Full text of "Kirtlandia"

See other formats

■%}

IISNI^NVINOSHIMS S3 I U VH 8 II LI B RAR ! ES^SMITHSONIAN^INSTITUTION
r- , Z r~ 2: r-

0/ -

m " W ^ VHAsvii/ rn ^

RIES $MITHSONIAN~!NSTITUTION NOIiniIl$NI~NVINOSHllWS S3IBVdan
co 21 f co z: co

2 X^TI T0>v < A 2 <

"5:

2 c/> 2 OT 2

I1SNI NVIN0SH1IWS S3iavaaiT LIBRARIES SMITHSONIAN INSTITUTION

5 \ W = 07 —

RIES SMITHSONIAN INSTITUTION NOlifUllSNI NVIN0SH1IINS S3iavaail

LI1SNI NVINOSH11WS S3iaVdan LIBRARIES SMITHSONIAN INSTITUTION

s ^ >

.RIES SMITHSONIAN INSTITUTION NOlilUllSNI NVINOSHillNS S3 I dVd 8 IT
c/> =; to — . to

LI1SNI NVINOSH1I1AIS SBIHVLiail LIBRARIES SMITHSONIAN_INSTITUTION
r - , z r~ z:

m >W' 2 m

CO — CO _ to

RIES SMITHSONIAN INSTITUTION NOIJ.nilJ.SNI NVIN0SH1IWS S3 I aVd8 IT

CO Z Co 2! CO

ITHSONIAN INSTITUTION NOlifllliSNI NVIN0SH1IWS S3ldVHan LIBRARIES SMITHSONIAN INSTITUTION NOIJLfUllSNI

niUSNI NVIN0SH1IWS S3ldVd8ll LIBRARIES SMITHSONIAN INSTITUTION _

CD
PD
>

VASV^X HI '' W' ^ X^VAS>j>^ m X^OS^>

co " = in tz co

ARIES SMITHSONIAN INSTITUTION NOUflJLIJLSNI NVINOSH1IWS S3IHVa ail
2: v co 2:

•S /C»I^TIT(yjN. ^ 2 * ~ S^\a Xj^n. ^

A z: ^ h MW/m- 2 /W“

^ o x fMf'W, 0

co sv co Mav/%A co

11I1SNI NVIN0SH1II/NS^ S 3 I dVd 8 ll\lBRAR I ES^SMITHSONIAN INSTITUTION ,
<f> x co x ___ ,c

UJ g L

O X^sv^ L. O

2 " -J 2

IARIES SMITHSONIAN INSTITUTION NOIlfUUSNI SWSN0SH1IWS S3IdVaail
x r* 2 r- 2

*,/ S'*' :n 1- t— />:

.. _ 'iM'tW - h

0y H

?*. dcJ/ <£> m X^_2C>' w x^Til^.

niUSNI NVIN0SH1I1AIS S3 I d V8 8 ll~LI B R AR I ES ^SMITHSONIAN” INSTITUTION
co 2 .... co 2

RAR I ES ^SMITHSON IAN INSTITUTION NOIlfUIlSNS^NVINOSHlIWS^Sa I d Vd 8 II

ARIES SMITHSONIAN INSTITUTION NOIlfUUSNI

SMITHSONIAN”* INSTITUTION '
r~ ;

nvinoshims saiyvyan

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO DECEMBER 28, 1967 NUMBER 1

A NEW MARINE MEMBER IN THE
CONEMAUGH GROUP OF OHIO

James L. Murphy and Larry Picking

Case Western Reserve University and
West Virginia Department of Highways

ABSTRACT

A previously unreported marine limestone from the Conemaugh
Group, Pennsylvanian, of Ohio is described as a new stratigraphic unit, the
Noble Limestone Member. The type section is designated as the shale pit
of the Ava Brick Co., in the SE^NW1^ sec. 31, Buffalo Twp., Noble Co.,
Ohio. The position of the new unit is between the Ewing and Rock Riffle
Limestones. At two localities the Noble Limestone overlies a thin carbo-
naceous deposit and an underclay which may represent the Upper Bakers-
town Coal and Clay. Fossils collected from the new member consist for
the most part of bryozoans, brachiopods, and dissociated crinoid plates and
stems. It is suggested that an unnamed marine shale that overlies the Up-
per Bakerstown Coal in the vicinity of Bakerstown, Pennsylvania may be
the stratigraphic equivalent of the Noble Limestone.

A previously undescribed marine unit in the Conemaugh
Group, Pennsylvanian, of Ohio was discovered in the spring of 1966
by Murphy and Mr. Harry Izenour of Salem, Ohio. The name Noble
Limestone Member is here proposed for the new unit; the name
is derived from Noble County, Ohio, where typical exposures of
the new member occur. The type section is designated as the shale
pit of the Ava Brick Company, in the SE^NW1^ sec. 31, Buffalo
Twp., Noble County, Ohio. The measured section follows:

Pennsylvanian: Ft in

Conemaugh Group:

Shale, buff, limonitic, thin-bedded, mostly covered, meas-
ured to highest float blocks of Ames Limestone 13 0

Harlem Coal and Clay: smut and underclay 6

2 JAMES L. MURPHY AND LARRY PICKING NO. 1

Shale, maroon, weathers easily; contains fresh- water lime-
stone nodules which probably represent the Rock Riffle
Limestone, partly covered near top 9 0

Noble Limestone:

Limestone, gray, nodular, thin-bedded, marine, brec-
ciated near top, sparingly fossiliferous, non-
persistent 4

Limestone, white to gray, weathers buff; nodular,
marine, sparingly fossiliferous; interbedded with
green to greenish -gray, dense, calcareous, fossil-
iferous marine shale 2 3

Clay shale, blue, nonfossiliferous 2 2

Shale, gray, with coal plant remains; probably represents

the Upper Bakerstown Coal 3

Clay shale, yellow, finely laminated; may represent the

Upper Bakerstown Underclay 1 9

Ewing Limestone: limestone, gray, nodular to thin-bedded,
fresh-water, with abundant Spirorbis, small pelecy-
pods, Xenacanthus, and other vertebrate remains;
interbedded with mottled maroon shale containing
hematite nodules ; 5 10

Cow Run Sandstone:

Shale, mottled buff-green, massive; includes calcareous

nodule layer with nodules up to 1 foot in diameter 13 0

Shale, mottled blue-brown, thin-bedded, with some

calcareous nodules 4 6

Covered interval 20 0

Portersville Limestone and Shale: shale, black, laminated,
calcareous, very fossiliferous limestone nodules; grades
upward into blue to gray, finely laminated clay shale,
sparingly fossiliferous, with nonfossiliferous lenticular
limestone nodules; upper contact covered 13 9

Anderson Coal 1 4

Clay shale, gray to buff, slumped and partly covered; scat-
tered limestone nodules near base probably represent
the Bloomfield Limestone 6 5

Cambridge Limestone: limestone, mottled, maroon, gray,

and green; marine, ferruginous, weathers dark brown 8

Covered interval 2 9

Another section exposing the units between the Ewing Lime-
stone and the Ames Limestone is in south-central sec. 30, Buffalo

1967 A NEW CONEMAUGH MARINE MEMBER 3

Pennsylvanian:

Conemaugh Group: Ft in

Ames Limestone 1 3

Shale and covered interval, including Harlem Coal float __ 28 7

Noble Limestone: limestone, white, nodular; interbedded
in olive clay shale, fossiliferous, marine; thin yellow
clay, possibly representing the Upper Bakerstown
Underclay, at base 1 3

Shale, mottled tan and buff 1 1

Shale, variegated, green and maroon, hematite nodules in

upper third, carbonaceous near top 5 1

Ewing Limestone: limestone, irregularly bedded, fossil-
iferous, fresh-water 8 3

The area between this section and the type section is very poor-
ly exposed, and the Noble Limestone could not be located. Just
north of the place of the second section, however, the interval in
which the Noble Member would be expected to occur is occupied
by tan, silty shale which contains plant fragments, estherids, and
fresh-water pelecypods ( Anthraconaia cf. A. arenacea (Dawson) ) .
The Ewing Limestone is exposed along the road in south-central
sec. 19 and along the highway in the SW^SE1/^ sec. 24, Buffalo
Twp., Noble County, Ohio, but no marine fossils were found in the
overlying shales.

The Ewing and Noble Limestones do not seem to be repre-
sented in adjoining parts of Spencer Township, Guernsey County,
and Noble Township, Noble County, where the interval between
the Ames and Portersville Limestones appears to consist for the
most part of channel sandstones or channel sandstones and inter-
bedded shales.

The Noble Limestone has been found at only one other locality,
in the King Quarry at Florence, 1 mile north of Caldwell, Ohio.
A measured section at this locality appears in Denton and others
(1961, p. 194, section 11), although the Noble Limestone Member
is not noted in this section. The limestone is present, however, at
least in the northern portion of the quarry, where it ranges between
four and five feet in thickness and consists of the usual fossiliferous,
white to gray, nodular limestone interbedded with greenish fossil-
iferous shale. In an abandoned quarry just southeast of the King
Quarry, marine fossils are absent at the top of the Ewing. In new

JAMES L. MURPHY AND LARRY PICKING

NO. 1

road cuts due west of the quarry, on the west side of Duck Creek,
the Noble Limestone Member is absent.

A faunal list derived from collections made at the Ava Brick
Company Shale Pit and the King Quarry is given below. Dr. Rich-
ard Hoare, Department of Geology, Bowling Green State Univer-
sity, and Mr. J. J. Burke, Cleveland Museum of Natural History,
have been of considerable help in identifying several of the brachio-
pods and the crinoid material. These collections have been pre-
sented to the Cleveland Museum of Natural History.

King

Ava

Shale

Species

Quarry

Pit

Anthozoa:

Stereostylus amesensis Bebout

Echinodermata :

Erisocrinus typus (Meek and Worthen)

Crinoid fragments

Bryozoa:

Unnamed solid ramose fistuliporid

(Vide Moore and Dudley, 1944, p. 265)

Polypora cf. P. valida Moore

Septopora robusta Ulrich

Megacanthopora cf. M. fallacis Moore

Brachiopoda:

Lingula carbonaria Shumard

Trigonoglossa nebrascensis (Meek)

Orbiculoidea missouriensis (Shumard)

Derbya crassa (Meek and Hayden)

Chonetinella flemingi (Norwood and Pratten)

C. flemingi alata (Dunbar and Condra)

C. verneuiliana (Norwood and Pratten)

Juresania nebrascensis (Owen)

Pulchratia ovalis (Dunbar and Condra)

Antiquatonia portlockianus crassicostatus

(Norwood and Pratten)

Linoproductus prattenianus

(Norwood and Pratten)

Neospirifer dunbari King

Punctospirifer kentuckensis amesi

Hoare and Sturgeon Ms

Crurithyris planoconvexa (Shumard)

Composita subtilita (Hall)

Pelecypoda:

Dunbar ella cf. D. striata (Stevens)

Acanthopecten carboniferus (Stevens)

Cirripedia:

Trypetesa caveata Tomlinson

1967

A NEW CONEMAUGH MARINE MEMBER

Vertebrata:

Petalodus ohioensis Safford
Deltodus sp.

Certain elements of the Noble Limestone fauna have not been
reported previously from Conemaugh beds older than the Ames
Limestone. These include the coral Stereostylus amesensis Bebout,
along with two brachiopods, Pulchratia ovalis (Dunbar and Con-
dra) and Punctospirifer kentuckensis amesi Hoare and Sturgeon
Ms. The unnamed solid fistuliporid bryozoan found in the Noble
Limestone is of interest, for the only other large fistuliporids known
from the Pennsylvanian of Ohio are specimens of Cyclotrypa that
occur in the Ames Limestone. The presence of the forms cited
above in the Noble Limestone indicates that the Noble fauna is
trending toward that of the Ames and is more advanced in that
direction than other Lower Conemaugh marine faunas. However,
the chonetid brachiopod Chonetinella, which is characteristic of
other Conemaugh marine limestones underlying the Ames, is com-
mon in the Noble fauna. This is in marked contrast with the fauna
of the Ames, in which Neochonetes is the common and characteris-
tic chonetid.

Unfortunately, present knowledge of the Conemaugh faunas
is not refined enough to permit precise correlation of units over
wide areas. The Portersville Limestone of Ohio has been corre-
lated with the Woods Run Limestone of western Pennsylvania
(Sturgeon and others, 1958, p. 118, 122; Flint, 1965, p. 70-71), but
this has been done solely on the basis of its stratigraphic position.
Burke (1958, p. 302) described three distinct marine members in
the interval between the Cambridge and Ames members: Woods
Run proper, the underlying Nadine Member, and the overlying
Carnahan Run Shale; as a consequence there is considerable doubt
about which particular unit corresponds to the Portersville Member.

The Noble Limestone definitely occurs above the Ewing Lime-
stone, which is known to underlie the coal distinguished in Ohio
as the Upper Bakerstown. The horizon of this coal appears to be
indicated by smut and underclay underlying the Noble Limestone
in the sections given above.

The coal which in western Pennsylvania is regarded as the
Upper Bakerstown is found in the vicinity of Bakerstown, Alle-
gheny County, Pennsylvania, and lies 50 to 60 feet below the Ames

JAMES L. MURPHY AND LARRY PICKING

NO. 1

Limestone. In the course of recent investigations we have dis-
covered that a marine shale immediately overlies that coal in the
vicinity of Bakerstown. The shale, carrying a predominately mol-
luscan fauna, was found at two localities, one just west of the junc-
tion of Pennsylvania State Routes 8 and 910, 1.2 miles south of
Bakerstown, and the other just east of the toll station at Inter-
change 4 on the Pennsylvania Turnpike, 2.9 miles south of Bakers-
town. The presence of a marine shale overlying the Upper Bakers-
town Coal in Pennsylvania, and apparently occupying the same
stratigraphic position as the marine Noble Limestone of Ohio,
strongly suggests that the two beds are correlative.

Burke (personal communication, September 1967) has found no
conclusive evidence that the Upper Bakerstown Coal is present in
his area of investigation in the Kiskiminetas Valley, and he is
inclined to believe that in that area this coal should occur higher
in the stratigraphic section than any of the marine beds underlying
the Ames Limestone that he distinguished in 1958.

In summary, the Noble Limestone in Ohio represents a pre-
viously unrecognized marine unit that occurs in the interval be-
tween the Portersville and the Ames Limestones. However, more
field study is required before the relationship of the Noble Lime-
stone to various Conemaugh marine units in Pennsylvania can be
determined.

1967

A NEW CONEMAUGH MARINE MEMBER

REFERENCES CITED

Burke, J. J., 1958, New marine horizon in the Conemaugh Formation: Science,
v. 128, no. 3319, p. 302.

Denton, G. H., and others, 1961, Pennsylvanian geology of eastern Ohio: Geol.
Soc. America Guidebook for Field Trips, Cincinnati Meeting, pp. 131-205,
1 pi., 20 figs.

Flint, N. K., 1965, Geology and mineral resources of southern Somerset County,
Pennsylvania: Pa. Geol. Survey Co. Report C56A, 267 pp., 13 pi., 55 figs.,
15 tables.

Hoare, R. D. and Sturgeon, M. T., Pennsylvanian Brachiopoda of Ohio: Ohio
Geol. Survey (in press).

Moore, R. C., and Dudley, R. M., 1944, Cheilotrypid bryozoans from Pennsyl-
vanian and Permian rocks of the midcontinent region: State Geol. Survey
of Kans. Bull. 52, Rept. of Studies, Pt. 6, pp. 229-408, pis. 1-48.

Sturgeon, M. T., and associates, 1958, The geology and mineral resources of
Athens County, Ohio: Ohio Geol. Survey Bull. 67, 600 pp., 11 pis., 34 figs.,
30 tables, maps.

MANUSCRIPT RECEIVED SEPTEMBER 15, 1967

K 4> I

^rMlA

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO DECEMBER 30, 1967 NUMBER 2

PRELIMINARY DESCRIPTION OF A
NEW GONIOPHOLID CROCODILIAN

Charles C. Mook*

ABSTRACT

A goniopholid crocodile, Eutretauranosuchus delfsi n.g.,
n.sp. from the Morrison Formation, Jurassic, near Canon City,
Colorado, is described. The new crocodilian is unique in show-
ing a palatal opening anterior to the internal nares and medial
to the palatal vacuities.

INTRODUCTION

In 1957, a party under the leadership of Edwin Delfs, M.D.
collected fossil reptilian remains from the Morrison Formation near
Canon City, Colorado. Included in this material was a fairly com-
plete skeleton of a small crocodilian. This, along with other rep-
tilian remains, is now in the collection of the Cleveland Museum
of Natural History. When the crocodilian material was prepared,
it was evident that it possessed goniopholid characters. Other
characters that are unique indicated a new genus for which Dr.
Delfs suggested the name Eutretauranosuchus signifying “doubly
pierced palate”) . The material was assigned to me for description.
I wish to express my thanks for this privilege to Dr. Delfs and to
the authorities of the Cleveland Museum.

The illustrations of the type are derived from photographs made
by the Photographic Department of the American Museum of
Natural History.

*Dr. Mook died in October, 1966, while this note was in press.

CHARLES C. MOOK

NO. 2

SYSTEMATIC DESCRIPTION
Order CROCODILIA
Suborder MESOSUCHIA
Family GONIOPHOLIDAE
Genus EUTRETAURANOSUCHUS, new genus.

Diagnostic characters: The skull is moderately long in pro-

portion to its breadth. There is a pronounced depression at the base
of the snout, and slightly developed preorbital ridges, suggesting
the condition in the living caimans. The postorbital bars were
clearly subdermal. The prefrontal bones extend farther forward
than the frontal. The nasal bones widen anterior to the prefrontal
tips. The nasal bones do not enter the external narial aperture at
the surface. The frontal bone extends forward to the level of the
eleventh maxillary teeth. The frontoparietal suture is located rather
far back, permitting a considerable participation of the frontal in
the anterior borders of the supratemporal fenestrae. The inter-
fenestral bar is relatively broad and flat, and its edges are slightly
uprolled. The supratemporal fenestrae are of moderate size. They
are elongate oval in shape. The quadratojugal bones have sharp
spines, resembling those of Crocodylus.

The internal narial aperture is unusually long and slender, its
length being several times its breadth, and relatively much longer
than in other crocodilians of its general size. It appears to be
divided, at the palatal surface, at least, by a slender bar of bone.
The composition of this bone is not clear, but it may be made of
slender anterior processes of the pterygoid.

Anterior to the internal narial aperture is another opening,
similar in character to it, but smaller. Like the internal nares it is
divided by two slender bones, which appear to be processes of the
palatines. The presence of this opening is unique among croco-
dilians.

The palatine fenestrae are large and are broad anteriorly.
Their anterior ends lie opposite the eighteenth maxillary teeth.
The pterygoid, omitting the anterior processes mentioned above, is
short and broad. The palatine bones extend forward to the level
of the thirteenth maxillary teeth.

1967

A NEW GONIOPHOLID CROCODILIAN

The lower jaw is long, slender and unusually low anteriorly.
The symphysis includes eight mandibular teeth on each side. The
splenial bones enter the symphysis. They extend forward to the
level of the seventh mandibular teeth.

Twenty-four alveoli are present in each ramus. Of these num-
bers 3 and 4 are slightly larger than the rest which are subequal
in size. Numbers 1 and 2 appear to open on the external surface
of the jaw. The first eighteen alveoli have separate walls, nineteen
to twenty-four merge together. Numbers 5 to 11 are visible from
the side. The external mandibular fenestrae are small and are
oblique in position. They are much longer than high. The anterior
internal fenestra is moderately large and the posterior one is small.

The articular process is excessively short and the articular
surface, that articulated with the quadrate, is also very short.

The teeth, such as they are preserved, are rather small, are
striated, and slightly bladed.

The posterior external portions of both rami are distinctly pit-
ted, the anterior portions only slightly so.

The vertebrae are amphicoelous, and are goniopholid in char-
acter.

The limb bones are typically crocodilian and appear to be
rather short for their breadth and in proportion to the dimensions
of the skull and vertebrae.

Type: Eutretauranosuchus delfsi Mook.

Eutretauranosuchus delfsi, new species

Type : Well-preserved skull and jaw; most of the precaudal

and a few caudal vertebrae; a few limb bones; scutes; Cleveland
Museum of Natural History, No. 8028.

Type locality and level: Red Canyon, north of Canon City,

Colorado; lower level of the Morrison Formation, upper Jurassic.

Diagnostic characters : As for the genus, of which this species
is the sole representative.

The name is given in honor of Dr. Edwin Delfs in recognition
of his services to paleontology in the discovery of this and other
specimens of fossil reptiles in the Canon City area.

CHARLES C. MOOK

NO. 2

MEASUREMENTS

Skull:

Length, tip of snout — occipital condyle 218 est.

Length, tip of snout — posterior border of cranial table 207 est.

Length, base of snout — occipital condyle 88

Length, base of snout — posterior border of cranial table 80

Breadth across base of snout 76

Breadth across quadratojugals 108

Breadth across cranial table, anterior end 61

Breadth across cranial table, posterior end 68

Length of right orbit 22.5

Breadth of right orbit 16

Length of left orbit 24

Breadth of left orbit 18

Length of right supratemporal fenestra 25

Breadth of right supratemporal fenestra 14

Length of left supratemporal fenestra 27

Breadth of left supratemporal fenestra 15

Length of right palatine fenestra 43

Breadth of right palatine fenestra 24

Length of left palatine fenestra 43

Breadth of left palatine fenestra 27

Breadth across pterygoids 85

Lower Jaw:

Right Left Both rami

ramus ramus together

mm mm mm

Length, total 256 261

Length, tip — posterior end of tooth row 159

Length, posterior end of tooth row —

end of articulae 111

Length, symphysis 54 53

Length, external mandibular fenestra 35 38

Length, articular (articulating surface

plus process) 39 37.5

Height, external mandibular fenestra 11 9

Breadth across symphysis

1967

A NEW GONIOPHOLID CROCODILIAN

COMMENTS

The characters of the palate present a considerable departure
from the condition usually seen in mesosuchian crocodiles. The
very large internal narial aperture, together with the anterior
opening, which must have entered the narial passage are unique.
They indicate specialization among the goniopholids comparable to
the wide range of specializations that may be noted among the
Crocodilidae in the Eusuchia.

MANUSCRIPT RECEIVED MARCH 6, 1965

CHARLES C. MOOK

NO. 2

Plate 1. Eutretauranosuchus delfsi n.gen., n.sp. Type, skull, C.M.N.H. No. 8028.
Superior view. Two -thirds natural size.

1967

A NEW GONIOPHOLID CROCODILIAN

Plate 2. Eutretauranosuchus delfsi n.gen., n.sp. Type, skull, C.M.N.H. No. 8028.
Inferior view. Two-thirds natural size.

CHARLES C. MOOK

NO. 2

Plate 3. Eutretauranosuchus delfsi n.gen., n.sp. Type, lower jaw, C.M.N.H. No.
8028. Superior view of the two rami, not joined together. Two-thirds natural
size.

1967

A NEW GONIOPHOLID CROCODILIAN

Plate 4. Eutretauranosuchus delfsi n.gen., n.sp. Type, lower iaw, C.M.N.H. No. 8028. Upper
figure: left ramus, external view; lower figure: right ramus, external view. Two-thirds
natural size.

NO. 2

CHARLES C. MOOK

Plate 5. Eutretauranosuchus delfsi n.gen., n.sp. Type, lower jaw, C.M.N.H. No. 8028. Uppe]
figure: left ramus, internal view; lower figure: right ramus, internal view. Two -third;
natural size.

' n

klRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO SEPTEMBER 12, 1968

NUMBER 3

PACHYLOCRINIDS FROM THE CONEMAUGH GROUP,
PENNSYLVANIAN

J. J. Burke
ABSTRACT

Four new species of pachylocrinid inadunate crinoids from the Cone-
maugh Group, Pennsylvanian, are described and attributed to the genus
Plummericrinus. Three of these forms, P. monongaliensis sp. nov., pre-
sumably from the Brush Creek Limestone, Monongalia County, West
Virginia, P. emilyae sp. nov., from the Ames Limestone, Carroll County,

Ohio, and P. pittsburghensis sp. nov. from the Ames Limestone, Allegheny
County, Pennsylvania, have dorsal cups with slight interradial notches,
pits at the corners of the cup plates, and generally resemble P. colubrosus
(Moore) although the cups are bowl-shaped rather than subpyriform; a
fourth species, P. nettingi sp. nov., from the Cambridge Limestone, Alle-
gheny County, Pennsylvania, has a truncate cone-shaped cup, prominent
interradial notches and the characteristic splay-toothed outline of
P. mcguiri (Moore).

The presence, in P. monongaliensis sp. nov., of muscular articulations
connecting anal X and the right tube plate with the two tube plates
overlying them is taken to indicate that some flexion of the anal tube
was possible in this and other pachylocrinid species where similar
articulations characterize these anal plates.

Four new species of pachylocrinid inadunate crinoids are de-
scribed in the present paper and attributed to the genus Plum-
mericrinus. The specimens on which the descriptions are based are
from the Conemaugh Group, Upper Pennsylvanian, of Ohio, West
Virginia and Pennsylvania. In addition to material from the Cleve-
land Museum of Natural History, specimens described herein or
used for comparative study were borrowed from the Carnegie Mu-
seum and the United States National Museum. I am grateful to the
authorities of these institutions for permission to study and to de-
scribe these specimens.

I wish to acknowledge the assistance of Mr. Bruce Frumker for
photography, and my wife, Emily, for arranging the illustrations.

( JUN 4 1969

J. J. BURKE

NO. 3

SYSTEMATIC PALEONTOLOGY

Family PACHYLOCRINIDAE Kirk, 1942
Genus PLUMMERICRINUS Moore and Laudon, 1943
PLUMMERICRINUS MON ON G ALIEN SIS i sp. nov.

Fig. 1

Diagnosis : Dorsal cup about as wide as that of Plummericrinus
colubrosus (Moore) but higher (form ratio .54) and bowl-shaped,
rather than subpyriform; walls of basal concavity less steep, infra-
basals and basals less reduced. Interradial notches slight, pits at
corners of cup plates, radianal strongly reduced, brachials without
keels.

Fig. 1. Plummericrinus monongaliensis sp. nov. Holotype, a dorsal cup, U.S.N.M.
no. 27488, from the ? Brush Creek Limestone, Conemaugh Group, Monongalia
County, West Virginia, a, dorsal view; b, posterior view; c, ventral view, X3.

Holotype : U.S. National Museum no. 27488, a dorsal cup with first
two primibrachs (B and C rays) and first two secundibrachs (B
ray).

Occurrence: ?Brush Creek Limestone, Conemaugh Group, Upper
Pennsylvanian.

Locality : Monongalia County, West Virginia.

Repository: United States National Museum, Washington, D.C.

Description: The dorsal cup of this species is truncate bowl-shaped;
the outline in dorsal and ventral view is subround and asym-
metrical because the plates of the posterior interradius and the right
posterior radial bulge outward to some extent. There is a distinct
basal impression, which is relatively deep for Plummericrinus ; its
diameter is about one-third that of the cup. The proximal portion
of the round stem is preserved; the articular surface of one colum-

1 Named for Monongalia County, West Virginia.

1968

PACHYLOCRINIDS FROM THE CONEMAUGH

nal shows 14 culmina and a subpentagonal lumen. The infrabasals
extend beyond the stem, flaring downward steeply to meet the
basals. The basals participate in the basal concavity and are strong-
ly curved proximally; beyond the basal plane they are rather gently
convex both longitudinally and transversely. These plates are only
slightly wider than high, and are not impressed along their common
sutures, but gentle furrows mark the sutures between the basals
and the radials. At the tips of the radials and the basals there are
definite deep pits at the plate junctions.

The radial plates are about a third wider than high. They are
more convex longitudinally than transversely, but not strongly con-
vex in any case. The furrows along the interradial sutures are
stronger than those along the radial-basal sutures, however. There
are definite but not pronounced interradial notches at the summits
of these plates.

The radial articular surface is nearly as wide as the radial
plate, although the transverse ridge is not. In lateral view, the
outer marginal ridge sags downward with a gentle arcuate curve.
In ventral view, the ridge has a stronger arcuate outline and bounds
a relatively deep outer ligament area. The marginal ridge is not
sharply delimited from the adjacent ligament area, which is den-
ticulate. The ligament pit furrow is slitlike and not strongly ex-
cavated; the ligament pit is also slitlike. The transverse ridge is
distinct and not denticulate. The inner ligament area is somewhat
deeper than the outer. The oblique fossae are prominent, and a
wide intermuscular notch separates the two triangular muscle
areas, which face outward strongly.

Three plates occupy the posterior interradial area. Pits and
furrows mark their junctions inter se and with adjoining basals and
radials. Of these plates, anal X is the largest. The radianal is about
half the size of anal X, and the right tube plate is about one-third
as large as anal X. Anal X abuts against the left posterior radial
and rests on the truncated tip of the posterior basal and the upper
side of the radianal. Below the right tube plate it makes contact
with the right posterior radial; above that place the triangular right
tube plate is wedged between anal X and the right posterior radial.
The radianal is an elongate triangle; below its contact with anal X it
is obliquely disposed between the posterior basal and the right pos-
terior radial; it narrows to a tip and barely contacts the right pos-
terior basal below. There is a distinct pit shared by the four plates

J. J. BURKE

NO. 3

at this place of contact. Between the anal X and the posterior basal,
there is a deep furrow, rather than a pit, and there is also a furrow
marking the contact between anal X and the right posterior radial.
Both anal X and the right tube plate extend only slightly higher
than the summits of the radials and their distal articular surfaces
lie in the same plane. Both plates resemble the radials in having
external ligament furrows and pits, but the furrows and pits are
less slitlike. These plates also exhibit transverse ridges and broad
flat internal ligament areas.

The primibrachs of the B and C rays are present. They are
wider than high, do not have median keels, and are axillary. The
axillary face of the B ray primibrach is showing; it is divided by
a median ridge into two articular surfaces, one for each first secun-
dibrach. Each of these articular surfaces is composed of inner and
outer ligament areas separated by a transverse ridge. The outer
ligament areas are denticulate; the outer ligament pits lie within
furrows. The two first secundibrachs of the C ray are also pre-
served. They are a little wider than high and nonaxillary. The
upper articular surfaces show definite transverse ridges and inner
and outer ligament areas resembling those of the radials. The outer
ligament areas show denticulation.

Very fine granulose ornamentation is visible on the cup and
brachial plates at a magnification of 20 X-

Discussion: It is of interest to note that in this species the distal
surfaces of anal X and the right tube plate are in essentially the
same plane. Strimple (1961, p. 98) noted that in Haerteocrinus,
Texacrinus, and Plummericrinus there is a trend toward this ar-
rangement of these plates. Strimple has noted also (1952, p. 246)
that in Haerteocrinus turbinatus “The upper surfaces of both RX
and X have muscular fossae somewhat comparable to those of the
RR. There is an outer ligament pit bordered by a transverse ridge.
The outer marginal ridge and transverse ridge possess denticles,
and other crenulations are found behind the muscle scar. Inter-
muscular notches are narrow and well defined, that of anal X being
to the left of center, and of RX to the right of center.”

This description in general applies to the facets of these anal
plates in P. monongaliensis also. Furthermore, I have removed the
matrix from the distal facets of these plates in the holotype of
P. mcguiri (U.S.N.M. no. 141074) ; they show the same structures

1968

PACHYLOCRINIDS FROM THE CONEMAUGH

noted by Strimple and the visible portion of the proximal facet of
the large tube plate that succeeds anal X has a similar articular
surface. A succeeding tube plate partly covers the distal surfaces
of anal X and the right tube plate in the holotype of P. colubrosus
(U.S.N.M. no. 141091) and I have not been able to make full prep-
aration of the articular structures, but it is evident that they are
much like those shown in P. mcguiri. In both of these Permian spe-
cies the distal articular surfaces of anal X and the right tube plate
are in approximately the same plane, as in P. monongaliensis. It is
also worth noting that the largest plates of the tegminal sacs of
P. mcguiri and P. colubrosus are in two rows, initiating with anal X
and the right tube plate.

It appears evident that in some species of Haerteocrinus, Tex-
acrinus and Plummericrinus, anal X and the right tube plate were
joined to the two tube plates distal to them by muscular articula-
tion. Furthermore, because in certain of these species the distal
articular surfaces of anal X and the right tube plate are in the
same plane, it is obvious that these two plates formed a common
hinge with the proximal articular surfaces of the two tube plates
immediately overlying them.

Because these anal plates were connected by muscular union
of the same type as that which connected the arm plates of these
species, it follows that some movement of the anal tube must have
been possible — approaching, at least, that which was attained by
the arms. In the case of Plummericrinus this is of special interest.
Moore (1939, p. 221) , because of the proximal location of what ap-
peared to be the anal vent in P. colubrosus, stated that “this sug-
gests that the main part of the tube functions mainly in connection
with respiration or the water circulatory system rather than as cov-
ering for an elongated convoluted gut.” Possibly, if such were the
case, movement of the tube contributed in some way to respiration
or circulation of water.

Muscular articulation of this type is confined to the opposed
faces of these pairs of anal plates of the anal tube in P. mcguiri, at
least. In making further preparation of the holotype specimen of
that species I uncovered two small plates of a row which originates
at the left distal corner of the tube plate which overlies anal X.
However, most of the distal surface of the latter tube plate articu-
lated with a second large plate above, to the right of which I ex-
posed another large plate which articulated with the plate which

J. J. BURKE

NO. 3

rested above the right tube plate. Of these additional plates, all of
the articular faces that are showing have moderate to deep fossae,
with rims which, on the external side, at least, are crenulated. This
type of articulation is usually interpreted as ligamentary. However,
this does not rule out the possibility that contractile fibers might
have been present, which would have contributed to flexion of the
tube.

Of the anal plates within the dorsal cup of Plummericrinus
monongaliensis, both anal X and the radianal are reduced, although
anal X is still functional, as indicated by the articular surface noted
previously. The separation of anal X from the radianal is unusual.
I have not seen this condition in any other specimen of Plummeri-
crinus. It may be only a simple case of variation and not character-
istic of the species. However, coupled with the reduction of the
radianal, it suggests that in this species we may be dealing with a
trend toward loss of the radianal through resorption.

The specific locality and horizon from which this excellent little
specimen was taken cannot be determined with certainty. I found
it in the fossil crinoid collection of the United States National Mu-
seum. It is identified on the original handwritten label as “Hydre-
ionocrinus discus (Meek & W.)” by “C.S.” with the notation “Upper
Garb. /locality lost.” Another label (typewritten) gives the same
information but adds “Found among Monongalia Co., W.Va. fossils
(C.S.).” I gather that the “C.S.” refers to Charles Schuchert.

It appears quite likely that this dorsal cup is the specimen
noted as “ Erisocrinus , undetermined species” by Meek (1871) in a
list of fossils reported by Stevenson (1871) to have been taken from
the Uffington Shale, Conemaugh Group, near Morgantown, W.Va.
The fossils listed were embodied in the United States National Mu-
seum collection, but many of them cannot be found at the present
time. It has since been determined (Price, 1917; Murphy, 1966)
and from my own field observations, that the Uffington Shale in the
Morgantown area does not carry a marine fauna, and that the fos-
sils identified by Meek probably came from the Brush Creek Lime-
stone. It was my impression, in the course of preparation of the
holotype of Plummericrinus monongaliensis that the matrix adher-
ing to the specimen was characteristic of the Brush Creek, and Mr.
James Murphy, who has collected marine fossils from the Brush
Creek in the Morgantown area, is of the same opinion (personal
communication, Feb. 16, 1968).

1968

PACHYLOCRINIDS FROM THE CONEMAUGH

Linear measurements of the holotype, in millimeters, are given
below:

Height of dorsal cup 3.7

Greatest width of cup 6.9

Ratio of height to width 0.54

Height of basal concavity 0.6

Width of basal concavity 2.3

Width of infrabasal circlet 1.9

Height of basal (raB) 2.6

Width of basal (raB) 2.8

Height of radial (aR) 2.3

Width of radial (aR) 3.5

Width of transverse ridge (aR) 3.3

Length of suture between basals 1.5

Length of suture between radials 1.1

Height of radianal 1.3

Width of radianal 1.3

Height of anal X 1.5

Width of anal X 1.3

Height of right tube plate 0.9

Width of right tube plate 0.8

Height of first primibrach (C ray) 2.3*

Width of first primibrach (B ray) 3.1

Height of first secundibrach (B ray) 2.0

Width of first secundibrach (B ray) 2.2

* Approximate

PLUMMERICRINUS NETTINGI2 sp. nov.

Fig. 2

Diagnosis : Dorsal cup resembling that of Plummericrinus mcguiri
(Moore) in having prominent interradial notches, outflaring radials
and lacking pits at corners of plates, but differing in smaller size
(width 13.8 mm) and in tending toward truncate-cone shape, with
shallower basal concavity and less elongate basals and radials.

Fig. 2. Plummericrinus nettingi sp. nov. Holotype, a dorsal cup, Carnegie Mu-
seum no. 29857, from the Cambridge Limestone, Conemaugh Group, near Vero-
na, Allegheny County, Pennsylvania, a, dorsal view; b, posterior view; c, ven-
tral view, X2.

2 Named for Dr. M. Graham Netting, Director, Carnegie Museum.

J. J. BURKE

NO. 3

Holotype : Carnegie Museum no. 29857, a dorsal cup.

Occurrence: Cambridge Limestone, Conemaugh Group, Upper

Pennsylvanian.

Locality: Sylvan Run, near Verona, Allegheny County, Pa. (Lat
40° 29' 25" N, Long 79° 50' 50" W) .

Repository: Carnegie Museum, Pittsburgh, Pa.

Description: The dorsal cup of this crinoid is modified truncate
cone-shaped. In dorsal view the rounded outline of the basal circlet
is in marked contrast with the scalloped outline of the radial circlet
resulting from prominent interradial notches at the summits of the
outflaring radial plates.

The outer ring of one columnal of the round stem is preserved
and shows faint traces of culmina. The basal concavity is distinct.
The width of the concavity is less than one-third that of the cup,
but the height is only about one-seventh of the cup height.

The infrabasals are relatively prominent, with tips extending
well beyond the stem. These plates are slightly convex and slope
downward gently to meet with the basals.

The basals participate in the basal concavity and are sharply
convex longitudinally in that region. Beyond the concavity the
basals are moderately convex. These plates are almost a fifth wider
than high. There are gentle furrows along the sutures between the
basals, and distally the tips of the basals are sharply incurved to
meet the interradial sutures.

The radials are about half as high as wide and flare outward
strongly in the midregion, where they show little curvature. Along
the sides, however, they are distinctly incurved, producing distinct
hollows that border the interradial sutures. There are prominent
interradial notches, with broad slopes, and the articular surface
does not occupy the full width of the radial.

In general, the radial articular surfaces are not strongly devel-
oped and they are not well defined. The outer ligament area may
be described as deep, although less so than the inner ligament area.
The outer marginal ridge is arcuate, and the ligament area is den-
ticulate. The ligament-pit furrow is slitlike, but I cannot distin-
guish a distinct ligament pit. There is a fairly strong transverse
ridge, which is not denticulate. The inner ligament area displays
moderate oblique fossae and outward-facing muscle areas, sepa-

1968

PACHYLOCRINIDS FROM THE CONEMAUGH

rated by a broad intermuscular notch. There are also traces of the
intermuscular furrow.

The radianal is the largest plate of the posterior interradius.
Proximally it is wedged in between the posterior basal and the
right posterior basal. Laterally it contacts anal X on the left and
the right posterior radial on the right. Distally it bears the right
tube plate, which intervenes between the right posterior radial and
anal X. Anal X is somewhat smaller than the radianal; in addition
to its contacts with the radianal and the right tube plate it rests on
the truncate tip of the posterior basal below and also abuts against
the left posterior radial on the left. The right tube plate is much
smaller than the other two anal plates. The superior surfaces of
anal X and the right tube plate are damaged and I cannot deter-
mine their original structure. The right tube plate is somewhat out
of place and overrides the underlying radianal slightly. Probably
the superior surfaces of anal X and the right tube plate were in
essentially the same plane.

The ornamentation consists of small irregular pustules, visible
at a magnification of 10 X- It appears to have been absent from the
infrabasals and proximal portions of the basals.

Discussion : As the diagnosis indicates, this form is quite distinct
from the other Conemaugh species described in this paper, par-
ticularly in showing strong interradial notches and outflaring radial
plates. In these respects it resembles the Permian Plummericrinus
mcguiri (Moore), although differing from that species in its smaller
size, different shape of the dorsal cup, shallower basal concavity
and less elongate basals and radials. However, none of these char-
acteristics would appear to bar it from the ancestry of the Permian
species.

Linear measurements of the holotype, in millimeters, are as
follows:

Height of dorsal cup 5.4

Greatest width of cup 13.8

Ratio of height to width 0.39

Height of basal concavity 0.8

Width of basal concavity 4.2

Width of infrabasal circlet 4.4

Height of basal (raB) 3.6

Width of basal (raB) . 4.7

Height of radial (aR) 3.4

Width of radial (aR) 6.0

Width of transverse ridge (aR) 4.8

Length of suture between basals 1.7

J. J. BURKE

NO. 3

Length of suture between radials 2.2

Height of radianal 4.0*

Width of radianal 3.7

Height of anal X 3.0

Width of anal X 3.4

Height of right tube plate 1.9

Width of right tube plate 2.5

* Estimated

PLUMMERICRINUS EMILYAE3 sp. nov.

Fig. 3

Diagnosis: A species resembling Plummericrinus colubrosus

(Moore) in showing slight development of interradial notches and
in having pits at corners of cup plates, but a larger form (estimated
crown height 34 mm, width of dorsal cup about 10 mm) having
cup more bowl-shaped and basal concavity quite shallow, infra-
basals and basals less reduced, summits of anal X and right tube
plate not in same plane and keels of brachials absent or insignifi-
cant.

Fig. 3. Plummericrinus emilyae sp. nov. Holotype, a dorsal cup with portions
of the arms attached, Cleveland Museum no. 4000, from the Ames Limestone,
Conemaugh Group, about 3 miles east of Carrollton, Carroll County, Ohio.
a, anterior view; b, posterior view; c, dorsal view, X2.

3 Named for my wife, Emily G. Burke.

1968

PACHYLOCRINIDS FROM THE CONEMAUGH

Holotype : Cleveland Museum no. 4000, a dorsal cup with portions
of the arms attached.

Occurrence: Ames Limestone, Conemaugh Group, Upper Penn-
sylvanian.

Locality: Joe Skinner Quarry, NW^NE1/^ sec. 13 (Lat 40° 34' 20"
N, Long 81° 01' 20" W) Center Township, about 3 miles east of Car-
rollton, Carroll County, Ohio.

Repository: Cleveland Museum of Natural History, Cleveland, Ohio.

Description: Unfortunately, this specimen is contained in a very
adherent matrix, and despite reasonable care in preparation, the
component plates of the crown suffered some abrasion, although
not to the extent that most of the salient characteristics were de-
stroyed.

None of the arms is complete, but by conservative estimate the
height of the crown was at least seven times that of the dorsal cup.
The dorsal cup is truncate bowl-shaped. The stem is not preserved.
Although the base of the cup has been abraded, it is evident that
there is a distinct basal impression, the width of which is slightly
more than one-third of the width of the cup; however, the max-
imum height of the impression, at most, is somewhat less than one
millimeter. The stem impression is relatively small, a little less than
a third as wide as the infrabasal circlet, consequently the infrabasal
plates extended well beyond the stem. The greatest height of the
basal impression is in the vicinity of junction of the infrabasal and
basal plates; the stem impression does not extend to the basal plane
of the cup, but the infrabasals flare upward slightly to meet the
basals, and most of the wall of the impression is formed by the
proximal portions of the basals.

The basal plates are a little more than one-fifth higher than
wide. The left posterior appears to retain approximately its original
slopes. Proximally, these plates show| strong upward curvature,
forming the walls of the basal impression. Distally, the longitudinal
curvature is less pronounced; laterally!, especially in the area be-
tween the basal sutures, the curvature is again strong, although
somewhat less so than in the proximal region. These slopes form
hollows bordering the basal sutures; there are also hollows, al-
though generally broader, between the basals and the radials.

J. J. BURKE

NO. 3

Apparently there were pits at all the angles where the basals
meet with the radials and the anal plates, but traces of some of
them have been obliterated.

The radials are nearly twice as high as wide. In general they
tend to flare outward, their slopes diverging somewhat from those
of the basals, although the tendency is not pronounced. These plates
for the most part are nearly as convex as the basals along their
height and are moderately convex transversely. Distally they curve
inward strongly along a lunate area that sags downward from the
summit corners of the plates; a similar upward arching area shows
in the proximal portion of the primibrachs, consequently the sutures
between the radials and the primibrachs are gaping. Interradial
notches are present, but they are very slight. Except for the left
anterior radial, which shows a trace of the outer ligament pit, de-
tails of the articular surfaces are concealed because the first primi-
brachs are all preserved.

The radianal plate is the largest of the anal plates; the right
and left plates are somewhat smaller and anal X is slightly smaller
than any of the others. The radianal extends proximally to the
suture between the right posterior basal and the posterior basal.
It abuts by a short side against the right posterior basal; a longer
side borders the right posterior radial, and another long side rests
against the posterior basal. A shorter side on the left contacts anal
X. Above, the radianal supports the right tube plate, which fits in
between the right posterior radial and anal X. The right tube plate
extends for about half its height above the summit of the right pos-
terior radial; on the left, above anal X, it rests against the left tube
plate. Anal X rests on the truncate tip of the posterior basal below,
and for most of its height contacts the left posterior radial, although
it extends slightly above the summit of that plate. Above, it sup-
ports the left tube plate.

The first primibrachs of the A and C rays are higher than wide;
those of the other rays are wider than high. These plates were
abraded in preparation, but some of them now show sharp ridges
along the midline that may be traces of keels. These plates are axil-
lary. The right division of the A ray is complete to a short distance
beyond the third isotomous division of the inner branch. There are
six secundibrachs; the sixth is axillary. Ten tertibrachs are pre-
served in the outer ray and no evidence of branching. However,

1968

PACHYLOCRINIDS FROM THE CONEMAUGH

the inner ray shows nine tertibrachs and the ninth is axillary.
Four quartibrachs remain on one side following the bifurcation, one
on the other. On the inner branch of the left division of the A ray,
both first quartibrachs are preserved, resting on the axillary terti-
brach. I find no evidence of tertiary division in either of the outer
branches of the A ray.

A few slender pinnules are showing along the sides of some of
the arms.

Discussion : This specimen is of interest because it shows the arm
structure of the species, at least in part. Branching takes place on
the first primibrach and again on the sixth secundibrach or there-
abouts. From this point on there is no indication, from what is
showing, of any further bifurcation on the outer branches, but the
inner branches bifurcate again at the ninth tertibrach, as shown in
one branch at least. This type of arm structure also characterizes
Texacrinus, wherein, because the outer branches fail to show fur-
ther bifurcation, the structure is termed exotomous. However, no
species of Plummericrinus of which I know differs from Texacrinus
in this respect, that is, they do not show any further bifurcation of
these outer branches, so exotomous arm structure does not consti-
tute a valid generic distinction between Texacrinus and Plum-
mericrinus.

As a matter of fact, generic distinction between Texacrinus and
Plummericrinus finds little support when based on other characters
as well. Moore (1940, p. 144) cited the absence of interradial
notches in the holotype of Texacrinus gracilis, but the presence or
absence of these notches is not regarded as a generic character in
other inadunate crinoids; Plaxocrinus, for example, is composed of
species some of which exhibit the notches, whereas others do not.
The interposition of the radianal between the posterior basal and
anal X appears to have been regarded as a generic character of
Texacrinus by Strimple (1961, p. 94). However, variations in the
arrangement of these anal plates will quite likely be found in species
of Texacrinus. As regards Plummericrinus, the radianal separates
anal X from the posterior basal in the holotype of P. bellirugosus.
It is also interesting to note that in four paratypes of P. mcguiri,
Moore (1939, p. 209) found the radianal separating anal X from the
posterior basal.

J. J. BURKE

NO 3

The slight basal impression shown in P. emilyae constitutes at
best a specific distinction that points up the transition in this re-
spect from genera such as Haerteocrinus, in which the impression
is lacking. Of more significance, it seems to me, is the fact that in
P. emilyae the distal surfaces of anal X and the right tube plate are
not in the same plane. In this respect P. emilyae appears to differ
from most, if not all, other species of Plummericrinus.

Linear measurements of the holotype, in millimeters, are given
in the following tabulation:

Height of crown 34.0*

Height of dorsal cup 4.9**

Greatest width of cup 10.1**

Height of basal concavity 0.7**

Width of basal concavity 3.7

Diameter of stem impression 1.0**

Width of infrabasal circlet 3.2

Height of basal (laB) 4.4

Width of basal 3.7

Height of radial (laR) 2.5

Width of radial (laR) 4.8

Length of suture between basals 1.6

Length of suture between radials 1.4

Height of radianal . 2.3

Width of radianal 2.3

Height of anal X 1.8

Width of anal X 1.8

Height of right tube plate 2.0

Width of right tube plate 2.0

Height of left tube plate 2.0

Width of left tube plate 2.0

Height of first primibach (A ray) 4.3

Width of first primibach (A ray) 4.2

Height of first secundibrach (A ray) 2.8

Width of first secundibrach (A ray) 2.5

Height of first tertibrach (A ray) 1.5

Width of first tertibrach (A ray) 1.6

* Estimated

** Approximate

1968

PACHYLOCRINIDS FROM THE CONEMAUGH

PLUMMERICRINUS PITTSBURGHEN SIS4 sp. nov.

Fig. 4

Diagnosis: Dorsal cup truncate bowl-shaped, estimated width 15
mm; interradial notches slight, pits at corners of plates; plates only
moderately convex; basals, radianal and anal X relatively large;
posterior basal makes narrow contact with anal X; primibrachs
without keels.

Fig. 4. Plummericrinus pittsburghensis sp. nov. Holotype, an obliquely crushed
dorsal cup, Carnegie Museum no. 29858, from the Ames Limestone, Conemaugh
Group, Brilliant Cutoff, Pittsburgh, Allegheny County, Pennsylvania. Dorsal
view, X2.

Holotype : Carnegie Museum no. 29858, a dorsal cup with primi-
brachs of the C and D rays attached.

Occurrence : Ames Limestone, Conemaugh Group, Upper Penn-
sylvanian.

Locality : Brilliant Cutoff (Lat 40° 29' N, Long 79° 54' 20" W) Pitts-
burgh, Allegheny County, Pennsylvania.

Description : Although the dorsal cup of the type has undergone
compression, which makes it extremely difficult to determine the
original outlines and dimensions, the specimen nevertheless pre-
serves many characters which mark it as a representative of a dis-
tinct species.

The dorsal cup was apparently truncate bowl-shaped. There is
a distinct basal concavity, the width of which, by rough estimate,
was about one-fifth that of the cup. The depth of the concavity
cannot be determined because portions of the stem are still in place.
Three columnals and part of a fourth are showing; the columnals
are thin, and each is estimated to bear about 35 culmina.

4 Named for Pittsburgh, Pennsylvania.

J. J. BURKE

NO. 3

The infrabasals project slightly beyond the stem, and slope
steeply downward. The circlet is evidently somewhat displaced in
the present specimen, but the tip of one of the plates appears to be
nearly in its original place, indicating that the plate leveled off dis-
tally into a triangular area that fitted in between adjacent basals.

Evidently the only participation of the basals in the basal con-
cavity was in the form of a sharp curvature of their proximal por-
tions that contributed to form a slight part of the wall of the con-
cavity. Despite compaction, it is apparent that the slopes of the
basal circlet were fairly gentle. The basal plates are a little wider
than long, and are relatively large. In general they are moderately
convex, although there are broad hollows bordering the basal su-
tures, and deep pits at the angles where they meet the radials, the
radianal, and anal X.

It is difficult to estimate the original slopes of the radials, but
from all appearances they did not deviate sharply from those of the
basals, and this would have made for a bowl-shaped cup. The radials
are gently convex longitudinally and transversely. These plates are
about two-fifths wider than long. The interradial notches are
slight. Details of the articular surfaces are poorly preserved be-
cause of wear. The outer ligament area was apparently shallow in
comparison with the inner area, and bore denticles. The outer
ligament furrow is slitlike; I cannot distinguish a distinct ligament
pit. There are traces of the transverse ridge, which quite evidently
did not extend the full width of the plate. There are indications of
fairly strong oblique fossae. The intermuscular notch is broad and
the slopes of the muscle areas faced outward.

The radianal is a large plate which approximates anal X in size;
the right tube plate is smaller. The proximal tip of the radianal
extends to the suture between the posterior basal and the right
posterior basal. On the left a sharp angle of the radianal limits,
but does not cut off, the contact of anal X with the posterior basal
below. On the right, the radianal extends to the tip of the right
radial; distally the radianal bears the right tube plate. The right
tube plate extends a short distance below the summit of the right
posterior radial and its entire left side abuts against anal X. The
left side of anal X contacts the left posterior radial. Details of the
articular surfaces of anal X and the right tube plate are obscure,
although the place of the outer ligament area in anal X is indicated

1968

PACHYLOCRINIDS FROM THE CONEMAUGH

by traces of denticulations. I think it is quite likely that the
articular surfaces of these two plates were in the same plane.

Two axillary primibrachs are preserved. These plates are wider
than high, somewhat constricted at the sides, and lack keels. The
articular surfaces for the secundibrachs resemble those of the ra-
dials, except that there is a definite ligament pit in the outer liga-
mentary area.

Discussion : In several respects this species resembles Plummeri-
crinus uddeni (Moore and Plummer) although in P. pittsburghensis
the interradial notches are apparently less prominent, pits are pres-
ent at the corners of plates, and there is a narrow contact between
the posterior basal and anal X. In the holotype of P. uddeni the
radianal is interposed between the posterior basal and anal X, con-
sequently the latter two plates are not in contact. This arrange-
ment of anal plates probably prompted Strimple (1961, p. 94) to
refer P. uddeni to Texacrinus. However, as I have pointed out
previously (p. 13) some species of Plummericrinus show this plate
arrangement as a variation, and it may be anticipated that the same
variation will be found in P. pittsburghensis when additional speci-
mens are available.

Linear measurements, in millimeters, of the holotype specimen
are summarized below. Because the specimen has been subject to
compaction, length, rather than height measurements of the plates
were taken.

Greatest width of cup : 15.0*

Width of basal concavity 3.0

Width of infrabasal circlet 3.9

Length of basal (raB) 4.5

Width of basal (raB) 4.8

Length of radial (aR) 3.7

Width of radial (aR) 6.2

Length of suture between basals 2.8

Length of suture between radials 2.1

Length of radianal 3.9

Width of radianal 3.7

Length of anal X 3.5

Width of anal X 4.0

Length of right tube plate 2.6

Width of right tube plate 2.9

Length of first primibrach (C ray) 4.1

Width of first primibrach (C ray) 4.9

* Estimated

J. J. BURKE

NO. 3

REFERENCES CITED

Meek, F. B., 1871, Lists of Carboniferous fossils from West Virginia — descrip-
tion of new species: In Stevenson, J. J., A geological examination of Mon-
ongalia County, West Virginia: West Virginia Univ. Board Regents Rept. 3,
p. 68-73.

Moore, R. C., 1939, New Crinoids from Upper Pennsylvanian and Lower Per-
mian rocks of Oklahoma, Kansas and Nebraska: Denison Univ. Bull.,
J. Sci. Lab., v. 34, p. 171-272.

1940, Crinoids from the Upper Carboniferous and Permian

strata in Texas: Univ. Texas Bull. 3945, p. 1-468.

Murphy, J. L., 1966, The Pennsylvanian pelecypod genus Palaeoneilo Hall and
Whitfield: Jour. Paleontology, v. 40, p. 867-876.

Price, J. M. Jr., 1917, The Uffington shale of West Virginia — absence of marine
fauna. In Hennen, R. V., Braxton and Clay Counties: West Virginia Geol.
Survey County Rept. p. 807-816.

Stevenson, J. J., 1871, A geological examination of Monongalia County, West
Virginia: West Virginia Univ. Board Regents Rept. 3, p. 40-67.

Strimple, H. L., 1952, The arms of Haerteocrinus: Washington Acad. Sci. Jour,
v. 42, p. 245-248.

1961, Late Desmoinesian crinoids: Oklahoma Geol. Surv.

Bull. 93, p. 3-189.

MANUSCRIPT RECEIVED JUNE 19, 1968

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO SEPTEMBER 27, 1968 NUMBER 4

THE HOBSON SITE: A FORT ANCIENT COMPONENT
NEAR MIDDLEPORT, MEIGS COUNTY, OHIO

James L. Murphy
ABSTRACT

Salvage archeology at the Hobson Site (33Ms-2) on the Ohio River
IV2 miles downstream frorp Middleport, Meigs County, Ohio, has re-
vealed an important Fort Ancient component. On the basis of the dom-
inant pottery types, the component is assigned to the Feurt Phase. It is
suggested that the site represents the early Feurt Phase, and the age of
the site is estimated as approximately 1100-1200 A.D. Minor traces of
Archaic, Woodland, and later Late Prehistoric components were also
noted.

INTRODUCTION

The Hobson Site (33Ms-2) located on the west bank of the Ohio
River, IV2 miles downstream from Middleport, Meigs County, Ohio,
was partly and hurriedly excavated in August, 1966. Mr. George
Orahood of Prospect, Ohio, the construction engineer at the site,
notified the author on the day that ground was broken. The follow-
ing three days were spent by the author and another Ohio Univer-
sity student, Mr. Tim Watkins, in salvaging as much material as
possible. The site is now the location of the new Middleport sewage
treatment plant. It is regrettable that, although the plant had been
in the planning stage for over ten years, we did not learn of the
existence of the archeological site until after construction had
started. On the other hand, it is fortunate that Mr. Orahood was
interested in the site and notified both the Ohio Historical Society
and Ohio University. In addition to Mr. Orahood, acknowledgment
should be made to Mr. Emmett Conway of the Institute for Re-
gional Development, Athens, Ohio, Mr. Ross Goodwin and Mr. Wat-
kins, who were of help in the salvage operation, and to Mr. Conway,
who photographed the burial found at the site.

JAMES L. MURPHY

NO. 4

LOCATION AND SITE DESCRIPTION

The site lies on the northern bank of Storys Run at the con-
fluence of that stream with the Ohio River, IV2 miles downstream
from Middleport, Ohio, and IV4 miles upstream from Lakin, West
Virginia, at an elevation of approximately 570 feet above sea level.
The bluff on which the site is located lies about 25 feet above the
present normal pool elevation of the Ohio River. The land is now
owned by the city of Middleport, and the site name is taken from
Hobson Junction, slightly less than half a mile upstream, on the
Chesapeake and Ohio Railroad.

Concentrated occupation occurred over an area of at least two
acres, the heaviest midden occurring near the edge of the river
bluff. Midden material averaged less than a foot in thickness, thin-
ning rapidly to the north and east until, on the periphery of the site,
only a few flint chips, potsherds, bone fragments, and clam shells
were seen, the latter being the most conspicuous in the freshly ex-
posed cuts made by the earth-moving equipment. Some areas of
fire-burned subsoil were found near the river bank and clam shells
were so abundant in the same area as to constitute a veritable shell
midden. The only other features encountered were the dozen or so
burials in the cemetery located at the northeast edge of the midden
area.

Excavation technique, if it can be called that, consisted largely
of surface collecting behind the earth-moving equipment as the ma-
chinery passed back and forth over the site. When the equipment
was not in operation, portions of the midden were scraped down to
the subsoil with mattocks. It was in this manner that the one com-
plete burial excavated was first found, although it was removed
with more care than we were able to give to excavation of the rest
of the site.

Flint and Stone Material
PI. 1

Relatively little chippage was recovered from the site, partly
because in our haste it was thought preferable to concentrate on
the bone, shell, and pottery refuse material. Of the 274 chips saved,
96 percent represented pebble chert collected from the river gravel.
The bulk of this river chert is composed of Devonian and Pennsyl-
vanian material, at least a third of it being dense black flint from

1968

HOBSON SITE, MEIGS COUNTY, OHIO

the Upper Mercer and Kanawha members. With the exception of
a few chips of Flint Ridge (Vanport) flint, none of the remaining
chippage could be identified.

Flint artifacts from the site include 13 triangular points of peb-
ble chert which fall into three relatively distinct types: large, crude,
possibly unfinished points (5) , small, thin, well-made points with
convex bases and concave sides (7) , and small, thin points with
straight sides and a straight base (1). Though admittedly a small
sample, the predominance of the convex base in association with
concave sides which frequently are produced to form basal ears
may prove to be a temporally distinct point type. It is a distinct
minority type on Feurt Phase Fort Ancient components in the
Hocking Valley.

There are also three large blanks of river pebble chert, two
broken elongate points (one of Brush Creek chert and the other of
Kanawha flint) , a broken, stemmed Adena point of an unidentified
flint, and a side-notched Archaic point of Flint Ridge flint. The base
of this last point is either fractured or else represents the original
surface of the striking platform; the lower portions of the notches
are moderately ground. A small, crude end-scraper on a blade, two
long blade-like spalls with retouch flaking along a portion of one
end, and four unidentifiable fragments of points complete the list
of worked flint material.

The remaining stone artifacts consist of a small hematite celt
and a roughly chipped basalt chopper. A single fragment of un-
worked cannel coal was also found.

Worked Bone Material
PI. 2

Bone artifacts were relatively uncommon at the site and con-
sisted mainly of awls and bone beads. There are two bird-bone
splinter awls, a deer-ulna awl, and the tip of a second deer-ulna
awl. Bird bone was utilized for beads, two of which were recov-
ered; there is also a fragmentary bead made from the radius of
a rabbit. Other bone artifacts are a small piece of a turtleshell cup,
part of a bone beamer, and a small, spatulate object apparently
made from a deer longbone.

JAMES L. MURPHY

NO. 4

Pottery
PL 3, 4

The potsherds collected at the Hobson Site may be divided into
five distinct types based on differences in surface finish, temper,
and rim decoration. Two of these types, however, are very poorly
represented.

Type 1: Limestone-tempered, cordmarked ware represented

by 25 body sherds and one rim sherd. Average thickness of ten
body sherds ranging from 4.6 mm to 8.2 mm is 6.6 mm. Temper
fragments measure up to 10 mm in diameter. The solitary rim
sherd has a slightly everted, cordmarked lip with the cordmarking
vertical on the rim area. Most of the sherds contain minor amounts
of sand, presumably accidental inclusions in the paste.

This type, as represented in the collection, cannot be distin-
guished from the late Middle Woodland Watson Ware of the Upper
Ohio Valley nor from the Late Woodland Peters Cordmarked Ware
from the Scioto and Hocking Valley drainages.

Type 2: Shell-tempered, cordmarked ware represented by 30

body sherds and a single rim sherd. Average thickness of ten sherds
ranging from 4.7 mm to 9.1 mm is 6.6 mm. The rim sherd is strong-
ly everted, the rim forming an angle with the body of approximate-
ly 145°. Cordmarking is vertical at the rim, the lower half of which
is decorated with crudely incised vertical lunules; the upper half
of the rim has the cordmarking obliterated by horizontal brushing
or combing. Insofar as is known, this decorated sherd is unique.
There is certainly no reason to assign it to Fort Ancient, although
the body sherds are indistinguishable from Fox Farm Cordmarked.

Type 3: Fox Farm Salt Pan: A single shell-tempered sherd
apparently comes from a Fox Farm salt pan. Although the lip is
missing, the curvature of the sherd makes such an assignment like-
ly. Depth of the pan is estimated as having been about 4 cm.

Type 4: Shell-tempered, plain vessels with plain rims. This

type is represented by 26 rim sherds which vary considerably in
rim profile; 21 are nearly vertical with rounded (12), flattened (6),
or incised (3) lips. Three sherds, possibly from the same vessel,
have sharply everted, slightly thickened rims 1 to 2 cm wide. Two
sherds have convex rims 35 to 40 mm wide, separated from the body
of the vessel by a rather abrupt flexure. Although those sherds
with sharply everted rims are indistinguishable from Madisonville

1968

HOBSON SITE, MEIGS COUNTY, OHIO

Plain rim sherds, the affinities of the bulk of the plain rim sherds
lie with Griffin’s Feurt Focus.

Type 5: Decorated rims which are considered to be variants

of the plain shell-tempered ware (Type 4). The 791 plain, shell-
tempered body sherds from the site cannot, of course, be separated
into the two rim types. Among the decorated rims there are seven
small sherds with multilinear incising and four small sherds with
cord-wrapped stick impressed punctates, sherds too small for one
to form an idea of the entire pattern. Sixteen sherds represent ves-
sels which had horizontal multiple incising around the rim; the
incised lines vary from four to six in number and from 1.9 mm to
3.3 mm in width. The incised lines vary from closely spaced (2.5
mm apart) to 15 mm apart. The incising is generally crude and the
lines are only roughly parallel. In four instances there is an addi-
tional incised design element below the horizontal banding: alter-
nating triangles filled with oblique incised lines, a crude, fine-lined
rectilinear guilloche, and two sherds on which the incised pattern
cannot be determined. Three rim sherds have the opposed chevron
design considered typical of the Feurt Phase (Murphy, ms.).

A single sherd combines cord-wrapped stick punctates as a rim
border and a rim decorated with oblique incised lines. The only
other known occurrence of this motif is a sherd of Chillicothe
Brushed from the McGraw Site (Prufer, et al., 1965, p. 55) , though
it would be rash to postulate any direct relationship between the
two sites.

Finally, there are ten strap handles or fragments of handles
and two lug handles. One of the lugs is a mammiform lug de-
tached from the vessel; the other is simply a small horizontal shelf
or ridge attached to the lip. The strap handles are crudely made,
large, with parallel sides. Two are punctate and have a castellated
lip. One of these has a row of punctates at the base of the rim, as
is the case with a third handle, which does not have castellations.
Another strap handle occurs with a similar raised rim area rather
than castellations, but it does not have punctates at the base of the
rim. The remaining handles are fragmentary or detached from the
vessels. Included in the count are two broken castellations which
do not belong to any of the rims collected.

JAMES L. MURPHY

NO. 4

POTTERY SUMMARY

The strap handles recovered from the site certainly do not re-
semble Madisonville or “Clover Complex” material. Nor does the
material strongly resemble Monongahela wares. It seems to fit best
the known characteristics of the type Feurt Plain. The large amount
of incised rim sherds would also suggest a relationship with the
Feurt Phase, especially when one considers the prevalence of the
opposed-chevron motif. The dominance at the site of smooth-
surfaced shell-tempered sherds is an even more striking similarity
between the Hobson Site and Feurt Phase components in the Hock-
ing Valley.

At variance with our present knowledge of Feurt ware is the
common use of horizontal incised motives at the Hobson Site. Such
sherds are indistinguishable from sherds labelled Monongahela In-
cised from the Speidel Site, Ohio County, West Virginia (Mayer-
Oakes, 1954, fig. 13, 14) . However, it is interesting to note that
Griffin (1943, pi. xx, fig. 1) illustrates an “atypical” Baum sherd
with horizontal incising from the Feurt Site. A horizontal rim motif
is also known from the McCune Site, Athens, Ohio, a Feurt com-
ponent (Murphy, ms.) . Even more significant is the dominance of
horizontal incising at the Blain Site, a Baum-like component cur-
rently under study by Dr. Olaf Prufer, and its presence on shell-
tempered ware from Cole Complex sites excavated by R. S. Baby.

It is tempting to suggest that horizontal incising is a carry-over
from the Baum Phase into the early Feurt Phase, but such an
hypothesis would certainly be premature. Unfortunately, because
of the hurried manner in which the Hobson Site was excavated, it
is a moot point whether one or two Late Prehistoric components
are represented and, if two, whether or not they are contempora-
neous. Until a definite Feurt site is discovered at which horizontal
incising is a common decorative element, the best course to follow
is to assign tentatively the incised Hobson sherds to Feurt Incised.

Historic Material
PI. 2

Numerous fragments of glass and china were noted at the site,
although unfortunately none of the china ware was saved. The only
recognizable metal objects found were a penknife and an un-
identifiable fragment of cast iron. The foundations of an old farm-

1968

HOBSON SITE, MEIGS COUNTY, OHIO

house lay immediately to the northeast of the site, and the historic
items may best be ascribed to that source.

The presence of an historic Delaware village site in this imme-
diate area should be noted, although it is not thought to have been
located at the Hobson Site. Lewis Evans’ 1755 map locates “Kish-
keminetas old T.” on the north side of the Ohio River, slightly less
than half way between “The big Bent” and the Kanawha River.
Hanna (1911, v. 2, p. 142) locates Kiskiminetas’ Town eight miles
above the mouth of the Kanawha, which would place it in the
vicinity of the town of Cheshire, 2Vz miles downstream from the
Hobson Site. The limited time available did not permit a general
survey of the area, but it is believed that such a survey would dis-
cover the exact location of Kiskiminetas’ Old Town.

Burials
PL 5, 6

The cemetery associated with this village site lay immediately
to the northeast of the midden deposit, possibly extending even as
far as the farmhouse mentioned above. At least nine burials were
noted, scattered over an area of about an acre, between the major
area of occupation and the foundations of the farm house. The
burials were uncovered by large earth-moving equipment which
removed them completely within the course of an hour. Already
badly damaged by the first swath cut by the excavating equipment,
none of the burials were salvaged. No grave goods were noted, and
all of the burials appeared to be flexed or semi-flexed.

The single burial retrieved from the Hobson Site was discov-
ered in the midden area, about ten yards from the river bank. It
lay from 6 to 12 inches below the surface but did not extend into
the yellow subsoil; there was no noticeable grave outline. The
burial was semiflexed, with legs folded, right arm at the side, and
left forearm placed on the abdomen. The patellae, distal ends of the
femora, and proximal ends of the tibiae had been removed by plow-
ing, but the rest of the burial was in good condition. Although no
grave goods accompanied the burial, the individual is unusual in
showing abundant evidence that he had been either murdered or
executed. The first and third lumbar vertebrae contain triangular
projectile points which are lodged in the ventral portion of the cen-
tra and must therefore have penetrated the abdomen. Two addi-

JAMES L. MURPHY

NO. 4

tional flint projectile points were found, one in the chest cavity and
the other touching the medial edge of the right scapula. Finally,
there was an antler projectile point located in the chest cavity.

Skeletal measurements are presented in table 1. It should be
noted that the skull was somewhat distorted by warping, so that
some of the measurements are only approximate at best. The indi-
vidual was male and, based on the dentition and features of the pu-
bic symphysis, 22 to 23 years of age. Dental caries had begun to de-
velop only on the upper right second molar, the upper first molars,
and the lower molars; even here decay was confined to small “pin-
hole” perforations, the largest measuring .9 mm in diameter. The
mandibular molar cusp pattern is of the Y-5 type only on the first
molar; that of the second is of the +5 type. Excessive crowding of
the incisors and canines had produced moderate malocclusion. The
presence of Wormian bones may also be noted, as well as the pres-
ence of an olecranon perforation in the left humerus. A very inter-
esting pathological condition of the skeleton is the ankylosis of
a portion of the vertebral column, all nine vertebrae from the
second cervical to the third thoracic being fused. When com-
pared with the skeletal data available from the Madisonville Site
(Hooten, 1922, p. 83-134) , three differences are noted: the basion-
bregma length is greater than that found in nearly all of the Madi-
sonville crania, the angle of the mandible is smaller than that of
any of the Madisonville mandibles, and the nasal index is slightly
lower than any of those given for the Madisonville skeletons. It is
unfortunate that none of the other burials were recovered, but if
the measurements of a single skeleton can be given any weight, it
is likely that the inhabitants of the Hobson Site were physically
distinct from the Madisonville people.

VERTEBRATE FAUNA

A list of the species identified in the bone refuse is given in
table 2 and requires little comment. Butchering marks were noted
on 29 deer bones: 7 astragali, 8 humeri, 4 ulnae, 2 scapulae, 2 radii,
5 calcanea, and 1 femur. The cuts on the humeri were more prox-
imal than the similar marks described by Guilday et al. (1962, p. 73
fig. 8) from the Eschelman Site and were generally located on the
shaft or the very base of the shaft. One of the right calcanea is
unusual in having cut marks on the anteroproximal surface of the
bone. The femur has two cut marks on the lateral surface of the

1968

HOBSON SITE, MEIGS COUNTY, OHIO

great trochanter. The wolf ulna has a sharp cut on the postero-
lateral surface of the olecranon, at the top of the semilunar notch.
Measurements of the 17 measured deer astragali, in milli-

meters, are as follows:

Length

Width

Thickness

Mean

39.8

25.9

22.6

Range

35.4-42.6

24.5-27.8

21.0-24.1

These measurements are

significantly lower than those from the

Eschelman Site, but the explanation probably lies in there being

a greater proportion of does and young animals at the site rather
than the presence of a subspecies characterized by its small size.
The 12 deer jaws that could be age-graded form a uniform series
ranging from about 8 months to 6 years of age, seven of the speci-
mens falling within the 3 to 4 year age bracket.

Naiad Material

As mentioned in the description of the site, freshwater clam
shells were so abundant in portions of the site that the term shell
midden was applicable to those areas. The list of species is given
in table 3. Noteworthy is the presence of Quadrula metanevra var.
wardi, considered a small river and creek form, which might sug-
gest that the shellfish were collected from both the Ohio River and
its tributary, Storys Run. Ortmann (1919, p. 49-50), however, notes
occurrences of this form as far down the Ohio River as Parkers-
burg. The present occurrence would merely extend the distribution
slightly farther downstream. Comparison with faunal lists from the
Childers, Globe Hill, and East Steubenville sites, 150 miles up-
stream, reveals relatively few differences. The complete absence of
Cyclonaias and Ptychobranchus in such a relatively large sample is
difficult to explain, but the only other differences are a compara-
tively greater percentage of Elliptio crassidens and Quadrula cylin-
drica at the Hobson Site.

CONCLUSIONS

The Hobson Site has yielded slight evidence of occupation dur-
ing Archaic, Late Woodland, and late Late Prehistoric times. The
major occupation occurred during middle Late Prehistoric times,
during the Feurt Phase, and probably rather early in that phase.
The only Feurt component having yielded an acceptable radio-

JAMES L. MURPHY

NO. 4

carbon date is the McCune Site at Athens, Ohio (Murphy, ms.)
which yielded a date of 1235 A.D. The 1180 A.D. date for the Gra-
ham Site, a Baum component at Logan, Hocking County, Ohio,
would seemingly pinpoint the date of the Hobson Site as very close
to 1200 A.D. If importance is attached to the similarity between the
Hobson ware and as yet undescribed pottery from the Blain Site,
the date might be estimated as closer to 1100 A.D., for the Blain
Site itself dates around 1000 A.D. (O. H. Prufer, personal commu-
nication, June, 1967). Affinities can also be seen with the Speidel
Site, which has been referred to the Monongahela Complex. Al-
though apparently close to the historic Kiskiminetas Town, it is
unlikely that the Hobson Site is the exact location of that settle-
ment.

It is regrettable that the Hobson Site could not have been ex-
cavated with the care and thoroughness it deserved. The material
salvaged from the construction site provides a glimpse of what
might be expected from a site transitional from the Baum to the
Feurt Phase. Griffin (1943, p. 209) speaks of the region between
the Madisonville and Feurt sites as “terra incognita,” and the same
might be said for that portion along the Ohio River between Proc-
torville and Marietta. If the Hobson Site is typical, this region must
contain numerous rich and important sites as yet untouched though
rapidly being destroyed.

TABLE 1

Skeletal Measurements and Indices
(Measurements in millimeters)

Cranial Measurements

a Glabello-occipital length
b Maximum breadth
c Basion -bregma height
d Mean thickness left parietal
e Minimum frontal diameter
c' Auricular height

183 1 Nasal breadth

136 m Orbital height — left

147 n Orbital breadth — left

4.4 r Interorbital breadth

91 s Biorbital breadth

121 t External palate length

507 u External palate width

355 v Condylo-symphyseal length

323 w Bicondylar width

124 x Height of symphysis

37
41
18

101

105

128

Horizontal circumference
Nasion-opisthion arc

Transverse arc
f Bizygomatic diameter

(Midfacial breadth)

g Total facial height
h Nasion-prosthion height
i Basion-nasion length
j Basion-prosthion length
k Nasal height

128 mh Mandibular height

73 y Bigonial diameter

110 Minimum breadth of left

96 ascending ramus

55 Mean angle of mandible

110

113

1968 HOBSON SITE, MEIGS COUNTY, OHIO

Cranial Indices

b/a Cranial index

74.2

y/e Fronto-gonial index

121

c/a Length/height

80.3

e/f Zygo-frontal index

c/b Breadth/height

108

m/n Left orbital index

90.2

a~r^)^C- Cranial module

155.3

r/s Interorbital index

17.8

e/b Fronto-parietal index

66.9

1/k Nasal index

43.6

g/f Total facial index

81.8

u/'t External palatal index

125

h/f Upper facial index

v/w Mandibular index (1)

f/b Cranio-facial index

91.2

mh/v Mandibular index (2)

55.2

y/f Zygo-gonial index

81.5

Postcranial Measurements

Humerus

Left

Right

Femur (cont.)

Left :

Right

Maximum length

320

340

Middle anteropost.

Maximum diameter

diam.

27.5

26.2

of head

Middle lateral

Maximum middle

diam.

25.7

diameter

21.3

21.8

Middle circum.

Minimum middle

Tibia

diameter

15.1

15.5

Maximum length

360

Middle circum.

Nutrient foramen

Ulna

anteropost. diam.

Maximum length

266

Nutrient foramen

Middle circum.

lateral diam.

22.7

Radius

Middle anteropost.

Maximum length
Middle circum.

251

diam.

Middle lateral

35.2

diam.

19.9

20.2

Femur

Middle circum.

Maximum length
Maximum diameter

—

Clavicle

of head

44.3

45.7

Maximum length

153

149

Subtrochanter

Middle circum.

anteropost. diam.

31.6

34.4

Innominate

Subtrochanter

Height

212

lat. diam.

25.8

26.4

Breadth

147

146

Postcranial Indices

Humerus

Fibula

Middle index

70.9

71.1

Robustness index

12.4

Robustness index

19.4

19.2

Tibia

Ulna

Platycnemic index

56.8

53.8

Robustness index

15.5

Middle index

58.9

57.4

Radius

Robustness index

14.3

Clavicle

Femur

Robustness index

Platymeric index

78.8

76.7

Innominate

Middle index

93.4

98.1

Innominate index

___

68.9

JAMES L. MURPHY

NO. 4

TABLE 2

Vertebrate Remains From the Hobson Site

Species No. of bones

Fishes:

Catastomid sp. 2

Aplodinotus grunniens Rafinesque 1

Ictalurus sp. 3

Unidentified 9

Reptiles:

Terrapene Carolina (Linnaeus) 27

Chelydra serpentina Linnaeus 6

Birds:

Meleagris gallopavo Linnaeus 95

Unidentified 11

Mammals:

Odocoileus virginianus (Zimmermann) 429

Procyon lotor (Linnaeus) 18

Sciurus sp. 9

Cervus canadensis Erxleben 9

Castor canadensis Kuhl 4

Tamias striatus (Linnaeus) 3

Sylvilagus floridanus (Allen) 2

Ursus americanus Pallas 2

Canis lupus Linnaeus 1

Unidentified large mammal bones

(probably deer) 122

Unidentified mammal bones 408

Total number of identified bones 611

Percentage

4.3

15.3

69.0

2.9

1.4

TABLE 3

Molluscan Remains From the Hobson Site

Species Minimum No*

Pelecypoda:

Amblema plicata (Say) 10

Quadrula cylindrica (Say) 18

Q. metanevra wardi (Lea) 9

Q. pustulosa (Lea) 2

Pleurobema cor datum (Rafinesque) 69

P. pyramidatum (Lea) 3

P. clava (Lamarck) 7

Elliptio dilatatus (Rafinesque) 3

E. crassidens (Lamarck) 67

Obovaria subrotunda (Rafinesque) 2

Proptera alata (Say) 1

Ligumia recta latissima (Rafinesque) 3

Lampsilis ovata (Say) 5

L. siliquoidea (Barnes) 1

Percentage

5.0

9.0

4.5

34.5

3.5

33.5

200

* The minimum number of individuals, based upon the maximum number
of either left or right valves of each species.

1968

HOBSON SITE, MEIGS COUNTY, OHIO

Gastropoda:

Anguispira alternata (Say)

A. kochi (Pfeiffer)

Mesodon clausus (Say)

Triodopsis tridentata (Say)

REFERENCES CITED

Evans, Lewis, 1755, A general map of the Middle British Colonies in America
. . . Sold by R. Dodsey, Pall-Mall, London, and by the author in Phila-
delphia.

Griffin, J. B., 1943, The Fort Ancient Aspect: Univ. Michigan Mus. Anthropol-
ogy Papers, no. 20, 392 p., 147 pis., 18 figs., 10 maps.

Guilday, J. E., Parmalee, P. W., and Tanner, D. P., 1962, Aboriginal butchering
techniques at the Eschelman Site (36La-12), Lancaster Co., Pa.: Pennsyl-
vania Archaeologist, v. 32, no. 2, p. 59-83, figs. 1-10.

Hanna, C. A., 1911, The wilderness trail: New York, G. P. Putnam’s Sons, 2 v.

Hooten, E. A., and Willoughby, C. C., 1920, Indian village site and cemetery
near Madisonville, Ohio: Papers of Peabody Mus. Amer. Archaeology and
Ethnology, v. 8, no. 1, 137 p., 30 pis.

Mayer-Oakes, W. J., 1954, The Speidel Site (46-Oh7) Ohio County, West Va.:
West Virginia Arch. Soc., Inc., Publ. Ser. no. 2, 30 p., 15 figs.

Murphy, J. L., ms., Two Feurt Phase Components near Athens, Ohio.

Ortmann, A. E., 1919, A monograph of the Naiades of Pennsylvania. Part III.
Systematic Account of the Genera and Species: Carnegie Mus. Mem., v. 8,
no. 1, 382 p., 21 pis.

Prufer, O. H., and McKenzie, D. H., 1967, Studies in Ohio archeology: Western
Reserve Univ. Press, 368 p., 34 pis., 55 figs.

MANUSCRIPT RECEIVED DECEMBER 29, 1967

JAMES L. MURPHY

NO. 4

EXPLANATION OF PLATES

Plate

1. Stone and flint artifacts. Bottom row: Triangular points and two flake
knives. Second row: Triangular points. Third row: Side-notched, stemmed,
and corner-notched points; two triangular blanks. Top row: Small hema-
tite celt, two large flint scrapers or choppers, and a crude triangular point.

2. Bone artifacts and historic material. Bottom row: Penknife and un-
identifiable fragment of cast iron; bird bone bead. Top row: Deer ulna
awl, bird bone bead, spatulate bone object, antler point, and two bird
bone awls.

3. Pottery. Bottom row: Incised and plain lips on rim sherds of Feurt Plain.
Note absence of flare on both sherds. Second row: Castellated, punctate
strap handle (Feurt) and everted rim sherd (Madisonville?). Top row:
Two Feurt strap handles, both with punctate borders, one with punctate
handle and castellations, the other with plain handle and raised rim area.

4. Pottery. Bottom row: Unusual incised sherd with basal punctations
(photographed obliquely to show decoration), punctate and cordmarked
body sherds. Middle row: Horizontal and diagonally incised sherds, Feurt
Incised (?). Top row: Watson Ware limestone-tempered, cordmarked
rim; plain convex, shell-tempered rim, and unique shell-tempered, cord-
marked rim with brushed and incised motif.

5. Single excavated burial: arrows show two associated triangular flint
points and location of two other associated points.

6. Four lumbar and one thoracic vertebrae with imbedded triangular flint
points.

KIRTLANDIA, NO. 4

PLATE 1

JAMES L. MURPHY

i A

KIRTLANDIA, NO. 4

PLATE 2

JAMES L. MURPHY

I i i I I 1 1 1 1 1 1 i I L 1 115cm

KIRTLANDIA, NO. 4

PLATE 3

JAMES L. MURPHY

KIRTLANDIA, NO. 4

PLATE 4

JAMES L. MURPHY

KIRTLANDIA, NO. 4

PLATE 5

JAMES L. MURPHY

J 1 OcM

KIRTLANDIA, NO. 4

PLATE 6

JAMES L. MURPHY

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO JANUARY 8, 1969 NUMBER 5

AN ANTIACODONT FROM THE GREEN RIVER EOCENE

OF UTAH

J. J. Burke

ABSTRACT

Report is made of a new occurrence of the dichobunid artiodactyl
genus Antiacodon following its discovery in the Green River Forma-
tion in northeastern Utah by a Cleveland Museum of Natural History
field party in 1967. A right ramus of the lower jaw with P4, Mi 3 is
described and identified as Antiacodon pygmaeus (Cope) .

In August, 1967, a field party from the Cleveland Museum of
Natural History made a search for vertebrate fossils in various
Eocene formations of the Uinta Basin, Utah. During a brief visit
to the Powder Wash collecting site in the Green River Formation
in Uintah County, Utah, near the Utah-Colorado state line, Mr. Wil-
liam Hlavin of the Museum party found part of a lower jaw of the
dichobunid artiodactyl Antiacodon, hitherto unreported from the
Utah Green River beds.

The specimen is described in the present paper. Originally a
more extensive study of Antiacodon was projected and was already
under way before I learned that Carnegie Museum also has some
Antiacodon material from the Powder Wash locality. Dr. Craig
Black of that institution is presently engaged in a study of various
Eocene artiodactyles of this type. Inasmuch as there is no point to
my duplicating Dr. Blacks work, I am limiting the present paper
to description of the specimen at hand.

The illustrations accompanying this paper are from superb
pencil drawings of the specimen prepared by Mr. Lawrence B.
Isham.

J. J. BURKE

NO. 5

SYSTEMATIC PALEONTOLOGY

Family DICHOBUNIDAE Gill, 1872
Subfamily ANTIACODONTINAE Gazin, 1958
Genus ANTIACODON Marsh, 1872
ANTIACODON PYGMAEUS (Cope), 1872

Fig. 1

The specimen, C.M.N.H. no. 10930, consists of most of the right
ramus of the mandible with P4, Mt ... It was contained in a block
of sandstone taken from the Powder Wash locality, designated by
Dawson (1968) as “two miles southeast of Powder Springs (sec. 8,

Fig. 1

Antiacodon pygmaeus (Cope). Right ramus of mandible (C.M.N.H. no. 10930),
lateral, occlusal and lingual views. Twice natural size. Douglas Creek Mem-
ber, Green River Formation, Middle Eocene, Powder Wash quarry, Uinta
Basin, Utah.

1969

GREEN RIVER ANTIACODONT FROM UTAH

T. 7 S., R 25 E., S.L.M.), Uintah County, Utah, on the basin side
of Raven Ridge in the eastern part of the Uinta Basin.” Dawson
also states that the United States Geological Survey has determined
that the mammal quarry is in the lower part (Douglas Creek Mem-
ber) of the Green River Formation. Although I indicated the as-
signment with question (Burke, 1935) , my attributing the sand-
stone of the mammal quarry to the upper part of the Green River
Formation remains a regrettable error.

Posteriorly, the lower jaw preserves part of the ascending
ramus and the anterior portion of the masseteric fossa. Anteriorly
it extends slightly in advance of the anterior border of the alveolus
of the canine. About 2 mm of the ventral border is missing be-
neath M4; the thin enamel walls of the hypoconids of M2 and M3
have been chipped, and most of the entoconid cusp of M1 has been
lost. All of these features have been restored in the illustrations

(fig- 1) •

The masseteric fossa is fairly well excavated, although its in-
ferior border is not well defined. The anterior border of the ascend-
ing ramus rises at an angle of about 65 degrees. The ventral border
of the ramus is slightly convex, and the anterior half curves gently
upward. There are three mental foramina, a slight slitlike one
beneath the posterior alveolus of P;;, another beneath the posterior
alveolus of POJ and the third and most prominent beneath the
diastema between P4 and P„.

The alveoli for the premolars anterior to P4 indicate two
diastemata, one between P0 and P3 and a second, slightly shorter,
between Pt and P.>. The single alveolus for Pj is separated from
the alveolus for the canine by about 1 mm of bone and the root of
P., appears to have been slightly larger than that of the canine.

The anterior portion of the ramus is slender because of the
upcurving of the ventral border and the progressive downbending
of the alveolar border. The symphysis is extensive and quite rough;
it narrows posteriorly, and extends back beneath the alveoli for P.,.

The protoconid of P4 is a strong cusp, attenuated anteriorly to
meet with a prominent paraconid. Internal to, and extending in
advance of the paraconid is a relatively large parastylid. The
metaconid, which arises on the posterointernal flank of the protoco-
nid, is a distinct cusp, but definitely smaller than is usual in
Antiacodon. The talonid is defined by the cingulum externally

J. J. BURKE

NO. 5

and is joined posteriorly by the strong posterior crest from the
protoconid. On the internal side of the talonid there is a small but
definite entoconid, from which a short crest extends to the meta-
conid.

All of the molars bear anterior and posterior cingula, although
the anterior cingulum of M , is barely distinguishable. Low cingula
also block the exits of the external valleys of the molars.

The trigonids of the molars are characteristic of Antiacodon.
The metaconids and paraconids are closely appressed, with the
paraconid the higher and larger cusp. The metaconid extends fur-
ther lingually than the paraconid. The protoconid is the lowest
trigonid cusp and might be termed subcrescentic. An anterior crest
from the protoconid joins the paraconid. A less elevated crest ex-
tends from the protoconid to the metaconid.

The molar talonids show large crescentic hypoconids and broad
central valleys, the exits of which are closed by the metaconid and
the entoconid walls, which form a broad V and meet low on the
lingual side. The valley slopes of the entoconids are decidedly flat
surfaces. From the hypoconid the crista obliqua extends to the
protoconid-metaconid crest in M2 and M ,; in M1 the crista obliqua
extends well up on the slope of the metaconid. The posterior crest
from the hypoconid connects with the hypoconulid, but not with
the entoconid, in M1 and M2. In M , this crest connects with the
posterior blade of the entoconid. The crestlike entoconid blade then
descends posteroaeternally toward the base of the hypoconulid. The
posterior cingulum originates along the posterior walls of the
hypoconid and entoconid and expands posteriorly to form the
prominent hypoconulid. An incipient crest extends forward and
downward along the anterior face of the hypoconulid to meet with
the descending blade of the entoconid. The posterior crest from the
hypoconid very definitely does not extend to the hypoconulid.

Discussion: Gazin (1952, 1955, 1958, 1962) has dealt rather ex-
tensively with Antiacodon in relation to other Eocene artiodactyl
genera. Based on his studies, I think there can be little doubt that
the present specimen represents Antiacodon pygmaeus (Cope) .

Although the small metaconid of P4 of C.M.N.H. no. 10930 is not
typical of Antiacodon pygmaeus, it appears to me that this may be
simply a matter of individual variation, and Dr. Gazin, who has

1989

GREEN RIVER ANTIACODONT FROM UTAH

had an opportunity to see the specimen, accords with me in this
conclusion (written communication, Feb. 1, 1968).

Also to be considered are variations in the length of diastemata
in specimens that have been referred to Antiacodon pygmaeus.
Gazin (1958, p. 3) has noted that the diastema between P2 and P3
in A.M.N.H. no. 12697 is about 2.5 mm and in U.S.N.M. no. 1800 it
is 3.9 mm. He also states (ibid., p. 4) that the jaw of A.M.N.H. no.
12697 is preserved to about 1.7 mm in advance of the roots of P2,
but doubtless he found no trace of an alveolus, otherwise he would
have noted it. The diastema between P2 and P., in C.M.N.H. no.
10930 is 1.9 mm, definitely shorter than in either of the above
specimens, and the diastema between P1 and P2 is 1.7 mm, so this
diastema also is apparently shorter than in A.M.N.H. no. 12697.
However, I am inclined to attach little weight to these variations;
they may, as a matter of fact, be related to age — the teeth of
U.S.N.M. no. 1800, for example, are considerably worn in com-
parison with those of C.M.N.H. no. 10930, and my specimen quite
obviously represents a much younger individual.

Because C.M.N.H. no. 10930 preserves more of the anterior
portion of the jaw than previously described specimens of Antiaco-
don, it is of particular interest. It demonstrates, for instance, that
by Middle Eocene time the P1 of Antiacodon had attained the same
size as the canine or was perhaps a little larger. But it also shows
that along with the enlarged Pn the anterior portion of the jaw
remained, as Gazin (1958, p. 4) characterized it, “relatively slender”
and the downbending of the alveolar border anteriorly contributes
to this slenderness to about the same extent as the upcurving of
the inferior border.

Starting with these characteristics, I would expect an Upper
Eocene descendant of Antiacodon to show much the same con-
struction of the lower jaw, and along with it no pronounced in-
crease in size of Pj — in effect, the type of jaw usually found in
Upper Eocene homacodonts.

Gazin’s (1958) Auxontodon combines a lower jaw with a
strongly convex inferior border and a much enlarged, perhaps
caniniform, P1 . The cusp-crest construction of the cheek teeth ap-
pears to conform in all respects with what one would expect of an
Upper Eocene antiacodont. Nevertheless, for the reasons cited

J. J. BURKE

NO. 5

above, I cannot visualize Auxontodon as a direct derivative of
Antiacodon, although there seems no doubt that the two forms
had common ancestry.

Possibly, as has been suggested in regard to other elements of
the North American Middle Eocene fauna, Antiacodon may have
been in some way too specialized for a particular environment to
survive its passing. Study of the skeleton might throw some light
on the matter, but despite years of collecting, skeletal material of
these Middle Eocene artiodactyles is still unknown. To judge from
teeth at least, Antiacodon pygmaeus appears to be the only
artiodactyl represented in the Powder Wash fauna. Almost any
artiodactyl skeletal material taken from that quarry might prove to
belong to Antiacodon pygmaeus and would be worthy of study on
this account.

Measurements of C.M.N.H. no. 10930, in millimeters, are given
below:

Depth of lower jaw at posterior margin of alveolus of Pi 5.0

Depth of lower jaw at anterior margin of first alveolus of P3 6.0

Depth of lower jaw beneath M2, lingually 7.5

Length of cheek tooth series from anterior margin of alveolus of C

to posterior margin of M3 35.5

Length of cheek tooth series from anterior margin of alveolus of Pi

to posterior margin of M3 33.9

Length of diastema between Pi and Pi 1.7

Length of diastema between P2 and P3 1.9

Length of lower premolar series from anterior margin of alveolus for

Pi to posterior margin of P4 18.9

Length of lower molar series, Mi-M3 inclusive 15.0

P4, length : greatest width 4.6 : 2.5

Mi, length : greatest width 4.5 : 2.9*

Mj, length : greatest width 4.4 : 3.4*

M3, length : greatest width 5.8 : 3.3

'Approximate

1969

GREEN RIVER ANTIACODONT FROM UTAH

REFERENCES CITED

Burke, J. J., 1935, Preliminary report on fossil mammals from the Green
River formation in Utah: Carnegie Mus. Annals, v. 25, no. 3, p. 13-14.

Dawson, M. R., 1968, Middle Eocene rodents (Mammalia) from northeastern
Utah: Carnegie Mus. Annals, v. 39, no. 20, p. 327-370.

Gazin, C. L„ 1952, The Lower Eocene Knight formation of western Wyoming
and its mammalian faunas: Smithsonian Misc. Coll., v. 117, no. 18, p. 1-57.

1955, A review of the Upper Eocene Artiodactyla of North

America: Smithsonian Misc. Coll., v. 128, no. 96, p. 1-96.

1958, A new dichobunid artiodactyl from the Uinta Eocene:

Breviora, no. 96, p. 1-6.

1962, A further study of the Lower Eocene mammalian

faunas of southwestern Wyoming: Smithsonian Misc. Coll., v. 144, no. 1,
p. 1-98.

MANUSCRIPT RECEIVED DEC. 5, 1968

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO MARCH 14, 1969 NUMBER 6

A TEMNOSPONDYLOUS LABYRINTHODONT FROM THE
LOWER CARBONIFEROUS

ALFRED SHERWOOD ROMER
Museum of Comparative Zoology, Harvard University

ABSTRACT

An amphibian skull and partial skeleton from the basal part
of the Mauch Chunk Group of the Mississippian of West Virginia
is that of a colosteid temnospondyl described as Greererpeton
burkemorani gen. et sp. nov.

