentale if Hayden (1857b,
p. 157) had not written: “In the autumn of 1856 I discovered on the
denuded summits of the same hills [Bijou] Hipparion occidentalis . . . These
remains have all been described by Dr. Leidy in the Proceedings of the
Philadelphia Academy.” It is evident that Hayden could not have col-
lected the type of H. occidentale some six or eight months after Leidy had
described it. Hayden misidentified these specimens, a fact that is clearly
shown by Leidy’s statement (1869, p. 282): “Shortly after the discovery of
the teeth of Hipparion occidentale on White River, Dr. Hayden obtained
several specimens of equine molars at Bijou Hills, situated east of the
Missouri River, below the outlet of the White River. In a notice of these
fossils in the Proceedings of the Academy for 1856, page 311, they were
referred to a species under the name of Hipparion speciosum.” Note that
Leidy, in his 1856 publication, had used “Hipparion (Hippodon) speciosum.”

MERYCHIPPUS LEIDY, 1857

Type Species: Merychippus insignis Leidy, 1857.

Type: Merychippus insignis is based on a fragment of a right maxilla
with dP?"3, A.N.S.P. No. 11276 (fig. 5A, B; measurements are given in
table 2).

Loca.ity: Collected by F. V. Hayden in 1856 from the Bijou Hills,
South Dakota.

StraTicraPpuic Unit: The fossilization of the type is similar to that of
the neotype of Merycodus necatus from the Fort Randall Formation, but
no positive allocation of level can be made.

Known Liruic, GEOGRAPHIC, AND TEMPORAL OCCURRENCE OF SPECIES:
From the Fort Randall Formation, Bijou Hills, South Dakota, late
Miocene or very early Pliocene; from the Lower Snake Creek beds! in
Sioux County, Nebraska, late Miocene; Pawnee Buttes area in north-
eastern Colorado, late Miocene; Trinity River Quarry, 7 miles north,
40° east, of the town of Coldspring, San Jacinto County, Texas, late
Miocene.

STATEMENT: The holotype of Merychippus insignis has long been an
enigma, because it is fragmentary and immature, and the over-all char-
acters of the two upper milk teeth are similar to those of many im-
mature equids in the late Tertiary. A detailed study reveals additional
characters, however, which restrict M. insignis to one taxon in the late

' Merychippus insignis has not been observed in the Sheep Creek beds as defined by
Matthew and Cook (1909).

1967 SKINNER AND TAYLOR: BIJOU HILLS 19

Miocene. We did not discover a match for the holotype at the type
locality in the Bijou Hills, but we did obtain a mature partial maxilla
with left P4-M2, F:A.M. No. 87011 (fig. 6A, B), an isolated M}, and
another badly worn molar from the Fort Randall Formation, which are
referable to Merychippus insignis. The fossilization of the type seems to
match that of other specimens collected from the Fort Randall Forma-
tion. Future collecting from this locality may produce a more complete
immature match for the holotype collected by Hayden.

Failing to find adequate topotypic material, we concentrated on
matching the type specimen with examples from an approximate tem-
poral unit. We were confronted with two, and possibly three, sets of de-
posits at the type locality, any one of which might have been the source
of the type. The large, stratigraphically controlled Valentine-Ash Hollow
collections in the Frick Laboratory have not produced an immature
example that can be matched exactly with the type of Merychippus msignis.

The type can be nearly duplicated in the fauna of one quarry in late
Miocene deposits in western Nebraska. A series of specimens ranging
from immaturity to old age was found in the fauna of Echo Quarry in
the Lower Snake Creek channel deposits of Sioux County, Nebraska.
These specimens demonstrate the skull and upper dental characters of
Merychippus insignis.+

The Echo Quarry deposits in Sioux County extend for about a quarter
of a mile along the bottom of an east-west trending draw, known in the
Frick Collection field records as “Antelope Draw.” The exact location
is in the SE. 14, sect. 33, and the SW. ¥/, sect. 34, T. 26 N., R. 55 W,,
Sioux County, Nebraska. The Echo Quarry fauna belongs to the Lower
Snake Creek biostratigraphic unit. We consider the Lower Snake Creek
beds of Matthew (1924) as an independent temporal and faunal unit
lithologically distinct from the Sheep Creek Formation of Matthew and
Cook (1909). By 1918 Matthew had profited from the research (then
unpublished) of Dr. J. C. Merriam on the later Tertiary mammalian
faunas of the Pacific coast. These data added largely to the correlation
studies of the faunas of the Plains. Matthew (1918, p. 184) stated that
Merriam’s studies “. . . afforded evidence strongly suggesting that the
Snake Creek fauna might in fact be a composite and not all of the same
age.” However, in his revised faunal list of the Snake Creek (1918, April,
pp. 185-188), Matthew made no faunal distinctions. Osborn (1918, June,

1 Although there are lower dentitions from Echo Quarry that we suspect might be
referable to M. insignis, we are not presenting the diagnosis of the lower jaws, because
of the lack of articulated upper and lower dentitions.

= 7 == = = = = = (0°01) i — 1YSIOyY
= = aad en = no cel C6l LLI 7 0'8T WIPIM
= = = = = = G'8T 661 9°61 = 8°61 yisusT
edP
a ac = oY Cs =— 7 = i re + 1ysI9H
7 = ae _ a = - Ei 691 ir o 91 PIM
ay hay ra a r = 8°E?N i A (0'92) = = yisusy]
zdP
aa: — = ae = = “7 88 = on = 1YSIOH
— aL £8 re 08 = = £8 — — — YIPIM
— o Sl OFT = OFT :— a! OST = a ris ysueT
1dP
c Olt 6°801 0'0eT O'6IT O'LTT VII 9811 = = ‘— me eW-ed
O'S 00S F9¢ CPS 1°GG 6'LS 69S = = =) = e-1N
0°09 c'6S 9°€9 0°S9 v'E9 L°¢9 L 19 cv im “F = ¢-2d
- = Mi rm = = r19 L'¢9 (0°49) rs i. +-edP
— _ 7 - — = OLE 88S OLE = 8°8E e-zdP
yisuay]
08 c9 BL BL . a a p(G°9) a = oa YIpIM sutuey
+ 4+ °sihk 7 + ‘SIA F + ‘SIA —'sik & + sik ¢ — sik ¢ =p sthk 7 IAT + ‘OuL Z + "sik ¢ ourg = =-seak ‘YUOUL :a8y
(9XLM) (-LM) = (+9M) (9M) (9M) (—9M) (GM) (6M) (-~M) — (9M) (bMé) p18aM Jo 281g
£ GOOL8 5 POOLS £ €00L8 £ ZO0L8 £ 10048 80048 LO0L8 £ 0008 90048 gl 10L8 v9Lell
‘ON “ON ‘ON ‘ON ‘ON ‘ON ‘ON "ON ‘ON ‘ON ‘ON

WV Wv-d WV WV Wv-d WV WV Wd WV WVd ‘dS NV

FIVNA J INV FIV “OV CIO OL HINOAX WO ‘VASVAGAN ‘ALNNOD XNOIg aNV
‘VLONV HLNOg ‘stu NOlig AHL woud studisur sndqrysGapyy 410 HAT], WAdd~) AO (SYTLAWITTIJ NI) SLNAWAUNSVAW

6 ATAV.L

— = GG 0& 0& Of O& O& —_ = — oInjeAING)

= = — — — = 1 74 Cv — VVC = IYSIOH
0°66 C61 8°8T OT? 8°81 LSI 9°61 £02 = csi = YIPIM
c9l 8°91 G6! £81 8°81 10é 0°02 OT? = c6l = yisueT
IW
= 0& Gc Lo ve 6b Gé 7 — rs = ainyeainy)
“ad 7. 7 on i BGC OLSY TT =, 8°8z a WYsIOH
6° 6é 606 602 OT? CsI 0°02 6°61 Lol " OT? 7 WPIM
€8l SLI v6! L’6l ¢6l 8°02 9°61 0°02 = 6'8¢ -— yisuay
rd
is GG GG GS GG Lo FE a = -s = ainyeainr)
=e i al = oa €°G¢ LS = = = = WYSIIP
GGG v6! 102 8°02 Lst T'¢ ¥ 02 L’6l i = arg YIPIM
Let CLI 0°02 F 02 G02 Ol? c 61 OTS = Pa = yisuay
ed
= Gc £6 GG GG ce == ar >= = — goIn}BaIny)
hon = a ft [. OIG 0°6é 7 = om WYSIOF]
a 691 Ls G8I OLI G61 Gs OLT in +e a WIPIM
G'S? 6 1c € ve 8°b6 G86 GES 8°66 L°&% = = = yysuay
; id
- _ + = = re or ma (001) = = ys
— — os 3 = = 9°81 L 61 V8 —s = WIPIM
a => i = = a 9°81 0°02 9°61 — sy yisuay
-dP
ee en ee re Se TL ee ge ———
2 €00L8 5 $0048 £ §00L8 ? Z00L8 ? 10048 80048 L00L8 £ 00028 90028 ql 10Z8 v9Lell
‘ON ‘ON ‘ON ‘ON ‘ON ‘ON ‘ON ‘ON ‘ON ; ON ae

WV'4 WV Wv'd Wv'i4 Wv-i Wv-d Wv-d WV WV WVvd ‘dS NV
eee

(panurju0))—Z 1942L

(99 ‘8y) xepul 94) Surutuejap 10j
sajejduray ind aaey 9A4 “YI00} sy) JUOposdAy aio PUL JOIYSrerIs sy} “KapuT oy} saySIy sy], ‘OANIVAIND Jo xapuT sy} st ‘sIaJ9UTTT]IUA UT passoudxa ‘oJouTO
yey) Jo sniper oy} jo wSua| 9YT, ‘epoulo w Jo yed st sJAIsOsauT DY} JO BAND oY], “9JOITO B JO SNIPeI 94} se passoidxa st oinyeaimd Jo xopul 24],