INTRODUCTION

Although amphibian remains are plentiful in the late Carbon-
iferous (the Pennsylvanian Period of American terminology), they
are exceedingly rare in the earlier Carboniferous (the Mississip-
pian) . In this paper there is described for the first time a labyrin-
thodont amphibian skull from the Mississippian of North America.
Until recent decades Lower Carboniferous amphibian remains were
absolutely unknown except in Scotland, and even there specimens
were few. All materials of that age then available were described
by Watson in 1929. Once there are excluded certain specimens now
known to be Upper rather than Lower Carboniferous in age (Pan-
chen and Walker, 1961), the list of finds is a meager one. There are
a few lepospondyls, mainly “adelospondyls” of Watson’s terminol-
ogy; of labyrinthodonts, one skeleton ( Pholidogaster); seven skulls,
most of the peculiar loxommid type, with keyhole-shaped orbits;
a few fragmentary remains.

In North America, no Mississippian amphibians were known
until relatively recently. In 1941, I reported the discovery of
remains of amphibians in the Hinton F ormation of the Mauch
Chunk Group in West Virginia; the bones present, however, were
disarticulated and generally fragmentary, and hence of little mor-

ALFRED SHERWOOD ROMER

NO. 6

phological or evolutionary value. In 1955 I decided to initiate a new
series of attempts to find sites and materials of American Carbon-
iferous amphibians. Considerable Pennsylvanian material was
found, but the only earlier find of value was that of a jaw of very
late Mississippian age, apparently of an anthracosaur, in the Point
Edwards Formation of the Canso Group in Nova Scotia (Romer,
1958) . Further exploration for Carboniferous amphibians has been
continued with success by Dr. Baird of Princeton and Dr. Carroll
of McGill, principally in Nova Scotia. Again, however, their finds
have been mostly Pennsylvanian, and the only Mississippian speci-
men reported is a partial skull of the loxommid from the Point
Edwards Formation (Baird, 1962).

Of especial interest for some time, in the Lower Carboniferous
of the Allegheny region, has been a quarry at Greer, West Virginia.
On several occasions this was visited by parties from the Museum
of Comparative Zoology; amphibian material was found, but of a
fragmentary nature. More successful have been Mr. John J. Burke
and Mr. William E. Moran who had earlier searched intensively in
the “tri-state area” of West Virginia-Pennsylvania-Ohio for Carbon-
iferous and early Permian vertebrates (Moran, 1952; Romer, 1952).
Materials collected by them at Greer, including a skeleton which
is apparently anthracosaurian, are in the U. S. National Museum
collections. The present specimen from Greer not only forms an
addition to our sparse representation of Lower Carboniferous laby-
rinthodonts, but also is important because it increases our knowl-
edge of the stratigraphic distribution of labyrinthodont types.

In his classic papers of 1919 and 1926, Watson, for the first time,
sorted out the then chaotic array of labyrinthodonts into a reasoned
series of subgroups. The Triassic members, he pointed out, form the
Stereospondyli, with intercentra enlarged and pleurocentra gen-
erally absent. These are clearly derivable from the Rhachitomi,
abundant in the Permian, in which in each segment there were
paired small pleurocentra and a fairly large intercentrum. The only
type of Carboniferous vertebrae known to Watson were of the em-
bolomere type, in which both intercentrum and pleurocentra form
complete rings. He therefore concluded that the embolomeres were
the ancestral labyrinthodonts, and that the evolutionary sequence
ran: embolomeres — - rhachitomes — stereospondyls.

This, when Watson proposed it, seemed a reasonable arrange-
ment. But with the passage of time and augmented knowledge of
fossil labyrinthodonts, the Watson classification became increasingly

1969

LOWER CARBONIFEROUS LABYRINTHODONT

unsatisfactory. Therefore, I proposed (Romer, 1947) a different phy-
logenetic scheme, with the labyrinthodonts arrayed dichotomous-
ly in two major subdivisions, Temnospondyli and Anthracosauria.
Since vertebrae similar to the rhachitomous type are found among
the ancestral crossopterygians, I suggested that this was the basic
vertebral structure among labyrinthodonts, and that a main central
group of these forms, to be termed Temnospondyli, continued on as
Rhachitomi through the Carboniferous and Permian, to terminate
in the Stereospondyli of the Triassic. A major side branch, to be
termed the Anthracosauria, consisted of forms in which, in contrast
to temnospondyls, the small paired pleurocentra of ancestral types
became enlarged and fused to form a solid ring-shaped centrum.
Here the major evolutionary line led, with eventual reduction of
the intercentrum, to the Reptilia, with the Seymouriamorpha as a
morphologically transitional group; the Embolomeri, instead of be-
ing truly primitive forms, appear to be an anthracosaurian side
branch in which the intercentrum forms a complete ring, as do the
pleurocentra.

When my scheme of labyrinthodont classification was first pro-
posed, its base in actually known materials was none too secure,
mainly because of the dearth of pre-Pennsylvanian finds. However,
over the last two decades new studies and new finds have tended
increasingly to support it. The description of postcranial remains
of the late Devonian ichthyostegids (Jarvik, 1952) strengthens the
conclusion that the rhachitomous vertebral type is primitive among
amphibians. Restudy of Pholidogaster from the Lower Carbonif-
erous of Scotland (Romer, 1964) beautifully illustrates a stage de-
manded by theory in the development of the pleurocentra toward
the “holospondylous” condition of advanced anthracosaurians.

There still remained, however, a major gap in the early history
of the Temnospondyli. Rhachitomes, contrary to Watson’s earlier
beliefs, are now known (mainly through studies by Baird and Car-
roll) to have been abundant and varied in the Pennsylvanian, but
were long thought to be absent in the early Carboniferous. At the
time I proposed the phylogenetic scheme here followed, I suggested
that the peculiar loxommids, present in the early as well as late
Carboniferous, were rhachitomous rather than embolomerous, as
Watson had believed. This has since been shown by Baird (1957)
to be the case. But surely other rhachitomes in addition to the
aberrant loxommids, with their peculiar keyhole-shaped orbits,
must already have been present in the Mississippian. The present

ALFRED SHERWOOD ROMER

NO. 6

find of a Mississippian rhachitome of more normal structure adds
major support to the belief that the Rhachitomi were already
flourishing in early Carboniferous times.

PROVENANCE OF THE SPECIMEN

The materials here described were collected from the quarry at
Greer, Monongalia County, West Virginia, by Mr. Burke and Mr.
Moran; this site is a commercial limestone quarry, located on
Deckers Creek, about 6V2 miles southeast of Morgantown. The
quarry region has been described by Tilton (1928), Coryell and
Sohn (1938) and McCue, Lucke and Woodward (1939) . The mate-
rial quarried is a massive limestone of the Greenbrier Group. Mr.
Thomas Arkle, Jr. of the West Virginia Geological Survey states
(written communication, Jan. 7, 1969) that this bed is identified as
the Union Limestone, which at Greer Quarry is overlain by the
Bluefield Formation of the Mauch Chunk Group. These determi-
nations are in keeping with studies by Wells (1950) and Flowers
(1956).

Mr. Burke informs me that the specimen is derived from the
Bickett Shale of the Bluefield Formation. The base of the Bickett
Shale is about 16 feet above the Union Limestone at the Greer
Quarry. The specimen is probably from the lower two to three feet
of the Bickett, but its exact horizon is uncertain, since the rock had
fallen from the face of the quarry.

The specimen is, thus, older than the fragmentary remains from
the Hinton district, derived from a horizon considerably higher in
the Mauch Chunk, and still older than the Nova Scotia remains
mentioned above. As noted by Weller and others (1948) , corre-
lation of Lower Carboniferous American formations and zones with
those of Europe is as yet none too certain, but the Greenbrier and
basal Mauch Chunk seem certainly to equate roughly with the up-
per part of the Visean of European terminology, and are pre-
Namurian. The oldest European Carboniferous labyrinthodonts are
from the “Carboniferous Limestone”, Namurian in age. (A few
lepospondyls come from the earlier Oil Shale Group.) Specimens
from the Greer locality thus appear to be the oldest labyrinthodonts
yet known except for the East Greenland ichthyostegids.

The remains here described are in all probability those of a
single individual. A principal block includes the skull and jaws
with, close behind, a disarticulated series of vertebrae and ribs

1969

LOWER CARBONIFEROUS LABYRINTHODONT

together with belly scales which, because of the presence of ele-
ments of the shoulder girdle and front limbs, obviously represent
the anterior part of the trunk. A second series of vertebrae, ribs,
and scales found to the rear of the main block appear to represent
the more posterior region of the trunk. Nearby were a few other
finds of limb and girdle bones, ribs and scales. Since vertebrae and
scales in all cases are of the same type and since the limb and girdle
elements form a nearly complete set of appropriately matching size
for a single individual, it is reasonable to conclude that we are deal-
ing with the disarticulated remains of one animal. The specimen is
entered as no. 10931 in the collection of the Cleveland Museum of
Natural History.

I am greatly indebted to the authorities of the Cleveland Mu-
seum of Natural History for the privilege and pleasure of describing
this specimen. I also wish to acknowledge help from Mr. Thomas
Arkle, Jr. and the West Virginia Geological Survey for information
concerning the stratigraphy of the Greer area.

SYSTEMATIC PALEONTOLOGY

Family COLGSTEXDAE Romer, 1930
Genus GREERERPETON1 gen. nov.

GREERERPETON BURKEMORANI ’ sp. nov.

Figs. 1-7

Diagnosis for genus and species : A colosteid rhachitome, similar in
skull roof pattern and proportions to Colosteus, but with a lesser
development of the anterior portions of the lateral line groove sys-
tem; premaxillary “tusks” present, as in Erpetosaurus.

Holotype : Cleveland Museum of Natural History no. 10931.

Occurrence : Bickett Shale of the Bluefield Formation, Mauch
Chunk Group, Mississippian.

Locality: Greer, Monongalia County, West Virginia, on Deckers
Creek, about 6V2 miles southeast of Morgantown.

Repository: Cleveland Museum of Natural History.

1 The generic name is derived from the locality.

2 The specific name is formed (reasonably if unorthodoxly) by combining the
surnames of the two discoverers of the specimen.

ALFRED SHERWOOD ROMER

NO. 6

Fig. 1. Greererpeton burkemorani Romer, C. M. N. H. no. 10931. Photograph
of the main block containing the skull and remains of the anterior part of the
trunk, X %•

1969

LOWER CARBONIFEROUS LABYRINTHODONT

Fig. 2. Greererpeton burkemorani Romer, C. M. N. H. no. 10931. Key to ele-
ments present on the skull, as preserved: a, angular; d, dentary; j, frontal;
?in, internasal; j, jugal; l, lacrimal; m, maxilla; n, nasal; p, parietal; pf, post-
frontal; pm, premaxilla; po, postorbital; pp, postparietal; prf, prefrontal; qj,
quadratojugal; sa, surangular; sq, squamosal; st, supratemporal; t, tabular,
X %.

ALFRED SHERWOOD ROMER

NO. 6

DESCRIPTION

Cranial remains : The principal block (figs. 1 and 2) contains the
entire cranial structure, with the skull, seen from above, crushed
flat and somewhat disarticulated. Most of the left jaw ramus is seen
to the left of the skull (the articular region and partial surangular
were removed during preparation) ; the right ramus is mainly con-
cealed by the skull, but the articular region is seen behind the right
tabular. Except for the right quadratojugal and the anterior end
of the left jugal, nearly the entire series of dorsal dermal elements
is present. The crushing of the skull has caused considerable frac-
turing and disarticulation. The stoutness of the premaxillae, to-
gether with the presence of large tusks in this region, has caused
these elements to be partially overturned and separated from the
roofing bones posterior to them. The upward tilting of the sides of
the skull into the horizontal plane has separated the lacrimals of
both sides from the prefrontals, and on the left side has separated
the “cheek” elements — squamosal and quadratojugal — from the
table, and the left parietal, as well as the “table” elements posterior
to it have been pushed some distance to the rear, together with the
left “cheek”.

In figure 3 I have attempted to articulate the roofing series in
natural relations. Since, however, the palatal structure is almost
unknown, I have been unable to determine the true width of the
skull and consequently have been forced to restore the whole roof
in an unnatural horizontal plane (fig. 3) . As a result, the articu-
lation between elements is to some extent distorted; more important
is the fact that this type of restoration makes the skull appear con-
siderably broader than it actually was in life.

The length, as restored, from snout to the back edge of the
skull table, is 128 mm. The general proportions of the skull (allow-
ing for the artificiality of the apparent width) are those of a mod-
erately long and rather narrow structure. The orbits, apparently
typically circular in outline, lie well toward the front of the skull,
giving a relatively short face and a much elongated postorbital seg-
ment. The skull table is broad; on either side the cheeks slant back-
ward gently toward the jaw articulation. The area of junction of
cheek and table is disturbed on both sides. That the suturing was
not too tight between the two regions is demonstrated by the clean
break between cheek and table seen on the left. There was ob-
viously little or no development of an otic notch.

1969

LOWER CARBONIFEROUS LABYRINTHODONT

Fig. 3. Greer erpeton burkemorani Romer, C. M. N. H. no. 10931. The skull roof
restored; the elements are shown as if all were in a horizontal plane, and hence
the true width is exaggerated, X %.

The sculpturing is of a typical labyrinthodont type: essentially
circular depressions surrounded by ridges near centers of ossifica-
tion, gradually changing to a series of long ridges with intervening
valleys in elements which extend some distance from the ossification
centers. A considerable amount of the potential pattern of grooves
for lateral line canals is present. The cheek line is seen extending
backward and upward on the posterior part of the jugal and ante-
rior part of the squamosal, and then curving downward posteriorly
to disappear at the posterior edge of the cheek near the squamosal-
quadratojugal suture. Of the longitudinal canals originally present
on the skull table, there is to be found only a pair of short grooves
on the supratemporals. The supraorbital lyrae are distinctly devel-
oped on postorbitals, jugals and prefrontals. A transverse post-

ALFRED SHERWOOD ROMER

NO. 6

orbital canal, forked medially, is present on both postorbitals. There
is no trace of a suborbital groove although it may have been present
along the suture between the (disarticulated) maxillae and the ele-
ments dorsal to them. I have seen no interpretable traces of canals
such as are found on the snouts of various other labyrinthodonts.

The premaxillae are stout elements whose outer edges are con-
siderably thickened and curve strongly downward from the level of
the snout roof to the upper jaw margins. The anterior ends of the
maxillae are broadened and obviously thickened dor sally; they ex-
tend as slender tooth-bearing strips far back below lacrimal and
jugal. The state of preservation makes it impossible to give details
of the region of the external naris; apparently premaxilla, maxilla,
nasal and prefrontal enter into its borders. The lacrimal appears to
have been excluded. The nasals are short but broad. Between two
areas clearly belonging to the pair of nasals is a median strip of
bone which I was at first inclined to consider a broken-off fragment
of one of the nasals. However, the sculpture pattern of this area
shows no relation to that adjacent to it on either nasal, and I think
it likely that we have here an internasal, a median unpaired ele-
ment such as is found occasionally in other labyrinthodonts.

The frontals are broad anteriorly, but narrow posteriorly be-
tween the orbits. The parietals are well developed, laterally occu-
pying (with the postorbitals) the area in which intertemporals are
present in many early forms. Postparietals are large and elongate.

Of the circumorbital series, the lacrimals are large, essentially
triangular elements broadly exposed along the anterior margin of
the orbits and tapering anteriorly; they appear not to have reached
the external nares. The prefrontals are elongate, narrow posteriorly
but broadening anteriorly. They appear to have but barely entered
the orbital margins anterodorsally. The prefrontals extend much
farther forward toward the nares than is typical of labyrinthodonts
generally. About opposite the anterior tips of the lacrimals there is
a crack separating the regions definitely pertaining to the prefrontals
from an area running forward toward the nares. I am none too cer-
tain of the identification of the element or elements present here.
Possibly the maxilla may extend medially here; possibly part or all
of this area may constitute an external exposure of a septomaxilla
(not otherwise identifiable in the specimen) . It seems, however,
more probable that we are dealing with a still further extension of
the prefrontal.

1969

LOWER CARBONIFEROUS LABYRINTHODONT

The upper margins of the orbits are formed by the postfrontals,
narrow anteriorly but broadening posteriorly where they extend
some distance back of the orbits to reach the parietals. The post-
orbitals are elongate anteroposteriorly, in conformance with the
general elongation of the postorbital region of the skull; they taper
posteriorly to terminate between supratemporal and squamosal.
The jugal is the largest in area of the dermal roofing elements, cov-
ering most of the anteroventral area of the “cheek” and extending
forward broadly beneath the orbit to the lacrimal.

Of the lateral elements of the skull table, the presumably prim-
itive intertemporal is absent. The supratemporal is a large element,
its anterior end lying between parietal and postorbital; posteriorly
the supratemporal tapers laterally to a point close to the otic notch
region. In temnospondyl fashion the tabulars are small, bounded
anteriorly by the supratemporal and medially by the postparietal;
there is, of course, no tabular-parietal contact. The squamosal is
broad, and rather elongate; and the quadratojugal well developed.
Although the articulation of squamosal with the skull table is ob-
viously relatively loose, as noted earlier, there can be seen a flange
of the squamosal which ran medially underneath the supratemporal.

It would be an extremely difficult task to remove the skull
from the thick block of impure limestone in which it is embedded;
in consequence little can be said of the palate, of which a few
glimpses can be had through the orbits and broken areas of the
skull roof. Part of the transverse flange of the left pterygoid has
pushed up to the surface in front of the left squamosal, and portions
of the anterior rami of the pterygoids are visible through the two
orbits. In the left orbit is seen in dorsal view the somewhat thick-
ened margin of the bone bordering the interpterygoid vacuity. On
the right the bone has apparently been broken and displaced, so
that the palatal surface of the same area is seen, the margin toward
the interpterygoid vacuity bearing a band of small denticles.

As mentioned above, the right jaw ramus is concealed by the
skull except for the articular region. On the left the block exhibits
a considerable part of the ramus, including most of the dentary and
parts of angular and surangular. A posterior fragment of this jaw,
removed from the block during preparation, is shown in figure 4.

Much of the dentary is visible on the left jaw ramus. At mid-
length of the bone there are teeth of labyrinthine structure and
modest height, spaced at intervals of about 5 mm, with frequent

ALFRED SHERWOOD ROMER

NO. 6

Fig. 4. Greererpeton burkemorani Romer, C. M. N. H. no. 10931. Left articular
region and partial surangular in dorsal and lateral view, X 1

empty sockets in alternate positions. As expected, tooth size and
spacing decrease posteriorly. A large tooth, with a length of about
9 mm, is seen projecting downward from the posterior margin of
the left premaxilla. The right premaxilla has been broken off and
overturned, revealing the presence here of several large tusks,
grooved in labyrinthine fashion, and with estimated lengths of well
over a centimeter. Two are broken off and one or both of these
may have been parasymphysial teeth of the lower jaw. One how-
ever, is definitely attached to the premaxilla — a most unusual situ-
ation for a large tooth of this sort.

Poster anial skeleton : Much of the vertebral column is present, but
disarticulated; as noted above, a considerable series of vertebrae of
the anterior part of the trunk is represented by materials posterior
to the skull on the main block (fig. 5) ; a second block carries a
series presumably from the posterior part of the trunk (fig. 6) .
The vertebral structure is typically rhachitomous. On both blocks
there are numerous intercentra; a total of about 30 are clearly
visible. They are stout structures although in many cases broken
in two by crushing. They have a typical crescentic shape, as seen
in end view; broad below, with an anteroposterior dimension of 5
to 7 mm, they taper upward on either side to form nearly a semi-
circle. Although moderately thick ventrally, there nevertheless re-
mains a large cavity for the presumably persistent notochord. The
curvature of the intercentra shows the diameter of the column to
have been about 20 mm. The pleurocentra are less well preserved,
but can be seen to have been the typical rhachitome wedge shape.
A relatively few neural arches are visible. The spines are low, with

1969

LOWER CARBONIFEROUS LABYRINTHODONT

Fig. 5. Greererpeton burkemorani Romer, C. M. N. H. no. 10931. Postcranial
remains present on main block: cl, clavicles; ?cth, possible cleithrum; h, hu-
merus; icl, interclavicle; r, radius; sc, left scapulocoracoid, X %.

the greatest height observed about 15 mm above the level of the
zygapophyses, but are stoutly built, with transverse widths of sev-
eral millimeters and anteroposterior dimensions of 5 to 7 mm. The
zygapophyses diverge moderately on either side of the arch base,
the width across them being 7 to 9 mm. On either side, the arch
bases send broad processes down and outward ending below in
horizontal surfaces for rib articulation. Scattered ribs are present;
the longest seen has a length of about 40 mm. The slender shafts
are nearly straight; the heads are expanded in triangular fashion.

I have not figured the abdominal scales, but a plentiful supply
is present below each of the two series of vertebrae. They have the
typical labyrinthodont oat-grain shape, and are generally about 10
mm in length.

As noted earlier, there was found with the specimen most of an
appropriate series of limb and girdle bones (fig. 7). Those present
appear to pertain to a single individual. All are of small dimensions
for an animal with a skull of this size, indicating feeble limbs and
an obviously aquatic habitat. As an indication of size, I may note
that the femoral length is only about 25 percent of skull length,
whereas in Eryops this figure averages about 35 percent, and in

ALFRED SHERWOOD ROMER

NO. 6

Fig. 6. Greererpeton burkemorani Romer, C. M. N. H. no. 10931. Disarticulated
elements — neural arches, intercentra, pleurocentra, ribs — pertaining to pos-
terior part of trunk, X V2.

1969

LOWER CARBONIFEROUS LABYRINTHODONT

Cacops about 55 percent. On the other hand, members of the Tri-
merorhachis group, which are known from the Pennsylvanian and
early Permian, are, like the present form, small limbed. Associated
materials are none too common, but in this group femoral length
appears to be rather less than 25 percent of skull length. At first
sight, the limb and girdle structures in the present specimen appear
to be generally comparable to those of trimerorhachoids.

The dorsal ends of both clavicles are visible behind the skull
in the main block, projecting upward in typical tapering fashion.
Their vertical position is due to their possession of expanded lower
portions lying in the plane of the block surface; the blade of the
right clavicle is exposed, that of the left concealed. I have not iden-
tified cleithra, although a long slender structure lying behind the
right side of the skull may be such an element. A portion of the
presumably expanded interclavicle is visible near the left clavicle.
A single scapulocoracoid of the left side, seen from the inner sur-

Fig. 7. Greererpeton burkemorani Romer, C. M. N. H. no. 10931. Limb and
girdle elements. A, left scapulocoracoid, inner surface. B. right radius, flexor
aspect. C, right ilium; ischium (seen only from inner surface) in outline.
D, right femur, dorsal aspect. E, left tibia, extensor aspect. F, left fibula, ex-
tensor aspect, all X 1.

ALFRED SHERWOOD ROMER

NO. 6

face, is preserved adjacent to the left clavicle. As in trimerorha-
choids, most of the coracoid plate and scapular blade is unossified;
the scapulocoracoid as preserved includes only the general region
of the supraglenoid buttress. The preservation is poor, and I am
unable to determine whether or not the typical supraglenoid fora-
men was present. There are remains of both humeri, but little can
be made out regarding their structure. Both radii are present; they
are short but broad elements about 19 mm long and in general re-
semble the corresponding elements of Eryops. The lateral margin
is thin. Posteriorly, toward the median side, there is a sharp lon-
gitudinal ridge. The upper end is very broad, with a width of 11
mm; the distal width is about 8 mm. Ulnae are not well preserved,
and I have not identified elements of the manus.

Both ilia are present: the right well preserved, the left im-
perfectly. The shape is that seen in many later rhachitomes, with
a tall vertical blade, and no trace of a posterior process (not even
the slight spur seen in Eryops) . There is a typical supra-acetabular
buttress. The right ilium has a total height of 30 mm. The two
ischia are present. Both are seen from the concave featureless inner
surface; their greatest lengths, from acetabulum to distal end, are
24 mm and 20 mm, as preserved. There is no trace of a pubis; very
probably, as in various other amphibians, this bone failed to ossify.

Of hind limb elements, the femur, tibia and fibula of the left
side were found associated with one another and with the left ilium;
the right femur and tibia were found isolated. The left femur has
a length of 40 mm; the right, apparently incomplete distally, is 35
mm long. As in temnospondyls generally the head is expanded
anteroposteriorly and thin dorsoventrally. From the articular sur-
face of the head, an unfinished surface runs downward and distally
along the outturned margin of the shaft for some distance, to be
continued by a pronounced ridge which extends down the antero-
ventral margin of the shaft to terminate at the ventral distal mid-
point. There is a typically gentle groove distally between the two
slightly convex condyles. There are, as usual, two distal condylar
facets for the tibia, the distal face of the bone giving the appearance
of a thickened V, with an apex ventrally at the termination of the
ventral ridge.

The two tibiae measure 22 and 21 mm in length. As with the
radius, we are dealing with a short but stout element. The head is
broadened to about 11 mm for apposition to the two femoral con-

1989

LOWER CARBONIFEROUS LABYRINTHODONT

dyles; the medial portion of the head is much the thicker of the two
areas, with a depth of 6 mm. Below the medial head there is a slight
development of a cnemial rugosity and crest. On the flexor aspect
below the head there is a low, short but rugose longitudinal crest
(a somewhat similar structure is present in Eryops) . Below the
head, the bone contracts, with a pronounced lateral convexity, to
a short shaft and then, twisting somewhat laterally, expands to a
distal articular surface clearly divided into separate areas for inter-
medium and fibulare. The distal end of the bone has a width of
about 9 mm; the articular face is broadest on the facet for the
fibulare, measuring here 4 mm. Except for lesser development of
a cnemial crest, the whole structure of the bone resembles that of
Eryops.

The left fibula is 18 mm long as preserved. Its proportions are
those of early tetrapods in general. The upper end is about 3 mm
wide; the broadened lower end has a width of about 7 mm. There
are no identifiable remains of the pes.

DISCUSSION

That the amphibian here described was primarily a water-
dwelling form rather than one leading a truly amphibious existence,
seems clear from the small limbs, which were obviously incapable
of supporting the body and heavy head on any extensive forays onto
land, and from the developed series of grooves indicating a func-
tioning lateral line system. A primarily aquatic existence in early
Carboniferous amphibians generally is to be expected (the rather
well developed limbs of ichthyostegids are a puzzling exception) in
the seeming absence of a terrestrial food supply of animal nature
before the radiation of insects in the Pennsylvanian.

Of greater interest, however, is the determination of the sys-
tematic position of this Mississippian form and its position in laby-
rinthodont evolution. Previously the only known Lower Carbonif-
erous temnospondyls had been the obviously aberrant loxommids.
We have here a form with orbits of normal shape, rather than the
peculiarly elongate openings of the loxommids, and I fondly hoped
at first, that the present form might prove to be a “generalized”
rhachitome, close to the ancestor of the abundant rhachitomous
amphibians of the Pennsylvanian and early Permian.

ALFRED SHERWOOD ROMER

NO. 6

This, sadly, proves not to be the ease. Typical later rhachitomes
(except for short-skulled types) have orbits posterior to the mid-
point of the skull length; here the eyes are well forward, with a
relatively short face and a long postorbital region. A shortness of
face combined with relatively small limbs suggests the trimero-
rhachoid rhachitomes — a group which, although off the main line
toward typical advanced rhachitomes and stereospondyls, never-
theless must have branched off at an early and primitive stage, since
they preserved the primitive movable basal articulation of brain-
case and palate and retained the intertemporal element. When,
however, the skull was reassembled, this possibility disappeared,
for there is no intertemporal.

Further thought and search led to the true but unexpected de-
termination of the specimen’s position. We have here a predecessor
of the Colosteidae, a small group of Pennsylvanian temnospondyls
of problematical relationships, whose members (Erpetosaurus and
Colosteus) were described by Romer (1930) and by Steen (1931) .

In every known regard the present form shows agreement with
the two colosteid genera. The general skull proportions with a short
face and a long postorbital region, are identical. Here, as in Colos-
teus., but in contrast to typical labyrinthodonts, there is no develop-
ment of a marked otic notch. The pattern of the lateral line canals
in our specimen is similar to that seen in Colosteus , except that the
anterior portions of the system are not as well developed (as far as
can be determined) as in the Pennsylvanian forms. Here, as in the
colosteids, the lacrimal enters broadly into the orbital margin and
tapers anteriorly, whereas in typical temnospondyls this bone enters
the orbit only over a short stretch and is often completely excluded.
The remarkable forward extension of the prefrontal found here is
also characteristic of colosteids, and contrasts with the situation
found in most labyrinthodonts. The pattern here of the long post-
orbital segment of the skull (with the intertemporal absent) agrees
well with that of the colosteids. To some degree a similar pattern
is present in other temnospondyls in which the posterior part of the
skull is lengthened, but the high development of postparietals and
supratemporals, in contrast with much reduced tabulars, is notable.

It is obvious that in this new form the palate, although little of
it is visible, was widely open, as in colosteids, and one may expect
that here, as in that family, the movable basal articulation of brain-
case and palate had been preserved. The most definitive point of

1969

LOWER CARBONIFEROUS LABYRINTHODONT

resemblance has to do with the presence, both in our form and in
the colosteids, of large tusks near the tip of the snout — a situation
unknown elsewhere among temnospondyls. The distribution of the
large anterior teeth in Colosteus is not known in detail, but in Erpe-
tosaurus, as figured by Steen (1931, figs. 4, 5) there are large tusks
on the premaxilla, and in the present specimen at least one such
premaxillary tusk is present.

Our new form, then, is definitely attributable to the Colosteidae.
It is of interest in showing that this family, formerly known from
the Pennsylvanian alone, is of great antiquity. But such attribution
does not place it with any degree of assurance in the general picture
of temnospondyl radiation, for the colosteids are themselves a prob-
lematical group. When the ichthyostegids were first described by
Save-Soderbergh (1932) I jumped to the conclusion, based on the
general skull pattern, that Otocratia of the Mississippian and the
colosteids of the Pennsylvanian were ichthyostegid relatives
(Romer, 1947). The assumption that Otocratia is indeed related
seems to be confirmed by the finding of ichthyostegids with com-
parable otic peculiarities (Jarvik, 1952), but that the colosteids are
also ichthyostegids is very doubtful. Apart from general skull pro-
portions and the loss of the intertemporal, there is at present no
reason to claim relationships, and it seems more probable that the
colosteids were, rather, an early offshoot of the Rhachitomi, paral-
leling the trimerorhachoids in skull proportions, small limbs, reten-
tion of a movable basal articulation and opening up of large inter-
pterygoid vacuities, but differing in loss of the intertemporal.

Our present specimen does not, thus, furnish us with a potential
ancestor of the typical rhachitomes of later times. But the presence
in the Mississippian of such an unusual type, together with the
equally aberrant loxommids, indicates that the radiation of the
Rhachitomi was well under way in early Carboniferous times.
Should future exploration, hopefully, result in additional finds of
Mississippian labyrinthodonts, we may expect that there will be
found in the faunal assemblages not only representatives of pro-
gressive stages in anthracosaurian evolution but also specimens
demonstrating an active radiation of early rhachitomes.

ALFRED SHERWOOD ROMER

NO. 6

REFERENCES CITED

Baird, D., 1957, Rhachitomous vertebrae in the loxommid amphibian Megalo-
cephalus : Geol. Soc. Amer. Bull., v. 68, p. 1698.

1962, A rhachitomous amphibian, Spathicephalus, from the

Mississippian of Nova Scotia: Breviora, no. 157, p. 1-9.

Coryell, H. N. and I. G. Sohn, 1938, Ostracoda from the Mauch Chunk (Missis-
sippian) of West Virginia: Jour. Paleontology, v. 12, p. 596-603.

Flowers, R. R., 1956, A subsurface study of the Greenbrier Limestone in West
Virginia: West Virginia Geol. Surv. Rept. Inv. no. 15, p. 1-17.

Jarvik, E., 1952, On the fish-like tail in the ichthyostegid stegocephalians:
Meddelelser om Gr^nland, v. 114, p. 1-90.

McCue, J. B., J. B. Lucke, and H. P. Woodward, 1939, Limestones of West Vir-
ginia: West Virginia Geol. Surv., v. 12, p. 1-560.

Moran, W. E., 1952, Location and stratigraphy of known occurrences of fossil
tetrapods in the Upper Pennsylvanian and Permian of Pennsylvania, West
Virginia, and Ohio: Carnegie Mus. Ann., v. 33, p. 1-44.

Panchen, A. L. and A. D. Walker, 1961, British Coal Measure labyrinthodont
localities: Ann. Mag. Nat. Hist., ser. 13, v. 3, p. 321-332.

Romer, A. S., 1930, The Pennsylvanian tetrapods of Linton, Ohio: Amer. Mus.
Nat. Hist. Bull., v. 59, p. 77-147.

1941, Earliest land vertebrates of this continent: Science,

v. 94, p. 279.

1947, Review of the Labyrinthodontia: Mus. Comp. Zool. Bull.,

v. 99, p. 1-368.

1952, Late Pennsylvanian and Early Permian vertebrates of

the Pittsburgh- West Virginia region: Carnegie Mus. Ann., v. 33, p. 47-112.

1958, An embolomere jaw from the Mid-Carboniferous of

Nova Scotia: Breviora, no. 87, p. 1-8.

1964, The skeleton of the Lower Carboniferous labyrinthodont

Pholidogaster pisciformis: Mus. Comp. Zool. Bull., v. 131, p. 129-159.

Save-Soderbergh, G., 1932, Preliminary note on Devonian stegocephalians from
East Greenland: Meddelelser om Gr^nland, v. 94, p. 1-107.

Steen, M. C., 1931, The British Museum collection of Amphibia from the Middle
Coal Measures of Linton, Ohio: Zool. Soc. London Proc., 1930 [1931], p.
849-891.

Tilton, J. L., 1928, Geology from Morgantown to Cascade, W. Va., along state
route number 7: West Virginia Univ. Sci. Assoc. Bull., p. 65-86.

Watson, D. M. S., 1919, The structure, evolution and origin of the Amphibia —
the “orders” Rhachitomi and Stereospondyli: Roy. Soc. [London] Philos.
Trans., ser. B, v. 209, p. 1-73.

1926, The evolution and origin of the Amphibia: Roy. Soc.

[London] Philos. Trans., ser. B, v. 214, p. 189-257.

1929, The Carboniferous Amphibia of Scotland: Palaeontologia

Hungarica, v. 1, p. 219-252.

Weller, J. M. (Chairman) and others, 1948, Correlation of the Mississippian
formations of North America: Geol. Soc. Amer. Bull., v. 59, p. 91-196.

Wells, D., 1950, Lower Middle Mississippian of southeastern West Virginia:
Am. Assoc. Petroleum Geologists Bull., v. 34, p. 882-922.

MANUSCRIPT RECEIVED JAN. 17, 1969

n* cv

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO MAY 2, 1969 NUMBER 7

A NEW AMIOID FISH FROM THE
UPPER CRETACEOUS OF KANSAS

DAVID H. DUNKLE

.^Ttmso
JUL s »s

ABSTRACT

The partial skeleton of a small fish from the Niobrara Forma-
tion (Upper Cretaceous) of Kansas possesses a hemiheterocercal
caudal fin, scales of rounded amioid type and other characteristics
denoting a new holostean described as Paraliodesmus guadagnii
gen. et sp. nov.

\n —

INTRODUCTION

There is a meagerness of published information about the
diminutive members of the paleoicthyological faunas of the classic
Upper Cretaceous formations of North America, coupled with a lack
of demonstrable systematic diversity among these forms. This is
quite evident when comparison is made with records from equiv-
alent strata in other parts of the world.

Prompted by the mode of occurrence of Leptecodon (Stewart,
1900) and Kansius (Hussakof, 1929) which are preserved in the
interiors of the shells of Inoceramus grandis (Conrad) , this writer
and G. Donald Guadagni made routine examination of eroding spec-
imens of this pelecypod while collecting in the Niobrara Formation
of western Kansas during the summer of 1954. The result of the
search was recovery of numerous additional specimens of the small
berycoid Kansius as well as one unfamiliar skeleton of modest size.
Although incomplete, the latter specimen possesses an abbreviated
heterocercal caudal appendage, scales of rounded amioid type and
other characteristics here interpreted as denoting a previously un-
known fish of holostean affinity. It is suggested that future search
of this sort might well add appreciably to knowledge of a little-
known faunal segment of the Kansas “Chalk”.

DAVID H. DUNKLE

NO. 7

I am indebted to the Smithsonian Institution for the privilege
of making this study. Dr. Colin Patterson of the British Museum
[Natural History], London and Dr. Richard Lund of the Carnegie
Museum, Pittsburgh, were gracious in making pertinent compara-
tive materials available for examination. The illustrations represent
the work of two members of the staff of the United States National
Museum: Lawrence B. Isham prepared the drawings and Jack Scott
made the photographs.

SYSTEMATIC PALEONTOLOGY

Order AMIIFORMES Hay, 1929
Suborder AMIOIDEI Schlosser, 1934
Family 7CATURIDAE Koken, 1911
Genus PARALIODESMUS gen. nov.

PARALIODESMUS GUADAGNII* 1 sp. nov.

PI. 1, figs. 1-3

Diagnosis for genus and species: An actinopterygian fish of holo-
stean affinity which differs from any of the more comparable mem-
bers of the families Caturidae and Amiidae in the following combi-
nation of characteristics: Dermal bones of the skull exceptionally
thick and with a noticeably coarse rugose ornamentation; notochord
persistent; dorsal fin originating remotely behind the pelvics and
largely opposed to a long-based anal; caudal fin equilobate, deeply
cleft and of apparent caturid internal structure; fins without evi-
dence of fulcra except for spinelets of accumulative length in ad-
vance of both upper and lower lobes of the caudal fin; scales minute
and of rounded amioid type.

1 Named in honor of Mr. G. Donald Guadagni, codiscoverer of the holotype.

EXPLANATION OF PLATE 1

Figs. 1-3. Paraliodesmus guadagnii Dunkle, holotype, U. S. N. M. 21083.

1. Habit sketch of specimen, X %; 2. Sketch of rounded scale, X 18 (approx.);
3. Photograph of specimen, X %.

KIRTLANDIA NO. 7

PLATE 1

DAVID H. DUNKLE

NO. 7

Holotype: United States National Museum no. 21083, a fish pre-
served in part as an impression.

Horizon and locality: Smoky Hill Member of the Niobrara Forma-
tion (Upper Cretaceous) in sec. 36, T. 15 S., R. 31 W., Gove County,
Kansas.

DESCRIPTION

Post-depositional compression of the Inoceramus valves with
consequent comminution of the enclosed skeleton and subsequent
exposure of the anterior half of the body has largely obliterated fine
structural detail. Clearly observable, however, is the outline of an
entire fish shown by (1) an impression of the left anterior side of
the body on the internal surface of the lower shell and (2) the right
aspect of preserved posterior scales, fins and axial structure which
were protected prior to discovery by portions of the upper shell
that had remained in place during weathering. An elongate fusi-
form fish is indicated with a maximum body depth of 35 mm, a
stout caudal pedicle of 15 mm depth, and head and opercular appa-
ratus occupying 40 mm of a total overall length approaching 170
mm.

The head is obliquely crushed and is seen in ventrolateral as-
pect. Only remnants of the skull bones remain but these are of re-
markable thickness, with coarse and radiating rugose ornamentation.
Whether or not an external covering of enamel is retained on the
cranial elements and scales has not been definitely ascertained, but
its presence is suspected. The suspensorium would appear to have
been vertical and the mouth terminal, with wide gape. The pre-
maxillaries are in contact with each other mesially and are suc-
ceeded dorsally in the snout by a reduced rostral bone which is
broader in front than long. Preserved fragments of the right maxil-
lary show an elongate element, somewhat deepened behind, with
a straight oral border. The separate bones of the mandibles cannot
be discerned. However, the dentary portion is relatively low ante-
riorly, progressively deepening to a coronary eminence posteriorly.
These visceral elements all bear a single labial row of robust teeth.
Those of the upper jaw are particularly stout and strongly recurved.
Dentary teeth are, in comparison, high, thin and styliform. Part of
a presumed splenial exposed beneath one of the lower jaws displays
a random arrangement of minute denticles.

1969

A NEW CRETACEOUS HOLOSTEAN

The number and arrangement of the opercular bones are char-
acteristic of the amioid fishes. The operculum is slightly the larger
element of the series with rounded posterior margin and antero-
ventral notch receiving an ascending process of the suboperculum.
The branchiostegal rays, in typical fanned pattern, are narrow and
numerous.

The notochord is persistent to its tip in the epichordal lobe of
the equilobate and noticeably notched caudal fin. The downcrush-
ing of scales over the expanded proximal processes of opposing
neural and haemal arches vaguely suggests segmentation of the
notochordal axis posteriorly, but no evidence of central ossifications
can be detected. If indeed such are present under the scale cover,
they must be very weak. Neural and haemal structures are acutely
recumbent upon the notochord throughout the caudal region. The
development of the posterior haemal spines is poorly visible. While
somewhat expanded, none that can be seen appears larger than an-
other. Nevertheless there is a definite overlap of the clustered and
forked proximal ends of the principal rays of the upper lobe of the
caudal fin across several of the posterior uptilted hypural bones as
in Caturus (Lund, 1967) .

The dorsal fin appears to have been about as long as high and
originates well behind the midlength of the specimen, posterior to
the plane of the pelvic fins and only slightly anterior to the origin
of the lower, longer-based anal.

The position of the pectoral fins has been distorted by the
oblique crushing of the head. It seems probable, however, that they
were somewhat elevated on the sides of the body. They are indi-
cated to have been larger than the pelvics. The separation of the
paired fins is about twice the distance between the pelvics and the
anal fin.

The forward edges of the upper and lower lobes of the caudal
fin display a number of undivided rays of posteriorly accumulative
length. These may also be present on the other fins, but evidence
of fulcral scales of more typical chondrostean or semionotoid form
is absent.

Scales of this fish are rounded, narrower apically than basally,
and only about 2 mm in length. The center of growth and exposed
portion of the scale, as marked externally by a few tubercles and
coarse, radiate rugae, is far apical and very short. The basal cov-
ered portion of each is of much greater area and displays fine, more

DAVID H. DUNKLE

NO. 7

or less parallel striae which radiate to the upper, lower and anterior
margins of the scales.

DISCUSSION

Known Cretaceous fishes of holostean stock are relatively few
in number. Among these, Paraliodesmus shows greater similarity
only to representatives of the families Caturidae and Amiidae. It
is distinguished from the essentially contemporaneous Lophiostomus
and N eorhombolepis (Woodward, 1895) , and Enneles (Santos,
1960) by scale type and size. Closer comparability of the characters
elucidated are in fact seen with the older and better known Jurassic
genera Caturus and Liodesmus. It differs most obviously from the
latter two in fin position and the thickness and coarse ornamentation
of the skull bones. Caudal structure suggests tentative reference
of Paraliodesmus to the family Caturidae rather than to Amiidae.
Such assignment necessitates assumption that the absence of fulcra
occurred independently in a number of phyletic lines, as did the
rounding and reduction of the scales (Schultze, 1966) .

REFERENCES CITED

Hussakof, L., 1929, A new teleostean fish from the Niobrara of Kansas: Am.
Mus. Novitates, no. 357, p. 1-4, 2 figs.

Lund, R., 1967, An analysis of the propulsive mechanisms of fishes, with refer-
ence to some fossil actinopterygians: Carnegie Mus. Ann., v. 39, art. 15,
p. 195-218, 12 figs.

Santos, R. da S., 1960, A posigao sistematica de Enneles audax Jordan e Bran-
ner de Chapada do Araripe, Brasil: Brazil, Div. Geol. e Miner., Mon. 17,
p. 1-25, 5 pis., 9 figs.

Schultze, H.-P., 1966, Morphologische und histologische Untersuchungen an
Schuppen mesozoischer Actinopterygier (Ubergang von Ganoid- zu Rund-
schuppen): Neues. Jahrb. Geologie u. Palaontologie Abh., v. 126, art. 3,
p. 232-314, pis. 49-53, 61 figs.

Stewart, A., 1900, Teleosts of the Upper Cretaceous: Kansas Univ. Geol. Sur-
vey [Rept.], v. 6, pt. 2, p. 257-403, pis. 33-73, 6 figs.

Woodward, A. S., 1895, Catalogue of the Fossil Fishes in the British Museum
(Natural History) : pt. 3, xlii + 544 p., 18 pis., 45 figs., London.

MANUSCRIPT RECEIVED APRIL 17, 1969

Kb l

&& MW

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO FEBRUARY 5, 1970 NUMBER 8

TREMATOPS STONEI SP. NOV. (TEMNOSPONDYLI:
AMPHIBIA) FROM THE WASHINGTON FORMATION,
DUNKARD GROUP, OHIO

EVERETT C. OLSON

University of California at Los Angeles

ABSTRACT

Trematops stonei, a new species of this genus of temnospon-
dylous, labyrinthodont Permian amphibian, is described and fig-
ured. The specimen upon which the description is based was
found in the Creston Shale, Dunkard Group, Lower Permian at
Marietta, Ohio. This genus, along with Melanothyris (previously
known from the Dunkard), and Dimetrodon and Diadectes, re-
cently obtained from a channel-fill deposit at Belpre, Ohio, indi-
cate the presence of a terrestrial component in the Early Permian
faunas of this region. This part of the faunal complex is rarely
represented in the usual pond -lake deposits of the Dunkard Group.

INTRODUCTION

The vertebrate fossils of the Dunkard Group have become fair-
ly well known through the studies of Moran (1952), Romer (1952)
and Beerbower (1963). Extensive collections have come from Ohio,
Pennsylvania and West Virginia. These include many as yet un-
described specimens so that faunal considerations, based as they
must be upon published information, are necessarily somewhat in-
complete. Among the collections of the Cleveland Museum of Nat-
ural History is the skull described in this paper. It is of sufficient
interest that its description cannot await studies of other collected
but undescribed materials.

Over the areas in which the Dunkard occurs, a more or less
continuous sequence of fossiliferous beds includes deposits formed
in the Late Pennsylvanian (Stephanian) and Early Permian (Autu-
nian) . Finds of vertebrate remains in these beds are widely scat-
tered geographically and the fossils are for the most part fragmen-

EVERETT C. OLSON

No. 8

tary. Outcrops are limited by the heavy cover of vegetation over
the area, and remains have in large part come from roadcuts or per-
sistent exposures created by resistant limestone. Nevertheless, rep-
resentatives of nearly 30 genera and species have come from the
Dunkard portion of this sequence, comprising several groups of
fishes, amphibians and reptiles. These, for the most part, are, at
a generic level, similar to the Late Pennsylvanian and Early Per-
mian vertebrates found in the better known areas of the Midcon-
tinent region, including northcentral Texas, Oklahoma, Kansas,
New Mexico, Colorado and Utah. In addition, there are striking
similarities to the Late Pennsylvanian vertebrates from the Oak-
wood locality of Illinois (see Olson, 1946) . A few genera, the best
known of which are Diploceraspis and Megamolgophis, are absent
from Midcontinent collections.

Both the Pennsylvanian deposits (Conemaugh and Mononga-
hela) and the Permian Dunkard are of the general “coal measures”
type, including sandstones, shales, clays and freshwater limestones,
along with coal. The gray dolomite and shale series of the Welling-
ton Formation of northcentral Oklahoma, except for the absence of
coal, approaches this depositional pattern most closely of Midcon-
tinent vertebrate-bearing formations. The predominant sediments
in both areas indicate deposition in shallow waters along the shores
of large freshwater lakes. In both areas occur sandstones and con-
glomerates deposited by current action. From such deposits, more
commonly in Oklahoma than in the eastern United States, have
come some fragmentary remains of kinds of animals not usually
encountered in the typical lake beds.

It is from such a deposit that the specimen that occasioned
this paper came. In February, 1969, Reed W. Irwin, a member of
a Marietta College field party investigating local Dunkard stratig-
raphy under the direction of Dr. Dwayne D. Stone, discovered the
specimen. It is part of a growing collection of Dunkard vertebrates
in the Cleveland Museum of Natural History. The specimen came
from the upper, green to gray beds of the Creston Shale (red beds)
of the Washington Formation, Dunkard Group, of Ohio. David
Dunkle turned the specimen over to me for study. It had been ex-
cellently prepared by Peter R. Hoover. The illustrations were pre-
pared by Eleanor Daly. I here express my appreciation to each of
these persons for their contributions to this study. The work was
supported in part by National Science Foundation Grant GB 13910.

1970

TREMATOPS FROM THE DUNKARD

SYSTEMATIC PALEONTOLOGY

Class AMPHIBIA
Subclass LABYRINTHODONTIA
Order TEMNOSPONDYLI
Family TREMATOPSIDAE Williston, 1910
Genus TREMATOPS Williston, 1909
TREMATOPS STONEP sp. nov.

Figs. 1, 2

Diagnosis: A moderately large species of Trematops with length of
skull of holotype (only known specimen) about 140 mm, as meas-
ured along the dorsal midline, slightly less than for known adult
specimens of T. milleri Williston. Marginal maxillary dentition
comprising relatively few, large, strongly labyrinthine teeth. Tusk
and pit on palatine bone strongly developed.

Holotype: Cleveland Museum of Natural History 10969, a partial
skull.

Horizon: Creston Shale (red beds) of the Washington Formation,
Dunkard Group, Lower Permian, immediately below the Upper
Marietta Sandstone.

Locality: Roadcut on Ohio Route 7, at Marietta, Washington

County, Ohio, 0.7 mile southwest of U.S. Route 50A.

Repository: Cleveland Museum of Natural History, Cleveland,
Ohio.

Description: The principal features of the skull are as shown in
figures 1 and 2. The specimen is somewhat distorted and the mar-
ginal portions, except forward from the midorbit on the left side,
much of the palate, and part of the snout are missing. A typically
trematopsid, large elongated narial fenestra is preserved on the left
side. The braincase, basicranium, parasphenoid, sphenethmoid and
interorbital-internasal septum are well preserved and show con-
siderable detail. All of these structures conform very closely in
form and size to those present in T. milleri.

1 The species name is given in recognition of Professor Dwayne D. Stone of
Marietta College, whose interest and inquiry about the specimen ultimately
resulted, with his cooperation, in its acquisition by the Department of Paleon-
tology of the Cleveland Museum of Natural History.

EVERETT C. OLSON

No. 8

Fig. 1. The skull of Trematops stonei sp. nov., X V2. A. dorsal view; B. ventral
view. Abbreviations: BO, basioccipital; BPT, basipterygoid process; F, fossa
on parietal platform; FR, frontal; IONS, interorbital-internasal septum; NA,
nasal bone; NR, naris; O, orbit; PA, parietal; PAL, palatine; PAS, para-
sphenoid; PF, postfrontal; PT, pterygoid; ST, supratemporal; STH, spheneth-
moid; T, tusk and pit on palatine.

The dorsal platform and lateral dermal surface of the skull
carry the usual reticulate, sculptured pattern found among the
labyrinthodonts. Although the bone is riddled by small, mineralized
fractures, the sutural pattern has remained moderately clear and
the reconstruction in figure 1 is reliable. On the left side of the
parietal shelf, a small, anomalous, elliptical fossa passes through the
bone (F in figure 1A) . It has well-formed, smooth finished edges.
It will be noted in the figure that the sutural pattern in the vicinity
of this opening is somewhat altered, showing that the fossa was
formed early in ontogeny and that it interrupted normal bone
growth. Whether it was a genetic defect or was induced by an early
mechanical interruption of normal growth cannot be determined.

Assignment : Reference of this specimen to the family Trematop-
sidae poses no difficulties for it is similar in all gross features to the

1970

TREMATOPS FROM THE DUNKARD

Fig. 2. Trematops stonei sp. nov., X V2. A. occipital view; B. braincase in
lateral view, lateral dermal bones removed. Abbreviations: BPT, basipterygoid
process; EPT, epipterygoid; FM, foramen magnum; PAP, paraoccipital process;
PT, pterygoid bone.

best known members of this family, Acheloma and Trematops. The
large, elongated external nares, the pattern of the dorsal and lateral
dermal bones, the strongly deflected basipterygoid processes, which
are fused to the pterygoids, the form of the braincase, including
the high, slender epipterygoid, and the narrow cultriform process
of the parasphenoid, which fails to meet the vomer anteriorly, are
the primary bases for this relationship (see Olson, 1941) .

Recently Vaughn (1969) considered the relationships of the
families Trematopsidae and Dissorophidae. On the basis of his new
genus Ecolsonia, which incorporates some features of dissorophids
although it is clearly trematopsid, he provided strong support for
the concept of close relationship between the two families. DeMar
(1966) described Longiscitula, an animal which has an elongated,
trematopsid-like narial opening, but is basically dissorophid in
many features. This, like Vaughn’s genus, indicates that the fam-
ilies are closely related but that they had undergone considerable
divergence from an unknown stock which must have existed well
prior to the Permian.

The only problem of generic assignment of T. stonei has been
whether it belongs to the genus Acheloma or to Trematops, which,
as discussed in an earlier review (Olson, 1941) , are very similar. As

EVERETT C. OLSON

No. 8

far as the structure of the skulls is concerned, the differences are
such that they could be judged to be either of specific or of generic
value. In instances where the structure is known, the frontal bone
is shorter in relationship to the parietal in Trematops than in Ache-
loma, the tabulars are relatively larger, the nasals relatively longer
and the squamosals relatively smaller. Postcranial differences are
somewhat more definitive. Most important is the complete ossifica-
tion of the hypocentra above the notochordal opening in Trematops
and the incomplete ossification in Acheloma. In addition the cora-
coids are better ossified in Trematops and the scapular blade is
erect in Trematops and reflected posteriorly in Acheloma (Olson,
1941) . The significant postcranial differences, of course, are not ap-
plicable at present to placement of the new species.

On the basis of the relatively short frontal, the moderately
large tabular and the relatively strong ossification of the braincase,
the Dunkard specimen is referred to the genus Trematops rather
than to Acheloma. It is considered to be distinct from T. milleri,
which it most closely resembles in size and proportions, primarily
on the basis of the marginal and palatal teeth. Although the full
tooth row is not present in the Dunkard specimen, it is probable
that the total number of marginal teeth did not exceed 20. The
teeth preserved in T. stonei are robust except for the most posterior
three, which show rapid reduction in size. This contrasts somewhat
with the teeth of T. milleri, which are comparably large only at the
level of the anterior part of the orbits and the posterior part of the
naris. In addition, the palatal tusk on the palatine bone of T. stonei
is very strong compared with that of other species of Trematops.
These are minor morphological differences and the few known spec-
imens do not indicate what role variation may have played. They
do, however, offer a morphological basis for separating the one
known specimen of T. stonei from those of T. milleri, a separation
that is suggested as well by the significant time interval between
the two, the one coming from the very beginning of the Permian
and the others from Clear Fork beds (Arroyo Formation) in the
middle Leonardian.

Measurements in millimeters of the skull dimensions of the
holotype and only known specimen of Trematops stonei sp. nov.
C.M.N.H. 10969 are as follows:

1970

TREMATOPS FROM THE DUNKARD

Skull length, along dorsal midline 140

Parietal length, along midline 33 (left) 37 (right)

Postparietal length, along midline 17 (left) 13 (right)

Frontal length, along midline 38

Skull width, least width between otic notch margins

on skull platform 86

Narial length, maximum 37

OCCURRENCE AND SIGNIFICANCE

The specimen of T. stonei sp. nov. was found near the top of
the “red shale” interval, the Creston Shale, between the lower and
the upper Marietta sandstones of the Washington Formation. Over
the years the status of the Washington as a formation or a more
comprehensive stratigraphic unit has been debated (see Hennen,
1911, McCue, et al. 1948, Hickock and Moyer, 1940, Moran, 1952,
and Nace and Bieber, 1958). For the purposes of this study the
conceptual differences involved in these discussions are unimportant
and for simplicity the Washington is considered to be a lithologi-
cally complex formation. The thickness of the formation ranges
from about 250 to 400 feet. Included, along with the sandstones
mentioned above, are five coals and four limestones.

The Creston Shale ranges between 35 and 60 feet in thickness
and lies between the Middle Washington Limestone below and the
Washington “A” Coal above. The latter is a reworked coal in the
area where the specimen was recovered. Overlying it is the Upper
Marietta Sandstone which has channeled to varying degrees into
the underlying beds. This sequence represents a variety of environ-
ments of deposition among which some, such as the Creston Shale,
give indications of being in part terrestrial. The only other verte-
brate remains from the Creston are a fragment of a spine of Edapho-
saurus found at Marietta, Ohio, by Stauffer (Stauffer, 1916, Stauf-
fer and Schroyer, 1920, Moran, 1952) and coprolites reported by
field parties from Ohio University. All specimens have come from
near the top of the beds.

The vertebrates add little to an understanding of the age of
the deposits. The Washington beds form the lower part of the Dun-
kard and are generally placed as very early Permian (Moran, 1952)

EVERETT C. OLSON

No. 8

or possibly as bridging the Stephanian-Autunian boundary (Dun-
bar, et al. 1960, Beerbower, 1963) .

The most fossiliferous locality in the Washington that has been
fully described to date is locality 6 of Moran (1952) located in the
SW14 sec. 18, T. 3 N., R. 4 W., Monroe County, Ohio, about 40 miles
northwest of Marietta. The fossils came from a series of limestones
and shales which ranges from 2 to 4 feet in thickness. The beds lie
about 7 to 10 feet below the Washington “A” Coal and are thus
more or less comparable in stratigraphic position to the site from
which T. stonei sp. nov. was taken. As identified by Homer (1952)
this assemblage includes:

Dittodus sp. ( =Xenacanthus sp.)

Sagenodus cf. S. periprion
Eryops cf. E. megacephalus
“Branchiosaurs”

Rhachitomi indet.

Diploceraspis burkei

Lysorophus dunkardensis (presence not certain)
Melanothyris morani (a jaw possibly pertaining to the genus
and species)

Edaphosaurus cf. E. boangeres
? Baldwinonus dunkardensis (jaw fragment)

Collections made during the last few years have produced an
extensive vertebrate assemblage from Belpre, Washington County,
Ohio. This has been reported by Hlavin, Windle and Wilcoxen
(1968). A heretofore unpublished faunal list from this locality sup-
plied by William Hlavin is as follows:

Ctenacanth-cladodont group (teeth and spines)

Xenacanthus sp.
cf. Ectosteorhachis sp.

Sagenodus sp.

Elonichthys sp.

Eryops cf. E. megacephalus
Rhachitomi indet.

Diploceraspis sp.

Lysorophus sp.

Megamolgophis sp.

Diadectes sp.

Edaphosaurus cf. E. boangeres
Dimetrodon sp.

Pelycosaurs, indet.

1970

TREMATOPS FROM THE DUNKARD

Some specimens from this site were kindly sent to me by Mr.
Hlavin and these give some added insight into the vertebrates pres-
ent at Belpre. A rhachitome, (probably Eryops) , Diploceraspis,
Diadectes, Edaphosaurus (cf. E. boangeres) , Ophiacodon and
Dimetrodon are readily identified in this assemblage. In addition
a large vertebral centrum probably is from Megamolgophis,
although it is not beyond the size range of the largest Texas and
Oklahoma specimens of Lysorophus. Furthermore, a very large,
rugose, tooth-bearing portion of the margin of the skull of a large
labyrinthodont is present. This represents Edops or a very Edops-
like amphibian.

The Belpre locality lies about 10 miles southwest of the
Marietta Trematops locality, and the specimens were found directly
above the Upper Marietta Sandstone. The bone-bearing deposit
represents a channel-fill and is composed of pebbles up to 2 inches
in diameter. The vertebrate remains were evidently washed in and,
although they are well preserved, they are fragmentary.

The fossils at locality 6 of Moran (1952) portray a typical fresh-
water pond and pond-margin array. Throughout much of the
Washington and the overlying Greene Formation such assemblages
predominate. Beerbower (1963) in his discussion of the paleo-
ecology of Diploceraspis , which covered most of the sites for which
substantial samples were known, indicated the predominant en-
vironment as that of “lakes and ponds.” The few identified stream
channels, except the one at Belpre, have yielded mostly unidentifi-
able scraps. None of the fossil-producing beds described by Moran
(1952) and Romer (1952) were formed by deposition in stream
channels.

This has resulted in a strong bias in preservation of types of
habitats, occasioned in large part, it would seem, by the fact that
resistant limestones provide the bulk of the outcrops. Natural
stream cuts and even road cuts in other materials are rapidly oblit-
erated. It seems highly likely that, as in the Midcontinent area,
more terrestrial environments existed adjacent to the standing
waters. Fragments and a few more complete specimens from the
Washington Formation have given indications of such environ-
ments. From Blacks ville, West Virginia, Melanothyris, a small
romeriid captorhinomorph, is indicative of a terrestrial habitat,
although the specimens were preserved in limestone nodules.

EVERETT C. OLSON

No. 8

Elsewhere specimens of Lysorophus are known, and Megamol-
gophis has come from one Washington locality. Lysorophus is quite
certainly an aestivator and, on the basis of great similarity of struc-
ture, Megamolgophis probably was as well. The distinctions between
the Dunkard Lysorophus specimens and those from Texas were
based primarily on size, expressed in length of vertebral centra.
It is now clear that the size ranges in the two regions completely
overlap and that there is no adequate morphological basis for sepa-
ration. The vertebrae of Megamolgophis are somewhat larger than
those of the largest Lysorophus from the Texas area, with the aver-
age central length as noted by Romer (1952) of 15 mm, about 2 mm
greater than the largest known from the Choza of Texas and the
Hennessey of Oklahoma. There is, as far as the vertebrae are
concerned, no reason to presume the habits to have been different.
The skull materials referred with some hesitation to Megamolgophis
by Romer (1952) indicate a quite different animal and, if the asso-
ciation is correct, may suggest a very different way of life.

If Lysorophus and perhaps Megamolgophis are indicative of
aestivation, they suggest the existence of seasonality in the climate
of the Dunkard area. It should be noted, however, that as yet no
evidence of the commonly aestivating dipnoan, Gnathorhiza, has
been reported.

Although much of the Belpre assemblage has the same general
cast as that from locality 6 of Moran (1952), the presence of Dime-
trodon and also of Diadectes suggests that somewhat different eco-
logical circumstances may have contributed elements to the total
assemblage. The large Ophiacodon at Belpre similarly suggests
that at least partially terrestrial pond and stream margins were
sources of parts of the faunal assemblage.

In this array of Washington specimens, however, even with
such genera as Diadectes, Ophiacodon and Dimetrodon present,
Trematops is something of an oddity. The genus is not a typical
representative of pond or lake assemblages in the Texas-Oklahoma
regions, where it is well known. Acheloma from the Wichita beds
perhaps comes slightly closer to filling such a role, but even it was
probably relatively highly terrestrial. Trematops in its typical
occurrences is associated with dissorophids, Seymouria, captorhinids
and Dimetrodon. It does not occur, except as possible fragments,
in stream deposits or in typical pond deposits characterized by

1970

TREMATOPS FROM THE DUNKARD

Xenacanthus, palaeoniscoids, T rimer orhachis, Eryops , Diplocaulus,
Edaphosaurus and, of course, the ubiquitous Dimetrodon.

The specimen of T. stonei in the Creston Shale of the Washing-
ton Formation gives very clear evidence, along with Melanothyris,
of the existence of a truly terrestrial life zone in the mideastern
region of the United States during the Early Permian. It would
appear that the various faunal subgroups, well known in the mid-
continent, occupying standing water, streams, margins of the ponds
and streams and the low divides or “uplands” persisted far to the
east from the places in which they were first recognized.

REFERENCES CITED

Beerbower, J. R., 1963, Morphology, paleontology and phylogeny of the Permo-
Carboniferous amphibian Diploceraspis : Mus. Comp. Zool. Bull., v. 130, p.
31-108.

DeMar, R. E., 1966, Longiscitula houghae , a new genus of dissorophid amphi-
bian from the Permian of Texas: Fieldiana, Geology, v. 16, p. 45-53.

Dunbar, C., et al., 1960, Correlation of the Permian Formations of North Amer-
ica: Geol. Soc. America Bull. 71, p. 1763-1806.

Hennen, R. V., 1911, Wirt, Roane and Calhoun Counties: West Virginia Geol.
Survey, County Rept., 573 p. (Creston Shale: p. 154).

Hickock, W. O. and Moyer, F. T., 1940, Geology and mineral resources of
Fayette County, Pa.: Pennsylvania Geol. Survey, 4th ser., Bull. C 26, 530 p.
(p. 149-151).

Hlavin, W. J., Windle, D. B. and Wilcoxen, J. M., 1968, A newly discovered
locality of Permian vertebrates in the Dunkard Series of Washington
County, Ohio [Unpublished abs.]: Ohio Acad. Sci., 77th Ann. Meeting.

McCue, J. B., et al., 1948, Clays of West Virginia: West Virginia Geol. Survey
Rept., v. 18, p. 9.

Moran, W. E., 1952, Location and stratigraphy of known occurrences of fossil
tetrapods in the Upper Pennsylvanian and Permian of Pennsylvania, West
Virginia and Ohio: Carnegie Mus. Ann., v. 33, p. 1-44.

Nace, R. L. and Bieber, P. P., 1958, Ground-water resources of Harrison
County, West Virginia: West Virginia Geol. Survey Bull. 14 (Creston
Shale: p. 18).

Olson, E. C., 1941, The family Trematopsidae: Jour. Geology, v. 49, p. 149-176.

, 1946, Fresh and brackish-water vertebrate-bearing deposits of the

Pennsylvanian of Illinois: Jour. Geology, v. 54, p. 281-305.

Romer, A. S., 1952, Late Pennsylvanian and early Permian vertebrates of the
Pittsburgh- West Virginia region: Carnegie Mus. Ann., v. 33, p. 47-113.

EVERETT C. OLSON

No. 8

Stauffer, C. R., 1916, Divisions and correlations of the Dunkard series of Ohio:
Geol. Soc. America Bull. 27, p. 86-88.

and Schroyer, C. R., 1920, The Dunkard series of Ohio: Ohio Geol.

Survey, 4th ser., Bull. 22, p. 1-167.

Vaughn, P. P., 1969, Further evidence of close relationship of the trematopsid
and dissorophid amphibians with a description of a new genus and species:
Southern California Acad. Sci. Bull., v. 68, p. 121-130.

MANUSCRIPT RECEIVED NOVEMBER 2, 1969

K<o\

si UV\

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO JANUARY 8, 1970

NUMBER 9

SOME ORNAMENTED ERISOCRINIDS
FROM THE AMES LIMESTONE

J. J. BURKE

ABSTRACT

Three new species of erisocrinid inadunate crinoids from the
Ames Limestone, Conemaugh Group, Pennsylvanian, are described.

All are ornamented forms. Delocrinus brookensis sp. nov. from
Brooke County, West Virginia, is more specialized than Delocrinus
tulsaensis (Strimple) from the Desmoinesian of Oklahoma, but is
probably a derivative of that species. Endelocrinus jpennsylvanicus
sp. nov. from Allegheny County, Pennsylvania, appears to be the
largest Pennsylvanian representative of Endelocrinus and is also
characterized by spine-bearing primibrachs. Paradelocrinus deco-
ratus sp. nov. from Brooke County, West Virginia, in addition to
being ornamented, is a relatively large species, but otherwise seems
to be a fairly conservative member of the genus.

In a discussion that follows, the recent proposal by Knapp to
remove from the Cladida forms with a dorsal cup having a basal
concavity and include them in a new order, the Declinida, is op-
posed because adequate grounds for such action are not evident.

Ornamented crinoids representative of the family Erisocrinidae
are fairly common in marine beds of the Conemaugh Group, Penn-
sylvanian, of the Allegheny region, and the three new species from
the Ames Limestone described in the present paper represent only
a few of the Conemaugh forms characterized by surface features
such as nodes, granules, or a combination of the two. Unfortunately,
ornamentation in many of these crinoids is quite variable, appar-
ently undergoes modifications during ontogeny, and is often poorly
preserved in fossil material. However, the specimens under de-
scription appear to represent mature forms, the ornamentation is
fairly well preserved, and my specific distinctions are not based on
ornamentation alone.

I wish to extend my appreciation to Mr, Bruce Frumker, staff
photographer of the Cleveland Museum of Natural History, for the

^THSO^?

FEB l 1

J. J. BURKE

No. 9

photographs from which the illustrations were taken, and to my
wife, Emily A. Burke, for aid in preparation of the manuscript and
arrangement of the illustrations.

SYSTEMATIC PALEONTOLOGY

Family ERISOCRINIDAE Miller, 1889
Genus DELOCRINUS Miller and Gurley, 1890
DELOCRINUS BROOKENSIS* sp. nov.
Fig. 1

Diagnosis : Dorsal cup in general resembling that of Delocrinus
tulsaensis (Strimple) but larger (width 23.4 mm) and higher (form
ratio .34) with deeper basal impression, less prominent nodose orna-
mentation, and anal X more reduced in size.

Fig. 1. Delocrinus brookensis sp. nov. Holotype, Cleveland Museum 3912, from
the Ames Limestone, Conemaugh Group, Brooke County, West Virginia.
a, dorsal view; b, posterior view; c, ventral view, Xl.

Holotype: Cleveland Museum of Natural History 3912, a complete
dorsal cup.

Occurrence: Ames Limestone, Conemaugh Group, Upper Pennsyl-
vanian.

Locality: Excavation (Tunnel Road Cut) for West Virginia Route
67 (lat. 40° 14' 24" N, long. 80° 35' 53" W.) near McKinley ville,
Brooke County, West Virginia.

Repository: Cleveland Museum of Natural History, Cleveland, Ohio.

Description: The dorsal cup of this species is low, truncate bowl-
shaped, and of medium size. In dorsal view the cup appears sub-
round, in ventral view pentagonal. The basal impression extends

1 Named for Brooke County, West Virginia.

1970

ORNAMENTED ERISOCRINIDS

somewhat beyond midheight and occupies about half the diameter
of the cup.

None of the stem is preserved. The infrabasals slope downward
steeply, and flare outward slightly. They extend for nearly three-
fifths the height of the basal impression. Proximally the curvature
of the basals is moderate, but distally, along the sides of the cup,
the curvature is more abrupt. Within the basal impression these
plates are gently concave transversely, and even slightly concave at
the basal plane, but become gently convex distally. Except for the
posterior basal, which is longer than wide, the basal plates are
slightly wider than long.

The radials are slightly less than twice as wide as long, have
fairly steep slopes, and are slightly convex along their length except
for a sharp incurvature distally to form a moderate forefacet. These
plates are broadly convex from side to side, contributing to the sub-
round outline of the cup in dorsal view. The articular facets show
denticles on the outer ligament areas and the transverse ridges,
prominent oblique fossae and broad intermuscular notches.

Anal X is relatively small, about a fourth longer than wide,
strongly incurved distally and slightly concave from side to side
proximally. It bears a single distal facet for articulation with a
second tube plate.

The major ornamentation consists of prominent nodes, the
largest of which are arranged in a festoon-like row bordering the
forefacet of each radial. Below these, scattered nodes occur on the
radials and on those portions of the basals that are part of the lateral
wall of the cup. In addition to the nodes, fine granular ornamenta-
tion is distributed over most of the cup, including the proximal por-
tion of the basals.

Linear measurements of the holotype, in millimeters, are given
in the table below:

Height of dorsal cup 7.9

Width of dorsal cup 23.4

Ratio of height to width 0.34

Height of basal impression 5.0

Width of basal impression 11.5

Length of basal (raB) 8.9

Width of basal (raB) 9.2

Length of radial (aR) 7.3

Width of radial (aR) 12.9

Length of suture between basals 6.0

Length of suture between radials 3.2

Length of anal X 4.8

Width of anal X 3.6

J. J. BURKE

No. 9

Remarks : As the diagnosis indicates, this species resembles Delo-
crinus tulsaensis (Strimple, 1962). It is probably a direct derivative
of the Oologah species and as such may be taken as an indicator of
the rate of evolution in this line of crinoids from Oologah to Ames
time.

Knapp (1969) advocates that the generic name Delocrinus
Miller and Gurley be treated as a nomen dubium and the name of
the genotype species, Poteriocrinus hemisphericus Shumard, 1858,
a nomen nudum. In support of these proposals he notes that Shum-
ard’s types have been lost, regards Shumard’s description as inade-
quate and indicates that no described specimens subsequently as-
signed to the species agree with the original description. However,
I am in agreement with Moore and Plummer (1940, p. 253) in feel-
ing that “the essential characters of the genus and its validity are
not in doubt.” Furthermore, although the species is probably a rare
one, I think that the description will prove adequate to distinguish
it, if additional specimens are discovered. In any case, as long as
there is a possibility that topotype material may be found, Delo-
crinus hemisphericus (Shumard) has claim as a valid species.

In a previous paper (Burke, 1966) I regarded Strimple’s (1961)
proposed genus Graffhamicrinus as a synonym of Delocrinus, be-
cause it was based on surface ornamentation, a feature insufficient
for generic distinction. Knapp (1989) recognizes Graffhamicrinus
as a genus but (p. 363) emends it to include several other species
previously included under Delocrinus “in which the proximal tips
of the radial plates are visible in side view, the cups may or may
not be ornamented, and in which the first primibrachs, which are
axillary, may or may not bear spines.”

As thus emended, Graffhamicrinus does not appear to me to
have any better claim to generic distinction than it had previously.
In this connection it is interesting to note that the measurements
given by Shumard (1858, p. 221) for Delocrinus hemisphericus indi-
cate that the proximal tips of the radials of that species are visible
in side view.

1970

ORNAMENTED ERISOCRINIDS

Genus ENDELOCRINUS Moore and Plummer, 1940
ENDELOCRINUS PENNSYLVANICUS sp. nov.

Fig. 2

Diagnosis : A large species, near Endelocrinus texanus (Weller) in
size (width of dorsal cup .21 mm) but cup higher (form ratio .36)
basal impression nearly twice as deep, walls steeper, and plates
ornamented with nodes and granules. Primibrachs spinose.

Fig. 2. Endelocrinus pennsylvanicus sp. nov. Holotype, Cleveland Museum
3913, from the Ames Limestone, Conemaugh Group, Allegheny County, Penn-
sylvania. a, dorsal view; b, posterior view; c, ventral view; d, E-ray view
showing IBn, IIBri_2 (A ray on left) ; e, oblique ventral view showing articular
facets of IBrx and IIBr2, E ray, Xl.

Holotype: Cleveland Museum of Natural History 3913, a dorsal cup
with the first primibrach and two secundibrachs of the E ray
attached.

Occurrence: Ames Limestone, Conemaugh Group, Upper Pennsyl-
vanian.

Locality: Excavation at Holiday Park (lat. 40° 27' 46" N., long. 79°
42' 57" W.) Plum Borough, Allegheny County, Pennsylvania.

Repository: Cleveland Museum of Natural History, Cleveland, Ohio.

Description: The dorsal cup is truncate bowl-shaped and of medium
height. The vertical extent of the basal impression is nearly three-
fifths the height of the cup, and the width of the impression is slight-
ly less than two-fifths the width of the cup. The infrabasal circlet

J. J. BURKE

No. 9

shows a pentalobate central canal within the stem impression. The
infrabasals are steep-walled, occupy more than half the height of
the basal impression, and flare outward to the extent that they
would be visible if the stem were attached.

The basals are about one-sixth wider than long. They flare
steeply outward and downward from the basal impression, but less
so than the infrabasals. However, the steepest longitudinal slopes
of the basals are in their proximal portions; the convexity in the
midregion is slight, but the upward (distal) slope is fairly steep.
These plates are strongly concave from side to side proximally.
They continue to be concave from side to side, although much more
gently, to their midregion and even beyond their proximal contact
with the radials along the lateral wall of the cup. Beyond the steep
wall of the basal impression the basal sutures are slightly impressed.
The distal margins of the basals are gently curved. At all angles
where the basals and radials meet, these plates are inflexed to form
the characteristic Endelocrinus pits.

The ornamentation of the basals is of two types. One type con-
sists of closely spaced rounded granules that cover the entire plate
except for the steep wall of the basal impression. The other is in
the form of coarse nodes that are confined to the distal part of each
plate. The most proximal of these nodes are paired and shared by
adjacent basals near the distal termination of each interbasal suture.
The basal plane is tangent to the tips of these paired nodes.

The radials are a little more than two-thirds wider than long.
They show slight convexity along their length, although each is
rather abruptly incurved near the summit, forming a distinct fore-
facet. The transverse curvature is moderate and most pronounced
about midheight. The ornamentation comprises nodes and granules
similar to those on the basals. The forefacet is free of nodes, but
there is a node at the lateral termination of each radial at the sum-
mit. Otherwise the radials bear no nodes closely adjacent to the
interradial sutures.

The radial facets are relatively shallow. The outer ligament
area is short and is denticulate in the region external to the slitlike
ligament pit which occupies most of its extent. The transverse
ridge extends essentially to the full width of the facet; it is fairly
well defined and denticulate. The inner ligament fossae are rather
deep and extend almost parallel to the transverse ridge to points
nearly opposite the terminations of the outer ligament area. The

1970

ORNAMENTED ERISOCRINIDS

oblique ridges are strong and denticulate. The intermuscular notch
is quite broad and there appear to be indications of muscle areas
flanking the intermuscular furrow. The lateral ridges face outward,
but are relatively low. The adsutural slopes are gentle and the ad-
sutural platforms extensive.

Anal X has been displaced; its distal end lies in the body cavity
and the entire plate has shifted inward between the posterior ra-
dials. Proximally, anal X rested on the truncated tip of the pos-
terior basal, and along with the basal and radials was indexed to
form a small pit on each side at the angles where it met those plates.
For most of its height anal X is slightly concave from side to side.
It bears a single distal facet for articulation with a second anal
plate.

A single axillary first primibrach, that of the E ray, remains
in place. It is more than twice as wide as long, and bears a short
but distinct spine. In addition to granulose ornamentation similar
to that of the dorsal cup, this plate bears a node on each lateral
shoulder, below and flanking the spine.

The distal articular surface of the primibrach is divided into
right and left facets by a prominent, partly denticulate ridge. The
right facet is fully exposed. It shows an outer ligament area that
is relatively longer and much less compressed than the outer liga-
ment areas of the radials. This ligament area is denticulate. The
external ligament pit is slightly elongate but not slitlike as in the
radials. It does not occur at midlength of the ligament area, but is
somewhat nearer the termination of the area in the vicinity of the
ridge that separates the facets. A strong denticulate transverse crest
is present, together with well-defined ligament fossae. Although
most of the articulation appears to have been of the ligamentary
type, structures on the inner side of the ridge that separates the
facets strongly suggest muscle scars.

The first two secundibrachs of the left arm of the E ray are
also present. The first secundibrach has been displaced outward,
carrying the second with it. Both of these plates show granulose
ornamentation. The first secundibrach bears a nodose swelling on
the right side. The second secundibrach displays a large node or
blunt spine in this region.

The first secundibrach is quadrangular and about three times
as wide as long. Because of the position of the plate I have not been

J. J. BURKE

No. 9

able to determine whether it bore a pinnule and interlocked with
adjacent plates, although presumably it did.

The second secundibrach is quite low and elongate from side
to side in external view. Because its lateral terminations are both
angular, it is not a typical cuneate plate. Laterally, the plate shows
structures for interlocking with plates of adjacent arms. These are
especially evident as bulbous swellings on the left (pinnular) side.
A pinnule socket is present on the internal side. The distal articu-
lar facet bears a prominent crenulated outer ligament area, traces
of an outer ligament pit, and a transverse crest. The inner ligament
area is more extensive on the left (pinnular) side, and extends in-
ward and laterally in that region, rising up to a rounded prominence
well above the rest of the articular surface. Along the outer border
of this left side, but internal to the transverse crest, are additional
crenulations.

The intermuscular notch of this second secundibrach is sharply
angular and the intermuscular furrow originates near the trans-
verse crest. The flexor muscle scar of the left (pinnular) side orig-
inates along the internal reaches of the intermuscular furrow and
extends as a channel bordering the pinnular side of the intermus-
cular notch for about half the length of that side of the notch. The
channel expands widely toward its termination.

The flexor muscle scar of the right (antipinnular) side of this
second secundibrach is separated from the muscle scar of the pin-
nular side by the intermuscular furrow, and runs parallel to the
latter to a point a little beyond the exit of the furrow. It is deeper
than the scar on the pinnular side, but much less elongate.

To the right of the antipinnular muscle scar the articular sur-
face is rounded, elevated, and bears crenulations suggestive of liga-
ment attachment areas. From that place to the transverse ridge the
antipinnular articular surface is concave and slopes gently laterad.

Linear measurements of the holotype specimen, in millimeters,
are shown in the following tabulation:

Height of dorsal cup

Width of dorsal cup

Ratio of height to width
Height of basal impression
Width of basal impression

Length of basal (raB)

Width of basal (raB)

Length of radial (laR) ___

7.5
21.0

0.36

4.2
8.8
7.1

8.3

6.5

1970

ORNAMENTED ERISOCRINIDS

Width of radial (laR) 11.2

Length of suture between basals 5.4

Length of suture between radials 3.4

Length of anal X 4.4

Width of anal X 3.6

Length of first primibrach (E ray) 5.0

Width of first primibrach (E ray) 11.6

Length of first secundibrach (E ray) 2.5

Width of first secundibrach (E ray) 7.5

Length of second secundibrach (E ray) 1.8

Width of second secundibrach (E ray) 5.8

Remarks: In ornamentation, Endelocrinus undulatus (Strimple,
1961) bears resemblance to this species, although the nodes are
apparently less prominent. In addition, Endelocrinus undulatus is
a smaller species, the dorsal cup is relatively higher, the basal im-
pression is wider, the primibrachs do not bear spines and both
primibrachs and secundibrachs have greater height.

From the standpoint of size, Endelocrinus pennsylvanicus ap-
pears to be the largest Pennsylvanian representative of the genus.

Knapp (1969) would restrict Endelocrinus to two species, En-
delocrinus fayettensis (Worthen) and Endelocrinus bifidus Moore
and Plummer, characterized by gently downflaring infrabasals. He
includes Endelocrinus texanus (Weller) in his proposed genus
Metarrectocrinus for the same reason. He emends Strimple’s genus
Tholiacrinus, which as originally proposed, was considered synony-
mous with Endelocrinus by Burke (1966) and Webster and Lane
(1967) because it was based solely on ornamentation, a feature in-
sufficient for generic distinction. As emended by Knapp, the sole
basis for distinction between Endelocrinus and Tholiacrinus would
be steeply downflaring infrabasals, which he assumes to character-
ize Tholiacrinus.

Dorsal cups of Endelocrinus are distinguished by the character-
istic pits at the angles where the radials and the basals meet, and
usually by convexity or even bulbosity of these plates of the cup.
These salient generic characters are shown in common by Endelo-
crinus fayettensis, Endelocrinus texanus , Endelocrinus bifidus, and
the various species of Endelocrinus which Knapp would lump to-
gether under Tholiacrinus. There is no evidence that the degree of
downflaring of the infrabasals modifies the dorsal cups of these spe-
cies to the extent that the generic characters cited above are not
evident. For this reason I regard Tholiacrinus as emended by
Knapp to be a synonym of Endelocrinus.

J. J. BURKE

No. 9

Genus PARADELOCRINUS Moore and Plummer, 1938
PARADELOCRINUS DECORATUS sp. nov.

Fig. 3

Diagnosis : A relatively large species of Paradelocrinus (width of
dorsal cup 25.7 mm); form ratio .37; cup pentagonal in ventral view;
basal impression deep and about half as wide as the cup; basal
plates concave from side to side at basal plane; basals and radials
ornamented with fine ridges and prominent nodes.

Fig. 3. Paradelocrinus decoratus sp. nov. Holotype, Cleveland Museum 3914,
from the Ames Limestone, Conemaugh Group, Brooke County, West Virginia.
a, dorsal view; b, posterior view; c, ventral view, Xl.

Holotype : Cleveland Museum of Natural History 3914, a dorsal cup
complete except for the infrabasal circlet.

Occurrence : Ames Limestone, Conemaugh Group, Upper Pennsyl-
vanian.

Locality: Excavation (Tunnel Road Cut) for West Virginia Route
67 (lat. 40° 14' 24" N, long. 80° 35' 53" W.) near McKinley ville,
Brooke County, West Virginia.

Repository: Cleveland Museum of Natural History, Cleveland, Ohio.

Description: The dorsal cup is low, truncate bowl-shaped and near-
ly three times wider than high. It is subpentagonal in dorsal view
and pentagonal in ventral view. The basal impression occupies
about half the diameter of the cup. The impression was evidently
quite deep, but the infrabasal circlet, which represented about one-
seventh of the cup diameter, is missing and the full height of the
impression cannot be determined.

The basal plates comprise most of the base of the cup; their
distal tips extend for more than half the height of the radials. All

1970

ORNAMENTED ERISOCRINIDS

of the basals, including the posterior basal, which is longer than the
others, are pentagonal in outline. These plates curve rather strongly
upward into the basal impression proximally; distally, beyond the
basal plane, they again curve upward, but more gently. Each basal
plate is also a little concave from side to side. The basal plane in
consequence, is tangent to points along the basal sutures, but a gen-
tle concavity characterizes these plates distally, well above the basal
plane. The distal margins of the basals are rounded to the tips,
rather than angular.

Each of the radial plates is pentagonal in outline and nearly
twice as wide as high. In keeping with the longer posterior basal
plate, the suture between the two posterior radials is shorter than
the other interradial sutures. Proximally the radials slope outward
and upward with little convexity, retaining the distal slope of the
basals. Toward the summit of the cup the radials become a little
more convex from side to side, and more convex along the length,
just below the outer ligament area.

The articular facets of the radials are of medium depth. The
outer ligament area is narrow and elongate, with a slitlike outer
ligament pit. The transverse crest extends to the lateral extremities
of the facet and is denticulate. In the inner ligament area the lat-
eral furrows are elongate and rather narrow and the oblique ridges
are denticulate. The adsutural slopes are abrupt and the adsutural
platforms wide. The lateral ridges slope outward fairly strongly.
The intermuscular notch is broad, leading into a short intermuscular
furrow. A broad, low, concave region, representing the two muscle
areas, is bordered by the notch and divided by the furrow.

Anal X is entirely shut off from the outer side of the cup by
the posterior radials, which share a CD interradial suture on that
side. At the summit of the cup the left lateral corner of the C radial
is slightly elevated and bent inward. At the termination of this pro-
longation of the radial corner, anal X fits between the lateral ridges
of the adjacent posterior radials. In ventral view the plate is tri-
angular in outline; it widens as it approaches the body cavity and
rises above the inner articular surfaces of the radials. Its ventral
surface bears a facet for articulation with a second anal plate.

The ornamentation is of two kinds, the most prominent consist-
ing of nodes on the radials and the basals. On the radials the nodes
are arranged in three rows. The proximal two rows tend to parallel

J. J. BURKE

No. 9

the sutures between the radials and the basals. The distal row is
somewhat transverse and the nodes are larger than those of the
proximal rows. There is also a row of nodes on each basal, which
borders the sutures between the radials and the basals, and nodose
swellings border the interbasal sutures distally. Irregular nodes are
also distributed over the portions of the basals that rise above the
basal plane. The sutures are not impressed, but where nodes border
the sutures they appear to be.

A second, finer kind of ornamentation comprises wrinkles or
ridges on the basals and the radials. These ridges also extend across
the nodes, where in general they show an arrangement at right
angles to the sutures. The wrinkled ornamentation persists part
way into the basal impression.

Linear measurements of the holotype, in millimeters, are as
follows:

Height of dorsal cup 9.5

Width of dorsal cup 25.7

Ratio of height to width 0.37

Width of basal impression 10.8

Length of basal (raB) 9.5

Width of basal (raB) 9.9

Length of radial (aR) 7.6

Width of radial (aR) 15.5

Length of suture between basals 7.1

Length of suture between radials 4.3

“Length” of anal X 5.3

Width of anal X 3.3

Remarks : In comparison with the genotype species, Paradelocrinus
aequabilis Moore and Plummer, 1938, Paradelocrinus decoratus ap-
pears to be a fairly typical representative of the genus. It is, of
course, a much larger species than Paradelocrinus aequabilis , the
cup is higher, the basals are more prominent, and the cup is orna-
mented. Nevertheless, in the outward flare of the cup walls, deep
basal impression and appearance of anal X, it is closer to Paradelo-
crinus aequabilis than many other forms that have been referred to
Paradelocrinus.

There are some resemblances between Paradelocrinus decoratus
and Paradelocrinus planus (White, 1880) in cup height, depth of
basal impression, and size of basal circlet, but the cup of Paradelo-
crinus decoratus is larger, more pentagonal, and ornamented rather
than smooth.

1970

ORNAMENTED ERISOCRINIDS

In ventral view, Paradelocrinus iolaensis Strimple, 1949, is
reminiscent of Paradelocrinus decoratus in its pentagonal outline
and appearance of anal X. The basal plates of Paradelocrinus
iolaensis are concave from side to side in the vicinity of the basal
plane, but more concave than those of Paradelocrinus decoratus.
The dorsal cup of Strimple’s species is also smaller, of lesser height,
more pentagonal in dorsal view, and unornamented.

The Atokan species described by Knapp (1969) as Atokacrinus
obscurus is likewise similar to Paradelocrinus decoratus, although
the dorsal cup is not as high. The dorsal cup of the Knapp type is
worn, and may have borne ornamentation, but the flattened de-
pressed areas that bound the sutures do not characterize the cup
of Paradelocrinus decoratus. Like those of Paradelocrinus decora-
tus, the basals of Knapp’s species are concave from side to side at
the basal plane.

In proposing Atokacrinus as a new genus, Knapp gives for a
diagnosis “Basal concavity deep, infrabasals steeply downflared.”
Knapp has also proposed a new genus, Sublobalocrinus, to include
Paradelocrinus iolaensis and Paradelocrinus planus, based on the
following diagnosis: “Basal concavity narrow and deep; infrabasals
steeply downflaring, basals transversely concave; arms unknown.”

Like Paradelocrinus aequabilis and Paradelocrinus decoratus,
the species comprising Knapp’s proposed new genera are fairly con-
servative forms, which, except for the characteristic curvature of
the cup in lateral view and absence of anal X from the outer side
of the cup, strongly resemble the dorsal cups of various species of
Delocrinus. For such forms, Knapp’s generic characters might have
specific, but not generic weight, and I regard Sublobalocrinus and
Atokacrinus as synonyms under Paradelocrinus.

DISCUSSION

This article was essentially completed before the appearance of
Knapp’s (1969) paper dealing with his proposed new order Decli-
nida. As a consequence some revision was necessary in order to deal
with new taxa introduced by Knapp and various taxonomic changes
which he advocates. I feel obligated to summarize certain of my
views concerning the content of Knapp’s paper, however, since I fail

J. J. BURKE

No. 9

to find justification for the establishment of the Declinida as a new
order.

Knapp (ibid. p. 343) contends that in the Inadunata, down-
flaring infrabasals cannot have evolved from infrabasals that were
originally upflaring, and states: “With the infrabasals established
as a stable element in the dorsal cups of crinoids having anal X and
radianal plates, the stage was set for the origin of a basal concavity
having steeply downflaring infrabasals. The origin of steeply down-
flaring plates can be envisioned by reference to the position at which
the infrabasal plates appear in the early growth stages of Recent
crinoids, that is, deep within the basal circlet and steeply flaring.
Rather than moving down from within the basal circlet as in Recent
crinoids and presumably inadunates incorporating upflaring infra-
basals, the infrabasal plates remained deep within the basal circlet
and a basal concavity was created in fossil inadunates. As in fossil
inadunates possessing upflaring infrabasal plates, evolution pro-
ceeded from steeply downflaring infrabasals to horizontal infra-
basals and a flat base.”

Knapp’s theory that the basal concavity and downflaring infra-
basals originated in the larval stage is a novel one, to say the least.
Such major evolutionary changes are usually developed in gradual
stages over long periods of time, and are traceable in adult forms.
However, it may be questioned whether the larvae of modern cri-
noids, which share the extreme specialization of the adult forms,
can be taken as indicators of an evolutionary sequence such as
Knapp outlines.

In Antedon adriatica Clark (1921, p. 414) , in a summary of
Seeliger (1893) reports that at the earliest larval stage at which the
infrabasals are found, they are anterior to the basals. On the other
hand, Mortensen (1920, p. 26, pi. 12, fig. 3) finds the first detected
infrabasal plate of Compsometra serrata lying within the basal cir-
clet. But it is not until later stages, when the infrabasals are sur-
rounded by the faster growing basal plates that they can be de-
scribed as “deep within the basal circlet and steeply flaring.” I fail
to find evidence to support Knapp’s view that the basal concavity
of inadunates arose as a spontaneous development in the larval
stage.

Knapp (1969, p. 351) stipulates that for inclusion in the Decli-
nida the crinoid must possess what he terms a “structural” basal

1970

ORNAMENTED ERISOCRINIDS

concavity — the infrabasals must flare downward. I gather from this
that if the walls of the concavity were composed of the basals and
were topped by a horizontal infrabasal circlet, an inadunate show-
ing such a concavity would be excluded from the Declinida. How-
ever, it is evident that if the one type of concavity could have arisen
in the larval stage, the other could have also.

It appears far more likely that the stages in evolution from up-
flaring to flat and finally to downflaring infrabasals illustrated by
Moore, Lalicker and Fischer (1952, p. 624) represent the true course
of change in attitude of these plates, rather than that suggested by
Knapp. The latter writer quite apparently holds no brief for the
derivation of downflaring from flat infrabasals. Attributing the de-
crease in steepness and flare and eventual “horizontality” of the
radial and basal plates of Antedon to differential accretion of cal-
cite during growth, Knapp (1969, p. 343) contends that these plates
cannot change their horizontal attitude, once it has been attained,
by accretion of calcite at their lateral margins. At this “horizontal”
stage the plates are probably joined by zygosynostosial articulations,
which, as Van Sant (1964, p. 39) states, are mostly immovable.

Now it is true that such articulations could prevent horizontal
infrabasal plates from progressing to a downflaring attitude, and
indeed zygosynostosial articulations appear to characterize the ad-
jacent sides of flat infrabasal plates of various Upper Carboniferous
inadunates. However, I have observed specimens of Polusocrinus,
Aesiocrinus, Parulocrinus and Etlnelocrinus in which individual
plates are separated from others in the circlet or shifted from
sutural position in such a manner as to suggest very slight calca-
reous deposits and the probability that at early stages of growth
they were bound together by connective tissue only.

It seems reasonable to me that a similar very loose sutural
union bound together the infrabasal plates of inadunates during
the evolutionary transition from upflared to downflared plates that
produced the basal concavity. Probably in early stages of growth
the circlet remained flat and was supported at the center by the
stem, but with the increase in weight of the expanding dorsal cup
the infrabasals sagged downward, extending below the top of the
stem peripherally, thus giving rise to the basal concavity. Calcite
deposition may have been arrested up to this point; the height of
the concavity in that case probably depended on (1) the rate of re-
newed calcite deposition leading to zygosynostosial articulation and

J. J. BURKE

No. 9

(2) the rate of upgrowth, if any, of the proximal portions of the
basals.

Upgrowing basals probably influenced the angle of flare of the
downflaring infrabasals before zygosynostosial articulations devel-
oped. Apparently synostosial articulation was not unusual between
the basals and infrabasals of some Upper Carboniferous inadunates.
In some species of Delocrinus, where such articulations occur, a
slender flange at the proximal tip of each basal inserts into a short
notch in the infrabasal circlet, preventing the closure of the infra-
basal suture at that place. This indicates that prior to the time that
the infrabasals were united by close suture they were separated, to
some extent at least, by the basal tips. It would appear that in such
specimens, if there is any possibility that the infrabasals are not yet
joined by close sutures, it would be unwise to assume that the angle
of flare of the infrabasals was stabilized and representative of the
particular species involved.

If the infrabasal circlet had remained flat, and was bound to
the proximal portions of the upcurving basals by loose suture, the
continued growth of the basals would have resulted in their forming
the walls of the basal concavity, with the infrabasal circlet roofing
the concavity. This is the type of basal concavity found in Plummer -
icrinus. Since from a mechanical standpoint such a concavity seems
to have served the same purpose as one with downflaring infra-
basals, I fail to see why Knapp considers the distinction between
the two of major taxonomic importance.

Obviously I do not agree with Knapp that evolution of the taxa
he includes in the Declinida proceeded from downflaring infra-
basals to horizontal infrabasals and a flat base. Of course it is pos-
sible, because all stages are represented in the sequence from flat
base to steeply downflaring infrabasals, to juggle various species of
Upper Carboniferous inadunates into groups arranged in chrono-
logical order supposedly illustrating the reverse of that sequence.
But many of these species are based on single specimens, and few
are represented by an adequate number of individuals to determine
extent of variation within a species, or to permit ontogenetic studies.
Such being the case, to assume constancy in the angle of flare of
the infrabasals in such species is not justified.

1970

ORNAMENTED ERISOCRINIDS

REFERENCES CITED

Burke, J. J., 1966, Endelocrinus kieri, a new crinoid from the Ames Limestone:
Ohio Jour. Sci., v. 66, p. 459-464.

Clark, A. H., 1921, A monograph of the existing crinoids: U. S. Nat. Mus. Bull.
82, v. 1, pt. 2, p. 1-795.

Knapp, W. D., 1969, Declinida, a new order of Late Paleozoic inadunate cri-
noids: Jour. Paleontology, v. 43, p. 340-391.

Moore, R. C., and Plummer, F. B., 1940, Crinoids from the Upper Carboniferous
and Permian strata in Texas: Univ. Texas Bull. 3945, 468 p.

Moore, R. C., Lalicker, C. G., and Fischer, A. G., 1952, Invertebrate fossils:
New York, McGraw-Hill, 766 p.

Mortensen, T., 1920, Studies in the development of crinoids: Carnegie Inst.
Washington Dept. Marine Biology Papers, v. 16, p. 1-94, pi. 1-28.

Seeliger, O., 1893, Studien zur Entwicklungsgeschichte der Crinoiden ( Antedon
rosacea ) : Zool. Jahrb. Abth. f. Morph., v. 6, p. 161-444. [Not seen]

Shumard, B. F., and Swallow, G. C., 1858, Descriptions of new fossils from the
Coal Measures of Missouri and Kansas: Acad. Sci. St. Louis Trans., v. 1,
p. 199-227.

Strimple, H. L., 1961, Late Desmoinesian crinoids: Oklahoma Geol. Surv. Bull.
93, 189 p.

, 1962, Crinoids from the Oologah Formation: Oklahoma Geol. Surv.

Circ. 60, 75 p.

Van Sant, J. F., and Lane, N. G., 1964, Crawfordsville (Indiana) crinoid
studies: Kansas Univ. Paleont. Contr. Echinodermata, Art. 7, p. 1-136.

Webster, G. D., and Lane, N. G., 1967, Additional Permian crinoids from south-
ern Nevada: Univ. Kansas Paleont. Contrib. Paper 27, 32 p.

MANUSCRIPT SUBMITTED DECEMBER 15, 1969

sx K1V\

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO APRIL 30, 1970 NUMBER 10

A NEW ANTHRACOSAURIAN LABYRINTHODONT,
PROTEROGYRINUS SCHEELEI,

FROM THE LOWER CARBONIFEROUS

ALFRED SHERWOOD ROMER
Museum of Comparative Zoology, Harvard University

ABSTRACT

Incomplete remains of the skull and skeleton of an anthraco-
saurian labyrinthodont from the basal part of the Mauch Chunk
Group of the Mississippian of Greer, West Virginia, are described
as Proterogyrinus scheelei, gen. et sp. nov. and made the type of
a new family Proterogyrinidae. The type resembles the Embolo-
meri in most regards but is more primitive in that both inter -
centra and pleurocentra are in the form of dorsally incomplete
rings.

INTRODUCTION

As noted in a previous publication in this series (Romer, 1969),
almost nothing has been known of labyrinthodont amphibians in the
earlier, Mississippian, portion of the Carboniferous. This lacuna is
in process of being filled in considerable measure by specimens from
the Greer quarry in West Virginia. In my previous paper I noted
the history and stratigraphy of the locality. In that paper I de-
scribed the skull and partial skeleton of a rhachitome from Greer;
specimens of an anthracosaur and a second rhachitome are in proc-
ess of description by Dr. Nicholas Hotton III; several further Greer
specimens are in process of study and collection. It is to be hoped
that before the possibilities of the Greer quarry are exhausted we
may attain a broad representation of the labyrinthodont fauna of
the Lower Carboniferous of North America.

Science is indebted to Mr. John J. Burke and Mr. William E.
Moran and, more recently, to Mr. William Hlavin for their success-

ALFRED SHERWOOD ROMER

No. 10

ful exploration of the Greer site, to the Greer Limestone Company,
owners of the property, for their cooperation, and to Mr. William E.
Scheele, Director of the Cleveland Museum of Natural History, for
his promotion of the work.

In the present paper is described an anthracosaur differing from
that under description by Dr. Hotton. The specimen is fragmentary
in nature, but is of interest, particularly, as showing a hitherto un-
known type of anthracosaurian vertebral structure.

SYSTEMATIC PALEONTOLOGY

Family PROTEROGYRINIDAE fam. nov.

Diagnosis : Anthracosaurian labyrinthodonts, resembling typical em-
bolomeres in most regards, but with both intercentra and pleuro-
centra in the form of incomplete rings, open dorsally. Type genus
Proterogyrinus.

Genus Proterogyrinus1 gen. nov.

Proterogyrinus scheelei2 sp. nov.

Figs. 1-8

Diagnosis for genus and species: Structure, so far as known, similar
in most regards to such an embolomere as Archeria. Snout mod-
erately elongate; length of frontal and nasal combined nearly twice
as long as parietal and postparietal. Skull roof lightly sculptured
with small shallow pits and short grooves.

Holotype : Cleveland Museum of Natural History 10950.

Occurrence : Bickett Shale of the Bluefield Formation, Mauch
Chunk Group, Mississippian.

Locality: Greer, Monongalia County, West Virginia, on Deckers
Creek, about 6V2 miles southeast of Morgantown.

Repository: Cleveland Museum of Natural History, Cleveland, Ohio.

1 The generic name continues the series of anthracosaurian names based on
“gyrinus” by Watson, and suggests the relative antiquity of the present form.

2 The specific name is in honor of Director William E. Scheele, who has enthu-
siastically promoted the search for Greer amphibians.

1970

A NEW ANTHRACOSAURIAN

DESCRIPTION

Cranial remains : Two large slabs show disarticulated and scattered
poster anial remains; a smaller block contains incomplete remains
of cranial structures (as well as several anterior vertebrae) (Figs.
1-2) •

Best preserved of skull materials are the dorsal series of roofing
bones, from nasals back to postparietals and tabulars. These are
preserved almost intact except for some disruption of the right side
of the table posteriorly. This type of preservation of the skull roof
is common in anthracosaurs, due principally to the loose connection
of skull table and cheek in typical members of this group, and aided
anteriorly by the apparently sharp drop of the sides of the snout
from the frontals and nasals. For example, in the Harvard collec-
tion of Archeria materials from the Geraldine bonebed of the Texas
Permian, no less than five specimens show a complete or nearly
complete series of dorsal roofing elements broken off from the ele-
ments of the side of the skull. The skull roof is but lightly sculp-
tured; near centers of ossification there are groups of small, shallow
pits; farther out one finds a series of short and shallow radiating
grooves. Of lateral line grooves, only faint and uncertain traces are
to be seen.

The skull table structures are comparable to those of embolo-
meres. Slender tabular “horns” are present, although broken off.
In the lateral series of table elements, the tabulars are somewhat
larger than typical, and in consequence the two temporal elements
are somewhat reduced in size. The suture between intertemporal
and supratemporal is obscure, but apparently the former element
is of small size. The parietal does not expand as much postero-
laterally as is usually the case.

Facial length is variable in anthracosaurians, but snout elonga-
tion is common in embolomeres. If we assume that the joint length
of parietals plus postparietals is relatively constant, we find that,
for example, the length of nasals and frontals together is about 180
percent of this figure in Palaeogyrinus, 260 percent or so in mem-
bers of the Pteroplax-Eogyrinus group, 300 percent and upward in
Archeria. Proterogyrinus is relatively short faced, with nasal and
frontal about twice the length of the posterior table elements.

Apart from the dorsal roofing elements, skull remains are few
and generally scattered. Fragments of both prefrontals and of the

ALFRED SHERWOOD ROMER

No. 10

Fig. 1. Proterogyrinus scheelei Romer, C.M.N.H. 10950. The block exhibiting
cranial materials, X %.

left postorbital are seen adjacent to the dorsal series. Far to the
right of the skull table is a roughly triangular mass of bone which
may represent the right cheek area of squamosal and quadrato jugal.
Several other pieces of bone lying to the right of the skull roof may
be part of the dermal elements of the right side of the face, but I
have not attempted to identify them.

To the right is found a crushed but nearly complete right ptery-
goid, seen from the inner or lower surface; in its extent it is com-
parable to that of Palaeogyrinus as figured by Watson (1926) and
by Panchen (1964) . The thickened portion of the epipterygoid sup-
porting the anteroventral surface of the basipterygoid process is
evident, but the more dorsal region of the socket for the process is
not clear. A bar of bone rising straight upward from this area is
presumably an imperfect columella cranii; whether further remains

1970

A NEW ANTHRACO SAURIAN

Fig. 2. Proterogyrinus scheelei Romer, C.M.N.H. 10950. To show, in outline,
identified elements present on the block shown in figure 1, X SA' clt, articular
region of right lower jaw; j, frontal; it, intertemporal; n, nasal; na, neural arch;
p, parietal; pf, postfrontal; po, postorbital; pp, postparietal; prf, prefrontal;
pt, pterygoid; sq, squamosal; st, supratemporal; sym, symphysial region of
right lower jaw; t, tabular.

of the epipterygoid are present, concealed beneath the pterygoid,
cannot be determined.

A bar of bone bearing about 40 teeth (plus a few empty alveoli)
is seen to the right of the series of dorsal elements.1 The tooth row
is of length appropriate to its being considered nearly a full denti-
tion for a maxilla or dentary of the present specimen. Most of the
teeth are about 3 mm long, subcircular in section, about % mm in
diameter and closely spaced. In some cases the basal section of the
tooth can be seen to be grooved in labyrinthine fashion. The tips
are blunt and when well preserved appear to be bevelled and tilted
slightly toward that end of the bar which lies anteriorly on the slab.

1 A section of this series running beneath the pterygoid has been developed
since the photograph of figure 1 was taken.

ALFRED SHERWOOD ROMER

No. 10

The nature of this tooth-bearing element is somewhat puzzling.
At first sight one would assume that it is the right maxilla, little
displaced. But there is a distinct longitudinal shelf the length of the
bone, superficial to the teeth as they lie on the bone. We are, hence,
looking at the inner surface of the bone. But if the bone is a max-
illa, it must be either the left element strongly displaced, or the
right maxilla rotated nearly 180°. Suggesting the latter interpreta-
tion is the fact that the teeth which lie most anteriorly are smaller
than most of the series, and hence may pertain to the posterior end
of the tooth series.

To add complexity to the situation, there are present, further to
the right in the slab and not far from the “front” end of the tooth-
bearing element, the articular end of a right lower jaw and, far to
the rear, the symphysial end of a right jaw. It is tempting to con-
sider that the major structure we are dealing with is a dentary,
rather than a maxilla. However, to make it a right dentary requires
not only that the smaller teeth be considered anterior, but necessi-
tates such a complex post-mortem juggling of parts that it seems

Fig. 3. Proterogyrinus scheelei Romer, C.M.N.H. 10950. Attempted restoration
of skull in dorsal view. Heavy lines indicate parts preserved in the specimen;
light lines, lateral portions of skull restored in embolomere pattern. Abbre-
viations as in fig. 2, X 3/4.

1970

A NEW ANTHRACOSAURIAN

better to keep to the assumption that we are dealing with a dis-
placed right maxilla.

Assuming normal proportions and arrangement of the dermal
roofing elements as found in other anthracosaurs, one can tenta-
tively restore the appearance of the skull as seen in dorsal view
(fig. 3) . Material is obviously insufficient for an attempt at a lateral
or ventral reconstruction.

POSTCRANIAL SKELETON

Axial skeleton : From the small block containing the remains of the
skull, disarticulated postcranial materials are sparsely spread over
two slabs extending for about 70 cm. Except for a partial Mega-
lichthys jaw near the far end of the slabs, all material visible is of
a sort which could have been derived, and presumably did derive,
from a single animal. All identifiable materials are appropriate to
a form with body and limb proportions similar to those of the em-
bolomere Archeria, and in various points the structure of limb and
girdle remains are comparable to those of typical embolomeres. De-
spite their disarticulated and scattered nature, the various preserved
fragments indicate that the individual had not completely “disinte-
grated” before burial; for example, the remains of the front leg are
close to the skull block, remains of the pelvic girdle are toward the
far end of the pair of slabs.

There are sparse scattered remains of the vertebral column.
Several neural arches, presumably from the cervical region, are
present on the skull slab. About a dozen arches from the trunk re-
gion can be seen; on the larger slabs most are crushed or incom-
plete, and surface detail is generally obscure. However, the general
structure can be made out (fig. 4 D,E) . It is of a normal anthra-
cosaurian type. The neural spine is moderately tall, thin from side
to side, and broad anteroposteriorly. Below, the arch expands ante-
riorly and posteriorly to the zygapophyses; further ventrally and
somewhat anteriorly the arch extends downward to, presumably,
afford tubercular attachment for the rib externally, and internally
shows a flat surface for “central” attachment.

Separated from the arches there are found some eight “central”
elements (fig. 5) . They are thin hoops of bone, forming the greater
part of a circle but incomplete at one point, presumably dorsally

ALFRED SHERWOOD ROMER

No. 10

Fig. 4. Proterogyrinus scheelei Romer, C.M.N.H. 10950. A, a caudal neural
arch, seen from the right side. B, a caudal intercentrum and haemal arch base,
seen from the side at the left, anteroventrally at the right. C, a rib; position
in column indeterminate. D, a dorsal neural arch in anterior view. E, a dorsal
neural arch in side view; central elements are restored in probable position,
X 3/2.

in the articulated position. The upper edges are bevelled, obviously
for neural arch articulation. Most of these central elements are
poorly preserved, but several, when seen in side view, have nearly
straight edges, with little indication of structural features except
for a slight indentation seen in two cases part way down the pre-
sumed anterior border. One element, however, is of a different
nature, and a second element appears to resemble it. Here, part
way down each side, there is a pronounced development of a semi-
circular area of articulation for a rib capitulum along the presumed

Fig. 5. Proterogyrinus scheelei Romer, C.M.N.H. 10950. “Central” elements of
the trunk. A, presumed intercentrum from the right side and posteriorly. B,
presumed pleurocentrum from the right side and posteriorly, X 3/2.

1970

A NEW ANTHRACOSAURIAN

posterior margin of the outer surface. It seems probable that we
have in these ring-shaped structures both pleurocentra and inter-
centra, those with the pronounced articular area being intercentra,
the others pleurocentra. I have ventured to restore a vertebra in
side view (fig. 4 E) . This restoration should, of course, be consid-
ered as tentative only, because of the sparsity and disarticulated
condition of the material. As seen in side view, the vertebra ap-
pears closely comparable to that of a typical embolomere. It must
be remembered, however, that the central elements are merely thin
shells, in strong contrast to the centra of such a typical embolomere
as Archeria and, further, that the “central” rings, as preserved, are
incomplete dorsally. It is not impossible that in a more mature
specimen of Proterogyrinus this dorsal gap might have been closed;
but it is highly improbable that, even so, the elements would have
been closely comparable to those of typical embolomeres, in which
ossification is as complete dorsally as around the rest of the circle
of the centrum. Of the caudal region I have been able to identify
a single neural arch, of relatively small size and with a slender
backwardly-slanting spine (fig. 4 A) . Adjacent to one of the ischia
there are badly preserved remains of the central elements of a
fraction of the tail region. There are here several intercentral ele-
ments, from which the haemal spines have broken off (fig. 4 B) .
The associated intercentra appear, as far as preserved, to be wedge-
shaped, as seen in side view, tapering to a point dorsally. A crushed
and poorly preserved element nearby appears to be a completely
circular structure. Possibly pleurocentral development may have
been more advanced in the caudal region than in the trunk.

In the neighborhood of the front limb are remains of two clus-
ters of ribs which presumably come from the anterior part of the
column. The rib heads are not visible. They are circular in section,
and show none of the flattening seen in various temnospondyls, and
there is no evidence of expansion of the shaft (except for a slight
distal expansion seen in one case) . A few further ribs are seen far-
ther posteriorly in the block; one is shown in figure 4 C.

There are a number of belly scales, poorly preserved, scattered
over the slabs.

Appendicular skeleton : There are no identifiable remains of the
shoulder girdle. Of the left pectoral limb there is only an imper-
fectly preserved humerus. Of the right leg, however, humerus,

ALFRED SHERWOOD ROMER

No. 10

radius and ulna are present close together in a semi-articulated
condition (fig. 6 A) .

The humerus, which measures 30 mm in overall
closely comparable to that of the embolomerous Archeria
build. The bone is less completely ossified than in most

Fig. 6. Proterogyrinus scheelei Romer, C.M.N.H. 10950. A, right humerus;
radius and ulna in position as found. The humerus is seen from the dorsal
surface. B, the same humerus in ventral view, X 1.

of that genus, so that the “unfinished” proximal surface extends
anteroventrally to include the region of the deltopectoral crest, and
distally the ectepicondyle, presumably projecting in an adult, is un-
ossified. As in Archeria, the entepicondyle is a very large sub-
quadrate structure, bearing, as is proper for anthracosaurians, an
entepicondylar foramen near its proximal inner corner. As in em-
bolomeres, a highly developed flange of bone extends from the re-
gion of the deltopectoral crest distally, without interruption, along
the anterior edge of the bone, to the ectepicondylar region.

Radius and ulna are seen from the dorsal (extensor) aspect;
the former is 18 mm in length, the latter 20 mm. It is obvious that
ossification was far from complete, for in the ulna there is no
olecranon and not even any trace of the articular surface for the
humerus.

Situated some 60 mm from the major limb bones is a series of
disarticulated foot elements which are not improbably part of the
right front foot (fig. 7) . Presumably, the four stouter elements are
metacarpals.

length, is
in general
specimens

1970

A NEW ANTHR AC O SAURIAN

Fig. 7. Proterogyrinus scheelei Romer, C.M.N.H. 10950. Scattered foot bones
found near right leg elements, X 1.

Of the pelvic girdle (fig. 8) the left ilium is present, and seen
from the inner surface, and there are both ischia, the right seen
from the inner surface, the left from the outer side. The greatest
length of the ilium, from the pubic articulation to the tip of the
posterior prong is 54 mm. The right and left ischia are, respectively,
33 and 32 mm in greatest length. The ilium is closely comparable

Fig. 8. Proterogyrinus scheelei Romer, C.M.N.H. 10950. Left ilium and ischium,
seen from inner surface. The ischium is that of the right side, reversed, X 1.

to that of Archeria. There obviously was a dorsal blade, for sacral
connection, but this is broken off. The posterior prong is as elongate
as that of Archeria. Since the element is seen from the inner side,
nothing can be said of acetabular structure. The base of the bone
presents a much thickened articular surface for pubis and ischium,
the pubic area being especially thick. Anteriorly there is a well

ALFRED SHERWOOD ROMER

No. 10

developed triangular buttress leading down to the region of pubic
articulation. This area faces nearly directly anteriorly, in contrast
to its somewhat more medial slant in Archeria.

The ischium, again, is comparable in proportions and build to
that of Archeria. The outer surface shows a gentle indentation for
the acetabular border. As in early tetrapods generally the upper
margin of the posterior projection of the bone is somewhat thick-
ened. The area of the ischiadic symphysis is striated in a fashion
seen in various early tetrapods.

No pubis is present in the material. Whether this is due to
chance, or whether (particularly taking into account the im-
maturity of the specimen) the pubes were unossified, is, of course,
uncertain.

Except for three phalanges or metapodials no identifiable re-
mains of the hind leg are present.

DISCUSSION

Despite the fact that the remains are incomplete, the nature of
Proterogyrinus seems clear as regards most features. The form is
clearly an anthracosaurian; further, in most regards it is in close
agreement with the Embolomeri of the Pennsylvanian and early
Permian. Such portions of the skull as are preserved show a close
approach to the structure seen in such representative Upper Car-
boniferous forms as Pteroplax, Eogyrinus, and Palaeogyrinus of the
English Coal Measures (Watson, 1926; Panchen, 1964) , Neoptero-
plax of the American Pennsylvanian (Romer, 1963) , and, except in
a lesser elongation of the snout, Archeria of the early Permian.1
Even in the nature of the marginal teeth — small, numerous, closely
crowded and with “chisel-like” tips — there is a clear comparison
with embolomeres far removed in time, such as Archeria of the
Permian. The humerus (except in the fact that ossification is less
advanced) is closely comparable to that of such an embolomere as
Archeria. The pelvis, too, is of a nature comparable to that of
known embolomeres.

1 Figures of the roof of the Archeria skull (as “ Cricotus ”) have been given by
Cope (1884), by Cope and Matthew (1915), and by Broom (1913). I hope to
give a more complete description of Archeria cranial anatomy in the near
future.

1970

A NEW ANTHRACOSAURIAN

Were it not for the structure of the central region of the ver-
tebrae, Proterogyrinus could well be considered a proper member
of the Embolomeri, hardly to be distinguished on a family basis
from various other members of that group. The central elements,
however, give one pause. All previously known embolomeres in
which vertebral material is present show both intercentra and
pleurocentra as complete rings, as well ossified dorsally as laterally
and ventrally, and with a thickness of ossification that reduces the
opening for the notochord to a fairly modest diameter. In the pres-
ent specimen the walls are relatively thin; most especially, both
intercentrum and centrum are incomplete dorsally, with a broad
gap in the region which in life lay below the neural canal. It must
be kept in mind that the present specimen is rather certainly im-
mature, and hence in an older specimen there may have been some
degree of ossification in this area, in which, presumably, cartilage
was already present. It is, however, rather certain that maturity
would not have brought this area to the highly ossified condition
seen in typical Embolomeri. This condition of the vertebrae has
led me to erect for this form the new family Proterogyrinidae.

It is possible that the Proterogyrinidae represents a generalized
type of anthracosaur, rather than a primitive group of embolomeres
or embolomere ancestors. But because of the similarity in most
features the Proterogyrinidae may, provisionally at least, be includ-
ed in the Embolomeri.

The finding of this form, however, suggests reconsideration of
generally accepted ideas of vertebral evolution in labyrinthodonts.
A first major attempt at sorting out the membership of this group
was that of Watson in his classic papers on the origin and evolution
of the Amphibia (1919, 1926), in which he distinguished between
the Rhachitomi, the Stereospondyli descended from them, and the
Embolomeri. He considered the embolomeres to be the basal group
of the entire Labyrinthodontia. In later years it became apparent
that the story was somewhat different, and I proposed (Romer,
1947) that, leaving out of consideration the Devonian ichthyoste-
gids, all labyrinthodonts could be divided into two major groups,
Temnospondyli (including Rhachitomi and Stereospondyli) and
Anthracosauria (including the Embolomeri and forms leading to-
ward and to the Reptilia) . This proposal assumed that in ancestral
forms there was a large single inter centrum, centered ventrally, and

ALFRED SHERWOOD ROMER

No. 10

small paired pleurocentra, dorsolateral in position; that in the tem-
nospondyls the pleurocentra remained small, and disappeared in
stereospondyls, while in the Anthracosauria the pleurocentra en-
larged to form the major central structure;1 the embolomeres, as
a side branch of the anthracosaurs, formed a variant in which inter-
centrum as well as pleurocentrum grew to form a second complete
ring.

Since the publication of this thesis, nearly all more recent dis-
coveries have tended to support it, although some variants in the
temnospondyl pattern have been discovered, such as the peculiar
plagiosaur group (Panchen, 1959) and Doleserpeton (Bolt, 1969) .
The nature of the centra in ichthyostegids (Jarvik, 1952) demon-
strated the high antiquity of the temnospondyl pattern. My belief
that the rhachitomous type was present at an early date has been
confirmed by the finding by Baird (1957) that the loxommids,
which had appeared before the close of the Mississippian, were
rhachitomes, and by the discovery of typical rhachitomes in the
Mississippian Greer quarry, including not only the specimen I have
already described (Romer, 1969) but also others awaiting de-
scription.

In 1964 I elaborated further on a probable evolutionary pattern
among anthracosaurs. I assumed that, beginning with the temno-
spondyl type of centrum, the pleurocentra grew downward, first in
the form of two half rings,2 and then consolidated into a complete
ring, while the inter centra remained ventral wedges. I termed this
the diplomerous condition. At this stage of phylogenetic develop-
ment, I believed, there occurred a dichotomy, the embolomeres
splitting off and developing the intercentra as well as pleurocentra

1 I consider here only the history of ossifications; as Panchen has pointed out
(1963), either intercentrum or pleurocentrum, if it tends to take over the en-
tire “central” area, includes in its substance the entire skeletogenous material
of the segment, earlier split between the two types of element.

2 A situation seen in Pholidog aster, which I interpreted as a primitive anthra-
cosaur. Panchen (in litteris, and cf. Carroll, 1969) has disputed this, suggest-
ing that Pholidogaster is a temnospondyl. The skull of the Pholidogaster type
is too crushed to be interpretable, and Panchen believes that the anthracosaur
skull in the Edinburgh collections which Watson and I thought to belong to
this genus can not so be assigned. My belief in the anthracosaurian nature of
this Lower Carboniferous form, however, was based mainly on the nature of
the pleurocentra, which form two half-rings reaching the ventral surface of
the column. Just this type of structure is seen in the immature specimens of
the anthracosaur Discosauriscus (Spinar, 1953); temnospondyl pleurocentra are
typically confined to a dorsal position, without ventral expansion.

1970

A NEW ANTHRACOSAURIAN

into complete rings, while in the “main line” tending toward and to
the reptilian condition (Seymouriamorpha in a broad sense) , the
intercentra failed to develop further and tended to be reduced.

Despite the restricted amount of material and its disassociated
condition, the apparent nature of the Proterogyrinus column indi-
cates that the story of anthracosaurian vertebral evolution needs
reconsideration. In nearly every feature that can be made out in
this new form, we are dealing with an animal closely allied to the
embolomeres and perhaps to be included in that group in a broad
sense. But, in conflict with my earlier beliefs, the pleurocentrum
is not yet a complete ring, while the intercentrum is in an equally
advanced condition. This suggests that the embolomeres split off
from the anthracosaur “main line” at an earlier stage than I had
thought was the case. Very probably further Mississippian dis-
coveries at Greer and elsewhere will show that anthracosaur evo-
lution was far more complex than I had assumed.

ALFRED SHERWOOD ROMER

No. 10

REFERENCES CITED

Baird, D., 1957, Rhachitomous vertebrae in the loxommid amphibian Megalo-
cephalus: Geol. Soc. Amer. Bull., v. 68, p. 1698.

Bolt, J. R., 1969, Lissamphibian origins: possible protolissamphibian from the
Lower Permian of Oklahoma: Science, v. 166, p. 888-891.

Broom, R., 1913, Studies on the Permian temnospondylous stegocephalians of
North America: Amer. Mus. Nat. Hist. Bull., v. 32, p. 563-595.

Carroll, R., 1969, Problems of the origin of reptiles: Biol. Rev., v. 44, p. 393-432.

Cope, E. D., 1884, The Batrachia of the Permian period of North America:
Amer. Nat., v. 18, p. 26-39.

Cope, E. D. and W. D. Matthew, 1915, Hitherto unpublished plates of Tertiary
Mammalia and Permian Vertebrata: Amer. Mus. Nat. Hist., Monogr. ser.,
no. 2.

Jarvik, E., 1952, On the fish-like tail in the ichthyostegid stegocephalians:
Meddelelser om Grpnland, v. 114, p. 1-90.

Panchen, A. L., 1959, A new armoured amphibian from the Upper Permian of
East Africa: Roy. Soc. [London] Philos. Trans., ser. B, v. 242, p. 207-281.

1963, The homologies of the labyrinthodont centrum: Internat.

Cong. Zoology, 16th, Washington, D.C., 1963, Proc., v. 1, p. 161.

1964, The cranial anatomy of two Coal Measure anthracosaurs:

Roy. Soc. [London] Philos. Trans., ser. B, v. 247, p. 593-637.

Romer, A, S., 1947, Review of the Labyrinthodontia: Mus. Comp. Zool. Bull.,
v. 99, p. 1-368.

1963, The larger embolomerous amphibians of the American Car-
boniferous: Mus. Comp. Zool. Bull., v. 128, p. 415-454.

1964, Problems in early amphibian history: Jour. Animal Morph.

Physiol., v. 2, p. 1-20.

1969, A temnospondylous labyrinthodont from the Lower Carbon-
iferous: Kirtlandia, no. 6, p. 1-20.

Spinar, Z., 1953, Revision of some Moravian Discosauriscidae: Roz. Ustred.
Ustav. Geol., v. 15, p. 1-159.

Watson, D. M. S., 1919, The structure, evolution and origin of the Amphibia —
the “orders” Rhachitomi and Stereospondyli: Roy. Soc. [London] Philos.
Trans., ser. B, v. 209, p. 1-73.

1926, The evolution and origin of the Amphibia: Roy. Soc. [Lon-
don] Philos. Trans., ser. B, v. 214, p. 189-257.

MANUSCRIPT RECEIVED MARCH 4, 1970

!<U

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO MAY 28, 1970 NUMBER 11

OBSERVATIONS ON THE PENNSYLVANIAN CRINOID
ENDELOCRINUS ARMATURA (STRIMPLE)

J. J. BURKE
ABSTRACT

Originally described as a species of Delocrinus, the Middle
Pennsylvanian crinoid Endelocrinus armatura (Strimple) shows
late attainment of biserial arm structure and has pits at the cor-
ners of the dorsal cup plates, both characteristic features of En-
delocrinus. Unique ornamentation, predominately nodose at ma-
turity, is traceable from pit-node structure in juveniles. Discovery
of new specimens extends the geographic range of the species from
Oklahoma to Ohio and Illinois.

Most species of Pennsylvanian inadunate crinoids are known
only from dorsal cups; the arms are seldom preserved, either as
a whole or in part. Consequently I have been fortunate in having
at hand for the present study two specimens of a species of Endelo-
crinus in which most of the arms are preserved. This crinoid is
characterized by unusual ornamentation, and in addition, one speci-
men represents a young, the other a mature stage of growth, per-
mitting us to trace the ontogeny of the arms in at least one species
of Endelocrinus.

For the opportunity to study these two specimens I am indebted
to Dr. Richard D. Hoare of Bowling Green State University, Mr.
James L. Murphy of Case Western Reserve University, and Dr.
Porter Kier of the United States National Museum. I also wish to
thank Dr. Kier for making the photographs from which the illustra-
tions for this paper were taken.

( 4UL * 5 ®70 J

J. J. BURKE

No. 11

SYSTEMATIC PALEONTOLOGY

Family ERISOCRINIDAE Miller, 1889
Genus ENDELOCRINUS Moore and Plummer, 1940
ENDELOCRINUS ARMATURA (Strimple), 1949
Plate 1, figs. 1-6

Delocrinus armatura Strimple, 1949, Paleontographica Americana,
v. 3, pt. 1, p. 18, pi. 3, figs. 3, 6.

Diagnosis: A large species of Endelocrinus (diameter of dorsal cup
at maturity exceeding 19 mm) ; form ratio 0.37 to about 0.41. Orna-
mentation distinctive, with crown plates characterized by pits and
nodes; pits dominant over nodes at juvenile stage, nodes predomi-
nant and granulose ornamentation characteristic of mature stage;
initiation of biseriality on third to fifth secundibrachs at cup width
of about 19 mm, with “normal biseriality” attained on seventh or
eighth secundibrachs. Primibrachs without spines.

Occurrence: Middle Pennsylvanian (Oklahoma, Ohio, and Illinois) .

The holotype of this species was derived from the Pumpkin
Creek Limestone, Dornick Hills Group, Pennsylvanian, in Love
County, Oklahoma.

In the original description of the holotype, U.S.N.M. S4689,
Strimple did not note the characteristic Endelocrinus pits at the
angles of the plates of the dorsal cup and attributed this species to
Delocrinus. At the same time he failed to indicate the unique orna-
ment of the cup, stating only that the ornamentation “consists of
irregular swollen tubercles which are more pronounced on the RR
than the BB.”

Of the two specimens which I am attributing to Endelocrinus
armatura, the one which approaches nearest to the holotype in size
and ornament is in the collection of the United States National Mu-
seum and was collected by Mr. James L. Murphy. It consists of a
dorsal cup with the greater part of the arms and a portion of the
stem attached. This specimen, U.S.N.M. 166578, was taken from the
Putnam Hill Limestone of the Allegheny Group, Pennsylvanian, in
the NE^SW1/^ sec. 13, Springfield Twp., Muskingum County, Ohio.
The Putnam Hill Limestone crops out about 200 feet west of U.S.
Route 22 on a small knoll west of a shopping center designated on
the Zanesville West 7.5' Quadrangle.

1970

PENNSYLVANIAN CRINOID

Because Strimple’s holotype has suffered damage from weath-
ering, breakage and dislocation of plates, it is difficult to make
strictly comparable measurements of the two specimens. Measure-
ments of height and width of U.S.N.M. S4689 are at best approxi-
mate, and I believe that the form ratio is closer to 0.41 than Strim-
ple’s finding of 0.45. The Ohio specimen is a larger and presumably
older individual, with a form ratio of 0.37. Breakage and distortion
may account to some extent for the apparent differences in width
and depth of the basal impressions of the two specimens, but the
holotype actually appears to have a wider and shallower concavity.
The proportions of the radials and basals of the two dorsal cups are
in close accord, and the cups are also similar in the outward flare
of their radials and in showing channels along the interradial su-
tures. Despite the variations noted, the remarkable agreement in
peculiar ornamentation argues strongly for the specimens being
conspecific.

As Strimple noted, the most prominent feature of the orna-
mentation of the cup is large nodes or tubercles. These are irregu-
larly distributed for the most part, although they tend toward a
festoon-like arrangement below the forefacets of the radials. The
nodes are more distinct on the radials, but they are found on the
basals also. However, the most striking ornamental feature of these
dorsal cups consists in angular depressions associated with the nodes
and connected by a network of slender and shallow canals. The
depressions are most numerous on the basals, but they also occur
along with the canals, on the radials. In the Ohio specimen, the
proximal region of the basal circlet is more extensively ornamented.
A single large, less angular depression occupies most of the surface
of the portion of anal X that rises above the summit of the cup, and
in U.S.N.M. 166578, the Ohio specimen, a similar depression shows
on the overlying tube plate. In this specimen there also appear to
be traces of incipient granular structure on portions of the cup, but
none is showing within the basal impression.

Unfortunately, the holotype of Endelocrinus armatura consists
of no more than the dorsal cup with a portion of the stem attached.
The Ohio specimen, U.S.N.M. 166578, preserves, in contrast, a con-
siderable portion of the arms. A glance at the arms leaves no doubt
that this is a very young individual. The height of the axillary first

J. J. BURKE

No. 11

primibrachs and the markedly high and quadrangular proximal
secundibrachs are evidence of this.

The primibrachs of this young specimen display the same orna-
mentation as the dorsal cup, with variations. The depressions are
more rounded and predominate over the nodes in diameter. There
are one or two crater-like impressions on each flank of a primi-
brach. Canals are also present. The primibrachs lack spines. Near
the distal tip of the primibrach of the A ray there is a depression
from the floor of which a small node originates. The nodes are
prominent and little trace of the depressions remains in the primi-
brachs of rays C, D, and E. The primibrach of the B ray is worn
at this place, but appears to have borne a small node in the de-
pression.

For the most part, the secundibrachs remain quadrangular up
to the sixth or seventh, with long and short sides alternating. Dis-
tally they become triangular or cuneate, and the initiation of the
biserial arrangement is apparent on the eighth or ninth, where the
cuneate plate fails to extend across the full width of the arm. None
of the arms are complete; there appear to be 13 secundibrachs on
one of them.

A few of the articular surfaces of the secundibrachs are show-
ing, but some of their details are obscure. The outer ligament area
is faintly denticulate and the outer ligament pit slitlike. The trans-
verse ridge is indistinct. The intermuscular notch is V-shaped and,
together with the intermuscular furrow, divides the two large flexor
muscle areas.

The sides of the secundibrachs show shallow hollows, indicat-
ing that even at this early stage some interlocking of the arms was
possible.

EXPLANATION OF PLATE 1
(All figures X 2)

Fig. 1. Endelocrinus armatura (Strimple), holotype, U.S.N.M. S4689, from
the Pumpkin Creek Limestone, Dornick Hills Group, Love County, Oklahoma.
a, basal view and b, posterior view of the dorsal cup.

Fig. 2. Endelocrinus armatura (Strimple), hypotype, U.S.N.M. 166578, from
the Putnam Hill Limestone, Allegheny Group, Muskingum County, Ohio.
a, basal view and b, posterior view of the crown.

Fig. 3. Endelocrinus armatura (Strimple), hypotype, B.G.S.U. 2540, from
the shale above the No. 6 Coal, Kewanee Group, St. Clair County, Illinois.
a, basal view and b, posterior view of the crown.

KIRTLANDIA NO. 11

PLATE 1

J. J. BURKE

No. 11

The specimen preserves a few stout pinnulars, but no complete
pinnules.

Depressions and nodes also characterize the ornamentation of
the secundibrachs, but only a few of these plates show canals. The
first secundibrach bears more prominent nodes than those distal
to it, but for the most part lacks the order in arrangement of de-
pressions and nodes that exists in the secundibrachs that follow.
This, in its simplest form, consists of a diamond-shaped depression
shared by pairs of secundibrachs along the length of the arm, ex-
tending from underlying to overlying plates, with its lateral angles
at their sutural junctions. The vertical angles of the depressions
are usually marked by nodes or incipient nodes. There are varia-
tions of this pattern, but as the secundibrachs approach or attain
the cuneate stage, the plates usually bear a single prominent node
nearly marking the midline of the arm.

Here and there on the arms there appear to be traces of granu-
lar ornamentation, but it is not well defined.

The second specimen which I am referring to Endelocrinus
armatura was collected from the shale over the No. 6 Coal of the
Carbondale Formation, Kewanee Group, Pennsylvanian, at the
Midwest Coal Company mine near Millstadt, St. Clair County,
Illinois. It was found by Miss Dorothy Lalonde, a student of Dr.
Richard D. Hoare of Bowling Green State University, who in turn
submitted the crinoid to me for study.

This specimen (B.G.S.U. 2540) is part of a crown with portions
of seven arms attached. Part of the stem is also preserved. On the
anterior side the arm and cup plates are dislocated and mashed. An
estimated width of about 19.5 mm for the dorsal cup indicates that
we are dealing with a large species of Endelocrinus. The form ratio
was probably about 0.40.

The basal impression is relatively deeper than that of the holo-
type of Endelocrinus armatura, but apparently less so than that of
the juvenile specimen from Ohio. The Illinois specimen is also
closer in accord with the holotype in showing less pit and node
ornamentation in the proximal portion of the basal circlet; the pits
and nodes are most evident in areas adjoining the distal reaches of
the interbasal sutures, although they are also present on other parts
of the distal extensions of the basals.

In this mature specimen the nodes predominate over the pits.

1970

PENNSYLVANIAN CRINOID

Growth of the nodes has eliminated the canals, and modified the
surface to such an extent that only one of the characteristic Endelo-
crinus pits, that of the DE interray, is still evident. Nodes occupy
the areas that show as pits in anal X and the overlying tube plate
of the Ohio specimen, although the outlines of the pits are still ap-
parent. The entire cup bears coarsely granulose ornament, which
extends to the proximal region of the basals.

The primibrachs, in keeping with the age of this large speci-
men, are relatively of less height and greater width than in the
juvenile specimen from Ohio. What remain of the craterlike de-
pressions shown on the primibrachs of that juvenile individual ap-
pear only as irregular borders separating a lumpy mass of nodes
from the relatively smooth lateral flanks of the plates. The most
distinct of these nodes are those at tips of the primibrachs, which
are relatively much larger than those characterizing the juvenile
specimen and much more irregular in outline. No traces of the
canals remain on the primibrachs.

The secundibrachs, for the most part, remain quadrangular up
to and including the third or fourth, although in one arm only the
first two secundibrachs are quadrangular. Triangular (cuneate)
plates succeed the quadrangular secundibrachs, and also mark the
initiation of biseriality, for these plates fail to extend the full width
of the arm. However, “normal biseriality,” as interpreted by Gra-
bau (1903, p. 290) apparently was not attained before the seventh
or eighth secundibrach. The arms are not complete; there are 18
secundibrachs present in each of the two with the most length pre-
served, and these arms must have been at least a third longer
originally.

Some of the biserial secundibrachs show the articular surfaces,
which are fairly typical for Endelocrinus. There are indications of
the external ligament area, transverse ridge, intermuscular notch
and intermuscular furrow. The principal flexor muscle scars are
fairly evident, but the minor scars are rather obscure.

The interlocking structures of the arms are traceable not only
on the sides of the arms, but also in external view, and the concave
borders of one arm are seen to mesh with the convex borders of
the adjacent arm with remarkable precision.

Along the midline of the arm each secundibrach bears a single
prominent node — the feature noted in the juvenile specimen at the

J. J. BURKE

No. 11

cuneate stage. These nodes occupy what were the diamond-shaped
depressions in the secundibrachs of the young crinoid. However,
although the nodes on the secundibrachs, as well as on other parts
of the crown of the more mature specimen have supplanted the pits
and other depressions, it is still possible to find indications of the
former structure in these secundibrachs.

Coarse granulose ornamentation, such as that found on the
dorsal cup, also characterizes the arms of the Illinois specimen.

Linear measurements, in millimeters, of the holotype of Endelo-
crinus armatura and the two specimens which I am referring to the
species are as follows:

Height dorsal cup
Width dorsal cup
H/W

Width basal concavity
Height basal concavity
Width stem
Length basal
Width basal
Length radial
Width radial

Length suture between BB
Length suture between RR
Length anal X
Width anal X
Length first tube plate
Width first tube plate
First primibrachs:

A, length

A, width

B, length

B, width

C, length

C, width

D, length

D, width

E, length
E, width

*Estimated
* * Approximate

Holotype

Hypotype

U.S.N.M.

B.G.S.U.

S4689

166578

2540

4.6*

4.6

8.2*

11.1*

12.3

19.5*

0.41*

0.37

0.40*

5.7*

5.0

8.4

1.8+

2.3+

1.8

1.9

2.9

3.6**

4.0

6.5

3.9

4.2

7.7

3.6

4.0

6.0

5.9

6.5

8.6

2.7**

2.7

4.7

2.3

2.4

3.4

2.2

3.6

3.9

2.0

2.7

3.6

—

1.4

2.1

—

1.7

2.4

4.9

5.7**

—

5.8

8.5**

—

3.8

4.8**

—

5.9

7.9

—

4.9

5.3

—

5.7

8.0

—

5.0

5.8

—

5.5

8.2

—

4.0

4.7

—

5.8

8.4

1970

PENNSYLVANIAN CRINOID

DISCUSSION

The unique major ornament of the three crinoid specimens
treated here, coupled with the fact that it can be traced step by
step from juvenile to mature stages of growth, constitutes the
strongest argument for regarding the three specimens as con-
specific. Greater size and presence of definite granular ornamen-
tation characterize the mature individual, B.G.S.U. 2540, and there
appear to be variations in depth and width of the basal impression
among the three specimens involved, but I have given these charac-
ters full consideration and fail to find them grounds for specific or
subspecific distinction.

In a previous paper (Burke, 1967) I suggested that fully bi-
serial arm structure may have been attained late in life in various
species of Endelocrinus. Study of these specimens of Endelocrinus
armatura and additional undescribed material representative of En-
delocrinus further supports this suggestion. However, biseriality in
the arms of B.G.S.U. 2540 has progressed at least to the extent that
I have observed the condition in certain presumed mature speci-
mens of Delocrinus. Nevertheless, biserial arm structure must have
been realized at a very early stage of growth in most species of
Delocrinus — possibly even before the arm segments were sufficient-
ly developed to remain associated after the death of the individual,
otherwise crowns showing earlier stages in attainment of biseriality
would be known.

There is, however, one species, either a Delocrinus or a form
closely related to that genus, which may compare with Endelo-
crinus in delayed acquirement of biseriality. Strimple and Knapp
(1966, pi. 36, figs. 1, 2) have illustrated the crown of a specimen
which they attribute to “ Diphuicrinus ” croneisi Moore and Plum-
mer, from which they conclude, because the arms are uniserial, that
“Diphuicrinus” is distinct from Delocrinus. However, the distal
secundibrachs of the crown appear to be cuneate, which strongly
suggests that the individual might have eventually developed bi-
serial arm structure. Furthermore, if Strimple and Knapp are cor-
rect in their identifications, this is presumably a young individual;
the diameter of the dorsal cup is given as 21 mm, that of the holo-
type of the species as 24 mm; cup diameters of two other specimens,
however, are noted as 31 and 43 mm respectively (ibid., p. 313) . If
the latter two specimens actually pertain to the species, there would

J. J. BURKE

No. 11

appear to be the possibility that in attaining such size “Diphui-
crinus ” croneisi might also have acquired some degree of biserial
arm structure.

REFERENCES CITED

Burke, J. J., 1967, A new Endelocrinus from the Brush Creek Limestone
(Pennsylvanian) of Pennsylvania: Carnegie Mus. Ann., v. 39, p. 75-83.

Grabau, A. W., 1903, Notes on the development of the biserial arm in certain
crinoids: Am. Jour. Sci., Ser. 4, v. 16, p. 289-300.

Strimple, H. L., and Knapp, W. D., 1966, Lower Pennsylvanian fauna from east-
ern Kentucky; Part 2, Crinoids: Jour. Paleontology, v. 40, p. 309-314.

MANUSCRIPT RECEIVED APRIL 6, 1970

KC/

SlRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO

SEPTEMBER 30, 1970

NUMBER 12

MAUCHCHUNKIA BASSA, GEN. ET SP. NOV.,

AN ANTHRACOSAUR (AMPHIBIA, LABYRINTHODONTIA)
FROM THE UPPER MISSISSIPPIAN

NICHOLAS HOTTON III

Smithsonian Institution

ABSTRACT

The skull, articulated presacral vertebral column, and several
limb elements of an anthracosaurian amphibian from the lowest
beds of the Mauch Chunk Group (Upper Mississippian) at Greer,

West Virginia, are described as Mauchchunkia bassa, gen. et sp.
nov., and made the type of a new family, the Mauchchunkiidae.

In most respects the type resembles the Embolomeri, but the pre-
sacral column is short, the intercentrum is a ventrally placed cres-
cent, the limbs are stout, and the snout is not elongate. Mauch-
chunkia appears to be the most generalized anthracosaur yet de-
scribed, and in many of its features it supports the view that the
primary adaptation of primitive tetrapods was toward a terrestrial
environment. Vertebral structure foreshadows reptilian conditions,
and the Mauchchunkiidae are proposed as early ancestors of all
reptiliomorph tetrapods.

INTRODUCTION

The specimen at hand, catalogue number 22573 in the National
Museum of Natural History (pi. 1) , is the third reasonably com-
plete fossil tetrapod to be described from the Upper Mississippian
deposits at Greer, Monongalia County, West Virginia. At the pres-
ent time the fossils from Greer are, except for Ichthyostega of the
latest Devonian (Save-Soderbergh, 1932), the oldest tetrapods of
which we have detailed information, and provide almost the only
record between Ichthyostega and the much better known tetrapods
of the Pennsylvanian (Panchen and Walker, 1961; Romer, 1969) .

The first tetrapod remains from Greer were evidently discov-
ered by an amateur, Mr. L. R. Collins, in 1948, whose find was fol-
lowed up successfully by Mr. John J. Burke and Mr. William E.
Moran. My attention was first drawn to the Greer locality by Mr.
Moran in 1960, and NMNH 22573 was collected during a trip that

KIRTLANDIA NO. 12

PLATE 1

NICHOLAS HOTTON III

1970

MAUCHCHUNKIA

Mr, Moran and I made to Greer in November of that year. It is
with pleasure that I acknowledge my debt to Messrs. Moran and
Burke for their generous cooperation, not only in showing me the
locality but also in educating me to the potential of late Paleozoic
deposits of West Virginia. Thanks are also due to officials of the
Greer Limestone Company for their friendly cooperation in permit-
ting access to the quarry, and to Dr. Alec Panchen of the Univer-
sity, Newcastle-upon-Tyne, to Dr. Robert Carroll of McGill Univer-
sity, and to Professor A. S. Romer for their open-handedness in
providing access to unpublished material.

PROVENANCE

The Greer locality is in a quarry operated by the Greer Lime-
stone Company in the valley of Deckers Creek, Monongalia County,
West Virginia, about 6.5 miles southeast of Morgantown. The
quarry lies just north of State Route 7, on the west side of the con-
fluence between a small tributary valley and Deckers Creek valley.
The actual spot in which the bones were found is about 0.5 miles
north of the highway along the west side of the tributary valley.
NMNH 22573 was found in place in dark greenish-gray shales over-
lying the massive limestone that is being worked commercially at
Greer.

Romer (1969) states that the rock being quarried commercially
is recognized by the West Virginia Geological Survey as the Union
Limestone of the Greenbrier Group (see also Weller, et al., 1948) .
The uppermost beds of the Greenbrier, the Cypress Sandstone and
Alderson Limestone, are missing at Greer (Tilton, 1928) , so that the
Union is directly overlain by the greenish and reddish shales and
blue to gray limestones that belong to the Bluefield Formation, the
lower subdivision of the M[auch Chunk Group.

The lowest 30 feet of the Bluefield are readily identified, on the
basis of Tilton’s detailed description, at the spot from which NMNH
22573 was collected (pi. 2) . In the center of the picture, the boy
is standing in the excavation left by removal of the specimen, about
6 feet above the bottom of the Bickett Shale. The Bickett (Bi) is
about 13 feet thick at Greer; its bottom is just below the top of the
vegetation in the foreground, and its top is about 1 foot below the
lower ends of the crossed tree trunks at the upper left of the pic-
ture. It is underlain by the Glenray Limestone (Gl) , the massive
layer near the lower right of plate 2. The Glenray is 7 to 10 feet
thick in various parts of the quarry; its base is buried at this spot.
Below the Glenray lies 6 feet of Lillydale Shale, covered by rubble
in the foreground but recognizable close by. The Lillydale lies di-
rectly upon the Union at Greer.

KIRTLANDIA NO. 12

PLATE 2

NICHOLAS HOTTON III

Bluefield Formation, Mauch Chunk Group, exposed at Greer, West Virginia.
Gl, Glenray Limestone; Bi, Bickett Shale; Re, Reynolds Limestone.

1970

MAUCHCHUNKIA

Above the Bickett, marked by the dark band below the crossed
tree trunks in plate 2, is a limy layer containing abundant brachio-
pods, some of which were identified by G. Arthur Cooper as An-
thracospirifer, Orthotetes, and Diaphragmus. This is evidently the
bottom of the Reynolds Limestone (Re) , identified as “Orthotetes
Zone” by Tilton (1928) .

Most of the vertebrate remains collected thus far evidently
come from three or four feet below the level of NMNH 22573. Fish
remains are generally fragmentary; partly articulated material is
restricted to lungfish (D. H. Dunkle, written communication, 1969)
and tetrapods. A few carbonized plant fragments are found at these
levels, and clay pebbles are common in the matrix immediately sur-
rounding articulated vertebrate elements. No ripple marks have yet
been noted in the Bickett at Greer. Invertebrate remains are rare
and fragmentary at the vertebrate-bearing levels, and no marine
forms have yet been identified with certainty. The Bickett becomes
more limy toward the top and marine vertebrates appear in it;
change from vertebrate levels to the overlying limestone is thus
gradational.

Clay pebbles and fragmentary fish remains indicate that the
environment in which the Greer tetrapods occur was one of flowing
water, and it is probable that all of the material suffered some
transportation before burial. However, the association of elements
of single individuals, and the articulation of NMNH 22573, suggest
that these specimens were not transported far. All of the articu-
lated or associated remains represent terrestrial, aquatic forms,
which indicates that the portion of the Bickett Shale that contains
them is an atypical facies of the normally marine Mauch Chunk
Group. At Greer the middle part of the Bickett evidently repre-
sents a local and momentary phase of terrestrial deposition, prob-
ably a consequence of the formation of a temporary bar in shallow
marine waters and not of any change in tectonic activity. It was
terminated gradually as continuing tectonic subsidence brought
about the return of more nearly normal marine conditions at the
site.

NICHOLAS HOTTON III

No. 12

SYSTEMATIC PALEONTOLOGY

Class AMPHIBIA
Order ANTHRACOSAURIA
Family MAUCHCHUNKIIDAE fam. nov.

Diagnosis: Anthracosaurian labyrinthodonts that resemble embolo-
meres in pattern and sculpture of dermal skull bones, and general
structure of skull, neural arches, and limbs. Differ from embolo-
meres and Proterogyrinidae (Romer, 1970) in that intercentrum is
crescentic in shape and unossified dorsally, attaining to not more
than half the height of the fully ossified pleurocentral disc. Differ
from embolomeres in that presacral column is short (not more than
28 segments) , limbs stout, snout not elongate. Supratemporal bone
contributes significantly to anterodorsal margin of otic notch. Type
genus Mauchchunkia.

Genus Mauchchunkia1 gen. nov.

Mauchchunkia bassa2 sp. nov.

PI. 1; Figs. 1-14

Diagnosis for genus and species: An anthracosaur of moderate size
in which pre- and postorbital moieties of the narrow skull are sub-
equal in length. Postparietal bone more than half the length of
parietal. Anterior maxillary teeth appreciably higher crowned than
posterior; diminution of crown height from front to rear gradual.
Approximately 26 marginal teeth. Palatal dentition closely similar
to that of the embolomere Eogyrinus attheyi as restored by Panchen
(written communication, 1969) , with two tusk-and-pit pairs on pala-
tine and a single smaller pair on ectopterygoid, followed by four
smaller teeth comparable in size to marginals. Pineal opening sub-
oval, rimmed; slight ridge formed along interparietal suture as in
the embolomere Pteroplax cornuta.

Holotype: National Museum of Natural History 22573. Nearly
complete skull in which dermal elements are partially disarticu-
lated and broken. Braincase badly damaged, partially hidden.
Lower jaw nearly complete but broken. Twenty-seven vertebrae,
including atlas-axis complex, articulated but with neural arches
displaced and broken. Dermal shoulder girdle in approximately
correct relationship to column but smashed; large fragments of bone
associated with it pertain to scapulocoracoid but are too poorly pre-

1 The generic name is derived from the stratigraphic occurrence.

2 Specific designation refers to the fact that the holotype occurs near the
bottom of the Mauch Chunk Group.

1970

MAUCHCHUNKIA

served to permit more than tentative analysis. Heads of right and
left humeri, and abundant fragments of ribs and dermal armor also
present.

Referred specimens: National Museum of Natural History 26368.
Left and right humeri, minus heads, left and right radii, left ulna,
four metapodials and four phalanges more closely associated with
left limb elements than with right, one metapodial associated with
right limb.

National Museum of Natural History 26369. Fragments of
three neural arches and right ilium and pubis, and proximal ends
of left and right femora.

Horizon and locality: Six feet above the bottom of the Bickett
Shale, Bluefield Formation, Mauch Chunk Group, Upper Mississip-
pian, at Greer, Monongalia County, West Virginia, in the face of
a quarry operated by the Greer Limestone Company, about 0.5
miles north of West Virginia Highway 7.

PRESERVATION AND RESTORATION

The holotype of Mauchchunkia was collected as a group of limy
nodules with bits of bone exposed on the surfaces. The skull had
come to rest right side up, but during burial it had collapsed to the
left, folding up along the typical anthracosaur hinge-line between
temporal series and squamosal, so that the left cheek and lower jaw
are folded underneath, covering part of the palate; the right cheek
and lower jaw are spread out to the right.

Dermal elements of the skull roof are in some disarray because
of maceration. Both squamosals and the premaxillary region have
been severely crushed so that details of their morphology cannot be
made out, and details of the area immediately in front of the orbits
are also obscure. Restoration of the remainder of the skull is based
upon actual sutures or upon patterns of dermal sculpture, and may
be accepted with confidence.

Most of the palate is represented merely by broad expanses of
bone covered with a shagreen of very fine denticles, and sutures
cannot be determined. Fortunately, a part of the pterygoid adjacent
to the basipterygoid articulation is preserved in proper relationship
to the most posterior ectopterygoid teeth, so that the width of the
pterygoid in this region can be determined, at least to an order of
magnitude. This width has provided the basis for establishing the
width and depth of the skull as restored.

In the axial skeleton, both central and neural arch elements
have been shifted variously, chiefly in a lateral direction; some

NICHOLAS HOTTON III

No. 12

intercentra are missing, and all of the neural arch elements are
more or less broken. To the extent that the broken and displaced
elements have been restored to their proper relationships, the illus-
trations represent reconstructions. All of the vertebrae are pre-
served in their proper sequence, however, and proportions of indi-
vidual bones are readily determinable.

The entire pectoral girdle is preserved in NMNH 22573 but was
badly smashed before and during fossilization. In addition, the ven-
tral surfaces of clavicles and interclavicle are partially obscured by
masses of abdominal scales, and the dorsal surfaces by a string of
about seven vertebrae and ribs. Restorations shown in figures 9
and 10 are reliable with respect to most dimensions, but the out-
lines of the interclavicle and the shape of its stem, and the shape
of the top of scapula and cleithrum are uncertain.

Except for the heads of the humeri and one anterior phalanx,
no limb elements are preserved with the holotype of Mauch-
chunkia. Restoration of the front limb is based upon NMNH 26368.
This specimen was found in a small tumble-block near the holotype,
but it was not in place and its association is open to question. It is
referred to Mauchchunkia because the headless humeri are pre-
cisely the right size for the humeral heads that belong to the type,
and their ends are broken at the proper angle to fit the broken ends
of the heads, although they do not make an exact “jigsaw puzzle”
fit. Moreover, the bones of NMNH 26368 were covered with a limy
crust before preparation, as were the bones of the type. Most other
specimens from Greer that I have examined lack a nodular crust,
and instead lie free in the shale matrix.

NMNH 26369 originally consisted of a small limy nodule with
broken bones exposed in its surface. It was forwarded to the writer
by Mr. Burke with the note that Mr. Moran had found it in the
excavation from which the holotype of Mauchchunkia had come. Its
association is thus better than that of NMNH 26368, and the pelvic
and femoral fragments that it contains represent an animal of the
same size as the holotype.

MORPHOLOGY

Skull roof: In dorsal aspect (fig. 1) the skull presents the appear-
ance of a typical anthracosaur, with characteristic sculpture, prom-
inent otic notches, and small but distinct tabular horns. The inter-
temporal bone is almost as large as the supratemporal, and the
tabular has a broad contact with the parietal. In dermal pattern
and sculpture the skull resembles that of “ Paleogyrinus ” decorus,
but the tabular horns, projecting as they do from the undersides of
the tabulars, are more nearly similar to the tabular horns of Ptero-

1970

MAUCHCHUNKIA

plax cornuta. The orbits lie about halfway along the length of the
skull, which is narrower relative to its length than the skull of
either “ Paleogyrinus ” or Pteroplax; gross proportions are more
nearly comparable to those of a form like Gephyrostegus than to
most embolomeres.

Fig. 1. Mauchchunkia bassa Hotton, NMNH 22573. Skull, dorsal aspect, some-
what restored. Key to elements: F, frontal; IT, intertemporal; J, jugal; L,
lacrimal; MX, maxilla; N, nasal; P, parietal; PF, postfrontal; PMX, premaxilla;
PO, postorbital; PP, postparietal; PRF, prefrontal; Q, quadrate; QJ, quadrato-
jugal; SQ, squamosal; ST, supratemporal; T, tabular, X %.

The otic notch is elongate in that its dorsal margin includes
part of the supratemporal as well as the tabular, instead of the tab-
ular alone as in embolomeres. In this respect it resembles Gephyro-
stegus and other reptiliomorph anthracosaurs. As preserved, the
otic notch is smashed down over the squamosal, so that its inferior
margins are indeterminate. The otic margin of the squamosal (fig.
2) is restored after that of “Paleogyrinus” decorus (Panchen, 1964).
It is possible, though not probable, that this margin in life was more
concave than is shown in the restoration, in which case the otic
notch would be larger and would bear a closer resemblance to that
of Gephyrostegus.

A slender process of the squamosal extends ventroposteriorly
between the quadratojugal and the quadrate, as in Gephyrostegus.
The quadrate is high, and is broadly exposed posteriorly between
the squamosal and the quadrate ramus of the pterygoid (fig. 1) .

NICHOLAS HOTTON III

No. 12

Its posterior surface is smooth and unsculptured, as though it con-
tributed to the floor of the middle ear.

The anterior margins of the nasals and the entire premaxilla
were badly smashed before fossilization; the anterior margin of the
premaxilla is marked only by a few displaced teeth. Configuration
of the external nares is unknown. They are restored as though they
were superficially continuous with the skull margin, following Pan-
chen’s treatment of “Paleogyrinus” (1964) , because the anterior
margin of the maxilla appears to be intact and to consist of finished
bone.

Fig. 2. Mauchchunkia bassa Hotton, NMNH 22573. Skull, lateral aspect, some-
what restored. For key to elements see Fig. 1. X %.

The frontals are long and narrow, widening anteriorly as in
“Paleogyrinus” , and the parietals are deeply notched laterally by
the large supratemporals.

Mauchchunkia is more primitive than any other anthracosaur
in the relative lengths of the dermal bones of the midline series.
Combined length of postparietals and parietals (skull table) is
about 73 percent of the combined length of frontals and nasals
(face) , and the postparietal is about % the length of the elongate
parietal, much longer than in any Paleozoic tetrapod except Ichthy-
ostega and its allies. For comparison, values of the ratio of skull
table to face in embolomeres and their close relative Protero-
gyrinus (Romer, 1970) are: “Paleogyrinus”, a relatively short-
faced form, 56 percent; Proterogyrinus, a contemporary of Mauch-
chunkia, 50 percent; Pteroplax, Eogyrinus, and other large forms,
38 percent; Archeria, 33 percent or less. In these forms the relative
shortening of the skull table is primarily a consequence of elonga-

1970

MAUCHCHUNKIA

tion of the snout, while in Gephyrostegus, in which the ratio is
about 38 percent, shortening does not involve modification of gross
proportions of the skull, but is related to a more deep-seated re-
structuring of the braincase (cf. Westoll, 1943) .

The parietals are the most heavily sculptured bones of the skull,
but even here the sculpture is shallow and ill defined. The frontals
and nasals are almost smooth, and sculpture on the postparietals and
tabulars consists of little more than a slight rugosity. Lateral line
canals are almost entirely lacking. The only structures that could
be so interpreted are a few linearly arranged, elongate pits sur-
rounding the orbit (fig. 1), on the front of the prefrontal, on the
postorbital, and on the jugal.

Occiput and braincase : The only relationships that have been pre-
served in the occiput (fig. 3A) are those of postparietals and tab-
ulars. Restoration of the positions of exoccipital, opisthotic, and
prootic (fig. 3B) must be regarded as tentative because of the dam-
age and dislocation that these elements have suffered. The bones
identified as exoccipital and opisthotic are stout, massive structures
that lie, disarticulated, in the matrix behind the posterior margin
of the skull, on either side of the midline.

Fig. 3. Mauchchunkia bassa Hotton, NMNH 22573. A, occiput; B, right lateral
aspect of braincase; both extensively restored. Key to elements: BO, basi-
occipital; BSP, basisphenoid; EO, exoccipital; OP, opisthotic; PRO, prootic;
PSP, parasphenoid; T, tabular; TF, facet on opisthotic for articulation with
tabular, X %.

The putative exoccipital lies just behind and a little below the
putative opisthotic. Ventrally it bears a posteriorly directed pedicel
that terminates in an elliptical articular facet that looks like (and
lies in the proper position for) the exoccipital moiety of the occipital
condyle. Anterior to the base of the pedicel it is pierced trans-
versely by a narrow canal, which, if the articular facet has been
properly identified, must be the hypoglossal foramen (fig. 3B) . The
medial margin of this bone is finished and provides a curved sur-
face that is plausibly interpreted as the lateral wall of the foramen

NICHOLAS HOTTON III

No. 12

magnum. The dorsal surface is expanded into an ovate facet of a
size to fit the broad posterior base of the bone identified as opis-
thotic (fig. 3A) .

The opisthotic is so identified because the surface that is upper-
most, as the bone lies in the matrix, is a falciform articular area
like the dorsolateral facet on the opisthotic of “Paleogyrinus”
(Panchen, 1964) . This facet is only slightly displaced from the ven-
tromedial margin of the occipital flange of the right tabular, and is
of the proper size to articulate with that margin.

Below the right tabular is a piece of broken bone which bears
a deep, well-finished notch in the margin that lies upward. If this
fragment is rotated counterclockwise about 90°, so that the notch
comes to face posteriorly, it fits well as a prootic (fig. 3B) . The
notch can then be interpreted as the anterior margin of the fenestra
ovalis, which now lies in the correct position with respect to tabular
and otic notch.

As restored, the occiput corresponds generally to that of
“Paleogyrinus”, except that the supraoccipital is unossified. The
lateral column formed by exoccipital and opisthotic is much taller,
making the braincase seem too tall and narrow in posterior aspect.
However, the dimensions of the occipital condyle produced by this
restoration fit the central part of the atlas-axis complex perfectly,
and there would scarcely be room above the condyle for a foramen
magnum of appropriate size if the braincase were not as tall as here
restored.

Palate: Exposed surfaces of the palate (fig. 4) are uniformly cov-
ered by a shagreen of fine denticles and no sutures are visible. Res-
toration is based upon general embolomere structure. The palatine-
ectopterygoid suture is placed in front of the posterior tusk-and-pit
pair because of the distance between it and the next anterior pair.
In distribution and form of palatal teeth Mauchchunkia resembles
Panchen’s restoration of Eogyrinus attheyi (written communication,
1969) . In front of the anterior tusk-and-pit pair there is a tiny
notch of what appears to be finished bone, which is interpreted as
the medioposterior margin of the internal naris.

The area in which the pterygoid articulates with the basis
cranii is identifiable by a finished medial margin and a small flange
turning upward and medially from the top of the pterygoid. This
flange is either the anterior root of the dorsally directed otic wing
of the pterygoid, or the base of the epipterygoid below the basi-
pterygoid articulation. But except for the flange, the entire area is
crushed flat, and the remainder of the epipterygoid and otic wing

1970

M AU CHCHUNKI A

Fig. 4. Mauchchunkia bassa Hotton, NMNH 22573. Palate, completely restored
on basis of distribution of palatal dentition, and medial pterygoid margin de-
picted by solid line. Key to elements: ECT, ectopterygoid; PAL, palatine; PT,
pterygoid; V, vomer, X %.

of the pterygoid are represented only by comminuted bits of flat
bone.

A small part of the margin of the subtemporal fossa can be
made out, and it is probable that the medial margin of the fossa
was turned down as a vertical flange as in labyrinthodonts gen-
erally. However, poor preservation renders the exact shape of the
subtemporal fossa extremely uncertain.

Teeth and lower jaw: There are about 26 marginal teeth including
those of the premaxilla, and the anterior maxillary teeth are the
largest. Crown height diminishes gradually from the anterior to the
most posterior maxillary teeth, the position of which is shown be-
neath the orbit in figure 2. The dentary tooth row is essentially
a mirror image of the maxillary tooth row. The lower jaw is typi-
cally anthracosaurian, lacking a retroarticular process and tapering
forward from its deepest point below the coronoid region. Sutures
are undeterminable. The deepest part of the jaw is slightly pitted
toward its lower margin, whence shallow grooves radiate in all
directions. The lateral face of the dentary is marked by shallow
longitudinal grooves and elongate pits, and the symphysial region
by very small, deep pits.

NICHOLAS HOTTON III

No. 12

Axial skeleton: The blocks in which the vertebral column was pre-
served can be joined as a continuous string with but one gap (pi. 1),
which resulted from damage during collection. It is doubtful that
any segments are missing in the region of poor contact, and the total
number of relatively complete, articulated vertebrae, including
atlas-axis complex, is 27. The intercentrum and part of the pleuro-
centrum of the 28th vertebral segment are also articulated to the
back of the last block.

Fig. 5. Mauchchunkia bassa Hotton, NMNH 22573. Dorsal vertebrae, left lateral
aspect, slightly restored. A, 19th segment, pleurocentrum and neural arch
duplicated; B, 24th and 25th segments, X %.

Except for the first four cervical vertebrae, changes in mor-
phology of the central elements are gradual from front to back over
the distance preserved. Anteriorly the pleurocentrum and inter-
centrum are subequal in length (fig. 5A) , and posteriorly the length
of the pleurocentrum is increased at the expense of the intercen-
trum (fig. 5B) . The dorsal pleurocentra are biconcave notochordal
discs in which the articulating surface is ovoid, the dorsoventral
axis being slightly longer than the transverse (fig. 6B) . The dis-
coidal portion of the pleurocentrum is surmounted by a low, stout
bony superstructure, the anterior face of which is convex and re-
ceives the pedicels of the neural arch (fig. 6 A, B) . The intercen-
trum forms a crescent lying below the notochord (fig. 6C) .
Throughout most of the column, the horns of the crescent do not
reach more than halfway up the face of the pleurocentrum, and
there does not appear to have been any osseous contact between
intercentrum and the pedicels of the neural arch.

The convex posterior surface of the intercentrum evidently
articulated with the slightly concave anterior face of the pleuro-
centrum of the same segment like a ball-and-socket joint, as sug-
gested by Panchen (1966) for the embolomere Eogyrinus. In any

1970

MAUCHCHUNKIA

case, the anterior and posterior faces of the intercentra of Mauch-
chunkia are identically shaped and finished, and whatever the arti-
culation was between the intercentrum and the pleurocentrum of
the next anterior segment, the articulation between the intercen-
trum and its own pleurocentrum must have been similar. It ap-
pears that in Mauchchunkia as in Eogyrinus about the same degree
of movement was possible between intercentrum and pleurocen-
trum of the same segment as between intercentrum and the pleuro-
centrum of the next anterior segment.

B C

Fig. 6. Mauchchunkia hassa Hotton, NMNH 22573. Nineteenth vertebra. A
and B, pleurocentrum, left lateral and anterior aspects respectively; C, entire
vertebra, anterior aspect. Key to elements: PC, pleurocentrum; IC, intercen-
trum, X 3/4.

Neural arches are generally like those of embolomeres. The
zygapophyses are pedicellate and lie close to the midline, and their
articular facets are markedly slanted (fig. 6C) . The spines are thin
(of small transverse dimension) and broad (of large anteroposterior
dimension) , so that in cross-section they are fusiform. They are
taller than the length of the arch from pre- to postzygapophysis
(table 1) . In this respect Mauchchunkia resembles such low-spined
pelycosaurs as Ophiacodon and Stereophallodon (Romer and Price,
1940) , and contrasts with Eogyrinus (Panchen, 1966) and Archeria
(NMNH 22811) , in which the spines are shorter than pre- to post-
zygapophyseal length.

Stout, blunt processes are developed in extremely variable
fashion near the tops of the lateral faces of most spines (fig. 5A) .
Their position is so variable that on a single spine the left process
may be close to the anterior margin while the right is close to the
posterior. Figure 5 A illustrates approximately the highest degree
of development, which grades downward to complete absence (fig.
5B). Distribution is evidently random; processes are certainly pres-

NICHOLAS HOTTON III

No. 12

ent on numbers 3-5, 7-9, and 19 (counting back from the atlas) , and
are certainly absent from numbers 6, 18, 20, 22, 25, and 27. In its
highest development the process is characterized by a dorsally
directed, unfinished ovoid surface that is separated from the unfin-
ished dorsal end of the spine by a narrow channel of unfinished
bone. Ventrally the process fairs into the lateral face of the spine,
but may be continued ventrally for a variable distance as a low
ridge directed toward the posterior margin of the transverse process.
This ridge may mark the attachment of the myoseptum, which pre-
sumably passed down the lateral face of the neural spine, crossed
the neural arch by running forward along the rib articulation (to
include the rib in the plane of the septum) , and terminated, via the
head of the rib, on the intercentrum, as Panchen (1967) has recon-
structed it in various early tetrapods.

The ribs of Mauchchunkia are fairly slender and appear to be
approximately cylindrical throughout their length, as in embolo-
meres. In any case, there is no evidence of distal flattening, devel-
opment of uncinate processes, or overlap in the fashion of Ichthy-
ostega or the terrestrial temnospondyles.

The number of presacral vertebrae cannot be determined by
the actual position of the sacrum or pelvis, for these elements are
missing in the type of Mauchchunkia, and there are no central ele-
ments or neural arch pedicels associated with the pelvic material of
NMNH 26369. However, the change in the structure of neural arch
pedicels and rib articulations from front to> back is closely com-
parable to Panchen’s restoration of Eogyrinus (1966) , in which the
position of pelvic fragments and sacral rib is known. In Mauch-
chunkia, the pedicels of anterior neural arches are markedly wid-
ened, and become in effect thick, stocky transverse processes (fig.
5 A, 6C) . Accordingly, the anterior ribs are characterized by an
elongate head that lies a considerable distance medial to the tuber-

Fig. 7. Mauchchunkia hassa Hotton, NMNH 22573. A, head of intermediate
rib; B, posterior (?) presacral rib, X %.

1970

M AU CHCHUNKI A

culum (fig. 7A) in order to articulate with the intercentrum. From
the 24th vertebra on back, the pedicels are expanded only slightly
beyond the margins of the centrum, and in the posterior ribs the
head does not project beyond the tuber culum (fig. 7B) . This type
of unexpanded neural arch pedicel and short-headed rib is said to
be restricted to a few segments in front of the sacrum in Eogyrinus.
The pedicels of the 24th and 25th vertebrae of Mauchchunkia (fig.
5B) are even less widely expanded than those of the immediately
presacral vertebrae of Eogyrinus (Panchen, 1966, fig. 5c, d) , and
therefore must be very close to the posterior end of the presacral
column. The pedicels of the 26th vertebra and possibly also those
of the 25th are very closely-knit, perhaps co-ossified, with their re-
spective pleurocentra.

The costal articulation of the 26th vertebra is not specialized
for a sacral rib, nor does the 26th pleurocentrum show any modifi-
cation for a sacral rib comparable to the sacral vertebra of Eogy-
rinus. The pedicel and pleurocentrum of the 27th vertebra are not
well enough preserved to determine presence or absence of such
specializations, but both the 26th and 27th pleurocentra are very
massive, while the 28th is shorter and slighter than the others. It
therefore seems probable that the sacral vertebra is number 27,
although it could even be number 26, in the unlikely event that the
sacral rib was not as highly specialized in Mauchchunkia as in
Eogyrinus. A presacral vertebral count of 26 is evidently close to
the mark for Mauchchunkia, and even if we allow for one or two
missing vertebrae in the column as preserved the count cannot ex-
ceed 28.

The atlas-axis complex is complete except for the atlantal neu-
ral arch and the proatlas (fig. 8A) . Central elements are little dif-
ferentiated from those of more posterior vertebrae, and such spe-
cializations as they show suggest the condition of pelycosaurs. The
atlantal pleurocentrum is much shorter than the others, the long
axis of its articular face is transverse rather than vertical (fig. 8B) ,

Fig. 8. Mauchchunkia hassa Hotton, NMNH 22573. A, atlas and axis and third
cervical vertebra, left lateral aspect, positions somewhat restored; B, atlas
pleurocentrum, anterior aspect; C, atlas intercentrum and pleurocentrum, ven-
tral aspect, X s/4.

NICHOLAS HOTTON III

No. 12

and its anterior surface is convex. However, it is a complete disc,
unfused with the axial intercentrum. The atlantal intercentrum
differs from all others in that its posterior face is markedly concave,
receiving the convex surface of its pleurocentrum (fig. 8A) . Its
ventroposterior margin forms a flange that wraps around the under-
side of the front of its pleurocentrum (fig. 8 A, C) . The lateral tips
of the atlantal intercentrum are lower than those of other inter-
centra.

The axial neural spine is broad, stout, and hatchet-shaped, very
like the axial neural spine of Ophiacodon. The axial prezygapo-
physis is “reversed”, its articular facet facing upward and outward,
which suggests that Mauchchunkia had an atlantal neural arch and
presumably a proatlas of primitive reptilian configuration.

The pleurocentra of the axis and third cervical are unspecial-
ized, and the intercentra of these two segments differ from those
more posterior only in the presence of prominent ventrolateral
bosses, which presumably provided articulation for cervical ribs.
No other costal articular surfaces are preserved on the cervical ver-
tebrae.

The neural spines of the third (fig. 8A) and fourth vertebrae
are somewhat shorter and narrower than more posterior spines, and
have a pronounced backward slant. The next two neural spines
(fifth and sixth vertebrae) are distinctly narrower than more pos-
terior spines and may retain something of the backward slant of the
third and fourth. This suggests that at least six vertebrae were dif-
ferentiated as cervicals to some degree. The third cervical bears
a small, posteriorly directed process on the pedicel of the post-
zygapophysis. Nothing of the sort is preserved on any other ver-
tebra.

Pectoral girdle: The clavicles are marked by a distinctive system
of arcuate grooves which terminate toward the lateral margins of
the bones in a series of shallow pits (fig. 9) . They lie rather far
apart, and the surface of the broad, flat interclavicle exposed be-
tween them is uniformly sculptured with shallow pits, indicating
that the wide separation of the clavicles is the condition that ob-
tained in life. Posteriorly the interclavicle is nearly smooth.

Dimensions of the cleithrum as restored (fig. 10) are reliable,
but details of its structure are obscure. The restoration of the
scapulocoracoid (fig. 10) is a composite, consisting of the blade of
the left scapula above the supraglenoid foramen and the battered
anterior margin of the right coracoid. The glenoid region is pre-
served only as abraded pieces of massive bone surrounding the head

1970

MAUCHCHUNKIA

of the right humerus, from which the dimensions of the glenoid
have been restored. The composite scapulocoracoid resembles that
of Archeria except that the blade of the scapula is somewhat lower
and narrower.

Fig. 9. Mauchchunkia bassa Hotton, NMNH 22573. Dermal shoulder girdle,
ventral aspect, extensively restored, X %.

Fig. 10. Mauchchunkia bassa Hotton, NMNH 22573. Cleithrum and scapula,
left lateral aspect, extensively restored, X 3A .

NICHOLAS HOTTON III

No. 12

Pelvic girdle: The only identifiable parts of the pelvis are the ace-
tabular contribution and stem of the ilium, and a small fragment
of the pubis which includes the obturator foramen. These fragments
are fitted into a restoration of the pelvis (fig. 11) based upon that
of Archeria (Romer, 1957) . A notable feature of the iliac portion
of the acetabulum is that its articular face is directed primarily
downward and is very heavily buttressed dorsally, as though to
support the weight of an animal that spent much of its time out of
water.

Fig. 11. cf. Mauchchunkia bassa Hotton, NMNH 26369. Pelvic girdle, left lat-
eral aspect, restored after Archeria, X %•

Humerus: Only the head of the humerus is included in the holotype
of Mauchchunkia. In dorsal aspect, it appears as a gently rounded,
relatively featureless knob that is strongly flattened in a dorso-
ventral plane. Its proximal end is largely unfinished and was capped
by cartilage in life. Posteriorly the unfinished area spreads onto the
dorsal surface of the humerus (fig. 12 A, LD) , and around its mar-
gin the finished bone is produced into a marked rim. This is evi-
dently the insertion of the latissimus dorsi muscle; it is larger and
more distinct than the same region of Archeria, more nearly com-
parable to the 1. dorsi insertion of pelycosaurs.

1970

MAUCHCHUNKIA

The exact shape of the proximal articular surface is undeter-
minable. However, the proximal end of the humerus is appreciably
thickest at its midpoint, as though the articular surface turned
obliquely around the end from an anterodorsal to a posteroventral
point as in primitive tetrapods generally, in contrast to the strap-
shaped proximal articulation of the humerus of Archeria as re-
stored by Romer (1957) . The unfinished surface extends anteriorly
along the proximal end of the bone almost to the deltopectoral
crest, from which it is separated by about 5 mm of finished bone.

The remainder of the humerus, and the front limb described
below, are based entirely upon material catalogued as NMNH 26368.
Although the association between this material and the holotype of
Mauchchunkia is not absolutely certain, and the front limb so re-
constructed is a composite, the results are self-consistent and are
consonant with the general structure of Mauchchunkia.

The deltopectoral crest (fig. 12B, DP) is a stout, prominent
process that extends about one-third of the way distally along the
ventral surface of the humerus. Its face is unfinished and very
broad. In shape it resembles that of the sphenacodont pelycosaurs
much more closely than it does the poorly developed deltopectoral
crest of Archeria, but the unfinished face is directed more anteriorly
than ventrally as in pelycosaurs. Except for the deltopectoral crest
the ventral surface of the humerus is flat and featureless.

Fig. 12. Mauchchunkia hassa Hotton, NMNH 22573, and cf. M. hassa, NMNH
26368. Left humerus, composite. A, dorsal aspect; B, ventral aspect; C, dorsal
aspect with anterior flange blocked out; all in plane of proximal end. Key to
processes: AF, anterodorsal flange; DP, deltopectoral crest; EC, ectepicondyle;
EN, entepicondyle; ENF entepicondylar foramen; RF, radial articular facet;
UF, ulnar articular facet, X %.

NICHOLAS HOTTON III

No. 12

In dorsal aspect (fig. 12A) the humerus resembles that of
Archeria more closely than it does the humerus of other early tetra-
pods. As in Archeria, the basic tetrahedral pattern is obscured by
the large size and quadrate shape of the entepicondyle, and by the
development of a broad flange lateroanterior to the ectepicondyle.

The posterior margin of the humerus is pierced by a clearly
defined entepicondylar foramen (fig. 12A, ENF) about halfway
along its length. The foramen slants in a distal direction from the
dorsal to the ventral surface, and lies very close to the proximal
root of the entepicondyle, as in Archeria. These relationships are
preserved in the left humerus of NMNH 26368, in which, despite
extensive destruction of the entepicondyle, a short spur of bone
projects to the rear, just proximal to the remains of the entepicon-
dylar foramen. This little spur of bone also demonstrates conclu-
sively that the proximal margin of the entepicondyle turns back-
ward to form a right angle with the axis of the humerus. Thus we
can be certain that the entepicondyle of Mauchchunkia has the
same broad, flat, quadrate appearance as that of Archeria. How
broad it was cannot be determined, but the broken medial edge is
very thick, which suggests that the entepicondyle was prominent.

The ectepicondyle is a tall, narrow ridge, the crest of which is
smoothly rounded (fig. 12A, EC) . It is most prominently developed
at its distal end and tapers gradually in a proximal direction for
about two-thirds the length of the humerus. Distally it overhangs
the unfinished radial articulation; its unfinished distal end is sepa-
rated from the radial articulation by about 4 mm of finished sur-
face.

The distal articular faces of the humerus are preserved only as
areas of unfinished bone. Their margins are very distinct, for the
finished bone that delimits them is produced into a fine rim, as it is
around the margins of the latissimus dorsi insertion. Although the
radial and ulnar surfaces are confluent, they can be distinguished
easily. The radial articulation lies below the distal end of the ecte-
picondyle. Its surface is extensive; a small part faces distally as in
Archeria, but a much larger part spreads onto the ventral surface
of the humerus and faces downward (fig. 12B, RF) , in contrast to
the arrangement in Archeria. The actual articular surface must
have been convex because it lay upon both distal and ventral sur-
faces, but if it were restored to resemble the radial articulation of
pelycosaurs, it would have to be a huge ball composed almost en-
tirely of cartilage. It is more likely that the cartilage cap was rela-
tively thin, so that the greatest convexity of the articular surface
occurred where it curved from the distal to the ventral surface of

1970

MAUCHCHUNKIA

the humerus, and that the large ventral part was only gently con-
vex. The articular surface for the ulna lies at the distal root of the
entepicondyle and faces distally (fig. 12B, UF).

The humerus is thickened along its axis, in a zone running
from the ulnar articular facet at the distal end, through the base
of the deltopectoral crest, to the middle of the humeral head. The
ends of the bone are “twisted” on the axis to form an angle of about
45° between the planes of the distal and proximal articulations, a
value closely comparable to that of terrestrial tetrapods and in
marked contrast to the 20° to 25° for this angle in Archeria, cited
by Homer (1957). This comparison reinforces the possibility that
the proximal articulation was more obliquely placed than in
Archeria.

The broad flange of bone produced anteriorly below the ecte-
picondyle is proportionately larger in Mauchchunkia than in Arche-
ria (fig. 12A and B, AF) . It arises proximally from the anterior
surface of the deltopectoral crest and passes distally to fade into the
base of the ectepicondyle. Proximally it lies in the same plane as
the head of the humerus and distally it lies in the same plane as the
distal articular surface; as a consequence it presents a distinctly
undulant surface because of the high angle between the ends of the
humerus.

Fig. 13. cf. Mauchchunkia hassa Hotton, NMNH 26368. Left forearm and hand,
extensor aspect, position of elements restored, X %.

NICHOLAS HOTTON III

No. 12

Radius: The radius is cylindrical, rather stouter in proportion to
its length than the radius of Archeria but otherwise very similar to
it. Its proximal articular facet is circular in outline and the surface
is almost flat, except for a shallow trough whose transverse orien-
tation corresponds to the plane of the distal articular surface of the
humerus. The distal articular surface of the radius is roughly the
shape of an isosceles triangle, its apex being directed medially. Ex-
tensor and flexor surfaces are essentially smooth. On the lateral
surface there is a fine arcuate line of rugosity occupying the distal
two-thirds of the bone, curving from the extensor to the flexor sur-
face, which may mark the attachment of an interosseous membrane.
Along the medial side there is a low, sharp-edged ridge developed
over most of the length of the bone, which probably marks the
medial boundary between extensor and flexor surfaces. The ridge
becomes more prominent at its proximal end, which is unfinished
and may mark the attachment of a biceps tendon.

Ulna: The ulna of Mauchchunkia is similar in all respects to that
of Archeria , except that like the radius it is somewhat stouter in
proportion to its length. Although the tip of the olecranon process
is not preserved because it was not ossified, the proximal articular
surface is obviously concave and faces medioproximally.

Elbow joint: The articulating surfaces of the ulno-humeral joint
are of standard tetrapod pattern, and the joint evidently functions
as a simple hinge, the ulna turning through an arc of about 90°.
The radio-humeral joint also allows an arc of about 90°, because
the humeral facet for the radius passes from the distal to the ven-
tral surface of the humerus. As a consequence the forearm can
turn, relative to the humerus, from a straight-line orientation in
full extension to a right-angle orientation in full flexion.

In the functional position of the limb during locomotion, the
humerus is oriented horizontally in such a way that the larger por-
tion of the radial articular facet faces downward, and the forearm
is fully flexed. In this position the radius stands vertically, with the
large ventral moiety of the radial articular facet of the humerus
resting on top of it. The radius is thus a weight-bearing column,
for which its short, stocky form is well adapted.

In its flatness, the radial component of the radio-humeral joint
of Mauchchunkia resembles the weight-bearing tibial component of
the knee of higher tetrapods. The flat surface of the radius is
matched incongruently to the convex surface of the humerus in the
elbow of Mauchchunkia, much as the flat tibial surface is matched
to the convex distal end of the femur in higher forms. Both of these

1970

MAU CHCHUNKI A

joints are anatomically unstable because of incongruence; stability
is established in the elbow of Mauchchunkia by the congruent bear-
ing surfaces of the ulno-humeral joint, and in the knee of higher
tetrapods by tendons and ligaments crossing the joint. The radio-
humeral joint of Mauchchunkia thus bears a closer resemblance to
the knee than to the elbow joint of higher tetrapods, and like the
mammalian knee appears to be a weight-bearer that must move
through a wide angle in a single plane.

The similarity of the radio-humeral joint to the knee of higher
tetrapods suggests that pronation and supination were of little func-
tional significance in the elbow of Mauchchunkia. Such rotation
of the radius as occurred during locomotion would have had about
the same magnitude and function as the rotation of the tibia on the
femur that takes place during flexion and extension of the knee in
generalized mammals.

Hand: Except for those elements here restored as third metacarpal
and proximal phalanx (fig. 13) , all of the elements of the hand were
disarticulated, and all were most closely associated with the left
humerus. Little detail can be added to what is shown in figure 13,
which indicates primarily that the hand of Mauchchunkia, like the
rest of the front limb, was large and stout in proportion to the size
of the animal, considerably more so than the front limb of Archeria.

Fig. 14. cf. Mauchchunkia hassa Holton, NMNH 26369. Right femur. A, dorsal
aspect; B, ventral aspect. Key to processes: ITR, internal trochanter; PI,
insertion of puboischiofemoralis or ischiotrochantericus muscle, or both; TR 4,
fourth trochanter, X %•

Femur: The heads of the right and left femora associated with
NMNH 26369 are nearly as broad as the femur of Archeria illus-
trated by Romer (1957, fig. 8C) , but the shafts taper to about one-
third the width of the femoral shaft of Archeria. Some of the broad-
ening of the femoral head of Mauchchunkia may be the result of
distortion after burial, but the femur does seem to be proportion-
ately larger and slimmer than that of Archeria.

NICHOLAS HOTTON III

No. 12

The distance from the internal trochanter to the proximal end
of the femur is approximately twice that of Archeria, indicating a
much higher degree of ossification. The shape of the articular sur-
face cannot be determined because the proximal end of the bone is
covered by fragments of pubis. The unfinished surface that occu-
pies most of the proximal end of the femur narrows abruptly ante-
riorly. It is separated from the unfinished end of the internal tro-
chanter by a sharp ridge of finished bone about 5 mm long.

The unfinished end of the internal trochanter is similar in shape
to that of Archeria, being short and broad rather than long and
narrow as in pelycosaurs. The entire internal trochanter, including
the unfinished end, is much more prominent than in Archeria. The
anterior wall of the intertrochanteric fossa is well developed, being
formed chiefly by the internal trochanter, but the posteridr wall is
poorly defined.

The fourth trochanter is very prominent. Proximally it consists
of an area of marked rugosity lying at the root of the internal tro-
chanter, and as it curves distally and posteriorly toward the middle
of the ventral surface of the shaft it becomes narrow, and is pro-
duced into a high, sharp ridge. It passes insensibly into the adduc-
tor ridge, which continues down the shaft of the femur with no evi-
dent diminution in height as far as preserved.

Dorsally the head of the femur is gently convex. It is nearly
featureless except for a well-defined patch of unfinished bone at the
proximal end of a low swelling near the posterior margin. This area
probably represents the confluent insertions of the ischiotrochan-
tericus and puboischiofemoralis internus muscles (cf. Romer and
Price, 1940, fig. 35) . The dorsal surface is marked along its prox-
imal margin by minor rugosity and fluting, the latter oriented more
or less radial to the margin; such sculpture may mark the site of
attachment of a joint capsule.

DISCUSSION

Mauchchunkia is a remarkable tetrapod, first because of its
extreme primitiveness, which coincides happily with its great geo-
logic age, and second because terrestrial adaptations can be identi-
fied in many aspects of its structure. It is short-coupled and stout-
limbed like Ichthyostega, which confirms the idea, suggested by
Panchen (1966) and elaborated by Carroll (1969) , that terrestrial
adaptation was fundamental to the structure of the earliest tetra-
pods. But the retention of a fish-like tail fin in Ichthyostega and the
evidence of aquatic larval stages in a variety of reptiliomorph an-
thracosaurs shows that these animals were not completely free of

1970

MAUCHCHUNKIA

water. Rather, their body and limb proportions suggest that they
were primarily walkers and waders of rather turtlelike habit, to
be contrasted with the long-bodied, short-limbed, swimming em-
bolomeres and the flattened, bottom-dwelling, persistently aquatic
temnospondyles.

Another anthracosaur from Greer, Proterogyrinus (Romer,
1970), is also primitive but is much closer than its contemporary
Mauchchunkia to true embolomeres; comparison with this animal
suggests that Mauchchunkia illustrates both the organization of
a basic anthracosaur stock and the stem from which arose the rep-
tiliomorph tetrapods, including gephyrostegids, seymouriamorphs,
and true reptiles. Some reptiliomorph tetrapods, such as Seymouria
and the early pelycosaurs, seem to have had much the same habits
as Mauchchunkia and Ichthyostega, while others, such as the ge-
phyrostegids and the earliest captorhinomorphs, may have been
more terrestrial, attaining a mode of life comparable to that of ter-
restrial salamanders or lizards. These differences in habit seem to
be correlated with size, for Ichthyostega, Mauchchunkia, the earliest
pelycosaurs, and Seymouria were all of moderate size; the embolo-
meres and bottom-dwelling temnospondyles tended to become very
large, while the gephyrostegids and captorhinomorphs were very
small (less than half the size of Mauchchunkia) . The origin of
various lines of early tetrapods, including the first reptiles, appears
to have been a matter of adaptive radiation controlled primarily by
the degree of dependence upon standing water, and a detailed com-
parison of Mauchchunkia with Proterogyrinus on the one hand, and
with more advanced reptiliomorph tetrapods on the other, affords
many data for conjecture about these origins.

Primitive characters and terrestrial adaptation: The primitiveness
of Mauchchunkia is indicated, more or less independently of ter-
restrial adaptation, by the length of skull table and postparietal bone
(cf. Westoll, 1943) , in which Mauchchunkia is more closely com-
parable than any other anthracosaur to Ichthyostega, and by its ven-
trally located, wedge-shaped or crescentic intercentra, in which it
resembles Ichthyostega and certain rhipidistian fishes (cf. Romer,
1947, 1964) . The crescentic shape of the intercentra probably repre-
sents a definitive adult condition in Mauchchunkia rather than an
ontogenetic stage in the development of a more conventional em-
bolomerous vertebral pattern, for the high degree of ossification of
other vertebral elements and the completeness of the skull bones
in the holotype indicate that the individual was essentially mature
when it died.

The inclusion of a part of the supratemporal in the dorsal mar-

NICHOLAS HOTTON III

No. 12

gin of the otic notch may also be a primitive character, although
in this respect the otic notch of Mauchchunkia resembles that of the
reptiliomorph anthracosaurs more closely than the otic notch of
embolomeres. In the traditional view of the origin of the amphibian
otic notch from the spiracular cleft of fishes, it is logical to suppose
that the short embolomere notch, with its dorsal margin restricted
to the tabular, is more primitive than the longer one of Mauch-
chunkia. But because the anthracosaur otic notch originated in
relation to a persistent hinge between skull table and cheek, it is
equally likely to have been elongate or ill-defined anteriorly at
some primitive stage, which may well be illustrated by Mauch-
chunkia.

The large size and quadrate shape of the humeral entepicondyle
(fig. 12, EN) , and the flange of bone produced anteriorly from the
shaft of the humerus (fig. 12, AF) , are probably conservative fea-
tures, perhaps held over from a fishlike stage, for they are retained
until the early Permian in the aquatic embolomere Archeria, in
which they are associated with small limbs and weak muscle attach-
ments. In Mauchchunkia, however, they are associated with rela-
tively large limbs and powerful muscle attachments. The entepi-
condyle of the pelycosaur Ophiacodon is proportionately smaller
than that of Mauchchunkia but retains something of the quadrate
shape, which only disappears in more advanced pelycosaurs as the
proximal half of the humerus increases in length. The anterior
flange contributes to the exotic appearance of the humerus of
Mauchchunkia, but its deletion (fig. 12C) clarifies the basic similar-
ity of the humerus to that of a primitive pelycosaur.

The similarity of development of the deltopectoral crest and the
latissimus dorsi insertion to that of pelycosaurs is convincing evi-
dence of terrestrial adaptation. The deltopectoral crest marks the
insertion of large and powerful muscles that maintained the hu-
merus in a horizontal position, supporting the body clear of the
ground during locomotion on land. The same muscles also provided
most of the force for flexion, adduction, and clockwise rotation of
the front limb, which collectively constituted the “power stroke”
in walking. The latissimus dorsi was an essential synergist of these
muscles and must have been correspondingly large and powerful.

Homer (1957) has related the low angle between the planes of
the proximal and distal ends, or “twist”, of the humerus of Archeria
to a primarily swimming mode of locomotion in that genus. Con-
trariwise, the high “twist” characteristic of Mauchchunkia is of the
same magnitude as that of terrestrial tetrapods of the Paleozoic, and
is correlated with highly developed muscle insertions in a complex
of terrestrial adaptation.

1970

MAUCHCHUNKIA

TABLE 1

Lengths of front limbs are expressed in terms of Romer’s ortho-
metric linear units (Romer and Price, 1940) : OLU = r2/'3, when
r = transverse radius of the pleurocentrum. OLU values are listed
as indices of gross size. Key to abbreviations: HS/LV, height of
neural spine/length, pre- to postzygapophysis, dorsal vertebrae;
LR/LH, length of radius /length of humerus; OLU, orthometric
linear units, measurements in millimeters. Data sources: Eogyri-
nus, Panchen, 1966; Archeria, Gephyrostegus, Carroll, 1970; Sey-
mouria, NMNH 21902; Proterogyrinus (estimated from figures),
Romer, 1970; Mauchchunkia, NMNH 22573, 26368; all pelycosaurs,
Romer and Price, 1940.