‘paydnioun JO UsxOIG SEM Y}00) JY} JVY} oJeOIPUT JUSUTEINSeoU B pUNoOIe sassy Uae] ,
“UIOM [JOM ST pA, PUB ZW Ysnory? Yo wom st usoyyed ay} 1VY} soywoIpUT ‘(9x/M) :snu) ‘x uy
‘TEOM UIOS SMOYS PUL Pajdne st SOLOS OY} Ul YOO} IxX9U BY} PUL TOM ]fOM ST YOO} 94} JO JTeY Jous}sod oy} yey) sayeotpur usIs snjd y ‘WioMuN jreYy
sowajsod {Jy = (—GM) “82 ‘uomun st y00) sy) Jo Jey soia}sod ay} yey) sayeorpul usts snutu Vy “gW ‘(ZM) ‘2 “OM) SW (GM) ‘sdP “(bM)
fodP “(GM) ‘222M Ur 2AdP “(ZM) ‘288 ITM payepeii0o st 7 ‘rea UT ST Saas [BIWApP 9y} UT YICC} YIYM MoOYs 07 PostAdP sem IeIM Jo a8vIs 94],
“eyseaqan ‘Ayunor xnorg ‘Auien() oyoy Woy sures Jequinu angsoreyes “PW: ue YIM pajst]
sugurioads 19420 TTY ‘STITH nofig oy} 3 UONeULIOY [[epuRY 307 oY) Wo psy2]]09 sem CW pue SW fd Wim eyixeur ys & “T1028 “ON (WV ¢
‘SITE nolig ay wos seusisu sndgrysGapy Jo odAy oy) St ‘p gdP YUM T[XeU IYI & Jo JUSUIsEYy & ‘Q/ZIT ON ‘d’'S'N'We
I

= 6c GZ co O& 66 ae 7 = = ae ainyeaAIny)
— csi C'E% €UC VIZ 6'S% = a = = a WYSIOH
CsI O'LI CLI O'LI cl (O'L1) O'LT — = == a PIM
SLI ¢91 VLI LL €'81 81 0'6I = — — — yisuery
eW
a = GG Of Of Lg O& Ge a - ce amnyeainy)
— — — = = 0'cZ 6'LZ = = VL2 = WSIIH
0% 0'6I O'sI 0°02 CLI € 81 L'61 C6l = 102 “of WIPIM.
9°91 VLI O'6! ‘81 C'6l 61 0°02 0°02 = Cél = yisueT
ZW
9coo.g P0048 E0048 20048 ? 10028 80028 LOOL8 £ 000L8 90028 gl 10L8 r9lzll
‘ON “ON ‘ON “ON ‘ON “ON ‘ON “ON ‘ONT ‘ONT ‘ON

WV WV: WYv'd Wve WV WV W'v-d WV: WV:d WVd ‘dS N'V

(panurquo))—z 21921,

‘ayeuntxoidde ase sasayjuered Ul pasopus s}usUTaInsvay] 5

"29 33000] ‘Z aIqQ¥") 99 ,

00L OL 0°02 O81 OSL OSL jesourenbs uo zed 3sapimm ‘uintuein
O°ES 02S 0°9S (0's) O19 (0'29) (APIA [eNqioysod ysomoueu “cunyuerc)
Oe a OFIT 0'6<dT (0'0€1) O'8TT (O'9IT) YIPIM asioasued ysoyeais “yore oeUIOSAT
oor 0°601 O'FIT (0°21) OFT (0911) vurod ysoprm “YIPIM Terqr0Isog
08 COL 0°98 (9'98) O'L8 (O'LZ) quiod jsomolseu 18 YIPIM [elIqIo JOUNIUYy
OF CLE OF 0'0F OIF OES e[[rxeuroid uo yurod jsapim ‘soreu JoUaUYy
0°6Z 0'1Z 0'8z 0°92 0°82 0'&Z apis [eyeyed uo uinsjsol Jo yred JsamoueN
0°66 0'0F O'S? 0'0F OCF 0°9¢ (19psiog IepOoale 1e a]ZznNUI jo YIPIA) gf O} oJ
HLAIAA
O'1S OIS 0°ZS 0'8F 0'6r 0'0S zd Jo Joluajue 0} ,] Jo s0119)sod ‘eurayseiq
O'0L OL 0°69 0°69 0'0L 0'0L JapIOg IAISIOUI 0} ,g JO JOLIOIUY
0°22 ae 0'9FZ — 0°€EZ — ainssy outjyeyedisod 0} umnuseur usureloj jo 10Ma}UYy
O'12ZZ = 0'°SEz — 0'91z — zd jO JOLIojUe 0} UINUSeW UsUIeIOJ Jo JOLIZ]UY
O'19I —- OFLI = O'ESI — zd JO A01a}sod 0) urnuseul usurel0j Jo OWa}UYy
OLLI = OFIT — 0°66 _ eW jo J014193s0d 0) wnuseur usureioy Jo JoUa}UYy
O'2ST = O'8ST = 0O-9OFT => Japsiog aurjyeyedjsod 0) umnuseul usuleloj Jo 10LWIa]UYy
0°€8 a 0'Z8 = 0°99 — JOWOA 0} WINUSeUT USUIeIOJ Jo 1OLIO]UY
(0-221) (O'LFI) O'ZLI — O'OLI O'EST BININS [VJUOT-Oseu 0} 3soI9 [eVd1I9I9Q
(0'26z) (0122) 0'16Z = 0'16z O'ILZ yojou Areyixeursid-oseu 0} 3sa10 yeyd1I099
(o'1z€) (0°662) 0'0¢E = (021g) = (uerpew) sfeseu jo dry 0} ysau0 [eydi9909
(O'8TT) = O'9IT = O'FIT 0°801 (jerues9) andtos0ysod 0} y1qG10380g
0'61¢ 0'8T¢ 0'9E¢ 0666 0°66 0°S0Z ([e19eJ) Jopsog eAtstour JOL9}Ue 0} IIqGI0I1s0g
(o'8ge) g(O'ZE8) O'ZSE = OSE O'ZIE (xoq19A) Japiog 9AIsIoUT JOLIA\Ue 0} yndIDIQ
0262 —_ 000 — 0°98z — (ae]Iseq) Jop1o0q oAISIOUT 0} UIMUSeU UDUIeIO]
NVIGAJJ
(9x2M) (—LM) (+9M) (9M) (9M) o(PM)
£ GOOL8 5 ¥00L8 2 S00L8 £ 20048 2 10048 £ 00048
ON WWE ON WW ON WWE ON WV ‘ON ‘WV ‘ON WV

(‘suondofoid teuosoy zo oe syUusWOINseoUI oy, ])
studisua sndqdiysGapy a0 (SAALAWITTIJ NI) SLNAWAWASVAYY TInAg

€ TTAVL

24 AMERICAN MUSEUM NOVITATES NO. 2300

-
@
<J
i
o
a
°

aod
-]
@
wo

a

Anterocone

Fic. 5. Merychippus insignis. A, B. Type, dP?-3, reversed, A.N.S.P. No. 11276;
from the Bijou Hills. A. Lateral buccal view. <4. B. Occlusal view, drawn from
cast. X 1. C, D. DP?4, reversed, F:A.M. No. 87006; from Echo Quarry, Sioux
County, Nebraska. C. Lateral buccal view. X'%. D. Occlusal view. Note cement
on dP*. <1. E. Left dP'!* and Mt}, occlusal view, F:A.M. No. 87000; from
Echo Quarry, Sioux County, Nebraska. Arrow at anterocone, new name. X 1.
F, P?_M&, reversed, occlusal view, F:A.M. No. 87002; from Echo Quarry, Sioux
County, Nebraska. Arrow at pseudoparastyle, new name. X 1.

1967 SKINNER AND TAYLOR: BIJOU HILLS 25

He! t36mmradive radius

: jeune index i on Weurvatore index se.

Fic. 6. A, B. Merychippus insignis, P*-M?, reversed, F:A.M. No. 87011; from
the Fort Randall Formation, south Bijou Hill (fig. 1C). A. Occlusal view. x 1.
B. Sectioned 10 mm. below crown. X 1. CG. Diagram for method of determining
the index of curvature of upper dentition of the Equidae.

pp. 23, 28) pointed out that the Snake Creek beds contained at least
two faunal levels of upper Miocene and lower Pliocene age and a third
from “Upper Pocket Levels” which yielded the type of Pliohippus lecdyanus.
In 1924 (pp. 64-68) Matthew gave a faunal list in which the Lower Snake
Creek, the Upper Snake Creek, and the Sheep Creek appeared as distinct
faunal units for the first time.

Matthew and Cook (1909, p. 386, fig. 11) listed four upper dentitions
which they considered matched the type of Merichippus insignis “fairly
well” and also exhibited the mature characters. The immature specimens

26 AMERICAN MUSEUM NOVITATES NO. 2300

Si} ves

ee ea

Fic. 7. Lateral views of Merychippus insignis from Echo Quarry, Lower Snake
Creek beds, Sioux County, Nebraska. A. F:A.M. No. 87000. B. F:A.M. No. 87001.
C. F:A.M. No. 87004. D. F:A.M. No. 87005. All approximately <%.

‘TX ‘uonnuap seddn jo mara esnp99 “g °% xX “MIIA [eIO}e7] “YW
“eyseiqon ‘AiUNO: xXnOIg ‘spaq yaaI-) syvug JaMOTT ‘ALENG) OYDy wo {100L8 “ON “W'Wig ‘SeuFisue sndgiyotiapy -Q “ot
YSeIqeN 2) IS. SPSq PMD SBS I i J L W'v-d fystiayy °3 “OT

28 AMERICAN MUSEUM NOVITATES NO. 2300

(A.M.N.H. Nos. 14010 and 14014) do not match the type of M. msignis.
The mature examples (A.M.N.H. Nos. 14003 and 14001) do not agree
with our series of skulls which range from immature (duplicating the
type) to adult individuals. Matthew (1924, p. 159) referred the bulk of
the equids from the Lower Snake Creek faunal level to Merychippus
paniensis. Two skulls were figured (fig. 47, AM.N.H. No. 18297; fig. 48,
A.M.N.H. No. 18299) as typical of M. paniensis, but these do not agree
in dental pattern with Cope’s lectotype of M. paniensis, A.M.N.H. No.
8249 (fide Gidley, 1907, p. 890). The partial female skull, A.M.N.H.
No. 18299, figured by Matthew (1924, fig. 48) is here considered a good
example of mature Merychippus insignis.