No. of
presacral
vertebrae

OLU

Length,

front

limb

LR/LH

HS/LV

Embolomeres

Eogyrinus

8.30

—

.66

Archeria

5.00

.54

.83

Seymouria

3.55

.65

—

Gephyrostegus

1.84

.53

.75

Proterogyrinus

—

3.90

.60

.90

Mauchchunkia

28 (max.)

3.97

.50

1.39

V aranosaurus
acutirostris

3.66

.70

Ophiacodon

navajovicus

3.83

.77

—

O. mirus

4.48

.74

—

O. retroversus

5.95

.78

2.00

Dimetrodon

limhatus

5.53

.87

Stereophallodon 27

6.08

—

1.25

In relative length of the front limb (table 1) , Mauchchunkia
appears to fall, together with Gephyrostegus, about halfway be-
tween Archeria and the terrestrially-adapted Seymouria. However,
this should not be interpreted without further consideration as a
morphological “halfway” stage, for as has been repeatedly noted
(Olson, 1951; Romer, 1957; Panchen, 1966), the basic assumption
of Romer’s use of orthometric linear units, that the radius of dorsal
vertebral centra provides an index of body mass, may be grossly
misleading in comparing animals of different habitus and distant
phyletic relationship. The centra of an evolved aquatic form like

NICHOLAS HOTTON III

No. 12

Archeria may have been smaller in proportion to body mass than
those of evolved terrestrial forms like Seymouria or the pelyco-
saurs, because Archeria could rely upon the supportive effect of the
circumambient water in which it lived, while terrestrial forms re-
quired anatomical reinforcement of the column for support on land.
On the other hand, the centra of such primitive terrestrial forms as
Mauchchunkia and Gephyrostegus may have been proportionately
larger than those of more advanced forms, because the need for
support in a terrestrial environment was probably met initially by
the crude expedient of increase in size, in contrast to a more so-
phisticated system of articulation developed in later terrestrial
animals.

To the extent that these possibilities are valid, Romer’s pro-
cedure gives an excessively high value for the limb proportion of
Archeria and an excessively low one for that of primitive terrestrial
forms, in comparison with evolved terrestrial tetrapods. Such dis-
crepancies cannot be dealt with by the procedure itself, but they
can be compensated for non-numerically by downgrading the nom-
inal value for Archeria and upgrading it for Mauchchunkia and
Gephyrostegus. Downgrading the nominal value for Archeria in-
creases the scale of difference in limb proportion between Archeria
and the demonstrably terrestrial tetrapods, and upgrading it for
Mauchchunkia and Gephyrostegus moves those genera further up
the scale toward Seymouria and the pelycosaurs than their nominal
halfway point. It is therefore evident that in proportion of the front
limb, Mauchchunkia is considerably closer to Seymouria and the
pelycosaurs than it is to Archeria.

The shortness of the radius, relative to length of humerus, is
probably a manifestation of primitiveness in Mauchchunkia, for the
morphological sequence Mauchchunkia-Gephyrostegus-Seymouria,
which reflects general evolutionary advance correlated with time in
a series of approximately common habitus, shows a consistent
increase in the relative length of the distal segment of the front
limb. The pelycosaurs exhibit still greater length of the distal seg-
ment; though they are contemporaneous with Seymouria and prob-
ably of somewhat different habitus, it is generally agreed that they
represent a more advanced evolutionary condition.

The initial shortness of the distal segment of the front limb may
express an early stage in the development of weight-bearing func-
tion, in probable correlation with the primitively weight-bearing
nature of the elbow joint in which pronation and supination were
not yet clearly defined (cf. p. 24) . By contrast, the greater length
of the pelycosaur radius and ulna is correlated with development
of a ball-shaped radial condyle of the humerus, similar to that of
more evolved tetrapods in which a greater range of pronation and

1970

MAUCHCHUNKIA

supination is possible. Lengthening of the distal segment probably
occurred as a means of increasing the length of stride, which was
selectively advantageous in a terrestrial environment, and it ap-
pears that development of pronation and supination accompanied
this advance as front limb motion became more complex.

The combination of primitiveness and terrestrial adaptation,
noted in the anatomical complex of stout limbs and short presacral
vertebral column, is also reflected in the basic structure and height
of the dorsal neural spines of Mauchchunkia. Basic structure is
probably primitive because it is similar to that of embolomeres on
the one hand and to that of pelycosaurs on the other, and contrasts
with the structure of advanced Permian seymouriamorphs and cap-
torhinomorphs. The remarkable height of the spines in Mauch-
chunkia indicates massive development of the dorsal axial muscula-
ture, which functioned in concert with intercostal and belly muscu-
lature to lend dynamic stability to the vertebral column. The very
massiveness of the dorsal musculature suggests that the column was
being stabilized under terrestrial rather than aquatic conditions.
Since stabilization of the column by means of muscular tension
would subject the centra to compressional forces, it is probable that
the high degree of ossification of the pleurocentra is associated with
emphasis on the dorsal axial musculature in the general terrestrial
adaptation of the vertebral column.

The atlas-axis complex is very generalized, but the atlantal
intercentrum and the large, hatchet-shaped axial neural spine are
both pelycosaurlike, the axial spine reflecting the former presence
of a well-developed nuchal ligament. These structures appear to be
adapted to support of the heavy head in a terrestrial environment,
and the distinctive if minor specialization of the four postaxial neu-
ral spines indicate differentiation of a neck, signifying that consider-
able movement of the head was possible.

Anthracosaur phytogeny: Associated with the primitive ground
plan of Mauchchunkia, but not particularly attributable to terres-
trial adaptation, are the features by which the genus is diagnosed
as a member of the Anthracosauria: tropitrabic skull; pattern and
sculpture of dermal skull bones; tabular horns; and pattern of
palate, dermal pectoral girdle, and neural arches. These features
are for the most part characteristic of the later embolomeres, but
their presence in Mauchchunkia suggests that they are also part of
the original anthracosaur heritage. They have been variously mod-
ified in later terrestrially adapted anthracosaurs, and in the em-
bolomeres by elongation of the snout and presacral column. Mauch-
chunkia obviously lies near the ancestry of both types, and for fur-

NICHOLAS HOTTON III

No. 12

ther assessment of its phylogenetic significance we must refer now
to its contemporary Proterogyrinus.

Proterogyrinus is much closer to true embolomeres in skull
proportion, for the skull table is only about half the length of the
face, and the contribution of the postparietal to the skull table is
smaller than in Mauchchunkia, being roughly comparable to that of
“ Paleogyrinus ” (cf. Panchen, 1964) . The otic notch is short, its
dorsal margin being restricted to the large tabular. The snout shows
some elongation. Length of the presacral column is unknown, but
the neural spines are more closely comparable in height to those of
embolomeres than to the spines of Mauchchunkia (table 1) . Spine
height and length of snout suggest that the column may have been
elongate, in correlation with the aquatic, piscivorous habit typical
of embolomeres. Romer interprets the central elements as subequal
in height and very similar in appearance; in this feature also Pro-
terogyrinus resembles the embolomeres more closely than does
Mauchchunkia. But pleurocentra as well as intercentra are de-
scribed as thin hoops of bone, unossified dorsally; in anterior aspect
both elements are U-shaped. Romer notes that although the type
of Proterogyrinus was probably not mature at death, maturity
would not have brought vertebral ossification to a point comparable
to that of true embolomeres, and therefore places Proterogyrinus
in a distinct family, morphologically antecedent to embolomerous
forms. Its proximity to true embolomeres is indicated not only by
skull structure, but also by the probability that only a genetically
simple increase in rate of ossification was necessary for its verte-
brae to become fully embolomerous, and it is very probably an
actual ancestor of definitive embolomeres.

The vertebrae of both Greer anthracosaurs are derivable from
a schizomerous pattern (Romer, 1964) in which the pleurocentrum
consisted of laterally placed half-rings and the intercentrum was
a ventrally located crescentic element. The vertebrae of Mauch-
chunkia conform in general to Romer’s diplomerous pattern, in
which the pleurocentrum is a complete perichordal ring derived by
dorsal and ventral coossification of schizomerous half-rings, with the
intercentrum remaining essentially unmodified. Most of the pleuro-
centra are advanced beyond a strictly diplomerous condition be-
cause they are ossified into complete discs, but a trace of schizo-
merous structure is retained in the atlantal pleurocentrum, in which
dorsal and ventral marginal notches (fig. 8B) suggest that the ele-
ment was formed by the fusion of lateral halves. The vertebrae of
Proterogyrinus are not diplomerous; the pleurocentrum seems to be
formed simply by coossification of schizomerous half-rings below
the notochord, and the intercentrum by dorsal ossification of the
horns of the original crescent. To distinguish these divergent pat-

1970

MAUCHCHUNKIA

terns in terms of their theoretical origins, the vertebrae of Protero-
gyrinus are styled “neoschizomerous” and those of Mauchchunkia
“neodiplomerous”.

The neoschizomerous vertebral pattern of Proterogyrinus is
a virtually ideal morphological intermediate between schizomerous
and embolomerous stages, and indicates that embolomerous verte-
brae originated directly from a schizomerous ancestral pattern with-
out going through a diplomerous stage at all. This, as Romer (1970)
notes, invalidates his earlier view (Romer, 1964) of the diplomerous
pattern as intermediate between schizomerous and embolomerous
stages. Moreover, it seems probable that the neodiplomerous struc-
ture of Mauchchunkia evolved from a schizomerous antecedent
during the same period of time that the neoschizomerous pattern
of Proterogyrinus was being developed. Thus the origin of the ver-
tebral patterns of Mauchchunkia and Proterogyrinus must be at-
tributed to independent trends that diverged from the level of a
putative schizomerous common ancestor a short time before the
Upper Mississippian. The possibility, suggested by Carroll (1970)
for Gephyrostegus, that neodiplomerous intercentra may have be-
come completely ossified dorsally in very old individuals, is not
known in actuality and in any case is probably not phylogenetically
significant.

Exclusion of diplomerous and neodiplomerous structure from
the line of embolomere descent greatly enhances the significance of
these patterns as indicators of the stem from which reptiliomorph
tetrapods sprang, which establishes Mauchchunkia as the earliest
known member of that stem, just as Proterogyrinus is the earliest
known member of the line that led to embolomeres. Assuming an
origin from schizomerous antecedents, the divergent trends in the
two lines may be interpreted in terms of ontogenetic acceleration
in the rate of ossification of vertebral centra, which in the line of
Mauchchunkia was rapid but affected primarily the pleurocentrum,
leaving the intercentrum little changed. In the descendants of
Mauchchunkia, ossification of the intercentrum was de-emphasized
and the element eventually disappeared. In the line of Protero-
gyrinus, acceleration of the rate of ossification was slower, but af-
fected pleurocentrum and intercentrum alike, leading ultimately to
complete ossification of both elements in the definitive embolomeres.

Since the terrestrial adaptations of Mauchchunkia appear to be
for the most part conservative in nature, establishment of diplo-
merous structure in that line probably represents a refinement of
the originally terrestrially oriented organization of the basic stock.
Emphasis on the pleurocentrum probably arose with emphasis on
the dorsal axial musculature as a means of stabilizing the column

NICHOLAS HOTTON III

No. 12

in a terrestrial environment, in response to selective pressure exert-
ed by the need for support inherent in such circumstances.

The origin of an embolomere stock, on the other hand, was
probably initiated as the ancestors of Proterogyrinus found them-
selves able to exploit a more completely aquatic mode of life. The
less well-ossified condition of the centra of Proterogyrinus indicates
that selective pressure occasioned by the need for support was not
as effective in the ancestry of Proterogyrinus as in that of Maucln-
chunkia, as would be expected if the former had taken to living
consistently in deeper water. Instead of requiring refinement of
structures utilized for support, the aquatic environment exerted
pressure toward improvement of a swimming habit, to which the
embolomere line responded by elongation of the column in the de-
velopment of a sinuous swimming motion. The selective advantage
of elongation of the column was presumably the increased flexibility
it afforded. If, as Panchen (1966) suggests, the pleurocentrum and
intercentrum of the same segment were movable on each other,
coeval ossification of the two elements would also enhance flex-
ibility without appreciable sacrifice of strength, and hence could
result from the same selective forces that produced column elon-
gation.

Whether elongation of the column preceded complete ossifica-
tion of the central elements or was concurrent with it cannot be
determined without a presacral vertebral count for Proterogyrinus.
In any case, by the early Pennsylvanian the embolomeres were
elongate swimmers with fully ossified central elements, and many
of them were quite large. Like most early tetrapods, these animals
were predaceous, and their increase in size from the Mississippian
to the Pennsylvanian was probably selected for as a consequence
of competition with predaceous fish. The utilization of sinuous mo-
tion by large swimming predators may have subjected the individual
components of an elongate column to unusual compressional and
tensional stresses, another factor that would select for complete and
coeval ossification of pleurocentrum and intercentrum.

Origin of reptiles: Although Mauchchunkia is clearly a member of
the reptiliomorph stem, as a generalized anthracosaur it exhibits no
closer morphological affinity to one reptiliomorph branch than to
another, and since it occurs nearly a full period earlier than any,
it represents a group that must have included the ancestors of ge-
phyrostegids, true reptiles, and seymouriamorphs alike. Members
of this group, which were primarily walkers and waders that lived
in shallow ponds and streams and along the margins of deeper
bodies of fresh water, were as generalized in habit as they were

1970

MAUCHCHUNKIA

in structure, and provide an excellent starting point for the con-
sideration of the origin of reptiles as an adaptive radiation.

The earliest tetrapods to occupy environments more highly
terrestrial than the margins of standing water were all very small
(Carroll, 1969), the largest of them less than half the size of Maucln-
chunkia. As examples Carroll cites the gephyrostegids and capto-
rhinomorphs of the Middle Pennsylvanian, but notes also (written
communication, 1970) that some of the earliest pelycosaurs on rec-
ord, which date from the same time, are the same size as gephyro-
stegids and primitive captorhinomorphs. He suggests that the suc-
cess of these animals under terrestrial conditions stemmed from
advantages conferred by small size, which mitigated problems of
support and enabled the animals to utilize secretive behavior to
conserve water, and which ultimately proved decisive in the origin
of the amniote egg. At a preamniote level, small size would reduce
the need for the egg to be laid in standing water because of reduced
need for support, greater facility for respiration, and the sufficiency
of local dampness to keep a small egg moist. Subsequent steps in
the evolution of the amniote egg required development of direct
internal fertilization and a large-yolked egg in which the larval
stage could be passed. Evolution of living amphibians provides
plausible parallels for this phase, for such features have appeared
independently a number of times in connection with increasing ter-
restrially. Noble (1931) points out that both direct internal fer-
tilization and large-yolked eggs are characteristic of the relatively
archaic caecelians, and implies that these features may have been
generally present in early tetrapods.

Carroll envisions the amniote egg as having originated in a line
of small progressive forms such as gephyrostegids, which, having
become highly terrestrial as adults, began to lay their tiny eggs in
damp places on land as do the living plethodont salamanders, and
later evolved direct internal fertilization and a large-yolked egg.
He implies that the final stages in the evolution of amnionic struc-
ture took place during the transition from gephyrostegids to capto-
rhinomorphs, and for the transition itself he presents a convincing
morphological argument. Uniformity of egg structure among living
amniotes indicates strongly that all are derived from a single type,
which in turn means either that the amniote egg arose only once
or that any other form that approximately duplicated amnionic
structure became extinct without issue.

Since the reasons for considering captorhinomorphs to be am-
niotes apply as well to pelycosaurs, it follows from the argument for
a single origin of the amniote egg that one group must have been
derived from the other. But though pelycosaur and captorhino-

NICHOLAS HOTTON III

No. 12

morph lines converge when traced backward from the Permian,
they are still distinct at the earliest appearance of reptiles in the
Middle Pennsylvanian. The possibility must therefore be admitted
that pelycosaurs arose from an anthracosaur level independently of
the origin of captorhinomorphs; they may have come from gephyro-
stegids, or perhaps even from animals of a mauchchunkiid level of
organization.

Such an eventuality demands examination of alternative pos-
sibilities for the origin of the amniote egg. As a first step, two fac-
tors must be noted: one, that a large-yolked egg may well have
been characteristic of a variety of primitive tetrapods as an inher-
itance from the archaic fishes from which they sprang; and two,
that seasonal water fouling and drying were probably characteristic
of the bodies of water in which members of the conservative anthra-
cosaur line lived. A large-yolked, sizeable egg would predispose its
possessors to pass their larval stages within the egg in any circum-
stances that were inimical to free larval life. The larger the egg,
the more vulnerable it would be to asphyxiation in oxygen-poor
surroundings because of the ratio of surface to volume, but by the
same token, the less vulnerable it would be to desiccation. In these
respects a large-yolked egg of appreciable size is well integrated
with the terrestrially oriented morphology of the earliest anthra-
cosaurs.

It is entirely conceivable that primitive tetrapods like Mauch-
chunkia, having established themselves in pond and river margin
environments as walkers, waders, and paddlers, would tend to
utilize the extreme margins or the damp banks of these bodies of
water as places in which to lay their eggs. This habit would be im-
mediately advantageous, for eggs laid in such places would be at
least partially exposed to air and so would be more likely to survive
the effects of water fouling. Quite possibly they would also be less
subject to predation. At the same time they would be subject to
desiccation, which they were predisposed to resist, but which would
introduce the same major selective factor that was operative on the
eggs of gephyrostegids and primitive captorhinomorphs that were
deposited in more completely terrestrial surroundings.

Conditions prerequisite to the origin of the amniote egg thus
probably obtained in conservative anthracosaurs of the reptilio-
morph line. It is doubtful that amnionic structure as such was
present as early as Mauchchunkia, for indications are that seymour-
iamorph derivatives of the mauchchunkiids went through a free-
living larval stage, but the amniote egg may well have appeared
before the establishment of definitive reptilian osteological struc-
ture. If this were the case, it would certainly have been a factor

1970

MAU CHCHUNKI A

in the success of the smaller forms that were making their way into
more highly terrestrial environments, and in addition it would ac-
count for the presence of conservative lines such as the limnosce-
loids and perhaps the diadectids, which at the same time were
evolving a reptilian morphology without being very small or being
markedly more highly adapted to terrestrial conditions.

Alternatively, it is possible that the last stages in the establish-
ment of amnionic structure occurred independently in small, highly
terrestrial ancestors of captorhinomorphs and pelycosaurs, in more
conservative ancestors of limnosceloids, and even, perhaps, in the
seymouriamorph line after the establishment of Seymouria- like
forms but before that of the family Diadectidae. This notion seems
to imply that the amniote egg arose several times, and brings to
mind the putative history of the later synapsid reptiles, in which
a wide variety of progressive characters evolved in tightly parallel
fashion under the pressure of an increasingly active mode of life.
However, the anthracosaurs in question were removed but a short
distance in time from their common ancestry, and must have been
much more closely interrelated than were the synapsids. The
greater part of the evolution of amnionic structure had already
taken place in what was essentially a single line, and the selective
pressure that had brought it along continued to affect the adaptive
branches to which the main line gave rise. The differences by
which these branches are identified foreshadow their great phylo-
genetic potential, but this should not lead us to exaggerate the dif-
ferences among them at the time of branching, with respect to the
genetic factors controlling the evolution of egg structure. Whatever
finishing touches were put upon amniote structure after the branch-
ing of reptiliomorph tetrapod lines were direct consequences of
their common history, and from an operational point of view the
origin of the amniote egg can be regarded as single. In this light,
the ease with which all reptiliomorph lines can be derived from
a hypothetical group no higher than family, whose basis is the genus
Mauchchunkia, obviates for the moment the vexed question of the
polyphyletic origin of major groups of reptiles.

NICHOLAS HOTTON III

No. 12

REFERENCES CITED

Carroll, R. L., 1969, Problems of the origin of reptiles: Biol. Rev., Cambridge,
v. 44, p. 393-432.

1970, The ancestry of reptiles: Roy. Soc. [London] Philos.

Trans., ser. B, v. 257, no. 814, p. 267-308.

Noble, G. K., 1931, The biology of the Amphibia: New York, McGraw-Hill,
573 p.

Olson, R., 1951, Size relations in the limb bones of Buettneria perfecta : Jour.
Paleont., v. 25, no. 4, p. 520-524.

Panchen, A. L., 1964, The cranial anatomy of two Coal Measure anthracosaurs:
Roy. Soc. [London] Philos. Trans., ser. B, v. 247, p. 593-637.

■ 1966, The axial skeleton of the labyrinthodont Eogyrinus

attheyi: Jour. Zool., London, v. 150, p. 199-222.

1967, The homologies of the labyrinthodont centrum: Evo-
lution, v. 21, no. 1, p. 24-33.

Panchen, A. L., and A. D. Walker, 1961, British Coal Measure labyrinthodont
localities: Ann. Mag. Nat. Hist., ser. 13, v. 3, p. 321-332.

Romer, A. S., 1947, Review of the Labyrinthodontia: Mus. Comp. Zool. Bull.,
v. 99, p. 1-368.

1957, The appendicular skeleton of the Permian embolomerous

amphibian Archeria : Mus. Paleont. Contrib., Univ. of Mich., v. 8, no. 5, p.
103-159.

1964, The skeleton of the Lower Carboniferous labyrinthodont

Pholidogaster pisciformis: Mus. Comp. Zool. Bull., v. 131, no. 6, p. 129-159.

1969, A temnospondylous labyrinthodont from the Lower

Carboniferous: Kirtlandia, no. 6, p. 1-20.

1970, A new anthracosaurian labyrinthodont, Proterogyrvnus

scheelei, from the Lower Carboniferous: Kirtlandia, no. 10, p. 1-16.

Romer, A. S., and L. I. Price, 1940, Review of the Pelycosauria: Geol. Soc.
Amer. Sp. Pap., no. 28, p. 1-538.

Save- Soderbergh, G., 1932, Preliminary note on Devonian stegocephalians from
East Greenland: Meddelelser om Gr0nland, v. 94, p. 1-107.

Tilton, J. L., 1928, Geology from Morgantown to Cascade, West Virginia, along
State Route number 7: West Virginia Univ. Sci. Assoc. Bull., v. 2, no. 3,
p. 65-86.

Weller, J. M. (Chairman) et al., 1948, Correlation of the Mississippian forma-
tions of North America: Geol. Soc. Amer. Bull., v. 59, p. 91-196.

Westoll, T. S., 1943, The origin of the tetrapods: Biol. Rev., Cambridge, v. 18,
p. 78-98.

MANUSCRIPT RECEIVED JUNE 29, 1970

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO APRIL 20, 1971 NUMBER 13

THE GIRDLED ROAD SITE, AN EARLY WOODLAND HUNT-
ING STATION IN LAKE COUNTY, OHIO

DAVID S, BROSE1

Associate Curator of Anthropology
Cleveland Museum of Natural History

ABSTRACT

Archaeological survey of a proposed reservoir area in Cascade
Valley, Lake County, Ohio, has revealed a small single -component
site along the bluffs bordering Big Creek. Analyses of the recov-
ered artifacts and the nature of their topographic and stratigraphic
location indicate that the site represents a small seasonal hunting
camp occupied early in the first millenium B.C.

In May, 1969, a preliminary archaeological survey of a portion
of southern Lake County, Ohio, was undertaken by the author for
the Cleveland Museum of Natural History. The area investigated
lay within the proposed expanded boundaries of the Girdled Road
Reservation of the Lake County Metropolitan Park District in
Leroy and Concord townships. The proposed reservation area is
bounded roughly by Winchell Road, State Route 608, and Williams
Road to the west, Huntoon Road to the north, Callow Road on the
east, and the Lake-Geauga county line to the south.

Topographically the area consists of a gently rolling portion of
the glaciated Allegheny Plateau, dominated by the deeply incised
valley of Big Creek and its tributaries, Aylworth Creek and East
Creek. The latter stream represents a more juvenile (less en-
trenched) drainage system which appears to have been captured
recently by Big Creek. East Creek presently joins Big Creek about
60 yards below the 18-foot Cascade Falls on Big Creek. This is the
most steeply entrenched portion of the valley, and Big Creek flows

1 Department of Anthropology, Case Western Reserve University

1971

THE GIRDLED ROAD SITE

approximately 150 feet below the plateau surface. This is also the
widest portion of the flood plain and is called Cascade Valley. It is
approximately a quarter of a mile wide from the junction of East
Creek for about three-quarters of a mile downstream (northwest) .
Two miles below this deep valley, Big Creek flows out onto the
lake plain and into Grand River approximately 5 miles above the
river’s entrance into Lake Erie (fig. 1) . This deep valley described
above was the proposed site of a dammed lake, and archaeological
survey was concentrated in that area.

ARCHAEOLOGICAL RECONNAISSANCE

Several days of surface collection and archaeological testing
were spent in Cascade Valley itself. With the exception of random-
ly distributed historic artifacts from the nineteenth and twentieth
centuries, no cultural material was recovered within the Cascade
Valley. The steep (over 40°) banks of the valley were too denuded
by sheet runoff to offer much hope of finding in situ cultural ma-
terial. It was also felt that the gradient itself should have prohib-
ited occupation. Several test-pits each measuring 5 by 5 feet were
excavated on the more gentle slopes along the western side of Cas-
cade Valley but produced no cultural material. Archaeological test-
ing along the east rim of the valley was limited to the undisturbed
southeastern portion lying north of East Creek. A total of four
5 by 5 feet test pits were excavated which produced no cultural
material. On the southernmost promontory on the eastern rim,
isolated on the south by East Creek and Cascade Falls and on the
north by a more gentle fifty-foot ravine cut by seasonal streams,
a single short-term component was encountered at a depth of 18
inches below the surface (fig. 2) .

GEOMORPHOLOGY

The occupation encountered in these tests could be followed
through three 5 by 5 feet excavation units where it was defined
as a thin discontinuous midden deposit lying in erosional depres-
sions on the surface of a poorly developed soil horizon (fig. 3) .
The underlying grayish-brown calcareous sands displayed a poorly

^ Fig. 1. Map showing location of Girdled Road site. Star indicates site.

DAVID S. BROSE

No. 13

Fig. 2. Cross section of Cascade Valley traversing Girdled Road site.

sorted particle-size distribution characteristic of lacustrine deposi-
tion (Friedman, 1961, p. 514-529) , probably representative of dune
formation along Lake Cuyahoga about 13,000 years ago (Rau, 1969,
p. 25-29) . With the draining of this proglacial lake after 11,000 B.P.
a period of warmer weather favored the rapid development of an
“inceptisol” with a weak illuvial horizon (U. S. Soil Conservation
Service, 1960) . It was on this surface that the human occupation
took place. Post-occupational geomorphological activity resulted in
the truncation of elevated portions of the midden and burial be-
neath the deposition of loam eroded from further upslope. A period
of lessened precipitation appears to have existed as seen in the de-
velopment of a clear A2 horizon containing fragile clay minerals
(Bunting, 1965) immediately overlying the cultural materials. The
present ground surface represents a historically disturbed (Ap)
horizon.

The only pedological data of immediate concern are the (1:1)
pH values for the soil horizon noted. The lowest (Cj) horizon has
a neutral value of 6.8 while the succeeding horizons range between
4.3 and 5.7. The occupational midden itself gave pH values from

1971

THE GIRDLED ROAD SITE

Fig. 3. Stratigraphic profile of test pits 17 and 18 excavated at Girdled Road
site. Dark lenses represent cultural materials.

4.9 to 5.2. These rather acid soils clearly are a recent result of
downward percolation of humic-acid-enriched rainwater (U.S.
Bureau of Plant Industry, Soils and Agricultural Engineering,
1962) and have had the result of destroying most of the faunal
material from this site.

CULTURAL MATERIALS

The distribution of cultural materials within the occupation
area was quite disappointing. The thickness of the deposits and
their lack of internal stratification indicate a single component of
rather limited duration. Erosion has removed all but isolated por-
tions of what was presumably a continuous deposit. While these
diagenetic processes have rendered meaningless any attempts at
analyzing the spatial distribution of the artifacts recovered, the
total observed extent of the area displaying these deposits may
provide some estimate of population. This point will be dealt with
in the interpretive section of this report.

No features were noted at the Girdled Road site. While areal
clusters of artifacts or charcoal were probably present, diagenetic
erosion has removed any indications of these. Subsurface features,

DAVID S. BROSE

No. 13

however, should have been recoverable. Their absence thus con-
firms the suggestion of an occupation of limited duration.

Recovered artifacts consisted of three stemmed projectile points
(fig. 4 A, C) ; two large ovate bifacial knives (fig. 4 B, D) , a large
bifacial sidescraper (fig. 4 E) ; a broken slate gorget or celt (fig.
4 F) ; 104 chert and flint chips; and two rolled copper beads (fig.
4 G).

5 CM.

Fig. 4. Artifacts from the Girdled Road site. A, C, projectile points; B, D,
bifacial blades; E, scraper; F, celt; G, rolled copper beads.

CHIPPED STONE

The chipped stone artifacts offer the best means of assigning
relative temporal placement to this component. Their metric attri-
butes are presented in table 1.

The entire lithic assemblage is typologically consistent and
argues for an Early Woodland occupation. The stemmed points are
morphologically and metrically analogous to the flat-based “Cresap
points” which Dragoo (1963, p. 109-113) recovered from the Cresap

METRIC ATTRIBUTES OF CHIPPED STONE ARTIFACTS
(In millimeters)

THE GIRDLED

ROAD SITE

■+->

’C

H-»

•S

-H

H->

-M

H->

M 0

W ft

ij +3

t—

1/5

■ft

-+-»

t-;

CM*

110

i— <

tip broken

DAVID S. BROSE

No. 13

Mound in West Virginia. Similar points are reported from numer-
ous Adena Phase burial mounds throughout the upper Ohio valley
(Dragoo, 1963, p. 118-121; Webb, 1942, p. 335). In northern Ohio
they are associated with components of the coeval Leimbach Phase
(Shane, 1967) . The ovate blades from the Girdled Road site are
morphologically and metrically analogous to “Adena leaf-shaped”
blades or points. Such artifacts are a common type on Adena Phase
sites in the Ohio valley (Dragoo, 1963, p. 107-108, Webb and Snow,
1945, p. 82) and in northern Ohio during the Leimbach Phase
(Shane, 1967, p. 114-116) . The large heavy sidescraper, although
less useful as a horizon marker, is also a common artifact on Adena
or Leimbach Phase sites throughout the state (Dragoo, 1963, p. 117-
121, Shane, 1967, p. 116-117, 156, 160) .

The chipping debris recovered from the Girdled Road site
(table 2) appeared randomly scattered throughout the truncated
midden deposit. No clusters of debitage were noted in the remain-
ing portions of the deposit. The entire midden level was hand-
screened through 3/32 inch hardware cloth and numerous small
chips, often missed in conventional inch screens, were recovered.
All of the typological categories of chippage are indicative of final
flaking or resharpening activities. None of the initial preparation
flake categories such as cores or decortication flakes were recovered
from this component (cf. Brose, 1970a).

All of the chipped stone artifacts were examined under a low
power (45 X) binocular microscope in an attempt to interpret their
function. The three stemmed points exhibited moderate to light use
on edges characterized by resolved or step-flaking. While all of
these points exhibited a slight sheen or polish, this was extremely
intermittent and occurred both on the flake-scar surfaces and on
the interlying ridges. No evidence of use striae was observed. The
ovate blades exhibited some use polish on both faces, usually con-
fined to within 7 mm of the edge. This gloss was rather continuous,
and most evident on the ridges separating flake-scars. Both ovate
blades were characterized by numerous more or less parallel longi-
tudinal striations on both faces of both edges. These striations were
confined to a zone within 5 mm of the edge, extending from the tip
about two-thirds of the distance along the edge of the blade. The
bifacial scraper showed numerous transverse striations on the
steeply retouched curved edge and faint parallel longitudinal stria-

1971

3 P

O <

>Ph

P H

C/2<
W P

ufe
£1*
C£3 00

O’

Pft

THE GIRDLED ROAD SITE

C/3

bJO

bfl

T~ 1

iri

i— i

v — ✓

> — ✓

t— 1

■"Cf

§§!

bD bfl
10 00

CO cO

bs bh

CM CM
P CM

HH Eh

pQp

bfl

«! M

s s

CO CO
ft f-i

be bD

00 O;
CM od

C/2

ffi

H W

& *
r p

H <j

S C

3 <
p fci
E 3

5 o

P c/2

TOTALS 46 (43.3 grams) 36 (15.7 grams) 14 (11.3 grams) 8 (2.6 grams) 104 (72.9 grams)

DAVID S. BROSE

No. 13

tions along the straight edge. Numerous areas of high-use polish
were observed on both faces of this artifact.

The comparison of wear patterns observed on these artifacts
with experimentally produced data (Semenov, 1964) provides the
functional interpretation of these artifacts with a considerable de-
gree of confidence. The stemmed points display evidence of use
fully consistent with their functional interpretation as projectile
points, probably used in mammal hunting. The ovate blades show
evidence of wear characteristic of hafted knives used for cutting
rather soft material which occasionally encountered a more resist-
ant material. The bifacial scraper seems to show evidence of use
characteristic of two separate functions. It appears to have been
used as a hand-held scraper for hides, and as a backed hand-held
knife, probably for skinning.

There is no evidence to support the contention that the ovate
blades represent preforms or an earlier stage in the production of
the stemmed projectile points. The ovate blades display distinct
wear patterns indicating that they functioned as a finished artifact.
In addition, the stemmed points display no wear patterns consistent
with their earlier use as knives. The two classes thus seem to rep-
resent functionally, as well as typologically, distinct artifacts.

The entire chipped-stone complex is clearly indicative of hunt-
ing and butchering activities with occasional refurbishing of knives
and projectile points but no actual tool manufacture. The analyses
of raw materials utilized in artifact manufacture has lead to the
recognition of four distinct sources. These sources are Flint Ridge
Flint, Plum Run Flint, Devonian Outcrop Cherts, and glacially de-
rived Pebble Cherts. The mottled gray and brown pebble cherts
occur commonly in the till and outwash formations of the glaciated
Allegheny Plateau region of northeastern Ohio. The Devonian Out-
crop Cherts represent a variety of light tan cherts common in nod-
ules or beds in formations pertaining to the Devonian System.
Around the western end of Lake Erie such outcrops in Dundee
Limestone formations are noted in Hillsdale, Lenawee, and Monroe
counties in Michigan, and in the Delaware Limestones in Lucas,
Ottawa, Erie, and Sandusky counties in Ohio (Brose [in press];
Stout and Schoenlaub, 1945) . The Plum Run Flint, which appears
to be a facies of Upper Mercer Flint, and the Flint Ridge Flints
have been discussed extensively in the literature of Ohio archae-

1971

THE GIRDLED ROAD SITE

ology. The comparison of raw materials utilized in artifact manu-
facture and raw materials exhibited in the chipping debris recov-
ered from the Girdled Road site (table 3) clearly indicates that

TABLE 3

LITHIC PRODUCTION AGAINST RAW MATERIAL CATEGORIES
BY WEIGHT (in grams)

RAW MATERIAL

FLINT

RIDGE

FLINT

PLUM

RUN

FLINT

GLACIAL

PEBBLE

CHERTS

DEVONIAN

OUTCROP

CHERTS

ARTIFACTS

Observed

62.7

18.5

16.2

9.5

Expected

63.0

20.4

16.4

7.2

DEBITAGE

Observed

43.3

15.7

11.3

2.6

Expected

43.0

13.9

11.1

4.9

X2 = 2.2339; df = 3; p = .50; 0 = .0124; n = 179.8

there is no statistical significance to the variations observed. Given
the size of the sample, the observed variations could be expected by
chance alone nearly fifty times out of a hundred if only a single
population were involved. The implications that the recovered
chipping debris was derived from the recovered artifacts seems to
be well founded. Certainly, neither all of the artifacts utilized at
the site, nor all of the debitage has been recovered. It is still quite
clear that we would not expect further samples to show radically
different proportions of the source materials already noted.

DAVID S. BROSE

No. 13

GROUND STONE

A single broken rectangular celt, manufactured from Bedford
Shale, was also recovered from the midden deposits. The celt is
65.5 mm long, with a maximum width of 44.0 mm and a maximum
thickness of 8.9 mm. Heavy transverse striations were noted on
both faces of the 11.2 mm portion of working edge which remained
intact, making it probable that the implement functioned more as
an axe than as an adze. The strong bedding planes within the
rather brittle shale make it appear unlikely that the celt was in-
tended for woodworking. Similar rectangular celts occur occasion-
ally throughout the Ohio area in Early Woodland contexts (Dragoo,
1963; Shane, 1967) although they are usually manufactured from
igneous or metamorphic rock. Small rectangular celts of sedimen-
tary rock are increasingly common in later periods in this region.

COPPER

Two small rolled copper beads recovered from this component
have analogs in the Great Lakes-Ohio valley region from the Late
Archaic (2000 B.C.) through the transition to Late Woodland (A.D.
500). Both beads are quite similar, having been manufactured by
cold-hammering native copper (common throughout the Upper
Great Lakes) into flat sheets, then folding and cold-hammering
these sheets into more or less rectangular strips. The rectangular
strips would then be wrapped about some sticklike object and cold-
hammered to form an overlapping rolled copper bead. All ductility
had been lost at this stage of the manufacturing process. The larger
bead with an internal diameter of 2.9 mm and an external diameter
of 11.1 mm had been formed of a rectangular strip which was 17.4
mm long, 7.0 mm wide, and 2.5 mm thick. The smaller bead with
a finished internal diameter of 2.5 mm and an external diameter of
10.2 mm had been rolled from a rectangular strip 16.8 mm long,
6.33 mm wide, and 2.3 mm thick.

Copper beads of this type represent a common artifact form in
Early Woodland sites throughout the region. They are ubiquitous
on those sites in the Ohio valley identified as Adena (Dragoo, 1963,
p. 121-123; Solecki, 1952, p. 370; Webb and Snow, 1945, p. 99-100;
Bache and Satterthwait, 1930, p. 140) . They have also been recov-
ered from most of the major components of the coeval Leimbach
Phase in northern Ohio (Shane, 1967, p. 23) .

1971

THE GIRDLED ROAD SITE

EXTERNAL CULTURAL RELATIONSHIPS
TYPOLOGY

The entire material culture assemblage from the Girdled Road
site is clearly related to the Early Woodland materials within Ohio.
Recently there has been an attempt made to denigrate McKernian
nomenclature and to refer to archaeological manifestations in terms
of phase and tradition as expounded by Caldwell (1958) . In this
region, Prufer and others (1965) have discussed the regional Scioto
Tradition which seems to extend over the entire Great Lakes-Ohio
River drainage system. Within this Tradition, Shane has distin-
guished two contemporary phases during the first millenium B.C.:
the Adena Phase in the Ohio drainage basin, and the Leimbach
Phase to the north, with the only distinction between these phases
based on minor ceramic variations. A critical discussion of these
taxonomic questions already exists (Fitting and Brose [in press])
and need not be repeated here. What is important is the recognition
of the close relationship between these two contemporaneous phases
which should probably be considered a single phase of two distinct
traditions within a Scioto Co-Tradition (cf. Willey and Phillips,
1958) . As Shane himself has noted, the differences are merely geo-
graphic.

Typologically, all of the Girdled Road site material can be con-
sidered characteristic of materials more generally referred to as
“Adena.” There are several problems with this approach, however.
As Webb and Baby (1957, p. 32) noted, one of the more salient
features of the Adena People on their own turf is their skill at
disguising information pertaining to their settlement system. Our
knowledge of Adena architecture is primarily from charnel houses.
Our understanding of the role which factors of topography and geog-
raphy played in the patterning of sites is largely confined to some
knowledge of the location of their more elaborate ceremonial
mounds. Even our conception of the material culture of the Adena
People is generally restricted to grave goods. Our appreciation of
the extractive economy (or economies) which supported the Adena
mortuary complex is severely limited and virtually worthless in
reconstructing any kind of cultural ecology. The supposition that
several Kentucky rock shelters were utilized as temporary hunting
camps is inferential. “What people, other than the Adena Indians,
could have been in these shelters, so close to the region of Adena

DAVID S. BROSE

No. 13

permanent residence, at that period?” (Webb and Baby, 1957, p.
34).

The Early Woodland period in northern Ohio was quite ob-
scure until the recent excavations and restudy of extant collections
by Orrin C. Shane III (1967) . Shane has documented a number of
sites located in the south-central portion of the Lake Erie drainage
basin which display clear affinities with the more southern classic
Adena mounds. The type site of this Phase is the Leimbach site on
the Vermilion River in Lorain County, Ohio. Here Shane exca-
vated a large midden (estimated to represent an occupation area
of more than 15,000 square feet) which varied between three and
eight inches in thickness. The midden contained numerous features,
one of which indicated a circular structure 40 feet in diameter.
This midden contained considerable amounts of ceramics which
Shane has assigned to Fayette Thick, Adena Plain, and two new
types, Leimbach Thick and Leimbach Cordmarked. The Leimbach
ceramics clearly are a local variation of the general Early Wood-
land ceramic assemblage throughout the northeast. This is the
sense in which Shane defines the Leimbach Phase and a coeval
Adena Phase.

One of the major contributions of Shane’s Leimbach excava-
tions is his demonstration that the number and size of features in
the Early Woodland occupation implies a semipermanent small
village in the settlement system. Of the three radiocarbon samples
submitted, the two noncontaminated ones provide dates of 540 B.C.
± 309 years (Shane, 1957, p. 136) thus establishing the chronolog-
ical position of the site. At this same time period (as determined
from ceramic affinities) Shane also recognizes several other related
components on northern Ohio. The lower levels of the Mixter site
in Erie County represent a hunting station contemporary with
Leimbach. The earlier collections from the Burrell Fort site on
French Creek in Lorain County, Ohio, were reanalyzed and a por-
tion of the site was re-excavated by Shane to reveal an early com-
ponent with material similar to the Leimbach site midden. The
Seaman’s Fort site on the Huron River, Erie County, Ohio, exca-
vated since 1945 by R. Vietzen was visited by Shane in the summer
of 1966. While the site was badly disturbed, Shane was able to
analyze earlier collections and concluded that there was an early
component at Seaman’s Fort, again coeval with the Early Woodland
occupation at Leimbach. He further noted (1967, p. 158) that the

1971

THE GIRDLED ROAD SITE

site was probably functionally similar to Leimbach in that it also
represented a number of seasonal semipermanent occupations. An
analysis of ceramics in the Ceramic Repository at the University
of Michigan Museum of Anthropology lead to the recognition of
another small component of the Leimbach Phase from the Mohawk
Park Rock Shelter in Geauga County, Ohio.

In addition to Shane’s analyses, numerous radiocarbon dates
exist for Adena burial mounds within the Ohio River drainage sys-
tem. Again it should be emphasized that Adena, as this term is
presently used, is certainly not a culture in spite of Shetrone’s
(1920) original suggestion that it represented one tribe or nation.
The succeeding decades, with the collection of data in a haphazard
and atheoretical manner, have done little for the Adena concept
other than to make Shetrone’s definition less acceptable. The
“traits” of Adena may be technological or economic, but for the
most part are ideological. As Griffin (1948) pointed out a quarter
of a century ago, a mortuary complex is not a culture.

Adena as a mortuary complex must be separated from Adena
as a settlement system and Adena as a nonmortuary style zone, the
latter perhaps representing some strain of ethnic unity similar to
that suggested by Shetrone. Even from preliminary analysis it is
clear that the distribution of these three aspects of Adena are not
coterminous.

Dragoo (1963, p. 288-297) has presented a large series of radio-
carbon determinations relating to the problem of placing Adena in
some absolute chronological framework. He strongly states that
“. . . on the basis of typology and stratigraphy, I seriously doubt
that any Adena site in the Ohio Valley would have been in exist-
ence much after A.D. 1 or perhaps even earlier.” (ibid. p. 289)
While this statement may be a bit too definite, one could note that
many of those dates after A.D. 1 were based on the carbon-black
laboratory procedure and are therefore rather untrustworthy. With
the exception of the Drake, Florence, and Cowan Creek Mounds,
all the Adena dates seem to fall within the first 1500 years B.C.,
and recent evaluations of fluctuations in the upper atmospheric
production of C14 make it likely that those radiocarbon dates around
the time of the birth of Christ should be pushed back, perhaps as
much as several hundred years. In New York, Ritchie (1969, p. 170-
178, 181-201) places his Orient Phase at 1000-700 B.C. and the
Meadowood Phase at 1000-500 B.C. The former phase has most of

DAVID S. BROSE

No. 13

the artifacts and burial customs common to Adena. The latter phase
is distinguishable from Red Ocher sites in Michigan only by (1)
the geographic location, (2) the fact that Meadowood burials con-
tain more indications of widespread trade in luxury items, and (3)
definite associations with Early Woodland ceramics. The Glacial
Kame “culture” also appears to fall within this period of the first
1500 years B.C. and it is becoming increasingly clear that many of
the “later” Old Copper manifestations do as well. (Halsey, 1966) .

It may be of some interest to note that the area of western
Lake Erie does not seem to show much evidence of occupation at
this time. Survey work in southeastern Michigan by Brose and
Fitting, and the surveys of northeastern Ohio by Earl J. Prahl of
the University of Toledo have revealed an apparent hiatus in the
occupation of the old lake plain area from the end of the Middle
Archaic period (ca. 1500 B.C.) until the transitional Middle-Late
Woodland period (ca. A.D. 700). The significance of this negative
evidence which is also duplicated in southwestern Ontario (John
Lee, personal communication) will perhaps become apparent when
some reasonable approximation of Early Woodland settlement sys-
tems is obtained.

At any rate, it seems clear that the Girdled Road site repre-
sents an occupation during this ill-defined and poorly understood
period when Late Archaic and Early Woodland burial cults were
flourishing throughout the eastern United States. The conclusions
which can be drawn from the analyses of this site may perhaps
dispel some of the confusion, at least in the limited area of north-
eastern Ohio.

CONCLUSIONS

While diagenetic soil conditions prevented the recovery of any
significant sample of faunal remains, some inferences as to economic
adaptation can be made from other data. The stratigraphic data,
the topographic location of the Girdled Road site, and the lithic
materials recovered all argue strongly for a limited-duration occu-
pation. Although the midden itself was discontinuous, the maxi-
mum boundaries of occupation were about 20 feet by 15 feet. The
rough ellipse appears to have enclosed an area of approximately
235 square feet. On the basis of modern ethnographic accounts
pertaining to the Indians of the Upper Great Lakes, this would
have been an area of floor space within a structure sufficient for

1971

THE GIRDLED ROAD SITE

from three to eight individuals (cf. Brose, 1970b). Since no post-
molds or features were encountered at the Girdled Road site, it
would seem that no such structure was erected.

The total absence of ceramics also argues for a short-term
occupation, probably by a group whose sexual composition was
unbalanced with few (if any) females represented. The above data
point rather clearly to the Girdled Road site as an example of
a temporary hunting station occupied by a group of males. There
was a surprisingly large number of unbroken artifacts recovered
from approximately one-third of the total available site area. This
probably indicates that the site was occupied several times for
short periods. In this respect one might note that Dragoo (1963)
has indicated that the variations in projectile point morphology
reflected in the sample from Girdled Road may be chronologically
significant.

The Girdled Road site thus represents an Early Woodland
hunting station similar to other northern Ohio sites such as the
lower levels at Mixter, and the Burrell Fort site (Shane, 1967).
The analysis of lithic sources indicates that the occupants of the
Girdled Road site either moved seasonally throughout the north-
eastern third of the state of Ohio to obtain these materials, or were
in contact with numerous localized groups located therein. While
it is probable that exotic material such as Lake Superior copper
was obtained by trade of some sort (Fitting and Brose [in press])
it is not likely that low-quality cherts and flints would have entered
such a network. The implications are that the group experienced
considerable seasonal mobility as a result of their economic adap-
tation. While this adaptation is imperfectly known, it must have
included fairly large semipermanent or permanent villages such as
Leimbach, as well as seasonal special-purpose extractive camps,
presumably for hunting large mammals. There is no evidence for
seasonal fishing or waterfowl-collecting camps, although too little
site survey has been performed along the lake shore to evaluate
this negative evidence. The evidence for horticultural or agricul-
tural activity at this time level is not only extremely limited
(Vickery, 1970) but is equivocal at best. The total settlement sys-
tem of these populations thus remains rather ambiguous. Not until
considerable further excavation has been done will we be able to
discuss the inter-relationship of culture and ecology during the
Early Woodland period.

DAVID S. BROSE

No. 13

REFERENCES CITED

Bache, C. and L. Satterthwait, Jr., 1930, The excavation of an Indian mound at
Beech Bottom, W. Va.: Univ. Pennsylvania Mus. Jour., v. 21, p. 132-187.

Brose, D. S., 1970 a, The archaeology of Summer Island: Changing settlement
systems in the northern Lake Michigan region: Univ. Michigan Mus. An-
thropology Anthrop. Papers, no. 41.

1970 b, Prehistoric cultural ecology and social organization in

the northern Lake Michigan area: Case Western Reserve Univ. Anthro-
pology Studies, v. 1, no. 1.

[in press] The archaeology of Monroe County, Michigan: The

Michigan Archaeologist.

Bunting, B., 1965, The geography of soils: Chicago, Aldine Press.

Caldwell, J., 1958, Trend and tradition in the prehistory of the eastern United
States: Am. Anthrop. Assoc. Mem. 88.

Cunningham, W. M., 1948, A study of the glacial kame culture in Michigan,
Ohio, and Indiana: Univ. Michigan Mus. Anthropology Occasional Contrib.,
no. 12.

Dragoo, D. W., 1963, Mounds for the dead: an analysis of the Adena Culture:
Carnegie Mus. Ann., v. 37.

Fitting, J. E. and D. S. Brose [in press] The northern periphery of Adena,
in Swartz, B. K. ed., The Adena Culture: Muncie, Ball State University
Press.

Friedman, G. M., 1961, Distinction between dune, beach, and river sands from
their textural characteristics: Jour. Sed. Petrology, v. 31, p. 514-529.

Griffin, J. B., 1948, An interpretation of the Glacial Kame Culture: Univ.
Michigan Mus. Anthropology Occasional Contrib., no. 12, p. 46-51.

Halsey, J. R., 1966, Radiocarbon dates from archaeological sites of Old Copper
and related cultures in the Great Lakes area: Artifacts, v. 4, no. 4, p. 7-11.

Prufer, O. and others, 1965, The McGraw Site: A study in Hopewellian dy-
namics: Cleveland Mus. Nat. Hist. Sci. Pubs, n.s., v. 3, no. 1.

Rau, J. L., 1969, The evolution of the Cuyahoga River: its geomorphology and
environmental geology, in Cooke, G. E. ed., The Cuyahoga River Water-
shed: Inst. Limnology and Dept. Biol. Sci., Kent State Univ., p. 9-40.

Ritchie, W. A., 1969, The archaeology of New York State: New York, Natural
History Press.

Semenov, S. A., 1964, Prehistoric technology. English translation by M. W.
Thompson: New York, Barnes and Noble.

Shane, O. C. Ill, 1967, The Leimbach Phase and its position in eastern North
American prehistory: Case Inst. Technology, PhD dissert, (unpub.).

Shetrone, H., 1920, The culture problem in Ohio archaeology: Am. Anthro-
pologist, v. 22, no. 2, p. 144-172.

Solecki, R., 1952, Exploration of an Adena mound at Natrium, W. Va.: Bur.
Am. Ethnol. Bull. 151, Anthrop. Paper no. 40, p. 313-395.

Stout, W. and R. A. Schoenlaub, 1945, The occurrence of flint in Ohio: Ohio
Geol. Surv. 4th ser. Bull. 46.

1971

THE GIRDLED ROAD SITE

U. S. Bureau of Plant Industry, Soils, and Agricultural Engineering, 1962, Soil
Survey Manual: U. S. Dept, of Agriculture.

U. S. Soil Conservation Service, 1960, Soil classification, a comprehensive sys-
tem: U. S. Dept, of Agriculture.

Vickery, K. D., 1970, Evidence supporting the theory of climatic change and
the decline of Hopewell: The Wisconsin Archaeologist, v. 51, no. 2, p.
57-76.

Webb, W. S., 1942, The C. and O. mounds at Paintsville, Sites Jo 2 and Jo 9,
Johnson County, Ky.: Univ. Kentucky Anthropology and Archaeology
Repts., v. 5, no. 6, p. 505-579.

Webb, W. S. and R. S. Baby, 1957, The Adena People, No. 2: Columbus, Ohio
Historical Society.

Webb, W. S. and C. E. Snow, 1945, The Adena People: Univ. Kentucky An-
thropology and Archaeology Repts., v. 6.

Willey, G. and P. Phillips, 1958, Method and theory in American archaeology:
Chicago, Univ. of Chicago Press.

MANUSCRIPT RECEIVED DECEMBER 17, 1970

< ui

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO MAY 3, 1971 NUMBER 14

NOTES ON AN ADDITION TO THE FISH FAUNA OF THE
MOWRY SHALE (CRETACEOUS) OF WYOMING

DAVID H. DUNKLE
ABSTRACT

The incomplete remains of the skull and skeleton of a diminu-
tive ray-finned fish from the Mo wry Shale (Middle Cretaceous)
near Cody, Park County, Wyoming, is described as a previously
unrecognized addition to the Mo wry fish fauna. The basic organi-
zation of morphologic characters suggests tentative assignment of
the specimen to the leptolepiform genus Clupavus.

INTRODUCTION

During the course of the 1970 American Western States Heri-
tage Tour, co-sponsored annually by the Cleveland Museum of Nat-
ural History and the Martha Holden Jennings Foundation of Cleve-
land, a brief examination was made of exposures of the Middle
Cretaceous Mo wry Shale in the vicinity of Cody, Wyoming. On that
occasion one of the youthful participants of the tour, Dale Shisler
of Bartlesville, Oklahoma, recovered a partial but articulated skele-
ton of a diminutive fish. The Mowry Formation and equivalent
strata of Wyoming and adjacent Rocky Mountain states have long
been known for their myriads of disassociated scales and other occa-
sional elements of fishes. The principal contributor to the descrip-
tion of some of these has been T. D. A. Cockerell (1919) . In view
of widely recognized difficulties in reasonably accurate interpreta-
tion of macerated skeletal parts, the presently described specimen
is indicative of the continuing incompleteness of information about
the Mowry fauna and emphasizes the importance of more complete,
associated, but undescribed materials from the formation also avail-
able in various of the nation’s museums.

DAVID H. DUNKLE

No. 14

The specimen (Cleveland Museum of Natural History 11045)
was encountered in Mowry outcrops immediately west of State
Route 120, 1.7 miles south of the Cody Airport in Park County.
Various stratigraphic sections of the formation in the Cody area and
their lithologies have been described by Reeside and Cobban (1960).
Displayed are the head, in left lateral aspect, with an attached
series of 27 vertebrae, recurved acutely and forwardly over the
skull roof. Little actual osseous tissue is preserved. However, the
sharp and distinct impressions in the highly siliceous, fine-grained
matrix of, in part, both internal and external surfaces of neuro-
cranial and visceral elements of the skeleton permits observation of
a significant number of basic morphological characteristics. These
are here interpreted as denoting definite leptolepiform affinity and,
in fact, within this order of halecostome fishes tentative reference
of the specimen to the genus Clupavus Arambourg (1950) is sug-
gested. Because of ignorance of caudal fin structure and other per-
tinent details, specific diagnosis of the fish is considered unwar-
ranted, at this time.

Deep appreciation is expressed to Dale Shisler for his donation
of this interesting specimen to the Cleveland Museum of Natural
History. The photograph and drawings have been provided by the
Cleveland Museum staff members Bruce Frumker and Barbara
Gardiner, respectively.

DESCRIPTION

The total preserved length of the specimen is 28.6 mm of which
the head occupies 10.6 mm. The poster odorsal border of the skull
roof is not preserved at the midline, but this transverse plane, as
reconstructed, is presumed to mark the deepest dimension of the
head, with a measurement little more than two-thirds of the head
length. Neural spines and ribs are delicate and short. These fea-
tures combine to indicate a very slenderly fusiform body habit.

Fig. 1. Clupavus sp. (CMNH 11045) from the Middle Cretaceous Mowry Shale
near Cody, Park County, Wyoming. (A) Photograph and (B) habit sketch of
the specimen, as preserved. Reproduction approx. X 6.6. Explanation of ab-
breviations: Ausph, autosphenotic; ?Brs, branchiostegal; Ecpt, ectopterygoid;
Enpt, entopterygoid; Fr, frontal; Io, infraorbital; lop, interoperculum; La,
lacrimal or preorbital; Mn, Mandible; Mx, Maxilla; Op, operculum; Pmx, pre-
maxilla; Pop, preoperculum; Psph, parasphenoid; Qu, quadrate; Smx, supra-
maxilla; Sop, suboperculum; Sym, symplectic.

1971

NEW FISH FROM THE MOWRY SHALE

DAVID H. DUNKLE

No. 14

In lateral aspect, the skull would appear to have only a slight
dorsal convexity. Preorbital and postorbital lengths of the neuro-
cranium are short. The relatively large, ovate orbit has an axial
length one-third that of the head from snout to posterior margin
of the opercular apparatus. The quadratomandibular articulation
is situated below the midorbital length and the mandibular sym-
physis is prominent.

Of the bones of the skull roof, the impression of the smooth
superior surface of the right frontal is to be observed almost in its
entirety. Indicated are extremely long and attenuated elements.
The bone bears the supraorbital sensory canal, which in the post-
orbital region branches into a mesial, parietal extension and a
lateral connection with the infraorbital and main supratemporal
canals. The lateral margin of the skull roof in the otic region is ob-
scured by remnants of adhering bone. These, however, undoubt-
edly represent portions of the dermosphenotic, autosphenotic and
pterotic bones. In this area there is rather wide separation and no
conjunction of the dorsal extension of the preopercular sensory
canal with the supraorbital and infraorbital canals. Dorsolaterally
the posterior edge of the pterotic is adjoined by an axially short and
somewhat transversely elongated extrascapular. Visible portions
of the dorsal margin of the parasphenoid exposed above the medial
edge of the entopterygoid traverse the orbit horizontally.

The bones surrounding the mouth, as preserved, are distended
upwardly and forwardly. The oral length of the premaxilla is
scarcely one-fifth that of the maxilla, which has a constricted
proximal part and a ventrally convex expanded distal portion, sur-
mounted with anterior and posterior supramaxillae. The dentition
as revealed by the internal impression of a maxilla is reduced to
a narrow oral band of minute, clustered teeth. The oral margin of
the mandible rises rapidly from the robust symphysis to a high
coronoid process.

The infraorbital series consists presumably of five elements.
The impressions of radiating sensory-canal grooves suggest a mod-
erately expanded lachrymal or preorbital bone as does the pre-
served evidence of the outline of the posteroventral second infra-
orbital. The third and fourth infraorbitals are much shorter and
there are no traces of suborbital elements.

A characteristic triangular expanse of a quadrate shows below
the second infraorbital. The anterior margin of this bone lies ver-

1971

NEW FISH FROM THE MOWRY SHALE

tically in a transverse plane. An oblique and anteroventrally direct-
ed articular facet for the symplectic suggests a vertical or only
slightly forward inclination of the hyomandibula. The anteromesial
and posteroventral limbs of the ectopterygoid meet in a nearly 90°
angle. Although anterior autopalatine and posterior metapterygoid
sutures are obscured, the entopterygoid would appear to be ex-
panded.

The opercular apparatus is complete. Ventral horizontal and
dorsal vertical rami of the preoperculum are of disparate length
and height. The length of the horizontal limb is somewhat less than
half the length of the skull. A mesial flange along the forward edge
of the preoperculum is indicated and suggests intimate internal
support of the element by the hyomandibula and quadrate as well
as the shift of origin of adductor mandibulae muscles which was
acquired by the advanced halecostomes (Gardiner, 1967) . The
operculum is the largest of the complex with an oblique ventral
margin meeting the vertical anterior border in an acute antero-
ventral angle. The posterior edge of the operculum is continued
downward and forward by that of the suboperculum in a broad
sweeping arc. The interoperculum is horizontally elongated along
with the ventral arm of the preoperculum which overlaps it widely.
An impression of a bone beneath the anterior extremity of the pre-
operculum suggests that the branchiostegal rays were expanded
and relatively few in number.

Of the 27 vertebrae preserved, the anterior 22 are considered
abdominal and the remaining 5 caudal. The centra show little re-
gional variation and are generally well ossified, with lengths slightly
greater than their diameters. They are hourglass-shaped with very
reduced notochordal perforations and bear laterally as many as
three axial strengthening laminae. The proximal attachments of
the neuropophyses with the centra are obscure but there appears
to be a progressive gain in the robustness of these structures pos-
teriorly. Epineural elements, as shown by delicate impressions
across the neural spines, are definitely present in the midabdominal
region. The proximal portions of the rib impressions recurve for-
wardly and lie parallel to the ventral borders of the vertebrae. Evi-
dences of well defined parapophyses are absent.

From the point of low ventral attachment of the pectoral fin
rays, the anteroventral limb of the cleithrum is short and com
stricted in comparison with the high and moderately expanded

DAVID H. DUNKLE

No. 14

dorsal ramus of the bone. The anterior margins of the two parts,
meeting in a gentle forward concavity, are thickened into a mesially
directed flange which forms the border of the opercular cleft.
Little else of other girdle elements and of the radials can be ascer-
tained. Pectoral fin rays number 7 in both the right and the left
appendages, but may have totaled a few more. The rays of both
fins are adpressed vertically along the posterior margin of the
cleithrum. As impressed, the rays indicate long narrow fins, with-
out evidence of axial jointing and with only the posterior rays
dichotomizing.

DISCUSSION

The generic assignment of the presently described and incom-
pletely revealed fish from the Mowry Shale may well be considered
as extremely tenuous by many. It is tentatively offered, however,
on the bases of general structures, proportions and relations of
parts. Of fundamental significance in these connections are the at-
tenuated frontals with the contained supraorbital sensory canals
having parietal and infraorbital branches; absence of a confluence
of the dorsal extension of the preopercular sensory canal with the
supraorbital and infraorbital canals and consequent lack of a reces-
sus lateralis (Greenwood and others, 1966) and similarities of
mouthparts, the infraorbital elements, the hyopalatine complex and
the opercular apparatus.

The genus Clupavus was defined by Arambourg (1950) for the
reception of various small fishes from the Upper Jurassic of Eng-
land and the Lower Cretaceous of Dalmatia, Jugoslavia and Gabon,
which had originally been attributed to Leptolepis. Subsequently,
Clupavus species have been recognized from the Cretaceous of
Portugal (Ferreira, 1961); the Lower Cretaceous of the Congo
(Casier, 1961) ; and the Upper Cretaceous (Cenomanian) of Moroc-
co (Arambourg, 1954) and Lebanon (Patterson, 1967) .

In the total brief literature concerning the genus there appears
to be a general concurrence of opinion that Clupavus is a morpho-
logic derivative of the halecostome fishes, through which narrow
front the holostean-teleostean transition is thought most probably
to have occurred. Within this framework Clupavus has been vari-
ously treated systematically: (1) together with the leptolepids

within an advanced holostean level of organization (Greenwood

1971

NEW FISH FROM THE MOWRY SHALE

Ptot

Fig. 2. Attempted reconstruction, in right lateral aspect, of the skull of the
Mowry specimen (CMNH 11045) of Clupavus. Reproduction approx. X 10.
Explanation of abbreviations: Dsph, dermosphenotic; Ecpt, ectopterygoid; Enpt,
entopterygoid; Exsc, extrascapular; Fr, frontal; lop, inter operculum; Io 2, 3,
and 4, second, third and fourth infraorbitals; La, lacrimal or preorbital; Mn,
mandible; Mx, maxilla; Op, operculum; Pmx, premaxilla; Pop, preoperculum;
Psph, parasphenoid; Ptot, pterotic; Qu, quadrate; Sm. a., anterior supramaxilla;
Sm. p., posterior supramaxilla; Sop, suboperculum; Sym, symplectic; i.so.s.c.,
infraorbital branch of supraorbital sensory canal; p.so.s.c., parietal branch of
supraorbital sensory canal; so.s.c., supraorbital sensory canal.

and others, 1966; Andrews and others, 1967) ; (2) at a teleostean
level of organization from which the leptolepids are excluded as
of lower level (Arambourg, 1954; Bertin and Arambourg, 1958) ;
and (3) at a teleostean level of organization in which the leptolepids
are included (Arambourg, 1950; Casier, 1961; Danirchenko, 1964;
Romer, 1966 and Patterson, 1967). The two most recent and com-
prehensive classifications of fishes (Greenwood and others, 1966;
Andrews and others, 1967) favor the first of these three alterna-
tives.

Romer (1966, p. 354) reports another North American occur-
rence of the genus Clupavus. No formal reference to such a prior
assignment has been found. It is, however, thought possibly to per-
tain to Leptolepis nevadensis David (1941) from the freshwater
Newark Canyon Formation of the Lower Cretaceous (Nolan, 1962),
since Miss David compared that species most favorably to the same
suite of diminutive species on which Arambourg erected Clupavus.

DAVID H. DUNKLE

No. 14

REFERENCES CITED

Andrews, S. M., Gardiner, B. G., Miles, R. S. and Patterson, C., 1967, Chapter
26, Pisces, in Harland, W. B. and others, The Fossil Record, a Symposium
with Documentation: Geol. Soc. London, p. 637-683.

Arambourg, C., 1950, Nouvelles observations sur les Halecostomes et Torigine
des Clupeidae: Acad. sci. (Paris) Comptes rendus, v. 231, no. 6, p. 416-
418, 1 fig.

, 1954, Les poissons cretaces du Jebel Tselfat (Maroc): Notes

Mem. Serv. Geol. Maroc, no. 118, p. 1-118, 20 pi., 68 figs.

Bertin, L. and Arambourg, C., 1958, Super-ordre des teleosteens (Teleostei),
in Grasse, P. P. Traite de Zoologie: Paris, Masson et Cie, v. 13, fasc. 3,
p. 2204-2500, 227 figs.

Casier, E., 1961, Materiaux pour la faune ichthyologique eocretacique du Con-
go: Ann. Mus. Congo beige., no. 39, xii-96 p., 12 pi.

Cockerell, T. D. A., 1919, Some American Cretaceous fish scales: U. S. Geol.
Survey Prof. Paper 120, p. 165-188, pis. 31-37.

Danil’chenko, P. G., 1964, Superorder Teleostei, in Obruchev, D. V., Osnovy
Paleontologii: Moscow, Acad. Nauk. S. S. S. R., v. 15, p. 396-472, pis. 5-14,
figs. 101-197.

Gardiner, B. G., 1967, The significance of the preoperculum in actinopterygian
evolution; Jour. Linn. Soc. London (Zool.), v. 47, p. 197-209, 8 figs.

Greenwood, P. H., Rosen, D. E., Weitzman, S. H. and Myers, G. S., 1966,
Phyletic studies of teleostean fishes with a provisional classification of
living forms: Amer. Mus. Nat. Hist. Bull., v. 131, art. 4, p. 339-456, pis. 21-
23, 9 figs., 32 charts.

Nolan, T. B., 1962, The Eureka mining district, Nevada: U. S. Geol. Survey
Prof. Paper 406, p. 1-78, 11 pi., 15 figs., 2 tables.

Patterson, C., 1967, Are the teleosts a polyphyletic group? in Problemes actuels
de Paleontologie (Evolution des Vertebres) Paris, Coll. Internat. Centre
Nat. Reserche Scient. no. 163, p. 93-109, 11 figs.

Reeside, J. B. and Cobban, W. A., 1960, Studies of the Mowry shale (Creta-
ceous) and contemporaneous formations in the United States and Canada:
U. S. Geol. Survey Prof. Paper 355, 127 p., illust.

Romer, A. S., 1966, Vertebrate Paleontology: Chicago, University of Chicago
Press, viii-468 p., 443 figs., 4 tables.

MANUSCRIPT RECEIVED JANUARY, 1971

VJe. • '

KIRTLANDI A

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO OCTOBER 1, 1971 NUMBER 15

DEVONIAN FISHES FROM CALIFORNIA

DAVID H. DUNKLE
Cleveland Museum of Natural History

and

N. GARY LANE

University of California, Los Angeles
ABSTRACT

The first occurrence of the well-known dinichthyid arthrodire
Dunkleosteus, represented by disassociated and weathered bones,
in the Quartz Spring Sandstone Member of the Lost Burro For-
mation (Late Devonian) in Inyo County, California, is reported.

Of particular interest is the presence of teeth of cladodont and
cochliodont sharks together with the Dunkleosteus bones.

This notice reports an original finding of weathered, disasso-
ciated remains of Devonian fishes from California. The specimens
were collected by one of us (N. G. L.) north and 100 feet above the
road passing through Lost Burro Gap (lat. 36° 44' 59"; long. 117°
31' 19", northeast corner of the Ubehebe Peak 15' quadrangle, Inyo
County, California) . The specimens were obtained from thin-
bedded cherty limestone and calcareous to dolomitic sandstone in
the upper 20 feet of the Quartz Spring Sandstone Member of the
Lost Burro Formation. This member constitutes the Cyrtospirifer
Zone of MacAllister (1952) and a late Upper Devonian age has been
indicated by studies of invertebrate fossils (Langenheim and Tisch-
ler, 1960) , and conodonts (Youngquist and Heinrich, 1966) . This
age can now be corroborated on the basis of the presently noted
vertebrate fossils.

The recognizable remains, deposited in the paleontological col-
lections of the Cleveland Museum of Natural History, include two
fragmentary elements of the dermal armor of a moderately large

D. H. DUNKLE AND N. G. LANE

No. 15

Fig. 1. Bones attributed to Dunkleosteus terrelli (Newberry) from the Quartz
Spring Sandstone Member of the Lost Burro Formation (Late Devonian) in
Inyo County, California. Sketches, as preserved, of right posterior dorsolateral
plate (CMNH 8231) in (A) external and (B) internal views and left intero-
lateral fragment (CMNH 8232) in (C) anterolateral aspect. Reproduction
approx. X .45 natural size. Abbreviations: ol. ADL, overlap area for anterior
dorsolateral; ol. MD, overlap area for median dorsal; ol. SP, articular facet for
spinale; ol. + PL, overlap area and sulcus for posterolateral.

ol.SP.

1971

DEVONIAN FISHES FROM CALIFORNIA

placodermatous fish and the teeth of a shark and of a cochliodont.
The placoderm bones are readily identifiable as the right posterior
dorsolateral (CMNH 8231) and the left interolateral (CMNH 8232)
plates of an arthrodire. In fact, comparative details of structure are
so closely similar that the bones cannot be distinguished from the
corresponding elements of the distinctive and well-known dinich-
thyid Dunkleosteus. The materials are illustrated (fig. 1) imposed
on the outlines of appropriate bones of Dunkleosteus terrelli (New-
berry) from the Ohio Shale. For present purposes further descrip-
tion is unnecessary.

The shark tooth (CMNH 8233) , typically cladodont although
generically indeterminate, is a relatively small example and consists
of an expanded root having rounded extremities and a width almost
equal to its length. The cross section of the principal cusp displayed
is robust and although a little flattened or slightly concave on its
labial side is without either lateral keels or vertical striation. A
single pair of diminutive lateral accessory cusps is indicated. The
recognized cochliodont nature of the other tooth (CMNH 8234) is
based on the histologic structure of a basal layer of osteodentine
and a superficial layer of tubular dentine whose parallel canals per-
forate the convex coronal surface perpendicularly.

Despite a lengthy awareness of the collective nature of the
arthrodiran genus Dinichthys only a few of the indicated revision-
ary studies of the many species either originally or secondarily
attributed to it have been made. Among those accomplished, how-
ever, Lehman’s (1956) removal of three forms (Dinichthys terrelli,
D. intermedins and D. curtus) (see also Eastman, 1907 and Dunkle
and Bungart, 1940, 1946) from the Upper Devonian Ohio Shales to
a new genus, Dunkleosteus, is of pertinent interest in the present
connection. The combination of structural characteristics which dis-
tinguish Dunkleosteus from all other arthrodiran associates have
been elaborated, especially by Heintz (1932), in detail. Of these and
prompting this similar generic allocation, the new materials from
California display the same overlap plus ingrown dentation of the
sutural articulation between posterolateral and the posterior dorso-
lateral bones. Also, the interolateral presents the similarly short-
ened ventral lamina and the well-defined articulatory facet pre-
served laterally on the ascending lamina denoting the retention of
a greatly reduced spinal element which among the presumably
closely related dinichthyid assemblage is a unique feature of
Dunkleosteus.

D. H. DUNKLE AND N. G. LANE

No. 15

Dunkleosteus is best known for numerous remains and many
individuals showing great variation in size in certain strata, notably
the black Ohio Shale (Famennian) of Ohio and Kentucky, equiv-
alent levels of the New Albany Shale in Indiana and north-central
Kentucky and the Chattanooga Shale of Tennessee. Some frag-
ments wholly reminiscent of the genus and possibly reworked into
the basal Mississippian have been reported from Texas (Dunkle
and Wilson, 1952) . In addition, other species from the Upper Devo-
nian of Morocco (Lehman, 1956) and of Europe (Obruchev, 1964)
have been referred to Dunkleosteus. It is probable that future
studies of other dinichthyid species (notably D. magnificus, Hussa-
kof and Bryant, 1918, from the Rhinestreet Shale, D. missouriensis
Branson, 1914, from the Grassy Creek Shale, among others) may
prove to have the same generic affinity but at the present time no
very great extension of the temporal distribution of Dunkleosteus
can be projected.

The finding in California of these fishes, all representative of
marine lineages, is not only of considerable paleogeographic impor-
tance but, in view of the meager available record of Devonian
cochliodonts, is definitive of an interesting and, as yet rather un-
common, faunal association.

1971

DEVONIAN FISHES FROM CALIFORNIA

REFERENCES CITED

Branson, E. B., 1914, Devonian fishes from Missouri: Univ. Missouri Bull., no. 4,
p. 59-75, 4 pi., 1 fig.

Dunkle, D. H. and Bungart, P. A., 1940, On one of the least known of the
Cleveland Shale Arthrodira: Cleveland Mus. Nat. Hist., Sci. Pubs. v. 8,
no. 2, p. 29-47, 2 pi., 7 figs.

1946, The anterosupragnathal of Gorgonichthys : Amer. Mus.

Novitates, no. 1316, 10 p., 4 figs.

Dunkle, D. H. and Wilson, J. A., 1952, Remains of Devonian fishes from Texas:
Jour. Wash. Acad. Sci., v. 42, no. 7, 3 p., 5 figs.

Eastman, C. R., 1907, Devonic fishes of the New York formations: New York
State Mus. Mem., no. 10, 235 p., 15 pi., 34 figs.

Heintz, A., 1932, The structure of Dinichthys, a contribution to our knowledge
of the Arthrodira: Amer. Mus. Nat. Hist., Bashford Dean Mem. Vol., art. 4,
p. 115-224, 9 pi, 91 figs.

Hussakof, L. and Bryant, W. L, 1918, Catalogue of the fossil fishes of the Buf-
falo Society of Natural Sciences: Buffalo Soc. Nat. Sci. Bull, v. 12, 198 p,

70 pi, 64 figs.

Langenheim, R. L, Jr. and Tischler, H, 1960, Mississippian and Devonian
paleontology and stratigraphy, Quartz Spring area, Inyo County, Califor-
nia: Univ. California Pubs, in Geol. Sci, v. 38, p. 89-152.

Lehman, J. P, 1956, Les arthrodires du Devonien Superieur du Tafilalet (Sud
Marocain) : Serv. Geol. Empire Cherifien (Rabat) Notes Mem, no. 129,

71 p, 24 pi, 15 figs.

MacAllister, J. F, 1952, Rocks and structure of the Quartz Spring area, north-
ern Panamint Range, California: Calif. Dept. Nat. Res, Div. Mines Spec.
Rept. 25, 38 p, 3 pi, 13 figs.

Obruchev, D. V, 1964, Class Placodermi, in Obruchev, D. V, Osnovy Paleon-
tologii: Acad. Nauk S.S.S.R, v. 11, p. 168-247, 6 pi, 82 figs.

Youngquist, W. and Heinrich, M.A, 1966, Late Devonian conodonts from the
Lost Burro Formation of California: Jour. Paleontology, v. 40, p. 974-975.

MANUSCRIPT RECEIVED AUGUST 24, 1971

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO DECEMBER 22, 1972 NUMBER 16

A CARBONIFEROUS LABYRINTHODONT AMPHIBIAN
WITH COMPLETE DERMAL ARMOR

ALFRED SHERWOOD ROMER
Museum of Comparative Zoology , Harvard University

ABSTRACT

Specimens of a labyrinthodont, Greererpeton burkemorani,
from the Lower Carboniferous of West Virginia, show p n
apparently complete body covering of dermal scales, indicating
the impossibility of ‘‘skin-breathing” in this early amphibian.

It has been suggested by many workers interested in the
history of vertebrate respiration that skin-breathing was a tran-
sitional stage in the shift from gill-breathing in fishes to lung-
breathing in higher terrestrial vertebrates. This hypothesis, of
course, arises from the fact that in modern amphibians the skin
is naked and moist and much of the oxygen-carbon dioxide ex-
change takes place through the skin. I have, however, recently
(Romer, 1972) pointed out that such an evolutionary succession
is highly improbable. To begin with, it is generally agreed that
lungs were already present in the rhipidistian crossopterygian
fishes from which land vertebrates quite certainly derived —
fishes completely ensheathed in an armor of thick bony scales.
However, an important question remains — what was the condi-
tion of the body covering in the ancient amphibians of the Paleo-
zoic, from which the modern orders were derived? It could be
argued that lungs, even if present in the ancestral tetrapods,
may have been too primitive and ineffectual in nature to carry
the whole burden of respiration, and that a r..ked skin may have
been early called upon as a breathing aid. To evaluate properly
the worth of such a concept, it is necessary to resort to paleon-
tology and attempt to determine the nature of the skin in older
amphibian types.

ALFRED SHERWOOD ROMER

No. 16

The living orders of amphibians — the Apoda (Gymnophi-
ona), Urodela (Caudata), and Anura (Salientia) — are rela-
tively modern groups. Almost nothing is known of fossils of the
first group ; urodeles go back only to the Cretaceous, last of the
Mesozoic periods ; frogs had developed by the mid-Mesozoic
Jurassic Period, and one “pre-frog” (Protobatrachus or Triado-
batrachus) is known from the Late Triassic. But amphibians
originated far back, in Late Devonian times, and for the Late
Paleozoic and even most of the Triassic Period — a stretch of
time of close to one hundred and fifty million years — we find
considerable faunas of amphibians, but amphibians quite distinct
from the more modern types. Within these groups quite surely
lay the ancestors of the later forms. What sort of body covering
did these ancestors possess?

Two ancient groups have been advocated as possible ances-
tors — the Microsauria and the Labyrinthodontia. The micro-
saurs were small forms, usually but a few inches in length, whose
remains are not infrequently preserved on slabs of shale which
show not only the skeleton but also the nature of the skin. The
skin is seen to have had a complete covering of bony scales ; hence
if the microsaurs were ancestors, the naked-skinned condition
was surely a late acquisition.

However, current opinion (Parsons and Williams, 1963;
Estes, 1965) favors the descent of the modern orders from the
Labyrinthodontia, particularly the subgroup of this great ordet
termed the Temnospondyli. The labyrinthodonts include a host
of varied forms which were very prominent in Late Paleozoic
and Triassic times. Some labyrinthodonts were small, but many
ranged up to the size of a modern alligator or crocodile. For
these forms, determination of the nature of the body covering is
in general a matter of difficulty. Seldom are their remains dis-
covered in flattened slab form, as are those of the little micro-
saurs. Usually they are found as three-dimensional burials in
clays and shales. After death and before burial, the soft parts
tended, of course, L rot, and skin as well as muscles and viscera
were generally separated from the skeleton of the cadaver ; fur-
ther, excavation of the skeleton and its preparation in the labora-
tory has tended to do away with any remnants of the body cover-
ing of the living animal.

1972

AMPHIBIAN DERMAL ARMOR

Almost invariably the ventral portion of the original fish-
scale covering is retained in labyrinthodonts as a series of gas-
tralia (or abdominal “ribs”), very useful in protecting the bellies
of these low-slung animals (fig. 1). As regards the rest of the

Fig. 1. Ventral view of head and part of trunk of a specimen of Greerer-
peton, CMNH 11090, showing the lower surface of the jaws, dermal shoulder
elements and most of the ventral abdominal shield of gastralia. About
1/3 natural size.

body covering, knowledge is, as expected, meagre. As I noted in
my recent paper (1972), dorsal squamation has been known in
only four labyrinthodonts : Eryops, T rimer or hachis, Archegosau-
rus, and Actinodon. All four are temnospondylous labyrintho-

ALFRED SHERWOOD ROMER

No. 16

donts — the generic group from which the modern orders are
very probably descended — but all four are, unfortunately, of
relatively late (Permian) age. It would be of considerable inter-
est to acquire data as to the nature of the body covering in the
older members of the group.

As we descend the geologic scale, however, labyrinthodont
remains become relatively rare. Numerous specimens have been
found in the Upper Carboniferous - — the Pennsylvanian Period
of American nomenclature — but for the most part they repre-
sent forerunners of the Permian groups. From the entire Lower
Carboniferous (Mississippian), there had been recovered, until
recently, only a single rather poorly preserved labyrinthodont
skeleton (Pholidog aster, Romer, 1964) ; a considerable amount
of material of the oldest amphibians, the ichthyostegids, of the
Late Devonian, has been collected but for the most part remains
unstudied. We thus had very little knowledge of the structure
of the older labyrinthodonts, and no adequate data upon which
to judge the presence or absence of bony squamation in the most
ancient members of the group.

This picture is now considerably altered by recent discover-
ies in the Mississippian of West Virginia. Most rocks of that
age are marine in nature, but at Greer, near Morgantown, West
Virginia, there occurs high up in the wall of a limestone quarry,
a layer containing continental sediments (Hotton, 1970, p. 1-5).
This layer is the Bickett Shale, a basal member of the Mauch
Chunk Group, representing the upper portion of the Mississip-
pian deposits of the area. In this shale there have been found
for many years fragmentary remains of amphibians, and in the
last few years the Cleveland Museum of Natural History has
undertaken, under the direction of Dr. David H. Dunkle and
with the assistance of William Hlavin and others, excavation of
the bone layer there. Most of the bones are in a single layer ex-
posed in the nearly vertical face of the quarry. Despite the diffi-
culties of the task, the Cleveland Museum has already excavated
to a variable depth (as much as 14 feet in one area) the main
bone-bearing layer for some 50 feet along the quarry face. This
layer proves to contain a nearly solid mass of amphibian remains,
skulls and complete skeletons, closely packed and flattened dorso-
ventrally so as to include, in essentially slab form, all the “hard”

1972

AMPHIBIAN DERMAL ARMOR

Fig-. 2. Dorsal view of a slab containing much of the skeletons of three
individuals of Greererpeton CMNH 11036, 11034 and 11082. Quantities of
dorsal scales are visible, particularly posteriorly. About 1/6 natural size.

ALFRER SHERWOOD ROMER

No. 16

remains of the animal — not merely the skeleton, but also the
skin if, in the process of burial, it had not been completely dis-
integrated and floated away from the main portion of the
cadaver.

The most common amphibian in the bed is Greererpeton, a
form of which I gave a preliminary description in 1969. The
exact systematic position of this genus is at present none too
certain. It is, however, a temnospondyl in vertebral structure
and hence, in at least a broad sense, a member of the general
group of labyrinthodonts from which modern orders are not
improbably descended, although Greererpeton is certainly not in
the direct ancestral line. In this genus the stout series of gas-
tralia is to be found (fig. 1) covering the belly region. But in
addition many of the specimens of this form show over the dorsal
region a highly developed series of scales, subcircular in form
and averaging 1.0 to 1.5 mm in diameter (figs. 2-4). The skin
in which they were embedded in life had quite surely become
more or less disintegrated before burial, and hence we cannot
be sure of the exact nature of the scale pattern in life. It is,
however, evident that even post mortem the disintegrating skin
often retained a definite pattern of the arrangement of the
scales in regular rows. There had rather certainly been a reduc-
tion in scale size from the situation seen in typical crossoptery-
gians (cf. for example, Romer, 1966, fig. 100), and a lessening
of the deep overlapping of scales seen in these ancestral forms;
however, in some cases (cf. fig. 4) there are indications that a
certain degree of overlapping persisted in the members of scale
rows.

Rather surely the general body squamation in Greererpeton
was not as complete as in crossopterygians, and some gaps may
have existed between scales. But certainly the scale covering
was such as to exclude the possibility of any great degree of skin
breathing. It is, of course, not impossible to argue that some
labyrinthodonts may have developed a naked skin. But the dis-
covery of a well-developed scale covering in one of the oldest
labyrinthodonts known, as well as evidence of a similar skin
structure in a number of later temnospondyls, adds weight to
the theory very generally accepted by paleontologists, that a
scaled body sheathing was present in the ancestors of the modern

1972 AMPHIBIAN DERMAL ARMOR 7

Fig. 3. Posterior portion of the block shown in figure 2, mainly skull and
skeleton of CMNH 11082. Dorsal scales, preserved in more or less of a
regular pattern, as seen along most of the region to the right of the ver-
tebral column. Disarticulated gastralia are seen along the left margin.
About 1/3 natural size.

ALFRED SHERWOOD ROMER

No. 16

Fig. 4. A small portion of the dorsal region of a skeleton, CMNH 11070,
about natural size, to show arrangement and possible overlapping of scales.

amphibians and that the development of a naked skin came about
at a relatively late period in amphibian evolution.

Credit is due Virginia Heisey of the Cleveland Museum of
Natural History for skillful preparation of the Greererpeton
specimens, and to Bruce Frumker, of the same institution, for
the excellent photographs from which plates 1-3 were derived.

REFERENCES CITED

Estes, R., 1965, Fossil salamanders and salamander origins: Amer. Zoologist,
v. 5, p. 319-334.

Hotton, N., 1970, Mauchchunkia bassa, gen. et sp. nov., an anthracosaur
(Amphibia, Labyrinthodontia) from the Upper Mississippian : Kirt-
landia, no. 12, p. 1-38.

Parsons, T. S. and Williams, E. E., 1963, The relationships of the modern
Amphibia: a re-examination: Quart. Rev. Biol., v. 38, p. 26-53.

Romer, A. S., 1964, The skeleton of the Lower Carboniferous labyrinthodont

Pholidogaster pisciformis : Mus. Comp. Zool. Bull., v. 131, no. 6, p.
129-159.

, 1966, Vertebrate Paleontology: 3rd ed., Chicago, Univ. Chicago

Press, 468 p.

, 1969, A temnospondylous labyrinthodont from the Lower Car-
boniferous: Kirtlandia, no. 6, p. 1-20.

, 1972, Skin-breathing — primary or secondary?: Respir.

Physiol., v. 14, p. 183-192.

MANUSCRIPT RECEIVED SEPT. 1, 1972

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO JUNE 29, 1973 NUMBER 17

NOTES ON THE MORPHOLOGY OF ACANTHERPESTES
(MYRIAPODA, ARCHYPOLYPODA) WITH THE DESCRIPTION
OF A NEW SPECIES FROM THE
PENNSYLVANIAN OF WEST VIRGINIA

J. J. BURKE

Senior Scientist, Cleveland Museum of Natural History and
Research Associate, West Virginia Geological Survey

ABSTRACT

Study of a large fossil myriapod from the Pennsylvanian Allegheny Group in Monon-
galia County, West Virginia, necessitated comparison with specimens representing
various species of the Upper Carboniferous euphoberiid genus Acantherpestes, including
Acantherpestes major, type species of the genus. This investigation determined that,
contrary to previous interpretations, Acantherpestes was a “flat-backed” myriapod,
characterized as follows: Tergites moderately arched transversely, with two rows of
spines on each side of the metazonite, one row comprising simple reduced subdorsal
spines flanking the dorsal midline, the second row consisting of long, stout, lateral spines
arising near the lateral border, subhorizontally or horizontally disposed, and bifurcate,
with basal spinelets. Lateral spines, prolonged beyond the body of tergite, sheltered the
laterally extended, elongate feet. Sternites entire, prosterna and metasterna not divided
medially, with “cups” housing exsertile sacs situated close to median line, and spiracles
adjoining the coxal region laterally. Coxal regions with sternal inflatations, terminating
in outward-facing coxal sockets. Feet composed of five podomeres, the second quite
elongate.

Scudder’s interpretation of Acantherpestes as an amphibious myriapod is disputed;
the feet are regarded as having been adapted for weight bearing and efficient locomotion,
rather than to serve as swimming appendages; exsertile sacs are considered to have
absorbed water to combat dessication, rather than having a gill-like function for under-
water respiration.

American species of Acantherpestes include Acantherpestes major Meek and Worthen,
Acantherpestes inequalis Scudder, and Acantherpestes clarkorum sp. nov. Also herewith
assigned to Acantherpestes is the American species Euphoberia hystricosa Scudder, and
the familiar English Coal Measures myriapod Euphoberia ferox (Salter). In addition, at
least one American species, and another from the English Coal Measures, both presently
unnamed, are attributable to Acantherpestes.

J. J. BURKE

NO. 17

Myriapod taxa from the Upper Carboniferous of Czechoslovakia, attributed by Fritsch
to Acantherpestes and Euphoberia, differ greatly from species comprising the latter genera,
having: (1) tergites more arcuate in transverse section; (2) flank spines more upright;
(3) subdorsal spines much longer and stouter; (4) sternites not entire — prosterna and
metasterna divided medially; and (5) sternal structures widely at variance with those
of Acantherpestes and Euphoberia. It is evident that new genera should be established for
the reception of these Fritsch species.

The myriapod from the Pennsylvanian of West Virginia, described as a new species,
Acantherpestes clarkorum, is distinguished mainly by its large size, subdorsal spines
reduced to nodes, small tubercle near outer termination of lateral furrow, and long
lateral spines, bifurcate at midlength, having the anterior prong about one-third the
length of the posterior, and prominent basal spinelets exceeding the anterior prong
in length.

INTRODUCTION

The present article stems from the discovery of a large fossil myriapod
in the Pennsylvanian Allegheny Group in Monongalia County, West
Virginia (Barlow, 1969). Study of this specimen showed it to be a repre-
sentative of the genus Acantherpestes Meek and Worth en, and a new
species, but prior to this determination it was necessary to make extensive
investigation of various fossil myriapods from the Upper Carboniferous
of North America and Europe. Completion of this work, in consequence,
has been delayed.

The paper is divided into two parts. The first embodies brief notes on
the morphology of Acantherpestes , sufficient, it is hoped, to furnish basic
information on the structure of the genus as we now know it. The second
part combines provenance and other details of the West Virginia speci-
men, followed by systematic paleontology, including an emended diag-
nosis of Acantherpestes, plus a diagnosis and description of the new species,
accompanied by pertinent discussions. A compilation of references cited
throughout the article follows the second part.

ACKNOWLEDGMENTS

Several persons and institutions, both here and abroad, have contributed in one way
or another to assist this project, and to all of them I extend hearty thanks. Specimens
have been loaned for study by Dr. Bernard Kummel and Miss Vickie Kohler of the
Harvard Museum of Comparative Zoology; Dr. H. W. Ball and Dr. S. F. Morris of the
British Museum (Natural History); Dr. Porter Kier of the National Museum of Natural
History; Dr. John Carter of the University of Illinois; Dr. Eugene S. Richardson of the
Field Museum; Mr. Stephen LeMay of Chicago, Illinois; and Mr. Walter Dabasinskas
of Monticello, Wisconsin. I am indebted to Dr. Alec Panchen of the University of New-
castle-upon-Tyne for information concerning fossil myriapod localities in the English
Coal Measures. Dr. John Hower of Case Western Reserve University conducted an
X-ray analysis of the specimen. I am grateful to the Thomas Clark family of Morgan-

1973

ACANTHERPESTES

town, West Virginia, for the opportunity to study this fossil, which they have since
donated to The Cleveland Museum of Natural History.

Initial preparation of the specimen was by Mr. Peter Hoover of Case Western Reserve
University. His help is fully appreciated, along with that of three staff members of The
Cleveland Museum of Natural History: Miss Virginia Heisey, for additional preparation;
Mr. Brant Gebhart, for illustrations; and Mr. Bruce Frumker, for photography.

TERMINOLOGY

An explanation of some of the terminology employed in this article is pertinent at this
point. A body segment (or diplosomite) of Acantherpestes is composed of a single dorsal
plate, the tergite, which overlies two ventral plates, the sternites, each of which bears a
single pair of legs. Two divisions of the tergite are recognized. The anterior of these, the
prozonite, is smooth, and is overlapped by a portion of the tergite anterior to it. The
posterior division of the tergite, the metazonite, is elevated above the prozonite and
overlaps the prozonite of the tergite posterior to it. The metazonite of Acantherpestes
bears spines and other distinctive features which are of use in making specific deter-
mination within the genus. Figure 1 is a diagrammatic sketch of a single tergite of
Acantherpestes, with significant details labeled.

If Pr PI Ls

Fig. 1. Diagrammatic dorsal view of a tergite of Acantherpestes, with salient details of
morphology indicated. Abbreviations: Apr, anterior prong of lateral spine; Ar, anterior
ridge; Ast, anterior spinelet; Lf, lateral furrow; Ls, lateral spine; Mtz, metazonite;
PI, posterior lobe; Ppr, posterior prong of lateral spine; Pr, posterior ridge; Prz, pro-
zonite; Pst, posterior spinelet; Sds, subdorsal spine.

Most of the designations used in figure 1 have been employed by previous writers in
dealing with these myriapods, or are self-explanatory. Two new terms are introduced.
A depression which arises posterior to the subdorsal spine and extends toward the
anterolateral corner of the metazonite is called the lateral furrow. From the standpoint
of orientation of these fossils, which may be damaged or fragmentary, the lateral furrow

J. J. BURKE

NO. 17

is the most important topological feature of the body of the tergite. In damaged speci-
mens, where the preservation is such that the prozonite-metazonite relations of successive
tergites are obscure, those furrows, extending obliquely outward and forward on opposite
sides of the metazonite, are a means of differentiating between the anterior and the pos-
terior regions of the body. In addition, in fragmentary specimens, if the subdorsal spine
and its accompanying lateral furrow are preserved, it is possible to determine whether
the right or the left side of the metazonite is represented.

The term posterior lobe is applied to a swollen area of the metazonite which borders
the lateral furrow posteriorly and merges laterally with the lateral spine. This swollen
area varies in prominence in different species of Acantherpestes, and is usually character-
ized by gridlike ornament.

ABBREVIATIONS

The following abbreviations of institution names are employed in this article: BM,
British Museum (Natural History); ISM, Illinois State Museum; MCZ, Harvard
Museum of Comparative Zoology; UI, University of Illinois; USNM, National Museum
of Natural History.

NOTES ON THE MORPHOLOGY OF ACANTHERPESTES

STRUCTURE OF ACANTHERPESTES AS INTERPRETED BY
MEEK AND WORTHEN (1868) AND SCUDDER (1882, 1890)

Up to the present, all described material from North America which
appears assignable to the genus Acantherpestes has been derived from the
siderite nodules of the Middle Pennsylvanian Francis Creek Shale of the
Carbondale Formation (the so-called Mazon Creek beds of Grundy
County, Illinois).

Meek and Worth en (1868a) in the course of describing Mazon Creek
specimens, established the myriapod genus Euphoberia, basing the genus
on a small species, Euphoberia armigera. A second species, Euphoberia
major, was distinguished from armigera on the basis of its larger size.

At that time, however, Meek and Worth en appear to have had at hand
at least two large specimens, one of which they compared with Salter’s
(1863) Eurypterus? ( Arthropleura ) ferox, stating that they had little doubt
that it was congeneric with that species.

Later (1868b) Meek and Worth en gave a fuller description of the larger
species under the name Euphoberia?? major. Much of the description is a
repetition of the original, but there are some additional observations on

1973

ACANTHERPESTES

features which the writers regarded as distinguishing Euphoberia?? major
from Euphoberia armigera. A figure, evidently intended to illustrate the
differences between the species, accompanied the description of Eupho-
beria?? major. Because the description and the figure gave rise to a mis-
understanding of the species which has persisted to the present, I am
reproducing the full text and figure below.

Euphoberia?? major, M. and W.

Euphoberia major M. and W., 1868. Am. Jour. Sci., vol. XLV, p. 26.

Fig. 2. (From Meek and Worthen, 1868b, p. 558) Cut illustrating Euphoberia'!' ! major
M & W = Acantherpestes major M & W. *

This name was proposed by us for a much larger fossil than the typical species of the
genus, though we unfortunately yet know it only from mere fragments, one of the best
of which is represented by the annexed cut. If as long in proportion as the other species,
it probably attained a length of 12 to 15 inches, and must have presented a formidable

* Original caption: “ Euphoberia ?? major / Cut of a fragment consisting of six of the
dorsal scutes, and parts of two others, with one of the dorsal spines (s) broken and lying
in the matrix. The nodes (n) are evidently spine bases. Some of the legs are seen below.”

J. J. BURKE

NO. 17

appearance. The node-like prominences, marked n in the figure, are evidently the bases
of spines that have been broken away. One of these, however, is seen lying in the matrix
at the point marked s. Another specimen (not figured) shows a direct view of the dorsal
side, compressed flat. In this, traces of two rows of these node-like prominences are seen
along the middle, while a row of spines can be seen projecting out into the matrix on
each side.

This latter specimen so nearly resembles a fossil figured by Mr. Salter in the Quarterly
Journal of the Geological Society of London, vol. XIX, p. 84, fig. 8, from the Stafford-
shire Coal Measures under the name Eurypterus? ( Arthropleura ) ferox, that we can
scarcely entertain a doubt that they are congeneric. Indeed, if it were not for the fact
that the species ferox has its spines each provided with three, instead of two, prongs, we
would even suspect that our specimen might possibly belong to the same species. Mr.
Salter thought his specimen probably a part of the central lobe of a trilobate Eurypterus,
or some allied genus, an opinion he would not have entertained for a moment (provided
we are right in our suggestion respecting its relations to our fossil) if he had seen a
specimen showing a side view of even a few of the segments, with their legs attached.
At any rate, our fossil is certainly distinct from the genus Arthropleura of Jordan and
von Meyer, which is almost beyond doubt a crustacean.

This larger type, for which we have proposed the specific name major, not only differs
in size from the typical species armigera, but also presents the marked difference of
having its dorsal scutes much shorter and deeper, in proportion to their size. Indeed, as
we are not positively sure that it has two segments below for each one of the dorsal
scutes, we are by no means clearly satisfied that it belongs to the same genus as armigera,
or that it may not even be even much more widely removed from that type. It is there-
fore only provisionally that we have placed it in this genus. This appearance, however,
may possibly be in part due to the oblique manner in which the specimen has been
compressed in the matrix.

If other specimens should be found, showing it not to agree with the typical species
of the genus Euphoberia, in having two segments below for each one above, it will of
course have to be removed from that genus, in which case it might be called Acantherpestes.

It is my feeling that Meek and Worthen were concerned about other
features of the morphology of Euphoberia?? major in addition to the
number of sternites per tergite. I have no doubt that the unfigured speci-
men, with the two rows of nodes running along the middle and a row of
spines projecting out into the matrix on either side, represented what we
now know as Acantherpestes. However, this specimen appears to have
been lost, and the description is so general that it might apply to almost
any species of the genus. Evidently the resemblance to Salter's Eurypterus?
(. Arthropleura ) ferox was much closer than Meek and Worthen thought—
Salter's drawing shows three large prongs, but there are really only two
prongs, and he greatly exaggerated the size of the anterior spinelet, which
is not at all prominent.

I suspect that Meek and Worthen did not figure this specimen showing

1973

ACANTHERPESTES

the two rows of lateral spines, nor declare it the type of their species,
because they were under the impression that Euphoberia?? major possessed
three rows of spines on each flank, and probably assumed that in the
unfigured specimen a third set was present, concealed in the matrix
beneath those that were projecting out on each side. Note that they felt
sure that Salter would not have confused his specimen with Eurypterus
or a similar form “. . . if he had seen a specimen showing a side view of
even a few segments with their legs attached.” Probably Meek and
Worthen had only one specimen preserved (as they thought) in this
fashion, and it is the subject of the drawing accompanying the description
(reproduced in my figure 2). It is apparent that Meek and Worthen
regarded this figure as showing the specimen in lateral view. It is only
when it is interpreted in this light that their statement (1868b, p. 559)
that in addition to differing from the species armigera in size, the species
Euphoberia?? major “. . . presents the marked difference of having its
dorsal scutes much shorter and deeper in proportion to their size” becomes
intelligible. This constitutes an adequate effort to diagnose the differences
between Euphoberia?? major and Euphoberia armigera, and I take the view
that this figured specimen is the type of Euphoberia?? major.

Woodward (1872) presented a drawing copied from that of Meek and
Worthen (1886b) along with most of the text of their description. He
alluded to Euphoberia?? major as Euphoberia? major , but applied the
generic name without question to the Salter species ferox, which he had
examined, using the combination Euphoberia ferox (Salter). However,
there is a curious omission in Woodward's quotation of the Meek and
Worthen text-— he does not include the portion dealing with the charac-
teristics which they felt distinguished Euphoberia?? major from Euphoberia
armigera. Neither does he allude to Meek and Worthen’s hypothetical
genus Acantherpestes.

Scudder (1882), although he was convinced that the tergites of Eupho-
beria?? major each bore two sternites, after expressing his displeasure at
the use of hypothetical names, nevertheless accepted the generic name
Acantherpestes and employed the combination Acantherpestes major Meek
and Worthen. He reproduced (1882, text fig. 5) the Meek and Worthen
illustration, at the same time enlarging upon their interpretation of the
species.

In effect, Scudder saw the Meek and Worthen figure as showing, in
lateral view, several segments of a deep-flanked myriapod which had a
cylindrical body and essentially circular cross-section. In Scudder’s con-

J. J. BURKE

NO. 17

cept, the animal bore three rows of spines (represented in the figure by
spine bases) on each flank. The lower row of spine bases he took to repre-
sent lateral spines, the second row he called “pleurodorsals,” and the row
at the top of the figure, subdorsals. This viewpoint of the structure of
Acantherpestes was illustrated by Scudder in 1882 (text figs. 3, 4, and
pi. 10).

MORPHOLOGY OF THE GENUS BASED ON RESTUDY OF
THE TYPE OF ACANTHERPESTES MAJOR MEEK AND WORTHEN

The specimen figured by Meek and Worthen as Euphoheria?? major,
which I take to be the type of that species, and consequently the type of
Acantherpestes major Meek and Worthen as well, is reposited in the
paleontological collection of the Department of Geology, University of
Illinois. Dr. John Carter, as curator of that collection, kindly loaned me
the specimen for study. The type bears the number UI X-504; formerly
it was part of the Illinois State Museum collection under the number
ISM 11120. The fossil was lightly coated with magnesium oxide and

Fig. 3. Acantherpestes major Meek and Worthen. Type, UI X-504, from the Francis
Creek Shale, Carbondale Formation, Pennsylvanian, at Mazon Creek, Grundy County,
Illinois. Anterior portion of specimen facing left. Slightly oblique dorsal view, X 1.

photographed (figs. 3, 4). In figure 4, various morphological features of
the specimen are labeled, using the terminology illustrated in figure 1.

1973

ACANTHERPESTES

Figures 3 and 4 indicate that the Meek and Worthen illustration, as
represented in my figure 2, is inaccurate in many respects, but prin-
cipally in failing to show details in the upper left portion (the anterior
part of the right side of the animal). Here, in addition to the spine bases
depicted by Meek and Worthen, my photograph shows at least five dis-
tinct lateral furrows on the right side, demonstrating beyond any doubt
that the spine bases associated with these furrows are the bases of the
right subdorsal spines of the animal. It is obvious that the dorsal midline
passes between this row of spines and the left subdorsals, which Scudder
mistakenly identified as “pleurodorsals.” Scudder correctly identified the
lower row of spine bases as laterals. On the opposite (right) side, the row
of lateral spines is not preserved; the specimen is not complete in this
region. Some portions of spines remain (one of which is shown in figure 2).
However, Meek and Worthen seem to have overlooked a damaged lateral
spine, which is displaced and lies on the right lateral flank of the posterior
half of the fossil. The spine is widely bifurcate, and the prongs appear to
deviate from the plane of the main shaft.

It is evident from the above that in UI X-504 the median line passes

Fig. 4. Acantherpesies major Meek and Worthen. Type, UI X-504, same as fig. 3, but
pertinent morphological features labeled. Abbreviations: Ar, anterior ridge; L, lateral
spine base; Lf, lateral furrow; Lg, leg; Lsp, lateral spine; Mtz, metazonite; Prz, pro-
zonite; Sd, subdorsal spine base; Spf, spine fragment; Trg, tergite.

between the two rows of subdorsal spines, and there were only two sets
of spines— the massive laterals and the less prominent subdorsals. The

J. J. BURKE

NO. 17

tergites of UI X-504 are only moderately arched from side to side, and
for the most part the specimen is showing in dorsal, rather than lateral,
view. It follows that Acantherpestes, as indicated by the type species,
was not, as Scudder thought, a deep-flanked myriapod, circular in cross-
section, with three rows of spines on each side. On the contrary, it was a
“flat-backed” myriapod, in the sense meant by Gill (1924). Two speci-
mens in the Harvard Collection, MCZ 7437/la/lb and 7437/2, seem
properly assigned to Acantherpestes major. Both consist mainly of sternal
segments that are gently convex ventrally, which would indicate that this
species is nearly elliptical in cross-section.

An investigation of the material described as Acantherpestes major by
Scudder in 1882 indicates that he was dealing with at least two species
of the genus, neither of which bears close resemblance to UI X-504.
Examination of his specimens shows that in no case did they conform
with his concept of Acantherpestes as a long-flanked myriapod with a
cylindrical body having three rows of spines on each flank. I have not
seen his specimens described and figured in 1890 as Acantherpestes in-
equaiis and Euphoberia hystricosa, but their affinities are evidently with
Acantherpestes as exemplified by the type species. The species hystricosa
is quite definitely an Acantherpestes, and it is here designated Acanther-
pestes hystricosus (Scudder) n. comb.

All of the American specimens and species cited above accord in the
features characteristic of Acantherpestes in keeping with my interpreta-
tion of the genus, and, when sufficiently complete, show tergites having
moderate curvature from side to side, the metazonites of which display
near each lateral border a single row of massive lateral spines, and on each
side of the dorsal midline, a row of shorter subdorsal spines.

ACANTHERPESTES IN THE ENGLISH COAL MEASURES

These same features hold also for representatives of the genus in the
English Coal Measures, where Acantherpestes is represented by at least
two species. Through the kindness of the authorities of the British Museum
(Natural History), I have been able to borrow a cas^JBM I. 1063) of the
specimen described by Salter (1863) as Eurypterus? (AMikhropleura) ferox,
and find it assignable to Acantherpestes, rather than to Euphoberia, as
suggested by Meek and Worthen (1868a, 1868b) and Woodward (1872).
(Actually, as I have noted previously, Meek and Worthen were probably
comparing Salter’s species with a specimen of Acantherpestes, but I have
not been able to find this specimen, which may be lost, and the description
could apply to almost any species of Acantherpestes.) In any case, I am here-

1973

ACANTHERPESTES

with designating the English species Acantherpestes ferox (Salter) n. comb.

In addition, I believe that the specimens described by Gill (1924) and
Brade-Birks (1928) which are derived from the Middle Coal Measures
Crow Coal at Crawcrook, near Pvyton-on-Tyne are referable to Acanther-
pestes as well. However, this small form, with distinctive lateral spines,
quite evidently represents a species other than ferox , to which it was
attributed, but apparently with some hesitation, by Gill and Brade-Birks.

The short papers by these English authors contributed much to clarifying the char-
acteristics of the taxon which I regard as constituting Acantherpestes. Gill (1924) stated:
“Some of the fossil millipedes at present known as species of Euphoberia do appear to
have been more or less cylindrical, but it may be suggested that that is a reason for
separating them generically from ferox rather than for assuming that ferox also was
cylindrical.” This observation followed his noting that the specimen he was describing
appeared to be a “flat-backed millipede,” and his contrasting the curvature of its tergites
and attitudes of the lateral spines with those represented in Woodward’s (1887, pi. 1,
fig. 11) restoration. As a mattter of fact, the tergites and the lateral spines of most of the
specimens figured in Woodward’s plate 1 do not appear to accord with the strongly arched
tergites and distinctly inclined lateral spines shown in the restoration. Furthermore, the
British Museum cast of Salter’s type of Acantherpestes ferox, which I have at hand, does
not indicate marked curvature of the tergites and shows that the lateral spines were sub-
horizontally disposed. Although Woodward (1887, p. 8) noted that he did not feel that
“these large Myriapods” were as round as indicated by Scudder, it is apparent that he
was much influenced by Scudder in preparing his restoration.

Brade-Birks (1928) gave further demonstration that the tergites of the Crawcrook
species were not strongly arched and that the spines were nearly horizontal in disposi-
tion. He also showed the structures of the sternites properly oriented; both Scudder
(1882) and Woodward (1887) confused anterior with posterior in specimens exhibiting
the sternites, and oriented these structures accordingly. However, I gather from Brade-
Birks’ description that he viewed the structures extending from the midline to the
spiracles as coxae, fused at the midline and penetrated closely adjacent to the midline
by the so-called branchial cups. Brade-Birks’ “walled pits” lateral to the “cups” he
considered bases of telepodite joints. Examination of USNM 33039, the specimen illus-
trated by Scudder (1882, pi. 11, figs. 1-4) would indicate otherwise. Scudder thought
that the portion illustrated in fig. 2 represented casts of portions of sternites; actually
these are fossilized exoskeletal structures seen in ventral view. Each plate appears to be
a fairly typical sternite, penetrated close to, and on each side of the midline, by the
“branchial cups.” Between the “cups” and the spiracles, the sternites are produced
ventrally as dilatations that terminate in obovate outward-facing coxal sockets. These
appear to be characteristic coxal sockets which in the American species of Acantherpestes
receive the relatively short but stout coxae, which in turn are joined to the very long
first telepodite joints. Woodward (1887) found two joints preceding the long joint, but I
suggest that restudy of his specimen will determine that only one, the coxa, precedes
the elongate podomere, as in modem Symphyla.

It is of interest to note that in the illustrations of all three authors, Scudder, Wood-

J. J. BURKE

NO. 17

ward, and Brade-Birks (who pointed it out in his specimen), the midlines of the sternites
deviate from those of the tergites, suggesting that after death the ventral and dorsal
segments of these animals slipped askew, tearing the sternites away from the tergites to
which they were probably joined by arthrodial membrane. I find no support for Wood-
ward’s (1887) inference that there was an “overhang” of the tergites beyond the ster-
nites. In one of the Harvard specimens, MCZ 7437/2, identified as Acantherpestes major,
some of the sternites and tergites are showing in such close proximity as to leave little
doubt that they were joined at their lateral extremities.

The species of Acantherpestes from the English Coal Measures do not attain the size,
nor display the specialized spines of some of the American forms, but this is in keeping
with their being possibly exclusively of Westphalian B age, whereas the American
representatives of the genus are from younger (Westphalian C and D) beds.

ACANTHERPESTES CLARKORUM SP. NOV. FROM THE
ALLEGHENY GROUP, PENNSYLVANIAN, OF WEST VIRGINIA

HISTORY OF THE SPECIMEN

The fossil myriapod described in the following pages was discovered
by Alan, Bruce, and Quentin Clark, the young sons of Mr. and Mrs.
Thomas Clark of Morgantown, West Virginia. It was found in the spoil
bank of an abandoned coal strip mine about 10 miles (16.9 km) south of
Morgantown. The specimen for the most part was contained in two pieces
of siltstone (since cemented together) with only the very tips of some of
the subdorsal spines penetrating an overlying piece of rock. Numerous
fossil leaves, mainly Neuropteris, were associated with the myriapod
specimen, which was covered with a very adherent matrix. The rock,
however, was transversed by cracks, and had been subjected to weather-
ing: beneath the matrix, the surface of the fossil consisted in many places
of powdery iron oxide.

The original skeleton of this myriapod was impregnated with calcium
carbonate, but diffraction X-ray analysis of the fossil, conducted by
Dr. John J^vrt^of the Department of Geology, Case Western Reserve
University, determined that the skeleton now consists of siderite with a
small percentage of chamosite.

Preparation was by means of an air abrasive unit. Although this
resulted in loss of the powdery oxide surface, I do not think that any
other method of preparation would have served much better. Despite
some evident damage otherwise, the ornament of the posterior lobes of

1973

ACANTHERPESTES

several metazonites is still showing— an indication that the effects of
preparation were not altogether too drastic.

A small exploratory opening on the underside of the stone containing
the fossil showed no trace of sternites nor legs. No further preparation
was attempted in this region because of the possibility of serious damage
to the specimen.

PROVENANCE

The abandoned strip mine in which the specimen was found lies about
0.8 mile (1.3 km) south of the village of Browns Chapel, in Clinton Dis-
trict, Monongalia County, West Virginia, on the south side of the Glades-
ville-Halleck road, 0.5 mile (0.8 km) east of the intersection of that road
and U. S. Route 119.

The coal that was strip mined at this site was previously identified by
the West Virginia Geological Survey (Hennen and Reger, 1913) as the
Lower Kittanning. However, Mr. Robert S. Reppert and Dr. James A.
Barlow, present members of the survey, on the basis of recent field
studies, informed me (letter, Feb. 2, 1973) that the 1913 designation was
in error, and that the coal is actually the Lower Freeport. At the time
of the 1913 report, a misidentification of the Brush Creek Coal of the
Conemaugh Group as the Upper Freeport Coal of the Allegheny Group
gave rise to the assumption that the coal at the site where the myriapod
fossil was found was separated from the presumed Upper Freeport by an
interval of nearly 200 feet (61 m), and consequently represented the
Lower Kittanning Coal of the Allegheny Group.

The Upper Freeport Coal is sparsely shown or missing in the area
where the fossil was found, but Reppert and Barlow state that its place
is indicated by the base of the Thornton flint clay, which they have traced
throughout the region. The coal of the strip mine lies about 70 feet
(21.3 m) below the base of the Thornton flint clay in that vicinity, an
interval that indicates that the coal in question, which is 4.5 feet (1.4 m)
thick at this place, represents the Lower Freeport Coal of the Allegheny
Group. The pieces of siltstone containing the fossil were not found in
place, but came from the spoil bank of the mine. However, the rock is
doubtless derived from the ferruginous siltstones associated with the
coal, and very likely came from a 35-foot (10.7-meter) siltstone unit
immediately overlying it.

J. J. BURKE

NO. 17

SYSTEMATIC PALEONTOLOGY

Class ARCHIPOLYPODA Scudder, 1882
Family EUPHOBERIIDAE Scudder, 1882
Genus Acantherpestes Meek and Worthen, 1868, emended

Diagnosis: Medium size to very large Upper Carboniferous myriapods.
Prozonites and metazonites fused to form single tergite. Tergites laterally
expanded; moderately arched. Prozonites smooth, overlapped by meta-
zonites. Metazonites elevated, with no more than a single row of large
lateral spines along each flank, and on each side of the dorsal midline a
single row of shorter subdorsal spines. Lateral spines long, massive, sub-
horizon tally to horizontally directed, evenly or unevenly bifurcate, bear-
ing two main prongs, and with spinelets at base. Subdorsal spines simple;
spikelike, curved laterad, or reduced to nodes. Metasternites and proster-
nites undivided medially, with spiracles lateral to coxal sockets and
openings for exsertile sacs near midline medial to coxal sockets. Sternites
with dilatations in coxal regions terminating in outward-facing coxal
sockets. Feet with five podomeres, and second podomere very elongate.

Type species: Acantherpestes major Meek and Worthen, 1868.

Referred species: Acantherpestes ferox (Salter) n. comb.; Acantherpestes
inequalis Scudder; Acantherpestes hystricosus (Scudder) n. comb.; and
Acantherpestes clarkorum sp. nov.

Distribution: Upper Carboniferous; Westphalian B and ?C, England;
Westphalian C and D, U.S.A.

Some anatomical features not included in my diagnosis which may embody details
limited only to a species rather than characterizing the genus as a whole, are also known.
Woodward (1887) described portions of three heads, apparently attributable to Acanther-
pestes ferox. The mouth parts are not preserved. The head exceeds the body segments
(exclusive of spines) in width. The front half is somewhat inflated and the posterior half
bears four tumid lobes. The two lateral and smaller of these lobes comprise the ocellaria,
which bear numerous ocelli. An antennal socket is found anterior to the ocellarium at
the anterolateral angle of the inner lobe. A deep median groove which separates the
inner lobes probably represents the epicranial suture. In the same paper Woodward
describes a telson that probably pertains to Acantherpestes also. Possibly two segments
are represented and only the posterior portion represents the telson proper. It bears
four spines directed posteriorly; the two nearest the median line are longer and more
robust. The anterior portion may comprise the metazonite of the penultimate segment;
the spines appear to be normal lateral spines which are directed posteriorly because of
breakage.

1973

ACANTHERPESTES

Nothing definite can be said of the segments immediately posterior to the head. How-
ever, Scudder’s (1890, pi. 33, fig. 2) figure of Acantherpestes inequalis, although plainly
poorly executed, is of much interest. The head is shown as somewhat wider than the
body segments exclusive of spines. The first four segments are represented as shorter
than those posterior to them, and the lateral spines progressively decrease in width from
the fifth to the first.

Euphoberia, as exemplified by the type species Euphoberia armigera Meek and Worthen,
bears closer resemblance to Acantherpestes than any other Carboniferous myriapod.
However, although specimens of Euphoberia may show the same sets of spines (lateral
and subdorsal) as Acantherpestes, the lateral spine in Euphoberia never attains the
extravagant development that characterizes it in Acantherpestes. Along with the short
lateral spine, the sternites of Euphoberia which, as in Acantherpestes, are not divided
medially, do not show dilatations, and the openings of the coxal sockets do not face
outward; in consequence the coxae were directly ventrally, rather than laterally or
dorsolaterally. Spiracles were present, situated essentially as in Acantherpestes, but if
there were also openings for exsertile sacs I have not been able to detect them in the few
specimens that I have at hand. As a rule, species of Acantherpestes greatly exceed those
of Euphoberia in size, but the small Acantherpestes from the English Coal Measures
described by Gill (1924) and Brade-Birks (1928) does not appear to have been much
larger than some examples of Euphoberia.

The genus Sandtheria Fritsch, 1899, shows some interesting euphoberiid resemblances.
The dorsal midline is flanked on each side by a row of simple subdorsal spines. However,
laterally, on each side, instead of the large lateral spine of Acantherpestes, the metazonites
of Sandtheria bear a single small node, smaller than the subdorsal spines. The ventral
side of Sandtheria is unknown. Apparently the spines and their arrangement in Chonio-
notus Jordan, 1856, are similar to those of Sandtheria, and in the absence of the char-
acteristic lateral spine of Acantherpestes, the Jordan genus bears no real resemblance to
the latter, despite the implications of Meek and Worthen (1868a) and Scudder (1882,
1885) revived more recently by Hoffman (1969). Chonionotus, contrary to Hoffman, has
not been reported from North America; the type species, Chonionotus lithanthraca is
derived from beds of Westphalian age near Saarbriicken, West Germany.

Species presently comprising the genus Paleosoma Jackson et al, 1919, from the English
Coal Measures, were originally attributed to Acantherpestes and Euphoberia by Baldwin
(1911). Paleosoma is clearly distinct from either of the latter genera, being extremely
“flat-backed” and having lateral extensions of the tergites in the form of keels, very
short prozonites, no subdorsal spines, and two distinct pleurites per tergite.

Ironically enough, the myriapods from Nyran which Fritsch (1899) attributed to
Acantherpestes come close to Scudder’s “long-flanked” concept of Acantherpestes, and
consequently differ in that respect from Acantherpestes proper. The subdorsal spines of
the Czechoslovakian species are long, robust and bifurcate, contrasting with the reduced,
simple or nodelike subdorsal spines of Acantherpestes, and their “lateral” spines are
directed dorsolaterally, rather than subhorizontally or horizontally, as in Acantherpestes.
As regards the sternites, neither Fritsch nor Verhoff (1926) appear to have taken into
account the fact that in Acantherpestes, as Scudder’s figures (1882, pi. 11, figs. 1-4)
indicate, the metasterna and prosterna are not divided, as they are in Nyran forms. Of
the three structures displayed in Verhoff’s “coxosternopleurites” the outermost certainly
has the appearance of a spiracle and the inner represents a coxal socket. The third

J. J. BURKE

NO. 17

feature, which occurs between the two just cited, but nearer the coxal socket, although
approximating in position the coxal sac opening in certain modern millipedes, is much
larger and more complex than that of any millipedes of which I know, and may mark
the location of an organ with a function other than those of respiration or water absorp-
tion. Certainly in position it does not correspond to the “cups” which are situated
medial to the coxal sockets of Acantherpestes, in essentially the same situation as the
structures in Symphyla that contain exsertile sacs.

It is obvious that these species which Fritsch attributed to Acantherpestes clearly
represent another and yet to be established genus. Also, a new genus is called for to in-
clude the Nyran taxa which Fritsch grouped under Euphoberia. These differ from both
Acantherpestes and Euphoberia in the rounding and depth of their flanks, in type and
disposition of spines, in showing medial separation of the sternites, and in having short
prosterna devoid of spiracles, along with long metasterna which carry sternal spines.

Acantherpestes clarkorum* sp. nov.

Figs. 5, 6

Diagnosis: A large species, approaching Acantherpestes major in size.
Anterior ridge occupies less than half the length of metazonite, and bears
two subdorsal spines, here reduced to nodes. Small tubercle at or near
outer termination of lateral furrow. Lateral spines large, length of each
nearly equal to width of body of metazonite, and bifurcate at midlength.
Posterior prong longest, bowed gently posterolaterally. Anterior prong
about one-third length of posterior, extends anterolaterally in gentle arc
recurving toward tip. Basal spinelets large, exceeding anterior prong in
length, the anterior recurved, the posterior nearly straight.

Holotype: CMNH 3917, a string of 25 diplosomites or portions of di-
plosomites preserved in dorsal view.

Occurrence: Siltstone overlying Lower Freeport Coal (Westphalian D),
Allegheny Group, Pennsylvanian Series, Upper Carboniferous.

Locality: Coal strip mine about 0.8 mile (1.3 km) south of the village of
Browns Chapel, Clinton District, Monongalia County, West Virginia,
on the south side of the Gladesville-Halleck road, 0.5 mile (0.8 km) east
of the intersection of that road and U.S. Route 119 (Lat. 39° 29' 15" N,
Long. 79° 54' 45" W) United States Geological Survey 7.5' Gladesville,
West Virginia quadrangle.

* The species name is in recognition of Bruce, Alan, and Quentin Clark, who found the
specimen upon which the species is based.

1973

ACANTHERPESTES

Description: The specimen exhibits, in dorsal view, and in various stages of preservation,
25 tergites disposed in a sinuous curve, and measures somewhat more than 25 cm over
the curvature. There is no definite indication of either head or telson, although an in-
determinate remnant beyond and to the right of the anterior end may represent a part
of the head. In general the segments of the posterior portion show better preservation.
All of the tergites have undergone compression to some extent, and some show longi-
tudinal cracks as well. Counting from the anterior end, to and including tergite 13 there
is noticeable flattening of these elements, and the surfaces of the segments are obscure,
although the lateral spines of the right side are well shown. However, all of the lateral
spines, which were once rounded in cross-section, are now flattened and almost paper
thin in places. In comparison with segments of Acantherpestes which have not been dis-
torted, tergites 14 to 20 appear to have suffered least damage and compression. The last
three tergites are much flattened, having split along the midline and spread apart; the
posteriormost is incomplete.

The surfaces of the metazonites are elevated above those of the prozonites, and each
metazonite along its anterior border is fused with a prozonite. The prozonites are smooth
and in life, probably to a considerable extent, each was overlapped by the metazonite
of the tergite anterior to it. Here, however, some of the prozonites show greater exposure
than normal and some are entirely exposed, possibly because after death, the decompos-
ing body of the animal was torn by water currents prior to burial. The prozonites are
less than the metazonites in length, and show their greatest length along the midline
where the anterior border comes to an apex.

The anterior ridge is not especially prominent and tends to diminish sharply in height
laterally. As a rule, it occupies less than half the length of the metazonite. On each side
of the midline it bears a single subdorsal spine, here reduced to a node. Most of these
nodes are broken away at the top, but the right subdorsal spines of the third tergite
from the posterior end of this specimen is complete. A few others are essentially complete,
and broken portions extracted from an overlying slab of rock into which the spines
extended confirm that they were low nodes, rather than spikelike spines. The subdorsals
are round to somewhat attenuate transversely, and their anterior slopes are continuous
with the anterior slopes of the metazonites.

The lateral furrows are shallow where they originate posterior to the subdorsal spines,
but expand and deepen in their anterolateral course. Anteriorly they are walled by the
posterolateral slopes of the anterior ridges. A small lateral tubercle is usually found at
the termination of the lateral furrow near the base of the anterior spinelet of the lateral
spine.

Bounding the lateral furrows posteriorly are the moderately developed posterior lobes.
Each is narrow and least swollen where it originates posterior to the subdorsal spines,
becoming inflated and gradually expanding anteriorly before merging with the lateral
spine. Most of the posterior lobes of this specimen have suffered extensive damage.
Nevertheless, several of them preserve the gridlike ornament which seems generally to
characterize these regions of the metazonites of Acantherpestes.

The midportion of the metazonite posterior to the anterior ridge is flattened or gently
concave and meets with a moderate posterior ridge which extends transversely, but not
beyond the posterior lobes. In general these ridges are poorly preserved in CMNH 3917.

The lateral spines arise from the sides of the metazonites. If the tergites of this fossil
retained their original curvature and could be viewed in cross-section, the lateral borders

J. J. BURKE

NO. 17

of the metazonites would be seen to extend beyond the spine bases. In view of the state
of preservation of this specimen, no reliable measurements of the width of metazonites
in relation to length of lateral spines can be obtained. It appears to me that the spine
length may have been nearly equal to the width of the metazonite, but this is only a
rough estimate.

These spines extend directly outward from the sides of the metazonites before bifur-
cating at midlength, although they expand slightly before branching into two prongs.
The posterior prong is the longer and indicates the greatest length of the spine. It pro-
longs the posterior border of the main shaft without interruption, although from the
place of bifurcation it bows gently posterolaterally. The anterior prong is about one-third
the length of the posterior, extends anterolaterally in a gentle arc, and is actually slightly
recurved near the tip.

The basal spinelets are relatively quite large. The anterior spinelet arises nearest to
the base of the spine proper. It is nearly two-fifths longer than the anterior prong of the
latter, and shows the same tendency to recurve. The anterior spinelet overlaps the
posterior spinelet of the spine preceding it. The posterior spinelets are about a millimeter
shorter than the anterior spinelets, but show slight curvature.

It is difficult to obtain meaningful measurements of a compressed specimen such as
this, consequently the following figures (in mm) are, at best, only approximate: Length,
tergite, 9.5; length, metazonite, 5.7; width, metazonite, 20.0 (estimated); length, lateral
spine, 20.0 (estimated); length, posterior prong, 9.6; length, anterior prong, 2.9; length,
anterior spinelet, 4.9; length, posterior spinelet, 4.0.

Discussion: The holotype of Acantherpestes clarkorum appears well differentiated from
certain previously described specimens which have been attributed to Acantherpestes,
although the specific relationships of some of the latter remain to be clarified. The strong
anterior ridges and the type of subdorsal spines (as indicated by spine bases) exhibited
by the metazonites of Acantherpestes major do not characterize the metazonites of Acan-
therpestes clarkorum, and the single lateral spine associated with the type of Acanther-
pestes major is entirely different from those of my West Virginia species. The superb
specimen in the collection of the National Museum of Natural History, USNM 33038,
described by Scudder (1882, p. 151-154, pi. 11, figs. 6, 8, 11) as Acantherpestes major
is clearly distinguishable from Acantherpestes clarkorum if only on the basis of its nearly
evenly bifurcate lateral spines and its characteristic laterally curving subdorsals.

Scudder also (1882, p. 154, 155, pi. 11, figs. 1-4) included under Acantherpestes major
another specimen, USNM 33039, which on examination proves to represent a species
entirely distinct from the latter, and also from the presumably yet to be established
species to which USNM 33038 should be attributed. Scudder did not orient USNM
33039 correctly; in his figure 1 (op. cit.) the four articulated tergites showing in dorsal
view are posteriorly disposed in relation to the rest of the segments. The lateral spines,
which Scudder called “pleurodorsals” are inaccurately represented. Two of them are
sufficiently preserved to show that they closely resembled those of Acantherpestes clark-

Fig. 5. Acantherpestes clarkorum sp. nov. Holotype, C.M.N.H. 3917, from siltstones over-
lying the Lower Freeport Coal, Allegheny Group, Pennsylvanian, near Browns Chapel,
Clinton District, Monongalia County, West Virginia. Dorsal view, X 1.

1973

ACANTHERPESTES

J. J. BURKE

NO. 17

orum. The subdorsals flank the midline and are reduced to nodes, as in my species, and
the resemblance extends even further, for in USNM 33039 small lateral tubercles also
mark the outer terminations of the lateral furrows. The tilted anteriormost metazonite
of USNM 33039 appears to have undergone little damage and its gentle curvature from
side to side indicates that the tergites were not strongly arched in cross-section. Unfor-
tunately, only small portions of the prozonites are preserved, and the compressed lateral
spines are difficult to trace with certainty, but I think this specimen may prove to be
conspecific with Acantherpestes clarkorum.

Scudder (1890, p. 424-426, pi. 33, figs. 1, 4) described three additional specimens,
all of which, despite the poor quality of his figures, seem assignable to Acantherpestes.
I have not been able to study this material at first hand, because I do not know where it
is reposited, if indeed it is still preserved. Under the name of Acantherpestes inequalis,
Scudder included two specimens. The first of these (op. cit. p. 424, 425, pi. 33, fig. 2)
shows several fragmentary lateral spines, and one nearly complete, which closely re-
semble those of Acantherpestes clarkorum. Other details of the figure are too vague for
comparison, however. The second specimen (idem. p. 426, pi. 33, fig. 4) does not seem
at all related to the first, but the spines illustrated suggest to some extent lateral spines
of the type which I attribute to Acantherpestes major.

As noted above, Scudder (1890, p. 426, pi. 33, figs. 1, 3) described a third specimen at
that time. To this he gave the name of Euphoberia hystricosa, but I have no doubt that
the species should properly be referred to Acantherpestes. Scudder was mistaken in his
orientation of the animal. The lateral furrows indicate that his “shorter anterior por-
tion” is the anterior portion of the metazonite, and the “longer and blunter” prong of
the lateral spine is the posterior, as is generally the case in Acantherpestes. The long,
robust anterior ridge and a lateral spine described as having prongs “. . . only slightly
divergent and subequal . . .” with basal spinelets “. . . apparently clearly separated . . .”
from the spine shaft clearly distinguish this species from Acantherpestes clarkorum.

Scudder was in error in stating that this tendency for the basal spinelets not to merge
with the shaft of the lateral spine, and the presence of an anterior ridge on the meta-
zonite, does not characterize Acantherpestes ferox (Salter). True, Salter’s (1863, fig. 8)
original illustration does not clearly indicate an anterior ridge per se, and it is quite mis-
leading in showing the anterior basal spinelets as greatly exaggerated in size and forming
integral parts of the lateral spines. However, the British Museum cast of Salter’s type
shows relatively small anterior spinelets, rather distinct from the main shaft, and small
but definite anterior ridges. It is also evident from the figures of Woodward (1887),
Gill (1924), and Brade-Birks (1928) that the somewhat disparate basal spinelets are
characteristic of English Coal Measures representatives of Acantherpestes. In this, as
well as in their smaller size, they differ from Acantherpestes clarkorum; Acantherpestes
ferox differs also in having spikelike subdorsal spines, and the form, from Crawcrook
described by Gill and Brade-Birks is distinguished from my species by the exception-
ally long posterior prongs of the lateral spines.

The environmental relationships of Acantherpestes have given rise to considerable
discussion. Scudder (1882) originated the concept that these were amphibious myriapods,
basing his conclusions on the structure of the feet, which he regarded as adapted for
swimming, and the presence on the sternites of so-called branchial cups, which he
interpreted as housing gill-like organs used for respiring under water. However, the
elongate foot of Acantherpestes, with podomeres described by Scudder (1882, p. 146)

1973

ACANTHERPESTES

as . . not cylindrical but compressed and slightly expanded, strengthened also on the
flattened surface by longitudinal ridges . . .” seems to me to be better interpreted as a
powerful and efficient walking limb, resembling in structure the walking legs of some
terrestrial insects, notably beetles, in being adapted both for bearing the weight of those
heavy arthropods and for efficient locomotion on land.

As for the “branchial cups,” they probably housed exsertile sacs of the type found in
Symphyla, and occupy the same position as the exsertile sac openings in the latter. In
Hansiella agilis, Teigs (1947) has demonstrated that these sacs are used for the absorp-
tion of water. Similar structures are found in Pauropoda, in primitive insects, and are

Fig. 6. Acantherpestes clarkorum sp. nov. Restoration of tergites (not corrected for com-
pression). Dorsal view, X 1. Abbreviations: Apr, anterior prong of lateral spine; Ar,
anterior ridge; Ast, anterior spinelet; Lf, lateral furrow; Ls, lateral spine; Mtz, meta-
zonite; PI, posterior lobe; Ppr, posterior prong of lateral spine; Pr, posterior ridge;
Prz, prozonite; Pst, posterior spinelet; Sds, subdorsal spine; Tb, tubercle.

J. J. BURKE

NO. 17

represented by coxal sacs in some modern millipedes. In the Myriapoda, one of their
functions seems to be that of absorbing water as a means of combating dessication, and
their presence in Acantherpestes does not demonstrate that representatives of that genus
were in consequence amphibious.

What appears to have been the evolutionary sequence leading up to Acantherpestes
also supports the conclusion that these were terrestrial animals. The stock from which
this line was derived were probably small myriapods, possibly near Euphoberia in size,
or even smaller. They must have been long-flanked, with cylindrical bodies, which bore
upright or nearly upright spines; the legs were probably relatively short and not par-
ticularly stout. Evolution evidently proceeded in the direction of increase in size and
development of legs suited to bear the increased weight, along with elongation of these
appendages, to provide speedier locomotion. With increased size and faster gait, pred-
ators became less of a problem, and there was less need for spines purely as a means of
protection. The long flanks were lost, the subdorsal spines became reduced, and the
body expanded laterally — probably initially to provide shelter for the lengthening legs.

What followed appears to have been one of the most fascinating developments in the
history of the Myriapoda. As the body expanded laterally to produce the “flat back”
characteristic of Acantherpestes, the lateral spines came to be directed essentially hori-
zontally, providing further protection for the lengthening legs, thus functioning in the
same way as the paranota or keels of modern millipedes. It is also likely that in species
such as Acantherpestes clarkorum, in which the subdorsal spines were much reduced, the
broad tergites and extended lateral spines were employed to separate masses of matted
leaves as the animal forced its way into them in search of food. The lifting and pene-
trating power in this case could have been supplied, as noted by Manton (1954, 1961)
in modern millipedes, by drawing in the legs and pushing upward and forward with
them. In this connection it might be noted that the anterior lateral spines of Acanther-
pestes inequalis, progressively decreasing in width cephalad, formed, together with the
head, a wedge that would have facilitated penetration of leaf litter by the animal.

Figure 7 represents an attempt at restoration of a diplosomite of Acantherpestes as
seen in posterior view, illustrating in cross-section the relationship of the essentially
horizontal lateral spines to the elongate legs.

Protected from most predators by sheer size, Acantherpestes was probably able to
move about freely. These myriapods may have ventured into open areas of the lowlands
bordering the Carboniferous swamps, and were probably able to withstand some exposure
to direct sunlight, as Causey noted for Brachycybe (Manton, 1961). Having retained the
water-absorbing exsertile sacs, it seems reasonable that, as Manton suggests for Brachy-
cybe and related millipedes, Acantherpestes may even have obtained water from drops
of dew. By this device the Carboniferous form could have staved off dessication under
dry conditions.

1973

ACANTHERPESTES

Fig. 7. Diagrammatic posterior view of a body segment of Acantherpestes (author’s
interpretation).

REFERENCES CITED

Baldwin, W., 1911, Fossil myriapods from the Middle Coal Measures of Sparth Bottom,
Rochdale, Lancashire: Geol. Mag. [Great Britain], v. 47, p. 74-80, 3 pis.

Barlow, J. A., 1969, New Paleozoic animal fossils found in West Virginia: West Virginia
Geol. Survey Newsletter, 13th issue, p. 3, 1 fig.

Brade-Birks, S. G., 1928, An important specimen of Euphoberia ferox from the Middle
Coal Measures of Crawcrook: Geol. Mag. [Great Britain], v. 65, p. 400-406, 1 pi.,
3 text figs.

Fritsch, A., 1899, Fauna der Gaskohle und der Kalksteine der Performation Bohmens:
Prague, v. 4, pt. 1, pi. 1-32, pt. 2, p. 33-64, 8 pis., figs.

Gill, E. L., 1924, Fossil arthropods from the Tyne coal field: Geol. Mag. [Great Britain],
v. 61, p. 455-471, 1 text fig.

Hennen, R. V. and Reger, D. B., 1913, Marion, Monongalia and Taylor Counties:
W. Va. Geol. Survey County Reports, p. 351.

Hoffman, R. L., 1969, Myriapoda, exclusive of Insecta: in Treatise on invertebrate
paleontology, R. C. Moore, ed., Part R, Arthropoda 4, v. 2, p. 572-606, 22 text
figs.: Lawrence, Kansas, Kansas Univ. Press and Geol. Soc. America.

Jackson, J. W., Brade-Birks, H. K., and Brade-Birks, S. G., 1919, Notes on Myriapoda.
XIX. A revision of some fossil material from Sparth Bottoms, Lancs.: Geol. Mag.
[Great Britain], v. 55, p. 406-411, 1 pi., 3 text figs.

J. J. BURKE

NO. 17

Jordan, H., and von Meyer, H., 1856, Ueber die Crustaceen der Steinkohlenformation
von Saarbriicken: Paleontographica, v. 4, p. 1-15, 2 pis.

Manton, S. M., 1954, The evolution of arthropodan locomotory mechanisms, Pt. 5.
The structure, habits and evolution of the Diplopoda: Linnean Soc. London (Zool-
ogy) Jour., v. 42, p. 299-364, 4 pis., 8 text figs.

1961, The evolution of arthropodan locomotory mechanisms, Pt. 7. Func-
tional requirements and body design in Colobgnatha (Diplopoda), together with a
comparative account of diplopod burrowing techniques, trunk musculature and
segmentation: Linnean Soc. London (Zoology) Jour., v. 44, p. 383-461, 3 pis.,
35 text figs.

Meek, F. B. and Worthen, A. H., 1868a, Preliminary notice of a scorpion, a Eurypterus?
and other fossils, from the Coal-measures of Illinois: American Jour. Sci., 2nd ser.,
v. 46, p. 25-27.

1868b, Articulate fossils of the Coal Measures: Illinois Geol. Survey Bull.

v. 3, p. 558-559, 1 text fig.

Salter, J. W., 1863, On some species of Eurypterus and allied forms: Geol. Soc. London
Quart. Jour., v. 19, p. 86, 87, fig. 8.

Scudder, S. H., 1882, Archipolypoda, a subordinal type of spined myriapods from the
Carboniferous Formation: Boston Soc. Nat. History, v. 3, 4 pis., 8 text figs.

1885, Myriopoda, in Zittel, K. A., Handbuch der Paleontologie: Munchen

and Leipzig, v. 2, abt. 1, p. 721-731, 12 text figs.

1890, New Carboniferous Myriapoda from Illinois: Boston Soc. Nat. History,

v. 4, p. 417-442, 6 pis.

Teigs, 0. W., 1947, The development and affinities of the Pauropoda, based on a study
of Pauropus silvaticus: Quart. Jour. Micro. Sci., v. 88, p. 165-336, 10 pis., 29 text figs.

Verhoff, K. W., 1926, Fossile Diplopoden: in Bronn, H. G., Klassen und Ordnungen des
Tierreichs: Leipzig, Acad. Velag., v. 5, pt. 2, bk. 2, p. 330-359, pis.

Woodward, H., 1872, A monograph of the British fossil Crustacea belonging to the order
Merostomata: London Paleontographical Soc., p. 172-174, 2 text figs.

1887, On some spined myriapods from the Carboniferous Series of England:

Geol. Mag. [Great Britain], v. 24, p. 1-10, 1 pi., 3 text figs.

MANUSCRIPT RECEIVED JUNE 8, 1973

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO JUNE 17, 1974 NUMBER 18

ADDITIONS TO

THE CRINOIB FAUNA OF THE AMES LIMESTONE,
BROOKE COUNTY, WEST VIRGINIA

J. J. BURKE

Senior Scientist, Cleveland Museum of Natural History, Cleveland, Ohio;

Research Associate, West Virginia Geological Survey

ABSTRACT

Descriptions of inadunate crinoids from the Ames Limestone, Conemaugh Group,
Upper Pennsylvanian, of Brooke County, West Virginia, and Guernsey County, Ohio,
pertain to the following taxa: (1) Delocrinus segedii sp. nov. related to Delocrinus broum-
villensis Strimple, but differing in lesser height of dorsal cup, deeper basal impression,
higher extent of basals on cup walls, and smaller primanal; (2) Appalachiacrinus erwini
gen. et sp. nov., a laudonocrinid, differing from other laudonocrinids in greater height of
dorsal cup, resembling Laudonocrinus in flatness of infrabasals, slight upflaring of in-
frabasal circlet, greater width and lesser curvature of C radial, and somewhat similar
axillary primibrachs, but also resembling species of Anchicrinus and Athlocrinus in
convexity of basal, radial and anal plates and presence of pits at angles of plates; and
(3) Parethelocrinus occultater sp. nov., characterized by a globose dorsal cup, with
shallow and wide basal concavity, convex downflaring infrabasals, radials having
prominent flat forefacets, tertanal concealed by C primibrach but partly within cup,
quartanal resting on secundanal; cup ornament finely granulose.

INTRODUCTION

As indicated in a previous study (Burke, 1973), several species of
Pennsylvanian crinoids from the Ames Limestone, Conemaugh Group,
of Brooke County, West Virginia, are known only from the Upper
Pennsylvanian of the Appalachian region. Subsequent study has deter-
mined three additional species from the Brooke County Ames, hitherto
undescribed and likewise unknown beyond the confines of the Appalachian
basin. These taxa, mainly represented by specimens in the collection of
the Cleveland Museum of Natural History (CMNH) but in one instance
also by material loaned by the Carnegie Museum (CM), are described in
the following pages.

J. J. BURKE

NO. 18

I wish to express my appreciation to Dr. John Carter and the Carnegie
Museum for the loan of specimens. I am also grateful to the West Virginia
Geological Survey for encouragement and financial aid in support of this
investigation. Dr. N. Gary Lane of the Department of Geology, Indiana
University, kindly loaned me literature pertinent to this study.

My thanks go to Mr. Bruce Frumker for the photographs from which
the illustrations were taken, and to my wife, Emily, for her aid in preparing
the manuscript.

SYSTEMATIC PALEONTOLOGY

Class CRINOIDEA Miller, 1821
Family CATACRINIDAE Knapp, 1969
Genus DELOCRINUS Miller and Gurley, 1890
Delocrinus segedii* sp. nov.

Figs. 1-8

Diagnosis: Dorsal cup resembles that of Delocrinus brownvillensis Strim-
ple, with subround to round outline in dorsal view, characteristic curvature
of lateral walls, impressed primanal, and fine granulose ornament, but
cup height shorter, basal impression deeper, basals extend higher on cup
walls, and primanal smaller.

Types: Holotype CMNH 3847 (C radial damaged— restored) ; paratypes
CMNH 3835, CM 33901 and CM 33902, all dorsal cups.

Occurrence: Ames Limestone, Conemaugh Group, Upper Pennsylvanian.

Localities: Holotype CMNH 3847 and paratype CMNH 3835 from road
cut on south side of Interstate 70 and east of junction with Ohio Route 513,
NW M sec. 25 (lat 40° 03' 07" N., long 81° 19' W.) Oxford Township,
near Middlebourne, Guernsey County, Ohio. Paratypes CM 33901 and

*Named for Mr. Robert Segedi, Science Instructor, Cleveland Museum of Natural
History, who collected the Ohio specimens.

Figures 1-8. Delocrinus segedii sp. nov. Figs. 1-4 holotype CMNH 3847 (damaged C radial
restored). Fig. 1, dorsal view. Fig. 2, posterior view. Fig. 3, anterior view. Fig. 4, ventral
view. Figs. 5-8 paratype CM 33901. Fig. 5, dorsal view. Fig. 6, posterior view. Fig. 7,
anterior view. Fig. 8, ventral view. All x 1.5.

1974

CRINOIDS FROM THE AMES LIMESTONE

J. J. BURKE

NO. 18

CM 33902 from Tunnel Road Cut, West Virginia Route 67 (lat 40° 14' 24"
N., long 80° 35' 53" W.) near McKinleyville, Brooke County, West
Virginia.

Description: Dorsal cup of medium size, low truncate bowl-shaped,
about % as high as wide. Outline subround in dorsal view, pentagonal
in ventral view. Height and width dimensions of basal impression less
than half those of the cup. Stem cicatrix small. Infrabasals extend for
slightly less than half the height of basal impression, steep walled prox-
imally, sloping outward and downward distally, merging with slopes of
basals without angularity. Basals with moderate downward slopes within
impression, flattening, but still slightly concave at basal plane; distal
slopes, along cup wall less steep, and gently convex. Tips of basals extend
to about half the cup height.

Radials flare outward and upward, moderately convex longitudinally,
less convex transversely. Forefacet very slight, broadly lunate. Transverse
ridge and outward-facing lateral lobes visible in lateral view (barely
showing in paratype CM 33901— result of compaction). Outer marginal
ridge sags downward; broadly bowed in ventral view. External ligament
pit slitlike, its outer ridge denticulate. Transverse ridge relatively delicate,
denticulate, compressed adjacent to outer ligament pit. Lateral furrow
moderately deep and broad, oblique ridge denticulate. Adsutural slopes
low, but prominent, expanding widely toward body cavity, where high
lateral lobes overhang them. Intermuscular notch broadly V-shaped,
intermuscular furrow narrow, flanked by round-basined muscular areas.

Primanal concave from side to side in midregion; portion within the
cup occupies a little less than a third of the cup height. Distally primanal
slopes inward and upward; in holotype, distal facet shows small basin on
each side of midline.

Ornament consists of fine granules, on primanal, radial and basal
plates; not shown on proximal portions of basals of Ohio types (CMNH
3847 and CMNH 3835).

Measurements: Linear measurements in mm taken on holotype, CMNH
3847: Dorsal cup height, 8.7, width, 23.5, H/W ratio, 0.35; basal impres-
sion, height, 3.8, width 10.5; stem impression width 1.9; basal (EA)
length, 9.6, width 9.4; radial (A) length, 7.8, width, 13.5; length suture
between basals, 6.7; length suture between radials, 4.3; primanal height,
5.5, width 4.1.

1974

CRINGIDS FROM THE AMES LIMESTONE

Discussion: This species belongs to a Delocrinus lineage fairly well rep-
resented in Upper Missouri and Virgil sediments, characterized by dorsal
cups showing granulose ornament and composed of rather massive plates
which are easily damaged. The holotype, CM 3847, had most of the C
radial broken away when found; the plate has been restored. Proximal
portions of the basals and all of the infrabasals of the West Virginia
paratypes were lost prior to deposition. 1 collected the latter specimens
several years ago, but it was not until the better Ohio material was found
that I felt that the species could be established securely.

Strimple (1949) pointed to the more rounded outline of the dorsal cup
in dorsal view, coupled with the less angular longitudinal slopes of the
basals, as characters distinguishing Delocrinus brownvillensis from Delo-
crinus vulgatus. These same features also appear to distinguish Delocrinus
segedii from both Delocrinus vulgatus and Delocrinus verus . The height/
width ratio is about the same for dorsal cups of the latter three species
(about 0.35), but this ratio for Delocrinus brownvillensis , based on
Strimple’s values for height and width, is 0.39, rather than 0.35, the
figure given by him. Height and width measurements of hypotype speci-
mens of Delocrinus brownvillensis indicate even higher height/width
ratios, up to 0.41 (Pabian and Strimple, 1973). The depth of the basal
impression is less in Delocrinus segedii than in Delocrinus verus and
Delocrinus vulgatus. The cup of Delocrinus segedii is ornamented; cups of
Delocrinus verus and Delocrinus vulgatus are smooth.

Family LAUDONOCRINXDAE Moore and Strimple, 1973
Genus APPALACHIACRINUS* gen. nov.

Diagnosis: Dorsal cup low truncate bowl-shaped, nearly times as
wide as high. Outline irregularly hexagonal in dorsal view. Posterior
interradius not impressed. Interradiai notches faint. Infrabasals flat, basal
and anal plates mildly convex, major portions of radials and first primi-
brachs swollen or bulbous. Ornament very fine irregular ridges. Infrabasal
circlet very slightly upflaring, not visible in lateral view. Basals curve

*The generic name is in reference to the Appalachian region where the type species
was found.

J. J. BURKE

NO. 18

downward to basal plane, then upward along cup wall. Slight pits or
impressions where distal extremities of basals meet corners of radials and
anals. C radial widest, least convex, with distal slopes relatively gentle.
Other radials more convex with distal slopes curving inward strongly.
Radial articular facets peneplenary. Primanal, secundanal and tertanal
plates in the cup, in normal arrangement. First primibrachs spinose and
axillary.

Type species: Appalachiacrinus erwini sp. nov., here designated.

Appalachiacrinus erwini** sp. nov.

Figs. 9-12

Diagnosis: As for the genus— see above.

Holotype: CMNH 3834, a dorsal cup retaining four first primibrachs.

Occurrence: Ames Limestone, Conemaugh Group, Upper Pennsylvanian.

Locality: Tunnel Road Cut, West Virginia Route 67 (lat 40° 14' 40" N.,
long 80° 35' 53" W.) near McKinleyville, Brooke County, West Virginia.

Description: Dorsal cup low truncate bowl-shaped, nearly 23^ times as
wide as high (H/W = 0.41); outline irregularly hexagonal in dorsal view.
Posterior interradius not impressed. Ornament very fine irregular ridges.

Infrabasal circlet nearly one-third diameter of cup, upflaring very
slightly, not visible in lateral view. Stem wide, occupying most of circlet;
closely surrounded by infrabasals. Infrabasals flat, with wide distal angles;
distal surfaces flush with those of basals along common sutures. Basals
gently convex, sloping downward from contact with infrabasals and then
upward along cup wall. Slight pits where distal terminations of basals
meet corners of radial and anal plates.

C radial widest and least convex of radial plates; distal slopes fairly
gentle. Remaining radials bulbous; least convex from side to side; distal

**The specific name is in honor of Dr. Robert B. Erwin, State Geologist of West
Virginia.

1974

CRINOIDS FROM THE AMES LIMESTONE

Figures 9-12. Appalachiacrinus erwini gen. et sp. nov. Holotype CMNH 3834. Fig. 9,
dorsal view. Fig. 10, posterior view. Fig. 11, anterior view. Fig. 12, ventral view. All x 3.

slopes curving inward sharply in relatively broad areas defined by strongly
arcuate forefacets. External surfaces slightly impressed in notches at
summits of interradial sutures; articular facets peneplenary and mod-
erately declinate. External marginal ridge bowed, sags downward and
faces outward together with slitlike external ligament pit. Transverse
crest low, denticulate; lateral furrows broad, shallow. Adsutural slopes
gentle but extensive. Internal notch V-shaped; muscle area basins round.

Primanal, secundanal, and tertanal gently convex; arrangement normal.
Two additional multifaceted anal plates present but dissociated.

First primibrachs axillary; all except B primibrach preserved. Plates
spinose, major portion bulbous, not compressed from side to side. Spines
arise high on plates, tips not preserved. Strong ridge separates right and
left f acetal areas. A single short secundibrach with part of articular surface
facing outward rests on the dorsal faces of the secundanal and tertanal.
It shows a wide V-shaped intermuscular notch and what appears to be a
short intermuscular furrow flanked by shallow basins on each side.

J. J. BURKE

NO. 18

Measurements: Linear measurements, in mm, taken on holotype, CMNH
3834: Dorsal cup height, 4.4, width 10.7; stem width, 2.2; infrabasal
circlet width, 3.4, basal (AB) length, 2.8, width 3.5; radial (A) length 3.5,
width 5.6; length suture between basals, 1.0; length suture between
radials, 2.2; primanal length, 3.2, width 1.7; secundanal length, 2.9,
width 2.3; tertanal length 3.0, width, 2.2; first primibrach (E) length
along lateral suture, 2.5, width 4.9.

Discussion : This interesting little crinoid bears resemblances to Laudono-
crinus, Anchicrinus, and Athlocrinus. The flat infrabasals and very
slightly upflared infrabasal circlet are also features of Laudonocrinus, but
Appalachiacrinus erwini differs from Laudonocrinus subsinuatus in not
showing the infrabasals in lateral view. As in Laudonocrinus, the C radial
is the widest and least convex, but the radials generally differ from those
of Laudonocrinus in being for the most part bulbous, although steep
walled distally. The axillary primibrachs are perhaps more like those of
Laudonocrinus than those of any other laudonocrinid genus, but they are
tumid and not compressed from side to side as in Laudonocrinus sub-
sinuatus.

However, no other laudonocrinid genus approaches Appalachiacrinus
in cup height. It resembles Anchicrinus and Athlocrinus, which have much
shallower cups, in the convexity of its cup plates, in showing pits at
the plate angles, and in the lack of an impressed posterior interradius,
such as characterizes some species of the latter genera.

Family CROMYOCRINIDAE Bather, 1890

Genus PARETHELOCRINUS Strimple, 1961
Parethelocrinus occultator* sp. nov.

Figs. 13-15

Diagnosis: Cup less than three times wider than high, constricted at
summit, basal impression shallow and wide, infrabasal circlet relatively
large, downflaring. Sutures impressed, ornament finely granulose. Radials
with prominent flat forefacets. Primanal, secundanal, and lower right
corner of tertanal in cup, quartanal rests on secundanal, not in cup.
Tertanal concealed by C ray primibrach.

*From the Latin occultator (concealer) in reference to the hidden tertanal.

1974

CRINOIDS FROM THE AMES LIMESTONE

Holotype: CMNH 3833, a crushed dorsal cup with three axillary primi-
brachs, one preserving portions of arms.

Referred specimen: CMNH 3802, associated cup and arm plates, including
an axillary first secundibrach.

Occurrence: Ames Limestone, Conemaugh Group, Upper Pennsylvanian.

Locality: Tunnel Road Cut, West Virginia Route 67 (lat 40° 14' 24" N.,
long 80° 35' 53" W.) near McKinley ville, Brooke County, West Virginia.

Description: Dorsal cup low truncate globe-shaped. Cup of holotype
crushed, but must have been less than three times as wide as high. Basal
impression shallow and wide, infrabasal circlet relatively large. Sutures
between cup plates in shallow V-shaped impressions; ornament very
finely granulose.

Stem prominent; infrabasals convex, downflaring. Basals gently convex
within basal impression, with moderate outward and downward slopes
to junction with radials; strongly convex beyond that region, curving
abruptly upward along outer wall of cup. Radials moderately convex,
widest at proximal lateral angles. Forefacets prominent in ventral view,
flattened. External ligament pit slitlike. Transverse ridge of moderate
height and denticulate. Internal notch V-shaped; intermuscular furrow
narrow. Muscle areas with shallow curved grooves.

Three anal plates in cup. Primanal large, quadrangular; secundanal
medium sized, pentagonal, its left side almost straight and nearly vertical
in arrangement, meeting D radial and D primibrach; bears quartanal
dorsally, extends in V-shaped angle to right, meeting with C radial above
and making very wide contact with primanal below; narrow proximal
side abuts against truncate CD basal. Surface exposure of quartanal
slight, plate bent inward, expands laterally to left and right in direction
of body cavity, contacts D first primibrach on left, firmly bound to
tertanal at right. Sides of tertanal and quartanal directed inward on the
right, bounding left side and corner of C primibrach. Tertanal sinks slightly
below left lateral lobe of C radial and consequently is within cup, but entire
plate would be concealed in external view if C first primibrach were in
place. Distally, quartanal and tertanal bear facets for two additional
tube plates. Another anal plate lies nearby in body cavity.

J. J. BURKE

NO. 18

Figures 13-15. Parethelocrinus occultator sp. nov. Holotype CMNH 3833. Fig. 13, dorsal
view. Fig. 14, slightly oblique ventral view. Fig. 15, posterior view. All x 1.

First primibrachs D, E, and A present; plates short in height, massive,
showing prominent forefacets proximally and having rounded tips that
project outward distally. Articular surfaces robust, transverse crests of
both articular faces of the A plate very strongly and broadly denticulate,
suggesting that this primibrach bore axillary secundibrachs. A single
axillary secundibrach is associated with plates of the referred specimen,
CMNH 3802, indicating that this species had more than 10 arms.

The E primibrach bore two arms; it retains a single first secundibrach
of the left ray and three secundibrachs of the right ray. These plates
appear to have attained biseriality on the second secundibrach. Isolated
secundibrachs show flat lateral sides, prolonged in typical ethelocrinid
fashion.

Measurements: Linear measurements, in mm, taken on holotype, CMNH
3833: Dorsal cup height, 13.0*, width, 34.0*; width stem, 4.0; width
infrabasal circlet, 11.3; basal (EA) length, 11.9, width, 15.5; radial (E)
length, 9.2, width (proximal lateral angle), 16.8, width (summit), 13.8;

1974

CRINOIDS FROM THE AMES LIMESTONE

length suture between basals, 8.7; length suture between radials, 4.6;
primanal length, 12.0*, width, 11.0*, secundanal length 9.0, width 7.0*,
quartanal length, 4.5, width 5.0+ ; axillary primibrach (A) length (along
lateral suture) 5.0, width 13.8.

Discussion: In general, the dorsal cup of this species appears more like
that of Parethelocrinus magnus (Strimple) than any other ethelocrinid
species (cf Strimple, 1949, pi. 2, fig. 1, 3, 5, 6). The basal concavities and
infrabasal circlets are much the same in the two species, and in ventral
view the prominent flat forefacets of the radials are strikingly alike. Even
the shallow grooves of the radial muscle areas appear the same in both
taxa. Parethelocrinus magnus also shows an inward extension of the
tertanal along the left lateral side of the C radial; this is the same region
where the tertanal of Parethelocrinus occultator is found, but the tertanal
and quartanal were not joined in the Strimple species and both plates
articulated with the secundanal.

In a dorsal cup of Dicromyocrinus geminatus (Trautschold) figured by
Yakovlev and Ivanov (1956, pi. 4, fig. 3) the tertanal occupies essentially
the same position as in Parethelocrinus occultator , and probably would not
have been visible in lateral view if the C primibrach were in place. How-
ever, in the Russian specimen, the tertanal is evidently a separate plate
wedged in between the C radial and the secundanal. The quartanal is not
shown. In the holotype of Parethelocrinus occultator, the quartanal and
the tertanal are very closely joined and I cannot find clear evidence of
sutural separation. If the two plates are fused, this may be an abnormal
condition. On the other hand, the position of the tertanal is such that
fusion with the quartanal would contribute to its function as a supporting
plate.

* Estimated.

REFERENCES CITED

Burke, J. J., 1973, Four new pirasocrinid crinoids from the Ames Limestone, Penn-
sylvanian, of Brooke County, West Virginia: Carnegie Mus. Ann. v. 44, p. 157-169.

Pabian, R. K., and Strimple, H. L., 1973, Delocrinus brownvillensis Strimple from the
vicinity of Fairfax, Oklahoma: Oklahoma Geol. Notes, v. 33, p. 17-20, fig. 1.

J. J. BURKE

NO. 18

Strimple, H. L., 1949, Studies of Carboniferous crinoids: I. A group of Pennsylvanian
crinoids from the Ardmore Basin: Paleont. Americana, v. 3, no. 23, p. 5-21, pi. 1-3.
II. Delocrinids of the Brownville formation of Oklahoma: Paleont. Americana, v. 3,
no. 23, p. 22-27, pi. 4.

Yakovlev, N. N., and Ivanov, A. P., 1956, Marine crinoids and blastoids of the Car-
boniferous and Permian deposits of Russia: Geol. Inst. Moscow, Sci. Invest., n. s.,
v. 11, p. 1-142, 21 pis., 23 figs.

MANUSCRIPT SUBMITTED MARCH 19, 1974

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO DECEMBER 30, 1974 NUMBER 19

A NEW SPECIES OF PARALLELOCRINUS FROM THE
VINLAND SHALE, PENNSYLVANIAN, OF KANSAS

J. J. BURKE

Senior Scientist , Cleveland Museum of Natural History;

Research Associate, West Virgina Geological Survey

ABSTRACT

An machinate crinoid, Parallelocrinus mercenarius sp. nov., from the Vinland Shale,
Douglas Group, Virgilian Series, (Pennsylvanian) of Kansas is described. The species is
distinguished by (1) a globe-shaped cup with round and shallowly scalloped outline in
dorsal and ventral views; (2) a wide, steep-walled basal impression; (3) prominent basal
plates with steep proximal slopes; (4) radials with proximal tips well above the basal
plane; (5) hollows that border the plates extend to distal reaches of interbasal borders;
and (6) distinctive rugose ornament. Comparison is made with two crinoid species from
the Permian of Timor. Possibility of relationship to Lopadiocrinus tuber culatus Wanner
is suggested. Close resemblances of species of Parallelocrinus to “ Delocrinus ” rugosus
Wanner in external structures of dorsal cups are noted, but attributed to homemorphy;
radial articular structure indicates that the Wanner species is referable to Apographio-
crinus, consequently it is herewith designated Apographiocrinus rugosus (Wanner, 1916)
comb. nov.

INTRODUCTION

The inadunate crinoid genus Parallelocrinus was proposed by Knapp
(1969) having for its type species Parallelocrinus typus from the Burgner
Formation, Atokan Series, of Missouri. Since then (Burke, 1971) I have
attributed to the genus a second species, Parallelocrinus sturgeoni from
the Ames Limestone, Conemaughan Series, of Ohio. In the present paper
a third species, Parallelocrinus mercenarius sp. nov. from the Vinland
Shale, Virgilian Series, of Kansas, is described.

I wish to thank Dr. Porter M. Kier and Dr. Richard E. Grant of the
National Museum of Natural History, and Dr. Eugene S. Richardson, Jr.
of the Field Museum for the privilege of studying specimens that were
pertinent to this investigation.

I am also grateful to my wife Emily for arranging the illustrations for
this paper, and to Bruce Frumker, Staff Photographer, for photographs
from which the illustrations were made.

J. J. BURKE

NO. 19

SYSTEMATIC PALEONTOLOGY

Class CRINOIDEA Miller, 1821
Family CATACRINIDAE Knapp, 1969
Genus PARALLELOCRINUS Knapp, 1969
Parallelocrinus mercenarius* sp. nov.

Figs. 1-4

Diagnosis: A species differing from Parallelocrinus typus and P. sturgeoni
mainly in the following features: Dorsal cup more rounded in dorsal and
ventral views and globe shaped, rather than bowl shaped, in lateral view,
lateral walls showing greater curvature; base more sharply truncate; basal
impression steeper walled and wider; basals more prominent in dorsal and
lateral views, with steeper proximal slopes; proximal tips of radials arising
higher on cup walls; hollows between plate borders extending beyond
proximal tips of radials and entering slightly along borders between
basals; depressions at proximal tips of radials shallower and broader; cup
surface appears shaggy, showing distinctive rugose ornament.

Holotype: CMNH 3977, a dorsal cup lacking the infrabasal circlet.

Repository: Cleveland Museum of Natural History, Cleveland, Ohio.

Occurrence: Vinland Shale, Stranger Formation, Douglas Group, Upper
Pennsylvanian (Virgilian).

Locality: Homewood, Franklin County, Kansas.

Description: Dorsal cup times as wide as high, and truncate low globe
shaped; outline round and shallowly scalloped in dorsal and ventral views;
right side of cup slightly higher than left. Basal impression steep walled
and wide. Infrabasal circlet missing; height of impression unknown.

Basal plates downflaring steeply for most of their height within the
basal impression, but bending rather abruptly outward on approach to
basal plane; beyond basal plane plates slope gently outward and upward
to their distal terminations.

*The holotype specimen was purchased from a dealer in geological supplies, hence the
specific name. The specimen had been identified as Graffhamicrinus magni ficus (Strimple).

1974

VIRGILIAM PARALLELOCRINUS

Figures 1-4. Parallelocrinus mercenarius, sp. nov. Holotype, CMNH 3977, from the
Vinland Shale, Douglas Group, Homewood, Franklin County, Kansas. Fig. 1, dorsal
view; fig. 2, posterior view; fig. 3, anterior view; fig. 4, ventral view. X 2

Radial plates with proximal tips well above the basal plane and sloping
upward and outward slightly steeper than basals to region of forefacets,
where their surfaces curve inward, with gentler slopes, to summit of cup.
Forefacets prominent, with depths about one-third those of the radial
articular surfaces.

Radial facets face outward. Outer ligament ridge slightly arcuate, sags
below transverse ridge; denticulate, at least in vicinity of slitlike ligament
pit. Transverse ridge sharply defined, denticulate, contracts in vicinity of
ligament pit, then expands on either side, but contracts again opposite
inner reaches of lateral furrows, narrowing sharply to lateral extremities.
Lateral furrows will defined, extend almost opposite extremities of liga-
ment pit, flanked on inner side by steep-walled denticulate oblique ridge.
Small central pit or foramen just below midline of transverse ridge on

J. J. BURKE

NO. 19

inner side. Adsutural slopes steep; adsutural channels widening and
deepening toward exits on internal side. Intermuscular notches broadly
V-shaped. Intermuscular furrows short, angular, terminating between
muscle areas. Muscle areas basin-like, separated also by rounded triang-
ular area that expands upward, terminating below transverse crest
opposite ligament pit.

Anal X convex from side to side, truncates distal tip of elongate CD
basal, is impressed between C and D radials and extents for about a
third of its height above the summits of those plates. It curves inward
and upward, narrowing above, with concave lateral surfaces; distal
articular surface faces both inward and upward and displays depressions
on each side of the midline.

Wide hollows border sutures between cup plates, extending along
interradial sutures, sutures between radials and basals, and entering
slightly along interbasal sutures. Hollows marked by smooth areas,
shared by adjacent plates along interradial and interbasal sutures; con-
fined to basals along sutures between radials and basals. Distal tips of
basals stand out in relief from floors of confluent hollows. Shallow de-
pressions mark common sutures at proximal tips of radials, but small
pits also present.

Ornament strikingly rugose, imparting shaggy appearance to cup;
consists predominately of elongate ridges running parallel to length of
plates (but opposite to length on anal X); ridges foreshortened or node-
like within basal impression, on forefacet, and on anal X.

The cup shows evidence of damage by boring organisms. An ill defined
subcircular depression with a central boss-like elevation, found near the
tip of the EA basal, may represent the work of a gastropod. On the DE
basal an elongate perforation is of the type usually attributed to the
burrowing barnacle Trypetesa. Small round pits on the BC basal prob-
ably represent boring also.

Measurements: Linear measurements, in mm, taken on holotype CMNH
3977: Dorsal cup height, 8.0, width (max.), 20.0, H/W ratio 0.40; basal
impression width, 8.9; EA basal length, 6.7 (appr.), width, 7.1; A radial
length, 6.6, width 10.6; suture between BB, length, 4.9; suture between
RR, length, 3.6; anal X height, 3.5, width, 2.7.

Discussion: Parallelocrinus mercenarius in several respects is more special-

1974

VIRGILIAN PARALLELOCRINUS

ized than Parallelocrinus typus and Paralielocrinus sturgeoni. This is
indicated by the steeper walled and wider proximal portion of the stem
impression (and by inference a more robust stem); more prominent
basals, having steeper proximal slopes; proximal tips of radials arising
higher on the cup walls; presence of shallow depressions where the
proximal tips of the radials meet subjacent basals; and hollows between
plates prolonged slightly, extending to distal reaches of interbasal borders.

The peculiarities which I noted previously (Burke, 1971, p. 201) as
characteristic of Parallelocrinus , notably . . the distal portions of the
basals standing out in relief, with their tips elevated above the hollows’7
are also to be found in at least two specimens of crinoids from the Permian
of Timor. The one, Lopadiocrinus tuberculatus Wanner (1937), not only
shows the features noted above, but also essentially all of the distinctive
surface structures of the cups of Parallelocrinus , including the deep
forefacet, although the hollows between the plates are evidently more
deeply entrenched and extend along the interbasal borders to merge with
the basal impression. The radial articular facets are similar to those of
Parallelocrinus and other Catacrinidae. It must be granted that the
prominent asymmetrical infrabasal circlet and shallow7 basal impression
would seem to exclude derivation of Lopadiocrinus tuberculatus from any
known species of Parallelocrinus , but the possibility of descent of the
Wanner species from Parallelocrinus or a Parallelocrinus- like ancestor
deserves consideration.

In over-all external structure of the dorsal cup, the most striking
similarity to species of Parallelocrinus is found in the Timor species
originally described as “Delocrinus” rugosus by Wanner (1916). I have
seen a specimen of this crinoid, USNM S3982, which is preserved in the
Springer Collection of the National Museum of Natural History. It is
labeled Ceriocrinus rugosus WTanner, and was probably identified by
Wanner himself. In this Timor species we again find the distal reaches of
the basals standing out in relief, and the hollows have the same distribu-
tion as in Parallelocrinus except that as in Lopadiocrinus tuberculatus
they also parallel the interbasal sutures, terminating at the basal im-
pression, just above the basal plane. The basal impression is moderate;
it resembles that of Parallelocrinus mercenarius in being steep walled and
relatively wide. The forefacet is prominent and deep. Anal X barely
nicks the summit of the dorsal cup ; it is wedged in between the shoulders
of the C and D radials; below it is separated from the top of the CD basal

J. J. BURKE

NO. 19

by these posterior radials, which have a common lateral suture. This is
similar to the structure and disposition of the plates in the posterior
interradius of Parallelocrinus sturgeoni, except that in that species, anal
X and the CD basal are only slightly separated. Even the dorsal-cup
ornament of USNM SB982 somewhat resembles that of Parallelocrinus
mercenarius; the rugae, although coarse, are oriented much as are the
more slender ridges of the latter species.

However, the articular surfaces of the radials of USNM S3982 are
definitely indicative of Apographiocrinus; two of the interfacet “prongs”
discussed by Moore and Plummer (1940, p. 117) are preserved in entirety,
and it is evident that they are merely exaggerations of the same structures
found in American species of the genus. It follows that the remarkable
correspondence in surface structures of the dorsal cups of species of
Parallelocrinus and those of Wanner’s “Delocrinus” rugosus is quite
evidently attributable to homeomorphy, and is not indicative of any close
relationship between Wanner's species and representatives of Parallelo-
crinus.

As noted previously, Wanner first described this species (1916) as
Delocrinus rugosus, then (1942) attributed it to Ceriocrinus, and finally
(1949) redescribed it as Graphiocrinus? rugosus. I am herewith desig-
nating it Apographiocrinus rugosus (Wanner, 1916) comb. nov. This adds
a third species of Apographiocrinus to the Timor fauna; Moore and
Plummer (1940) also placed Graphiocrinus quinquelobus Wanner and
Delocrinus pumilus Wanner in synonomy under Apographiocrinus.

Moore and Plummer (1940) having relegated Poteriocrinus rugosus
Shumard (1858) to synonomy under Delocrinus, argued (1940, footnote
p. 272) that “D rugosus Wanner, 1916 is a homonym of D. rugosus
(Shumard) 1858 . . Inasmuch as Shumard’s type was never illustrated
and is no longer in existence, there is no way of determining what its
generic affiliation may have been, and in any case there would now seem
to be no grounds for applying a new trivial name to Apographiocrinus
rugosus.

1974

VIRGILIAN PARALLELOCRINUS

REFERENCES CITED

Burke, J, J., 1971, Parallelocrinus (Crinoidea, Inadunata) in the Ames Limestone,
Pennsylvanian, of Ohio: Ohio Jour. Sei., v. 71, no. 4, p. 198-201.

Knapp, W. D., 1969, Declinida, a new order of Late Paleozoic inadunate crinoids: Jour.
Paleontology, v. 43, no. 2, p. 340-391.

Moore, R. C. and Plummer, F. B., 1940, Crinoids from the Upper Carboniferous and
Permian strata in Texas: Univ. Texas Pub. 3945, p. 1-468.

Shumard, B. F. and Swallow, G. E., 1858, Descriptions of new fossils from the Coal
Measures of Missouri and Kansas: Acad. Sci. St. Louis Trans., v. 1, p. 198-227.

Wanner, J., 1916, Die permischen Echinodermen von Timor, Teil 1: Paleontologie von
Timor, Lief 6, Teil 11, p. 1-329.

1924, Die Permischen Krinoiden von Timor: 2e Nederlandsche Timor-

Expeditie 1916, II p. 1-348.

1949, Neue Beitrage zur Kentnis der permischen Echinodermen von

Timor, XVI. Poteriocrinidae, Teil 4: Paleontographica Supp., bd, 4, p. 1-56.

MANUSCRIPT SUBMITTED NOVEMBER 4, 1974

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO FEBRUARY 27, 1975 NUMBER 20

A NEW PERMIAN CIBOLOCRINUS FROM BOLIVIA

J. J. Burke

Senior Scientist, Cleveland Museum of Natural History ;

Research Associate, West Virginia Geological Survey

ABSTRACT

A flexible crinoid, Cibolocrinus patriciae sp. nov. from the Permian
Copaeabana Group of Bolivia is described. The holotype specimen was found
at the original D’Orbigny collecting site near Yaurichambi, 10 km east of
Lake Titicaca. The species is distinguished principally by the upflaring
infrabasal plates, which are slightly convex, show impressed sutures,
and are visible in lateral view of the dorsal cup. An inadunate crinoid
from the same locality, previously described as Delocrinus titicara by
Strimple and Moore, is here designated Endelocrinus titicara (Strimple
and Moore, 1971) comb. nov.

INTRODUCTION

In 1969 Drs. Frank Stehli and James Helwig of Case Western
Reserve University conducted a field study of Carboniferous and
Permian strata of Bolivia. Mrs. Patricia Helwig accompanied the
party as a representative of the Cleveland Museum of Natural History.
The party visited the locality in the vicinity of Yaurichambi, 10 km
east of Lake Titicaca, where D’Orbigny (1842) made the initial collec-
tion of Permian fossils from Bolivia. At that place Patricia Helwig
collected fossil invertebrates from the Copaeabana Group (Wolf-
campian) and found a specimen of the flexible crinoid Cibolocrinus
that proves to be a new species, which is described in the following
pages. A paper by James Helwig (1972, map, fig. 1; Yaurichambi
section, fig. 2) shows the location of Yaurichambi and a stratigraphic
section taken at the collecting site.

Previously (1971) Strimple and Moore, in the first formal descrip-
tion of a crinoid from the Bolivian Permian, proposed the name

J. J. BURKE

NO. 20

Delocrinus titicara for an inadunate crinoid that was also derived
from the Copacabana Group at the Yaurichambi locality. However,
their holotype specimen shows pits at the corners of the dorsal cup
plates and the arms attain normal biseriality well above the summits
of the primibrachs. These are diagnostic features of the genus Endel-
ocrinus Moore and Plummer, 1940. In consequence I am herewith
designating the taxon Endelocrinus titicara (Strimple and Moore,
1971) comb. nov. Further discussion of this species is reserved for
a future publication.

Other references to Permian crinoids of Bolivia are few. Ahlfeld
and Branisa (1960, p. 104) listed Icthyocrinidae indet., Lecythiocrinus
cf olivaeformis* and Delocrinus sp. from the Cocacabana Group at
Lake Titicaca and Yaurichambi, and from the same beds at Zudanez,
Delocrinus sp. and Aulocrinus? sp. Of the latter, which is figured
(ibid. pi. 7, fig. 20) Webster (1970) notes “(probably not an Aulo-
crinus)”; judging from the illustration, I take the specimen to be a
cromyocrinid. In another publication Branisa (1965, pi. 54) illustrated
additional crinoid material from the Copacabana, including Lecythio-
crinus cf. olliculaeformis from Zudanez (fig. 19), part of a biserial
arm from Yaurichambi (fig. 43), and portions of crinoid stems (figs.
1-27), from Zudanez, Apillipampa, Colquencha, and Yaurichambi.

ACKNOWLEDGMENTS

For the privilege of studying comparative material in their care,
I am indebted to Drs. Porter M. Kier of the National Museum of
Natural History, Stig Bergstrom of the Orton Museum, Ohio State
University, and Eugene S. Richardson, Jr. of the Field Museum. Dr.
G. D. Webster of Washington State University furnished bibliographic
data and Dr. James Helwig of Case Western Reserve University
loaned me literature pertinent to this study. I thank my wife Emily
for preparing the illustrations, and Bruce Frumker, Staff Photogra-
pher, for making the photographs from which the illustrations were
made.

Evidently a nomen nudum

1975

BOLIVIAN CIBOLOCRINUS

SYSTEMATIC PALEONTOLOGY

CLASS CRINOIDEA Miller, 1821
Family MESPILOCRINIDAE Jaekel, 1918
Genus CIBOLOCRINUS Weller, 1909
Cibolocrinus patriciae* sp. nov.

Figs. 1-5

Diagnosis : Walls of dorsal cup outflaring from base, but more erect
in region of radial circlet. Infrabasals visible in lateral view upflaring,
and convex, with sutures between them impressed. Basals elongate,
with sharply angular tips and straight distal borders.

Holotype: CMNH 3801, a mashed dorsal cup with complete B and
C and damaged A and D IBrn ; also damaged A and B IBm.

Repository : Cleveland Museum of Natural History, Cleveland, Ohio.

Occurrence : Copacabana Group, Lower Permian (Wolfcampian) .

Locality: Hogback called “Cerro Vacha Kahtawi” 3 km WNW of
village of Yaurichambi and 10 km east of Lake Titicaca, just south
of small church school (lat 40°17'56" S., long 68°29'10" W.) northern

Bolivia.

Description: The dorsal cup is mashed and it is difficult to get a
clear concept of its original proportions. The portion in the vicinity of
the E ray seems to be the least distorted. Apparently the cup was
low bowl shaped, and probably a little less than twice as wide as
high. In lateral view the walls flare outward from the base, becoming
steeper and more rounded a little above midheight. The posterior
slope was evidently gentler than the anterior. In dorsal or ventral
view the cup was probably pentagonal in outline.

The three infrabasals are clearly upflaring in lateral view of the
cup. They are slightly but definitely convex, most so longitudinally,
and a little indented along their common sutures. The A-ray plate
is the smallest of the three. Part of the stem, showing a round lumen,
is preserved. Its impression occupies nearly 3/5 of the diameter of
the infrabasal circlet.

The five basals are slightly convex, with short interbasal sutures.

*The species is named for Patricia Helwig.

J. J. BURKE

NO. 20

4 5

Figures 1-5. Cibolocrinus patriciae sp. nov. Holotype, CMNH 3801 from the Copa-
cabana Group, Lower Permian, near Yaurichambi, Bolivia. Fig. 1, dorsal view;
fig. 2, posterior view; fig. 3, ventral view; fig. 4, C-ray view (primanal at left) ;
fig. 5, EA-interray view. X2

Their slopes are moderate. Except for the CD plate, each is hexagonal
and extended distally, having a sharp tip. Their proximal angles are
quite broad; that of the CD basal is least so. All of these plates are
wider than long, including the CD basal, but the latter is the largest,
with greatest distal extent. The CD basal is also heptagonal, being
truncated distally where it meets the primanal plate.

The A, B, and E radials are about 3/5, but the C and D radials
are only slightly more than 1/2, wider than long. The interradial
sutures are nearly as long as the interbasal sutures, but the radial-
primanal sutures are half again as long. The radials are slightly convex
longitudinally and transversely, but are more erect than the basals,
and the cup walls are perceptibly steeper in the region of these plates.
Interradial notches are present at the summits of the radials. The
articular surface of the E radial is exposed, revealing the typical

1975

BOLIVIAN CIBOLOCRINUS

Cibolocrinus structure, with fan-shaped depressions at the lateral
extremities and the surface contracted medially opposite the slitlike
ligament pit.

The primanal is a large plate, nearly as long as the CD basal and
a little more than half as wide. More than half its height is below
the summits of the posterior radials. The plate is heptagonal; it has
been damaged distally, but still shows four facets above the radials.
Laterally it is notched in areas adjacent to the summit regions of
the C and D first primibrachs.

Only portions of the A and D first primibachs are preserved, and
they show no details of the articular surfaces. The shattered E first
primibrach was removed to expose the articular surface of the under-
lying radial. The C first primibrach is entire, and the adjacent B
plate is essentially complete also. At midwidth these plates are about
as long as the interradial sutures, but they are longer laterally, where
they project downward along the interradial notches, although there
is some distal increase also along the sides of the C plate, and the
right side of the D plate shows marked elevation. The distal articular
surface of the C first primibrach shows a slitlike external ligament
pit, similar to that of the radial, but the lateral depressions are deeper
and more basin-like than those of the radial. These plates bear notches
that extend below their lateral summits (similar notches are found
on the primanal, as noted above). Lateral extension of the second
primibrachs extended downward along these notches.

Part of the articular surface of the left side of the B second
primibrach is preserved; the lateral depression is similar to that of
the first primibrach, although shallower. The remainder of this second
primibrach is shattered and crushed. The adjacent A second primi-
brach is also too distorted by crushing to yield reliable data.

The cup surface appears finely granulose, but the ornament is not
clearly distinguishable.

Measurements : Linear measurements, in mm, taken on holotype,
CMNH 3801: Dorsal cup height, 8.5 (est.), width 16.0 (est.) ; H/W
ratio, ca. 0.53; stem impression width, 3.9; infrabasal circlet width,
6.8; basal (EA) length, 5.4, width 5.8; radial (E) length, 4.9, width,
8.6; length suture between basals, 2.2; length suture between radials,

J. J. BURKE

NO. 20

2.0; primanal height, 5.2 (appr.), width, 4.8.

Discussion : Cibolocrinus patriciae resembles certain Pennsylvanian
representatives of the genus to the extent that its infrabasal plates
are visible in lateral view of the dorsal cup. In the Morrowan Cibolo-
crinus circulus the lateral exposure of these plates is much less than in
C. patriciae , and the cup outline has been described as “. . . subhemis-
pherical when viewed from the side” (Moore and Strimple, 1973,
p. 33). Although in its proximal half the cup of C. patriciae flares
outward from the base, it is steeper walled and slightly rounded
distally, thus differing from C. circulus and from the Missourian
species C. erectus Strimple, 1951a and C. conicus Strimple, 1951b.
In both of the latter forms the cup walls flare outward strongly from
base to summit, and the dorsal cup has less height than that of C.
patriciae , which in turn is surpassed in height by the cup of C. circulus.
Cibolocrinus patriciae shows greater exposure of the infrabasal plates
in lateral view than we find in Strimple’s taxa, along with somewhat
greater width of its infrabasal circlet and slightly greater diameter
of the stem impression. Among all four species, C. patriciae is unique
in displaying convex infrabasals with impressed sutures. The elongate
basals of C. patriciae compare to some extent with those of C. circulus ,
but in C. circulus the distal borders of the basals are curved, rather
than straight. The interradial notches noted by Strimple in C. erectus
are present in C. patriciae; they are also apparently present in C.
conicus , but seem to be absent in C. circulus, which would be consistent
with Strimple’s (1951a) observation of their absence in Morrowan
forms.

In competition with species of Cibolocrinus having flat-based dorsal
cups, these forms with upflared infrabasals may have had the ad-
vantage of offering less resistance to strong water currents, and thus
might have survived under conditions where species with the more
specialized type of cup could not. There is insufficient evidence to
support the argument that species of Cibolocrinus showing upflared
infrabasals are examples of regressive evolution. Already present in
Lower Pennsylvanian time, species with infrabasals of this type prob-
ably occupied an environmental niche where this feature was an ad-
vantage, and they persisted into the Lower Permian with little further
specialization.

1975

BOLIVIAN CIBOLOCRINUS

REFERENCES CITED

Ahlfeld, F. and Branisa, L., 1960, Geologia de Bolivia: Boliviano Inst. Petroleo,
p. 1-245.

Branisa, L., 1965, Las fosiles guyas de Bolivia: Bolivia Serv. Geol., p. 1-282.

D’Orbigny, A., 1842, Voyage dans l’Amerique meridionale de 1826-1833: v. 3, pt. 4,
Paleontologie.

Helwig, J., 1972, Stratigraphy, sedimentation, paleogeography, and paleoclimates
of Carboniferous (“Gondwana”) and Permian of Bolivia: Am. Assoc. Petrol-
eum Geologists Bull. v. 56, no. 6, p. 1008-1033.

Moore, R. C. and Strimple, H. L., 1973, Lower Pennsylvanian (Morrowan) crinoids
from Arkansas, Oklahoma, and Texas: Univ. Kansas Paleont. Contrib. art.
60, p. 1-84.

Strimple, H. L., 1951a, Pennsylvanian crinoids from Lake Bridgeport, Texas: Jour.
Paleontology, v. 25, no. 2, p. 200-207.

, 1951b, Some new species of Carboniferous crinoids: Bull. Am.

Paleontology, v. 33, no. 33, p. 1-41.

, and Moore, R. C., 1971, A crinoid crown from D’Orbigny’s famous
fossil locality at Yaurichampi, Bolivia: Univ. Kansas Paleont. Contrib. Paper
56, p. 33-35.

Webster, G. D., 1973, Bibliography and index of Paleozoic crinoids 1942-1968: Geol.
Soc. America Mem. 137, p. 1-341.

MANUSCRIPT SUBMITTED JANUARY 17, 1975

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO MARCH 30, 1976 NUMBER 21

SOME MISSISSIPPIAN CERAMICS FROM ARKANSAS,
^ CLEVELAND MUSEUM OF NATURAL HISTORY

\ DAVID S. BROSE

TJie Cleveland Museum of Natural History

$ $ 18/ O

with

^^ALAsIS OF THE CARBONIZED PLANT REMAINS
RICHARD I. FORD

Ethnobotanical Laboratory, University of Michigan

ABSTRACT

Reanalysis of Cleveland Museum of Natural History collections has re-
vealed evidence for an early agricultural occupation of an archaeological
site in northeastern Arkansas. This Mississippian culture mound, dated to
A.D. 1050, showed distinctive ceramics with ethnobotanical material sug-
gesting a revision of earlier theories of prehistoric diffusion in the eastern
United States.

During December of 1973, as part of my duties as Curator of Ar-
chaeology for The Cleveland Museum of Natural History, I assisted
in the transfer of a considerable amount of archaeological and ethno-
logical material from a soon-to-be demolished storage building into
the Museum's permanent research collections. Much of this material
had been donated by private parties during the nineteen thirties, and
had been accessioned prior to 1936. It was never really looked at by
any professional archaeologist until the spring of 1974.

In one large cardboard box, taped shut, and located in a deep
storage midden in a back closet, were two shell-tempered, looped-
handled bowls: a large intact bowl (fig. 1) with an average lip thick-
ness of about 4.7 mm (s2 =1.9 mm) and a smaller, similar partially

DAVID S. BROSE

NO. 21

reconstructed bowl (fig. 2) with an average lip thickness of 3.9 mm
(s2 = 1.8 mm). The smaller vessel has traces of red fugitive slip. The
vessels both bore the C.M.N.H. catalog number 9674. References to
the accession catalog indicated only that the two vessels were from
Cross County, Arkansas, and had been donated together to The Cleve-
land Museum of Natural History by a Mr. H.P. Shaw some time
between 1927 and 1932. Interesting, but not really significant.

Working with a graduate assistant during February of 1974, I
continued sorting through the older materials. Within a large card-
board box marked as coming from a mound on Weeden Island in
Tampa Bay, and containing a series of shell (Busy con?) implements,
fragments of human skeletal material, and a few Franklin Plain and
Carabelle Incised sherds (Willey 1949:v.ll3, p.479) was a small metal
box about 9x4x2 inches rusted shut. Inside this small metal cigar-
ette or cigar box were three ceramic sherds, a series of botanical
materials including nut shell, maize cob, various seeds which I could
not recognize even to family, and several pieces of wood fiber about
3 inches long and approximately 3/4 inches in diameter. Several of
the latter appeared to have been varnished. Included within the metal
box was a worn, folded, and somewhat rust-stained note, which reads :

9674

— Two pots and seeds
acca; H.P. Shaw

collected by — Jones [1 — 7] below ground at More’s Md. about four miles
S. of Parkman in Cross Co., Arkansas on E. Side of the river.

These pots were found together and were dug out w — the seeds and
shavings in the bigger pot made of mixed clay and [clam] shells. The small
clay shards were found under these pots and the big broken shard with
lines scratched was in the dirt above.

While there is a village named Parkman in Ohio (about 25 miles
ESE of Cleveland) there is only a Parkin in Cross County, Arkansas.
While there are several rivers in Cross County, there is only one
major one: the St. Francis. While C.B. Moore operated upon local
mounds, the only mound he reported which was about four miles
south of Parkin on the east side of the St. Francis was the Rose
Mound (Moore, 1910 :276-303) . The note in The Archaeological Survey
of the Lower Mississippi Valley (Phillips, Ford and Griffin, 1951 :278)
that Rose had been severely potted since Moore’s day lent some sup-
port to the notion that The Cleveland Museum of Natural History had
obtained material probably assignable to the Rose Mound.

1976

SOME MISSISSIPPIAN CERAMICS FROM ARKANSAS

Figure 1. Large shell-tempered vessel probably of Mississippian Plain type, Mound
Field variety, within which the carbonized botanical remains were found. CMNH
Accession No. 9674.

Figure 2. Small red-slipped vessel (partially reconstructed) of Old Town Red type,
associated with the large bowl within the mound. CMNH Accession No. 9674.

DAVID S. BROSE

NO. 21

The two undecorated small rimsherds in the metal box were clay
or grog tempered and can probably be considered representative of
some varient of Baytown Plain (Phillips, Ford and Griffin, 1951 :270) .
Both of these rimsherds have a relatively thin lip thickness (x = 9.0
mm) and display a hard compact paste. They seem equivalent to
what Phillips (1970:57) has called the West Lake variety of Bay-
town Plain. The larger sherd was coarsely shell-tempered as were
the two complete vessels. It is not a complete rim but probably
comes from just below the lip — it is broken along a medium wide
incised line. Below this break are two parallel horizontal (?)
incised lines about 1.5 mm wide, 0.8 mm deep, and 8.8 mm and 9.2 mm
apart. Below these the sherd curves out toward the shoulder and
is roughly broken about 17 mm below the lowest incised line. This
neck sherd has a very Coles Creek look although if it truly overlies
the Neeleys Ferry Plain vessels that would suggest a stratigraphic
problem. A more reasonable attribution may be suggested, however,
as Coles Creek is not generally found as far north as Cross County,
Arkansas (J.B. Griffin: personal communication). This sherd can
easily be encompassed within what Phillips has characterized as
the Mound Place variety of the type Mound Place Incised (Phillips,
1970: 185, fig. 59c).

There is little question that the two complete shell-tempered ves-
sels with loop handles can be considered some variety of Neeleys
Ferry Plain (Phillips, Ford and Griffin, 1951 :287). These Mississip-
pian Plain vessels do not appear quite thin enough to meet Williams
(1954) or Phillips (1970:132) criterion for Mississippian Plain,
var. Coker. They both might fall within what Phillips would consider
Mississippian Plain, var. Mound Field although adequate quantified
data for certain attribution are not presented in that monumental
report (Phillips, 1970:132-3). The smaller vessel with traces of red
slipping should thus represent an example of Phillips’ Old Town
Red, var. Old Town (Phillips, 1970:145).

On both of these vessels, the paired loop handles are undecorated
and are diametrically opposed on the vessels. The lower loop end is
riveted through the vessel wall and luted on the interior. The upper
handle end is attached by luting on the rim at the lip. On both vessels
the loop handles rise slightly higher than the rim itself although erosion
and breakage along much of the vessel rim make any more quantified
statement spurious. These loop handles are virtually identical to one
illustrated by Phillips, Ford and Griffin (1951 :228) which came from

1976

SOME MISSISSIPPI AN CERAMICS FROM ARKANSAS

the lower levels of the Rose Mound. They are similar to the loop-
handled Neeleys Ferry Plain vessels reported from the early Missi-
sippian Banks Mound 3 dated A.D. 1075 ± 75. Similar vessels formed
part of a richer ceramic assemblage with dates of A.D. 710 ± 150,
A.D. 930 ± 150, and A.D. 1100 ± 110 in Crittenden County, Arkansas
(Perino, 1967:69).