Examples from youth to old age, male and female, of M. msignis are
presented in this paper (figs. 5, 7) to illustrate the uniformity of dental
and facial morphology. These specimens indicate that the morphology
of the preorbital fossae is as genetically fixed as that of the dentition.
Within one quarry sample of a specific population the preorbital fossae
do not exhibit wide ranges of individual variation because of sex or age,
_ but are part of a set of morphological characters that remain constant
with the dentition. The lateral views (fig. 7) of a colt, a young male, a
young female, and an old male show the general outlines of the skulls
of Merychippus insignis. Nearly uncrushed skulls were selected in order to
show consistency in measurements, placement of preorbital fossae, crown
height of cheek teeth, and so on. As noted in table 2, it is not possible
to give the crown height of the entire dental series of one individual.

In table 2 the order of tooth eruption and wear of M. insignis may be
compared with that of modern Equus (Sisson and Grossman, 1953, p.
405). We believe that the dental development of M. insignis proceeded at
approximately the same rate as that of living Equus. The optimum dental
pattern is developed shortly before the third molar is fully worn, after
which the dental pattern changes rapidly on the preceding teeth (see
fig. 5F). The normal life span of Merychippus is presumed to have been
in direct ratio to the crown height of the cheek teeth. Since Merychippus
was mesodont, its life span would presumably have been shorter than
that of Equus, which is hypsodont.

IMMATURE Upper Dentition oF Merychippus insignis

Table 2 and figure 5 supplement this description. The holotype,
A.N.S.P. No. 11276 (fig. 5A, B), consisting of dP?-3, was compared with
immature dentitions from Echo Quarry: F:A.M. No. 87006 (fig. 5C, D),
a right maxilla with dP?-4 very slightly less worn than the type; F:A.M.

1967 SKINNER AND TAYLOR: BIJOU HILLS 29

No. 87009, a right maxilla with dP2-4; and F:A.M. No. 87010, a left
maxilla with dP?~‘. All nearly duplicate the wear of the holotype. F:A.M.
No. 87000 (figs. 5E, 7A) is an immature partial skull lacking the decid-
uous incisors. The left maxilla of the skull is complete, with deciduous
premolars about to be shed, M}, and the unerupted germ of M?. This
skull (F:A.M. No. 87000) serves as a link between the deciduous denti-
tion exhibited in the holotype and referred immature examples and the
mature dentition of M. insignis.

The milk premolars of the holotype (A.N.S.P. No. 11276) and the
referred examples, F:A.M. Nos. 87006, 87009, and 87010, are in an
early stage of wear with well-defined crown patterns. The dP?* of
F:A.M. No. 87000 show a slightly advanced stage of wear. The referred
examples all have the anterocone,! new name, preserved, which is broken.
off in the type. The deciduous premolars are low-crowned, with the
ectoloph and the hypostylar region of dP* usually coated lightly with
cement. Occasionally there are traces of cement on dP? and, more rarely,
on dP?. The milk teeth of the holotype are devoid of cement. On the
buccal side of the milk premolars there is an ectocingulum which is con-
fluent with the mesostyle and metastyle. There are fairly prominent
anterior and posterior ribs. On the lingual side there is usually an in-
cipient style or slight swelling around the base of the enamel anterior
to the protocone or at the base of the medivallum. This may or may
not be present.

pP?: The mesostyle is prominent, but the metastyle is weakly defined.
A crescent-shaped protoloph with one small pliprefossette is united to
the anterior portion of the base of the paracone and to the posterior
part of the protocone near the crown of the tooth. Posteriorly the pro-
toloph has a slight elongation which might be interpreted as an antero-
crochet, but this does not unite with the metaloph. The protocone is
round, placed antero-internally to the mesostyle, and is united to the
posterior part of the protoloph by an enamel isthmus or pliprotocone.
The metaloph unites to the ectoloph on the posterior part of the paracone,
anterior to the mesostyle, and has a distinct crochet, suggesting a pli-
caballin. This crochet bifurcates lingually into the medivallum and in-
ternally into the prefossette toward the most posterior extension of the
protoloph. On the anterior side the metaloph has two small enamel
plicae entering the prefossette between the crochet and ectoloph. On the
posterior side there is a cluster of three small enamel plicae extending

1 The anterocone (fig. 5E) is a conelike structure developed on the dP? of equids be-
tween the true parastyle and the paracone.

30 AMERICAN MUSEUM NOVITATES NO. 2300

into the postfossette near the junction of the metaloph and ectoloph.
Posteriorly the metaloph is united to the hypocone slightly below the
crown and is confluent with the hypoconule or hypoloph, or with both.
The metaloph is united to the internal posterior base of the metacone.
There is one small plihypoloph extending into the postfossette. The
hypostyle is small, distinct, and united anteriorly to the metaloph. It is
developed from the cingulum on the posterior median side of the tooth.
Anteriorly the hypostyle unites with the metaloph and forms the internal
wall of the hypoconal groove which extends to the base of the tooth.
There is a weak cingulum extending from the crown of the hypostyle
to the base of the hypoconal groove. One small plication extends from the
hypostyle into the hypoconal groove.

pP?: The parastyle is prominent and overlaps dP; the mesostyle is
also prominent, and the metastyle is distinct but small. A crescent-
shaped protoloph with one very small plica on the antero-internal side
is united to the internal portion of the parastyle and posteriorly to the
protocone near the crown of the tooth. The posterior part of the pro-
toloph extends slightly beyond its junction with the protocone but
differs from dP? in that no plication is developed.

The crown pattern of the protocone is tear-shaped, lingually rounded,
and narrows toward its junction with the protoloph. There is a marked
preprotoconal groove. The metaloph is crescentic, confluent with the
hypoloph, and is united to the ectoloph slightly anterior to the meso-
style. It is also united to the posterior part of the paracone. On the
anterior side a prominent crochet bifurcates the medivallum; the lingual
fork produces a small plicaballin, and the buccal fork tends toward a
union with the posterior part of the protoloph. There are two prominent
internal plicae extending into the prefossette, the lingual plica extending
toward the posterior extension of the protoloph and the more buccal one
into the prefossette. On the posterior side the metaloph has a cluster
of three small plicae extending into the postfossette toward the meta-
cone at its junction with the ectoloph. Posteriorly the metaloph is united
to the hypocone slightly below the crown of the tooth. The metaloph
is confluent with the hypoconule or hypoloph, or both, the hypoloph ex-
tending posteriorly and externally to unite with the midposterior wall
of the tooth. There is one small plihypoloph extending forward into the
postfossette. The hypostyle is small but distinct and unites anteriorly to
the metaloph. It is situated on the posterior side of the tooth. The
anterior extension of the hypostyle and its junction with the metaloph
form the internal wall of the hypoconal groove which extends to the
base of the tooth. There is a distinct lingual cingulum extending from

1967 SKINNER AND TAYLOR: BIJOU HILLS 31

the crown of the hypostyle to the base of the hypocone. Two plihypo-
styles are present in the hypoconal groove.

Mature Upper DeEntitTion oF Merychippus insignis

There are 47 mature dental examples of M. msignis from Echo Quarry.
The following description is supplemented by figures 5F and 8B in which
the ultimate in dental complexity for this species is expressed in the
occlusal patterns of F:A.M. Nos. 87001 and 87002. The dental formula
of Merychippus insignis is 3, C, dP!, P?-M3. In the Equidae we have
never observed a replacement for the first premolar that erupts with
the milk dentition. Therefore, we refer to this tooth as dP! in order to
distinguish it from the permanent premolars. In Merychippus insignis the
dP! is double-rooted. In later equids the dP! is single-rooted, peglike,
and frequently absent. The mature cheek teeth of Merychippus are meso-
dont, moderately curved (table 2, footnote e; fig. 6C), and cement-
covered, except the centers of the fossettes which are not, in every case,
completely filled.

There are no examples of I+? from Echo Quarry, but I? has a thin
layer of cement in the bottom of the cups. Canines, which are heavy in
males and smaller and lighter in females, are covered with enamel caps.
Cement covers the external surfaces of P?-M? but does not invariably
fill the postprotoconal valley and the hypoconal groove in P?-M?, Cement
partially fills the fossettes. Deciduous premolars two and three are usually
cement-free.

The greatest degree of dental complexity is observed on the occlusal
pattern just before the M? comes into wear (fig. 5F). After the M® is in
wear, the occlusal pattern of the cheek teeth becomes increasingly sim-
plified as the result of attrition. The numerous plications that develop
at right angles to the plane of the metaloph are leaflike outfoldings which
change rapidly as the tooth wears.

The parastyle of P? is centrally placed and elongate, with a small
pseudoparastyle, new name? (fig. 5F), posterior to it. The parastyles of
P? are large and overlap the metastyles of the preceding teeth. On the
molars the parastyles are not so large and do not overlap the preceding
metastyles as much as in the premolars. The mesostyles of the premolars
are heavier than those of the molars. There is a prominent anterior rib,
and a weak posterior external rib on P?. The external ribs of P3-M3
are all weak.

1 Pseudoparastyle refers to a fold that develops on the ectoloph posterior to the para-
style and anterior to the paracone.

32 AMERICAN MUSEUM NOVITATES NO. 2300

On P? the protoloph unites with the ectoloph posterior to the para-
style. Toward the base of P?, the protoloph is united to the true para-
style by a lingual cingulum that produces a small parastylar fossette.
On P3-M3 the protoloph is united to the parastyle. The pliprotoloph is
weakly developed on the crown of the premolars, is seldom observed
on the molars, and disappears altogether toward the base of the tooth.
There is a weak pliprotoconule on P? which becomes increasingly
stronger on P®-M®. The protoloph tends to remain separate from the
metaloph on P?, but on P?-M® the protoloph is united to the metaloph
by plicae that originate from a crochet.