The wood shavings from the large pot were identified by Dr. O.
Elzam of the Case Western Reserve University Biology Department
as most likely some non-twig portion of Carya, possibly ovata. I
shaved off all the varnished surfaces and ended up with a handful
of match-stick-sized fragments. These were then washed with a variety
of solvents, washed with acid and base, and rewashed several weeks
with distilled inert water by Dr. A. Sumodi of the C.W.R.U. Radio-
carbon Lab. These wood fragments were then combusted and the car-
bon content converted to benzene, and the C14 content counted. The re-
sultant date (CWRU-172) is A.D. 1050 ± 65 or BP 900. Until this
date was returned it was uncertain whether the wood represented
some rather recent contamination, or whether the association re-
ported on the C.M.N.H. accession note was valid. It now appears most
reasonable to assume that the date on the wood shavings also refers
to the ceramics within which Mr. Jones claimed to have found them.

This suggests that the other ethnobotanical materials can also be
referred to an early Mississippian period of A.D. 900-1100, and that
there might be some value in obtaining the evaluation of a competent
ethnobotanist concerning their specific identification and signifi-
cance. Dr. Richard I. Ford of the University of Michigan Museum of
Anthropology Ethnobotanical Laboratory was kind enough to consent
to analyze these materials. His description and discussion are ap-
pended to this report.

Before beginning any extended discussion on the archaeological
significance of these rediscovered ceramics and their contents, it is
necessary to inquire more closely as to their context. Dr. J.B. Griffin
(personal communication 21 May 1974) has pointed out that there is

. . . very little control over the time at which the seeds and shavings were
utilized or how they got into the vessel. While it is not unreasonable
to suggest that the site could be the Rose Mound, I would say that if the
gentleman went by road from Parkin to the site that it would not be much
more than two miles but if he went by water it would be probably closer to
four miles. Down river from the Parkin site there is another site (12-N-6)
which we identified as the Westmorland site. I don’t see how you could be
sure what site these materials come from.

DAVID S. BROSE

NO. 21

In addition to Westmorland, the Lower Mississippi Valley Survey
also identified the Welshans Place site (12-N-5) also on the east side
and about two miles further down river from the Westmorland site.
The Westmorland site however yielded only scanty amounts of middle
Baytown ceramics (Phillips, Ford and Griffin, 1951: fig. 20) while
the Welshans Place site, aside from lying beyond the 4-mile limit
was apparently neither visited by Moore (Moore 1910), nor con-
tained any ceramics which predated the early Mississippian horizon
(Phillips, Ford and Griffin, 1951: fig 20). The Rose Mound site not
only produced materials from late Baytown through late Mississip-
pian, but gave evidence of an early Mississippian component with
shell-tempered plain ceramics which differed somewhat from typical
Neeleys Ferry Plain (Phillips, Ford and Griffin, 1951 : 105-110, 287-
88; Phillips, 1970:938). As Phillips (1970:246) later noted,

In a large surface collection from the Rose Mound (12-N-3) made in 1940,
there were only five clay-tempered sherds out of a total of 1416 (Phillips,
Ford and Griffin 1951: fig. 21, sample no 12-N-3/A). A later test excava-
tion in 1947 revealed a 100% pure ‘clay-tempered’ component with sherd
yield per level as high as those in the overlying shell-tempered levels and
plow-zone {ibid. fig. 53)

Given available information, it seems quite reasonable to accept
the information in the C.M.N.H accession note accompanying the
ethnobotanical material and the three sherds in the metal box. The
stratigraphic sequence thus indicated would suggest a late Baytown
component overlain by an early Mississippian occupation with plain
loop-handled vessels dated at A.D. 1050 ± 110, below incised Missis-
sippian Ceramics. This sequence certainly conforms to the earlier
ceramic chronology developed by Phillips, Ford and Griffin (1951)
and should be capable of being placed within the more detailed frame-
work recently proposed by Phillips (1970). If there are indeed se-
quent occupations, the only ceramics which can be assigned to a
particular phase would be the Mississippian Plain (var. Mound Field,
and var. Old Town) bowls. These could easily represent a portion of
Phillips’ revised Parkin Complex, the absence of Parkin Punctate
and Barton Incised being the result of sampling error. With the asso-
ciation of the radiocarbon of A.D. 1050 ± 110 these shell-tempered ves-
sels would fall into the Cherry Valley Phase (Phillips 1970, 930-33).
While this date may seem somewhat early for Mississippian Plain
ceramics in the St. Francis area it agrees with Perino’s dates at

1976

SOME MISSISSIPPIAN CERAMICS FROM ARKANSAS

Cherry Valley and Banks (Perino, 1967:67). Phillips, discussing
this Cherry Valley phase, has predicted that as new data continue
to accumulate, such dates will seem less a special case (1970:930).

It is not, unfortunately, certain from the Cleveland Museum’s note
of Mr. Shaw’s recollection of Mr. Jones’ report, that three distinct
components are indeed present. Phillips (1970:914), following earlier
suggestions by Williams for southeast Missouri, recognizes a late pre-
Mississippian Black Bluff horizon marked by a plain “clay-tempered”
ware with a harder, more compact paste than is normal for Baytown
Plain in this region. Indeed, Phillips suggests two distinct late Bay-
town phases flanking the lower St. Francis area, both characterized
by a preponderance of Baytown Plain with Coles Creek “earmarks.”
Phillips (1970: 916-17) cites Hester Davis’s 1967 characterization
of the Toltec phase occupying the area between the St. Francis and the
Mississippi as containing components with Mississippian settlement
patterns associated with clay-tempered Baytowm and shell-tempered
Mississippian Plain ceramics.

If the materials herein described from the Cleveland Museum
which are tentatively assigned to the Rose Mound site represent the
ceramic assemblage from a single occupation, their relative strati-
graphic occurrence within the component may be considerd unim-
portant. In that case they should probably be assigned to this Toltec
phase. In this event the radiocarbon date of A.D. 1050 ± 110 should
represent a reasonable (if not conservative) temporal position for
the early Mississippian transition in the St. Francis basin. Phillips
has pointed out (1970 :960) that the present state of calendrical
dating in the lower Mississippi valley need not be taken too seriously
in its present state of development. He postulates the first appearance
of Mississippian ceramics in the Yazoo region (at about A.D. 1000)
in his Crippen Point Phase (1970 :fig 450A). This suggestion has re-
ceived some support from the recent excavations at the Winterville
site in the Yazoo basin where Brain has recovered Mississippian
ceramics late in a Crippen Point phase dated A.D. 1050-1200 (Brain,
1970: 276,304). While one might presume that such ceramics would
be earlier upstream, it seems preferable at this point to accept Phillips’
(1970:930) cautious prediction that,

As other early Mississippi data accumulate, as they are bound to do, . . .
it may turn out that Mississippian Culture was developing over a broad
front with interconnections that were not exclusively in a one-way, north-
south direction.

RICHARD I. FORD

NO. 21

CARBONIZED PLANT REMAINS FROM
CROSS COUNTY, ARKANSAS

Museums are fascinating places. Their lure captivates the dreams
of children and stimulates the imagination of adults. But to the sci-
entist they are the foster homes of displaced discoveries. It is only in
their confines that the archaeologist can reexamine the excavated arti-
facts of a predecessor and more often than not “excavate” once again
long forgotten and unreported relics of man’s prehistoric past. Such
was the fortuitous discovery by Dr. David S. Brose of The Cleveland
Museum of Natural History when he salvaged from an old museum
warehouse an unpublished Neeleys Ferry Plain vessel containing
several carbonized plant fragments. They were submitted to the Eth-
nobotanical Laboratory for identification.

While these charred remains are an unrepresentative sample of
the plants used by early Mississippian peoples, nevertheless they do
contribute to an accumulation of similar data that together are en-
lightening our knowledge of subsistence patterns 1000 years ago.

In his report Brose documents the inadequate records accompanying
these plant parts. It is evident that they do come from a mound lo-
cated south of Parkin in Cross County, Arkansas. Whether the actual
site situated on the east side of the St. Francis River was the famous
Rose Mound remains uncertain.

The remains consist of two butternut shells, one pawpaw seed,
three bean cotyledons, three kernels of corn, and three pieces of cobs.
Why they were in this pot is not clear ; perhaps their excellent state
of preservation attracted the attention of the excavator and he placed
them there out of harm’s way.

The two nut shells, weighing 3.3 g, are butternuts, Juglans cinerea
L. From a phytogeographical point of view this is an interesting
identification. Butternuts are common throughout the midwest, in-
cluding southeastern Missouri, but are rare in Arkansas. Aside from
two or three disjunct populations their main occurrence in this state
is along the riverbottom and alluvial benches of the St. Francis
River. Here butternuts are a common element in the bottomland
hardwood forests. This archaeological find supports the hypothesis
that the modern pattern of distribution existed in the early part of
the tenth century A.D. as well.

The nut, an ovoid fruit 50 to 80 mm long, ripens in September and

1976

SOME MISSISSIPPIAN CERAMICS FROM ARKANSAS

October. It contains a good vegetable protein and is an excellent
source of protein but does not store well for long periods of time with-
out becoming rancid. As a consequence it may have been necessary
to process these tasty nuts soon after they ripened by removing the
thin hull and extracting the meat from the cracked shell. The oil could
be separated from the meat by boiling. Butternuts undoubtedly con-
tributed only a minor portion of food to the prehistoric diet. The
trees are rarely found in groves^ most often singly, and each mature
tree produces a good crop of nuts every second or third year at most.
A thrifty tree yields only 1/4 to 1 bushel of nuts (Fowells 1965:208-
10). Unless simultaneous access to many trees is available, a family,
much less a village, would benefit little if it attempted to rely on
butternuts.

The pawpaw, Asimina triloba L. is another component of bottom-
land vegetation although it will also occur as a mature fruit-bearing
tree in later stages of old field succession. The fruits ripen in the
early fall and are a delightful treat. The evidence for their consump-
tion at this site is represented by only one seed measuring 22 mm in
length and 11 mm in width. It is well within the range of modern
examples. Pawpaw seeds have been identified from a number of
Mississippian sites, including the neighboring Parkin Site (Cutler and
Blake 1973:9-10).

Beans and corn represent the only evidence of cultivated plants.
The three beans, Phaseolus vulgaris L., are all broken. The most
complete measures 10 mm by 6.2 mm. The remainder have widths of
6.7 mm and 5.6 mm; their lengths are indeterminable.

Beans are not found in every excavation of Mississippian sites,
and actually they may not have been grown by every community.
They were a late introduction into the eastern United States, perhaps
reaching this area as late as A.D. 900. Long recognized for their pro-
tein content and complementary relationship with corn’s amino
acids, they were not essential to the Mississippian economy every-
where. Other native foods such as acorn (Wilma Wetter strom, per-
sonal communication) have a lysine value that will bolster the
nutritional value of corn as well.

The fragmentary condition and small sample of corn permits only
a superficial discussion. We know that ethnographic cultures living
in the Southeast raised a number of varieties of corn, but to date the
phenotypic traits used to define these types have not been distin-

RICHARD I. FORD

NO. 21

guished for analyzing carbonized archaeological remains. There-
fore our categories are artifacts of limited analytical procedures.

The three kernels of corn are similar in size and shape. Although
two kernels are lacking their “germ,” all have a crescent outline and
are wider than they are high (width x height: 11 x 7.9; 7.9 x 6.2;
9.2 x 7.3 mm).

Of the three cobs, one is a segment of connected cupules 15.6 mm
long. There are 2.4 cupules per 10 mm of cob length on this and the
other two cobs. The broken basal portion has 10 kernel rows and an
elliptical cross-section. The kernel rows are even and the butt is un-
flared. Its median cupule width is 10.1 mm. The tip portion (upper
end) has 8 kernel rows and a circular cross-section. The kernel rows
are regular and are not strongly paired. Kernel facets are present to
the very top of the cob. The median cupule width is 7.8 mm.

Assigning a racial type to this corn can be done only in an evolu-
tionary framework. The low row number (8 and 10 rows), ratio of
kernel width to height, and the wide cupules recommend Eastern
Complex corn ; yet the lack of a flared butt and the broadly separated
cupules do not conform with this designation. The reason, of course,
is that the classic features of this variety evolved in the east and
these samples represent an earlier stage of its development. This corn
is not unlike pre-A.D.-1300 maize previously identified from north-
eastern Arkansas and southeastern Missouri (Cutler and Blake 1973:
9-10, 41).

REFERENCES CITED

Brain, Jeffrey P., 1970, Winterville, a case study of prehistoric culture contact in the
lower Mississippi valley: Yale University Ph.D. Dissertation in Anthropology,
372 p.

Cutler, H.C. and Blake, L.W., 1973, Plants from archaeological sites east of the
Rockies: St. Louis: Missouri Botanical Gardens, 174 p.

Fowells, H.A., 1965, “Butternut ( Juglans cinerea L.),” Silvics of forest trees of the
United States: U.S. Dept, of Agriculture Handbook no. 271, p. 208-210

Moore, Clarence B., 1910, Antiquities of the St. Francis, White and Black rivers,
Arkansas: Academy of Natural Sciences of Philadelphia, Journal, Second
Series, v. 14, p. 254-364

Perino, Gregory, 1967, The Cherry Valley Mounds, and Banks Mound 3: Central
States Archaeological Societies, Inc., memoir no. 1, 67 p., St. Louis

1976

SOME MISSISSIPPIAN CERAMICS FROM ARKANSAS

Phillips, Philip, 1970, Archaeological survey in the lower Yazoo basin: Papers of
the Peabody Museum of American Archaeology and Ethnology, v. 60, 965 p. :
Harvard University

Phillips, Philip, Ford, James A., and Griffin, James B., 1951, Archaeological survey
in the lower Mississippi valley 1940-1947 : Papers of the Peabody Museum of
American Archaeology and Ethnology, v. 25, 422 p. : Harvard University.

Willey, Gordon, 1949, Archaeology of the Florida Gulf coast: Smithsonian Miscel-
laneous Collections, v. 113, 479 p.

Williams, Stephen, 1954, An archaeological study of the Mississippian culture in
southeastern Missouri: Yale University Ph.D. Dissertation in Anthropology,
297 p.

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO MAY 14, 1976 NUMBER 22

THE FIRST AMERICANS:

A STUDY OF THE ORIGIN, EVOLUTION
AND VARIATION OF THE AMERICAN INDIANS

Mahmoud Y. El -Najjar
Department of Anthropology
Case Western Reserve University

ABSTRACT

A well -documented date of entry for the first New World inhabitants has not been es-
tablished. Evidence shows that man’s initial crossing into the New World was by way of
the Bering Land Bridge at least 20,000 to 30,000 years ago. Earlier hypotheses attributing
biological differences among native Americans to successive waves of migration are not
supported by the present findings. There is no evidence that Australoid, Melanesian, Cau-
casoid or Negroid admixtures have contributed to the formation of the American Indian
physical variety. Cultural and/or linguistic similarities between the American Indian and
Asiatic Mongoloids have not been fully demonstrated. Studies of living and skeletal mor-
phology have been more useful in reconstructing the past biological affinity of the New
World natives. Differences exhibited among modern American Indian groups suggest the
possibility that differences in the original Mongoloid stock from which they came have
been retained. Since inhabiting the New World, physiological adaptation occurred under
a wide range of environmental conditions including subarctic, desert and tropical rain
forests, and therefore many of the biological differences can be due to environmental ex-
tremes.

INTRODUCTION

The origin and evolutionary history of the American Indians is still a contro-
versial subject argued by archaeologists, prehistorians, geologists, and evolution-
ary biologists. Despite years of search in the Old and New Worlds for evidence of
the origin of the American Indians, a well-documented synthesis of the available
information is yet to be offered.

In considering the first New World inhabitants a major problem is the lack
of well-dated, comparative osteological material from both Asia and the Ameri-
cas. When such material is recovered, analysis is limited by several factors. Few

MAHMOUD Y. EL NAJJAR

NO. 22

of the reported finds of Paleo-Indian skeletal remains are complete. Often, these
finds consist of fragmentary skeletal material that is difficult to reconstruct; so
cranial and post -cranial morphology is difficult to determine. Analysis has been
further handicapped by the failure of Western scientists to adequately integrate
(due to political and linguistic barriers) published data, particularly from Japan
and Russia.

It is the intention of this investigation to critically review the available evi-
dence of Paleo-Indian studies and to offer a model by which the present infor-
mation can be placed into a more useful perspective.

ACKNOWLEDGMENTS

I would like to thank Drs. Allen Young, Charles Callender and David Brose
for their valuable comments and suggestions; H. Donna Brown for her editorial
comments; and Mrs. Patricia Ann El-Najjar for editing and final typing of the
manuscript. Mrs. Joy A. O’Connell assisted in editing the original manuscript.
My thanks to Miss Deborah Vaiksnoras who made the map.

HISTORICAL REVIEW

Columbus and many of the early Spanish settlers viewed American Indians as
less than fully human since they were not mentioned in the Bible. With Las
Casa’s (1474-1566) appeal, the church gave Indians a human status and agreed
that they originated in the Old World. Another widely held theory was that the
American Indians were descendants of the “Ten Lost Tribes of Israel.” This was
first proposed by Amerigo Vespucci after his voyage of 1497. James Adair (Wil-
liams, 1930), one of the advocates of this theory, based his hypothesis on pho-
netics instead of structural similarities between the languages of the two groups.
Upon surveying the literature, no biological, linguistic or cultural evidence sup-
porting this theory was found. Jennings, who recently reviewed this theory, con-
cludes,

“Complete lack of tangible evidence — such as the wheel, Old World grains or domestic
animals — makes the theory untenable, to say nothing of the common sense problem of
how a group of herdsmen and gardeners with no recorded skills of seamanship could have
voyaged to the Americas all the way from the dry hills of Asia Minor.” (Jennings, 1968:
44-45).

Father Jose de Acosta in his Historia Natural y Moral de las Indians, first
published in Seville in 1590, considered the various theories relating to the
American Indian origin. While rejecting Atlantis, the Lost Continent of Mu,

1976

THE FIRST AMERICANS

and the Ten Lost Tribes of Israel as possible Indian homelands (Beals, 1957;
Wilmsen, 1965), he suggested the possibility of a land bridge or a narrow strait
in high northern latitudes, over which small groups of hunters crossed to the
New World. This appears to be the first time that an Asiatic origin was hypothe-
sized for the American Indian natives. In 1650 Thomas Gage also suggested a
Bering Strait route because the American Indians living near the strait were sim-
ilar to Asiatics in their customs and habits (Thompson, 1958). Meanwhile, Euro-
pean authors were advancing their own hypotheses with regard to the origin of
native American Indians. Hugo Grotius suggested Scandinavians, Ethiopians,
Chinese and Moluccans as the American Indian’s ancestors. Johannes de Laet
and George Horn hypothesized that Scythian, pre-Columbian Spaniards,
Moduc’s Welsh and Polynesians were the forerunners of the American Indians
(Winsor, 1889:369-370).

During the 18th century, even more invalid theories for the origin of the
American Indians were advanced. Cotton Mather in his Magnalia Christi Amer-
icana writes,

“Probably the devil decoyed these miserable savages hither, in hope that the gospel of the
Lord Jesus Christ would never come here to disturb or destroy his absolute empire over
them.” (Drake, 1880:24-25).

In 1784 Thomas Jefferson (1801) excavated a small site near his home and re-
covered a large amount of mastodon bones and teeth. This discovery developed
his interest in the American Indians. After further study of Indian languages, he
concluded that American Indians and northern Asiatics had a common linguis-
tic origin. Similar conclusions were reached by Father Ignaz Pfefferkorn, S.J.
(1794-95) who states,

“ ... in the northern regions America and Asia met or are separated by such a narrow
strait that people and animals might pass without difficulty from one continent to the
other. To me it is almost certain that the first inhabitants of America really came by way
of this strait . ” (T reutlein , 1949:161).

Lubbock’s Prehistoric Times was probably the first major work dealing with
the history of American Indians. Lubbock (1878) divided their history into
four stages: original barbarism, mounds, garden beds, and relapse into partial
barbarism. To Lubbock, man’s arrival in the New World came as a result of
slow population movement, with a few groups entering the New World at any
one time. In 1873, Abbott described what may have been a post- Archaic
campsite and used it as a basis for explaining the origin of American Indians.
After analyzing several hundred artifacts, Abbott suggested an early postgla-
cial date but later (1889:304) concluded, “We are pretty sure of twenty or
even thirty thousand years now.”

MAHMOUD Y. EL NAJJAR

NO. 22

Advocates of the multiple migration hypothesis predominated at the Inter-
national Congress of Americanists, held at Nancy, France in 1875. Winchell
(1880), for example, traced all of mankind from a submerged continent he
called Lemuria where he at times even identified the specific tribes who were
their descendants. Others gave more valid accounts of the American Indian ori-
gin. Dali (1877:93-98) favored a crossing on ice at the Bering Strait and Rau
(1822) suggested a land bridge over which man made his crossing to the New
World.

After the turn of the century, the fact was established that American In-
dians did not originate in the New World (Hrdlicka, 1920). For the first time,
the question of their origin and evolutionary history was discussed by students
with formal anthropological training. New skeletal material was recovered, an-
alyzed and reported. This increased the opportunity for comparative skeletal
studies and for communication between researchers interested in American
Indian studies. Methods, techniques, and professional treatment and preserva-
tion of human skeletal material and artifacts were improved. Systematic site
excavations were also employed and more knowledge of past human adapta-
tion became possible. Most important was the rise of anthropological theory
with a definite trend toward the integration of various anthropological subdis-
ciplines in Paleo-Indian studies.

During the first half of the twentieth century, polyracialists advanced
several theories to explain the origin and physical variability of the New World
natives. These theories were based on a typological approach developed in the
early days of physical anthropology using cranial measurements and indicial
resemblances. Taylor (1946), Gladwin (1947), Howells (1946), Hooton (1947),
Imbelloni (1943, 1958) and Rivet (1958) are among such authors.

Taylor (Birdsell, 1951) hypothesized Australoid, Mediterranean and Ne-
groid migrations. Imbelloni (1943) postulated seven distinct human groups en-
tering the New World in the following sequence: Tasmanians, Australians,
Melanesians, Proto-Indonesians, Mongoloids, Indonesians and Eskimos. In
1958 Imbelloni revised his earlier hypothesis to include four additional groups.
His study was based on stature, robusticity of bone structure, cephalic and
head height indices, nasal and facial indices, hair color and form, and skin
color. Both skeletal material and observations on living American Indians
were used. The eleven varieties were also assigned a chronological order of
entry into the New World. Hooton (1947) also attributed physical variability
among the New World natives to different migrations. According to Hooton,
Eskimos are the most Mongoloid in appearance with a smooth forehead,
marked epicanthic eye-folds, a low-rooted and saddled infantile nose and a
yellow skin. In North America, and presumably preceding the Eskimo as im-

1976

THE FIRST AMERICANS

migrants to the New World, American Indians were characterized by more re-
ceding brow ridges, boldly arched noses and coppery skins; in Central and
South America are those with wavy hair, very dark skins, and short straight
noses. These distributions suggest a series of separate migrations from Asia to
the New World, with the earlier waves of immigration being non-Mongoloid
(Hooton, 1947).

One of the more controversial arguments to explain the origin of the
American Indian natives is that proposed by Birdsell in 1951. Birdsell hy-
pothesized an eastern Asiatic population known as Archaic Caucasoid that was
composed of three groups (Murrayians, Carpentarians, and Amurians) during
the later part of the Pleistocene. Due to population pressure, Murrayians and
Carpenterians were pushed southward where their descendants are today’s
Australian aborigines. Coon, Garn and Birdsell (1950) suggest that only the
eastern branch of the Caucasoids (the Amurians) were represented in north-
eastern Asia. They further indicate that late in the fourth glacial period in re-
sponse to stringent environmental conditions, the Mongoloid people evolved
from an Archaic Caucasoid stock and spread rapidly. According to Birdsell,
the American Indians are hybrids produced by an admixture of Amurian and
Mongoloid varieties in which the Mongoloid features became predominant and
masked the Caucasoid element. Such admixture, according to Birdsell, is
found in groups such as Coahuila tribes of inland southern California and to a
certain extent in the Porno and Yuki of northern California.

Neumann (1952) viewed the differences between American Indian groups
as a result of successive migrations. Neumann believes that all but one of his
hypothesized eight varieties represent separate migrations to the New World
from northeastern Asia. These migrations began with the Otamid variety, a
rugged, long-headed people with large mandibles, and continued up to late
prehistoric times with the Deneid (Athabaskans) and Inuid (Eskimos) being
the latest. Later, however, Neumann (1960) modified his earlier views and of-
fered an evolutionary interpretation to explain the observed physical variability
between American Indian groups.

Mourant (1954) and Simmons (1956) suggested that Polynesians and North
American Indians could have shared a common gene pool in the not too dis-
tant past. Rivet (1958) suggested that the American Indians were the result of
four migrations: Mongoloids and Eskimoids (coming through the Bering Strait)
and Australoid and Melanesians (coming through the Arctic and Pacific re-
spectively).

W. W. Howells (1946) is not in agreement with the above hypotheses of a
non-Mongoloid origin of the American Indians. He argues that the early
American Indians were descendants of a generalized Mongoloid stock which

MAHMOUD Y. EL-NAJJAR

NO. 22

was present during Pleistocene times in northern Asia before the more spe-
cialized Mongoloids, such as the Chinese, had developed. Howells further indi-
cates that the groups most similar to the American Indians are those of In-
donesia, central Asia and Tibet. Evidence of this generalized American
Indian-like Mongoloid variety in Asia is cited by T. D. Stewart (1960) who
compared the Late Pleistocene Tzeyang and Liukang crania of western and
southern China with American Indian skulls from Florida and California.
Stewart also stated that the skulls from the upper cave of Choukoutien (Late
Pleistocene) near Peking, China, are similar to those of the American Indians.

Turner (1971), using the incidence of three-rooted mandibular first per-
manent molars (3RM1), suggested three separate migrations from Asia. The
first arrivals were the ancestors of all American Indians except for the Na-
Dene (Athabaskans and related groups) who were the second migration. The
third group (Proto-Aleut-Eskimo), according to Turner, could have entered
the New World at any time during the Upper Paleolithic.

Although linguistic and blood group data support Turner’s contention, the
use of such data has not been reliable. Swadesh (1960, 1962), on questionable
grounds derived from glottochronology, views the majority of American Indian
languages to have developed out of the single speech community “Proto-
Ancient American.” Aleut-Eskimo and Nadenean languages formed a distinc-
tive linguistic group. All North and South American Indians, with the excep-
tion of the Athabaskan speakers, show a high incidence of blood group genes
O with an extremely low incidence of blood group genes B and A. Athabas-
kans and related groups show the world’s highest known incidence of blood
type A. Eskimos and Aleuts are more like Asiatics, particularly in the high in-
cidence of blood type B (Zolatoreva, 1965), with almost equal distributions of
blood group genes A, B and O when compared to other American Indian
groups. The uncertainty of using blood group gene frequencies in elucidating
population affinities and for tracing historical relationships between closely re-
lated groups is well-documented (Hanna, 1962; Merbs, 1965). For example,
data on the ABO blood systems on Southwestern Indians and Athabaskans
(Merbs, 1965) shows Pima Indians to have an incidence more similar to the
Ramah Navajo than the latter to other Navajo tribes. Some of the Arizona
Apaches of Cibecue and East Fork show distribution more similar to the Pima
tribes in Arizona and the Tewan-speaking groups of New Mexico than to the
Cedar Creek and San Carlos Apache tribes, also of Arizona.

There are other physical characteristics which clearly distinguish the Amer-
ican Indians from Aleuts and Eskimos. According to Laughlin (1967), Eskimos
and Aleuts, along with the Chuckchi, Koryak, and possibly the Kamchadal en-
circle the Bering Sea and compose a biologically related group, the Bering Sea

1976

THE FIRST AMERICANS

Mongoloid. In their physical characteristics, Eskimos and Aleuts (Laughlin,
1950, 1963, 1966, 1967) have many common elements that establish a close af-
finity with the Chuckchi and Asiatic Mongoloids, rather than with American
Indians. Common features are large heads and faces, large mandibles, high
frequency of mandibular torii, thickening of the tympanic plate which is often
pronounced, and narrow nasal bones often achieving a world extreme in East-
ern Eskimos. There are many dental traits common to Eskimos and Aleuts,
but not American Indians. Among these are the frequent absence of third mo-
lars, the large lateral as well as medial incisors, and the three-rooted mandibu-
lar first permanent molars. Physiologically, the Eskimos display differences
from Indians in their cold adaptations, especially in their elevated basal me-
tabolism (Milan, 1963). Eskimos also have a high incidence of separate neural
arches and other anomalies, again showing a greater affinity with Asiatic Mon-
goloids than with the American Indians (Merbs, 1963).

THE BERING STRAIT

It is now accepted that man’s first crossing from Asia into the New World was
by way of the Bering Land Bridge connecting Siberia and Alaska some time
during the Late Pleistocene (Hopkins, 1967). Who were these people? How did
they get here?

Where the Bering Strait now exists, a broad land area called Beringia (west-
ern Alaska, northeastern Siberia and the shallow parts of the Bering Sea and
Chuckchi Sea) was present several times during the Wisconsin glacial stage
(Butzer, 1971). The Bering Strait is only 56 miles wide at its narrowest point. A
land bridge would form if sea level dropped 120 feet below its present level (So-
lecki, 1951a, 1951b; Creager and McManus, 1967) connecting Siberia and
Alaska by way of St. Lawrence Island (Fig. 1).

Geological evidence shows that the Wisconsin glacier reached its maximum
about 40,000 years ago and lowered the sea level by about 460 feet (Haag, 1962).
The bridge was submerged again approximately 28,000 to 25,000 years ago. On
the basis of climatic evidence, the land bridge could have lasted from about
25,000 to about 11,000 years ago. The lowering of the sea level exposed nearly
all the Bering-Chuckchi platform connecting Alaska and Siberia by a plain ex-
tending from the north shore of the shrunken Bering Sea to the south shore of
the Arctic Ocean (Hopkins, 1959). According to Chard (1959), northeastern
Asia and Alaska were glaciated only in the mountainous regions even during the
maximum extension of the ice sheets. The coastal plains of northeastern Alaska
were free of ice and much easier to travel than the mountainous region to the
south. Furthermore, neither the Chuckchi Peninsula in Siberia nor the Seward

ARCTIC OCEAN

MAHMOUD Y. EL-NAJJAR

NO. 22

t p

OCEAN

1976

THE FIRST AMERICANS

Peninsula of Alaska were glaciated during the Wisconsin period. The Bering
Land Bridge which also existed at the same time was glacier-free, making the
crossing even easier. With no physiographic barrier, Asiatic people and ani-
mals could have moved freely across the land bridge during most of the Wiscon-
sin Stage. Haag (1962) indicates that fossil evidence for the origin and geo-
graphic distribution of North American mammals shows that many animals
crossed the Bering Land Bridge to the New World during the Wisconsin Stage.
These include mastodon, mammoth, musk oxen, moose, bear, bison, mountain
sheep, goats, elk, camels, fox, wolves and horses (Haag, 1962:114).

During the Late Pleistocene the environmental conditions of Beringia were
very similar to those of northeastern Siberia (Colinvaux, 1964). The abundance
of horse, bison, woolly mammoth, antelope and yak made this area economically
attractive to the Asiatic hunters. Following the game, these hunters moved east-
ward until this pursuit led them unknowingly across the Bering Land Bridge into
Alaska and on to North and South America. These migrants may have moved
into the New World on the winter pack ice even when the land bridge was sub-
merged. It is also possible these first Americans had developed small sea crafts
capable of negotiating the Bering Strait.

It is conceivable that Asiatic hunters deliberately undertook a long journey
into an unknown land. Population movement appears to have been sporadic and
to have spread slowly. Population pressure, abundance of game and climatic con-
ditions were both motivating and limiting factors which determined the magni-
tude and rate at which the first migrants moved out of Asia. Those who made
the crossing were probably adapted to the conditions of the peri -glacial ecosys-
tem, and had developed adequate tools and hunting techniques. These people
spread eastward along the Alaskan foothill country, then southward. The ad-
vantages of traveling such a route have been outlined by Irving (1953). These
areas (1) are comparatively dry in contrast with the wet lowland tundras, (2) are
good for sighting game, (3) offer the advantage of both forest and tundra, being
open enough for pursuit of game but with enough tree and bush cover to provide
shelter, (4) are preferred routes for major game trails that are otherwise re-
stricted by rugged mountains and soggy grounds. Wilmsen (1965) suggests it is
important to note that this was the only type of environment which might be vir-
tually continuous from central Siberia to central North America during glacial
periods. Once on Beringia, with the advantages of sea, rivers and nearby forests,
some of these human bands may have become established with permanent occu-
pations. As the Bering Sea Platform slowly submerged due to warmer tempera-
tures, some group dispersal to the interior may have occurred. These hunters
would not have the tendency to go back to Siberia but rather to push forward
toward the interior.

MAHMOUD Y. EL-NAJJAR

NO. 22

EVIDENCE FROM THE NEW WORLD

The oldest carbon 14 dated New World human skeleton (skull) is Los An-
geles Man at 23,600 B.P. (Wormington, 1971), and the Laguna Beach skull
dated at about 17,000 B.P. (Berger and Libby, 1969). The best dated skeletal
remains are the 10,750 ± 500 B.P. -year-old cranium fragments from

Marmes, Washington.

In North America, Indian cultural traditions are usually subdivided into
three successive though overlapping horizons on the basis of their style: Llano,
Folsom and Plano. The Llano complex, including Clovis and Sandia points,
dates some time around 10,000 to 12,000 years ago. The Llano is an early
plains hunting complex known mainly from sites in the southwestern United
States and Mexico, of which Lehner Ranch (Haury et al, 1959) and Black-
water Draw (Sellards, 1952) are typical. Clovis points have been dated at 9250
± 300 B.P. at the Naco Site in Arizona, and seven other radiocarbon dates at
Lehner Site range from 7022 ± 450 to 12,000 ± 450 B.P. (Haury et al,
1959). Folsom points seem to have been the regional development of the
Llano. A shift from mammoth to bison hunting is evident between the Llano
and Folsom Sites, but there seem to be no major structural changes between
the artifact assemblages (Willey, 1966). Folsom points have been dated at
three places. The Lubbock Site in the Texas Panhandle 4s dated at 9883 +
350 B.P. (Sellards, 1952), the Lindenmeier in Colorado at 10,780 ± 375
(Haynes and Agogino, 1960), and Brewster in eastern Wyoming at 10,375 +
700 (Krieger, 1964). At Graham Cabe, Missouri, and Modoc Rock Shelter in
Illinois, points of “piano” types range from 8830 + 500 to 10,651 ± 651 B.P.
A date of 11,200 ± 800 comes from the bottom of the Modoc Rock Shelter
(Krieger, 1964).

There is other evidence suggesting that man’s arrival in the New World
could have occurred much earlier. The evidence for such an early appearance
comes from several sites in North and South America, and is based on artifact
assemblages. In general, these “hypothesized” tools are crude, percussion-
flaked tools, scrapers, and choppers. Most of these claims for the “pre-projec-
tile point” cultural tradition rests on typology alone.

The earliest of the radiocarbon dates for these suggested tools in the New
World come from Lewisville, in northern Texas. Twenty-one hearths, fossil
bones, charred hackberry seeds and a crude chopper or scraper show some evi-
dence of human occupation (Crook and Harris, 1957). A date of 37,000 B.P.
has been determined. The antiquity of this site has been challenged and the
hearths are not believed to have been man-made (Heizer and Brooks, 1965).
The clovis-type projectile points recovered suggest either a mixture of later
material with earlier geological strata or the points were simply “planted”

1976

THE FIRST AMERICANS

there. On Santa Rosa Island dwarf mammoth bones and burned bone frag-
ments gave a radiocarbon date ranging from 30,000 to more than 37,000 years
ago. Only one specimen of what possibly is a crude chipped-stone has been
found with any of the hearths or bone deposits (Orr, 1968). Radiocarbon dates
on “presumed” charcoal from Tule Spring in southern Nevada were more than
23,000 and 28,000 years old (Harrington and Simpson, 1961). Recent excava-
tions at Tule Springs do not lend support to the original claims. Laboratory
analysis has shown that the dark, carbonaceous materials from the supposed
“hearths” may not have been entirely charcoal. The earliest evidence of man’s
presence at Tule Springs is now placed at about B.C. 11,000 (Bryan, 1964;
Shutler, 1965).

The earliest definite proof of man’s presence in South America comes from
two localities: the Chivateros I complex in the Chillon Valley of the central
coast of Peru and Lagoa Santa, Brazil. Radiocarbon dates have placed the
end of the Chivateros I occupation at B.C. 8500 (Willey, 1966). At the Lagoa
Santa caves and rock shelter, fauna, artifacts and human remains were re-
covered. Those from levels 2 and 3 averaged 9311 ± 120 B.P. and levels 6
and 7 averaged 10,024 ± 127 B.P. (Hurt, 1962). Cruxent (1968) has proposed
an arrival date in South America of 15,000 to 20,000 years ago. This hypothesis
was based on the following finds: El Jobo, 10,000 years; Las Lagunas and El
Camare, more than 16,000 years; and Muaco, 14,740 to 16,580 years (Rouse
and Cruxent, 1962). Lanning and Patterson (1967), estimated the Chuqui com-
plex in Chile and the Tortuga and Red Zone complexes in Peru as the most
ancient in the Pacific Andean region, both dating between 13,000 and 14,000
B.P. Other rough stone tool complexes from South America which have been
suggested as belonging to a pre-projectile point horizon include: Viscachani in
Bolivia (Krieger, 1964); Ghatchi I in northern Chile (La Paige, 1958, 1960;
Krieger, 1964); Ampajango in northwestern Argentina (Cigliano, 1961); Tandi-
lense in Argentine Pampas (Menghin and Bormida, 1950); Oliviense in Argen-
tine Patagonia (Menghin, 1952) and early Rio Chico, Tierra del Fuego (Viganti,
1927). An excellent survey of early man in the New World is given by Worming -
ton (1971).

EVIDENCE FROM ASIA

Several lines of evidence (Stewart, 1960) show that during the middle Late
Pleistocene, early forms of primitive Mongoloids were present in northeastern
Asia. Many of these have been implicated in the origin and evolution of recent
Mongoloids and Mongoloid- affiliated human groups. The most primitive, con-

MAHMOUD Y. EL-NAJJAR

NO. 22

sisting only of a skull cap, was found in 1958 in a limestone cave near the vil-
lage of Mapa in Kwangtung province in southern China. According to Woo
and Peng (Coon, 1962) this is the earliest fossil so far found in China with the
exception of the Homo erectus material from Choukoutien. Woo believes that
the Mapa skull had evolved to the same grade as the European Classic Nean-
derthals. According to Coon (1962) the Mapa skull stands at the threshold be-
tween the two grades of Homo and that it is essentially Mongoloid in its mor-
phology. The second find (Liukiang man) was discovered in a cave near Liu-
chow in the Kwangsi Chuang Autonomous region, also in southern China. Ac-
cording to Woo (Coon, 1962), the Liukiang man represents an early form of
the evolving Mongoloid and is the earliest fossil representative of modern man-
kind so far found in China. The third (Tze-Yang) was found in 1951 in the
Szechuan province about 700 to 800 miles southwest of Peking. Woo described
the Tze-Yang find as an early form of Homo sapiens more primitive than the
European Cro-Magnon and the upper cave people of Choukoutien. According
to Coon (1962) the Tze-Yang skull falls within the female range of both Metal
Age Prehistoric and recent North Chinese series and is essentially a Homo sa-
piens.

There are no archaeological sites in Siberia having carbon 14 dates in
excess of 25,000 years (Skimkin, 1968). According to Debetz (1960) all early
Siberian sites are of the Upper Paleolithic tradition and are concentrated in
the Lake Baikal region. Neanderthal-Mousterian sites have not been found so
far in Siberia, even though Mousterian culture remains are the best established
ancient occupation in Russia, along the Volga River and in Turkestan (De-
betz, 1960).

The earliest Paleolithic sites in eastern Siberia are located around the Lake
Baikal region. The oldest of these sites date around B.C. 20,000 (Bushnell and
McBurney, 1959). Of the Paleolithic sites, the best known are those reported
from the Lake Baikal region including Malta (14,750 B.P., Butzer, 1971) and
Buret. Artifacts include points, sidescrapers, knives, burins, semi-subterranean
structures employing mammoth bones, a variety of venus figurines and bone
needles. Two sites are of particular interest: Duiktai cave with a single date of
13,070 B.P. including bifacial, pressure -flaked projectile points and knives in
association with a mammoth fauna, and Uski with a date of 14,300 and
13,600 for levels VI and V with bifacial foliate points and knives.

Other finds in China, particularly those from the Upper Cave of Choukou-
tien, are of great interest. These remains provide the only information of the
terminal Pleistocene population of eastern Asia from which the New World
natives most likely came.

The two female skulls were described by Weidenreich (Coon, 1962) as a

1976

THE FIRST AMERICANS

Melanesian (102) and an Eskimo (103). This conclusion was based on a pre-
liminary interpretation of the unrestored skulls. His assumption that the male
skull (101) is an Ainu has also been questioned. Weidenreich made his
comparison on the basis of photographs sent him by S. Kodanei (Coon, 1962)
who at the time was working on Ainu craniology in Japan. Comparing skull
number 101 with those from a series of Ainu skulls from Hokkaido, Sakhalin
and the Kuriles, Coon reports many significant differences. For example, the
cranial length of the Upper Cave skull Is 16 mm greater than the largest Ainu
mean. The minimum breadth of the Upper Cave skull is 11 mm greater and
the nasal height is 5 mm higher than any Ainu average, and bi-orbital di-
ameter is 9 mm beyond any Ainu mean. According to Coon (1962), the Upper
Cave skull resembles the large-faced tribes of the American Plains Indians.
Coon concludes that this is particularly visible in the upper part of the nasal
skeleton and the lateral borders of the orbits, but the molars and the lower
part of the nasal skeleton are fully Mongoloid in the eastern Asiatic sense.
Morphological traits of the Upper Cave skull also appear commonly in various
American Indians and the differences may simply reflect the range of varia-
bility of these groups. Indeed, these skulls have been referred to by W. W.
Howells as “Unmigrated American Indians” (Howells, 1940).

DISCUSSION

The Asiatic origin of the New World natives Is now a generally accepted
fact. There are few serious students who any longer question either the general
genetic or geographic origin of the first human inhabitants of the Americas or
the basic routes of their initial entry. All remains recovered thus far in the New
World are Homo sapiens.

Culturally, linguistically and genetically, American Indians are more close-
ly related to Asiatics than to any other human group. Earlier hypotheses at-
tributing biological differences between American Indian groups to waves of
migrations from diverse parts of the Old World are not supported by the pres-
ent findings. There Is no evidence of any element other than Mongoloid in the
formation of the American Indian physical variety. All evidence points to an
Asiatic homeland for the New World natives.

There is no valid evidence that Australoid, Caucasoid, Negroid and
Melanesian migration to and/or admixture in the New World contributed to
the American Indian physical variety. For example, if the American Indians are
derived from a Mongoloid- Australoid admixture then they should have blood
group N which is very common among Australoids. Actually the American In-

MAHMOUD Y. EL-NAJJAR

NO. 22

dians have one of the lowest incidences of N in the world. Moreover, Austra-
loids have facial and body hair, large teeth (often exceeding those of classic
Neanderthal) and they lack the wide, flat faces, heavy noses, and pronounced
cheek bones characteristic of the American Indians. The Australoid skin color
and hair form are also very different from those of the American Indians. If
Negroid genes were present in the American Indians then R0, the sickle cell
and thalassemia genes, the African form of G6PDD, the rare gene V and the
Duffy variant, as well as other hemoglobin polymorphisms should be present.
None of the above blood characteristics are found in the American Indian.
Moreover, skull form characteristics also differentiate American Indians from
Negroids.

If Caucasian genes are present, the Rh -negative and blood group type A
should be common. These genes are nonexistent in the American Indians with
the exception of Athabaskans and related groups who have a high incidence of
blood type A. Caucasoid features not found in the American Indian are small
teeth, high incidence of Carabelli’s cusp, delayed tooth eruption, skin and hair
color, sickle cell and thalassemia genes, G6PDD and familial Mediterranean
fever. Melanesians differ in physical characteristics from American Indians by
having darker skin, and hair that curls, twists and frizzes. There is a higher in-
cidence of G6PDD, blood groups B and N. Melanesians possess the thalas-
semia gene and are subject to constant selection by a vast number of virulent
diseases (Garn, 1972).

On the other hand, the evidence strongly indicates that eastern Asiatics are
the most closely related to the American Indians. The straight, dark hair,
wide, flat faces, heavy noses, the tendency toward a Mongoloid eye, scant body
hair, and the prominence of the cheek bones are characteristics of eastern
Asiatics which are always present among American Indian natives.

That today’s American Indians differ from living Asiatics in the incidence
of certain blood group genes can be explained as follows. First, American In-
dians crossed the Bering Land Bridge at the time when genetic differentiation
in the original Mongoloid stock was taking place, and thus original differences
within this stock have been retained. Second, a number of migrations did take
place at different times and from separate areas, but still from groups falling
within the range of variation of the generalized Mongoloid stock. Third, since
the peopling of the New World, American Indians have been evolving on their
own, and differences between them and their Asiatic relatives can be at-
tributed to genetic drift and natural selection operating under variable envi-
ronmental conditions. Fourth, it is only in the simply inherited blood group
genes that the differences are most apparent. Multifactorial (polygenic) traits,
e.g. hair form, color, facial characteristics, shovel-shaped teeth, etc., do not

1976

THE FIRST AMERICANS

show such differences to exist among American Indian groups or between
them and their Asiatic relatives.

Viewed as a geographic entity and a physical variety of its own, American
Indians consistently show extreme values of several traits such as high shovel-
shaped teeth, low Carabelli’s cusp, the world’s highest incidence of blood
group O, Rh -positive gene, high incidence of blood group M, the secretor
gene, the Diego positive gene (Dia) which set them apart from other major
geographic groups and suggest basic genetic similarities producing a unique
constellation of physical characteristics. T. D. Stewart (1960:262) states, “In-
deed, it is safe to say that no population of comparable size has remained so
uniform after expanding in whatever time has been involved, over such a large
area.”

There are no clear-cut cultural or linguistic similarities between American
Indians and Asiatics. Boas (1940) indicates some similarity of the absolute
pitch of South American and eastern Asiatics’ musical instruments, the use of
birch bark for making vessels, canoes and for building houses, and the use of
slat armor and flat drums. Similarities in religious ceremonials, beliefs and
traditions have also been suggested by Boas (1940). Recently, Chard (1960) sug-
gested an apparent late “North Pacific Continuum” from Kamchatka to Puget
Sound.

Wilmsen (1964) has considered the possible cultural relationships between
the Old and New Worlds. He proposed a cultural-ecological continuum and a
technological relationship that extends from Siberia all the way into the inter-
ior of North America. Two New WTorld archaeological assemblages were con-
sidered as follows (see also MacNeish, 1959): the Kogruk Complex from Anak-
tuvuk Pass in north-central Alaska, and the British Mountain Complex from
the fifth River Delta on the Arctic Coast of Yukon Territory, Canada. Be-
tween Siberian and British Mountain materials MacNeish (1959:46) states,

“The earliest occurrences of these resemblances is the Buryet (Buret)-Malta complex of
the Trans-Baikal and perhaps it also occurs at the Chastino site of the Middle Lena.

Here are also found tools struck from discoidal cores that include unifacial points both
lenticular and lanceolate, hooked gravers, scrapers and central convex-type burins . . .
end of blade scrapers and blades and pebble choppers.”

According to Campbell (1961:16-17):

“ . . . Kogruk implements somewhat resemble points, perforators, scrapers and blades
from the earliest levels of the Malta site . . . Siberian Paleolithic sites in the Lena River
Valley have produced artifacts quite closely akin to Kogruk flakecores and blades . . .
There are, apparently, even closer connections between the British Mountain complex
and these Asian collections, especially in the categories of flake burins and bifaces.”

MAHMOUD Y. EL-NAJJAR

NO. 22

Since little information of the skeletal biology of the Mongoloid stock from
which American Indians originated is available, most archaeologists find
themselves highly dependent on lithic material for questions of origin. There
are certain important limitations in using lithic material for evidence of trac-
ing and reconstructing past biological relationships. Skeletal and genetic data
are more useful.

Cultural similarities can be indicative of biological affinity. This relation-
ship is not always valid. Similar cultural developments are known to exist in
several parts of the world without any evidence of biological resemblance.
Archaeologists must recognize that several thousand years may have elapsed
before Asiatic migrants reached the New World. New tools and techniques
could have developed en route. Tolstoy (1958) has made an extensive study of
Old and New World relationships and has concluded that many Paleo-Indian
traits, especially parallel-flaking and fluting, were of New World origin.

Chard (1959a, 1959b) suggests that the only Siberian and far eastern Asi-
atic Paleolithic cultures that were clearly old enough to have provided the cul-
tural heritage for the early immigrants were the chopping tool industries char-
acterized by rough core tools, choppers and scrapers but lacking both bifacial
blades and points. It was this kind of technology, according to Chard, that
was carried to the New World, and it was in the Americas over a span of
several thousand years that the distinctive bifaced, lanceolate projectile point
types evolved independently of any further Asiatic influence. As Wormington
(1962) has suggested, archaeologists should not look for duplications of New
and Old World tools but rather for prototypes from which New World tools
were derived. After all, the American continent was essentially isolated from
the Old World for several thousand years. This was time enough for Paleo-
Indians to develop a diverse variety of tools which were compatible with the
changing environmental conditions ranging from subarctic to high mountains
and low deserts.

At present, no evidence of a relationship between the American Indian
and Old World languages has been demonstrated (Willey, 1966). Indian-
Asiatic linguistic affinity is very distant, if it exists at all. The fundamental
structural differences in Siberian languages make it difficult to trace the origin
of the American Indian languages. In North America alone, at least six major
linguistic stocks are known to exist. The total separate and mutually unintelli-
gible languages exceed 200 (Jennings, 1968:4). The influx of the Tungus and
Turkish tribes into Siberia, although recent, also disturbed the earlier distri-
bution in one way or another, making the tracing of such relationships impos-
sible (many aboriginal languages in Siberia are no longer spoken). Shafer’s
(1952) hypothesis that the Athabaskan language is related to Sino-Tibetan is

1976

THE FIRST AMERICANS

only weakly supported. Kiparsky’s (1968) suggestion that there was contact be-
tween Sahaptian and Chuckchi - Kamcha d a speakers is also weak, resting on
the shared trait of diagonal vowel harmony. Thus it appears that until more
information Is available any statement concerning linguistic affinity can be no
more than speculative.

At this point it is Important to recognize that Asiatic Mongoloids and those
who migrated to the New World have been Independently evolving culturally,
linguistically, and biologically since their geographic separation. Individual
American Indian groups were generally small and isolated, particularly during
pre-Columbian times, thus maximizing the chance for genetic diversity between
these groups and decreasing variability within such groups. The picture of
only a few small groups actually completing the crossing into the New World
is compatible both with geological and environmental conditions and with the
observed degree of biological differences among the New World natives. The
high frequency of blood group gene O and virtual absence of B and A in
North and South America with the exception of Athabaskan speakers and the
presence of A, B and O' in Eskimos and Aleuts could suggest that early differ-
ences have been retained. A more plausible hypothesis is that the simply inher-
ited blood group genes change at a much faster rate than the multifactorial
(polygenic) traits, which show similar incidence among all New World natives.
Other characteristics of certain American Indian natives, such as large chest,
lungs and hearts among Andean groups in South America are primary adaptive
responses to the environment. The prominent nose and projecting chin of the
Plains Indians, the high frequency of dislocated hip among the Navajos and
Apaches, the beard hair among the Palute and the Coahuila, the albinism
among Hopi, Zuni, Jemez and San Bias, and the obesity among Pima and
Papago are traits either environmentally determined or were brought about by
the action of selection, genetic drift and other genetic determinants acting on
small isolated groups.

At present, virtually nothing is known about the rate of evolution among
human populations. Genetic differences often observed among prehistoric and
recent historic skeletal and/or living American Indian groups can be ex-
plained without Invoking hybridization and/or multiple migrations. Genetic
drift and/or natural selection may have operated singly or in combination to
produce the observed differences. The first New World inhabitants arrived
during the phase of evolution in which differentiation in the original Asiatic
stock was taking place. Since then, physiological adaptations occurred under
an extreme range of environmental conditions including subarctic, desert,
and tropical rain forests, and therefore, biological differences can simply be
attributed to environmental extremes. Under such conditions, genetic varia-

MAHMOUD Y. EL-NAJJAR

NO. 22

tion, without obscuring the basic assumption of the genetic homogeneity of
the American Indians, is expected. Reconstruction of the general Mongoloid
physical type, as well as environmental conditions which existed at the same time
as the hypothesized migrations, is crucial and must be thoroughly investigated.
There is a further need for the recovery of skeletal and cultural materials from
submerged sites on the continental shelf, which would shed more light on the
nature of migration(s) into the New World.

REFERENCES CITED

Abbot, C. C., 1873, Relics of a Homestead of the Stone Age: The American Naturalist, v. 7, no.
5, pp. 271-278.

, 1889, Evidence of the antiquity of man in eastern North America: Proceedings of the

American Association for the Advancement of Science, v. 37, pp. 293-315.

Beals, R. L., 1957, Father Acosta on the first peopling of the New World: American Antiquity, v.
23, no. 2, pp. 182-183.

Berger, R. and Libby, W. F., 1969, U.C.L.A. Radiocarbon Dates IX: Radiocarbon, v. 11, no. 1,
pp. 194-209.

Birdsell, J. B., 1951, The problem of early peopling of the Americas as viewed from Asia. In:
W. S. Laughlin (ed.), Papers on the Physical Anthropology of the American Indian: Viking
Fund Publication, pp. 1-68.

Boas, F., 1940, Race, Language and Culture: The Free Press.

Bryan, A. L., 1964, New evidence concerning early man in North America: Man, v. 186, pp. 152-
153.

Bushnell, G. and McBurney, C., 1959, New World origins seen from the Old World: Antiquity,
v. 33, no. 133, pp. 93-101 .

Butzer, K. W., 1971, Environment and Archeology: An ecological approach to prehistory: Aldine
Atherton Inc.

Campbell, J. M., 1961, The Kogruk Complex of Anaktuvuk Pass, Alaska: Anthropologica, n.s., v.
3, no. 1 , pp. 3-20.

Carter, G. F., 1957, Pleistocene man at San Diego: The Johns Hopkins Press.

Chard, C. S., 1959a, New World Origins, A reappraisal: Antiquity, v. 33, no. 129, pp. 44-49.

, 1959b, Old World sources for early lithic cultures: Actas del 33rd Congreso Inter-
national de Americanists, pp. 314-320.

, 1960, Northwest-coast northeast Asiatic similarities: a new hypothesis. In: Men and

Culture, A. F. C. Wallace, ed: University of Pennsylvania Press.

, 1963, The Old World roots: Review and speculation: Anthropological Papers, v. 2,

no. 2, University of Alaska.

1976

THE FIRST AMERICANS

Cigliano, C. M., 1961, Noticia sobre una nueva industria preceramica en el valle de Santa Maria
(catamarca): El Ampajanguense, Anales de Arqueologica y ethnologia, v. 16, pp. 169-179.

Colinvaux, P. A., 1964, The environment of the Bering Land Bridge: Ecological Monographs, v.
34, pp. 297-329.

Coon, C. S., 1962, The origin of races: A. A. Knopf.

Coon, C. S., Garn, S. M. and Birdsell, J. B., 1950, Races — A study of the problem of race forma-
tion in man: C. C. Thomas.

Creager, J. S. and McManus, D. A., 1967, Geology of the floor of Bering and Chuckchi Seas. In:
The Bering Land Bridge, D. M. Hopkins, ed: Stanford University Press.

' Crook, W. W. and Harris, R. K., 1957, Hearths and artifacts of early man near Lewisville, Texas
and associated faunal material: Bull. Texas Arch. Soc., v. 28, pp. 7-97.

Cruxent, J. M., 1968, Theses for meditation on the origin and distribution of man in South Amer-
ica. In: Biomedical challenges presented by the American Indian: Pan American Health Or-
ganization Scientific Publication No. 165, pp. 11-16.

Dali, W. H., 1877, Tribes of the extreme northwest: Bur. Amer. Ethnol: Contributions to North
American Ethnology, v. 1 , pp. 1 -121 .

Damon, P. E. and Long, A., 1962, Arizona radiocarbon dates III: Radiocarbon, v. 4, pp. 239-
349.

Debetz, G. F., 1960, Summary of the paleo-anthropological and ethnological investigation in the
U.S.S.R. In: Men and Culture, A. F. C. Wallace, ed: University of Pennsylvania Press.

Drake, S. C., 1880, The aboriginal races of North America: Hurst and Company.

Driver, H. E., 1961, Indians of North America: University of Chicago Press.

Garn, S. M., 1972, Human Races: C. C. Thomas, Springfield, 111.

Gladwin, H. S., 1947, Men out of Asia: McGraw-Hill Book Co. Inc.

Green, F. E., 1963, The Clovis blades: An important addition to the Llano complex: American
Antiquity, v. 29, no. 2, pp. 145-165.

Greenberg, J. H., 1960, The general classification of Central and South American languages. In:
Men and Culture, A. F. C. Wallace, ed: University of Pennsylvania Press, pp. 791-794.

Haag, W. G., 1962, The Bering Strait Land Bridge: Scientific American, v. 206, no. 1, pp.
112-123.

Hanna, B. L., 1962, The biological relationships among Indian groups of the Southwest: Am. J.
Phys. Anthrop., v. 20, pp. 499-508.

Harrington, M. R. and Simpson, R. D., 1961, Tule Springs, Nevada, with other evidences of
Pleistocene man in North America: Southwest Museum Papers no. 18.

Haury, E. W. , Sayles, E. B. and Wasley, W. W., 1959, The Lehner mammoth site, southeastern
Arizona: American Antiquity, v. 25, no. 1, pp. 2-30.

Haynes, C. V. and Agogino, G., 1960, Geological significance of a new radiocarbon date from the
Lindenmeier site: Denver Museum of Natural History Proceedings, no. 9.

Heizer, R. F. and Brooks, R. A., 1965, Lewisville — ancient campsite or wood rat house?: MS.

MAHMOUD Y. EL-NAJJAR

NO. 22

Hoijer, H., 1946, Introduction, In: Linguistic structure of native America. H. Hoijer, ed: Viking
Fund Publications in Anthropology, no. 6, pp. 9-29.

Hooton, E. A., 1947, Up from the ape: The Macmillan Company.

Hopkins, D. M., 1959, Cenozoic history of the Bering land bridge: Science, v. 129, no. 3362,
pp. 1519-1528.

, editor, 1967, The Bering Land Bridge: Stanford University Press.

Howells, W. W., 1940, The origin of American Indian race type. In: The Maya and their neigh-
bors: D. Appleton-Century Co. Inc., pp. 3-9.

, 1946, Mankind so far: The American Museum of Natural History Science Series,

Doubleday.

Hrdlicka, A., 1920, Shovel-shaped teeth: Am. J. Phys. Anthrop., v. 3, pp. 429-465.

Hurt, W. R., 1962, New and revised radiocarbon dates from Brazil: Museum News, W. H. Over
Mus., State University of South Dakota, 23, nos. 11-12, pp. 1-4.

Imbelloni, J. , 1943, The peopling of America: Acta Americana, v. 1, no. 3, pp. 309-330.

, 1958, Nouveaux apports a la classification de l’homme americain: Miscellanea Paul

Rivet Octogenario Dicatal. Universidad Nacional Autonoma de Mexico, Mexico, D. F., v. 1,
pp. 107-136.

Irving, W., 1953, Evidence of early tundra cultures in northern Alaska: University of Alaska
Anthropological Papers, v. 1, no. 2, pp. 55-85.

Jefferson, T., 1801, Notes on the state of Virginia.

Jennings, J. D., 1968, Prehistory of North America: McGraw-Hill Book Co.

Kiparsky, P., 1968, How abstract is phonology: Indiana University Linguistic Club.

Krieger, A. D., 1962, The earliest cultures in the western United States: American Antiquity, v.
28, no. 2, pp. 138-143.

, 1964, Early man in the New World. In: Prehistoric man in the New World, J. D.

Jennings and E. Norbeck, eds: The University of Chicago Press, pp. 23-81 .

Lanning, E. P., and Patterson, T. C., 1967, Early man in South America: Scientific American, v.
217, no. 5, pp. 44-50.

Laughlin, W. S., 1950, Blood groups, morphology and population size of the Eskimos: Cold
Spring Harbor Symposia on Quantitative Biology, v. 15, pp. 165-173.

, 1963, Eskimos and Aleuts, their origins and evolution: Science, v. 142, pp. 633-645.

, 1966, Genetical and anthropological characteristics of the arctic populations. In:

The biology of human adaptability, P. T. Baker and J. S. Weiner, eds: Clarendon Press,
Oxford.

, 1967, Human migration and permanent occupation in the Bering Sea. In: The

Bering Land Bridge, D. M. Hopkins, ed: Stanford University Press, pp. 409-450.

LePaige, G., 1958, Antiguas Culturas atacamenas en la cordillera chilena: Revista Universitaria
Catolica de Chile, v. 43, pp. 139-165.

, 1960, Antiguas Culturas atacamenas en la cordillera chilena: Revista Universitaria

Catolica de Chile, v. 44-45, pp. 191-206.

1976

THE FIRST AMERICANS

Lubbock, Sir John (Lord Avebury), 1878, Prehistoric times: D. Appleton -Century Co. Inc.

MacNeish, R. S., 1959, Men out of Asia, as seen from the northwest Yukon: Anthropological
Papers of the University of Alaska, v. 7, no. 2, pp. 41-70.

Menghin, O. F. A., 1952, Fundamentos cronologicos de la prehistoria de Patagonia: Runna, v.
5, pp. 23-43.

Menghin, O. F. A. and Bormida, M., 1950, Investigaciones prehistoricas en cuevas de Tandilia,
Provincia de Buenos Aires: Runna, v. 3, parts 1 and 2.

Merbs, C. F. , 1963, The Sadlermiut Eskimo vertebral column: Master’s thesis, University of Wis-
consin.

' , 1965, Blood group distribution in the American Southwest: Unpublished manu-

script.

Milan, F. A., 1963, An experimental study of the thermoregulation in two arctic races: Ph.D. dis-
sertation, University Microfilm Inc.

Mourant, A. E., 1954, The distribution of the human blood groups: C. C. Thomas.

Neumann, G. K., 1952, Archeology and race in the American Indian. In: Archeology of East-
ern United States, J. B. Griffin, ed: The University of Chicago Press, pp. 13-34.

, 1960, Origins of the Indians of the middle Mississippi area: Proceedings of the

Indiana Academy of Sciences, v. 69, pp. 66-68.

Orr, P. C., 1956, Radiocarbon dates from Santa Rosa Island: Santa Barbara Museum of Natural
History, Department of Anthropology Bulletin No. 2.

, 1968, Prehistory of Santa Rosa Island: Santa Barbara Museum of Natural History,

Department of Anthropology.

Rau, C. , 1882, Observations on cup-shaped and other lapidarian sculptures in the Old World
and in America: Bureau of American Ethnology, Contributions to North American Eth-
nology, v. 5.

Rivet, P. . 1957, Les origins de l’homme American: Paris, Gallimond.

Rouse, I. and Cruxent, J. M., 1962, Arquelogia Venezolana: Caracas, Ed. Espanola.

Sellards, E. H., 1952, Early man in America. A study in prehistory: University of Texas Press.

Shafer, R., 1952, Athabaskan and Sino-Tibetan: Inter. J. Am. Linguistics, v. 18, no. 1, pp.
12-19.

Shimkin, D. B., 1960, Western Siberia archeology, an interpretative summary. In: Men and
Culture, A. F. C. Wallace, ed: University of Pennsylvania Press, pp. 648-661.

Shimkin, E. M., (translator), 1968, Radiocarbon dates from Soviet laboratories: Radiocarbon
10, no. 2, pp. 417-467.

Shutler, R., 1965, Tule Springs Expedition: Current Anthropology, v. 6, no. 1, pp. 110-111.

Simmons, R. T. , 1956, A report of blood groups genetical survey in eastern Asia, Indonesia,
Melanesia, Micronesia, Polynesia, and Australia in the study of man: Anthropos, v. 51,
pp. 500-512,

Solecki, R. S., 1951a, Archeology and ecology of the Arctic slope of Alaska: Smithsonian Annual
Report for 1950, pp. 469-495.

MAHMOUD Y. EL-NAJJAR

NO. 22

, 1951b, How man came to North America: Scientific American, v. 184, no. 1,

pp. 11-15.

Stewart, T. D., 1953, Ameriques: Catalogue des hommes fossiles: Fascicule V des compte rendus
de la XIX Session des Congress Geologique International, pp. 354-364.

, 1960, A physical anthropologist’s view of the peopling of the New World: South-
western J. Anthrop., v. 16, no. 3, pp. 259-273.

Sulimirski, T., 1970, Prehistoric Russia: J. Baker, London.

Swadesh, M., 1960, La linguistica como instrumento de la prehistorica: Instituto Nacional de
Anthropologica e Historia, Direccion de Prehistorica, pub. 9, Mexico, D,F.

, 1962, Linguistic relations across the Bering Strait: American Anthropologist, v. 64,

no. 6, pp. 1262-1291.

Thompson, J. E. S., 1958, The first peopling of the New World: American Antiquity, v. 24, no.
2,p. 184.

Tolstoy, P., 1958, The archeology of Lena Basin and its New World relationships, Part II: Amer-
ican Antiquity, v. 24, no. 1, pp. 63-81.

Treutlein, T. E. (translator), 1949, Pfefferkorn’s description ol the province of Sonora: Coro-
nado Quarto, Centennial Publications, 1540-1940, v. 12, University of New Mexico Press.

Turber, C. G. II, 1971, Three-rooted mandibular first permanent molars and the question of
American Indian origins: Am. J. Phys. Anthrop., v. 34, no. 2, pp. 229-241.

Viganyi, M. A., 1927, Arquelogia y anthropologia de los Conchales Gueginos: Revista del Museo
de la Plata, v. 30, pp. 79-143. Buenos Aires.

Wendorf, F. and Krieger, A.D., 1959, New light on the Midland Discovery: American Antiquity,
v. 25, pp. 66-78.

Willey, G. R., 1966, An introduction to American archeology: Prentice Hall.

Williams, S. C. (ed.), 1930, Adair’s history of the American Indians: Wataugo Press.

Wilmsen, E. N., 1964, Flake tools in the American Arctic, some speculations: American An-
tiquity, v. 29, no. 3, pp. 338-345.

, 1965, An outline of early man studies in the New World: American Antiquity, v.

31, no. 2, pp. 172-192.

Winchell, A., 1880, Preadamites: S. C. Griggs and Company.

Winsor, J. (ed.), 1889, Narrative and critical history of North America: Aboriginal America, v. 1,
Houghton Mifflin Company.

Wormington, H. M., 1961, Prehistoric cultural stages of Alberta, Canada: Homenja a Pablo
Martinez de Rio, 25 Anniversario de la edicion de los Origines Americanos, pp. 163-171 .

, 1962, A survey of early American prehistory: American Scientist, v. 50, no. 1,

pp. 230-242.

, 1971, Comments on early man in North America, 1960-1970: Arctic Anthropology,

v. 8, no. 2, pp. 83-91 .

Zolotareva, I. M., 1965, Blood group distribution of the peoples of northern Siberia: Arctic An-
thropology, v. 3, no. 1, pp. 26-33.

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO OCTOBER 22, 1976 NUMBER 23

A NEW CRINOID FROM THE PUTNAM HILL
LIMESTONE MEMBER
(ALLEGHENY GROUP, PENNSYLVANIAN)

OF OHIO

J.J. BURKE

Senior Scientist, Cleveland Museum of Natural History
Research Associate, West Virginia Geological Survey

ABSTRACT

An inadunate crinoid, Diphuicrinus ohioensis sp. nov., from the Putnam Hill Lime-
stone, Allegheny Group, Middle Pennsylvanian of Ohio is described. Features of the dor-
sal cup that characterize the species are (1) width more than three times the height, (2)
similarity to cups of D. patina Strimple and Knapp and D. coalensis Strimple and Moore
in lateral profile, (3) pentagonal outline in ventral view, (4) radial forefacets less deep
than internal facetal areas, (5) prominence of anal X, and (6) small nodes on infrabasals
and proximal portions of basals; large discrete nodes on remainder of cup.

The genus Diphuicrinus Moore and Plummer, 1938, is considered to compose an
aberrant line of Pennsylvanian crinoids and to be the sole representative of the family
Diphuicrinidae. The genus Graffhamicrinus Strimple, 1961, is regarded as invalid, be-
cause based on surface ornament, an inadequate criterion for generic distinction; the
genus is here treated as a junior subjective synonym of Delocrinus Miller and Gurley,
1890. Study of the holotype of Delocrinus aristatus Strimple shows it to be a typical rep-
resentative of Diphuicrinus-, the species is here designated Diphuicrinus aristatus (Strimple,
1949) comb. nov.

INTRODUCTION

Specimens comprising the types of a new species of the Pennsylvanian crin-
oid genus Diphuicrinus described in the following pages are reposited in the
Cleveland Museum of Natural History (CMNH), the National Museum of
Natural History (USNM) and Orton Museum, Ohio State University (OSU). The
Orton Museum specimens were formerly in the collection of the Geology Depart-

J.J. BURKE

No. 23

ment of Ohio University. Dr. Myron Sturgeon of that institution kindly per-
mitted me to describe them.

I am indebted to Dr. Patrick Sutherland of the University of Oklahoma
(OU), to Mr. Harrel Strimple of the University of Iowa (SUI), and to Mr. James
Murphy of Case Western Reserve University for the opportunity to study crinoid
material pertinent to this investigation. I thank Dr. Mildred Walmsley for tech-
nical assistance in preparation of this paper.

All of the crinoid specimens designated as types of the new species described
herein were collected from shaly portions of the Putnam Hill Limestone Mem-
ber, Allegheny Group, Pennsylvanian, in Elk Township, Vinton County, Ohio,
near the town of McArthur. Morningstar (1922) applied the name McArthur
Limestone to the unit from which these crinoids were taken, but the name Put-
nam Hill Limestone (Andrews, 1870) has priority over Morningstar’s appella-
tion and is the term currently used in Ohio. Locations of the collecting sites are
given in the following summary:

Locality 1: Abandoned strip mine east of township road 17, in the SE V4
sec. 7, Elk Twp. , Vinton County, Ohio (lat 39°16 '45 " N. , long 82°30 '31 " W.)

4.4 km northwest of McArthur, Ohio.

Locality 2: Abandoned strip mine NE of township road 17, in the NE Vi
sec. 17, Elk Twp., Vinton County, Ohio (lat 39°16'26"N., long 82°30 TO " W.)

3.5 km northwest of McArthur, Ohio.

Locality 3: Abandoned strip mine in the NE Vi sec. 17, Elk Twp., Vinton
County, Ohio (lat 39°16 '35" N., long 82°29 '40" W.) north and east of a tribu-
tary of Elk Fork, and 3.4 km northwest of McArthur, Ohio. The original label
notes that the locality is “across the ravine from the old Moore mine.” The
Moore mine was designated the type locality of the McArthur Limestone by
Morningstar (1922).

SYSTEMATIC PALEONTOLOGY
Class CRINOIDEA Miller, 1821
Family DIPHUICRINIDAE Strimple and Knapp, 1966
Genus DIPHUICRINUS Moore and Plummer, 1938
Diphuicrinus ohioensis sp. nov.

Figs. 1-9

Diagnosis: Dorsal cup more than three times wider than high (H/W ratio
about 0.30) and near that of D. coalensis Strimple and Moore in size; resembles
cups of D. coalensis Strimple and Moore and D. patina Strimple and Knapp
in lateral profile; outline somewhat rounded in dorsal view, pentagonal in ven-

Figures 1-3. Diphuicrinus ohioensis, sp. nov. Holotype, CMNH 3800, from the Putnam
Hill Limestone, Allegheny Group, Vinton County, Ohio. Fig. 1 , dorsal view; fig. 2;
posterior view; fig. 3, ventral view. All x2.

4 6

Figures 4-6. Diphuicrinus ohioensis, sp. nov. Paratype, OSU 31504, from the Putnam
Hill Limestone, Allegheny Group, Vinton County, Ohio. Fig. 4, dorsal view; fig. 5,
posterior view (dorsal side up). Note collapsed basals, and radials still in place. Fig. 6,
ventral view. All x2.

J.J. BURKE

No. 23

tral view; radial forefacet less deep than internal facetal area; anal X a promi-
nent component of cup. Principal ornament large discrete nodes on anal X,
radials and distal portions of basals; smaller nodes on infrabasals and proximal
region of basals.

Types: Holotype CMNH 3800, collected by Delbert Windle. Paratypes: OSU
31504, collected by Myron Sturgeon and Richard Hoare; OSU 31503, col-
lected by Don Crissinger; USNM 166575 and USNM 166576 (partial dorsal cup),
both collected by James Murphy.

Occurrence: Putnam Hill Limestone Member, Allegheny Group, Pennsylvani-
an.

Localities: Elk Township, Vinton County, Ohio (see Register of Localities).
Holotype, CMNH 3800 and paratype OSU 31503 from Locality 1. Paratype
OSU 31504 from Locality 2. Paratypes USNM 166575 and USNM 166576 from
Locality 3.

Description: Dorsal cup low truncate bowl -shaped, more than three times wider
than high, rounded pentagonal in dorsal outline, pentagonal in ventral. Lateral
walls steep. Height and width dimensions of basal impression more than half
those of the cup. Slopes of impression moderate. Infrabasals convex, discrete,
gently downflaring, of medium size.

Basals with moderate but steepest slopes within basal impression; nearly
flat proximally, maximum longitudinal curvature in vicinity of basal plane,
less along lateral wall; slightly concave transversely in region of basal plane.
Interbasal sutures inconspicuous. Distal borders of basals slightly curved. Except
for CD basal, little exposure of basals on lateral wall. CD basal quite elongate,
truncate distally.

Proximal tips of radials mark basal plane. Radial slopes gentle proximally,
conforming with those of basals, but steep along lateral wall. Transverse con-
vexity of radials slight. Forefacet less deep than internal facetal area. Outer
facetal ridge sturdy in holotype. External ligament pit slitlike, transverse ridge
elevated, remarkably slender and sharp, denticulate. Inner facetal area faces
inward. Lateral furrows narrow. Adsutural slopes steep, adsutural valley
floors rising with lateral ridges. Lateral lobes with winglike borders. Intermuscu-
lar notch very broad. Intermuscular furrow short, extends to intermuscular
elevation separating rounded muscular-basin areas.

Anal X slightly damaged in holotype, partly recumbent and moderately
elevated on internal side, with about one-third of height above summits of radi-
als; shows two distal facets; proximally rests on truncated tip of CD basal. In all
the types anal X is prominent component of cup.

Ornament of basals, radials and anal X consists mainly of fairly large, well-
separated nodes, with a few small nodes scattered among them. Small nodes

1976

NEW SPECIES OF DIPHUICRINUS

predominant on proximal parts of basals. Small nodes also present on infra-
basals.

Measurements: Linear measurements, in mm, taken on two specimens are as
follows: CMNH 3800 (holotype) - Dorsal cup height, 7.1 (appr.), width, 24.0
(appr.), H/W ratio, 0.30, basal impression height, 4.0, width, 13.0; infrabasal
circlet, width, 5.7; basal (EA) length, 8.4, width, 8.7; radial (A) length, 6.7
width, 12.8; suture between basals, length, 4.7 (appr.), suture between radials,
length, 4.0; anal X, height, 5.9, width 3.9. OSU 31504 (paratype) - Dorsal cup
height, 7.1, width 22.2, H/W ratio, 0.32; basal impression width, 13.8; radial
(E) length, 6.6, width, 12.6.

Comparisons: Dorsal cups of Diphuicrinus ohioensis sp. nov., D. patina Strim-
ple and Knapp, and D. coalensis Strimple and Moore agree in showing, in later-
al profile, subparallel, almost plane, dorsal and ventral borders, steep lateral
walls (least steep in D. coalensis) and width of cup more than three times the
height. In dorsal and ventral views, D. patina is more rounded; in ventral view,
D. ohioensis and D. coalensis are definitely pentagonal. In D. patina the radial
forefacet is as deep as the internal ligament area; in D. coalensis and D. ohio-
ensis it is less. In D. ohioensis the large nodes are more prominent; they are also
more discrete than in D. patina, less sparse than in D. coalensis. The infra -
basals are nodose in D. ohioensis, not ornamented in D. patina, and evidently
not ornamented in D. coalensis also. Anal X is a prominent plate in D. ohio-
ensis, of moderate size in D. patina, and reduced in D. coalensis. The holotype
of D. ohioensis greatly exceeds that of D. patina in size, but is nearly the same
size as the holotype of D. coalensis. The figured paratype of D. patina, SUI
11901 (Strimple and Knapp, 1966, fig. 23) which is larger than the holotype of
D. patina, quite clearly does not belong to that species, and possibly does not
pertain to Diphuicrinus. It shows peculiar depressions along the interbasal su-
tures that are not characteristic of Diphuicrinus, but the dorsal cup is too poorly
preserved for definite allocation.

Discussion: Moore and Plummer (1937, p. 311) indicated that they based the
genus Diphuicrinus on “the structure of the calyx, including the deep ligament
fossae of the suture faces . . The deep ligament fossae are a characteristic
feature of Diphuicrinus, it is true, as shown both by isolated plates and by plates
still in association within the cup. However, the fossae do not persist throughout
the life span of the animal. Some isolated radial and basal plates display them
on all suture faces. These plates must pertain to young animals, for other plates
show the fossae filled with calcareous upgrowths that make flush contacts with
similar deposits on suture faces of adjoining plates. Along such contacts the
plates tend to be bound together; apparently at this stage they were connected,
not only by ligaments, but also by calcareous deposits. This last stage is first

J.J. BURKE

No. 23

traceable along the interradial sutures (and along the sutures between the pos-
terior radials and anal X). It is not unusual to find specimens of Diphuicrinus
in which the basals, although still lying within the dorsal cup, have fallen away
from sutural contact with the radials and infrabasals. My paratypes of Diphui-
crinus ohioensis demonstrate this (see Figures 4-5). The radials of these dorsal
cups continue to remain upright, bound together interradially in normal posi-
tion in the cup. At later growth stages, however, the other fossae of the radials
and those of the basals became filled with calcareous deposits also, and if the cup
is preserved the plates are usually found in place, as they were in the living ani-
mal. This seems to hold for the three dorsal cups included in the type suite of
Diphuicrinus croneisi Moore and Plummer, the type species of the genus, and
may be taken as indicating a degree of maturity for these specimens.

Now it is of great interest that this final stage in sutural contact seems near-
ly to have been attained by the cup plates of the holotype of Diphuicrinus
faustus Moore and Strimple, OU 7511. Portions of the articular surfaces of the
D radial and the DE basal of this specimen are exposed because the E radial has
slipped inward. Both the interradial and the radial-basal suture faces show
surfaces that are practically flush, with elevations occupying the places of the
fossae and crenellae showing along the ridges.

The arms of Diphuicrinus are known from two specimens, both attribut-
ed to Diphuicrinus faustus Moore and Strimple by those authors (1973). The
arms of the holotype, OU 7511, were illustrated by Moore and Strimple (1973,
pi. 14, fig. la; pi. 15, fig. 4a). The crown of the paratype of D. faustus, OU
4597, was figured by Strimple and Knapp (1966, pi. 36, figs. 1, 2) but the speci-
men, as depicted more recently by Moore and Strimple (1973, pi. 14, figs, lb,
Id) no longer retains all of the surface detail of the arms, probably because of
rash use of an air abrasive machine in preparation.

Study of these two specimens gives support to my previous suggestion
(Burke, 1970, p. 9-10) that mature individuals of Diphuicrinus would be found
to have biserial arm structure. In the midregions of the arms of these two crowns
of Diphuicrinus faustus, biseriality appears to be already under way. The
secundibrachs are quite cuneate; some of them are reduced to feather edges on
the antipinnular side and do not extend the full width of the arm. But this is
not fully apparent in external view, because the pinnular ends of some of the
brachs are intruded between the pinnular ends of brachs of adjoining arms,
which makes for a remarkably robust interlocking mechanism, but the interlock-
ing brachs conceal the underlying structure of the arm. It is only the external
portions of the brachs which mesh with and slightly overlap the brachs of ad-
joining arms. This is evident when the arms are separated. Actually, at depth,
subjacent and superjacent pinnular ends of some of the secundibrachs are in

1976

NEW SPECIES OF DIPHUICRINUS

contact, and where they are in contact they fence off the tips of the antipinnular
wedges, preventing them from extending the full width of the arm.

I have attempted to illustrate the arm structure of the holotype of Diphui-
crinus faustus, OU 7511, in Figures 7, 8, and 9. In Figure 7, a right posterior
view of the crown, the interlocking secundibrachs are shown. Notches are evi-
dent in the right side of the arm of the D ray, which is elevated above and does
not mesh with the adjoining arm. The region between the arrows on the left side
of the figure includes the portions of the arms of the D ray which are illustrated
in Figures 8 and 9.

Figures 7-9. Diphuicrinus faustus Moore and Strimple. Holotype, OU 7511. Fig. 7, right
posterior view of crown, x2. Arrows delimit portions of D ray illustrated in figs. 8 and 9.
Fig. 8 , external view of portions of both arms of D ray, x 4 . 6 . Fig. 9 , lateral view of
portion of left arm of D ray, x 4.6.

J. J. BURKE

No. 23

The proximal secundibrachs of these arms of the D ray of OU 7511 are not
satisfactorily preserved, consequently in Figures 8 and 9 only that part of the
left arm beginning with the 15th secundibrach, and of the right arm the por-
tion originating with the 17th secundibrach, are illustrated. Neither arm is quite
complete distally. In Figure 8, an external view, both arms display the notches
between secundibrachs, which become increasingly prominent proximad; distad
the notches are less distinct and are not evident on the terminal parts of the
arms. The left arm interlocks with the right, but it is free on the left side, be-
cause the right arm of the E ray is displaced and lies a greater depth. On this
free side of the left arm it is evident that the antipinnular ends of the secundi-
brachs are exposed laterally, but there is no indication of the extent of their
exposure at depth. In Figure 9, the extent of their lateral exposure is evident.

Figure 9 is a lateral view of this same part of the left arm of the D ray. Dis-
tad, some of the pinnules are preserved. But proximad lies the region of great-
est interest, for here are exposed the pinnular ends of five secundibrachs with
subjacent and superjacent surfaces in contact at depth. External to the junc-
tures of the pinnular ends of these secundibrachs, four antipinnular tips of
secundibrachs are seen, cut off from lateral exposure at depth, and progressive-
ly decreasing in size and extent of lateral exposure proximad (an indication that
the antipinnular sides of the secundibrachs were being resorbed). There seems
no escaping the conclusion that this arm was becoming biserial, and in fact ex-
hibits biseriality in this region, where five successive secundibrachs on one side
are in contact and pinnulate. In advance of this region the antipinnular ends
of secundibrachs intervene between pinnular ends in typical uniserial fashion,
and a few notches are evident. It is apparent that the interlocking structures of
these arms were developed before the attainment of biseriality.

Burdick and Strimple (1973) have demonstrated that arms of fully ma-
ture specimens of Phanocrinus attain a stage of incipient biseriality, with all
the brachs on each side pinnulate, but the arms are not biserial in any strict
sense of the word. The cuneate brachs of Diphuicrinus are far more advanced
than those of Phanocrinus, but whether any species of Diphuicrinus ever ac-
quired biserial arms comparable to those of Delocrinus — or to those of Morrow -
an contemporaries of Diphuicrinus, such as Endelocrinus, remains to be dem-
onstrated.

Whatever its ultimate assignment, Diphuicrinus is quite evidently a valid
and distinct genus. Apparently the taxon is composed of species remarkable for
their slow attainment of maturity, as demonstrated by the dorsal cup, which in
some specimens appears literally to be “falling apart at the seams,” and also by
the arms, which in the two crowns that are known show only the beginnings of
biseriality, although Strimple and Moore (1971) regarded one of them (OU

1976

NEW SPECIES OF DIPHUICRINUS

4597) as “young but mature.” Another feature of Diphuicrinus , the coarse orna-
ment, suggested “specialization” to Strimple and Knapp (1966). Coarse orna-
ment does not characterize all of the species, but when present it is so marked
that it might better be regarded as evidence of overspecialization. In general,
there is much about Diphuicrinus that seems to me to indicate a crinoid stock
in decline; something akin to phylogerontism appears to be in evidence here. I
am inclined to view the taxon as a sterile offshoot of the Phanocrinus stock,
which gave rise to no other Pennsylvanian genera, and which probably became
extinct in Desmoinesian time. I consider it to be the sole member of the family
Diphuicrinidae, as did Strimple and Knapp (1966). In view of the various
unique characteristics of Diphuicrinus, the family deserves full recognition.
However, it was originally assumed that Diphuicrinus was distinguished by uni-
serial arm structure, and this was the sole basis for establishment of the family.
It is now evident that uniseriality simply represents a stage in the ontogeny of
the diphuicrind arm and does not constitute grounds for family distinction.

Almost any attempt to clarify the relationship of Diphuicrinus to several
other genera contemporary with it would involve detailed taxonomic discus-
sion beyond the scope of this paper. Nevertheless, one instance of generic con-
fusion relates directly to Diphuicrinus and needs citation here. It stems from
Strimple’s (1961) proposal of the genus Graff hamicrinus , with Graff hamicrinus
acutus as the type species. This crinoid is an ornate form, bearing a few super-
ficial nodes on the cup plates and brachials; otherwise there is little to distin-
guish it from smooth-plated species of Delocrinus. Strimple founded his genus,
and distinguished it from Delocrinus, on the basis of surface ornament of vari-
ous types. Inasmuch as surface ornament alone, such as typifies Graff hamicrinus
acutus, is not regarded by most crinoid authorities as sufficient grounds for gen-
eric distinction, I have refused to recognize Gra ffha m icrinus as a valid genus
(Burke, 1966, 1970), relegating it to synonomy under Delocrinus, and continue
to advocate its suppression.

However, within the omnium-gatherum of species which Strimple in-
cluded under Gra ff hamicrinus there are some which in addition to surface orna-
ment, show diagnostic features which entitle them to recognition as representa-
tives of distinct genera. One of these was originally described as Delocrinus
aristatus by Strimple (1949). Probably this species, more than any other, has
been regarded as the connecting link between Diphuicrinus on the one hand
and Delocrinus ( = Gra ffhamicrinus ) on the other. Of this taxon Pabian and
Strimple (1974, p. 15) state: Graff ha micrinus aristatus appears to be a very

primitive representative of this genus. The radial facets, contour of the cup, and
attitude of the anal plate strongly suggest a relationship to Diphuicrinus Moore
and Plummer.” I have recently examined the holotype of this species, an incom-

J.J. BURKE

No. 23

plete dorsal cup, USNM S 4690, and find it a typical example of Diphuicrinus ;
it shows a steep-walled basal impression, tips of radials in the basal plane, and a
distinct forefacet. In lateral profile it resembles OU 6446, the specimen figured
by Strimple and Moore (1971, figs. 9-11) and identified as Diphuicrinus coalen-
sis ? Strimple and Moore. It also resembles OU 6445, the holotype of Diphui-
crinus coalensis, in showing a much reduced stem. I am herewith designating the
species Diphuicrinus aristatus (Strimple, 1949) comb. nov. If this species, or any
other species of Diphuicrinus, is assumed to be either closely related or ancestral
to Delocrinus ( = Graff hamicrinus) I fail to find evidence to support that as-
sumption.

REFERENCES CITED

Andrews, E. B. , 1870, Report of progress in the second district: Ohio Geol. Survey Rept. Prog. 1869,
p. 55-142.

Burdick, D. W., and Strimple, H. L., 1973, New late Mississippian crinoids from northern Arkan-
sas: Jour. Paleontology, V. 47, p. 231-243.

Burke, J. J., 1966, Endelocrinus kieri, a new crinoid from the Ames Limestone: Ohio Jour. Sci., V.
66, p. 459-464.

1970, Observations of the Pennsylvanian crinoid Endelocrinus armatura (Strimple):

Kirtlandia, No. 11, p. 1-10.

Moore, R. C., and Plummer, F. B., 1937, Upper Carboniferous crinoids from the Morrow sub-
series of Arkansas, Oklahoma and Texas: Denison Univ. Bull., V. 37, No. 20, p. 209-313.

Moore, R. C. and Strimple, H. L., 1973, Lower Pennsylvanian (Morrowan) crinoids from Arkan-
sas, Oklahoma, and Texas: Univ. Kansas Paleont. Contrib. Art. 60, p. 1-84.

Morningstar, H., 1922, Pottsville fauna of Ohio: Ohio Geol. Survey Bull. 25, p. 1-312.

Pabian, R. K. and Strimple, H. L., 1974, Miscellaneous Pennsylvanian crinoids from Kansas, Okla-
homa, and Nebraska in Fossil Crinoid Studies: Univ. Kansas Paleont. Contrib. Paper 73, Pt.
1, p. 2-19.

Strimple, H. L., 1949, Studies of Carboniferous crinoids: I. A group of Pennsylvanian crinoids
from the Ardmore Basin: Paleont. Americana, V. 3, No. 23, p. 5-22.

1961, Late Desmoinesian crinoids: Oklahoma Geol. Surv. Bull. 93, 189 p.

Strimple, H. L., and Knapp, W. D., 1966, Lower Pennsylvanian fauna from eastern Kentucky;
Part 2, Crinoids: Jour. Paleontology, V. 40, p.309-314.

Strimple, H. L., and Moore, R. C., 1971, The family Diphuicrinidae in Fossil Crinoid Studies:
Univ. Kansas Paleont. Contrib. Paper 56, Pt. 1, p. 2-9.

MANUSCRIPT SUBMITTED AUGUST 27, 1975

KIRTLANDIA

CLEVELAND, OHIO NUMBERS 24, 25, 26, 27

•NATURAL HISTORY*

KIRTLANDIA
David S. Brose, Editor

Kirtlandia is an occasional publication of The Cleveland Museum of Natural History and is
devoted to scientific papers in the various fields of inquiry within the Museum’s sphere of interest.
Published at least twice a year, issues will vary between collections of short papers and single
issue-length studies.

Kirtlandia is distributed by The Kent State University Press, Kent, Ohio 44242. Price $2.00 per
issue domestic, $2.50 foreign.

The appearance of the code at the bottom of the first page of an article in this serial indicates the
copyright owner’s consent that copies of the article may be made for personal or internal use, or for
the personal or internal use of specific clients. This consent is given on the condition, however, that
the copier pay the stated per-copy fee through the Copyright Clearance Center, Inc. for copying
beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not
extend to other kinds of copying, such as copying for general distribution, for advertising or
promotional purposes, for creating new collective works, or for resale.

Kirtlandia is abstracted in Zoological Record and Biological Abstracts, and indexed in Bibliog-
raphy and Index of Geology.

ISSN: 0075-6245

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO NUMBER 24

MAMMALS UTILIZED AS FOOD BY OWLS IN REFERENCE
TO THE LOCAL FAUNA OF NORTHEASTERN OHIO

RALPH W. DEXTER

Department of Biological Sciences
Kent State University

Abstract

Food habits of 6 species of owls living in northeastern Ohio were studied
between 1949 and 1969 by pellet and stomach analyses. Small mammals constituted
the bulk of the diet. In a sample of 1839 pellets of the Bam Owl from 5 counties,
96.86% of the food consisted of 3 species ( Microtus pennsylvanicus, 77.27%;
Blarina brevicauda, 16.94%; Peromyscus leucopus, 2.65%). Altogether, 14
species were utilized. These constituted 50% of the small mammal fauna of this
area. Limited data indicate that the Great Homed Owl, Barred Owl, Long-eared
Owl, and Screech Owl utilized essentially the same species, but the larger owls took
more cottontails than the smaller owls. A sample of 85 pellets of the Short-eared
Owl living at a city dump produced a different result, with 96.3% of the food
consisting of introduced pest species — the Norway Rat ( Rattus norvegicus, 75.9%)
and the House Mouse ( Mus musculus, 20.4%), with very little utilization of the
common wild species of small mammals.

Introduction

Food habits of owls have been studied in northeastern Ohio over a period
of 20 years by the analysis of disgorged pellets and stomach contents. For the 6
species of owls studied, small mammals constituted the bulk of the diet for
each. While many studies have been published on pellet analyses for Barn
Owls, this report gives stress to the relation of prey species to the available
mammalian fauna.

Publications of food habits of owls in the area of the present study have
been issued by Stupka (c. 1932) and Phillips (1951) for Ohio; by Price (1942)

0075-6245/78/1978-0024 $00.50/0

RALPH W. DEXTER

No. 24

for northwestern Ohio and northeastern Indiana; by Kirkpatrick and Conway
(1947) for Indiana; by Wilson (1938), Wallace (1948), and Reed (1959) for
Michigan; and Pearson and Pearson (1947) for Pennsylvania. The Wallace
report (1948) is the most complete for this area and contains an excellent
bibliography on the Barn Owl, including food habits.

The most intensive study for this report was made on the Barn Owl ( Tyto
alba), formerly a common species, but now becoming rare in the area. Wil-
liams (1950) classified the Bam Owl as a “not uncommon permanent resi-
dent.” Formerly, the writer and his students banded Barn Owl nestlings and
collected pellets in many bams of northeastern Ohio. In recent years, however,
no Barn Owls were reported to the Cleveland Bird Calendar from 1964 until 13
December 1967, when one was seen by B. P. Bole, Jr., at Kirtland Hills.
During the annual Christmas Bird Counts reported in Audubon Field Notes and
American Birds, only 2 records of Barn Owls are given for northeastern Ohio
since 1964. The exceptions were single birds at Burton and Mentor in January
1976.

Results

A total of 1839 pellets collected from 5 counties of northeastern Ohio
contained 5586 mammal skulls, which were identified through the keys of Katz
(1941) and Glass (1958). This averages approximately 3 skulls per pellet, the
same found by Wallace (1948) in his sample of 6742 pellets. Fourteen species
of small mammals were included in the sample (Table 1). The two most
common prey species, M. pennsylvanicus , the Common Field Mouse, and B.
brevicauda, the Short-tailed Shrew, together account for 94.21% of the prey.
The top 3 species account for nearly 97% of the food, while the remaining 1 1
species make up only slightly more than 3% of the food. The Meadow Mouse,
M. pennsylvanicus, was the chief item of food for the Barn Owl in this area.
This species comprised nearly 80% of Ohio pellets analyzed by Stupka (c.
1932) and a little over 85% reported by Phillips (1951). Both of these studies,
however, found B. brevicauda, the Short-tailed Shrew, comprised a little more
than 6%, in contrast to the present study, which found that it comprised nearly
17%. While Price (1942) found a similar utilization of the Meadow Mouse, he
found the Short-tailed Shrew made up only 4% of the food in Williams County,
Ohio, but he found Cryptotis parva, the Least Shrew, made up 27% as the
second most important food item in that area. Wilson (1948) found M. pennsyl-
vanicus and B. brevicauda to be the first and second most important in his study
in Michigan, although the latter was less important than in the present study for
northeastern Ohio. Wallace (1948) found a larger percentage for M. pennsyl-

1978

MAMMALS UTILIZED AS FOOD

hTtinooooooonmn-'t

NO\'Ooo'tl;^nHHHq

r^vdcMooooooooo

r- ^

r-~

OOOOhf'VO'tl'-'

■*t

(N -H o

1 — 1

ir>

"St

'— 1

co as as

b o o on

2 m N
co — '

UO — i

r-% ^

~ n in vo in in in

? k IO

vo -h ro

O O
<N — '

ro r-
CM

•3

to m n m O
■ CM CO

i ^ -

(N m N

rf co
os oo
00 CM

'q OO (N VO

os r-~
(N ©
t-~ CO

.Cj

’5

J -f
I!
II

S ■“

b a

^ -g

to to 0-
X "Q *«.

& s g

5 *« c

S' 05 5

i |*

*5-2 tu J3 -<3

< 03 a, N d;

2 to

a 3
~~
c a

5U Cj

•S §

* s
2: §
% s

s ■=> ,

C J 3

, > a, ^

2 35 -5

I £ s
S ^ I
S’ & £

RALPH W. DEXTER

No. 24

vanicus (84.97%), but a smaller percentage for B. brevicauda (6.53%), in his
study in Michigan, than reported here, but these 2 species made up the bulk of
the diet in Michigan as well as in Ohio.