Early in wear the protocones of P?-M® are free, narrow, and elongate,
with prominent plicae on the antero-internal borders. Later in wear
these plicae unite the protocone to the protoloph. This is in antithesis
to the type of Merychippus paniensis which has a rounded protocone with
no pliprotoconal folds. In M. imsignis the occlusal pattern of the pro-
tocone becomes rounded with wear. There is a shallow protoconal cleft.
In P?_M? the protocones and hypocones are subequal, and in M® the
hypocone is reduced. The protocones are either slightly anterior to, or
opposite, the mesostyle, with the long axis nearly parallel to the antero-
posterior length of the tooth. The protocones tend to be more posteriorly
situated on the molars than on the premolars. There is a well-defined
preprotoconal groove, and a deep postprotoconal valley.

The hypocones are united to the metaloph and are long and narrow
on the crown but are oval-shaped or round toward the base. The axis
of the hypocone is nearly parallel with the anteroposterior length of the
tooth and is at right angles or oblique to the axis of the metaloph. There
is a weak prehypoconal fold, and there is a deep posthypoconal groove
which closes near the base of the tooth, producing a hypoconal fossette
observed only on a well-worn tooth. On P?-M? prominent lingualpli-
hypostyles extend into the hypoconal groove and disappear toward the
base.

The main portion of the metaloph is anterior to the hypocone and,
on the crown, unites with the ectoloph near the mesostyle by a junction
with a medicrista. With wear the medicrista becomes confluent with the
metaloph. In early wear the borders of the fossettes are moderately com-
plex. There is a well-developed plihypoloph into the postfossette, and
in some specimens there is a very small shallow hypolophal fossette in-
ternal to the hypoconal groove (fig. 8B, M1 of F:A.M. No. 87001). A
caballin fold is produced by a crochet that develops from the metaloph.
In the premolars it is bifurcate. In the molars the caballin fold is usually
single. |

1967 SKINNER AND TAYLOR: BIJOU HILLS 33

Mature SKULL oF Merychippus insignis

There are five mature skulls of M. insignis from Echo Quarry: F:A.M.
Nos. 87001 (figs. 7B, 8), 87002, 87003, and 87005 (fig. 7D), males; and
F:A.M. No. 87004 (fig. 7C), a female. The measurements of these five
mature skulls and those of one immature skull (F:A.M. No. 87000) are
given in table 3. The over-all length of the skulls of M. isignis ranges
from 312 mm. to 352 mm., and the basilar length, from 286 mm. to
300 mm. The females exhibit a smaller canine than the males; otherwise
no sexual dimorphism, such as variations in the size, shape, and position
of the preorbital fossae, has been observed.

DorsaL AND LATERAL Views OF SKULL: The muzzle is medium-sized
in proportion to the skull. Anteriorly the nasals extend forward beyond
the canines and overhang a comparatively small nasal opening. A deep-
ening of the anterior part of the maxilla (buccinator fossa) extends back,
but does not unite with the preorbital fossa. The nasal process of the pre-
maxilla is broad, ending above posterior P? or anterior P?, and joins
the nasal and the maxilla in a well-defined suture.

The preorbital fossa is separated from the orbit by a preorbital bar
which averages 16 mm. wide and involves the lacrimal, nasal, and maxil-
lary bones. The fossa is oval-shaped and high on the face, ending anteri-
orly above P? where it is separated from the buccinator fossa by a gently
rounded ridge. The mid-superior border is sharp, and the posterior border
is sharp and deeply rounded, whereas the inferior portion slopes into the
maxilla with no defined border. There is no malar fossa, but there is a
slight depression on the malar bone for the attachment of the levator
labii superioris proprius muscle. The infraorbital foramen opens above
the posterior half of P*, outside the preorbital fossa, and moderately low
on the face. The opening is round, with a deeply defined sulcus fading
out below the anterior portion of the preorbital fossa.

Well-rounded orbits sit about 15 mm. below the plane of the frontals
immediately posterior to the midportion of the skull. The orbits are
closed by a broad, well-developed, supraorbital process which unites
strongly with the zygomatic process of the temporal bone. A large, com-
pletely closed supraorbital foramen is placed medianly on the supra-
orbital process. On the anterior internal part of the orbit there is a deep,
well-defined fossa for the lacrimal sac. External to and slightly above this
fossa there is a distinct lacrimal duct leading forward and opening in-
ternally into the nasal cavity. This duct leads forward and down behind
the preorbital fossa.

A prominent, well-rounded, facial crest begins abruptly above post P*4

34 AMERICAN MUSEUM NOVITATES NO. 2300

or mid M? and, in confluence with the zygomatic process of the malar
bone, forms a heavy crest which is, in no case, dorsoventrally compressed
as in examples of Pliohippus. The malar bone is anteriorly expanded
and forms the lower anterior third of the orbit. The zygomatic process
of the temporal bone is fairly light and well separated from the cranium
and ends just ahead of the auditory meatus.

The frontal bones are flat, without any indication of a dome as in
Griphippus, and curve sharply downward to the low-placed orbits. The
naso-maxillary and the naso-lacrimal sutures are in the superior and
posterior portions of the preorbital fossa. The naso-frontal suture ends
over the anterior half of the orbit and forms a distinct V at the mid-
line. The parieto-frontal suture is about 23.0 mm. anterior to a sharply
defined, moderately high, parietal crest. The parietal region is rounded
and, at the maximum transverse width, is about equal to the width of
the skull at the posterior borders of the preorbital fossae. Posteriorly
the parietal crest unites with a moderately extended, narrow, nuchal
crest which unites laterally with well-defined temporal crests. The para-
mastoid processes are at a 15-degree angle to the long axis of the skull,
are wide posteriorly, with a broad lateral aspect, but are thin and nar-
row anteriorly.

VENTRAL VIEW OF THE SKULL: The incisive border forms a broad U.
The palatine surface is concave, deepest at the posterior part of the
palatine processes. A small foramen incisivum opens about 10 mm. be-
hind the median line between the first incisors. Large palatine fissures
open at the anterior end of the palatine processes and extend as far for-
ward as the foramen incisivum and as far back as opposite the middle
part of the canine. In Equus the palatine fissures are long, narrow slits,
much narrower than in Merychippus.

The posterior nares are widest opposite M? in adult skulls and are
widest opposite M? in immature specimens. In modern horses the posterior
nares are widest opposite M® and post M®. The palatine foramen opens
anteriorly opposite the anterior part of M?, whereas in modern horses
the foramen is opposite M23.

The basilar part of the occipital and the body of the sphenoid are
broad and heavy, with well-developed basilar tubercles. In the available
specimens the vomer is not well preserved but appears light and may
extend back as far as the alar canal of the pterygoid process.

The occipital condyles are moderate in size and roundness and have
shallow, wide condyloid fossae; in Equus these condyloid fossae are deeply
concave. The auditory bullae are moderately inflated, with deep hyoidal
pits and grooves on the lateral external sides of the bullae which end at

1967 SKINNER AND TAYLOR: BIJOU HILLS 35

the base of the auditory meatus. The canal of the meatus extends upward
and back, opening posteriorly with a slight outward flare. The temporal
canal is placed far back on the temporal fossa and opens posteriorly
toward the wall of the auditory meatus.

ORDER ARTIODACTYLA
SUBORDER RUMINANTIA
FAMILY ANTILOCAPRIDAE

SUBFAMILY MERYCODONTINAE

MERYCODUS LEIDY, 1854

Type Species: Merycodus necatus Leidy, 1854.

Type: Lost, a ramal fragment with P,-M,; collected by Hayden and
Meek in 1853 from the Bijou Hills, South Dakota.

Neotype: F:A.M. No. 51790 (fig. 9B; table 4), a right partial ramus
with dP, (P, germ)-M,; dP,, P, germ, and M, are broken; M,_, are
damaged, but the crown height of both can be measured. (See horn core,
F:A.M. No. 51127, probably the same individual.)

ParatyPe!: F:A.M. No. 51127 (fig. 10A). A left incomplete horn with
burr; both prongs are broken. (See partial ramus, F:A.M. No. 51790,
probably the same individual.)

Locatity: The neotype and paratype of Merycodus necatus came from
the south Bijou Hill in the extreme northwest corner of Charles Mix
County, South Dakota, in the NW. %, sect. 8, T. 100 N., R. 71 W.; col-
lected by Morris F. Skinner and party in May, 1937.

STRATIGRAPHIC Unit: F:A.M. Nos. 51790 and 51127 came from a 5-
foot zone of brownish pink clay that carries scattered barite crystals or
“rock roses.” These beds are part of a unit that has been named the
Fort Randall Formation (this paper, p. 11) and are of late Miocene age,

1 In the case of the horn core (F:A.M. No. 51127), we use the term “paratype” with
some hesitation, because, although the neotype has all the status of a holotype (Stoll and
others, 1961, Article 75), there is no term in the Code to designate a paratype for a
neotype, such as “neoparatype.” The neotype (F:A.M. No. 51790) and the paratype
(F:A.M, No. 51127) were found by the senior author within 6 inches of each other. There
is good reason to believe that the jaw and the horn core belonged to one individual.
Because the supposition cannot be proved, they are treated as a syntypic series, in which
the jaw becomes a neotype and the horn core a paratype, or a “neoparatype,” a term
that we do not wish to introduce. [See also Stoll and others, 1961, Article 73 (c) and
Recommendation 73D.] The present diagnosis is based on both the neotype and the
paratype and adds considerably to the concept of the genus.

36 AMERICAN MUSEUM NOVITATES NO. 2300

Fic. 9. Lateral buccal views. A. Merycodus sabulonis, type, reversed, A.M.N.H.
No. 14109; from 23 miles south of Agate, Sioux County, Nebraska. B. Merycodus
necatus, neotype, reversed, F:A.M. No. 51790; from south Bijou Hill (fig. 1C).
C. Cosoryx furcatus, reversed (after Leidy, 1869, pl. 14, fig. 9); from Niobrara
River, Nebraska. D. Cosoryx furcatus, F:A.M. No. 51024 (cf. fig. 10E, one indi-
vidual); from June Quarry, Burge Member of the Valentine Formation, Brown
County, Nebraska. All x 1.

post Lower Snake Creek and pre-Lower Valentine (fig. 1C at 1995 feet
above mean sea level).