Pellets and stomach contents of the Great Horned Owl ( Bubo virginianus).
Barred Owl (Strix varia ), Long-eared Owl (Asio otus ), and Screech Owl (Otus
asio) indicate utilization of the same common species of mammals used by the
Barn Owl. However, insufficient data were acquired to establish meaningful
ratios of prey. The only notable difference is the more common occurrence of
the Cottontail Rabbit in the diet of the larger species of owls.

A sample of 85 pellets of the Short-eared Owl (Asio flammeus ) collected 8
February 1956 from an owl living at the Cleveland dump demonstrated an
unusual relationship in the food of the owls. Synder and Hope (1938) found M.
pennsylvanicus to make up the vast bulk of the diet of the Short-eared Owl. The
House Mouse, M. musculus, represented only 0.1% of the prey in their study
made in the Toronto region. Hendrickson and Swan (1938) found the winter
food of this owl in Iowa to be almost entirely M. pennsylvanicus and P.
leucopus. Terres and Jameson (1943) also found M. pennsylvanicus to make up
the bulk (82. 17%) of the diet for this owl near Perry City in New York, while
M. musculus composed nearly 12%. Stegeman (1957) found M. pennsyl-
vanicus to make up 97.2% of the winter food while M. musculus accounted for
only 0.15% and R. norvegicus only 0.07% in central New York. Reed (1959)
found only M. pennsylvanicus (74.1%) andP. leucopus (25.9%) in the sample
he studied in Michigan. In this study there were 41 skulls (75.9%) of R.
norvegicus , and 1 1 skulls (20.4%) of M. musculus. There were only 2 skulls of
M. pennsylvanicus. This is a reversal of the usual ratio and is a reflection of the
specialized habitat of this particular owl. While most owls live in rural habitats,
this one lived in a city dump.

Discussion

Five orders of small mammals were utilized by the owls examined in this
study. Rodentia (5 species) and Insectivora (6 species) composed over 99% of
the food. Chiroptera, Lagomorpha, and Carnivora were each represented by a
single species.

Phillips ( 1 95 1 ) listed many local species of small mammals not utilized by
the Bam Owl in his area. Bole and Moulthrop (1942) recorded 20 species of
small mammals from northeastern Ohio. Only half of these were found in the
diets of owls studied here. The other half, however, are either uncommon
species for the most part, or they are chiefly diurnal in their activities. Dexter
(1955) recorded 19 species of small mammals on the Kent State University

1978

MAMMALS UTILIZED AS FOOD

campus in Portage County, Ohio. Again, only half of these were found in owl
pellets collected in that area, and those species not utilized are relatively
uncommon and, in some cases, diurnal in their activities.

Pearson and Pearson (1947) concluded that, “Neither owls nor trappers
catch a representative sample of the small mammal population.” Stegeman
(1957) also concluded that the prey found in pellets did not reflect the relative
abundance of the prey species in nature, and Weller et al. (1963) demonstrated
that “Mammal trapping in the owl roosting areas produced quite different data
on species composition of the mammalian fauna than was implied from the
remains in owl pellets.”

Apparently, owl pellet analysis cannot be used to give reliable data on the
relative abundance of local small mammals. It is clear from the present study,
however, that the great bulk of mammals utilized by local owls for food consists
of only 3 common species, with only 50% of local species in the area being
utilized.

Acknowledgements

Acknowledgement is made to my former students in ornithology who
assisted with the study, especially David Gerrick, who supplied a great many of
the pellets for analysis.

References

Bole, B. P., Jr. and P. N. Moulthrop, 1942, The Ohio recent mammal collection in the
Cleveland Museum of Natural History, Sci. Pub. Cleve. Mus. Nat. Hist. 5(6):
83-181.

Dexter, P. W., 1955, The vertebrate fauna on the campus of Kent State University,
Biologist 37: 84-88.

Glass, P., 1958, A key to the skulls of North American mammals. Private pub., 53 pp.

Hendrickson, G. O. andC. Swan, 1938, Winter notes on the Short-eared Owl, Ecology
19: 584-588.

Katz, D. T., 1941, A key to the mammals of Ohio, Ohio Wildlife Research Sta. Release
No. 158, 34 pp.

Kirkpatrick, C. M. and C. H. Conway, 1947, The winter foods of some Indiana owls,
Amer. Midi. Nat. 38: 755-766.

Pearson, O. P. and A. K. Pearson, 1947, Owl predation in Pennsylvania, with notes on
the small mammals of Delaware County, Jour. Mammal. 28: 137-147.

Phillips, R. S., 1951, Food of the barn owl Tyto alba pratincola, in Hancock County,
Ohio, Auk 68: 239-241.

Price, H. F., 1942, Contents of owl pellets, Amer. Midi. Nat. 28: 524-525.

Reed, S. A., 1959, An analysis of 111 pellets from the Short-eared Owl, Jack-Pine
Warbler 37: 19-23.

RALPH W. DEXTER

No. 24

Synder, L. L. and C. E. Hope, 1938, A predator-prey relationship between Short-eared
Owl and the Meadow Mouse, Wilson Bull. 50: 110-112.

Stegeman, L. C., 1957, Winter food of the Short-eared Owl in Central New York,
Amer. Midi. Nat. 57: 120-124.

Stupka, A., n.d. [c. 1932], The dietary habits of Bam Owls, Bull. Ohio Dept. Agric.,
Div. of Conser. Bur. of Sci. Research No. 6, 5. pp.

Terres, J. K. and E. W. Jameson, Jr., 1943, Plague of mice as food for Short-eared
Owls, Wilson Bull. 55: 131.

Wallace, G. J., 1948, The Barn Owl in Michigan. Its distribution, natural history, and
food habits, Mich. State Coll. Agric. Exper. Sta. Tech. Bull. No. 208, 61 pp.

Weller, M. W., L. H. Frederickson, and F. W. Kent, 1963, Small mammal prey of some
owls wintering in Iowa, Iowa State Jour. Sci. 38: 151-160.

Williams, A. B., 1950, Birds of the Cleveland region, Sci. Pub. Cleve. Mus. Nat. Hist.
(Vol. 10) 215 pp.

Wilson, K. A., 1938, Owl studies at Ann Arbor, Michigan, Auk 55: 187-197.

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO

NUMBER 25

LIPID DEPOSITION IN THE HOUSE SPARROW AND
RED- WINGED BLACKBIRD

TIMOTHY O. MATSON AND LARRY D. CALDWELL

Cleveland Museum of Natural History
Michigan State University

Abstract

In Central Michigan the mean fat to dry-lean weight ratios of juvenal and adult
House Sparrows and Red-winged Blackbirds ranged from 0.09 to 0.35 during the
summer and fall of 1973. The lipid index of adults was lowest during the breeding
season and highest in the fall. Female red-wings migrated by early September
shortly after completion of the postnuptial molt with no apparent increase in fatness.
In late autumn other red-wings began premigratory fattening and increased their fat
to dry-lean index from 0. 16 in late October to 0.35 in mid-November. The sparrow
fat index increased gradually from 0.09 early in the season to 0.17 in November.

Odum (1960), Caldwell et al. (1964), and Johnston (1966) have stated that
some avian species begin migration in a relatively lean state and gradually
increment their fat reserves through hyperphagia as they move toward their
spring breeding areas or fall wintering grounds. In contrast, nonmigratory
species presumably do not exhibit any appreciable vernal or autumnal increase
in body fats.

In the present study amounts of whole body lipid values were compared
between the migratory Red- winged Blackbird ( Agelaius phoeniceus ) and the
nonmigratory House Sparrow ( Passer domesticus). Whole body lipids were
measured through the summer and into the fall migratory season.

Methods

Collections of the two species were made in Isabella County, Michigan,
from May into December 1973 with a shotgun and number nine shot. The

0075-6245/78/1978-0025 $00.50/0

TIMOTHY O. MATSON AND LARRY D. CALDWELL

No. 25

blackbirds (N = 130) were taken from both upland and marshland habitats and
the sparrows (N = 88) were collected near human dwellings or farm outbuild-
ings. All specimens were weighed and then frozen until the time of analysis.

Skull ossification was used to separate birds of the year from older birds
(Nero, 1951; DeHaven et al., 1974). Each specimen was dissected, the sex
determined, and the gut and crop contents removed before fat extraction. The
fat extraction technique involved the use of a food blender with petroleum ether
and ethyl alcohol solvents and is described elsewhere (Matson and Caldwell,
Kirtlandia, No. 26).

Results

Fledged immature House Sparrows were of a smaller body mass than the
adults. The mean dry-lean weight of the fledglings at 6.5 ±0.12 (1 standard
error of the mean) grams was significantly smaller than the adults 7.61 ±0.06
grams (P<0.01). However, since total body fats averaged 0.96 grams for each
group, the immature birds had a higher fat to dry- lean ratio (0.15) than did the
adults (0.12) because of a smaller dry-lean body mass (P<0.05).

A fat to dry-lean comparison in the red-wing is more complex because this
species exhibits sexual dimorphism and premigratory fattening. The females
average 11.18 grams of dry-lean weight and were smaller than the males at
18.68 grams (P<0.01).

With regard to age differences, immature female red-wings were signifi-
cantly smaller in dry-lean body mass at 10. 17 grams than adult females ( 1 1 .49
grams, P<0.05), as would be expected. Likewise, immature males at 16.70
grams were smaller than the adult males at 19.34 grams (P<0.01). There was
likewise a tendency for immatures to be more obese than the adults. During the
summer and early fall months the 0. 18 fat index of immature red-wings of both
sexes combined was significantly higher than the 0. 14 value found in the adults
(P <0.01). Just prior to migration, the ratio of fat to dry lean of the immatures
rose to 0.29 but there was no difference from the ratio value of the adult birds at
0.27 (P>0.50).

Adult House Sparrows in the early part of the nesting season had a fat
index of 0.09 with a variance of only 0.0002 (Fig. 1). In the middle and later
parts of the nesting season the variance was much larger (S 2 = 0.003), but the
mean index (0. 12) had not risen significantly (P <0. 10). By late October, after
the postnuptial molt, the index had increased to 0.17 while the variance had
decreased to 0.0005. The postnuptial fat index was markedly higher than the
early nesting season index (P <0.001). The fat index of the immatures averaged
0.14 throughout most of the nesting season, but in early October it declined to
0.11 before increasing to the same value (0.17) as found in the adults in late
October.

o o © o

CSJ

1AJ

1VJ

X30NI

Fig. 1 . Changes in the fat index of adult and juvenal House Sparrows in Central Michigan during the
summer and fall of 1973.

IMMATURES

X30NI 1VJ

- o

J O H

CSi

J °

_ o

“9

Fig. 2. Changes in the fat index of adult and juvenal Red-winged Blackbirds on their breeding
grounds in Central Michigan during the summer and fall of 1973.

1978

LIPID DEPOSITION

The 0.09 fat index for adult red-wings during the June and early July
nesting season (Fig. 2) was almost identical with the index of the adult House
Sparrows ( P >0.50). Then by 15 August the red-wing index had risen to 0.14,
where it remained relatively constant until the start of the fall premigratory fat
build-up beginning about 27 October. The variance of the index throughout the
above mentioned interval was moderate (S2= 0.0009). Then, between 27
October and 1 1 November, the index rose rapidly fromO. 16 to 0.35. The linear
regression equation for the increase of fat in grams per day for adult Red-
winged Blackbirds during this period of rapid premigratory fattening is
Y =0.27X-36.78 (S2 = 0.009). In other words, 0.27 grams of fat were stored
per day for approximately 15 days, at which time the migrants departed.

TABLE 1

Comparison of Major Body Components of Juvenal and
Adult House Sparrows and Red- winged Blackbirds

Age

Mean Dry-lean
Weight ( Grams)

House Sparrow
Fat

( Grams)

Fat

Index

Water
( Grams)

Water

Index

Juvenal

6.5

1.0

0.2

20.1

3.1

Adult

7.6

1.0

Red-winged Blackbird

0.1

21.0

2.8

Juvenal

15.1

3.1

0.2

42.2

2.8

Adult

17.3

3.0

0.2

43.1

2.5

The data that have been analyzed in this study are linear rather than
discrete in nature. For summary purposes, however, a set of discrete data in
tabular form is valuable for reference (Table 1). Notice in Table 1 that water is
also a variable body component. The water index (water weight divided by
dry-lean weight) is lowest in the adults. The 3.1 index of immature sparrows is
higher than that of the immature red-wings (2.8) and other avian species for
which water indices have been determined. For example, Ricklefs (1967) found
the index of fledgling Bam Swallows to be about 2.3 and for Dickcissels it is
about 2.3 (Zimmerman, 1965). The value for adult House Sparrows (2.8) and
adult Red- winged Blackbirds (2.5) is quite close to the index of 2.6 in Long-
billed Marsh Wrens (Kale, 1965).

TIMOTHY O. MATSON AND LARRY D. CALDWELL

No. 25

Discussion

Both House Sparrows and adult Red-winged Blackbirds were relatively
lean with a low fat index (0.09) and low variance during the early part of the
nesting season (Figs. 1 and 2). The low index and variance in the adult is
probably related to a uniformly high expenditure of energy in nesting activities
once breeding and caring for the young begins. Once the first clutch has been
raised, however, the variance increases as some adults increase their lipid
reserves. Increased variance in the ratio of fat to dry-lean during the middle and
latter part of the House Sparrow nesting season should be expected since not all
adult birds breed the same number of times (Weaver, 1943) nor is the second
breeding as synchronized as the first breeding. Folk and Novotny (1970) in their
year-long study of the body weight of the House Sparrow noted that the body
weight decreased in May. They attributed this loss to parental care for the young
of the first clutches. In June, however, they found that the body weight
increased in accordance with the production of sexual products for the next
breeding. It seems that in addition to the increase in sexual products our data
indicate that some of the body weight increase is due to the increase in fat
reserves following the period of parental care for the young, especially in the
individuals that breed only once during the season. Similarly, the index of the
red-wings increased after the period of parental care ended in early July.

By late October, after the breeding season was finished and the postnuptial
molt had been completed, the ratio of fat to dry-lean of the House Sparrow was
again uniform. Folk and Novotny (1970) found that by October the body weight
of the male and female components of their sample had risen significantly over
that of the birds taken during the breeding season. They attributed this increase
in body weight to the increase of fat prior to the winter months. This study
indicates that their proposal is correct and that House Sparrows do indeed
increase their fat reserves prior to the winter season. Note that the 5 House
Sparrows collected in December were not involved in the data analysis or
interpretation since they were collected after the red-wings had departed for the
winter.

Molting requires energy and could result in a drop in the fat index if food
intake is rather low. The effect of the postnuptial molt on the fat index of the
adult House Sparrow is uncertain. Adults were found replacing flight feathers
from about the middle of August through the latter part of October. The data on
adults during this period are, however, too few to make any meaningful
analysis.

Adult red-wings were observed molting from late July into the early part of
October. A statistical comparison of the fat index between molting adults and

1978

LIPID DEPOSITION

nonmolting adults reveals no significant differences between the two groupings
(P<0.40). Unfortunately, only 4 adult birds were observed not molting during
this period. Likewise, the sample number (N = 2) of nonmolting immatures
was too small to warrant a comparison with molting immatures.

It should be noted that after 1 September only two female red-wings were
collected. Shortly after completing the postnuptial molt, both adult and imma-
ture females migrated to other areas. A comparison of the fat indices between
the last two females collected in September (0.19) with the only two females
collected in November (0.34) reveals that females found in the area late in the
fall had increased their fat reserves considerably over those leaving earlier. The
fat index of 0.34 found in the November females is comparable with the 0.27
value found in adult males during this premigratory fattening period. In other
words, most female red-wings begin some form of migration early in the fall or
late summer after completion of the postnuptial molt, and they do so without
increasing their fat reserves to any appreciable degree. Yet those few females
that are found in the area in the late fall increment their fat depots to levels equal
with those of males, which normally migrate in November.

During the premigratory fattening period that began around 27 October the
fat index of the Red-winged Blackbirds increased from 0.16 to 0.35, thus
doubling in value from the summer period. The index values of 0.29 and 0.27
for immature and adult male red-wings during premigratory fattening were not
significantly different (P>0.50). The equation expressing the relationship
between the fat index and time during premigration for both age groups is
Y = 0.01 X — 1.11, where Y is the fat index and X is the number of the collection
day based on 31 May as day one. In other words the ratio of fat to dry-lean
weight increased 0.01 each day from 27 October to 14 November. Thus, as
mentioned earlier, an average of 0.27 grams of fat were added to the lipid depot
each day throughout the premigratory fattening period. For comparison, Mor-
ton et al. (1973) estimated the rate of fattening of transient White-crowned
Sparrows ( Zonotrichia leucophrys gambelii ) as 0.3 grams per day. He also
estimated the rate of fattening of captive Zonotrichia leucophrys oriantha to be
about 0.1 grams per day for the few days following the molting period, and
preceding the autumn migration. From the data in Table 1 of Norris et al. (1957)
it is possible to calculate the rate of lipid accumulation of a transient flock of
Ruby-throated Hummingbirds during a period of intense preflight fattening.
The mean fat content increased from about 1.07 grams on 16 September to a
value of 2. 18 grams on 23 September. In other words, fat was deposited at a rate
of 0. 16 grams per day throughout a seven day period before the birds left the
area. Of course the Ruby-throated Hummingbird, by virtue of its small size,
does not need to accumulate as much fat as the red-wing. But for comparative

TIMOTHY O. MATSON AND LARRY D. CALDWELL

No. 25

purposes it is interesting to note that the red-wing, which has a dry-lean weight
18 times that of a Ruby-throated Hummingbird, adds almost twice as much fat
per day. It is important to realize that the premigratory fattening period of the
red-wing lasts about 16 days in contrast to the amazingly short time of about 7
days in the transient Ruby-throated Hummingbird population studied by Nor-
ris.

References

Caldwell, L. D., E. P. Odum, and S. G. Marshall, 1964, Comparison of fat levels in
migrating birds killed at a Central Michigan and a Florida Gulf Coast television
tower, Wilson Bull., 75:428-434.

DeHaven, R. W., F. T. Crase, andM. R. Miller, 1974, Aging Tricolored Blackbirds by
cranial ossification, Bird-Banding, 45:156-159.

Folk, C. , and I. Novotny, 1 970, Variations in body weight and wing length in the House
Sparrow, Passer domesticus L., in the course of a year, Zoologicke Listy,
19:333-342.

Johnston, D. W., 1966, A review ofthe vernal fat deposition picture in overland migrant
birds, Bird-Banding, 37:172-183.

Kale, H. W., II, Bioenergetics of the Long-billed Marsh Wren, Telmatodytes palustris
griseus, in a salt marsh ecosystem, Diss. Abstr. 25, 1964:3173.

Matson, T. O. and L. D. Caldwell, 1976, Lipid deposition in nestlings of the House
Sparrow and Red-winged Blackbird. Kirtlandia, No. 26.

Morton, M. L., J. L. Horstman, and C. Carey, 1973, Body weights and lipids of
summering mountain White-throated Sparrows in California, Auk, 90:83-93.

Nero, R. W. , 1951, Pattern and rate of cranial ossification in the House Sparrow, Wilson
Bull., 63:85-88.

Norris, R. A., C. E. Connell, and D. W. Johnston, 1957, Notes on fall plumages,
weights, and fat condition in the Ruby-throated Hummingbird, Wilson Bull.,
69:155-163.

Odum, E. P., 1960, Lipid deposition in nocturnal migrant birds. Proc. 12th Intern.
Ornithol. Congr., 563-567.

Odum, E. P., 1960, Premigratory hyperphagia in birds, Amer. Jour, of Clinical
Nutrition, 8:621-627.

Ricklefs, R. E., 1967,, Relative growth, body constituents, and energy content of
nestling Barn Swallows and Red-winged Blackbirds, Auk, 84:560-570.

Weaver, R. L., 1943, Reproduction in English Sparrows, Auk, 60:62-73.

Zimmerman, J. L., 1965, Carcass analysis of wild and thermal-stressed Dickcissels,
Wilson Bull. 77:55-70.

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO NUMBER 26

LIPID DEPOSITION IN NESTLINGS OF THE HOUSE
SPARROW AND RED- WINGED BLACKBIRD

TIMOTHY O. MATSON AND LARRY D. CALDWELL

Cleveland Museum of Natural History
Michigan State University

Abstract

It is well known that lipid depots serve as an energy source for birds during
migratory flight. Furthermore, studies by Wolfson (1954), Odum and Connell
(1956), King and Famer( 1959) and others have shown that lipid volumes vary from
one phase of the annual life cycle to the next. Little of the work on avian lipids to
date, however, has been concerned with nestlings. According to Brenner (1964),
nestling Red-winged Blackbirds ( Agelaius phoeniceus ) deposit about 75 mgms per
day. Brenner’s measurements were made from body organ sections and the major
lipid depots.

In the spring and summer of 1973 nestlings of the House Sparrow (j Passer
domesticus) and Red- winged Blackbird were collected in Isabella County of central
Michigan. The House Sparrow as a nonmigrant serves as a basis of comparison with
the Red-wing which is a medium-range migrant. The Red-wing winters in the
United States as far south as Florida and Texas.

Methods

An entire clutch of nestlings was removed at each nest site rather than
removing part of a clutch and thus increasing the quantity of nutrient available
to the remaining nestlings. We collected 62 House Sparrows and46 Red-wings.
Weights were taken after collection and specimens were frozen in plastic bags
until the time of analysis. Nestlings were aged to the nearest day by using the
weight and feather tract development criteria of Weaver (1942) and Williams
(1940). Each specimen was dissected, the sex determined, and the gut and crop
contents removed before fat extraction.

0075-6245/78/1978-0026 $00.50/0

TIMOTHY O. MATSON AND LARRY D. CALDWELL

No. 26

The fat extraction technique involved the use of a food blender with
petroleum ether and ethyl alcohol solvents. Each specimen was covered with
alcohol and then macerated in the blender for 45 seconds. Petroleum ether was
used to wash the residue into a beaker, which was then placed over a steam bath
for several minutes. The two-solvent mixture was subsequently filtered through
a strainer and filter paper into a separatory funnel. The residue that was trapped
in the strainer was again extracted with petroleum ether in the beaker, heated,
and filtered through the filtering sy stem a second time . A total of 1 0-22 volumes
of solvent was used relative to body mass for the extraction process. A biphasic
system subsequently resulted upon acidification of the two-solvent mixture
(Cratin, 1970). The ether phase, containing the fat, was drawn off, whereas the
alcohol phase was washed 2-3 times with petroleum ether. All ether phases for
each specimen were finally pooled and evaporated to near dryness. The concen-
trated fat was then transferred to a pre weighed aluminum pan for drying over a
steam bath . Final drying was achieved in an oven at 1 1 5° C for 1 2- 1 8 hours . The
dried weights of the residue on the filter paper and the residue on the strainer
combined represents the dry-lean weight or nonfat dry weight. The method of
Cratin (1970) when properly applied gives fat values virtually identical to
results obtained by the popular Soxhlet method of fat extraction.

Results

Nestling sparrows exhibited a linear relationship between fat and dry-lean
weight as expressed by the equation Y = 0. 18X -0.03 (Fig. 1 , line A). In other
words, 0.18 grams of fat were added for each 1.00 gram of dry-lean weight
increase in body mass. However, this line of best fit is separable into two
stages, with a tendency for the smaller nestlings (Fig. 1, line B) to accumulate
fat more slowly than the larger nestlings (Fig. 1, C) which have a dry-lean
weight of 2.49 grams or greater (P<0.01 between slopes). Small sparrows
exhibited much less variance about the regression than did the large individuals
(0.001 and 0.007 respectively). Notice that 2 nestlings in the lower right-hand
corner of the figure were very lean when compared to all other nestlings of a
similar dry-lean weight. The above 2 nestlings were not included in the
calculations of the regression line. Each of the 2 lean nestlings came from a
different clutch.

In contrast, the blackbird nestlings were considerably more variable in the
rate of fat increment with respect to dry-lean weight (Fig. 1 , line D). In spite of
the large overall variance in nestling blackbirds, the smallest individuals of both
species were very similar. The eight smallest Red-wings ( X = 0.67 grams) had
0.06 grams of fat. Accordingly, the fat/dry-lean ratio was 0.09 and was

1978

LIPID DEPOSITION IN NESTLINGS

(s muo) ivd

Fig. 1. Changes in the fat content of nestling House Sparrows and Red-winged Blackbirds as a
function of dry-lean weight. The equation of the regression line describing all the House Sparrow
data (line A) is Y = Q.18X-0.03. The equation for nestling sparrows with a dry-lean weight of less
than 2.49 grams (line B) is Y— 0.12X-0.00, whereas the relationship of nestlings 2.49 grams and
more in weight (line C) is given by the equation Y — . 17X-0.01 . The equation of the regression line
describing nestling Red-winged Blackbirds (line D) is Y = 0.12X-0.04.

identical to that of the 8 smallest sparrows. The blackbird variance from the
regression increased rapidly to a maximum in the 6.00 to 7.50 gram weight
class, at which point body fat ranged from 0.58 to 1 . 15 grams. The overall fat

FAT INDEX

4 TIMOTHY O. MATSON AND LARRY D. CALDWELL No. 26

Fig. 2. Changes in the fat index of nestling House Sparrows and Red-winged Blackbirds as a
function of age. The equation of the regression line describing the House Sparrow is Y—
0.01 X+ 0.10, whereas the equation describing the blackbird is 7= 0.01X + 0.08.

index equation for the blackbird is 0. 12X + 0.02 with a variance of 0.02.
Thus, sparrow nestlings with a regression coefficient of 0.18 had one-third
more lipid per gram of body mass than did the blackbirds at 0. 12. An analysis of
covariance between species regression lines reveals a significant difference
between slopes (i.e., fat index, P<0.01).

Our fat index versus age-in-days analysis shows a large scatter of data.
This variance is partly related to the difficulty of aging nestlings more accu-
rately than to within a 24 hour interval (Figure 2). The 0.002 variance for
nestling sparrows during the first 6 days was twice as large as the 0. 00 1 variance
of the older individuals. By contrast, the blackbird nestlings exhibit no correla-

1978

LIPID DEPOSITION IN NESTLINGS

C/3

Cl,

C3
CX
GO

3
O

* o

llh Vh

o o

C OX)
O a)

f 1

>, O

T3 C/3

O OX)

PQ .g

lh T3
O ™
•rr> <^>

£ Z

^ T3

« .fa

-S x)

n W
O 03

+2 5

J3 _,

.SP12

<L> OX)

£ .s

«+H -o

O G
C «*

c/i

r-~ m
U-i Tt

t m to
ON VO Tf

to ON

o 6 r-

1—1

3/ >— i (N CN

I odd

~£s © o o

03 £

s; a

J! O

Q .op

cn 00
— cn d

omo m
© c-i to

NO On

UO . ^
dn NO -H

TIMOTHY O. MATSON AND LARRY D. CALDWELL

No. 26

tion (P <0.10) between fat index and age. In other words, the ratio of fat to
dry-lean changed little during development. Variance in fat index is, however,
large at all stages of blackbird nestling development.

Water is also a dynamic body component of developing nestlings (Table
1). As would be expected, the water content of both species increased with age
up to the time of fledging. A water index (water content divided by the dry-lean
weight), however, is often used to express body water content. Since the
dry-lean mass increases at a high rate in developing birds, one can observe in
Table 1 that the water index declines with age of the individual. In terms of total
body mass, the water content of the sparrow and blackbird decreased from 87
and 89 percent respectively down to 79 and 80 percent. Body water of adult
birds would likely approach 66 percent.

Discussion

The fat to dry-lean ratio in nestling sparrows just prior to fledging was
almost one-half higher than the ratio in Red-wing nestlings. The difference in
fat between nestlings is probably related to the different time intervals spent in
the nest by each species plus the size difference between species. House
Sparrows fledge at about 14.4 days of age (Weaver, 1942). Body mass in-
creases from 2.8 grams at hatching to about 26 grams at fledging for a growth
rate of 1 .6 grams per day. In contrast, female and male Red-wings respectively
spend only 9.2 and 9.7 days in the nest (Holcomb and Twiest, 1970). The
growth rate for females is 2.8 grams per day, whereas males grow at 3 .0 grams
per day. Body mass for the two sexes respectively increases from 5.0 grams at
hatching (Williams, 1940) to 33 and 35 grams at fledging (Holcomb and
Twiest, 1970). Thus both sexes of the blackbird achieve a larger body mass
increment than the sparrows during a shorter time span (almost a twofold
difference in growth rates). Since growth of the Red-wing nestling is relatively
fast, it seems reasonable to find fat storage to be minimal. Energy intake in the
blackbird is evidently channeled more into skeletal and nonfat tissue growth . In
contrast, excess lipid deposition can occur in the nestling House Sparrow
apparently because of its lesser energy demand for tissue growth.

The variance in the ratio of fat to dry-lean weight for the nestling House
Sparrows (Fig. 1) was much lower for the nestlings under 2.49 grams than for
those nestlings above that amount. Weaver (1942) suggests that as fledging
approaches, the larger nestlings are found high in the nest and accordingly
receive more food than do any smaller siblings. Since competition for food
between siblings probably increases with age, it follows that some individuals
receive more food than do others. As a result, some siblings would be able to

1978

LIPID DEPOSITION IN NESTLINGS

store more fat (higher fat index), thus possibly accounting for the increased
variance observed in the larger nestling size cohorts.

Although the data on Red-wings are few in the 5.00 to 7.00 gram dry-lean
weight category (Fig. 1), it appears that the rate of fat deposition increases just
prior to fledging, thus making the regression line somewhat curvilinear. As
mentioned earlier, Brenner (1964) employed different methods on nestling
Red-wings and obtained results similar to those of the present study. According
to Brenner, the amount of fat deposited per gram of body weight decreased up to
day seven and then increased rapidly until the time of fledging.

The large variance encountered in the fat index of nestling House Spar-
rows 6 days and under in age was undoubtedly due to the difficulty in correctly
aging individuals to within a 24-hour period. Weaver ( 1943) found that the first
2 or 3 eggs in a clutch were incubated for about the same length of time. The
other eggs usually hatched 1 2 to 24 hours later but always within 48 hours. With
such a large time interval between the hatching of eggs, the first nestlings to
hatch have a competitive developmental advantage over their siblings with re-
spect to begging food from the parents. The nestlings hatching later may receive
enough food to sustain life and increase their body mass, but may not receive
enough food to store much energy in fat depots. As a result, the fat index of the
nestlings to hatch last would be lower than that in the nestlings to hatch first, and
hence one observes a large variance in the data. It appears that the nestlings
hatching later, however, catch up with their older siblings in fatness during the
last half of the nestling period. This probably occurs because the rapid increase
in body mass appears to slow after about 6 days (Table 1). Slowing of the
growth rate of the older siblings would allow the younger nestlings to receive
more food and deposit more fat, thereby achieving about the same fat index as
their older siblings.

Fat index did not increase with age for red-wing nestlings, whereas the
sparrow regression was quite significant (P<0.01). Ricklefs (1967), while
comparing the lipid index of nestling Barn Swallows to that of nestling Red-
winged Blackbirds, was puzzled by the low levels of lipid reserves found in the
nestling blackbirds. It seems likely that the low lipid levels observed by
Ricklefs as well as those observed in the present study were due to the rapid
increase in nonfat body mass. In Table 1 the index for sparrows increased from
0.08 to 0. 16, whereas the blackbird fat index increased from 0.08 to only 0.11.
As mentioned previously, nestling Red-winged Blackbirds display a very high
growth rate, and, therefore, do not accumulate fat to any appreciable extent. If
the energy intake is just adequate to promote tissue growth, then the ratio of fat
to metabolic body mass will likely remain the same with time. Any excess
energy results in fat storage and reflects either large food energy sources or only

TIMOTHY O. MATSON AND LARRY D. CALDWELL

No. 26

moderate energy demands for tissue growth or a combination of these two
factors. From Figure 2 it is evident that the Red-wing is of the first category with
tissue growth being the highest metabolic priority, and the House Sparrow,
with an increase in fat index with age, is of the second category, where surplus
energy for fat storage is available.

References

Brenner, F. J., 1964, Growth, fat deposition, and development of endothermy in
nestling Red-winged Blackbirds, Jour, of Sci. Laboratories, 46: 81-89.

Connell, C. E., E. P. Odum, andH. Kale, 1960, Fat-free weights of birds, Auk, 77: 1-9.

Cratin, P. D., 1970, Interfacial behavior of asphaltenes. Highway Research Record,
340: 29-37.

DeHaven, R. W., F. T. Crase, and M. R. Miller, 1974, Aging Tricolored Blackbirds by
cranial ossification, Bird-Banding, 45: 156-159.

Helms, C. W., W. H. Aussiker, E. B. Bower, and S. D. Fretwell, 1967, A biometric
study of major body components of the Slate-colored Junco, Junco hy emails,
Condor, 69: 560-578.

Hicks, D. L., 1967, Adipose tissue composition and cell size in fall migratory thrushes
(Turdidae), Condor, 69: 387-399.

Holcomb, C., and G. Twiest, 1970, Growth rates and sex ratios of Red-winged
Blackbird nestlings, Wilson Bull., 82: 294-303.

Nero, R. W., 1951, Pattern and rate of cranial ossification in the House Sparrow, Wilson
Bull., 63: 85-88.

Odum, E. P., D. T. Rogers, and D. L. Hicks, 1964, Homeostasis of the nonfat
components of migrating birds, Science, 143: 1037-1039.

Quigley, P., Methods of fat extraction, Unfinished M. S. thesis, Central Michigan
University.

Ricklefs, R. E., 1967, Relative growth, body constituents, and energy content of
nestling Barn Swallows and Red-winged Blackbirds, Auk, 84: 560-570.

Stewart, R. M., 1972, The reliability of aging some fall migrants by skull pneumatiza-
tion, Bird-Banding, 43: 9-14.

Weaver, R. L. , 1942, Growth and development of English Sparrows, Wilson Bull., 54:
183-191.

Weaver, R. L., 1943, Reproduction in English Sparrows, Auk, 60: 62-73.

Williams, J. F., 1940, The sex ratio in nestling Eastern Red-wings, Wilson Bull. 52:
267-277.

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO NUMBER 27

BENTHIC RECOLONIZATION PATTERNS IN THE
VERMILION RIVER, OHIO

MICHAEL J. S. TEVESZ

Department of Geological Sciences
The Cleveland State University

Abstract

Two recolonization studies were performed in the Vermilion River, Ohio, by
planting invertebrate- free sedimentary “islands’ ’ in the substratum and then collect-
ing them at predetermined intervals. Every macroinvertebrate taxon found living on
the undisturbed sedimentary bottom appeared on the islands by the 5th week in the
first study and by the 24th day in the second study.

Drift was likely an important source of organisms in the early stages of
recolonization. Nevertheless, the taxonomic composition of the sedimentary islands
was not entirely predicted by the total composition of the drift fauna.

Compared to the marine environment, the pattern of recolonization in the
Vermilion River is not established mainly through reproductive events but rather by
the relative mobility and abundance of the organisms involved. Also, the time
involved in complete recolonization is much less for the Vermilion River. This rapid
recolonization ability of lotic benthos helps insure survival of the various species by
aiding their wide dispersal within their habitat.

Introduction

By a series of experiments in which he observed the recolonization by
benthic macroinvertebrates of sedimentary “islands” planted on the sublittoral
sediments of Long Island Sound, McCall (1977) conclusively demonstrated
that studying recolonizing processes in aqueous environments can provide
important insights into the structure and dynamics of benthic communities.
Furthermore, he showed that such experiments can also provide information
regarding the patterns and rates by which benthic communities recover from
local disasters such as pollution events.

0075-6245/78/1978-0027 $00.85/0

MICHAEL J. S. TEVESZ

No. 27

Like McCall’s, most colonization or recolonization studies which are
based in aquatic environments are generally concerned with macrobenthos
(particularly invertebrates) and involve the marine realm. A summary of the
scope and significance of many of these marine-based studies is currently being
prepared (McCall et al., in prepr.). By contrast, there are noticeably fewer
macrobenthos colonization or recolonization studies for freshwater envi-
ronments. Those involving lentic environments are often concerned with
changes in species composition, richness, and abundance that occur when a
new lake is created or an existing dry lake refilled (e.g., McLachlan and
McLachlan, 1971; Paterson and Fernando, 1969; McLachlan, 1975). Experi-
ments based in lotic environments generally fall into one of the following three
categories: 1) the recovery of benthos after pollution abatement (e.g., Brink-
hurst, 1965; Crisp and Gledhill, 1970); 2) the colonization of introduced
artificial or natural substrata as a means of benthic sampling (e.g. , Mason et al. ,
1970; Coleman and Hynes, 1970; Glime and Cleman, 1972); and 3) the sources
(i.e., by drift, upstream migration, etc.) of recolonizing organisms (Waters,
1964; Williams and Hynes, 1976). A single study involving macroinverte-
brates was concerned with colonization patterns on artificial substrata in
relation to the Mac Arthur- Wilson equilibrium model (Dickson and Cairns,
1972; see also Mac Arthur and Wilson, 1963). Lotic environments in general
are not extensively studied in terms of macrobenthic recolonization.

The main purpose of this paper is to provide new information concerning
little-known aspects of recolonizing processes involving lotic macroinverteb-
rates, including the sequence of appearance of different organisms during
recolonization, the relative efficiency of various organism sources with respect
to rates of recolonization, the time involved in a complete recolonization cycle,
and the way in which the overall aspect of lotic recolonization is comparable to
recolonization patterns in other environments. It is hoped that this information
will not only provide a more comprehensive view of recolonization processes in
aquatic systems in general, but will also provide a more complete understand-
ing of how rapidly river beds may recover biologically from local disasters.
Additionally, this paper presents the first published account of the dynamics of
benthos of the Vermilion River, Ohio. The Vermilion is one of several similar,
little-studied northern Ohio rivers that empty into Lake Erie. Information for
this study was collected by the author during July-September 1975 from the
Vermilion River, Ohio.

Study Area

The Vermilion River arises in the community of Bailey Lake (Ashland
Co.), Ohio, where it is the main outlet of Mud Lake (40° 57 'N, 82° 21 'W).

1978

BENTHIC RECOLONIZATION

According to the Ohio Division of Water (1954), the river is 94.4 km long and
drains an area of 703.4 km2. The elevation at the source is 313.9 m, and the
average fall is 1 .5 m/km (all units here converted to the metric). The mouth of
the river is in the town of Vermilion (Erie Co.), Ohio, where the river empties
into Lake Erie. By interpolating figures received from the National Weather
Service (personal communication) it is reasonable to estimate that the drainage
basin of the Vermilion River receives about 86 cm of rainfall annually and that
the average yearly temperature for the same area is approximately 10.4°C.

Excepting the last 5-6 km of the river, where it flows through suburban and
urban areas and is in places noticeably polluted, all the upper reaches of the
Vermilion flow through rural countryside where the chief pollutants are fine
clastic sediments and fertilizers derived from agricultural operations. Inferring
from the presence in the river of a variety of pollution intolerant organisms
(e.g., diverse Plecoptera, Ephemeroptera, Trichoptera; Gaufin and Tarzwell,
1952, 1956; Beck, 1954), most of the upper reaches of the river are fairly clean.

The Vermilion’s bed is a complex mosaic of sediments. Nevertheless, in
the upper reaches of the river, the bottom is usually composed of one of two
broad sedimentary suites. Where currents are more rapid, the bottom is largely
composed of gravels consisting of shale, siltstone, or sandstone clasts that
overlie coarse to fine sands and muds. Where currents are slower, the pre-
dominating substratum often consists of sandy muds. Species richness is
usually greater on the gravelly substrata.

A small area of the river ( —5 x 15 m) located —16.5 km (river distance)
from the mouth and —0.6 km south of the end of Banks Road was selected for
its remoteness from human disturbance and homogenous sedimentary bottom.
This section has the gravelly bottom characteristic of large sections of the river
and moreover, is about as rich in benthic macroinvertebrate species for any
particular time of year as any other sampled area of the river (excluding riffles).
During July-September 1975 the observed bottom water temperature in this
area ranged from 23 to 27°C, water depth varied from 30 to 70 cm, and current
velocity ranged from 4 to 18 cm/ sec. Two successive recolonization experi-
ments were subsequently performed at this site (Fig. 1.).

Materials and Methods

The river was sampled in three ways. Grab samples were taken by
inserting into the river bottom a plastic container open on one end. The
container was placed open-end-downward and pushed vertically into the sedi-
ment until it was full. Then the surrounding sediment was dug away from the
sides and bottom of the container, a cap was placed over the open end, and the

MICHAEL J. S. TEVESZ

No. 27

container was removed from the sediment. The open end of the container
measured 10 cm x 10 cm, and its height was 14 cm. Thus each “grab” sampled
about 100 cm2 of bottom to a depth of 14 cm (slight flexing of the sides and
bottom of the container during sampling make these figures close approxima-
tions).

Williams and Hynes (1974) show that numerous macroinvertebrates occur
to depths of about 30 cm within the substrata of certain streams, with maximum
organism densities occuring at 10 cm. Additionally they report a few species
living at depths of up to 80 cm. Deep excavations into the substratum of the
Vermilion River showed that in June -September 1975 most macroinvertebrates
occurring in the study area were living in the upper 5-10 cm of the substratum.
While oligochaetes were occasionally found below this level to depths of 20 cm
or more, aquarium studies showed that they usually penetrated to this depth as
an avoidance response to physical disturbance, and were not characteristically

1978

BENTHIC RECOLONIZATION

living there. Since the grab samples penetrated to a depth of 14 cm, they were
likely an adequate means of determining for the area they covered the abun-
dance and taxonomic composition of most macroinvertebrate taxa excluding
oligochaetes.

Recolonization samples were obtained by filling containers of the dimen-
sions mentioned above with river sediments from the experiment site. These
sediments had previously been dried, treated with boiling water, and dried
again to remove all macroinvertebrates and their eggs and resting stages. The
containers were capped and placed in the river in an area of substratum
upstream from any previous sampling disturbance. The containers were sunk in
the bottom until the lip was nearly flush with the river bed. By extending
slightly above the substratum, the containers represented a small positive area
on the bottom even when covered by subsequently deposited sediments. This
facilitated locating the samples. The lids were then removed and the individual
containers collected at predetermined intervals. Covers were placed on the
containers just prior to sampling.

Removing, treating, and replanting the substratum obviously altered such
properties as porosity, organic content, and probably grain size. Nevertheless,
the gross physical aspects of the substratum remained the same as that of the
river bed, save that it lacked macrofauna.

Rapid sedimentary accumulation (within 1-2 days) up to and over the
edges of the containers meant that they were susceptible to recolonization from
the four main recolonizing avenues employed by river organisms (upstream
migration within the water, drift, within substratum migration, and oviposition
[Williams and Hynes, 1976]).

Drift samples were obtained by attaching a net (7 threads/cm) with an
opening of 32 cm to an iron support and placing it for 24 hours in the river with
the open end facing upstream. The net was elevated appriximately 4 cm off the
bottom to exclude organisms crawling along the substratum.

Fouling of the drift net by sediments, leaves, and wood frequently oc-
curred during the sampling interval and caused resistance to water flow through
the net. This impedence to flow sometimes caused eddying near the mouth of
the net, which may have lowered sampling efficiency. Thus the drift samples
likely provide only a very rough estimate of the relative abundance of drifting
organisms and may have excluded rare species. Although the net material
consisted of 7 threads/cm, the material was fastened to the net in a double layer,
and organisms less than Vi mm in length were occasionally retained.

Approximately forty grab samples were taken in late June 1975 near the
experimental sites to determine the taxonomic composition of the bottom in this
section of the river. Analysis of these samples showed that over 90% of the

MICHAEL J. S. TEVESZ

No. 27

species sampled occurred in any given set of four samples. Also, the relative
abundance of organisms in these four samples correspond to the relative
abundance in all samples. So for the purposes of the experiment, the contents of
four grab samples were considered representative of the fauna for this small
area of river (see also Cain, 1938). Similarly, the contents of four simultane-
ously collected recolonization samples were considered representative of how
this area would appear at a given instant of time after defaunation.

All benthic macro in vertebrates ( = size > 1 mm) were picked by hand
from all samples within 12 hours of collection. The organisms were preserved
in 70% ethanol.

Two consecutive recolonization experiments were run in the following
manner: For the first experiment, 20 recolonization sample “islands” were
planted in the river bed on 2 J uly 1975 . The original plan was to collect 4 boxes
at a time, once a week, for 5 consecutive weeks. However, problems in locating
the islands on the 3rd week of the experiment extended the experiment an extra
week. The final set of 4 recolonization samples was collected on 13 August.
The 4 recolonization samples were always selected at random. One drift sample
was also taken simultaneously with each set of 4 recolonization samples.
Experimentation with multiple drift samples revealed no noticeable differences
in taxonomic composition or relative abundance of organisms among different
samples. Two sets of 4 grab samples of the surrounding natural bottom were
also taken during the course of the 5 week experiment.

A second experiment was initiated on 13 August when 32 defaunated
sediment samples were planted in the river upstream and laterally displaced
from the previous experiment site. Four recolonization samples were taken
simultaneously every 3rd day. One drift sample was also taken at each sampling
interval. Two sets of 4 grab samples were also taken during this time. The final
set of 4 recolonization samples was collected on 6 September 1975.

Results

The recolonization samples from both experiments (Tables 1 , 2) show that
the earliest colonizers were mostly immature Insecta. In the first experiment, all
insect taxa found in the grab samples were present in the recolonization samples
by the end of the first week (Tables 1 , 3). In the second experiment, 4 of the 5
insect taxa found in the grab samples were also found in the recolonization
samples taken 3 days after the beginning of the experiment (Tables 2, 4).
Besides the immature insects, the gastropod Oxytrema was the only other taxon
to colonize the samples during the first week. All these early colonizers were
represented during the course of the experiments in the drift samples (see also
Tables 5,6).

Description of Recolonization Samples for Experiment I

1978

BENTHIC RECOLONIZATION

O — <

3 *
60 > *

^ 6 ^

|-S

6
eu

2 8

43 co

t; «
< ~

o 2

D-.S
P K

C3 55

u o
u .is

p , 43

X U

2-2 173 S jy

g J § | §■ § S^-Sts * &s.s F

* § s § &•§! 1 s-»| s 2

xs S
o S
cx x
p &

S <3 ?*

1°

* = Present in drift samples taken simultaneously.

** = No samples collected during week 3 of experiment.
All Insecta represented as aquatic immatures.

Organism numbers = #individuals/400 cm2

Description of Recolonization Samples for Experiment II

No. 27

MICHAEL J. S. TEVESZ

— < co

— <N — i

_ * * _
°

* * _ *

o o © o

^ &

^ 00
Qh .£

fl
li

Co ^
00

_ * * *

0 — I o OC

Ph O O — < Tt

00 <N

O O CO

03
T3

O - «

O O rv-G

b <o O
t; c Q
< ~

2 .^2

<u c

H CV

^ 5;
6^

C <D
<D 03
00 T3

Oh 0)
<D 03

03 -53
T3 Cj
03 O
0) ,0c

O o

£> 03 C

OU.^3-

Key: * denotes taxa also found in corresponding drift sample.

Organism numbers = #individuals/400 cm2

1978

BENTHIC RECOLONIZATION

TABLE 3

Description of Grab Samples for Experiment I

Taxa

Arthropoda

Insecta

Diptera

Chironomidae
Coleoptera
Stenelmis
' Plecoptera
Acroneuria
Megaloptera
Sialis

Ephemeroptera

Ephemera

Annelida

Oligochaeta

Mollusca

Gastropoda

Oxytrema

Ferrissia

Bivalvia

Sphaerium

Sampling Week

All Insecta represented as aquatic immatures.
Organism numbers = # individuals/400cm2

3
2
2
9

The last organisms to appear in the recolonization samples were
oligochaetes, bivalves, and the gastropod Ferrissia. In the first experiment,
oligochaetes colonized the samples during the interval between the 2nd and 4th
week, and bivalves appeared by the 5th week. In the second experiment, the
oligochaetes and Ferrissia appeared between the sampling times of the 21st and
24th days. None of these organisms ever occurred in the drift samples. Al-
though all taxa from the grab samples eventually appeared in the recolonization
samples, these 2 kinds of samples were never identical taxonomically for any
sampling period. Nevertheless, recolonization samples always grew biologi-
cally to resemble most closely the grab samples instead of the drift samples, and
organisms that occurred abundantly in drift samples were sometimes uncom-
mon or absent entirely from the recolonization samples (Compare Tables 1,2,
to Tables 5, 6).

The pattern of colonization with respect to the overall abundance of
organisms was marked by large fluctuations (Figs. 2,3). Both for the abun-
dance pattern as a whole and with respect to any particular taxon, there were no

MICHAEL J. S. TEVESZ

No. 27

Week

Fig. 2. Organism abundance fluctuations for the first experiment. The dashed line represents the
summation of the average number of individuals per taxon collected from the grab samples.

long, uniform trends. Also, peak abundances were often reached, for each
taxon and collectively, prior to the last sampling interval.

Although not all sampled organisms were measured, a qualitative assess-
ment of their size patterns showed that this often varied widely for particular
taxa within each sample and among samples taken at different times. Hence
there was no discemable pattern of organism size (or age) with respect to the
sequence of recolonization.

Discussion

One of the main distinctions between the early and late recolonizers is that
the early recolonizers are drift-prone and the later recolonizers are not. Thus
drift is likely an important mechanism in the early recolonization process in the

1978

BENTHIC RECOLONIZATION

Sampling period

Fig. 3. Organism abundance fluctuations for the second experiment. The dashed line represents the
summation of the average number of individuals per taxon collected from the grab samples.

Vermilion River. This inference is supported by Waters (1964), who shows
drift to be responsible for rapid recolonization by Baetis vagans (mayfly) and
Gammarus limnaeus (amphipod) in Valley Creek, Minnesota.

Nevertheless, as noted previously, the recolonization samples always
more closely resembled the taxonomic composition of the grab, rather than the
drift samples. This indicates that while drift is important in the early recoloniz-
ing process in this river, drift-aided recolonization is not a haphazard process
dictated by the total composition of the drift fauna. Moreover, it suggests that
particular organisms have an ability to select particular sites at which they may
terminate drifting.

Of the other possible sources of recolonizing organisms, oviposition was
likely not a major source for the Vermilion River during this time of year.
Inspection of the recolonization samples showed that while the colonizing
organisms varied widely intraspecifically in size, they were generally larger
than they would have been had they hatched from eggs at any time during the
experiment.

MICHAEL J. S. TEVESZ

No. 27

TABLE 4

Description of Grab Samples for Experiment II

Taxa Sampling period

Arthropoda 1 4

Insecta
Diptera

Chironomidae 25 17

Coleoptera

Stenelmis 4 3

Megaloptera

Sialis 1 0

Ephemeroptera

Ephemera 1 0

Heptageniidae 1 1

Caenidae 5 2

Annelida

Oligochaeta 4 2

Mollusca

Gastropoda

Ferrissia 1 1

Oxytrema 1 3

Bivalvia

Sphaerium 1 1

Lasmigona 0 1

All Insecta represented as aquatic immatures.

Organism numbers = # individuals/400cm2

By process of elimination, the absence of oligochaetes, bivalves, and the
gastropod Ferrissia from the drift samples indicates these organisms colonized
the experimental islands by migration within the substratum or by upstream
migration within the water.

Some of the early recolonizers are also more abundant on the undisturbed
river bottom than are the later recolonizers (Tables 3, 4). This suggests that the
higher the relative abundance of a particular taxon, the greater the chance of its
being a successful early recolonizer. Thus it is entirely conceivable that highly
mobile but nondrifting organisms (e.g., Sphaerium, Oligochaeta) may have
appeared on the islands earlier had these taxa been more abundant.

In the sense of having 1 ) all taxa in the grab samples represented at some
time on the islands, and 2) total organism number in the grab samples at least
equaled on the islands, recolonization was complete for both experiments by
the 5th week and 24th day, respectively. Although exact equivalence in
taxonomic richness and organism number was not achieved between the grab

1978

BENTHIC RECOLONIZATION

TABLE 5

Description of Drift Samples for Experiment I

Sampling week

(Samples taken at weekly intervals July 9- Aug. 13, 1975)

Taxa 1

Arthropoda
Insecta
Diptera

Chironomidae A

, Coleoptera

Psephenus C

Stenelmis (a) A

Stenelmis

Plecoptera

Acroneuria

Megaloptera

Sialis R

Ephemeroptera
Ephemera A

Callibaetis A

Heptageniidae C

Caenidae C

Trichoptera

Hydropsyche C

Chimarra R

Odonata
Clithemis

Heteroptera C

Corixidae
Mollusca
Gastropoda
Oxytrema

Physa R

2 4* 5

AAA

R C

C R

C R C

C C R

C R

Key: *No samples collected during week 3 of experiment
A = Abundant> 10 individuals
C = Common =2-9 individuals
R = 1 individual

All Insecta represented are immatures unless denoted by “(a)” signifying adult.

samples and islands by the end of the experiment, the discrepancy could
possibly be attributed to one or both of the following causes: 1) differences
between the microenvironments of the islands and area where the grab samples
were taken, and 2) unequal biotic modification of the different areas by the
numerous observed storm events. These possible causes, plus the additional

MICHAEL J. S. TEVESZ

No. 27

TABLE 6

Description of Drift Samples for Experiment II

Sampling period

(Samples taken every third day beginning Aug. 16 — ending Sept. 6, 1975)

Taxa 1

Arthropoda

Insecta

Diptera

Chironomidae A

Coleoptera
Stenelmis (a) R

Stenelmis
Plecoptera
Acroneuria R

Ephemeroptera
Ephemera

Callibaetis A

Heptageniidae A

Caenidae C

Trichoptera
Hydropsyche
Odonata
Clithemis
Heteroptera
Corixidae
Mollusca
Gastropoda
Oxytrema C

Key: A = Abundant > 10 individuals
C = Common = 2-9 individuals
R = Rare = 1 individual

5 6 7 8

A A A A

R R

C R

C C R

All Insecta represented are immatures unless denoted by “(a)” signifying adult.

factors of emergence of adult insects and possible population density regulation
through drift (Waters, 1966; Dimond, 1967), might also explain the pro-
nounced fluctuations in organism number during both experiments.

The time scale involved in the recolonization of this area of the Vermilion
River is roughly similar to the findings of other workers for different streams.
For example, Waters ( 1964) finds 1-2 days sometimes a sufficient recolonizing
time for numerically dominant invertebrates in a Minnesota stream. Mason et
al. (1967) suggest “about six weeks” is an adequate time for recolonization by

1978

BENTHIC RECOLONIZATION

rock-adhering or rock-clinging invertebrates in large streams. Williams and
Hynes (1976) drawing upon previously published sources, believe 28 days to be
a reasonable average time for their field area in Ontario.

For the Vermilion River, this all suggests that the benthic macroinverte-
brate population may rapidly recover from a local disaster if there are 1) no
permanent, major, physical and chemical alterations in the environment, and 2)
if there are organisms available nearby in the river to repopulate the affected
area. For rivers in general, the overall kind, extent, and duration of the disaster
can greatly increase this recovery time (Cairns et al. 1971).

Colonization studies that are precisely comparable to the present one are
lacking for lentic environments. This is because most lentic colonization
studies performed to date involve a situation in which the environment is
extensively physically modified by abiotic factors prior to the completion of
colonization (e.g., as in the filling of an impoundment or a dry lake; see
Paterson and Fernando, 1969; McLachlan, 1975). These physical changes,
such as modification of water depth and a change from a lotic to lentic system,
are time consuming and may help retard the completion of colonization by a
year or more.

For the marine realm, the experiments performed by McCall (1977) are
not only comparable to the present study, but also provide notably contrasting
results. McCall showed for Long Island Sound that most early colonizers settle
from the water column onto defaunated bottom as larvae or very young
juveniles. Mobility subsequent to settling is restricted. An easily recognizable
succession occurs wherein opportunistic, or “r”-strategist, species initially
colonize the area, only to be replaced later by more diverse assemblages of
“K” -strategist species. The time for complete recolonization is approximately
one year.

In the Vermilion River, recolonization takes place by either immature
stages of varying age or by adults. Thus the colonization process is more of a
function of relative mobility (drift, bottom migration) and relative abundance
on the natural bottom and is not mainly a reproductive event. Also, no obvious
pattern of succession could be identified for the Vermilion River, and the time
involved in complete recolonization was much less than that for Long Island
Sound.

This rapid recolonizing ability of lotic benthos is highly adaptive to the
extremely rigorous physical nature of their habitat. Removal of organisms
from areas of river bottom is likely a continual but spatially patchy phenomenon
caused by such factors as floods, ice scour, and sediment transport. The ability
of organisms to quickly recolonize areas following natural defaunation helps
insure maximal dispersal within their habitat. This ability would help promote

MICHAEL J. S. TEVESZ

No. 27

the maintenance of high population levels within the habitat by reducing the
potential for intraspecific competition. Also, wide dispersal would mean that as
local populations are wiped out by changing environmental conditions, undis-
turbed populations of organisms are present elsewhere in the river to help insure
the species’ survival.

Acknowledgements

The author thanks the following people for contributing to this study: Dr.
Peter L. McCall, Case Western Reserve University, critically reviewed the
manuscript; Mr. Paul Swaidner, Cleveland State University, and Dr. Wilson
Britt, Ohio State University, provided taxonomic identifications of immature
aquatic Insecta; Mr. James Blaser, Amherst, Ohio, provided editorial and field
assistance.

This study was funded by grants from Cleveland State University.

REFERENCES

Beck, W. M., 1954, Studies in stream pollution biology, Q. Jour. Florida Acad. Sci.
17:211-227.

Brinkhurst, R. O., 1965, Observations on the recovery of a British river from gross
organic pollution, Hydrobiologia 25:9-51.

Cain, S. A., 1938, The species-area curve, Am. Mid. Nat. 19:573-581.

Caims, J., J. S. Crossman, K. L. Dickson, and E. E. Herricks, 1971, The recovery of
damaged streams, Assoc, of Southeast. Biol. Bull. 18:79-106.

Coleman, M. J. and H. B. N. Hynes, 1970, The vertical distribution of the invertebrate
fauna in the bed of a stream, Limnol. Oceanogr. 15:31-40.

Crisp, D. T., and T. Gledhill, 1970, A quantitative description of the recovery of the
bottom fauna in a muddy reach of a mill stream in Southern England after draining
and dredging, Arch. Hydrobiol. 67:502-541.

Dickson, K. L. and J. Cairns, 1972, The relationship of fresh- water macro-invertebrate
communities collected by floating artificial substrates to the MacArthur-Wilson
equilibrium model, Amer. Mid. Nat. 88:68-75.

Dimond, John B., 1967, Evidence that drift of stream benthos is density related, Ecol.
48:855-857.

Gaufin, A. R. and C. M. Tarzwell, 1952, Aquatic invertebrates as indicators of stream
pollution, Publ. Hlth. Rep. Wash. 67:57-64.

Gaufin, A. R. and C. M. Tarzwell, 1956, Aquatic macro-invertebrate communities as
indicators of pollution in Lytle Creek, Sewage Industr. Wastes 28:906-924.

1978

BENTHIC RECOLONIZATION

Glime, J. M. and R. M. Clemans, 1972, Species diversity of stream insects on
Fontinalis spp. compared to diversity on artificial substrates, Ecol. 53:458-464.

Mac Arthur, R. and E. O. Wilson, 1963, An equilibrium theory of insular zoogeog-
raphy, Evolution 17:373-387.

Mason, W. T., J. B. Anderson, andG. E. Morrison, 1967, A limestone-filled artificial
substrate sampler-float unit for collecting macroinvertebrates in large streams.
Prog. Fish-Cultur. 29:74.

Mason, W. T., J. B. Anderson, R. Kreis, and W. C. Johnson, 1970, Artificial substrate
sampling, macro invertebrates in a polluted reach of the Klamath River, Oregon, J.
Water Pollution Control Federation, 42:R315-R328.

McCall, P. L., 1977, Community patterns and adaptive strategies of the infaunal
benthos of Long Island Sound, Jour. Mar. Res. 35:221-266.

McLachlan, A. J., 1975, The role of aquatic macrophytes in the recovery of the benthic
fauna of a tropical lake after a dry phase, Limn. Oceanogr. 20:54-63.

McLachlan, A. J. and S. M. McLachlan, 1971, Benthic fauna and sediments in the
newly created Lake Kariba (Central Africa) Ecol. 52:800-809.

Ohio Division of Water, 1954, Gazetteer of Ohio Streams, Ohio Dept. Natur. Re-
sources, Div. Water, 175pp.

Paterson, C. G. and C. H. Fernando, 1969, Macroinvertebrate colonization of the
marginal zone of a small impoundment in Eastern Canada, Can. J. Zool.
47:1229-1238.

Waters, T. F., 1964, Recolonization of denuded stream bottom areas by drift, Trans.
Am. Fish. Soc. 93:311-315.

Waters, Thomas F., 1966, Production rate, population density, and drift of a stream
invertebrate, Ecol. 47:595-604.

Williams, D. D. and H. B. N. Hynes, 1974, The occurrence of benthos deep in the
substratum of a stream, Freshwater Biol. 4:233-256.

Williams, D. D. and H. B. N. Hynes, 1976, The recolonization mechanisms of stream
benthos, Oikos 27:265-272.

CONTENTS

No. 24. Mammals Utilized as Food by Owls in Reference to the Local
Fauna of Northeastern Ohio — Ralph W. Dexter

No. 25. Lipid Deposition in the House Sparrow and Red-Winged
Blackbird — Timothy O. Matson and Larry D. Caldwell

No. 26. Lipid Deposition in Nestlings of the House Sparrow and Red-
Winged Blackbird — Timothy O. Matson and Larry D. Caldwell

No. 27. Benthic Recolonization Patterns in the Vermillion River, Ohio—
Michael J. S. Tevesz

KIRTLANDIA

•NATURAL HISTORY*

KIRTLANDIA

David S. Brose, Editor

Kirtlandia is distributed by The Kent State University Press, Kent, Ohio 44242. Price $2.00 per
issue domestic, $2.50 foreign.

Kirtlandia is abstracted in Zoological Record and Biological Abstracts, and indexed in Bibliog-
raphy and Index of Geology.

ISSN: 0075-6245

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO NUMBER 28

A NEW SPECIES OF THE GENUS AUSTRALOPITHECUS
(PRIMATES: HOMINIDAE) FROM THE
PLIOCENE OF EASTERN AFRICA

DONALD C. JOHANSON

Curator of Physical Anthropology
The Cleveland Museum of Natural History
Cleveland, Ohio

TIM D. WHITE

Department of Anthropology
University of Califomia-Berkeley
Berkeley, California

YVES COPPENS
Directeur

Laboratoire d’Anthropologie
Musee de V Homme
Paris, France

0075-6245/78/1978-0028 $00.70/0

D. JOHANSON et al.

No. 28

Abstract

Hominid fossils have recently been recovered from Pliocene age deposits at
Hadar, Ethiopia, and Laetolil, Tanzania. These fossils share an array of distinctive
morphological characteristics which suggests that they belong to a single species of
the genus Australopithecus, differing significantly from those previously described.

The binomen Australopithecus afarensis sp. nov. is therefore assigned to this
collection of early hominid remains.

A substantial collection of hominid fossils has recently been recovered
from two Pliocene sites in eastern Africa. Hominid specimens from Hadar in
Ethiopia (11°N, 40°30'E) and Laetolil in Tanzania (3°12'S, 35°11'E) have
been dated to between ca. 2.9 and ca. 3.7 million years before present (Aronson
et al., 1977; Leakey et al., 1977). The strong morphological continuity be-
tween these two samples suggests that they are best considered as representing
a single taxon; hence, the Hadar and Laetolil fossils currently constitute the
oldest indisputable evidence of the family Hominidae.

Some of these specimens have been provisionally allocated to Homo sp.
indet. (Johanson and Taieb, 1976; Leakey et al., 1977) while others have been
referred to Australopithecus aff. africanus (Johanson and Taieb, 1976). Sub-
sequent to this preliminary assessment, more detailed study of the entire
hominid sample from Laetolil and Hadar has provided us with new information
indicating that 1) the specimens belong to only a single taxon, and 2) they differ
significantly from previously recognized species of Plio/Pleistocene
Hominidae. The Hadar and Laetolil hominids exhibit many morphological
features found in specimens attributed to the genus Australopithecus (sensu la-
to) (as defined by Le Gros Clark, 1955) and they are therefore assigned to this
taxon. Careful evaluation of the material has led to the recognition of a distinc-
tive suite of morphological traits distinguishing the Laetolil and Hadar remains
from other hominid taxa. Such study indicates the necessity of assigning these
fossils to a new and more primitive species of Australopithecus.

Order primates Linnaeus 1758

Superfamily hominoidea Simpson 1931
Family hominidae (Le Gros Clark 1955)

Genus Australopithecus Dart 1925

Australopithecus afarensis sp. nov.

Synonomy:

1950 Meganthropus africanus Weinert, H.: 139

1955 Praeanthropus africanus §enyiirek, M.: 33

1978 A NEW SPECIES 3

Holotype:

Laetolil Hominid (L.H.-) 4, mandibular corpus with broken RC,
M,, M2; intact R and LP4; RP3, M3; LMl9 M2.

Locality:

Locality 7 of the Laetolil Site, Tanzania, collected in 1974 by
M. Muluila.

Horizon:

Laetolil Beds between Aeolian Tuffs b and c, Pliocene age
(3. 6-3. 7 m.y. b.p.)

Paratypes:

Laetolil Beds, Tanzania:

L.H.-l, RP4; L.H. -2, immature mandibular corpus with perma-
nent and deciduous teeth; L.H.-3(a-t), isolated upper and lower
deciduous and permanent teeth; L.H.-3/6a, b, Rdc-, Ldm1;
L.H. -5, R. maxillary row, P-M1; L.H.-6(a-e), isolated perma-
nent and deciduous upper teeth; L.H. -7, RM- frag.; L.H. -8,
RM2, RM3; L.H. -10, L. edentulous mandibular frag.; L.H.-l 1,
LM1/2;L.H.-12, LM2/3 frag.; L.H.-13, R. edentulous mandibu-
lar corpus frag.; L.H.-14(a-h), isolated lower teeth; Garusi
maxilla, RP3-P4.

Hadar Formation, Ethiopia:

Sidi Hakoma Member:

A.L. 128-1, L. prox. femur frag.; A.L. 128-23, R. mandibular
corpus, C-M2; A.L. 129-la, b, c, femur and tibia frags.; A.L.
129-52, L. ischium; A.L. 137-48a, R. distal humerus; A.L.
137-48b, R. distal ulna; A.L. 145-35, L. mandibular corpus,
P3-M2; A.L. 166-9, L. temporal frag.; A.L. 198-1, L. mandibu-
lar corpus, C-M3; A.L. 198-17a, b, LI\LP;A.L. 198-18, RL>;
A.L. 199-1, R. maxilla, C-M3; A.L. 200- la, maxilla, complete
dentition; A.L. 200-lb, RM^ A.L. 211-1, R. prox. femur frag.;
A.L. 228-1, R. diaphysis femur; A.L. 266-1, mandibular cor-
pus, LP3-M!, RP3-M3; A.L. 277-1, L. mandibular corpus,
C-M2; A.L. 311-1, L. mandibular corpus, P3; A.L. 322-1, L.
distal humerus; A.L. 400-la, mandibular corpus, LIrM3,
RI2-M3; A.L. 400-lb, RC-; A.L. 411-1 , R. mandibular corpus,
MrM3.

Denan Dora Member:

A.L. 161-40, LM3; A.L. 188-1, R. mandibular corpus, M2-M3;
A.L. 207-13, L. mandibular corpus, P3-M2; A.L. 241-14,
LM_; A.L. 366-1, LM3; A.L. 388-1, LM3.

D. JOHANSON et al.

No. 28

A.L. 333-1, facial frag, and maxilla, RP3-P4, LC-P3; -2,
maxilla, RC-M1, LI2-P4; -3, R. prox. femur; -4, R. distal femur;
-5, L. prox. tibia frag.; -6, L. distal tibia; -7, L. distal tibia; -8,
R. calcaneum frag.; -9a, -9b, R. and L. distal fibulae; -10, L.
mandibular corpus frag., P3; -11, R. prox. ulna frag.; -12, R.
distal ulna; -13, L. prox. V metatarsal (MT); -14, R. V metacar-
pal (MC); -15, L. prox. II MC; -16, L. Ill MC; -17, R. distal V
MC; -18, R. distal IV MC; -19, prox. hand phalanx; -20, prox.
hand phalanx frag.; -21, distal MT; -22, prox. hand phalanx
frag.; -23, cranial frag.; -25, intermed. hand phalanx; -26, prox.
phalanx; -27, R. distal II MC; -28, R. medial cuneiform frag.;
-29, L. distal humerus; -30, Rdm^ -31, prox. hand phalanx
frag.; -32, intermed. hand phalanx; -33, prox. hand phalanx;
-34, immature metapodial; -35, Rdc_; -36, R. foot navicular;
-37, R. calcaneum frag.; -38, L. immature distal ulna; -39, L.
immature prox. tibia; -40, R. capitate; -41, R. med. femoral
condyle frag.; -42, L. prox. tibia; -43a, b, L. and R. mandibular
corpi, R. andL. dmrdm2; -44, LM_; -45, partial cranium; -46,
intermed. hand phalanx; -47, R. foot navicular; -48, L. II MC;
-49, prox. hand phalanx frag.; -50, R. hamate; -51, body
thoracic vertebra; -52, frag. M_; -53, thoracic vertebra frag.;
-54, L. prox. I MT; -55, L. calcaneum frag.; -56, L. IV MC;
-57, prox. hand phalanx; -58, R. prox. I MC frag.; -59, R.
mandibular corpus frag., M2-M3; -60, prox. phalanx; -61, L.
distal femur frag.; -62, prox. hand phalanx; -63, prox. hand
phalanx; -64, intermed. hand phalanx; -65, R. prox. Ill MC;
-66, Ldc~; -67, Rdi2; -68, Ldi2; -69, L. prox. hand phalanx; -70,
immature metapodial; -71, prox. foot phalanx; -72, MT frag;
-73, body lumbar vertebra; -74, L. mandibular corpus frag.,
MrM3; -75, headR. talus; -76, Ldi2; -77, Ldc_; -78, L. prox. V
MT; -79, L. lateral cuneiform; -80, R. trapezium; -81, body
immature thoracic vertebra; -82 LI1; -83 atlas vertebra frag.;
-84, R. temporal frag; -85, L. distal fibula; -86, maxilla, L. and
R. dnV-dm2, M1; -87, L. prox. humerus frag.; -88, intermed.
hand phalanx; -89, L. V MC; -90, LC_; -92, immature long
bone frag.; -93, prox. hand phalanx; -94, L. clavicle frag.; -95,
R. prox. femur; -96, L. distal tibia; -97, L. mandibular corpus
frag.; -98, R. prox. radius; -99, Ldc~; -100. L. coronoid process
mandible; -101, axis vertebra; -102, prox. hand phalanx; -103,
RC_; -104, Rdc~; -105, partial immature cranium, Rdm4-dm2;

1978

A NEW SPECIES

-106, cervical vertebra; -107, R. prox. humerus; -108, L. as-
cending ramus; -109, humerus shaft frag.; -110, L. immature
distal femur frag.; -Ill, R. immature distal femur frag.; -113,
immature long bone frag.; -115, associated foot bones.

A.L. 333w-la-e, R. andL. mandibular corpi, LP3-M2, RP3-M2,
RM3 frag., R. condyle; -2, LC-; -3, LI2; -4, prox. hand
phalanx; -5, R. distal II MC; -6, R. prox. Ill MC; -7, immature
prox. hand phalanx; -8, vertebra frag.; -9a, b, LIl9 LI2; -10,
RC_; -11, distal hand phalanx; -12, R. mandibular corpus frag.,
Mj; -13, prox. fibula; -15, R. coronoid process mandible; -16,
L. mandibular condyle; -17, -18, -19, rib frags.; -20, immature
prox. hand phalanx frag.; -21, immature phalanx; -22 R. distal
humerus frag.; -23, R. immature II MC; -25, prox. hand
phalanx frag.; -26, L. prox. V MC; -27, L. mandibular corpus,
M2; -28, RI2; -29, immature prox. phalanx frag.; -30, rib; -31,
L. distal humerus frag.; -32, R. mandibular corpus, M3; -33, R.
prox. radius; -34, intermed. hand phalanx frag.; -35, L. prox. V
MC; -36, L. prox. ulna; -37, L. distal fibula -38, intermed. hand
phalanx; -39, R. IMC; -40, R. prox. femur frag.; -41, rib frag.;
-42, RP4; -43, immature prox. I MT frag.; -45, rib frag.; -46, R.
mandibular corpus, P3; -47, rib frag.; -48, RM2; -50, distal hand
phalanx; -51, prox. hand phalanx frag.; -52, L. mandibular
condyle; -53, intermed. hand phalanx frag.; -54, prox. hand
phalanx frag.; -55, MT frag.; -56, R. distal femur; -57, L.
mandibular corpus, M2-M3; -58, mandibular corpus frag.,
LI,-P4, RIrC_; -59, L. mandibular corpus, M2-M3; -60, man-
dibular corpus, LP3-M3, RIrC_.

A.L. 333x-l, RM3; -2, LI2; -3, LC~; -4, RI1; -5, R. prox. ulna;
-6, -9, R. clavicle frags.; -12, thoracic vertebra; -13a, prox.
hand phalanx; -13b, intermed. hand phalanx; -14, -15, prox.
radial epiphyses; -17, RI2; -18, intermed. hand phalanx; -20,
RI1; -21a, b, intermed. hand phalanges; -25, di1/2; -26, R. prox.
tibia.

Kada Hadar Member:

A. L. 288-1, partial skeleton.

Horizon:

Laetolil Beds, Tanzania. Known hominid sample from between
strata dated to 3.59 and 3.77 m.y.

D. JOHANSON et al.

No. 28

Hadar Formation, Ethiopia. Sidi Hakoma Member dated to
older than ca. 3.0 m.y., but less than ca. 3.3 m.y. Denan Dora
and Kada Hadar Members dated to younger than ca. 3.0 and
older than ca. 2.6 m.y. with the latter member stratigraphically
above the former.

Diagnosis:

A species of Australopithecus distinguished by the following characters:

Dentition

Upper central incisors relatively and absolutely large; upper
central and diminutive lateral incisors with strong lingual basal
tubercles, upper incisors with flexed roots; strong variation in
canine size, canines asymmetric, lowers with strong lingual
ridge, uppers usually with exposed dentine strip along distal
edge when worn; P3 occlusal outline elongate oval in shape with
main axis mesiobuccal to distolingual at 45°-60° to tooth row,
dominant mesiodistally-elongate buccal cusp, small lingual
cusp often expressed only as inflated lingual ridge; diastemata
often present between I2/C_ and C_/P3; C/P3 complex not
functionally analogous to pongid condition.

Mandible

Ascending ramus broad, not high; corpus of larger specimens
relatively deep anteriorly and hollowed in region of low mental
foramen which usually opens anterosuperiorly; moderate supe-
rior transverse torus; low rounded inferior transverse torus; an-
terior corpus rounded and bulbous; strong posterior angulation
of symphyseal axis; postcanine teeth aligned in straight rows;
arcade tends to be sub-rectangular, smaller mandibles with rela-
tively narrow incisor region.

Cranium

Strong alveolar prognathism with convex clivus; palate shallow,
especially anteriorly; dental arcade long, narrow, straight sided;
facial skelton exhibiting large, pillar-like canine juga separated
from zygomatic processes by deep hollows, large zygomatic
processes located above P4/Mx and oriented at right angles to
tooth row with inferior margins flared anteriorly and laterally;
occipital region characterized by compound temporal/nuchal
crest (in larger specimens), concave nuchal plane short an-
te ro posteriorly; large, flattened mastoids; shallow mandibular
fossae with weak articular eminences placed only partly under
braincase; occipital condyles with strong ventral angulation.

1978

A NEW SPECIES

Postcranium

See remarks.

Description:

Dentition

Large canines project beyond tooth rows and possess massive,
long roots; buccal face of P3’s often with vertical wear striae
caused by occlusion with upper canines; P3’s often with two
distinct roots, the mesial one round and angulated mesio-
buccally, the distal one plate-like and oriented transverse to the
tooth row; P3’s sometimes three rooted, with pointed buccal
cusp, extensive and asymmetric buccal face, buccal cer-
viocoenamel line projecting towards mesio-buccal root, and the
lingual cusp situated mesial to buccal cusp, P3’s tend to be larger
than P4’s and the latter do not show mesiodistal elongation of the
buccal crown half; lower molars, especially M, and M2 tend to
be square with cusps arranged in Y-5 pattern; wide occlusal
foveae on all molars; strong molar size gradient of
M3 > M2 > Ml; hypocones and hypoconulids large; deciduous
canines similar to the permanent ones in form and occlusal
projection; dm,’s molarized, with lingually facing anterior
foveae and deep buccal grooves; substantial variation in tooth
size.

Mandible

Ascending ramus slopes posteriorly and joins corpus at high
position defining narrow extramolar sulcus; broad condyles;
mandibular canal immediately below distal M3 root; base of
corpus everted.

Cranium

Incisors procumbent; lower margin of pyriform aperture marked
laterally by raised borders; tooth rows tend to converge
posteriorly; strong muscle markings on vault and cranial
base, temporal lines converge anteriorly, but presence of
sagittal cresting unknown; lateral portion of cranial base highly
pneumatised; occipital condyles placed below external auditory
meatus in lateral view; estimated cranial capacity small relative
to Homo sp.; broad mandibular fossae, laterally projecting
postglenoid process; pyramid process angles anteriorly relative
to more transverse tympanic plate.

Postcranium

Strong dimorphism in body size; all skeletal elements with high

D. JOHANSON et al.

No. 28

level of robusticity in muscle and tendon insertions; pelvic
region and lower limbs indicate adaptation to bipedal locomo-
tion; “waisted” appearance of capitate; third metacarpal lack-
ing styloid process; phalanges strongly longitudinally curved;
foot navicular with cuboideonavicular facet; deep peroneal
grooves on distal fibulae; anterior margin of ilium between
anterior superior and inferior spines relatively straight; cervical
vertebrae with long spinous process; relatively high
humero femoral index compared to modem humans.

Etymology:

The species name afarensis derives from the Afar depression of
northeastern Ethiopia, where the largest portion of the paratype
series was recovered.

Remarks

Laetolil Hominid-4 was selected as the holotype both because of its distinc-
tive, diagnostic morphology and because it has previously been fully described
and illustrated (White, 1977). The generic name Praeanthropus originally
proposed by Hennig (1948) is invalid because no species designation was
given. Senyurek (1955) used the generic nomen Praeanthropus and utilized
Weinert’s (1950) specific name africanus, designating the original Garusi
maxillary fragment as Praeanthropus africanus. The present authors do not
consider the original Garusi maxillary fragment or the new Laetolil and Hadar
material to represent a hominid genus distinct from Australopithecus.

The authors recognize that individual traits and even single specimens in the
new collections can be matched in other samples representing different taxa
(e.g., Australopithecus africanus Dart 1925, Homo habilis Leakey, Tobias and
Napier 1964). However, the overall character complex seen in the Hadar and
Laetolil fossils is distinct from other previously found and described species.
Care has been taken in the diagnosis to follow Mayr’s suggestion to “list the
most important characters or character combinations that are peculiar to the
given taxon and by which it can be differentiated from other similar or closely
related ones” (1969: p. 266). In the description of Australopithecus afarensis
we have chosen to present a characterization of the entire hypodigm. This
should insure that the presentation not be viewed as typological and should also
given some indication of the variation recognized in this new taxon.

It is important to recognize that certain traits or complexes were not
considered in the diagnosis but placed in the description due to the lack of
comparable anatomical specimens from other species of Australopithecus.

1978

A NEW SPECIES

Some of the traits, such as the morphology of the hand and foot bones, may be
diagnostic of the new species, but this cannot be ascertained until pertinent new
material is recovered from other sites.

The Hadar and Laetolil fossils appear to represent a distinctive early hominid
form characterized by substantial size variation which is interpreted as reflect-
ing sexual dimorphism. Members of this new taxon display a complex of
primitive dental, cranial, and possibly postcranial characteristics. Recognition
of the new species Australopithecus afarensis has important implications for
interpretations of early hominid phylogeny. These implications will be consid-
ered in forthcoming publications.

Acknowledgments

Fieldwork at Hadar and Laetolil was undertaken with the kind permission
and cooperation of the Provisional Military Government of Socialist Ethiopia
and the United Republic of Tanzania respectively. We thank the following
institutions for financial support: the National Geographic Society, the National
Science Foundation, the L. S. B. Leakey Foundation, the Wenner-Gren Foun-
dation, the Cleveland Museum of Natural History, the Centre National de la
Recherche Scientifique, and the Singer- Polignac Foundation.

We thank Dr. Owen Lovejoy for helpful comments and Mr. Anson Laufer
and Mr. Bruce Frumker for assistance with the photographs. Thanks are also
due Mr. William H. Kimbel and Mr. B. Thomas Gray for providing invaluable
comments, suggestions, as well as editorial and photographic assistance.

Appreciation is expressed to Dr. Maurice Taieb for his role in the discovery
of Hadar and for his initiation and successful development of the International
Afar Research Expedition. Special thanks are extended to Professor Ernst
Mayr for critically reviewing the manuscript.

D. JOHAN SON et al.

No. 28

1000

Kilometers

Fig. 1. Map of eastern Africa showing the locations of Hadar and Laetolil.

1978

A NEW SPECIES

Fig. 2. Type specimen of the new species Australopithecus afarensis, the mandible L.H. -4 from
Laetolil, Tanzania. Occlusal view. Natural size.

D. JOHANSON et al.

No. 28

Fig. 3. Two distal femora from Hadar, Ethiopia (A. L. 333-4, left; A.L. 129-la, right) indicating
the size variation within the new species. Anterior view.

1978

A NEW SPECIES

Fig. 4. The partial skeleton from Hadar, Ethiopia A.L. 288-1 . The total length of the left femur is
approximately 280 mm.

D. JOHANSON et al.

No. 28

References

Aronson, J. A., T. J. Schmitt, R. C. Walter, M. Taieb, J. J. Tiercelin, D. C. Johanson,
C. W. Naeser and A. E. M. Nairn, 1977, New geochronologic and
palaeomagnetic data for the hominid-bearing Hadar Formation of Ethiopia,
Nature, 267: 323-327.

Hennig, E., 1948, Quartarfaunen und Urgeschichte Ostafrikas, Naturwiss. Rdsch.
Jahrg. 1, 5: 212-217.

Johanson, D. C. and M. Taieb, 1976, Plio-Pleistocene hominid discoveries in Hadar,
Ethiopia, Nature, 260: 293-297.

Leakey, M. D., R. L. Hay, G. H. Curtis, R. E. Drake, M. K. JackesandT. D. White,
1977, Fossil hominids from the Laetolil Beds, Nature, 262: 460-466.

Le Gros Clark, W. E., 1955, The fossil evidence for human evolution, Chicago: The
University of Chicago Press.

Mayr, E., 1969, Principles of systematic zoology, New York: McGraw-Hill Book
Company.

§enyurek, M., 1955, A note on the teeth of Meganthropus africanus Weinert from
Tanganyika Territory, Belleten (Ankara), 19: 1-54.

Weinert, H., 1950, Uber die Neuen Vor-und Fruhmenschenfunde aus Afrika, Java,
China und Frankreich, Zeit. Morph. Anthrop., 42: 113-148.

White, T. D., 1977, New fossil hominids from Laetolil, Tanzania. Am. J. Phys.
Anthrop., 46: 197-230.

KIRTLANDIA

THE CLEVELAND MUSEUM OF NATURAL HISTORY

CLEVELAND, OHIO

NUMBER 29

PLIOCENE-PLEISTOCENE SUIDAE FROM HADAR,

ETHIOPIA

H. B. S. COOKE

Dalhousie University
Halifax, Nova Scotia, Canada

Abstract

Three different suids occur in the Hadar Formation. Nyanzachoerus pattersoni is
plentiful in the Sidi Hakoma member, especially in the lower part, and morphologi-
cally matches the type material from Kanapoi very closely. Notochoerus euilus
ranges throughout, making up 61% of the total sample. The skull architecture is
described for the first time. A small suid occurring throughout is regarded as a new
species of Kolpoc hoe r us, K. afarensis, distinguished from/C limnetes by its smaller
size, simpler molars and more Sus-like features of the premolars. It is, in all
probability, ancestral to AT. limnetes. Comparison of the Notochoerus euilus mate-
rial with that from the Usno Formation and the lower part of the Shungura Formation
suggest that the lower Sidi Hakoma member may be in the range of 3 .0 to 3 . 5 million
years old.

Introduction

Addis Ababa, the capital of Ethiopia, lies in the center of the country at an
elevation of 2,355 metres above sea level, not far from the source of the Awash
river. This river drains northeastward into the Afar province, but derives much
of its water from the highlands near the source area and from seasonal streams
that flow into it from the western flanks of the Afar triangle. One of these
streams, Kada Hadar, meets the Awash in an area that has yielded substantial
amounts of fossil material, including important hominid remains (Taieb et al.,
1972; Taieb, Johanson and Coppens, 1975). The broad framework of the
geology has been considered by Taieb (1974) and the particular succession

0075-6245/78/1978-0029 $03.15/0

H.B.S. COOKE

No. 29

exposed in the Hadar area has been described recently (Taieb et al., 1976),
together with a preliminary account of the fossil remains and hominid
discoveries (Johanson and Taieb, 1976). The Awash at Hadar is close to 500 m
above sea level, with local relief of some 100 m and badland type dissection.

The Hadar Formation has a thickness of about 180-200 m, although the
base of the lowest unit is not yet firmly established. The sediments are largely
arenaceous and argillaceous and represent various phases in a fluctuating
complex of lacustrine, lake margin and fluvial deposits, apparently related to a
Pliocene-Pleistocene lake that occupied much of the Afar basin. Several minor
erosional unconformities can be seen, but they do not appear to represent major
breaks. A few steep normal faults, with displacements of 5 to 40 m occur. The
Hadar Formation is capped unconformably by an unnamed unit of Pleistocene
gravels and sands that contain hand axes.

The Hadar Formation has been divided into four members, conveniently
separated by volcanic tuff horizons that are laterally extensive and usually
altered to a chalky white material, forming useful “markers.” The lowest tuff
complex (SHT) divides the Basal Member (BM) from the Sidi Hakoma
Member (SH) and in places it has been preserved in channels, where the
material is not altered and consists largely of fresh glass shards. The Sidi
Hakoma Member has been subdivided into four submembers (SH-1, SH-2,
SH-3, SH-4) and within the upper submember there is at the eastern end of the
Hadar area a magnetically reversed basalt flow, 1 to 4 m thick, that dies out
towards the northwest. Its resistance to erosion resulted in the development of a
minor plateau in this area, capped directly by the Pleistocene gravels. Prelimi-
nary K/Ar determinations on the basalt give ages of 2.9 ±0.2 m.y. and
3.0±0.2m.y. (Aronson et al. , 1977). The TT marker lies 20 m above the basalt
level and is a very persistent zone of thin tuffs defining the base of the Denen
Dora Member (DD); within this member, three submembers are recognised
(DD-1, DD-2, DD-3). The KHT tuff is a single deposit that is laterally less
extensive than the older tuffs but serves to define the base of the fourth member,
the Kada Hadar Member (KH). Like the SHT tuff, it occurs also in channels
where glass shards are well preserved. Four submembers have been recognized
within the Kada Hadar Member (KH-1, KH-2, KH-3, KH-upper), although the
validity of these divisions has more recently been questioned (Johanson, et al. ,
1978). About 9 m above the KHT tuff is another distinctive marker, CC,
consisting of a green argillite that produces small flakes as a characteristic
alteration product of weathering.

The occurrence of complete and partial skeletons suggests rapid deposi-
tion, in a relatively low energy environment. The preservation is usually
excellent, with some specimens looking almost like fresh bone, but many are

1978

PLIOCENE- PLEISTOCENE SUIDAE

encased in a thin layer of hard calcified mud that may make cleaning and
preparation difficult. Nearly all the specimens are surface finds, but their
condition makes it clear that they have not been transported for any significant
distance and the range of uncertainty regarding the horizons of derivation is
usually small. However, it is not always possible to place material within one of
the submembers, so localities which lie close to a boundary are designated by
recording the two possible units (eg. DD2/3). The fossil localities are assigned
numbers, prefixed by “AL” and each specimen from that locality is given an
individual number (for example, the hominid skeleton “Lucy” is AL 288-1).
The stratigraphic positions at present assigned to those localities that have
furnished suid remains are indicated below. As will be seen, there are many
localities in SH-1 to SH-3 and again in DD-1 to DD-3, but relatively few in
SH-4 and in KH-1 to KH-3. It is convenient to group the suids stratigraphic ally
into four divisions as follows:

D. Lower Kada Hadar Member, together with two localities which span
DD-3 and KH-1; includes localities: 120, 157, 164, 186, 310, 359,
361 and 367;

C. Denen Dora Member, covering DD-1 (except those localities which
also span SH-4), DD-2 and DD-3, including localities: 58, 116, 118,
121, 133, 134, 161, 162, 167, 168, 169, 171, 172, 174, 182, 184,

185, 187, 188, 190, 191, 194, 195, 201, 220, 233, 239, 241, 246,

247, 250, 259, 260, 287, 291, 296, 307, 309, 315, 316, 317, 321,

325, 332, 337, 342, 344, 358, 362, 378, 379 and 385;

B. Upper Sidi Hakoma Member, embracing the localities within SH-3,
together with those which span SH-4 and DD-1; includes localities:
53, 214, 226, 264, 266, 319, 330, 345, 347, 348, 380 and 384;

A. Lower Sidi Hakoma Member, comprising localities assigned to SHT,
SH-1, SH-2 and SH-3, together with those spanning both SH-3 and
SH-4; includes localities: 107, 108, 109, 124, 125, 126, 127, 128,
129, 130, 131, 137, 138, 141, 142, 145, 147, 148, 165, 166, 175,

198, 199, 200, 204, 208, 217, 218, 222, 224, 225, 229, 232, 235,

248, 251 , 252, 254, 255,257, 263, 277, 327, 353,360, 365 and 374.

There are at present no suids from the Upper Kada Hadar.

Much of the material is housed in the Ethiopian National Museum in Addis
Ababa and the author is indebted to the Director, Ato Mamo, for allowing him

H.B.S. COOKE

No. 29

Fig. 1 . Stratigraphic distribution of sites that have yielded suid fossil material from the Hadar area.

1978

PLIOCENE- PLEISTOCENE SUIDAE

to work there, and to Ato Woldesenbet for considerable help in handling the
material. Thanks are also due to Maurice Taieb, Yves Coppens and Don
Johanson for inviting the author to study the suids from Hadar, and to Tom Gray
for providing background data and invaluable help with the text. Support
funds have been provided by the National Research Council of Canada, the
Wenner-Gren Foundation for Anthropological Research, and the International
Afar Research Expedition (IARE) and are gratefully acknowledged. A substan-
tial debt is also due to Mr. Richard Wilding of the Archaeology Unit, University
of Addis Ababa, for much help, advice and hospitality.

Description

Genus Nyanzachoerus Leakey 1958
TYPE SPECIES: Nyanzachoerus kanamensis Leakey 1958
Revised diagnosis: See Cooke and Ewer, 1972

Nyanzachoerus patter soni Cooke and Ewer 1972

Material

Basal member. Mandibular fragment with LM2.3, AL 272-1.

Lower Sidi Hakoma. Almost complete skull with associated atlas vertebra,
presumed male, AL 137-4 (SH-2); slightly damaged skull, presumed female,
AL 107-13 (SH-3); incomplete skull, teeth broken, AL 235-2 (SH-2); palate
and part maxilla with cheek teeth, except RP2, AL 145-26 (SH-2); palate and
part maxilla with partial dentition, AL 235-10 (SH-2); four maxilla fragments
with two or more cheek teeth; four single upper teeth; good upper canine, AL
131-8 (SH-2). Damaged mandible with most of the teeth, AL 218-2 (SH-1/3);
mandible with damaged symphysis, AL 126-8 (SH-2); symphysis and both
mandibular rami with P3-M2 on both sides, LM3 erupting, AL 365-9 (SH-3);
left and right mandibular rami with complete cheek teeth AL 137-5 (SH-2);
symphysis and left mandibular ramus with teeth, AL 127-15, AL 127-7 (SH-2);
right mandibular ramus with RP4-M3, AL 165-14 (SH-2); four partial mandi-
bles with several teeth, AL 142-13 (SH-2), AL 137-16 (SH-2), AL 126-65
(SH-2), AL 166-5 (SH-2); six mandibular fragments with two or more teeth,
from SH-1 to SH-3, and one from SH-3/4; six mandibular fragments with single
teeth; seven single lower teeth and two fragments.

Upper Sidi Hakoma. Partial skull with RP3-M3, AL 347-8 (SH-4); parts of
broken skull with LM1'3, AL 384-7 (SH-4); RM3, incompletely erupted, AL
325-3 (SH-4).

H.B.S. COOKE

No. 29

Denen Dora. Maxilla fragment with LM3, AL 134-14 (DD-2/3); isolated
RM,, AL 20 1 - 1 A (DD-2).

Description

Nyanzachoerus is well represented in the collections up to the end of 1975 , with
approximately fifty cranial and mandibular specimens and a small amount of
postcranial material, although the association is uncertain. The sample is larger
than that from Kanapoi in northwestern Kenya, but matches very well in all
respects. The two best skulls, AL 107-13 and AL 137-4, agree very closely with
the female type and male paratype from Kanapoi, and both amplify and amend
the interpretation of those specimens. They thus warrant individual description.

AL 137-4

This skull is very well preserved and has suffered only minor damage
(Plate 1). The atlas vertebra is associated with the skull. The tip of the
premaxilla and the tips of the nasals are missing. The top of the skull and most
of the nasal area are intact, but both zygomatic arches have been damaged. The
occiput is complete but there is damage to the auditory bulla and the ear region
on the left side. The left upper canine is intact and the right upper canine is
broken off at the alveolar margin. Both upper second premolars are broken off
and the left dentition is damaged from the back of M1 to the middle of M3. On
the right side P3 is slightly damaged, as is M2, but the dentition is reasonably
complete. The dimensions and morphology of the skull agree almost perfectly
with the paratype from Kanapoi wherever there are corresponding parts. The
dimensions of the teeth match well with those of the male, and, as at Kanapoi,
are somewhat more robust than the teeth of the female. They are less worn than
the teeth of the paratype. Although the zygomatic arches in the Hadar specimen
are broken, their roots are present sufficiently to show that the zygomatic arch
flared out and was certainly as large as in the Kanapoi specimen. Comparative
measurements are given in Table 1.

Of particular value in the Hadar specimen is the excellent preservation of
the occipital and back of the cranial region of the skull, which were detached or
missing in the Kanapoi paratype. In the Kanapoi specimen there was a part of
the fronto-parietal area preserved and it is now clear that the restoration made by
Cooke and Ewer (1972) is incorrect, as the skull is longer than had been
expected. The parietal area is broad behind the orbits and is still broad at the
constriction, about three-quarters of the way to the lambdoid crest. The occiput
itself is long and the angle formed between the occiput and the parietal surface is
only about 45°. In consequence, the occipital condyles do not lie very far behind

1978

PLIOCENE- PLEISTOCENE SUIDAE

TABLE 1

Measurements of Skulls of Nyanzachoerus pattersoni (in mm)

Hadar Specimens Kanapoi Specimens

AL 137-4

AL 107-13

KP 239

KP 264

Holotype Paratype

Vertex length

580+(633e)

+457(530e)

510e

620e

Basilar length

447+(490e)

+364(440e)

425

500e

Bizygomatic breadth

+ 270+

308

c270

c480

Frontal breadth

166

147e

Parietal constriction

72.5

—

Crest breadth

155

108e

Binaural breadth

202

180e

Greatest breadth across nasals

—

Breadth of muzzle between

infraorbital foramina

56e

—

Palatal length

320+(353e)

+230(305e)

301

352

Breadth of palate

— between M3-M3

43.0

35.0

33.0

40.0

— between P2-P2

71.5

40.2

52.5

73.0

Premaxilla breadth over I3-I3

76e

—

Diastema I3-P2

—

101

Diastema C-P2

—

39.5

c35.0

Length of premolar series

56.5

50.3

51.9

68.0

Length of molar series

105.5

92.0e

95.2

99.1

Length of P2-M3

162.0

142.0e

147.0

161.6

e = estimated measurement; c = approximate measurement

the orbits when the skull is resting on the occlusal plane. The distance from the
lambdoid crest to the condyles is not much less than the distance from the
condyles to the back of M3. The specimen is a little crushed and it is possible
that this sharp angle of the occiput is exaggerated by deformation, but even if
that is true, the angle of slope is unusually sharp. The back of the occiput is
fairly deeply scooped, much as inSus and there is no sign of any median vertical
rib. The upper part of the occiput is relatively flat and the wings do not sweep
backwards as sharply as in Sus, but rather more like the condition in
Hylochoerus . The ear region is notably wider than the rather narrow occiput,
and the auditory canal emerges at an angle of approximately 45° from the
horizontal. The bullae have been eroded away but the right paroccipital process
is present and stout. The top of the braincase is depressed, and this was regarded

H.B.S. COOKE

No. 29

initially as an artifact. However, it has been noted in several skulls and may be a
real feature of the genus.