1967 SKINNER AND TAYLOR: BIJOU HILLS 37

History: Leidy’s type of Merycodus necatus from the Bijou Hills has an
involved history. The type was never figured and disappeared after 1918.
Nevertheless, the concept, based on this lost ramus, of a group of primi-
tive, antelope-like creatures has been carried through the writings of most
subsequent authors. Even though the characters of the type species were
in doubt, the genus Merycodus was selected by Matthew (1909, p. 115)
as the type genus for the subfamily Merycodontinae. Frick (1937, p. 371)
made two divisions of the Antilocapridae—Merycodontini and Antilo-
caprini. The division was intermediate between the family and sub-
family. Frick’s “-ini” ending for the division is not in accordance with
the recommendation in the 1961 International Code (Stoll and others,
1961, Article 29A). If the type had been figured in 1854, Leidy’s genus
and species Merycodus necatus would now have fulfilled the requirements of
a nominal taxon, but his failure to do so and the subsequent loss of the
type left the genus and species indeterminate.

Matthew (1918, p. 219) stated, “The type of M. necatus is from the
Bijou Hills and is in the American Museum but, in absence of topo-
types, it affords no good specific characters except size.” Matthew gave
no catalogue number for the specimen. Frick (1937, p. 314) wrote that
the whereabouts of the specimen was unknown and stated in a footnote,
“The statement of Matthew (1918, p. 219) that the type is in the Ameri-
can Museum collection appears to be incorrect as the specimen has never
been entered in the Museum catalogue.” As of 1966 we have not found the
specimen in the American Museum of Natural History collection or in
that of the Academy of Natural Sciences of Philadelphia. The specimen
is not recorded in either collection.

Henshaw (1939, p. 26) stated: “It is important to recognize that the
species, being intangible, is not lost, and hence the genus Merycodus can-
not be discarded. The only loss is the type specimen for the species. To
rectify this, a neotype, providing that more material can be obtained
at Bijou Hill, or a neoholotype from another locality should be set up.
It is the contention of the present writer that the material described by
Leidy in 1869 fulfills just such a category.” Although Henshaw suggested
that the material described by Leidy be considered a neoholotype, he
did not formally designate it as such. Moreover, the ramus figured by Leidy
(1869, pl. 14, figs. 9, 10) from the Niobrara does not have the qualifica-
tions for a neotype because it was not derived from the type locality.
The specimens figured in the present report (M1. necatus neotype and
paratype) fulfill the requirements as set forth by the 1961 International
Code (Stoll and others, 1961, Article 75C).

DescripTion: Concerning the lost type of Merycodus necatus, Leidy

“ABME WIOM a}assoy JOLIOJUE ‘AA f9}]OSSO} JOLIOJUY BuTUTEaI "Py ‘AA ‘JUasoid srejourord yyrur ‘J :(6g1 “d ‘ZPET) JAUUTAG Woy “IeIM ,

OST = = OZ! OST aseq 1e fy
6ST ard — Gol Cal uUMOID UO FPL
38 = ¢'8 BL G8 aseq ye *W
0'ZI = OTT £6 C6 UMOID UO * Py
BL = aL L9 aL aseq ye "WW
C6 08 86 WL 08 uMmoio uo TA
ZL 0's — et = rd
19 = = 69 = *d
vr = = ZF = 2d
y1sug] 101Wajsodo1ajuy
(M) (1) (TM) (M) PLN
COFIG O6LIS 6014 1691
‘ON (WV Jaquiny] ON ‘ON WW: ‘ON ‘H'N'W'V ‘ON ‘W'V:d
POLIIJOY addy, adAq09Ny odd J, Poet1aJOy
snyoosnf x40so') snjojau “Ww SnyoIau ‘WW stuojngDns “Ww stuojngns "jy

xdioso7) NW Snpootiay JO HILAL] WAMO'] AHL JO (SUALAWITIJY NI) SLNAWAUNSVA PY
¥ ATAVL

1967 SKINNER AND TAYLOR: BIJOU HILLS 39

(1854a, p. 90), stated: “. . . Fragment of a lower jaw, containing a last
premolar and the first true molar of a small ruminant allied to the
Musk. The last premolar on the triturating surface presents a series of
five folds projecting inwardly. Its measurement antero-posteriorly is 8
mm. The inner side of the true molar is nearly as plain as in those of
Poebrotherium. Its measurement antero-posteriorly is also 8 mm.” The last
premolar of the lost type of M. necatus, as Leidy described it, was un-
worn, because the five folds on the crown surface disappear with wear.
The neotype, F:A.M. No. 51790 (fig. 9B), has an unerupted broken P,

TABLE 5
ComPaRISON OF LENGTH AND HEIGHT oF M,

Length of First Crown Height

Two Lobes¢ of M, of M, as

Merycodus sabulonis

Referred, F:A.M. No. 51691 9.3 12.2 76%

Type, A.M.N.H. No. 14109 8.8 116 80%
Merycodus necatus

Neotype, F:A.M. No. 51790 9.8 12.0 55%

Type (no number) — — —
Cosoryx furcatus

Referred, F:A.M. No. 51405 10.3 21.7 47%

“The third lobe is missing from M. necatus neotype (F:A.M. No. 51790).
’Estimated, because of wear.

and is therefore less mature than the lost type. Comparisons with Leidy’s
description of his lost type are limited to the M,.

The measurements of the neotype of M. necatus bracket the measure-
ments of the lost type. Leidy gave the anteroposterior length of M, of
the type as 8 mm. In the neotype the M, is 9.5 mm. long on the
crown and 7 mm. long at the base; the lingual surface is smooth, as
Leidy described his type. The crown wear of the lost type is unknown.
However, Leidy’s anteroposterior measurement of M, indicates that it
was within the size range of the M, of the neotype. The M, of the
neotype is broken, but enough is present to show that it was high-
crowned and had a well-developed median external ridge between the
protoconid and hypoconid. The first two lobes of the M3 are unworn,
revealing the degree of hypsodonty; the third lobe of Mz, is missing.

Table 5 gives the crown height and the ratio of the height to the
length of the first two lobes of M, in the neotype of M. necatus as com-

40 AMERICAN MUSEUM NOVITATES NO. 2300

pared with those of Merycodus sabulonis! (fig. 9A) and Cosoryx furcatus, re-
ferred (fig. 9C, D). The crown height of the Mz, in the neotype of
Merycodus necatus is intermediate between that of Merycodus sabulonis from
the Lower Snake Creek beds of Sioux County, Nebraska, and that of
Cosoryx furcatus, referred, from the Burge Member of the Valentine
Formation in Brown and Cherry counties, Nebraska.

Matthew and Cook (1909, p. 411, fig. 24) based Merycodus necutus sabu-
lonis on a ramus (A.M.N.H. No. 14109) and referred a series of lower
jaws to the subspecies from the “Pliocene deposits” of Sioux County,
Nebraska. In this publication Matthew and Cook considered the channel
deposits of the area as the Snake Creek beds, not realizing that there
was a temporal difference. However, Matthew (1924, p. 68) recorded
Merycodus necatus in a faunal list as occurring in the Sheep Creek and the
Lower Snake Creek deposits but not from the Upper Snake Creek beds.
Subsequent Frick collections of Merycodus sabulonis, referred, from both
Sheep Creek and Lower Snake Creek beds, verify Matthew’s 1924 cor-
rected allocation.

A comparison between the neotype of Merycodus necatus and the type
of M. sabulonis must take into account the difference in dental wear.
Merycodus sabuloms has well-worn dentition; the dentition of M. necatus
neotype is approaching maturity. If the differences in age be allowed for,
the type of M4. sabulonis has smaller, less hypsodont molars and a rela-
tively larger P, than the incomplete P, in the dental series of M. necatus
neotype, very possibly the result of temporal differences. Gregory (1942,
pp. 405-406) also correlated dental changes with time.

The paratype of Merycodus necatus, F:A.M. No. 51127, is a small horn
with a relatively short, straight shaft (fig. 10A). The burr is centrally
placed, and the prongs are long relative to the length of the shaft.
Although a portion of each prong is missing, the prongs appear sub-
equal. A comparison of the paratype of M. necatus with the type and re-
ferred horns and skulls of Cosoryx furcatus (fig. 10D-F), shows that the
horn of M. necatus is smaller, the shaft is much shorter, with prongs pro-
portionately longer relative to the length of the shaft, and the burr is more
centrally situated and comparatively higher than in Cosoryx _furcatus (fig.
10F). The referred specimens of Cosoryx furcatus (fig. 10E, F) are from
June Quarry, Brown County, Nebraska. The stratigraphic relationship
and lithology of this quarry are typical of the Burge Member of the
Valentine Formation. The size and variation of the horn cores shown

-. |Cosoryx (Subparacosoryx) savoronis Frick, 1937 = Merycodus sabulonis Matthew and Cook,
1909. See synonymy, page 41.

1967 SKINNER AND TAYLOR: BIJOU HILLS 41

in figure 10E and F illustrate the range of individual variation in a
large population sample from a single faunal deposit.

Comparison of the paratype of M. necatus (fig. 10A) with Frick’s type
of “Cosoryx (Subparacosoryx) savaronis,” A.M.N.H. No. 17339 (fig. 11B)
[= Merycodus sabulonis|, reveals that the horn shaft of M. necatus is shorter
and the prongs longer. In both M. necatus and M. sabulonis, the burr is
centrally placed on the shaft and is relatively higher than that in
Cosoryx furcatus and Meryceros warreni (Leidy, 1858), Frick, 1937. There is,
however, a great deal of variation in the length of the shaft and prongs
in specimens referred to M. sabulonis Matthew and Cook. A referred M.
sabulonis horn, F:A.M. No. 51259 (fig. 11A), from New Surface Quarry,
East Sinclair Draw, in the Lower Snake Creek deposits, is an extreme
variant of the species and differs from that of Merycodus necatus in having
a much longer shaft and reduced prongs in comparison to shaft length.
This referred specimen of Merycodus sabulonis is distinctively different from
the type of Cosoryx furcatus and referred remains from the Burge Member
of the Valentine Formation, the probable source of Leidy’s type of C.
furcatus. On the basis of the new horn evidence, the paratype of M. necatus
is more comparable to the horns referred to the Lower Snake Creek M.
sabulons population and morphologically distinct from Cosoryx furcatus.

In the types of Meryceros warren, U.S.N.M. No. 149 (fig. 10C), and
Meryceros nenzelensis, the horns are larger and the horn shafts are heavier
and more laterally compressed than in the paratype of Merycodus necatus.
The type of Meryceros (Submeryceros) crucianus, F:A.M. No. 31495 (fig. 10B),
is a small horn in which the prongs emerge from the fork at a more acute
angle than in the paratype of M. necatus. In M. (S.) crucianus the burrs
are developed separately at the base of each prong instead of on the
main shaft as in Merycodus necatus and all other observed merycodont
horns.

Merycodus sabulonis Matthew and Cook, 1909

Merycodus necatus sabulons MatrHEw AnD Cook, 1909, p. 411, fig. 24. The
type is a ramus, A.M.N.H. No. 14109.

Merycodus necatus Leidy; referred: MatTHEew, 1918, p. 219, fig. 18. The re-
ferred skull is ALM.N.H. No. 17339; later selected by Frick, 1937, as the holo-
type of Cosoryx (Subparacosoryx) savaronis.

Merycodus necatus sabulonis Matthew and Cook: Matruew, 1924, p. 200.

Cosoryx (Paracosoryx) sabulonis (Matthew and Cook): Frick, 1937, p. 353,
fig. 44. The type is a ramus, A.M.N.H. No. 14109.

Cosoryx (Subparacosoryx) savaronis Frick, 1937, p. 353, figs. 28, 28A.

Discussion: Frick (1937, p. 353) described ‘“Cosoryx (Subparacosoryx)
savaronis” and designated it the subgenotypic species. The type is a partial

42 AMERICAN MUSEUM NOVITATES NO. 2300

= MAG)
Ty A
inn Hl

4
AN

i
qi

Fic. 10. Lateral views. A. Merycodus necatus, paratype, F:A.M. No. 51127;
from south Bijou Hill (fig. 1C). B. Meryceros (Submeryceros) crucianus, type,
F:A.M. No. 31495 (after Frick, 1937, fig. 35); from Santa Cruz, New Mexico.
C. Meryceros warreni, type, U.S.N.M. No. 149, from cast, reversed; from Nio-
brara River, Nebraska. D. Cosoryx furcatus, type, U.S.N.M. No. 148, from cast;
from Niobrara River, Nebraska. E, F. Cosoryx furcatus; from June Quarry, Brown
County, Nebraska. E. F:A.M. No. 51024 (cf. fig. 9D, one individual). F. F:A.M.
No. 51081, reversed. All X'%.

skull (A.M.N.H. No. 17339, fig. 11B). On the same page Frick trans-
ferred Merycodus necatus sabulons Matthew and Cook (1909) to Cosoryx

1967 SKINNER AND TAYLOR: BIJOU HILLS 43

(Paracosoryx). Concerning the change, Frick (1937, p. 353) stated, “It
seems necessary, pending the securing of more satisfactory evidence, to
hold the possibly equivalent crania from the general area in a separate
species, §. savaronis.” Through an error, Frick (1937, p. 353) included
figure 18 from Matthew (1918), which was the type of his new species,
“C. (S.) savaronis,” in the synonymy of the referred material of “Cosoryx
(Paracosoryx) sabulonts.”

Fic. 11. Lateral views of Merycodus sabulonis. A. F:A.M. No. 51259; from New
Surface Quarry, Sioux County, Nebraska. B. A.M.N.H. No. 17339 (after Frick,
1937, fig. 28A); Lower Snake Creek beds, Sioux County, Nebraska. Both x.

The subgenotypic species, Cosoryx (Paracosoryx) wilsoni, based on a par-
tial skull, is from the Sheep Creek deposits in Sioux County, Nebraska.
The dentitions of Merycodus sabulonis from the Lower Snake Creek deposits
are smaller and lower-crowned than dentitions referred to Cosoryx (Para-
cosoryx) wilson. Unpublished stratigraphic studies by Skinner reveal that
Long Quarry, the type locality of C. (P.) wilsoni, is in the lower part of
the Sheep Creek deposits and therefore earlier than the Lower Snake
Creek deposits from which Merycodus sabulonis came. The type ramus and
referred dentitions of M. sabulonis occurred in faunal association with
crania, horns, and dentitions of “Cosoryx (Subparacosoryx) savaronis” Frick,
1937, and appear to be the same species. On this basis, the species “C.
(Subparacosoryx) savaronis,” based on a partial skull (see fig. 11B), is here
considered a synonym of Merycodus sabulonis. The type of M. sabulonis
(fig. 9A) is a ramus, A.M.N.H. No. 14109.

The neotype of Merycodus necatus, which is a ramal fragment (F:A.M.

44 AMERICAN MUSEUM NOVITATES NO. 2300

No. 51790), and the paratype, which is a horn core (F:A.M. No. 51127),
were found in such close association that the two are almost certainly
one individual. Leidy’s description of the type ramus (now lost) gave
no diagnostic characters, but gave measurements that can be duplicated
in almost any of the Merycodontinae. The neotype (F:A.M. No. 51790)
is from the type locality and approximates the measurements given by
Leidy for the lost type (table 4). The associated horn core adds charac-
ters to the genotypic species, M. necatus, which have been heretofore un-
known. It is presumed that Leidy’s lost type of M. necatus came from the
lower beds of the Fort Randall Formation where we obtained the neo-
type and paratype (see above, p. 4).

Frick (1937, pp. 271, 276, 292, 345) was unable to establish objective
characters for Merycodus as a genus but did not drop the name, as wit-
nessed by his use of the division Merycodontini for all included genera.

Frick (1937, p. 334) commented, “The genus Cosoryx . . . may prove
to be synonymous with Merycodus Leidy, should it ever be possible to
definitely determine the latter.” However, the relatively short-shafted
horn cited in this report as the paratype of M. necatus (F:A.M. No. 51127)
is notably different from the taller and more slender-shafted incomplete
horn (U.S.N.M. No. 148), the type of Cosoryx furcatus Leidy, genotypic
species (fig. 10D).

Henshaw (1939, p. 26) contended that the ramus from the Niobrara,
which Leidy figured as Cosoryx furcatus referred (this paper, fig. 9C), should
be set up as a neoholotype of Merycodus necatus. However, Leidy’s figures
(1869, pl. 14, figs. 9, 10) show that the molars are larger and taller-
crowned than those of the neotype (F:A.M. No. 51790) from the Bijou
Hills. The taller-crowned molars and reduced premolars are comparable
to dentitions referred to Cosoryx furcatus which is the dominantly occurring
merycodont from the Burge Member of the Valentine Formation along
the Niobrara River.

Gregory (1942, p. 407) predicated: “. .. M. necatus is known only from
teeth, and is not positively associated with any horn pattern. There is
some evidence, moreover, that ‘Cervus’ warreni may be a synonym of M.
necatus, and represents the horn pattern of that species. In the Niobrara
River fauna only ‘C.’ warrent horns occur, and this fauna appears to be
the same as that at Bijou Hills, from which the type of Merycodus was
obtained.” There are observable differences in the horn cores that tend
to refute Gregory’s statement. The type of Meryceros warreni from the
Niobrara River differs from the paratype of Merycodus necatus from the
Bijou Hills in that M. warreni is larger and has a heavier, laterally more
compressed shaft (see fig. 10C).

1967 SKINNER AND TAYLOR: BIJOU HILLS 45

According to the unpublished diary of Warren, and that of Snowden,
Warren’s cartographer, the expedition that collected the type of Mery-
ceros warreni passed the mouth of the Snake River on October 22, 1857.
The night of October 22-23, the Warren-Hayden expedition camped 8
miles below the mouth of the Snake River on the Niobrara. On the fol-
lowing day, October 23, the party crossed the Minnecheduza Creek and
ascended the canyons north of the present town of Valentine, Nebraska.
They camped for the night on a high bluff on the north side of the
Niobrara River approximately 1'’2 miles northeast of the site of the
present Fort Niobrara buildings.! There are exposures of the Lower
Valentine Formation throughout this area, and it is possible that the
type of Meryceros warrent could have been found at this time. ‘The type
horn of Meryceros warreni is characteristic of horns from the Lower
Valentine Formation, whereas the paratype of Merycodus necatus more
closely resembles Merycodus sabulonis from the Lower Snake Creek deposits
of Sioux County, Nebraska, which occurred earlier.

CONCLUSIONS

A revision of faunal names and geologic terms, such as the present
paper attempts by supplying factual data, is rewarding only when it
appears that a step forward has been made and that something useful
has been added for future studies. McKenna’s statement (1965, p. 17),
“The use of the written word to denote various aspects of nature demands
that concepts be disciplined by facts and that these facts be readily ac-
cessible to any investigator,” is particularly applicable to a revision of
this nature. For every long-known group of fossil animals, a revision of
some sort should be made in order to establish a firm foundation for an
evolutionary concept. In this paper, we have placed Leidy’s fauna from the
Bijou Hills where we believe it belongs in time and have supplied a strati-
graphic framework. A long and cloudy history has been built around
Leidy’s types, and either we have diagnosed these so that the characters
they exhibit are identified with more complete material, or we have left
them in the category of nomina dubia (i.e., Hippodon speciosus).

We believe that a step forward has been taken by the introduction
of a new formation in the Bijou Hills area. It would serve little purpose
to name a new geologic formation if it did not represent deposits and
time missing from the record elsewhere. The deposits of the Fort Randall
Formation (new) have yielded a fauna which, in part, represents the

1 Fort Niobrara was not established until 1880. The town of Valentine, Nebraska, was
not in existence in 1857.

46 AMERICAN MUSEUM NOVITATES NO. 2300

hiatus between the Lower Snake Creek and the Lower Valentine forma-
tions. Certain other rocks in this region, which have been variously
mapped as “Bijou Formation” and “Bijou facies,” are here considered
as representing neither a formation nor a facies but rather a sedimentary
condition.

We have designated a type set of rocks in the type locality for the
Rosebud Formation in order to stabilize the nomenclature for this lithic
unit. There is still no published description of the Rosebud Formation,
and it is beyond the scope of this paper to supply one.