The canine flanges are more or less intermediate in form between those of
Sus and Hylochoerus . There is slight expansion of the nasals just behind the
canine flanges, or, more correctly, above the back of the canine flanges, and
there are slight signs of rugosity in this area reminiscent of the condition in the
bush pig. The upper canine emerges practically parallel to the palatal plane, but
sweeps outwards and upwards at the tip. It is truncated along much of the front
face by a vertical wear facet. A broad ventral band of enamel occurs and is
slightly striated parallel to the length, but not deeply grooved. There is also a
posterior narrow inset enamel band but no indication of such a band on the
upper anterior face. In cross-section, the canine is almost equidimensional with
a squarish U-shape or broad heart shape with a shallow dorsal groove. Consid-
ering the massive size of the skull, the canines are relatively small. The cheek
teeth conform to the pattern seen in the Kanapoi material except that the enamel
is a little more complicated in P3 and P4. The upper left M1 is worn to a flat
surface of dentine, but the right one still has a perimeter ring of enamel. The
enamel in the upper M2 and upper M3 is thicker than it is in the female,
corresponding again with features noted in the paratype material. The palatine
notch lies only very slightly behind the back of M3 and is broadly rounded,
although in some respects a little like a gothic arch. The basioccipital is fairly
short, and the basisphenoid descends from it at a steep angle.

AL 107-13

This skull is well preserved, although the superficial bone is in a bad state.
In size and morphology it is virtually a twin to the holotype from Kanapoi
(Table 1). Most fortunately, it has preserved portions of the skull which were
missing in the holotype (Plate 2 A/B). As in the holotype, the palate and cheek
teeth are perfectly preserved, although the anterior premolars are missing in the
Hadar specimen. The premaxilla, well preserved in the holotype, is missing
from AL 107-13. The basicranial region, well preserved in the holotype, is
badly damaged and only part of one occipital condyle remains in the Hadar
specimen. The right side of the zygoma is missing, but the left zygoma is in
excellent preservation. It differs from the type only in that the inflated marginal
knob is better developed, and thus the front part of the zygoma projects more at
right angles to the axis of the skull. The extra inflation is on the anterior external
part of the knob, making it more like that of Hylochoerus and adding consid-
erably to the broad platy area below and anterior to the orbit. In the holotype the
whole of the upper surface, including the upper part of the occiput, the whole of

1978

PLIOCENE- PLEISTOCENE SUIDAE

the top of the braincase and orbits, and the entire nasal region were missing.
These are intact in the Hadar specimen and show that the nasal region was rathei
narrow, with the nasal bones up-arched; there is some lateral crushing and it is
possible that the arching in the specimen may thus be exaggerated.

The parietal is moderately broad and the constriction between it and the
lambdoid crest is wide. As in the male specimen, the braincase is hollowed and
this increases the probability that this is the natural condition, somewhat
resembling the braincase of Hylochoerus. Again as in the male, the upper part
of the occiput is not deeply scooped, resembling that of Sus in general morphol-
ogy, but the angle between the upper part of the occipital and the parietal
surface is abnormally sharp. The rim of the orbits is raised a little above the
level of the parietal, recalling the condition seen both in Hylochoerus and in
Phacochoerus . From these raised orbits ridges extend down to at least the end
of the lacrimals. The condition of the bone unfortunately makes it impossible to
see the lacrimal sutures properly. Although the tip of the snout is missing, it
would seem that the narrowest part of the maxillary area is in the vicinity of the
infraorbital foramina, and the nasals widen anteriorly and tend to overhang the
maxilla. The canine flanges are very similar to those in a maleSws . In the Hadar
fossil the canine flanges are much more strongly developed than they were in
the Kanapoi type.

The cheek teeth are slightly less worn than those of the holotype, but all the
features are essentially similar. A minor difference is that in the holotype P3 the
tip of the main cone is worn flat parallel to the grinding surface, whereas in the
Hadar specimen the wear island runs obliquely from the tip back towards the P4.
The P3 and P4 are also a little stouter and the anterior cingulum is not as strongly
developed. The palate is a little wider at the positions of the second upper
pre-molars. Essentially, therefore, the only difference between this specimen
and the holotype is in the stronger inflation of the zygomatic arches. The
bizygomatic width is consequently also a little greater. Both this skull and the
male skull show that original estimates of the vertex length were too low, and
the skulls are actually a little longer than was indicated in the publication by
Cooke and Ewer (1972).

Upper dentition

No specimen occurs in which the upper incisors are preserved, so these
teeth are still known only from the Kanapoi holotype.

The upper canine in the male skull AL 137-4 has already been described
above. There are a few fragments of other upper canines and one almost
complete left upper canine, AL 131-8, and the matching tip of the right canine,

H.B.S. COOKE

No. 29

AL 131-7. The canine in the skull measures 40.5 mm (vertical) by 38.8 mm
(antero-posterior), whereas the isolated specimen is only 3 1 .9 by 29.0 mm and
has a length (in a straight line across the arc) of 130 mm. Although smaller than
the male tusk, it is morphologically similar, with a broad ventral band of

Nyanzachoerus pattersoni

§ -

< 5-

o:
m

1 p2 251

25-

x X

XX xx

•

. X XX

X X <

20-

x . 20-

5 20

E 20

Ixl

I—

35-

30-

. X
X

5 x

— i — i — i — i — i — i — i — i— i — r—i

25 30 35

KEY

▼ Upper Denen Dora
o Lower Denen Dora

* Upper Sidi Hakoma

• Kanapoi sample

Fig. 2.

40 50 60 70

LENGTH (mm)

Nyanzachoerus pattersoni. Length-breadth measurements on upper cheek teeth.

1978

PLIOCENE- PLEISTOCENE SUIDAE

grooved enamel and a narrow inset posterior band. However, in this specimen
there are traces of an anterior inset band as in typical Sus. The tooth is obliquely
truncated by a long and slightly wavy wear facet.

Only two specimens of P2 are known, and they are slightly smaller than in
the Kanapoi material. P3 and P4 are well represented and their morphology and
dimensions agree closely with the Kanapoi specimens. The same is true also for
the molars, as will be seen from the plots in Fig. 2. A few of the upper third
molars from Hadar are a little larger than those from Kanapoi , but the difference
does not seem to be significant. The upper cheek teeth of AL 1 37-4 are shown in
Plate 2 C.

Mandible and lower dentition

There are several good lower jaws that warrant particular description; all
are from the Lower Sidi Hakoma unit. The most complete is AL 218-2, which
corresponds in general with that of the paratype from Kanapoi, but is not quite
as massive (Plate 3 A/B). It compares very well with the specimen KNM-KP
219, illustrated in Plate 4 of Cooke and Ewer (1972). However, it is more
complete in that the entire symphysis is preserved with the two canines broken
off a little way above their bases. The four central lower incisors are present, in
very advanced wear, and the broken roots of the lateral incisors are also
preserved. The anterior pre molars are lost on both sides and there is some
damage to the teeth, which are in advanced wear. There are no significant
differences in the cheek teeth as compared with the Kanapoi material. The
incisors are somewhat curious in that they are heavily covered in cement and
show a ring of enamel surrounding a core of dentine. The same condition is
observed in the incisor teeth of Notochoerus and suggests a link between these
two genera. The symphysis is not wide but the incisors are arranged in a shallow
arc. There is only a small gap between the lateral incisor and the canine but there
is a long diastema between the back of the canine and P2. In cross-section the
canines have a blunted V-shape, which is almost heart-shaped. They emerge
from the symphysis at a very shallow angle. The symphysis itself is long, with
the lower junction level with the back of P2. The ascending ramus is broken off
and the back of the mandible seems very narrow for the massive jaw.

AL 126-8 is an excellent match for the holotype mandible in size, includ-
ing the presence of the bases of rather small canines. The incisor area has been
broken away and cannot be seen. The cheek teeth are well preserved on both
sides and in fairly early wear (Plate 2 D). Both lower P2’s are lost. The lower
P3’s have only the tips in wear and show a remarkably smooth anterior ridge
with a distinct backward slope so that the tip of the crown lies over the rear

H.B.S. COOKE

No. 29

roots. The broken canine has a flat rear surface and is U-shaped rather than
V-shaped, with enamel on all but the posterior surface.

AL 127-15 consists of a complete symphysis with the right P2 and the
whole of the left ramus with all the cheek teeth, while AL 1 27-7 is the matching
right lower third molar in a fragment of mandible. The specimen is in excellent
preservation and shows the symphysis extremely well. This is slightly larger
than in the holotype mandible and the canines are already somewhat larger than
in the previous specimen. The teeth are in early wear and it is possible that this
dentition belongs to a young male, although it is not nearly as massive as the
other specimen ascribed to a male. Five of the incisors are preserved. The two
central incisors are smaller than the second incisors, as was the case with the
holotype. The left I3 is missing, but the right I3 is present and is a very small
tooth, relatively smaller than in the holotype. The central incisors have a strong
median ridge on the upper surface and this is flanked by shallow grooves. The
second incisors have a strong ridge on the outer side of the midline and the
surface from the ridge to the medial side is a gentle slope, whereas on the outer
side there is a decided groove. The canines are of moderate size and U-shaped in
cross-section. They emerge at a strongly lateral angle, rising only gently above
the horizontal plane. The cheek teeth are in early wear. P2 is small and rather
equidimensional. P3 has the tip only just in wear and the peak lies above the
center of the posterior roots. There are no other distinctive features which differ
from previous material.

The lower cheek teeth are all well represented and the lengths and breadths
of those complete enough to be measured are shown in Figure 3. The dimen-
sions are closely comparable with the Kanapoi sample, although the third
molars are, on the average, a little longer.

Discussion

The Hadar material increases very substantially the size of the sample of
Nyanzachoerus pattersoni and adds to knowledge of some features of the
species, especially in providing better evidence for the length of the skull and
morphology of the dorsal surface. Although the third molars are slightly longer
in the Hadar sample, this cannot be regarded as an indication of evolutionary or
temporal trends. The combined Kanapoi-Hadar sample consistently displays a
lower third premolar that is shorter and stouter than in the type of N. kanamen-
sis , and M 1 is also notably longer in the latter species. In the Kanapoi sample the
diastema between the lower canine and P2 was only about 48-50 mm, compared
with 67.5 mm in the type of N. kanamensis. In the Hadar sample, three of the
lower jaws have diastemata in the range 62-64 mm, although others approxi-

BREADTH (mm) BREADTH (mm) BREADTH(mm)

1978

PLIOCENE-PLEISTOCENE SUIDAE

Nyanzachoerus pattersoni

25-> P,

20-

x* x.
A •

x /
Xx * *

15 20

• J? X

X X

XX x; . x X

x .x . x

25 30

35-

30-

25-

LENGTH (mm)

Fig. 3. Nyanzachoerus pattersoni. Length-breadth measurements on lower cheek teeth. Key as in
Fig. 2.

mate the Kanapoi range. This diminishes the difference but until the elongate,
narrow P3 of the Kanam type can be matched, it seems preferable not to
synonymize the two species.

H.B.S. COOKE

No. 29

Genus Notochoerus Broom 1925

SYNONYMY: Gerontochoerus Leakey 1943
TYPE Species: Notochoerus capensis Broom 1925

New Diagnosis: A genus of Suidae of large size, possessing hypsodont third
molars in which the main lateral pillars are strongly folded, tending to produce
dumbbell or H-shaped enamel islands, particularly in the lowers. Pre molars
reduced, with only the third and fourth premolars normally retained in the adult.
P4 with small, closely apressed, paracone and metacone that are well separated
from a small rounded protocone. Zygomatic arches robust and possessing
thimble- shaped lateral projections in the male. Upper canines dorso-ventrally
flattened, carrying a ventral enamel band, and flaring outwards in a strong flat
curve not much above the palatal plane. Mandible robust with long, wide
symphysis; anterior border almost straight and incisors small and well sepa-
rated. Lower canines heart-shaped to U-shaped, at least in early stages of
growth, and flaring laterally parallel to the uppers.

Remarks

The genotype species was founded by Broom (1925) on an isolated upper
right third molar from the Vaal River gravels, South Africa. The tooth lacks an
unknown amount from the anterior and was restored by Broom with an addi-
tional pair of laterals, making a total of five pairs (the last rather small) and six
medians. The main lateral pillars have elongate, flattened outer walls and those
on the lingual side are stouter and more complex than those on the buccal side,
which are also displaced a little towards the front of the tooth. The tooth is
moderately hypsodont, with the crown height more than one and a half times the
maximum basal breadth. Notochoerus euilus differs from N. capensis in
having fewer pairs of laterals and a lesser degree of hypsodonty.

Notochoerus euilus (Hopwood, 1926)

1926 Hylochoerus euilus Hopwood: 21; text-fig. 7; pi. 2, figs. 7-10.

1942 Hylochoerus euilus Hopwood; Dietrich: 108; pi. 17, figs. 114-126, 128,
130-136; pi. 18, fig. 149.

1958 Notochoerus euilus (Hopwood) (partim ); Leakey: 31.

1958 NON Notochoerus ( Gerontochoerus ) euilus (Hopwood); Ewer: 357;
text-fig. 11; pi. 4C, 5, 6.

1970 Notochoerus euilus (Hopwood); Cooke and Coryndon: 147; text- figs.
11, 12; pi. 7, A-E.

1978

PLIOCENE- PLEISTOCENE SUIDAE

Emended diagnosis

A Notochoerus considerably bigger than the extant Hylochoerus . Upper third
molars normally with three or four pairs of lateral pillars, lower third molars
normally with four pairs of lateral pillars, plus a small terminal pillar or
complex. Successive pairs of laterals separated by single strong median pillars;
lateral pillars well separated from each other for most of their height, producing
deep lateral valleys between them; cement well developed. Pillars strongly
folded near the tips, producing stellate enamel islands in early wear; outer
borders of lateral pillars rather flat and longer than the median lobe. Second
molars much expanded antero-posteriorly above the base, somewhat as in
Phacochoerus . Cheek teeth only moderately hypsodont, with maximum crown
height in unworn upper third molars normally less than 1.3 times anterior basal
breadth and in unworn lower third molars less than 1.5 times anterior basal
breadth.

Material

Approximately 150 specimens represent substantial parts of skulls and
mandibles, fragments, or isolated teeth that are measurable and there are
also some 50 partial teeth. Listed below is a selection of the more important
specimens in the collection.

Lower Sidi Hakoma. Damaged skull with RM13, LM3, part LM2, AL 108-3
(SH-2/3); palate with P3-M3 both sides, AL 127-46 (SH-1/2); pair of mandibu-
lar rami with LP4-M3, RM2.3, broken RP4, RM1? AL 108-2 (SH-2/3); right
mandibular ramus with RP3-M3, AL 127-13 (SH-2); right mandibular ramus
with RP4-M3, AL 126-64 (SH-2).

Upper Sidi Hakoma. Two halves of palate with LP3-M3, RP3-M2, and left
mandibular ramus with LP3-M3, young adult, AL 53-44 (SH-3); damaged skull
with LP4-M3, RP4-M3, snout lost, AL 171-1 (SH-3); left mandible with
LP3-M3 (advanced wear), AL 58-13 (SH-4/DD-2).

Denen Dora. ( DD-2 ): Fine skull, occiput slightly damaged, with P3-M3, both
sides, AL 172-1; skull, damaged at occiput and lacking front of snout, and
associated mandibles, lacking ascending rami, AL 342-9; damaged skull,
lacking snout but with most cheek teeth, AL 167-15; maxilla with LP3-M3 and
fragment with RM3, AL 158-28/29; left maxilla with LP3-M3, and right maxilla
with RP3-M3 (advanced wear), AL 330-1 ; damaged mandible with most cheek
teeth, AL 174-2; mandible with symphysis, parts of canines and most cheek
teeth, AL 174-4; damaged mandible with symphysis and incisors, AL 153-9;
partial mandible with incisor sockets, LP4-M3, RM2.3, AL 168-2; (. DD-2/3 ):

H.B.S. COOKE

No. 29

Left and right mandibular rami with LP3-M3, RP4-M2, AL 158-34; (. DD-3 ):
Partial skull, most teeth broken, AL 241-5; palate with P3-M3, both sides AL
116-93; maxilla fragments with LP3'4, LM2'3, RM3, AL 1 16-54; fine mandible,
almost complete AL 116-28 (may belong with AL 116-54); broken mandible
with nearly all the teeth, AL 168-12; pair of mandibular rami with P3-M3 both
sides, cracked, AL 185-4; right mandibular ramus with RP3-M3, AL 174-1;
broken mandible with LP3-M3, RP3-M2, partial RM3, AL 185-3; mandible,
damaged anteriorly, with LMrM3, RP4-M3, AL 378-1; damaged mandible
with both canines, LP3_4, LM3, RP4-M3, AL 157-3; anterior palate with RP3 4,
AL 317-3A.

Kada Hadar. Damaged mandible with LM2_3, RM2_3, other teeth broken, AL
164-2 (KH-1).

Description

Notochoerus is not well represented in the Sidi Hakoma member but is
abundant in the Denen Dora member, from which the best and most complete
specimens have come. Although the form and structure of the molars have been
known for some time, the present account constitutes the first description of the
skull and mandible, and of the entire dentition.

The Skull

There are five good crania in the collection, as well as several less
complete specimens or substantial parts of skulls. The most complete, which
might be regarded as a sort of informal “type,” is AL 172-1, from DD-2, which
is virtually complete except for damage at the occiput and the loss of the canines
and incisors (Plate 4). The vertex length is estimated as close to 580 mm when
intact, as compared to about 430 mm in a good adult Hylochoerus. One other
cranium, AL 108-3, from SH-2/3, appears to be roughly comparable in size,
but the other three good specimens are more robust and have an estimated
vertex length of approximately 610-620 mm. The more important mea-
surements are given in Table 2.

The overall character of the skull is not particularly like any of the living
suids, although in a general way it resembles the warthog, but without the
exaggerated elevation of the braincase and orbits. There are also resemblances
to Nyanzachoerus pattersoni and to the Kanapoi material referred to N.jaegeri
(Coppens, 1971 ; = plicatus , Cooke and Ewer, 1972). The skull table is broad,
with the frontal breadth ranging from a minimum of 1 66 mm in AL 108-3 , to an
estimated 230 mm in AL 167-15. The upper margins of the orbits are very

Measurements of Skulls of Notochoerus euilus from Hadar (in mm)

1978 PLIOCENE- PLEISTOCENE SUIDAE 17

t-~

<N O

— Tt

of co

— -i

+ +

£2co

x 2

c/o

wo O
On <N
co co
+ +

NO ON (N

NO O’ rO On

cn on

Q S

<u ^

o o

CM CM

NO WO

cj cj

88 8 8
of of oo O
of ro co cn
+ + ° °

^ s 8

NO OO CN

o o CO

ON ON

wo r-

cn — 1

<u ^
2o

I - I I 22

wo r"
oo r-

wo O

<N <N

Q j

(N NO (N| h

of ON ON 00 WO

WO Of CO —

I 2

co r-
oo wo

u ^

WO CO ON
NO NO NO
CJ — CO

t —* WO
NO CO O
of wo wo

J3 -5

e *-* D
« cd 6
JH g x>

% c
03 QJ

C 1)
C/5 ^

C/5 +-*

— v- X)

o & «

v- "O T3 3

if Ji*

O ca £ £ ^

P — w °

«_ =3 w TD

w ca ta ca

<u c y <u

o
£}

^ CUD *2
•'s >* C

g .N o „ „

CQffl££GfflOffl

I , ■£

sS 2 I

5 ° £ £ s

Diastema L-P3 175

Diastema C-P3 74

Length of premolar series 29.8 31.5 33.0

Length of molar series 111.6 119.1 129e 131.7

Length of P3-M3 141.4 150.0 164.0

H.B.S. COOKE

No. 29

slightly elevated above the top of the braincase, which is gently concave
between the eyes and flat behind them. Behind the eyes, the temporal crest
sweeps rather sharply inwards, not quite as abruptly as in the antero-posteriorly
compressed braincase of the warthog, but a little more abruptly than in the
bushpig. As in both these suids, the breadth at the occipital crest is wide, unlike
the condition in Sus. The parietal constriction is narrow and lies far back on the
braincase, quite unlike the condition in Hylochoerus . The occiput itself is
deeply scooped and resembles that of Potamochoerus porcus. There is no
median ridge (Plate 3 D).

The slight ridges that form above the orbits are continued in front of the
eyes as ridges that flank the supraorbital foramina, very much as in Nyan-
zachoerus and also in the forest hog. The nasals are straight and parallel-sided
almost to the level of the canines before tapering to their tips. The muzzle is
shaped very like that of Nyanzachoerus, with the nasals convexly curved and
joining the upper edge of the maxillae to form a fairly sharp edge that may
overhang the sides of the snout. The maxillae then curve outwards gently to the
border of the palate. Anteriorly, the lower part of the maxillae expands progres-
sively to form the sheath that surrounds the canines, in general rather like that of
the forest hog. The resemblance to the forest hog lies in the basic shape of the
flange, particularly in the form of the dorsal surface, which is broad with a low
lateral crest. However, the canine emerges in a less lateral direction, as the
alveolar margin is oblique to the axis of the skull instead of almost parallel to it.
Anteriorly the palatal part of the flange extends forwards and is continued as an
unusual feature in the form of a lateral shelf that borders the front of the snout,
including part of the premaxilla. Posteriorly, the flanges curve smoothly into
the root of the zygoma, much as in the warthog, with the infraobrital foramina at
the junction; however, the flanges do not have the tubular character of those in
Phacochoerus. On the palatal aspect, the expanded area has much the shape
seen in the forest hog but is tapered forwards by the anterior shelf described
above; this makes the premaxilla look shorter than is actually the case.

The root of the zygoma begins just behind the infraorbital foramen and
curves smoothly outwards very much as in the warthog until it is at an angle of
about 45° to the axis of the skull. It then changes direction and commences a
second, rather gentle, curve recalling the form seen in a more exaggerated
manner in the warthog. This is complicated by the presence of large thimble-
shaped lateral protrusions (or knobs) that project from the jugal area and these
tend to obscure the fact that the temporal portion of the zygoma is in a normal
position; without the ‘ ‘knobs’ ’ the zygomatic arch would be much as in a female
Hylochoerus . These curious and characteristic lateral protrusions are approxi-
mately at right angles to the axis of the skull when viewed from the dorsal

1978

PLIOCENE-PLEISTOCENE SUIDAE

aspect, but from the front or back they can be seen to droop down until their
rounded tips are close to the palatal plane (see Plates 3 C, 4). The protrusions
seem to consist of hollow bone, and when they are broken away, as is usually
the case, it is very difficult to be sure whether they existed or not. Conse-
quently, it is not certain whether they occur only in the males, as might be
expected; in one of the skulls AL 171-1, the lower border of the right zygomatic
arch seems to be intact and it appears as if no knob existed. In all the other
skulls, the survival of the outwardly flared bases of the knobs betrays their
existence.

As a result of some elevation of the back of the skull, although not nearly to
the extent seen in the warthog, the area below the orbits is broad and platelike.
In most of the specimens the area where there would normally be the hollow for
the origins of the levator rostri muscles is crushed in and seemed to consist of
very thin bone with a sinus space below it. However, two of the skulls show
that, far from being scooped out for the muscle insertions, this area was inflated
in a manner resembling that of the so-called “lacrimal bulla” in the ox. The
same feature is shown by specimens from the Omo area and appears to be a
diagnostic characteristic. The attachment areas for the depressor rostri muscles
are weakly scooped and it must be inferred that Notochoerus did not use its
snout for rooting.

One of the skulls, AL 108-3, although lacking the whole of the snout and
premolars, is important as it belongs to a young adult; the sutures in the lacrimal
area are still open and it is possible to examine the shapes of the major bones.
The lacrimal is quite narrow and the lacrimal/parietal suture is high up on the
eye socket. The orbit has a strong indentation at the antero-lower edge, often
also seen in Sus, and the lacrimal extends down into this notch. The lacrimal
then widens and runs along the upper crest line in an arch and does not
apparently come into contact with the nasals until well in front of the nasal-
frontal suture. As the skull is broken at the critical point, it is not clear whether
the lacrimal ever does come into contact with the nasals or whether it stops
short, although the latter interpretation seems the more likely. The lacrimal is
thus a rather elongate, somewhat oval- shaped bone, and it does not show any
area of depression for the attachment of the narial muscles. The suture between
the maxillary and the jugal is very far forward and lies across the bulging area
that has already been noted above. The bulge has collapsed because of the
thin nature of the bone.

This same skull has the occiput well preserved. The frontal area is flat, as
in other specimens, and the parietal is short and narrowed, much as in Sus. The
parietal constriction is narrow and Sus- like, with the braincase expanding
below it. The occiput itself is not very high and is relatively wider than in Sus,

H.B.S. COOKE

No. 29

but not as deeply scooped. The wings do not project diagonally backwards as
they do in the modem pig, but spread laterally as in the bushpig and forest hog;
there is no median ridge. When the skull is set in the occlusal plane, the occiput
is only about 20-25° off the vertical, and the surface of the temporal condyles
lies about level with the top of the foramen magnum. The temporal condyles
trend almost perpendicular to the axis of the skull and in consequence, the back
of the zygoma, where it bends up to the auditory region, is farther behind the
orbit than in Sus. The occipital condyles are not elevated abnormally above the
palatal plane and the general appearance of braincase height in the skull is
achieved by a relative increase in the elevation of the orbits, as well as by the
height of the zygomatic area of the maxilla. The ear region is also high and
wide, though not to the extent seen in the warthog, and the auditory canal points
diagonally upwards at a rather steep angle, perhaps as much as 50° from the
horizontal.

The basicranium is poorly preserved in most of the specimens. As far as
can be seen, the auditory bullae are relatively small and are strongly com-
pressed laterally. They are very narrow and pointed anteriorly and are directed
obliquely towards the center of the palatal notch. The mastoid and paramastoid
processes are strong and the paramastoids very stout (and probably also long).

The palate has a general resemblance in shape to that of the forest hog, no
doubt dictated by the need to accommodate the extremely large canines. The
premaxilla is relatively narrower than in the forest hog and resembles to some
extent the corresponding region in the warthog. However, the border of the
palate curves smoothly into the expanded area in front of the canines and does
not show the abrupt change of angle or notch seen in most other suids (Plate 4).
Only one pair of incisors was present and there is no trace of sockets for other
incisors. The teeth themselves are not known but the sockets indicate that they
were relatively smaller than in the forest hog and comparable with those in the
warthog. The anterior palatal foramina are of moderate length and are laterally
compressed. The canines emerge very much as in forest hog or warthog, but lie
very nearly in the palatal plane. Behind this expanded area, the palate narrows
gradually to the point where the canine sockets end and the palate is narrowest
just in front of the anterior premolar (normally P3). The posterior palatal
foramina lie about opposite the midpoint of the third molars and from them the
usual blood vessel grooves run forwards, diverging at the level of the canines
and then converging again to their termination at the anterior palatal foramina.
The palatine notch lies well behind the back of the third molars.

The canines are large and fairly strongly curved, comprising an arc that
may be as much as one third of a circle, as is also the case in the male forest hog
or warthog. However, their orientation is different as they extend horizontally

1978

PLIOCENE- PLEISTOCENE SUIDAE

almost in the plane of the palate, with only a small rise at the tips. In
cross-section they are unusual as the normal suine structure has been distorted
by oblique flattening, with the dorso- ventral dimension only about two thirds to
three quarters of the antero- posterior one. On the dorsal surface there is a broad
shallow groove and on the antero -ventral surface there is a strongly marked
groove that lies almost directly below the dorsal one, making the cross-section
rather like a figure 8 on its side, with the smaller loop anteriorly. The posterior
face is weakly grooved, thus turning the “lazy 8” into a form that can best be
likened to an inverted squashed bell shape (see Figure 4). The ventral band is

Do r s a I

Vent ra I

Fig. 4. Outline cross-sections of upper right canine (top) and lower left canines (below) of
Notocherus euilus. The top and lower right drawings are taken near the alveolar margin; the lower
left drawing is near the tip. Natural size.

made of ribbed enamel and there are narrow inset bands of enamel on the
antero-dorsal and postero-dorsal “edges” of the tooth. However, in some
specimens the whole tooth is longitudinally ribbed, although only the ventral
part has enamel.

H.B.S. COOKE

No. 29

In all but one of the skulls and palates, there is no sign of an upper P2 and
no scars that might represent P1 are found in the fairly long diastema between
the canine and P3. The same is true of the mandibles and it is clear that only P3/3
and P4/4 are normally retained in the adult. However, AL 342-9 (from DD-1/2)
consists of a good skull and associated mandible, a little distorted by oblique
pressure, and it is a young adult with the third molars in wear anteriorly but not
fully erupted posteriorly. The P2 is retained on both sides in the lower jaw but
the palate is broken across just in front of the upper P3’s and the presence or
absence of P2 cannot be stated with certainty. If P2 was present, as seems likely,
it must have been separated from P3 by a gap of at least a few millimetres.

In this skull, AL 342-9, the bone is badly cracked and postmortem distor-
tion has separated the teeth so that the contact facets between them are now from
2 to 8 mm apart (RM3 has been crushed back into its socket and is 19 mm from
the back of RM2). The P3 and P4 in this specimen are in early wear and are fairly
typical. P3 is a subtriangular tooth comprising two external elements (paracone
and metacone), lying obliquely to the axis of the tooth row so that the paracone
is on the midline. The postero-internal comer is buttressed by a strong oval
protocone, separated from the paracone/metacone by a wide and deep fovea,
open on the antero-internal side. The back of the fovea is closed by one or two
high cingulum cusps. Anteriorly there is a moderately low cingulum band. P4 is
rounded in outline and the crown is about as high as the transverse diameter.
The external face shows only the faintest sign of a groove that divides the crown
at the tip into paracone and metacone islands. The protocone is rounded and
relatively small and is separated from the paracone/metacone by a narrow but
well developed transverse valley, closed off posteriorly by a high but thin
cingulum. There is a modest anterior cingulum. From the transverse valley, a
distinct indentation marks the division between paracone and metacone and
there is a matching shallow groove in the protocone, giving a slight + -like form
to the valley. In other specimens, the paracone/metacone in P3 may appear
virtually as a single cone. In P4 the area in the center of the valley may form an
isolated pit as advancing wear of the anterior and posterior cingula closes off the
valley. There is also a tendency for the three cones to become more columnar in
some specimens.

In most of the skulls, even in the young adult AL 342-9, the upper first
molar is very worn and the pattern is difficult to discern. It soon becomes a
stump of dentine and M2 moves forward as M3 erupts behind it, much as is the
case with the warthog. In teeth in early wear, M1 is seen to be formed of two
well-developed pairs of laterals, with a single median between them, a weak
anterior cingulum complex, and a modest posterior cingulum complex. The

1978

PLIOCENE- PLEISTOCENE SUIDAE

tooth is longest in early wear and becomes almost equidimensional as wear
proceeds.

The upper second molar is moderately high crowned and has its maximum
length in early wear, when only the tips of the enamel pillars have been abraded.
From the base, the anterior face rises forward and the back bulges upward,
much as is the case with the more hypsodont second molars of the warthog
(Fig. 5). The third molar is initially “tucked in” under the greatly expanded

Fig. 5. Sketches to show the lateral aspects of upper (top) and lower (bottom) second molars of
Notocherus euilus, indicating the substantial posterior bulge of the large cingulum complex.
Natural size.

posterior cingulum complex. The crown length thus decreases with advancing
wear and the posterior complex shortens, but the second molar remains an elon-
gate tooth even in advanced wear. As far as possible, measurements are taken
at the base of the crown so as to facilitate comparison by minimizing changes
in the occlusal length with age. The basic structure of the tooth is like that of M1
but the cingula are better developed, especially the posterior one. The form of
the lateral pillars is essentially like the anterior part of the M3.

The upper third molars consist essentially of three well-developed pairs of
large lateral pillars, followed by a terminal complex that ranges from three
small pillars to a good fourth pair, the former being more normal. Very often
there is a well-developed fourth lateral pillar on the external (buccal) side but
only a small one on the lingual side and the lingual pillars are not level with their

H.B.S. COOKE

No. 29

mates but are displaced a little posteriorly. The space between the first and
second laterals is wider on the buccal side than on the lingual one. A single
median pillar lies between the first and second pairs of laterals, but there may be
one or two medians between the second and third pairs. The first and second
pairs of laterals are in contact at the midline but the third and fourth pairs may be
separated by the partial intervention of the medians. There is an anterior
cingulum complex, broken up into several small columns, the center one being
almost the size of the medians and forming a wedge between the front pair of
laterals. There is usually a stout central terminal pillar at the back of the tooth,
but this may sometimes be displaced towards the buccal side and be accom-
panied by an additional small terminal pillar. In early wear, the lateral pillars
show stellate enamel islands, but as wear proceeds, two of the grooves become
dominant and divide the pillar into a lateral lobe and a median lobe. The lateral
lobe has a relatively smoother outer rim of thicker enamel than is the case in the
remainder of the island, although there is sometimes a thin external groove or
indentation in the wall of the lateral pillar. The median lobe is less regular and
retains some of the stellate appearance by having three or even four “points”
initially; but with wear the grooves diminish and the enamel island as a whole
becomes progressively more like a dumbbell or H- shape, although minor
irregularities may persist on the median lobes, especially in the lingual pillars.
The H- shape is best developed at the front of the crown and tends to become
distorted in the columns nearer the back, with the median arm of the H displaced
forwards . In general, the median area of the H tends to be a little shorter than the
lateral one. Minor folds diminish with advancing wear and the pattern becomes
very simple. The central, or median, pillars also form stellate enamel islands
initially but these become oval or irregularly oval as wear proceeds. Mea-
surements are plotted in Fig. 6 and examples are illustrated in Plates 5 and 6.

Another feature of variation within the Hadar sample is in the degree of
upward taper of the crown from the base to the occlusal surface. In general, the
occlusal breadth is a good deal less than the maximum breadth of the tooth at the
base of the enamel. As the crown height of unworn third molars is not much
greater than the anterior basal breadth, the lower crowned variants tend to have
a more or less trapezoidal transverse cross-section at the front of the tooth, the
broader the base the greater the tendency. In the narrower molars, the sides of
the crown are more nearly vertical, although there is always some degree of
upward narrowing of the crown. Some individual teeth may resemble the
molars of Nyanzachoerus jaegeri but are usually distinguishable by the greater
development of the talon ( N . jaegeri has only three pairs of lateral pillars, the
third often weak) and by the shape of the enamel islands, which in N. jaegeri
are more M- or W-shaped than H-shaped. If premolars are present in the

1978

PLIOCENE- PLEISTOCENE SUIDAE

Notochoerus euilus
_ 2

20-

15-

▲

15-

io-

— T-. IO'

% x

20 IO

35-

20-

KEY

v Kada Hodar
a Denen Dora
x Upper Sidi Hakoma
a Lower Sidi Hakoma

60 70 80

LENGTH (mm)

Fig. 6. Notochoerus euilus. Length-breadth measurements on upper cheek teeth.

specimen, the P3 is substantially smaller and P4 somewhat smaller in
Notochoerus euilus. However, the morphological resemblances suggest a close
genetic relationship between N. euilus and Nyanzachoerus jaegeri, and there
may be some risk of misidentification in certain isolated specimens.

H.B.S. COOKE

No. 29

The Mandible

The lower jaw is represented by several specimens that are reasonably
complete except for the invariable loss of the ascending rami. AL 1 16-28 (from
DD-2/3) is probably the best specimen, with the incisors and canines preserved
on both sides and a full complement of cheek teeth (Plate 7). The symphysis is
long, although there are other specimens that are longer, and it is very wide and
spatulate across the canines, very much as in the forest hog. The anterior border
is almost straight — even more so than in forest hog — and all six incisors are
present (Plate 8 A). They are very much reduced, almost peglike in character,
and are strengthened by abundant cementum. The lateral incisor is an absurdly
small tooth for a jaw of this size and it looks almost like a first premolar,
embedded in a thick shaft of cement. The central incisors are very worn and
were probably smaller and shorter initially than those in a modern domestic pig,
(the skull of which is only half as large as that of Notochoerus)\ only the bases
remain as stumps, about the size and shape of a human incisor and thickly
covered with cement. The second incisors are morphologically like shortened
Sus teeth, of which only the bases remain. Only a few other lower incisors are
known, including an RI, belonging to the associated skull and mandible AL
342-9 (DD-2) which is a young adult so that the incisor is very little worn. The
length of enamel on the anterior (lower) surface is only 14.5 mm and on the
upper surface is 22.3 mm along the broad median ridge, while the transverse
diameter at the base of the enamel is 8 . 0 mm and the antero-posterior dimension
at this position is 9.3 mm. This is a little more robust than in Sus scrofa or
Potamochoerus porcus, but very much shorter. The median posterior rib is
broader and less marked than in Sus and the morphology is more like the I2 than
I j in the living form. It is more nearly comparable with the corresponding tooth
in the forest hog. In this specimen the heavy cement coating is not present, but
may have been lost. The whole symphysis in this specimen is very like that of
AL 116-28, as also is AL 153-9 (DD-2) in which the broken roots of the four
central incisors confirm the small size of these teeth and the two tiny lateral
incisors are preserved, with their premolarlike morphology. Ventrally, the
symphysis is also rather like that of the forest hog.

In AL 116-28 the canines are intact and emerge from the symphysis at a
very low angle, not much above the occlusal plane. They are robust and
essentially flattened oval in shape, measuring 42 mm by 27 mm at the alveolar
margin, with the long axis antero- posteriorly oriented. The surface is covered
with thin enamel, except on a slightly flattened posterior surface. The teeth are
fairly sharply curved and carry only a very small anterior wear facet that may be
due to use rather than to attrition against the uppers (Plate 8). Other specimens
confirm both the form and the attitude of the canines, although in AL 174-4

1978

PLIOCENE-PLEISTOCENE SUIDAE

(DD-2) there is no enamel on the upper face or on the flattened area at the back
of the tooth.

In AL 116-28, and in most of the other specimens that include canines,
there is a long diastema between the back of the canine and the anterior
premolar, which is normally P3. This tooth lies just about level with the junction
of the two halves of the symphysis. In this specimen there is no trace of P2 or of
a healed scar, despite the excellent state of the bone. The same is true of other
specimens, although in most of them the bone is damaged and a scar might
easily be missed. However, in the mandible of the young adult, AL 342-9, P2 is
present on both sides. It is a very small tooth, about half the length of the P3, and
is relatively wide. It consists of a single stout cone with a smooth anterior face,
but having a thin ridge running to the back of the crown. The disparity in size
between the P2 and the P3 is reminiscent of the relationship found in Nyan-
zachoerus. The P2 is a little stouter than P2 in Sus scrofa, but is of comparable
absolute size, whereas P3 is 50% larger.

The P3 is represented by a good number of specimens, for which length-
breadth plots are given in Figure 7. Also shown are plots for P3 in Nyan-
zachoerus jaegeri and it will be seen that this tooth is considerably smaller in
Notochoerus euilus. However, two specimens from the Sidi Hakoma unit are of
comparable size. AL 124-5, from the lower Sidi Hakoma, consists only of P3
and P4 and its identity is not certain. The other, AL 174-1, comes from the
upper Sidi Hakoma, but the third molars are larger than those of Nyanzachoerus
jaegeri and their morphology is typical for Notochoerus , and its length (82 mm)
suggests that it is a large N. euilus , although this is not certain. Morphologi-
cally, the P3 is moderately elongate and comprises a stout single cone with a
smooth anterior face that slopes backwards so that the tip lies over the front of
the rear root. There is a well-formed posterior cingulum, half the height of the
crown, marked off from the main cone by a weak groove lingually and a
stronger one buccally . The P4 is wider than the P3 but often slightly shorter, so
that it appears almost equidimensional. It also consists of a stout main cone,
with the tip practically central, but there is a well-formed, fairly narrow,
anterior cingulum and a strong posterior cingulum, or complex, that extends at
least three quarters of the height of the crown, much as in Nyanzachoerus . In
some specimens, especially those in early wear, the tip of the main cone is
divided into two elements, the anterior one slightly on the buccal side of the
midline, and the posterior one slightly on the lingual side. Because of this slight
asymmetry, the posterior cingulum is marked off from the main cone by deep
lateral grooves, of which that on the buccal side is the deeper, whereas weaker
grooves flank this anterior cingulum and the deeper lies on the lingual side.
There is a distinct anterior ridge in most specimens, but it disappears with wear
and an anterior cingulum pillar may appear.

BREADTH (mm) BREADTH (mm) BREADTH(mm)

H.B.S. COOKE

No. 29

Notochoerus euilus

15-

A * * X

J a □

. aa A f

AA a a a

> i 7”" t- i — i rn r

20 25

Fig. 7. Notochoerus euilus. Length-breadth measurements on lower cheek teeth. Key as in Figure
6. Data for the P3 of Nyanzachoerus jaeger i and for the Kaiso type of Notochoerus euilus are also
shown.

The first molar is usually worn to dentine with remnants of enamel but its
structure is essentially like that of M2, except that the posterior cingulum is

1978

PLIOCENE- PLEISTOCENE SUIDAE

weaker. In early wear it is a narrow, elongate tooth, but it broadens and shortens
with advancing wear. Dimensions are given in Figure 7.

The M2 is an elongate tooth with a well-developed posterior cingulum
complex comprising a median pillar and two posterior columns that look almost
like an incipient third pair of laterals. There is also commonly an accessory
basal pillar on the buccal side behind the second external lateral pillar. There is
a weak anterior cingulum complex but it is obliterated when the tooth is only
half worn. The first paired laterals are usually not quite mirror images, the
lingual one being the larger and wearing to an enamel island that may be more
X-shaped than H-shaped, while the buccal one sometimes becomes mushroom
shaped through lack of the median arm of the H. The second pair of laterals is
usually more symmetrical and the islands approximate an H-shape, although
the median arm is normally shorter than the lateral one and there is a distinct
indentation, or groove, on the median side. The lateral pillars may be angled
slightly forwards relative to the base of the crown, when viewed from the side.
The anterior and posterior complexes bulge outwards, somewhat in the manner
seen in the more hyposodont teeth of the warthog, so the maximum crown
length is reached when only about one-fifth of the original height has been worn
away (Fig. 5). The anterior complex of M3 fits under the posterior bulge of M2
and comes into wear after M2 has passed its maximum length. The M3 rotates
into position by completion and elevation of the posterior columns and as it
grows and abrades, the space between it and the back of P4 diminishes
progressively until the whole of M3 is in full wear. From a functional view-
point, the effect is to maintain a reasonably consistent length of molar grinding
surface from the young adult, with only the front half of M3 in wear, to the older
adult, with Mx-M2 shortened, but the whole of M3 in use.

The lower third molar normally consists of four well-developed pairs of
lateral columns and a terminal complex consisting of three columns, two of
which form an incipient fifth pair of laterals; in some specimens these two
columns are sufficiently developed to constitute a fifth pair of laterals. Each
pair of laterals is well separated from the next pair by wide lateral grooves and
by a stout median pillar. As in the uppers, the laterals form moderately stellate
enamel islands in early wear, although dominant anterior and posterior clefts
soon divide the pillars into external and median lobes. On the first two pairs, an
H-shaped enamel island develops with further wear, initially with the two lobes
of the island sub equal in size but, as wear proceeds, the lateral lobe may
become relatively longer than the median one. On the succeeding pairs, the
stellate form persists longer and the H-shape is less obvious. In some teeth, the
posterior cleft may dominate so that the shape becomes almost an inverted U.
The disparity between basal breadth and occlusal breadth is not as great as in the

H.B.S. COOKE

No. 29

uppers, so that the sides of the crown do not taper markedly. The greatest
breadth at the base of the crown is often at the second pair of laterals rather than
the first. The median pillars form more or less equidimensional enamel islands,
initially stellate but becoming oval with wear. The anterior complex is weakly
developed. The hypsodonty is moderate with the maximum crown height less
than 1.5 times the maximum basal breadth. Length/breadth ratios are plotted in
Figure 7 and examples of the lower dentition are shown in Plates 7 and 8.

Discussion

The Hadar collection provides a substantial sample of material that can be
ascribed with some confidence to Notochoerus euilus as the features of the
original syntypes from Kaiso fall well within the range of variation encountered
in the Hadar material. If the reconstruction of the holotype RM3 by Cooke and
Coryndon (1970) is approximately correct, it would lie about in the middle of
the range observed for M3 (see Fig. 7). However, those authors considered that
material from Laetolil, in Tanzania, differed from the Kaiso species and they
referred it tentatively to Notochoerus cf. capensis. The writer has re-examined
this material in East Berlin and, in the light of this study and the variations
encountered in the Hadar sample, it is concluded that Dietrich’s (1942) original
identification as ‘ ‘ Hylochoerus" euilus should stand.

The Hadar sample, although large, comes mainly from the Denen Dora
member of the succession and the samples from the other members are rather
inadequate for statistical comparisons to be useful. In general, the range for the
molars from the Denen Dora would include all the samples from the earlier and
later horizons, the only exception being an unusually small M3 from the upper
Sidi Hakoma member. However, for the upper premolars this is less true and
there is a suspicion that the specimens from the Sidi Hakoma member may be
larger than those from the Denen Dora. From a morphological aspect, the
premolars from the Sidi Hakoma are more robust and Nyanzachoerus-likc than
those from the higher members, and there seems to be a change towards higher
crowned and more delicate pre molars in the upper part of the sequence. The
frequency of occurrence of third molars with a strong upward tapering of the
crowns also diminishes in the successively higher horizons and the molar
structure seems to become more stable and “typical.”

From the Usno Formation in the Omo area of southwestern Ethiopia, there
is a large sample of material assigned to Notochoerus euilus by the present
writer but the description is still unpublished, apart from some measurements
on the third molars (Cooke, 1976). The range of size variation in the third
molars in the Usno sample is almost identical with that for the Hadar material.

1978

PLIOCENE- PLEISTOCENE SUIDAE

However, in the Hadar specimens the enamel is generally thick and does not
show the thinning in the folded portions of the lobes that is seen in the Usno
material. As a result, the Hadar teeth seem to be less complex in their folding,
even in the early stages of wear.

The structural features of the dentition suggest that Notochoerus is related
to Nyanzachoerus and is closest to N. jaegeri. The premolars are further
reduced in size compared with the latter, but the disparity between P2 and P3
remains as evidence for the trend and relationship. The third molars are further
enlarged, as compared with N. jaegeri , thus increasing the grinding area, but
they remain relatively low crowned.

Genus Kolpochoerus E. C. N. and H. E. van Hoepen 1932

SYNONOMY: Mesochoerus Shaw and Cooke 1941; Omochoerus Arambourg
1942; Promesochoerus Leakey 1965; Ectopotamochoerus Leakey 1965.
Type Species: Kolpochoerus paiceae (Broom 1931) (syn. K. sinuosus E. C. N.
and H. E. van Hoepen 1932).

Diagnosis: Suidae of moderate to large size with skull architecture generally
resembling that of Potamochoerus in early forms, but zygoma expanded
laterally and drooping, especially in the male. Male canines resemble those of
Hylochoerus in structure and cross-section, but relatively shorter and stouter;
female canines much smaller than in the male and primitively rooted in some
species. Cheek teeth resemble those of Sus or Potamochoerus in general
structure but molars higher crowned and have lateral columns that are distinct
and well separated. Talon of third molar tends to become more developed than
in Sus scrofa or Potamochoerus porcus, exceeding the length of the main body
of the crown in advanced species. Molar brachydont or moderately hypsodont,
always strongly rooted. Little cement in more brachydont forms, abundant in
hypsodont molars. Premolars rather moreSws-like than Potamochoerus -like; P2
and P3 triangular with well-developed protocone; P4 equidimensional with a
strong protocone and tendency toward the development of multituberculate and
complex islands; P4 has elevated anterior and posterior cingulum cusps and a
double central cusp with the two elements displaced laterally relative to one
another.

Remarks. Kolpochoerus is widely distributed in the later Pliocene and Pleis-
tocene, ranging from South Africa to North Africa. This generic name has
priority over Mesochoerus , which has been used in most of the literature. The
type species is one of the most advanced in the genus.

H.B.S. COOKE

No. 29

Kolpochoerus afarensis sp. nov.

Diagnosis

A suine about the size of the living bush pig, recognizable as Kol-
pochoerus by the structure of the zygomatic arch; differing from Pota-
mochoerus porcus in possessing premolars that are less reduced and
morphologically somewhat intermediate between premolars of bush pig and
wild boar; P1 normally, and Pj sometimes retained in adult; lower canines
verrucose; third molars relatively larger than in bush pig, tending to be more
columnar and with smoother enamel on outer faces of main pillars; distin-
guished from Kolpochoerus limnetes by smaller size and by possession of only
two pairs of fully developed laterals in the lower third molars.

Holotype. AL 147-10 in the Ethiopian National Museum, Addis Ababa,
cranium with occiput and basicranium damaged, zygomatic arches and tips of
premaxillae lost, P3-M3 present on both sides in early wear; some lateral
distortion and crushing of the cranium.

Locality: Hadar, Ethiopia

Hypodigm: SlDI Hakoma. Type (SH-2); palate of juvenile with erupting
canines, RP1, RM1 2, LM1'2 intact, AL 224-3 (SH- 1/2); maxilla fragment with
RP4-M2, AL 222-4 (SH-1/3); maxilla fragment with RM2'3, AL 218-1 (SH-1/
3); isolated RM3, AL 200-13 (7SH-1); isolated RP4, AL 165-12 (SH-2).
Mandibular ramus with base of canine, LP4, LM2.3, roots LP^, LMb AL
109-1 (SH-2); mandibular ramus with roots of LP3, LM, broken, LP4, LM2.3
intact, AL 127-38 (SH-1); mandibular ramus with LM2, partial LM3, broken
LP4-M2, A1 165-5 (SH-2); mandibular ramus with LMj.3, AL 125-4 (SH-1/2);
mandibular fragment with damaged LM,-LM3, AL 217-3 (SH-1/3); mandibu-
lar fragment with RdP4-M2, AL 248-3 (SH-2); mandibular fragment with
LMj_2, AL 214-1 (SH-4); mandibular fragment with LM2_3, AL 148-101
(SH-1); mandibular fragments with M,: AL 147-24 (SH-2), AL 266-3 (SH-3),
AL 277-8 (SH-2/3); mandibular fragments with M3: AL 131-4 (SH-2), AL
229-3 (SH-3), AL 199-2 (7SH-1), AL 251-34 (SHT), AL 255-1 (SH-1/2), AL
165-4 (SH-2), AL 327-19 (7SH-2), AL 325-9 (SH-4), AL 259-2 (SH-2);
isolated M3, AL 233-3 (SH-4).

Referred Material: Denen Dora. Partial cranium, lacking snout and occiput
but with zygomatic arches, teeth broken, AL 154-34 (DD-2); pair of maxillae
with LP3-M2, RP4-M2, AL 116-1 (DD-3); maxilla fragment with RP3-MX, AL
118-7 (DD-3); maxilla fragment with LP4-M3, AL 385-2B (DD-3); palate with
P4-M3 both sides and two fragments of mandible with M2_3 both sides, AL
56-16 (7DD-2/3); maxilla fragments: with P4-M!, AL 332-35 (DD-1), AL

1978

PLIOCENE- PLEISTOCENE SUIDAE

358-12 (7DD-2); and M1'2, AL 1 16-114 (DD-2); with M2'3, AL 287-2 (DD-1);
isolated LP4, AL 342-13 (DD-2); isolated M1, AL 116-42 (DD-2/3); isolated
M3, AL 116-7 (DD-2), AL 201-1B (DD-2), AL 220-3 (DD-2). Incomplete
mandible with roots of incisors and canine, LP3-M3, parts of right ramus with
RM2-3, AL 134-7 (DD-2); symphysis with base of left canine, AL 220-2
(DD-2/3); symphysis, right ramus with RM2.3, anterior teeth broken, left
mandibular fragment with LM2.3, AL 321-10 (DD-3); mandibular ramus with
LP3-M3, AL 385-2A (DD-3); mandibular ramus with RP4-M3, AL 168-13
(DD-2); symphysis fragment with RI, , LIj.2, left ramus with LP2-M3, part right
ramus with RP2.4, AL 186-20 (DD-3); juvenile mandibular ramus with RP4-
M2, M3 in crypt, AL 185-20 (DD-3); mandibular fragments with LM3, RM,_3,
AL 116-60 (DD-3); mandibular ramus with RP4-M3, AL 116-15 (DD-3);
mandibular ramus with RP4-M3, AL 183-44 (DD-3); mandibular fragment with
LMj.3, AL 169-16 (DD-2); mandibular fragments: with M2.3, AL 379-3 (DD-
2/3), AL 1 18-6 (DD-2/3), AL 291-10B (DD-3); with M2, AL 302-4 (DD-2),
AL 169-19 (DD-2); with M3, AL 169-18 (DD-2), AL 187-12 (DD-3), AL
309-2 (DD-3).

Kada Hadar. Symphysis with four incisors and both canines, parts of
rami with damaged LP3_4, RP3, roots of RP4-M2, damaged RM3, AL 157-2
(KH-1); mandibular fragment with LP3-M3, roots LP2, AL 359-1 (KH-2);
mandible fragment with RdP4, AL 367-1 (KH-2/3).

Hadar South. Maxilla fragments with LM2 3, RM3, AL 273- 1 ; mandible
fragment with RdP4, RMl9 and erupting RM2, AL 273-2.

Uncertain Horizon. Isolated M3, AL 246-5.

Description

The holotype comes from the middle part of the Sidi Hakoma member but
unfortunately there are few other specimens of the upper dentition from this
stratigraphic unit. The material from the Denen Dora unit seems to be a little
larger, but the sample from the Sidi Hakoma is too small for it to be clear
whether this is a real trend or a chance variation. The holotype lies near the
lower limit for the entire sample but the morphology is characteristic. The
samples for the lower dentition are bigger and show only a small trend towards
larger size.

The Holotype

The holotype is an incomplete cranium from site AL 147, which also

H.B.S. COOKE

No. 29

furnished one isolated lower first molar. The cranium has lost the tip of the
snout in front of the third incisors and also the upper part of the occiput, the
occipital condyles, much of the auditory bullae and paramastoid processes, and
most of the zygomatic arches. The specimen is deformed by pressure from the
right side so that the midline of the skull is now almost directly above the left
tooth row. In part, this has been accomplished by shearing of the maxilla and
translocation of the snout and frontal area with relatively little actual distortion
of the facial bones, although there is a mosaic of fracturing. The maxillary root
of the right zygoma is partly preserved. There is also some backward compres-
sion of the skull. The palate is virtually intact and P3-M3 are present on both sides
and almost complete, except for slight damage to RM1 and LM3. The third
molars are still erupting and the main pillars are just coming into wear, so that
the dentition is in an almost ideal state. Measurements on the skull are necessar-
ily a little speculative but some are given in Table 3, together with reliable
measurements on the cheek teeth. The skull and teeth are illustrated in Plates 9
and 10 A.

TABLE 3

Measurements on Holotype Cranium of Kolpochoerus afarensis (in. mm)

AL 147-10

Vertex length
Nasal breadth
Muzzle breadth
Ocular breadth
Frontal breadth
Posterior palatal breadth
Posterior maxillary breadth
Anterior palatal breadth

+ 296 (c340-350e)
c51
c55
c75
cllO
33.5
80.0
c41e

length

breadth

height

LP3

14.7

9.8

10.5+

LP4

12.7

13.6

10.0+

T A ..occlusal
LM basal

18.6

15.4

15.2

9.5 +

T x k ‘i occlusal
LM basal

23.9

12.7

21.0

19.2

12.5 +

LM3

34.3

20.2

16.0e

length

breadth

height

RP3

14.0

10.0

11.5 +

RP4

12.8

13.2

9.5 +

occlusal
RM basal

18.2

14.9

15.1

7.0 +

occlusal
RM basal

24.1

13.0

21.4

19.4

12.5 +

RM3

33.9

19.6

16.0e+

In overall size and general morphology, the holotype cranium is very
similar to the corresponding parts of the skull in Sus scrofa or Potamochoerus
porcus and, in the absence of the zygomatic arches, might belong to either

1978

PLIOCENE- PLEISTOCENE SUIDAE

genus. The nasals are moderately wide and appear to have been gently arched
and to have curved smoothly into the side walls of the muzzle. The supraorbital
foramina are about in line with the front of the orbits, and the canals leading
from these are moderately deep. The orbits were well below the frontal surface,
which was gently arched, and the frontal breadth is normal. The parietal
constriction is narrow, being substantially less than the breadth of the snout, as
is common in Potamochoerus whereas in S us it is usually wider than the snout.
The braincase probably bulged outwards below the parietal crest, as in the
bush pig, but it is too crushed for this to be certain. The occiput was apparently
narrow, but this area is too badly damaged for details to be seen. However, the
squamous temporal bone was clearly broad at the base, with the external
auditory meatus fairly low down, as in Potamochoerus pore us, not narrow and
rather high as in Sus scrofa. The infraorbital foramina lie above the junction
between P4 and M1; in Sus scrofa they are normally above P4 and in
Potamochoerus porcus above M1. The maxillary root of the zygoma has been
pushed back into the orbit, but it would seem that there was only a moderately
scooped area for the levator rostri attachments, not as deep as in typical Sus.
The remnants of the auditory bullae and the paramastoid process appear
normal.

The front of the premaxilla is broken through the sockets of the second left
and third right incisors, but it is clear that the front of the snout was moderately
long and narrow, much as in Sus scrofa. The left canine has been broken off
with part of the bone, but the form of the flange is clearly seen and is small with
an antero- posterior measurement of 15.3 mm and a transverse measurement of
8.9 mm. It is laterally more compressed than in a bush pig of the same age and
like that of Kolpochoerus limnetes, but smaller. There is a good deal of
distortion, but it would seem that the tusk pointed downwards much as in
females of the wild boar and did not curve sideways even to the extent seen in
modem bush pig females. There is a gap of approximately 10 mm between the
front of the canine and the back of the socket for LI3, but crushing makes this
measurement unreliable. About 1 cm behind the canine is the front of the
anterior root of LP2, but it is not possible to determine whether P1 was present or
not, as this area is distorted and damaged. Judging by the roots, P2 was about 1 2
mm long.

Of the cheek teeth, only P3-M3 of both sides are known in the holotype.
Measurements are given in Table 3 and length/breadth ratios are plotted in Fig.
8. The dimensions are closely comparable with those of Sus scrofa. Compared
with Potamochoerus porcus, the third molars are larger than is usual, but the
other teeth are similar in size. Morphologically, they are somewhat inter-
mediate in character, but are less columnar than in Kolpochoerus limnetes.

H.B.S. COOKE

No. 29

Kolpochoerus afarensis

L±J

IO 15 20 25

KEY

v Kada Hadar
a Denen Dora
x Upper Sidi Hakoma
□ Lower Sidi Hakoma

* Hadar South

1 0— f — i — i — i — i — i — i — i — i — i — i — i — . — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i

20 30 40 50

LENGTH (mm)

Fig. 8. Kolpochoerus afarensis. Length-breadth measurements on upper cheek teeth. The circled
measurement is for an unusually large specimen.

P3 is a sub-triangular tooth with the main cone (paracone) developed more
strongly than in Sus scrofa, but not as robust as in typical bush pig. As in the
latter, the anterior cingulum is not very well developed, but is strong enough for

1978

PLIOCENE- PLEISTOCENE SUIDAE

an antero- internal fossa to develop in Sus- like fashion, though weaker. The
posterior cingulum is also lower than in Sus and the postero-intemal cusp
(metaconule) is small and rather isolated, but does not lead to the formation of a
strong fovea in front of it, as is usual in the bush pig. The main cone is linked to
the cingulum by well-developed ridges, making the tooth more sectorial than in
bush pig at the same dental age. This structure is normal in Kolpochoerus .

P4 has the two outer main cusps (paracone and metacone) well developed
and well separated, as in Sms scrofa , and the paracone is also somewhat larger
than the metacone, whereas in the bush pig the two cones are roughly equal in
size and are closer together. The protocone is large and antero- posteriorly
elongate, and lacks any trace of the median spur that occurs typically in Sus
scrofa, but not normally in the bush pig. In the former a wide and persistent
fossa separates the paracone/metacone ridge from the protocone, but a narrow
spur from the back part of the paracone juts into the fossa and tends to form two
foveae, of which the anterior one is open to the front as the anterior cingulum
does not block it. In the bush pig the fossa is narrow but the anterior cingulum
extends partly around the protocone and blocks off the anterior fovea, while the
posterior fovea is blocked by a strong cingulum that is tied to the back of the
paracone. In the fossil, the paracone has a strong median spur that meets the
front of the protocone and there are well developed cingulum cusps at the back
of the crown that block the fossa. The result is the formation of a deep, wide,
conical fovea that looks very like a volcanic crater. The anterior cingulum
extends part way around the protocone, as in the bush pig, and a very small
fovea is formed between it and the paracone and protocone. The cingulum
development has resemblances to the bush pig condition, but the isolated
protocone and very large fovea is more like the wide fossa in Sus scrofa. The
distinctive structure of later Kolpochoerus is not yet developed.

M1 and M2 are very similar to the corresponding teeth of the wild boar and
of the bush pig, which can be difficult to distinguish from one another. In Sus
scrofa the enamel is thinner than in Potamochoerus porcus, and in the earlier
stages of wear the clefts (or furrows of Hunermann, 1968) in the main pillars are
wider and more open in Sus than in Potamochoerus , where they are hairline
cracks. In the fossil teeth they conform to the latter rather than to|he former.
The successive pairs of laterals are more widely spaced than in Sms scrofa. The
latter also tends to have the space between the first an’d seeOnd lateral columns
on the inner (lingual) side filled by well-developed 'accessory basal pillars,
which are less strongly formed in the bush pig. The fossil has only weak basal
pillars on the inner side but does possess a single short basal pillar on the
external (buccal) side between the well-spaced laterals; this sometimes occurs
in second molars both in the wild boar and in the bush pig.

H.B.S. COOKE

No. 29

The third molars are relatively larger than in the bush pig but match closely
in dimensions with the wild boar. There are two pairs of laterals and a single
stout terminal pillar, as in both these living suids, although in both the talon may
sometimes be more complex. As was the case with the first and second molars,
the main pillars are smoother and less plicate than in Sus scrofa, and are also
more distinctly separated and columnar than in the wild boar; indeed they are
also rather simpler and better separated even than in living Potamochoerus
porcus , presenting a less crowded appearance to the crown, as might be
expected in a very early stage of Kolpochoerus .

Other cranial material

AL 224-3, from SH-1/2, is an important specimen as it provides informa-
tion about the anterior cheek teeth. It is an incomplete palate, with a narrow rim
of maxilla, and is broken across in front of the canines and through the palatal
notch. The third molars are still completely in the crypt and the deciduous teeth
have just been shed. The first and second molars are intact on both sides and the
two canines are almost intact. RP1 is preserved but LP1 is broken off at the roots
and the remaining premolars are represented only by the sockets. The root
impressions of the last deciduous tooth on the right side are still visible and the
distance from the front of P1 to the front of M1 on this side is about 5 mm longer
than on the left side, so it may be assumed that the teeth will close up a little.
There are no obvious gaps and it il? considered that the arrangement was very
much like that in Sus scrofa, with P1 slightly separated from the canine, and the
premolars just in mutual contact. Each of the P^s has a length of 9.8 mm and
estimated measurements for the other premolars are approximately 14 mm for
P2, 15 mm for P3 and 14 mm for P4, an aggregate of 53 mm. The actual distance
from the front of LP1 to the front of LM1 is 56.0 mm. This is greater than P1-?4
in Sus scrofa at the same stage of dental maturity, which ranged from 45 to 48
mm in three specimens available for comparison, but the difference is not great.
In Potamochoerus porcus, P1 is normally absent, but it is sometimes present in
juveniles and is shed very early in life, at approximately the same time as the
last milk tooth. In the fossil it is clear that P1 is firmly rooted and it is expected
that it would remain as a functioning tooth in the adult. In size and morphology
it is almost identical with that in the living wild boar. In dimensions, the P1
corresponds closely to P2 in the bush pig, but the tooth is narrower and more
bladelike in appearance; it is longer than the P1 occasionally found in the bush
pig in young animals. The M1 and M2 have the same characters as the
corresponding teeth in the holotype cranium, but M2 lacks the small external
basal pillar between the laterals. The canines have only just emerged but the tip
is already worn to a small facet about three quarters of a centimetre in length

1978

PLIOCENE- PLEISTOCENE SUIDAE

(Plate 12 A). The anterior and posterior crests typical of a young canine are still
visible and the morphology does not differ significantly from that of wild boar
or bush pig at this stage of eruption.

The two maxillary fragments, AL 222-4, AL 218-1, and the isolated RM3,
AL 200-13, confirm the general characters shown by the molars in the other
specimens described above. The isolated RP4, AL 165-12, and the RP4 in the
maxilla AL 222-4, do not exhibit the large craterlike fovea seen in the holotype
and the pattern is more like that of the bush pig and early Kolpochoerus
limnetes.

Mandible

The lower jaw is not very well known from the Sidi Hakoma member. The
best specimen is AL 109- 1 , which is a left mandibular ramus broken through the
canine socket, but retaining the left canine (Plate 12 B). The canine is verrucose
and thus closer to the characteristic tusk of Potamochoerus porcus than to the
scrofic one of the wild boar. The posterior face at the alveolar margin measures
15.0 mm, the internal face 19.7 mm, and the external face 17.5 mm. The three
anterior premolars are all broken off at the roots, but their sizes can be
estimated. Pj lies about 7.5 mm behind the canine, as in Sus, and there is a
gap of 25 mm between it and P2, which must have been approximately 12-13
mm long, while P3 had a length close to 15 or 15.5 mm. Pj is thus about the size
of the corresponding tooth in Sus scrofa, as also are P2 and P3. P! is about the
same size as P2 in the bush pig and both P2 and P3 are larger than in
Potamochoerus porcus. In the fossil, P4 is intact and is a little larger than in the
bush pig, but close to the wild boar in size. The P4 in Sus scrofa has a strong
main cone consisting of two elements that are laterally displaced, so that there is
a marked “kink” in the crest; in Potamochoerus porcus the protocone is stout
and undivided. In the fossil there is a distinct deviation in the crest. The crest is
also bladelike, as in Sus scrofa, and less expanded than in the bush pig; the
internal buttress of later Kolpochoerus is not developed. The anterior and
posterior cingula, although quite well developed, are not as high as in the wild
boar. The tooth is thus more or less intermediate in character. In this jaw, the
crown of Mj is broken away but M2 and M3 are preserved. As was the case with
the upper molars, they resemble the corresponding teeth of Sus scrofa and
Potamochoerus porcus in structure but the pillars are less complexly folded
than in the wild boar and are more columnar and better separated even than in
the average bush pig. However, the third molar clearly has only two pairs of
laterals and a stout median terminal pillar, flanked by accessory basal pillars,
and is thus simpler than the specimen from Kaiso that Cooke and Coryndon
(1970) named “ Sus ” waylandi, in which there are three pairs of laterals, as in

H.B.S. COOKE

No. 29

Kolpochoerus afarensis

E
E

I- 5
Q
<

1 R 2

l5-i

15-

IO-

.*0

IO-

15 IO

1 M,

20-

“ ©

□

15-

IO 15 20 25 15 20 25 30

e®

"Sus waylandi" Type

□

IO I-

20 30 40 50

LENGTH (mm)

Fig. 9. Kolpochoerus afarensis. Length-breadth measurements on lower cheek teeth. Key as in
Figure 8. The type of “Sus waylandi’’ (now placed in K. limnetes) is also shown.

typical Kolpochoerus limnetes (to which waylandi is now referred).

The other mandibular fragments do not add significantly to the informa-
tion given above, but serve to confirm that AL 109-1 is a fairly representative
specimen. Only AL 127-38 has the P4 preserved, and this is slightly smaller

1978

PLIOCENE- PLEISTOCENE SUIDAE

than in AL 109- 1 , but morphologically similar. There are ten third molars from
the Sidi Hakoma member, and these range in length from 33.5 mm to 39. 8 mm,
but all have essentially similar morphology. Length/breadth ratios are shown in
Fig. 9.

Referred material

The Denen Dora member has yielded a substantial amount of material that
can be referred without hesitation to the same species as that occurring in the
Sidi Hakoma member. However, there are also found in the Denen Dora
member four specimens that seem to be distinctly larger than the rest of the
sample and also show slight differences in morphology. Although it is very
likely that they are only variants, they will be discussed separately below. It is
possible that they represent the beginning of an actual dichotomy. Under the
circumstances, it has been thought best to restrict the hypodigm to the Sidi
Hakoma sample. Many of the Denen Dora specimens contain parts that were
not represented in the Sidi Hakoma material and thus warrant particular descrip-
tion. Length/breadth plots for all the referred cheek teeth are shown in Figs. 8
and 9.

AL 1 54-34 (DD- 1/2) is a partial cranium, lacking the snout, the back of the
braincase and basioccipital area, but apparently undistorted and having both
zygomatic arches well preserved. It is illustrated in Plate 1 1 . Most unfortu-
nately, the M2 and M3 on both sides were in very advanced wear and have been
damaged as well, so measurements on these are only approximate. RM2 has a
basal length of close to 23.5 mm and a breadth of 19.0 mm; LM3 has a basal
length of 38.5 mm and a breadth of 22.5, in good agreement with other
material in better preservation. Both orbits are preserved and the frontal is
broken across the braincase just behind them, so it is possible tc^measure the
frontal breadth as close to 1 10 mm, which is virtually the same as the estimate
for the holotype cranium. The interocular breadth is 85 mm, which is also close
to that for the holotype, and the morphology of this part of the snout is very
similar. There can be no reasonable doubt about the identity of the specimen as
representing the same species as the holotype. The top of the orbit lies approxi-
mately 150 mm above the occlusal plane and the front of the orbit lies very
slightly in front of the back of M3, as is also the case in Potamochoerus porcus,
whereas in Sus scrofa (wild) and Kolpochoerus limnetes it is usually slightly
behind the back of M3. The lateral profile is decidedly steeper than in wild Sus
scrofa, or even Sus verrucosus, and is perhaps a little steeper than in
Potomochoerus porcus, which it otherwise resembles closely. As in the bush
pig, there are elevated ridges flanking the supraorbital canals, clearly seen as
bumps in the profile. The flat naso-frontal area overhangs the maxilla, forming

H.B.S. COOKE

No. 29

a rather sharp edge above the scooped out area for the levator rostri muscles, as
in the bush pig; the scooping is not quite as deep as in the wild boar. The origin
of the depressor rostri, however, is relatively weak and shallow, as in S us scrofa
and Kolpochoerus limnetes, and not as deeply excavated as in the bush pig. The
ridge between these two areas, which forms the maxillary root of the zygoma, is
only moderately developed and rounded, unlike the strong sharp ridge of Sus
scrofa. The root of the zygoma curves rapidly outwards until it is perpendicular
to the skull axis and the jugal is greatly expanded laterally into large bosses that
are rugose over much of the inflated area. Although there is a basic resemblance
to the zygoma in old male animals of Potamochoerus porcus, there is less of a
forward bulge and the lateral expansion is very typical of Kolpochoerus lim-
netes. There is also a resemblance to the zygomatic enlargement in the forest
hog, Hylochoerus meinertzhageni, but in that species the zygomatic arches
droop almost to the occlusal plane, whereas in the fossil and in the bush pig they
lie at a much higher level. In the forest hog, the maxillary root of the zygoma is
quite different; the broad parietal area is also quite unlike the braincase in the
holotype. The bizygomatic breadth is 223 mm, which is greater than is usual in
Potamochoerus porcus, for which 200 mm is rare and 175-180 mm more
normal. The structure of the zygoma, coupled with the other morphological
features already discussed, seem to place afarensis firmly in the genus Kol-
pochoerus, despite the existence of some more Sus- like characters and some
Potamochoerus resemblances.

The upper dentition is well represented in the referred material, which is
listed above, and includes two specimens in which P3 is preserved, AL 1 16-1
(Plate 10 B) and AL 1 1 8-7 (DD-2/3). Both these P3’s are more robust than in the
holotype and correspondingly a little less like those of Sus, but larger than the
somewhat reduced P3 of Potamochoerus porcus (Fig. 8). The P4’s are also
larger than in the holotype, but not quite as large as in specimen AL 165-12
from the upper Sidi Hakoma member. The protocone blocks the fossa, as in the
bush pig, but the paracone and metacone are more like those in Sus scrofa and
Kolpochoerus limnetes. The third molars are also generally larger than the Sidi
Hakoma sample, and there is a tendency for the molars to be more columnar and
to have less secondary plication than in wild boar or in bush pig. In particular,
the outer lateral pillars tend to become more rounded and smooth walled.

The best mandible is AL 134-7 (DD-2/3), comprising the symphysis and
complete left ramus and also a fragment of the right ramus with RM2_3. The
canines are broken off at the alveolar margin and the incisors are badly broken.
The left side of the jaw is damaged in front of P3 and it is thus impossible to
determine the size or form of P2. There is no sign of Pl9 or of a scar where it
might have been (Plate 13 A/C). The general shape of the jaw and the sym-

1978

PLIOCENE- PLEISTOCENE SUIDAE

physis are like those of the wild boar and the bush pig, but closer to the latter
than the former. The canine flanges are more prominent than in the wild boar,
and in profile the symphysis is flat or concave rather than slightly convex, as is
the case with Sus scrofa. The back of the symphysis is about level with P3,
whereas in Sus scrofa it is level with P2, but the small shelf developed in the
bush pig for the insertions of the genioglossus and geniohyoideus muscles is not
apparent. The mandibular ramus is thicker than in Sus scrofa and has a swollen
appearance lateral to M, and M2, as in the forest hog and in Kolpochoerus
limnetes. The incisors are broken but were generally like those of Sus scrofa or
Potamochoerus porcus. The P3 and P4 are narrow cutting teeth, with P3 very
Sws-like. In P4 (which is lying obliquely in the tooth row) the displacement of
the central cusp is not as marked as in Sus scrofa , but it is also not as stout a
tooth as in the bush pig. The first and second molars have less complex folding
than in Sus scrofa and are more like those of the bush pig. However, the third
molars are further simplified and the lateral columns are more distinct than in
the bush pig and the outer walls of the main pillars are smoothly rounded,
although the basic structure is still close to Potamochoerus porcus in respect of
the number of columns present and in the limited development of the talonid.

Other material from the members DD and KH serves to confirm that AL
134-7 is a reasonably representative specimen. However, in a few of the third
molars the large terminal pillar is accompanied by a second, smaller one. AL
157-2 (DD-3/KH-1) also includes the symphysis with broken incisors and
canines, and this specimen also lacks any evidence for a lower P, in the
diastema; the cheek teeth are too damaged to be informative. AL 1 16-60
(DD-2/3) is a right mandibular ramus with RM,_3 well preserved and a
fragment of the left ramus with LM3. In this specimen the teeth are distinctly
simpler in enamel folding, and the columnar laterals have a smooth, rounded
external face, with thick enamel.

The Larger Specimens

Five specimens differ sufficiently in size from the other material to warrant
special consideration. Three are from DD-1 and are: AL 287-1 A, a left man-
dibular ramus with LP3-M2 and part of LM3; AL 296-1 , a damaged mandible
with the symphysis and some intact teeth; AL 325-8, a left mandibular ramus
with LP4-M2. From DD-1/2 there is a fragment of mandible with LM3, AL
182-48, and from KH- 1/2 a fragment of maxilla with RM3, AL310-12. InFigs.
8 and 9 the plots for these specimens have been circled. The RM3 lies just over
two standard deviations from the mean for the remainder of the sample. The
mean for the four lower third molars is almost exactly two standard deviations

H.B.S. COOKE

No. 29

distant from the mean for the remaining sample, but lies closely along the same
regression line. Similar size differences apply to the other teeth, which is
simply consistent with the fact that they belonged to larger-than-average indi-
viduals.

From a morphological point of view, the third molars in these specimens
have thick enamel and the secondary fissures are merely thin cracks, while the
main lateral pillars tend to be columnar and have smooth, rounded external
faces. In all these respects, the teeth resemble those of Kolpochoerus limnetes,
especially the material from the lowest members in the Omo sequence. How-
ever, the crown still consists of only two pairs of laterals and a single large
median terminal pillar, whereas in K . limnetes there is normally a third lateral
pillar on the lingual side in the uppers and on both sides in the lowers; in some
early specimens, the third pair of laterals constitutes the terminal talonid, in
place of the single median terminal pillar. Clearly these larger specimens point
towards K. limnetes, while retaining the basic characteristics defined for the
species afarensis. This is what one might expect if the latter is ancestral to
limnetes, as seems very probable, and it is perhaps natural that the larger
specimens should be most limnetes- like.

Discussion

The discovery of this small suid is of particular interest in furnishing a
possible clue to the ancestry of the living bush pig, which is almost unknown as
a fossil until the later Pleistocene. It may also provide a very reasonable
ancestor for Kolpochoerus limnetes, whose primary origin has been uncertain.
On purely dental grounds, the small Hadar suid might have been placed almost
equally well in either Sus or Potamochoerus, although perhaps with a bias
towards the latter, but the structure of the zygomatic arches is very different
from that of Sus and so like that of Kolpochoerus that it is logical to assign it to
the latter genus. There is a basic, but not exact, resemblance to the zygoma in
the bush pig.

Consideration must be given to the possibility that Kolpochoerus afarensis
is actually the direct ancestor of the living P. porcus, or even that it should be
regarded merely as a fossil representative of that species. It has not been
regarded here as a synonym of P. porcus because of the following consid-
erations:

(a) the mutual relationships between the premolars and molars are differ-
ent and have the same proportions as in Sus scrofa: P1 is present in the adult,
which is rare in bush pig, and P1 and P2 are virtually the same as in the wild
boar, P1 being almost as big as P2 in the bush pig; M3 is larger than in the bush

1978

PLIOCENE- PLEISTOCENE SUIDAE

pig but the same size as in Sus scrofa. Similar relationships apply to the lower
dentition.

(b) the premolar morphology is somewhat different from that of
Potamochoerus porcus, not only in the rather more sectorial character of the
second and third premolars, but also in the possession of a Sus-like displaced
double crest in P4. The P4 is more Sus- like and, despite the variability of this
tooth encountered in living suids, the difference must be regarded as grounds
for not placing the fossil into Potamochoerus.

(c) the molar teeth in the fossil tend to have thicker enamel and less folding
than in the living bush pig. This would suggest some degree of specialization
away from a bush- pig-like tooth rather than a truly ancestral type . However, the
earliest specimens have molar teeth that are most like those ofP. porcus, and
thus early Kolpochoerus afarensis may well be very close to the branching
point. Modification of the premolars would doubtless take place during the
descent of P. porcus, leading to its present characteristics. It seems likely that
the premolar characteristics that help to separate Potamochoerus porcus from
Sus are not fully developed and it might be inferred that a still earlier ancestor
might well be even more Sus- like in the premolar dentition. There is thus no
particular reason to demand descent of Potamochoerus from Pro-
potamochoerus.

As far as Kolpochoerus limnetes is concerned, K. afarensis seems to be a
very suitable ancestor, with the degree of resemblance increasing in the higher
stratigraphic horizons. It could be argued that it should therefore be placed in
that species as a very early representative and that a distinction based on the
number of lateral columns in the third molars is too arbitrary to be valid. Of
course, in a more-or-less unbroken lineage of this type it is indeed difficult to
draw a hard and fast boundary and any definition must be somewhat arbitrary.
The literature teems with arguments about the problems involved in attempting
to subdivide a continuously evolving lineage, especially in the absence of those
frequent time gaps that often serve so conveniently to break up stratigraphic
sequences into neat and manageable parcels. As W. S. McKerrow (1956) has
pointed out, it is necessary “to realise that specific names in palaeontology are
frequently only applicable to fixed points in a plexus of evolving organisms.”
Extension of a single species name to a morphologically changing complex
over a long period of time avoids (or evades) the problem, but ceases to be
useful when the end members in the lineage are so different that it is almost
inconceivable that they would, or could, belong to the same biospecies if they
had lived at the same time. As T. Neville George (1956) put it: ‘ ‘If a spade were
always called a spade no doubt correlation would be greatly eased, though it
might not be very exact if the spade were anything from a steam shovel to a

H.B.S. COOKE

No. 29

teaspoon.” Palaeontological nomenclature should be stratigraphically useful
as well as biologically reasonable. Accordingly, the present writer prefers to
define Kolpochoerus afarensis as a species, recognizing that it will grade into
K. limnetes and that some specimens may have to be dealt with by indicating
their borderline character (easily done by use of the prefix aff with one of the
species names). K . afarensis is thus regarded as close to the point of divergence
of theP. porcus and P. limnetes lineages. Perhaps also belonging to this species
are two third molars from Laetolil, Tanzania, described and figured by
Dietrich (1942, figs. 150, 157) as Potamochoerus sp. cf major.

Discussion

The identifiable material assigned to each of the three species described
here has been analysed in Table 4 to show the numbers of specimens from each
stratigraphic submember, divided into upper and lower dentitions. Much frag-
mentary material has been omitted. On the right are shown the total numbers of
specimens for each stratigraphic subdivision. It should be borne in mind that the
units do not represent equal intervals of time and, accordingly, that the numbers
do not represent original relative abundances. It is of interest to note that for
each of the species, skulls and upper dentitions make up only one quarter to one
third of the material, while the lower jaws and teeth are much better repre-
sented.

From a stratigraphic viewpoint, Nyanzachoerus pattersoni is almost con-
fined to the Lower Sidi Hakoma member, with only three specimens from the
Upper Sidi Hakoma. One isolated upper third molar from locality AL 134 is
supposedly from DD-2/3 and if this is correct, then N. pattersoni still existed at
that level, but it might be a derived or misplaced specimen. In the lower Sidi
Hakoma member, Notochoerus euilus is almost as abundant as Nyan-
zachoerus, but Kolpochoerus afarensis is slightly less abundant, making up
one quarter of the suid sample from this level. In the Denen Dora member,
where Nyanzachoerus is effectively absent, Notochoerus euilus makes up 75%
of the suid material and Kolpochoerus afarensis maintains its proportion of
25% of the suid sample. From the Kada Hadar there are too few suids for
numbers to be significant, but both Notochoerus euilus and Kolpochoerus
afarensis are represented.

As far as comparison with other sites is concerned, the Nyanzachoerus
pattersoni material matches very closely with the sample from Kanapoi, where
the age is estimated as close to 4.0 million years. The Notochoerus euilus
material is generally similar to the large sample from the Usno Formation in the
Omo area, which has an estimated age of close to 3.0 million years. However,

1978

PLIOCENE- PLEISTOCENE SUIDAE

TABLE 4

Stratigraphic distribution of Hadar suid specimens
(individuals - upper and lower elements)

Stratigraphic Nyanzachoerus Notochoerus Kolpochoerus
Unit pattersoni euilus afarensis

Upper Lower

Upper

Lower

Total

KH-2/3

—

KH-2

—

KH-1/2

—

KH-1

—

DD-3/KH-1

—

DD-3

—

DD-2/3

—

DD-2

— -

DD-1/2

—

DD-1

—

SH4/DD-1

—

SH-4

SH-3/4

—

SH-3

—

SH-2/3

—

SH-2

—

SH-1/3

—

SH-1

—

Basal

—

Uncertain

—

Totals

136

217

298

As %
of species

23.8

76.2

25.7

74.3

32.9

67.1

27.2

72.8

Summary

Kada Hadar

Denen Dora
Upper Sidi

116

164

Hakoma
Lower Sidi

Hakoma

Miscellaneous

As % of
Grand

3.4

10.7

15.8

45.6

8.1

16.4

27.2

72.8

Total

14.1%

61.4%

24.5%

100.0%

H.B.S. COOKE

No. 29

the Hadar material from the Sidi Hakoma member seems to be a little more
“primitive” and closer to its presumed ancestor, Nyanzachoerus jaegeri. An
age for the Sidi Hakoma member of 3. 0-3. 5 million years would give a better
fit than the 2. 9-3.0 m. y. age presently obtained from the basalt, while an age
close to 3 .0 million years for the Denen Dora member would be best in line with
the stage of evolution of the Notochoerus euilus material. Thus, the suid
evidence confirms in general the age provisionally assigned to the Hadar
Formation but suggests that the lower part may be a little older than the present
radiometric determinations indicate.

1978

PLIOCENE- PLEISTOCENE SUIDAE

Plate 1 . Nyanzachoerus pattersoni, male AL 137-4, dorsal (A), right lateral (B) and palatal views.
One-fifth natural size.

H.B.S. COOKE

No. 29

>-} c

<< <D
^ <U
^ t/i

« c

C3 O

£ £
o .22

co "O
£ J3
<L> .«
> *

& I ^
c/5 1 rr

ss J

* < |

If ^

13 <* |

J2 u

« &
<Sh ^
<u j*
— <u
- <L>

< fi

1978

PLIOCENE- PLEISTOCENE SUIDAE

AL 218-2: A, mandible of old individual, one-fifth natural

Plate 3. Nyanzachoerus pattersoni,
size; and B, right lower cheek teeth, natural size. Notochoerus e
to show the form of the zygomatic protuberances; D, occiput.

H.B.S. COOKE

No. 29

PLIOCENE- PLEISTOCENE SUIDAE

Plate 5. Notochoerus euilus, AL 172-1, palate and cheek teeth on both sides. Natural size

H.B.S. COOKE

No. 29

- g
« a,
o

oo o
O T3

, OO

« Ph
M "4

» 2
4= 8

<o I <

ISc

1978

PLIOCENE- PLEISTOCENE SUIDAE

Plate 7. Notochoerus euilus, AL 1 16-28. A, mandible with both incisors, one-fifth natural size; B,
cheek teeth of both sides, natural size.

H.B.S. COOKE

No. 29

Plate 8. Notochoerus euilus. A, anterior border of symphysis with LI2.3 intact and roots of other
incisors, one-half natural size. B, AL 122-5, LM3 in early wear, occlusal and outer lateral views,
natural size; C, AL 173-4, LM3 in moderate wear, inner lateral and occlusal views, natural size.

1978

PLIOCENE-PLEISTOCENE SUIDAE

Plate 9. Kolpochoerus afarensis, Holotype, AL 147-10, left lateral (A) and palatal (B) views,
one-third natural size.

H.B.S. COOKE

No. 29

,Vnnv

Plate 10. Kolpochoerus afarensis. A, palate and cheek teeth of Holotype, AL 147-10. B, palate
with P3-M2 on both sides, AL 116-1. Natural size.

1978

PLIOCENE- PLEISTOCENE SUIDAE

Plate 1 1 . Kolpochoerus afarensis. AL 154-34, partial craiiium of old individual, dorsal (A), right
lateral (B), and palatal (C) views. One-third natural size.

H.B.S. COOKE

No. 29

Plate 12. Kolpochoerus afarensis. A, palate of juvenile, AL 224-3, with canines erupting, RP1 and
M1'2 on both sides; B, AL 109-1, left mandibular ramus with broken canine, socket of LP, near
canine, roots of LP2.3 and LM,, LP4, LM2„3 intact. All natural size.

• •

1978

PLIOCENE- PLEISTOCENE SUIDAE

Plate 13. Kolpochoerus afarensis. A, AL 134-7, incomplete mandible with good symphysis,
one-fifth natural size; B, left cheek teeth of A, showing the oblique LP4, and C, right M2.3, natural
size; D, AL 186-20, left mandibular ramus with good LP2.4 and well-worn LM,_3, natural size.

H.B.S. COOKE

No. 29

References

Aronson, J. L., T. J. Schmitt, R. C. Walter, M. Taieb, J. J. Tiercelin, D. C. Johanson,
C. W. Naeser, and A. E. M. Nairn, 1977, New geochronologic and paleo magne-
tic data for the hominid-bearing Hadar formation of Ethiopia, Nature, London,
267: 323-327.

Broom, R., 1925, On evidence of a giant pig from the late Tertiaries of South Africa,
Records Albany Mus., 3: 307-308.

Cooke, H. B. S., 1976, Suidae from Plio-Pleistocene strata of the Rudolf basin, in
Earliest man and environments in the Lake Rudolph Basin, edited by Y. Cop-
pens, F. C. Howell, G. L. Isaac, and R. E. F. Leakey, Chicago: University of
Chicago Press, 251-263.

Cooke, H. B. S. and S. C. Coryndon, 1970, Pleistocene mammals from the Kaiso
Formation and other related deposits in Uganda, Fossil Vertebrates of Africa, 2:
109-224.

Cooke, H. B. S. and R. F. Ewer, 1972, Fossil Suidae from Kanapoi and Lothagam,
northwestern Kenya. Bull. Mus. Comp. Zool., Harvard, 143 (3): 149-295.

Coppens, Y. 1971, Une nouvelle espece de suide du Villafranchien de Tunisie, Nyan-
zachoerus jaegeri nov. sp., C. R. Acad. Sci. Paris, 272D: 3264-3267.

Dietrich, W. O., 1942, Altesquart'are Saugetiere aus der siidlichen Serengeti,
Deutsch-Ost-Afrika, Palaontographica, 94A: 43-133.

Ewer, R. F., 1958, The fossil Suidae of Makapansgat. Proc. Zool. Soc. London, 130:
329-372.

George, T. N., 1956, Biospecies, chronospecies and morphospecies, in The species
concept in palaeontology, by P. C. Sylvester-Bradley (ed.), London:. The Sys-
tematics Society, 1956: 123-137.

Hoepen, E. C. N. and H. E. van, 1932, Vrystaatse wilde varke, Paleont. Navors. nas.
Mus. Bloemfontein, 2 (4): 39-62.

Hopwood, A. T., 1926, Fossil mammalia, in The geology and palaeontology of the
Kaiso Bone Beds, by E. J. Wayland, Occ. Pap. Geol. Surv. Uganda, 2: 13-36.

Hiinermann, K. A., 1968, Die Suidae (Mammalia, Artiodactyla) aus den Dinotherien-
sanden (Unterplioz'an = Pont.) Rheinhessens (Siidwestdeutschland), Schweiz,
palaeont. Abh. 86: 1-96.

Johanson, D. C. and M. Taieb, 1976, Plio-Pleistocene hominid discoveries in Hadar,
Ethiopia, Nature, London, 260: 293-297.

Johanson, D. C., M. Taieb, B. T. Gray, and Y. Coppens, 1978, Geological framework
of the Pliocene Hadar Formation (Afar, Ethiopia), with notes on paleontology
including hominids, in Geological background to fossil man, by W. W. Bishop
(ed.), Edinburgh: Scottish Academic Press.

1978

PLIOCENE- PLEISTOCENE SUIDAE

Leakey, L. S. B., 1958, Some East African Pleistocene Suidae, Fossil Mammals of
Africa, No. 14, British Mus. (Nat. Hist.): 1-133.

McKerrow, W. S., 1956, Fossil species and the rules of nomenclature (Discussion), in
The species concept in palaeontology, by P. C. Sylvester-Bradley (ed), London:
The Systematics Society, 1956, p. 122.

Taieb, M. , 1974, Evolution Quatemaire du bassin de T Awash (Rift ethiopien et Afar),
Universite de Paris VI, doctoral thesis, Vol. 1 — Text, 390 pp., Vol. 2 — Plates
and maps.

Taieb, M., Y. Coppens, and D. C. Johanson, 1972, Depots sedimentaire et faunes du
plio-Pleistocene de la basse vallee de l’Awash (Afar central, Ethiopie), C. R.
Acad. Sci. Paris, 275D: 819-822.

Taieb, M., D. C. Johanson, Y. Coppens, and J. L. Aronson, 1976, Geological and
palaeontological background of Hadar hominid site. Afar, Ethiopia, Nature,
London, 260: 293-297.

CONTENTS

No. 28. A New Species of the Genus Australopithecus (Primates:
Hominidae) From the Pliocene of Eastern Africa — Donald C.
Johanson , Tim D. White , and Yves Coppens
No. 29. Pliocene-Pleistocene Suidae From Hadar, Ethiopia —H.B.S.
Cooke

DlinillSN! NVINOSHJLISAIS S3I8V88I1 LIBRARIES SMITHSONIAN INSTITU

' > Z r~ z r~

AIMKSv — — y'jmAkmr.i*/ r- -y '

ro s 2: rn ^ m N^/osv^

CO CO f~ £/}

BR ARIES SMITHSONIAN INSTITUTION NOlirUIXSNI NVINOSH1IWS S3 1 HVB

Z > CO 21 C/3

- • - / ^ t

> ' 2 X^VQSV^^ >

HifUliSNI NVINOSHimS^ S 3 I 8V8 8 ll\l B RAR ! ES^SMITHSONIAN JNSTITU

v j§y ^

2 5 ' / y 5

«j 2 j 2

B RAR I ES SMITHSONIAN INSTITUTION NOIlfUllSN! NVINOSHJLIWS $3I8Vfc

33
>

W ± £ . _

)linillSNI NVINOSHilWS S3IUVUan LIBRARIES SMITHSONIAN INSTITU
2 CO 2: ... CO 2

< 2 , < .. 2 <r‘

S S _ _

BRARIES SMITHSONIAN“ INSTITUTION NOlinillSNI^NVINOSHilWS^SS I BV

^ ^ ^ ~ 00

lii ><avaSOaJ>\ LU ~ x5vS - Ld

HlfUllSNI NVINOSHillNS S3I8V88H LIBRARIES SMITHSONIAN INSTITU
^ * Z r~ 2: r*

rn ^ rn ^ m ^v^osvi

CO — CO S CO

BRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHiMS S3 1 8 Vi

> g ^ z \ CO 2 CO

s < \v 2 Ji#'. 5

t « 1 ,

> '^VASV^- g > ’* ’ i

liniIISNI__NVJN0SHiIWS^S3 I 8V8 8 Il\l B RAR I ES^SMITHSONIAN^INSTITU

*\ ^ CO

S IIE, JiJ s;>

q xQftTgC^ “ O x£* p^ — x^oxas^ O

z N0!iniIXSNI^NVIN0SHillMSzS3 I UVLI a IT^LI B rar i es^smithsonian^insti

§ S^&zs. 5 V _ 2 5 /^S^x I Jb». 5 ^

, > I

o/ 33 V5tJ

m n^vqsv^x ^ m Ss<^

"LI B R AR ! ESC/,SMITHSONIAN~INSTITUTIONC,>NOIinillSNl""NVINOSHimS S3 I d

2 '* ^ ^ ^ 2 \ in Z V

I /*Z fis'

:> s ^ >' 2

C/3 2 </} •*** 2 CO 2

NOliniliSNI NVIN0SH1IWS S3ldVd8n LIBRARIES SMITHSONIAN INSTI

^ ^ co ... x

O Xi^osv^X Z ' O X

2 _J 2

LI B RAR I ES_ SMITHSON IAN _JN ST I TUT I ON NOliniliSNI NVIN0SH1IWS S3 I d

2 -5 /£g£&\ 2 5 ° ^

tr /.•>/ A 30 t !23

I— — XtMiKoV h- UWiiW/ — >'■ !—

“ m rn X 2

CO — C o £ co \ ±

NOlinilJLSNI NVIN0SH1IWS S3ldVd8ll LIBRARIES SMITHSONIAN INSTI

CO 2 ^ ^ CO Z »y • <£ 2

2 §* mav> ^ ><u5vXr\

x :/^2

1 i 5 <*<■' 5 ^ > '<ZjnSZ®' 5

2 to z CO * Z to

LIBRARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHilWS S3 I *i

~ C/3 X ■:•■■■ C/3 ZI .... C/3

Q X^Y P OX _ O

2 NOliniliSNI ^NVINOSHilWS S3 IdVdOll

LIBRARIES SMITHSONIAN INSTI

_ m ^

'Ll B RAR I ES SMITHSONIAN”’lNSTITUTIONWNOIinillSNl“"NVINOSHllWS S3 I h

2 t co 2

< \v 2 <

,/3 'sVS’ co

W 'W;/ X VJ,

> 2 _

CO 2 CO 2

S3ldVdail LIBRARIES SMITHSONIAN INSTI

CO ^ “ CO

4x1 m LU

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Kirtlandia"

See other formats