In the case of Merychippus insignis, we have duplicated as nearly as
possible the fragmentary holotype and associated it with an extensive
population sample from the Lower Snake Creek deposits of western
Nebraska. The morphological characters of the type taxon are here re-
defined and enlarged upon.

In the case of Merycodus necatus, the neotype (a ramal fragment) was
found in such close association with the paratype (an incomplete horn)
that the two are believed to represent one individual. The paratype,
therefore, enables us to give, for the genus Merycodus, diagnostic horn
characters which have been heretofore unknown. These characters, taken
with all the known temporal aspects of the Bijou Hills fauna, show that
Merycodus is morphologically distinct from Cosoryx.

The stratigraphic nomenclature of the Miocene of western Nebraska
has been summarized by McKenna (1965, fig. 1). McKenna clearly
stated the need for more precise studies and adequate definitions of the
sediments from which faunas have been obtained. For the past 30 years
at least, Frick Laboratory parties have been accumulating factual data
about the sequence of sediments in the Great Plains and their con-
tained faunas.

ADDENDUM: RODENT IDENTIFICATIONS

By THompson M. Stout!

Among the many interesting specimens collected by Morris F. Skinner
and associates from the Bijou Hills are teeth of several rodents, listed
below. Of these, only the beaver and mylagaulid seem presently useful
for correlation purposes. The beaver teeth, of which one is illustrated
(fig. 12A), compare best with specimens from the highest-known level
of the “Lower Snake Creek” (Hemingford Group) sediments of western

1 Research Associate, Frick Laboratory, the American Museum of Natural History;
Department of Geology and State Museum, University of Nebraska, Lincoln, Nebraska.

1967 SKINNER AND TAYLOR: BIJOU HILLS 47

Nebraska, the highest unit in the Great Plains Miocene in Nebraska
usage.! The mylagaulid teeth seem also to relate best to specimens from
this same “Lower Snake Creek” horizon.

FAMILY CASTORIDAE

“Eucastor” cf. curtus (Matthew and Cook)

Dipoides curtus MATTHEW AND Cook, 1909, p. 381, fig. 9 [= “Dipordes brevis”
Matthew and Cook, a nomen nudum in the faunal list of this same paper, p. 364;
type, A.M.N.H. No. 13871]. MatrHew, 1918, p. 186; 1924, pp. 66, 74.

Monosaulax curtis [sic] (Matthew and Cook): Stirron, 1935, pp. 420-421, figs.
74-75, chart 2.

SPECIMENS: Three cheek teeth (F:A.M. Nos. 42922A-—42922C), apparently
of different individuals: left P, (fig. 12A), right P4, and left M°.

Locatity AND LEvEL: South Bijou Hill (F:A.M. Prospect No. 55),
Charles Mix County, South Dakota, at “Rosestone level”; collected by
Morris F. Skinner and associates, 1950.

DenTAL MEasuREMENTS: P,, wear surface, length, 4.4 mm.; width
(anterior), 2.9; width (posterior), 3.8; external crown height, 7.6; ex-
ternal groove height, 5.75; central internal groove height, 0.8; other
two internal grooves worn out to form enamel lakes in pattern; moderate
mature wear. P*, wear surface, length, 3.7; width (anterior), 3.75; width
(posterior), 3.7; internal crown height not measured; internal groove
height, 4.1; anterior external groove height, 0.2; other two external
grooves worn out to form enamel lakes in pattern; moderate mature
wear. M®, wear surface, length 2.9; width (anterior), 3.0; width (posterior),
2.6; internal crown height, 5.0; internal groove height, 1.9; the three
external grooves worn out to form enamel lakes in pattern; moderate
mature wear.

Comparisons: A specimen from the highest level of the “Lower Snake
Creek” (Mill Quarry of the Frick Laboratory), in the “Sheep Creek-
Snake Creek collecting area,” Sioux County, Nebraska, compares favor-
ably (fig. 12C), except for a slight difference in the stage of wear: it is
a mandible with both incisors (broken), both right and left P,-M,

'The present writer considers the Hemingford Group to be of late Miocene age,
with the Marsland an approximate early Vindobonian (Helvetian) correlative, the prin-
cipal or restricted part of the Sheep Creek equivalent to the lower part of the late
Vindobonian (Tortonian), and the “Lower Snake Creek” equal to that part of the
French Miocene usually termed late Vindobonian and in central Europe late Vindo-
bonian to (or) Sarmatian. The evidence from the fossil beavers for these correlations,
together with a report on recently completed European studies, will be presented in full
elsewhere.

48 AMERICAN MUSEUM NOVITATES NO. 2300

Fic. 12. Comparison of beaver specimens, and a cricetine dentition.
A. “Eucastor” cf. curtus (Matthew and Cook), left P,, F:A.M. No. 42922A;
from the south Bijou Hill (fig. 1C). x 5. B. “Eucastor” curtus (Matthew and Cook),
holotype, right P,-M., reversed, A.M.N.H. No. 13871; from the “Lower Snake
Creek,” Sioux County, Nebraska. <5. C. “Eucastor” cf. curtus (Matthew and
Cook), left P,-M3, F:A.M. No. 42926; from the “Lower Snake Creek,” Sioux
County, Nebraska. 5. D. Peromyscus (Copemys) sp., right M,_3, F:A.M. No.
42923; from the south Bijou Hill (fig. 1C). 10. Each bar equals 1 mm.

(F:A.M. No. 42926), of which the left ramus is illustrated.? Further com-
parison with the holotype of this species (fig. 12B), a right ramus with

1 This “Lower Snake Creek” specimen was also collected by Morris F. Skinner and
associates.

1967 SKINNER AND TAYLOR: BIJOU HILLS 49

P,-M,, alveolus of M, (A.M.N.H. No. 13871), probably from a some-
what older level of the “Lower Snake Creek” of the same collecting
area, Sioux County, Nebraska,’ emphasizes the smaller size of the type.
A new subspecific name may be necessary eventually to accommodate
the Mill Quarry and Bijou Hills specimens.

Discussion: This species cannot now be assigned to “Monosaulax,”’ fol-
lowing Stirton (1935, pp. 420-421), because this genus proves to be in-
valid and a synonym of LEucastor, with their type species apparently
conspecific also. Since the Bijou Hills teeth and the abundant materials
from the “Lower Snake Creek” here under consideration presumably
belong in the lineage ancestral to the Pliocene Eucastor, this generic allo-
cation is employed here but in quotation marks (“Fucastor’’.

FAMILY MYLAGAULIDAE

Mylagaulus sp.

SPECIMEN: Skull fragment, chiefly basicranium, with left P*, and man-
dibular fragments with right incisor (broken), left and right P,, together
with some fragments of the appendicular skeleton; all taken as pertain-
ing to a single individual (F:A.M. No. 42924).

Locairy AND LeveL: “From the extreme north Bijou Hill, near the
Missouri River, in Brule County, South Dakota” (F:A.M. Prospect
No. 55), in the “Pink deposits,” referred to in the 1949 field geologic
section”; collected by Morris F. Skinner and associates.

DeEnTAL MEAsureMENTs: Left and right P,, wear surfaces, lengths, 10.4
and 10.5 mm.; widths (anterior), 6.8 and 6.6; widths (posterior), 7.45
and 6.8; external crown heights, 13.3 and 15.5, the former approximate;
early mature wear, with crown cementum. P*, wear surface, length, 11.3;

' The holotype was collected by the American Museum of Natural History Expedition
of 1908, from the “Snake Cr’k Beds, 23 mi. S. of Agate, Nebr.” After careful investiga-
tion of the field records and long familiarity with the collecting history of this area,
Morris F. Skinner (personal communication) has determined that the “1908 or Old
Workings” are probably the same as the “Princeton Locality 1000-C” and as the site
worked by the Frick Laboratory and called “West Surface Quarry.” In any case, speci-
mens from the “West Sinclair Draw” in Sioux County, and the truly remarkable series
collected by Ted Galusha at the “Observation Quarry” in Dawes County, Nebraska (all
in the Frick Collection), can be matched almost identically with the holotype. The
typical horizon is therefore also “Lower Snake Creek.’

2 [F:A.M. No. 42924 was collected at an elevation of 1987 feet, 27 feet above the base
of the Fort Randall Formation (fig. 1B)}.—SxKInNER.

50 AMERICAN MUSEUM NOVITATES NO, 2300

width (anterior), 7.1; width (posterior), 8.0; internal crown height, 13.4;
early mature wear, with crown cementum.

Comparisons: As remarked above, the Bijou Hills specimen of myla-
gaulid also compares favorably with specimens from the “Lower Snake
Creek.”

Discussion: This is the largest of the three mylagaulids of the “Lower
Snake Creek,” for which a new generic name and possibly a new specific
name may be necessary. The first mylagaulid specimens to receive names
were loose cheek teeth. Later, generic weight was given to the presence
or absence of horned processes on the nasals, together with the relative
heights of the horns, if present. Since it is now evident that the presence
or absence of horns is a sex-linked character, a reinvestigation of all
available materials was undertaken and is now nearing completion,
parallel to a revision of the beavers.

FAMILY MURIDAE
SUBFAMILY CRICETINAE

Peromyscus (Copemys) sp.

SPECIMEN: Fragment of right ramus (F:A.M. No. 42923), with M,_,
(fig. 12D).

Locatity AND LEvEL: South Bijou Hill (F:A.M. Prospect No. 55),
Charles Mix County, South Dakota, at “Rosestone level”; collected by
Morris F. Skinner and associates, 1950.

DenTAL MEAsuREMENTS: M,_3, total crown length at base, 4.8 mm.
M,, total crown length, 1.7; total width (anterior), 1.1; total width
(posterior), 1.25. M,, total crown length, 1.65; total width (anterior),
1.4; total width (posterior), 1.35. M3, total crown length, 1.35; total
width (anterior), 1.2; total width (posterior), 1.05.

Comparisons: A study of probably topotype specimens of Peromyscus
(Copemys) loxodon (Cope), collected in New Mexico for the Frick Labora-
tory by Ted Galusha, reveals apparent minor differences from the Bijou
Hills specimen. A specimen from the Pliocene of Cherry County, Ne-
braska, collected by Morris F. Skinner, is also available, and it likewise
differs only slightly. More study, but also additional specimens, will be
necessary before this problem of significant differences can be settled.

Discussion: The affinities of the New Mexico materials with Peromyscus
seem evident, but the subgeneric name of Copemys Wood (1936) is tenta-
tively retained. A recent discussion by Clark, Dawson, and Wood (1964)
must be given careful study.

1967 SKINNER AND TAYLOR: BIJOU HILLS 51

INDETERMINATE

SPECIMENS: Two small rodent incisors, F:A.M. No. 42924A (fig. 1B),
found associated with the above-listed mylagaulid specimen from the
Bijou Hills; also a matrix block, F:A.M. No. 42925 (fig. 1C), still em-
bedded in plaster, with a mass of broken bones, presently undetermined.

Locatiry AND LEveEL: The incisors were associated with the myla-
gaulid (see under Mylagaulus sp., above). The block (a “concentration of
rodent bones”) is from the south Bijou Hill (F:A.M. Prospect No. 55),
Charles Mix County, South Dakota, in the “Zone above the Rock
Roses’’; collected by Morris F. Skinner and associates, 1949.

REFERENCES

AGNEw, ALLEN F.
1958. Bijou Formation—a stream deposit? Proc. South Dakota Acad. Sci.,
for 1957, vol. 36, pp. 129-133, fig. 1.
AGNEw, ALLEN F., AND Pau. C. TycuHsEen
1965. A guide to the stratigraphy of South Dakota. South Dakota State
Geol. Surv., Bull. 14, ii+ 195 pp., figs. 1-2.
CLARK, JOHN B., Mary R. Dawson, AND ALBERT E. Woop
1964. Fossil mammals from the Lower Pliocene of Fish Lake Valley, Nevada.
Bull. Mus. Comp. Zool., Harvard Univ., vol. 131, no. 2, pp. 29-63,
figs. 1-11.
Frick, CHILDS
1937. Horned ruminants of North America. Bull. Amer. Mus. Nat. Hist.,
vol. 69, xxviii + 669 pp., figs. 1-103.
GivLey, J. W.
1907. Revision of the Miocene and Pliocene Equidae of North America.
Bull. Amer. Mus. Nat. Hist., vol. 23, art. 35, pp. 865-934.
GREEN, MorTON
1958. Gomphothertum fricki from the Bijou Hills, South Dakota. Proc. South
Dakota Acad. Sci., for 1957, vol. 36, pp. 139-142, figs. 1-2.
GrecoryY, JOsEPH TRACY
1942. Pliocene vertebrates from Big Springs Canyon, South Dakota. Univ.
California Publ., Bull. Dept. Geol. Sci., vol. 26, no. 4, pp. 307-466,
figs. 1-51, pls. 49-51.
Hayopen, F. V.
1857a. Notes explanatory of a map and section illustrating the geological
structure of the country bordering on the Missouri River, from the
mouth of the Platte River to Fort Benton, in lat. 47° 30’ N., long.
110° 30’ W. Proc. Acad. Nat. Sci. Philadelphia, vol. 9, pp. 109-116,
map.
1857b. Notes on the geology of the Mauvaises Terres of White River, Ne-
braska. [bid., vol. 9, pp. 151-158.
1869. Geology of the Tertiary formations of Dakota and Nebraska [Intro-
duction]. Jn Leidy, Joseph, The extinct mammalian fauna of Dakota

52 AMERICAN MUSEUM NOVITATES NO. 2300

and Nebraska. Jour. Acad. Nat. Sci. Philadelphia, ser. 2, vol. 7,
viit472 pp., pls. 1-30.
HeEnsHaAw, Paut C.
1939. A Tertiary mammalian fauna from the Avawatz Mountains, San
Bernardino County, California. Publ. Carnegie Inst. Washington,
no. 514, pp. 1-30, figs. 1-3, pls. 1-6.
Jounson, F. WALKER
1936. The status of the name “Valentine” in Tertiary geology and paleon-
tology. Amer. Jour. Sci., vol. 31, pp. 467-475, figs. 1-2.
Leipy, JOSEPH
1854a. [On Bison latifrons, Arctodus pristinus, Hippodon speciosus, and Merycodus
necatus.| Proc. Acad. Nat. Sci. Philadelphia, vol. 7, pp. 89-90.
1854b. Synopsis of extinct Mammalia, the remains of which have been
discovered in the Eocene formations of Nebraska. Jbid, vol. 7, pp.
156-157.
1856 (Feb. 19). Notices of some remains of extinct Mammalia, recently
discovered by Dr. F. V. Hayden in the bad lands of Nebraska. Jb:d.,
vol. 8, p. 59.
1857a (read December, 1856). Notices of extinct Vertebrata discovered by
Dr. F. V. Hayden, during the expedition to the Sioux country under
the command of Lieut. G. K. Warren. Jé:d., vol. 8, pp. 311-312.
1857b. List of extinct Vertebrata, the remains of which have been discovered
in the region of the Missouri River: with remarks on their geological
age. Ibid, vol. 9, pp. 89-91.
1858. Notice of remains of extinct Vertebrata, from the valley of the
Niobrara River, collected during the exploring expedition of 1857,
in Nebraska, under the command of Lieut. G. K. Warren, U. S.
Top. Eng. by Dr. F. V. Hayden, Geologist to the expedition. Jdd,
vol. 10, p. 23.
1869. The extinct mammalian fauna of Dakota and Nebraska. Jour. Acad.
Nat. Sci. Philadelphia, ser. 2, vol. 7, viit+472 pp., pls. 1-30.
McKenna, Marco” C.
1965. Stratigraphic nomenclature of the Miocene Hemingford group, Ne-
braska. Amer. Mus. Novitates, no. 2228, pp. 1-21, fig. 1.
MatTTHEw, WILLIAM DILLER
1909. Faunal lists of the Tertiary Mammalia of the west. Jn Osborn, Henry
Fairfield, Cenozoic mammal horizons of western North America.
Bull. U. S. Geol. Surv., no. 361, pp. 1-138, figs. 1-15, 1 pl.
1918. Contributions to the Snake Creek fauna, with notes upon the Pleis-
tocene of western Nebraska, American Museum Expedition of 1916.
Bull. Amer. Mus. Nat. Hist., vol. 38, art. 7, pp. 183-229, figs. 1-20,
pls. 4-10.
1924. Third contribution to the Snake Creek fauna. Jbid, vol. 50, art. 2,
pp. 59-210, figs. 1-63.
MatTHew, WILLIAM DILLER, AND Haroip J. Goox
1909. A Pliocene fauna from western Nebraska. Bull. Amer. Mus. Nat.
Hist., vol. 26, art. 27, pp. 361-414, figs. 1-27.
MatTTHEw, WILLIAM DILLER, AND J. W. GIDLEY
1904. New or little known mammals from the Miocene of South Dakota,

1967 SKINNER AND TAYLOR: BIJOU HILLS 53

- American Museum expedition of 1903. Bull. Amer. Mus. Nat. Hist.,
vol. 20, art. 22, pt. 1, pp. 241-246.
' Merk, F. B., anp F. V. HaypDENn

1857. Descriptions of new species and genera of fossils, collected by Dr.
F, V. Hayden in Nebraska Territory, under the direction of Lieut.
G. K. Warren, U. 8S. Topographical Engineer; with some remarks on
the Tertiary and Cretaceous formations of the northwest, and the
parallelism of the latter with those of other portions of the United
States and territories. Proc. Acad. Nat. Sci. Philadelphia, vol. 9,
pp. 117-133.

Osporn, HEnry FaIRFIELD

1912. Craniometry of the Equidae. Mem. Amer. Mus. Nat. Hist., new ser.,
vol. 1, pt. 3, pp. 57-100, figs. 1-17.

1918. Equidae of the Oligocene, Miocene, and Pliocene of North America,
iconographic type revision. /bid, new ser., vol. 2, pt. 1, pp. 1-217,
figs. 1-173, pls. 1-54.

Quinn, JAMES HARRISON

1955. Miocene Equidae of the Texas Gulf Coast Plain. Publ. Bur. Econ.

Geol., Univ. Texas, no. 5516, pp. 1-82, figs. 1-5, pls. 1-14.
SIssON, SEPTIMUS, AND JAMES DANIELS GROSSMAN

1953. The anatomy of the domestic animals. [Fourth edition.] Philadelphia

and London, W. B. Saunders Co., 972 pp., figs. 1-736.
SKINNER, Morris F.

1942. The fauna of Papago Springs Cave, Arizona, and a study of Stocko-
ceros; with three new antilocaprines from Nebraska and Arizona.
Bull. Amer. Mus. Nat. Hist., vol. 80, pp. 143-220, figs. 1-19.

STEVENSON, ROBERT E.

1954. The Bijou Formation in southcentral South Dakota. Proc. South
Dakota Acad. Sci., for 1953, vol. 32, pp. 86-90, 1 fig.

1958. Revision and interpretation of the Bijou Formation. Jb:d, for 1957,
vol. 36, pp. 134-138, 1 fig.

STIRTON, REUBEN A.

1935. <A review of the Tertiary beavers. Univ. California Publ., Dept. Geol.

Sci., vol. 23, no. 13, pp. 391-458, figs. 1-142, map, charts 1-2.
STOLL, N. R., AND OTHERS (EDS.)

1961. International Code of Zoological Nomenclature adopted by the XV
International Congress of Zoology. London, International Trust for
Zoological Nomenclature, London, xvii+ 176 pp.

Unitep STaTEs ArmMy, Corps OF ENGINEERS

1947. Missouri River, Gavins Point near Yankton, South Dakota, to Stanton,

North Dakota, Omaha District, Sheet No. 38.
Warren, G. K.

1858. Preliminary report of explorations in Nebraska and Dakota in the
years 1855-’56-’57. Rept. Secretary of War, for 1858, appendixes.
Reprinted 1875, Washington, United States Army, Engineer Depart-
ment, pp. 1-125, 1 map.

Woop, ALBERT E.

1936. The cricetid rodents described by Leidy and Cope from the Tertiary
of North America. Amer. Mus. Novitates, no. 822, pp. 1-8, figs. 1-5,
tables 1-2.